|
MODBUS RTU communication protocol
implementation on TM4C123 |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The The
purpose of this project is to develop embedded software that implements
MODBUS communication with a TMC123 processor. With the protocol
implementation, it should be possible not only to read data from the
processor, but also to modify desired system parameters. |
|||||
|
|
||||||
|
IEC60870-5-103 communication protocol
implementation on TM4C123 |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The purpose of this project is to develop
embedded software that implements IEC60870-5-103 communication with a TMC123 processor. With
the protocol implementation, it should be possible not only to read data from
the processor, but also to modify desired system parameters. |
|||||
|
|
||||||
|
IRIG-B protocol implementation |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The purpose of this project is to design a
circuit that allows IRIG-B communication. Software should also be developped
to test that communication is possible with another IRIG-B device. |
|||||
|
|
||||||
|
PRU Software Development for
Continuous Data Acquisition using ARM335x processors |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The purpose of this project is to develop
embedded PRU libraries that can be used for continuous sensor measurements
using an ARM335x development board. The PRU firmware should allow capture of multiple ADC channels at desired sampling rates upto
10KHz. The CPU use should be minimized so that ample time is left to actually
deal with the data. It is also desired that dropped readings be reported. |
|||||
|
|
||||||
|
Inrush Current Detection and Filtering
using Harmonic Techniques |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The purpose of this project is to detect
and filter inrush currents that arise due to transformer energization.
Algorithms to be developed should be implemented in Matlab SIMULINK to test
the efficacy of the proposed techniques. |
|||||
|
|
||||||
|
Benchmarking GNNs for Brain
Connectomes |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The purpose of this project is to
investigate how to effectively design graph neural networks (GNNs) for brain
connectomes to identify different stages of dementia. Brain connectomes are
simplified descriptions of brain regions and their inter-connections
organizing the human brain, hence it is a graph. Recently GNNs have gained
increasing popularity for modeling graph data. Despite the promising results
in many fields, how to design effective GNNs for brain connectomes to
classify dementia stages is an open question. The project will investigate
the use of GNNs for identifying different stages of dementia, such as SCI,
MCI and AD, from brain connectomes. Structural and functional brain
connectomes will be used in this project. Approach and Methodology: The project is offered to one student. The
first phase of the project is a learning phase where the student is expected
to learn the theoretical basis of GNNs. A strong mathematical and coding
background are required to reach the goals. This first phase also includes
learning the implementation of the different GNN architectures with Python
libraries. The second phase will be an experimental stage where the student
will compare different GNN architectures, different feature selections etc.
to effectively design GNNs for structural and functional connectomes to
classify dementia. Expected Results and Accomplishments: At the end of the project, it is expected
to get an insight into the different GNN models, how to build GNNs for brain
connectomes, and how to preprocess brain connectomes to make them suitable
for GNNs. |
|||||
|
|
||||||
|
Blind Source Separation for
Neuroscience |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2023-2024 |
|||||
|
Description |
Description of the project: The purpose of this project is to
investigate the potential of Blind Source Separation (BSS) in identifying
cognitive subnetworks in human brain. Brain is thought to operate in terms of
subnetworks, ie. multiple regions co-operate in fulfilling certain tasks.
These subnetworks are believed to be sparse and overlapping, meaning that
they are composed of few number of brain regions and that each brain region
can be simultaneously active in more than one operation. The project will
investigate the use of BSS in identifying such subnetworks via decomposing
nodal (cortical) signals into the underlying source signals which are
associated with distinct subnetworks. The brain signals may well be
fMRI/BOLD, EEG, MEG, etc. Approach and Methodology: The project is offered to multiple
students. The first phase of the project is a learning phase where the
student(s) is(are) expected to learn the theoretical basis of BSS and major
matrix factorization methods relevant to BSS, namely PCA, ICA, NMF and PMF. A
strong mathematical background and interest in theory is required to reach
the goals. This first phase also includes implementation and application of
the above methods to signals of the student’(s) choice to get hands on
experience. The second phase will be differentiated for
each participating student. Based on the number of student(s) taking the
project one or more of the following problems (or any similar problem that
may be defined in due course) will be addressed: • BSS
on fMRI/BOLD and/or EEG signals with sparsity constraint on mixing matrix and
comparison to known functional brain subnetworks. • Application
of information maximization methods, adopted from PMF, to brain network
(connectome) factorization • Optimum
polytope (PMF) for fMRI/BOLD and/or EEG source signals • Effect
of signal sample size and number of source signals on BSS for fMRI/BOLD and/or EEG signalsExpected
Results and Accomplishments: Expected Results and Accomplishments: At the end of the project, it is expected
to get an insight into the functional organization of brain from a signal
decomposition perspective, as opposed to pure graph theoretical approaches
where the signals are used to build graphs/networks and not utilized further.
Such source signals and associated subnetworks can potentially serve as
diagnostic biomarkers. |
|||||
|
|
||||||
|
Generating Fake Image Data for Unseen
Cases Using Image Based Defect Detection Model Training |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2022-2023 |
|||||
|
Description |
Description of the project: A common bottleneck in training deep ANN
models for detection of certain structures (not objects) in images is the
lack of training data. Depending on the application domain, these structures
of interest can be defects in industrial products or disease related
structures (eg. tumors) in medical images. The purpose of this project is to
design, implement and test deep ANN models for generating fake training image
data for training defect detection networks. The project has 2 subtasks: 1. Design
and train a generative model to generate fake positive samples of a
previously seen type of defected / diseased images for which negative samples
are also available. 2. Design
and train a generative model to generate fake positive samples of a
previously unseen (yet a similar) type of defected / diseased images for
which only negative samples are available. Positive samples: Images with defects /
diseases Negative samples: Images without defects /
diseases Approach and Methodology: • Search
for available open datasets and determine an application domain • Literature
survey on generative deep ANN models, most prominently VAEs and GANs. • Design,
implement and test a deep ANN model to address the first problem above. • Design,
implement and test a deep ANN model to extend the first solution to address
the second problem above.Expected Results and Accomplishments: Expected Results and Accomplishments: It is expected to improve deep defect
detection model training with few training samples by augmenting the data set
with fake samples. |
|||||
|
|
||||||
|
Graph Neural Networks for Graph
Classification and Application to Brain Connectomes |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2021-2022 |
|||||
|
Description |
Description of the project: This project aims to investigate new class
of Graph Neural Networks (GNNs) : Graph Isomorphism Inspired Neural Networks.
These type of models leverage Weisfeiler-Lehman Graph Isomorphism test in
order to increase the expressive power of GNNs and offer a novel way to deal
with graph classification tasks. Theoretical background of such GNNs will be
explained and classification performance on different common datasets will be
compared in this work. Approach and Methodology: Among Graph Isomoprhism Inspired Neural
Networks, two popular model will be chosen and implemented : Graph
Isomorphism Network (GIN) and Folklore Graph Neural Network (F-GNN). These
two GNN model will be experimented on common datasets including chemical
compunds (MUTAG,Proteins,etc.) and social networks (IMDB,Reddit,etc.). They
will also be used in the classification of structural brain networks, i.e. to
distinguish patients with different clinical labels. The student is expected to use Julia
programming language and relevant packages. Expected Results and Accomplishments: At the end of the project, it is expected
to demonstrate the pros and cons of the two architectures in different tasks
including brain network classification. |
|||||
|
|
||||||
|
Dual X-ray Image Generation From a
Single Image for Material Discrimination Purposes |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2021-2022 |
|||||
|
Description |
Description of the project: Material discrimination in cargo scanners
using dual energy x-ray is a well-known method. Despite its theoretical
simplicity, the practical application has several obstacles to tackle with.
The major ones are wide energy spectrum of x-ray sources, overlapping
objects, detector sensitivity and fan-beam geometry. These are dealt with
using approximations. Modern ANN approaches offer a black-box approach to
material discrimination which is limited by the training set and has shown
acceptable performance in a recent VAVlab – TUBITAK project. The purpose of
this project, on the other hand, is to achieve material discrimination using
mono-energy scanners, by means of ANNs. The project involves working in
collaboration with TUBITAK-BILGEM. Despite the common goal and data, this
project is intended to pursue an alternative method to the TUBITAK-BILGEM
team in a mutually supportive collaboration. Approach and Methodology: The approach to be pursued in this EE492
project has 2 stages: 1. Train a deep NN to generate the second
(high/low energy) x-ray image from a given (low/high energy) image using GAN
training. The generator thus obtained will be used to artificially complete
the “pair of dual energy images”. 2. The real (low/high energy) x-ray image
and its artificially generated (high/low energy) pair will be fed into a
material discriminator ANN and the whole pipeline will be trained end-to-end. X-ray simulators and real scanner images
will be available. The coding will be done in Python, though it is not a
requirement. Expected Results and Accomplishments: At
the end of the project, it is expected to demonstrate an acceptable material
separation performance using a single x-ray image and report on the
challenges. |
|||||
|
|
||||||
|
Brain Connectome Factorization using
Deep N |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2021-2022 |
|||||
|
Description |
Description of the project: The purpose of this project is to adapt and
apply deep matrix factorization methods to brain connectomes for the purpose
of learning basis connectomes and low dimensional connectome representations. Approach and Methodology: This project falls in the field of medical
image analysis will be run in parallel with of the BRAINet project series at
VAVlab. The human brain structural and functional connectome factorizations
have recently been proposed in a PhD thesis at VAVlab. This project involves
adapting and applying a recently proposed deep matrix factorization method
(Deep Non-Negative Matrix Factorization Architecture Based on Underlying
Basis Images Learning, Zhao et al., IEEE PAMI, 43(6), 2021) to brain
connectome factorization problem. The project will use data from the spectrum
dementia. Public data as well as VAVlab’s own data will be used. The
discriminative power (across the continuum of dementia) of the learned low
dimensional brain connectome representations, as well as the medical
relevance of the learned basis networks is expected to be assessed as the
project output. Expected Results and Accomplishments: At the end of the project, 1. Deep
NMF will have been adapted and applied to brain connectomes, 2. A
learned brain connectome space will have been proposed, 3. The
novel set of basis brain networks learned and will have been compared to the
recent set of networks that were proposed in a recent PhD thesis, 4. Discriminative
power and clinical relevance of the learned connectome space will have been
discussed Should the project be completed
successfully, the project results can potentially be published. |
|||||
|
|
||||||
|
Topic Models and Their Application on
Different Data Types |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2017-2018/1 |
|||||
|
Description |
Description of the project: The aim of this project is to implement
topic models on different data types for clustering. Latent Dirichlet Allocation(LDA) and
Non-negative Matrix Factorization (NMF) methods will be the major focus of
the project. These algorithms are to be implemented for image processing and
text mining to cluster the data. Further purpose of the project is to
investigate whether topic models can be used on brain images to identify
functional regions in brain. Approach and Methodology: This project falls into the field of topic
modeling, which is a probabilistic modeling, on diferent data types (image
and text data for this project). Topic modeling provides methods for
automatically organizing, understanding, searching, and summarizing large
electronic archives. In this project LDA and NMF literature will be reviewed.
Mainly, the hidden themes that pervade the collection of images and text data
will be discovered and the data will be annotated according to those themes
based on LDA and NMF algorithms using Python. As to further implementation of
those algorithm, brain images will be studied and related literature will be
reviewed. In this context, the models are to extract a set of neural
“topics”, where each topic corresponds to a functional brain region. For each
topic, the model provides a description of its spatial extent (captured via
probability distributions over neural activation) and cognitive function
(captured via probability distributions over linguistic terms). The
efficiency of these models in providing a novel approach for jointly
identifying the spatial location and cognitive mapping of functional brain
regions will be investigated. There will be weekly project meetings. Expected Results and Accomplishments: When completed, this project is expected to
provide: 1. a
model and/or models that organize and cluster image and text data, 2. efficiency
investigation of those models to identify functional regions of brain. |
|||||
|
|
||||||
|
3D Cortex Parcellation |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2017-2018/1 |
|||||
|
Description |
Project
Class(es): EE491/2 Number
of students: 1 Description of the project: The purpose of this project is to improve the 3D human brain
cortex parcellation by means of joint structural and functional segmentation
and registration. The goal of cortex parcellation is to achieve structurally
compact and plausible, functionally homogeneous 3D segmentation of the human
brain cortex using multi-modal MRI data. The project involves elastically and
iteratively refining an existing cortex parcellation by maximizing the
intra-parcel correlation of fMRI BOLD signals. Approach
and Methodology: This
project falls in the field of medical image analysis will be run in parallel
with of the BRAINet project series at VAVlab. The BRAINet platform and
preprocessing routines provide 3D co-registered T1w MRI, T2w MRI, fMRI and
parcellation map (Destrieux atlas). The project involves applying elastic
free-form deformations on the parcellation map to maximize intra-parcel BOLD
signal correlation (using fMRI data), constrained to be on the cortex (by
comparison with T1w and T2w images). A literature survey on 3D elastic
registration / deformation methods will be followed by choosing an
appropriate algorithm, implementing it on BRAINet platform (C++, Linux) and
running on sample data. The student is expected to join regular BRAINet
meeting at VAVlab and collaborate with graduate students. Expected Results and Accomplishments: At the end of the
project, it is expected to have 1. A
novel method for structurally compact and plausible, functionally homogeneous
3D segmentation of the human brain cortex, 2. A
detailed report on the effects of new parcellation on the human brain
connectome parameters. Should the project be completed successfully, the project results will be publishable in a high-profile conference, leading potentially to a journal paper. |
|||||
|
|
||||||
|
Comparison of
Deterministic and Probabilistic Tractography in Brain Connectome Analysis |
||||||
|
Principal Investigator (PI) |
: Burak Acar - acarbu@boun.edu.tr |
|||||
|
Semester |
: 2017-2018/1 |
|||||
|
Description |
Project Class(es): EE491/2 Number of students: 1 Description of the
project: The
purpose of this project is to compare the use of deterministic streamline
tractography (RK4), SMT and probabilistic tractography on brain connectomes
of healthy individuals. Streamline tractography is the basic deterministic
tractography approach that generates artificial fibers representing brain
connections. Probabilistic tractography, on the other hand generates maps of
connectivity between any 2 points in the brain without explicitly depicting
any connection. SMT (Split and Merge Tractography), on the other hand, is a hybrid method of deterministic and
probabilistic tractography on dMRI (Diffusion MRI) data, that follows the
MCMC approach. The project aims at comparing these algorithms’ effects on
brain connectome (network) in terms of the network parameters. Approach and Methodology: This project falls in the field of medical image analysis will
be run in parallel with of the BRAINet project series. As part of this
project, the SMT algorithm is to be improved by higher order diffusion
models, a state-of-art probabilistic tractography algorithm will be
implemented, RK4, SMT and probabilistic tractography algorithms are to be
compared wrt human connectome parameters. The algorithms will be implemented
in BRAINet platform (C++, Linux). There will be weekly project meetings in
addition to the monthly TUBITAK project meetings to which the student is
expected to attend. Expected Results and
Accomplishments: At
the end of the project, it is expected to have 1.
A reliably and
efficiently running tractography module under BRAINet platform, 2.
A detailed report
regarding the the pros/cons of three different tractography approaches on
brain connectome analysis. Should the project be completed successfully, the project results will be publishable in a high-profile conference, leading potentially to a journal paper. |
|||||
|
|
||||||
|
Predictive energy management in
textile processes |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2017-2018/1 |
|||||
|
Description |
: Traditional energy management systems monitor and report energy flow
and cost. On the other hand, predictive energy management, which takes action
long before limits are hit, can be employed in textile dyeing/finishing
processes, not only to minimize production disruptions, but also energy
costs. An interesting problem in this context is to avoid electrical and gas
peaks that substantially increase financial expense. Additionally, it might
be quite advantageous to take decisions based on varying electricity prices
during the day. The objective of this project is to develop a predictive energy
management software module that will balance energy loads. As an example, the system may be configured
with parameters for gas and electricity load per time period. In developing the predictive energy management software module, some
of the tasks that have to be handled are listed below: -
Determine
the optimal energy management schedule for nominal conditions -
Determine
conditions under which a critical steam load is encountered -
Determine
and execute the best set of actions with the least impact on production time
and cost, when a critical steam is detected (e.g., production processes can
be altered during production runtime to run alternatively, without
influencing quality and planning. The action chain might include dosing
systems, hot water boilers, machine tanks, the step of each active machine
program, production process situation, etc) Profile: -
You are expected to
develop software in MATLAB and possibly using big data tools. -
You are expected to
have/acquire knowledge on systems theory, signal processing and artificial
intelligent techniques. -
You may need to
conduct some field tests with ELIAR. It will be required to allow time for
these experiments. If
you are interested, please do not hesitate to contact Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
|
||||||
|
Predictive maintenance of textile
dyeing and finishing machines |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2017-2018/1 |
|||||
|
Description |
: Textile
dyeing and finishing machines require regular maintenance of various
components including sensors, actuators, PLCs and software. Unforeseen
downtimes on production machines may adversely affect production schedules
and increase operation cost. The
objective of this project is to develop a software module that will manage
the required data, predict and plan maintenance actions and manage
interaction with daily production planning. This predictive maintenance
planning is to be carried out by analyzing system data from PLCs and from any
other available sensors. In
developing the predictive maintenance software module, the following tasks
will have to be handled: -
Develop and implement cyber-physical model of the product machine -
Investigate mathematical models of reliability in the literature -
Implement maintenance rules supplied by the manufacturer or any
calibration rule that may be required -
Develop and implement a predictive maintenance plan by data analysis With the predictive maintenance software module
implemented as part of the Manufacturing Execution System, longer and more
reliable machine availability is expected, which will in turn translate into
cost minimization benefits. Profile: -
You are expected to
develop software in MATLAB and using big data tools. -
You are expected to
have/acquire knowledge on systems theory, signal processing and artificial
intelligent techniques. -
You may need to
conduct some field tests with ELIAR. It will be required to allow time for
these experiments. If
you are interested, please do not hesitate to contact Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
|
||||||
|
|
||||||
|
Dyeing quality estimation based on
machine and fabric type |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar – mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2017-2018/1 |
|||||
|
Description |
: The manufacture of textiles is not only a major global industry but
also one of the main components of the local economy. This project is on
textile coloration processes. More specifically, given a coloring machine,
recipe, and fabric type, the objective of this project is to develop a
software module that will decide whether fabric dyeing is as desired. This prediction module is to be realized by analyzing system input and
output measurements based on machine learning tools. In developing the software module, the following tasks will have to be
handled: -
Collect input/output data relevant to the
prediction module -
Implement alternative techniques and
compare their performances -
Investigate the applicability of big data
analysis tools -
Decide whether correct dyeing is achieved The output of this software module
can be utilized to optimize dyeing procedures by removing the need to take
samples which will in turn reduce production time. Additional improvement can
be achieved by skipping some of the subsequent stages for incorrectly dyed
fabric. Profile: -
You are expected to
develop software in MATLAB and using big data tools. -
You are expected to
have/acquire knowledge on systems theory, signal processing and artificial
intelligent techniques. -
You may need to conduct
some field tests with ELIAR. It will be required to allow time for these
experiments. If
you are interested, please do not hesitate to contact Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
Auto-identification of PMSM Electrical
Parameters |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2016-2017/1 |
|||||
|
Description |
: Field oriented control of a permanent magnet
synchronous motor (PMSM) is the algorithm very often exploited in today’s
advanced motor control drives. The electrical parameters such as stator
resistances, synchronous inductances and pole pairs of PMSMs are required to
adapt the PI controller gains in order to get the desired performance. The
objective of this project is to automatically identify the electrical
parameters of PMSMs, rather than measuring them using devices such as RLC
meters. Profile: - Scope for graduate degree students in Electrical and
Electronics Engineering. - You must be interested in embedded hardware and
software design. - You must have knowledge about digital signal processing. -
You may need to conduct some experiments in the laboratories which are
located on Arcelik Cayirova Washing Machine Plant. It will be required to
allow time for these experiments. If you are interested, please do not
hesitate to contact Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
|
||||||
|
Sensor Design for Conductivity and
Turbidity Detection |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2016-2017/1 |
|||||
|
Description |
: Conductivity and turbidity sensors are generally
exploited to assess the quality of water. The main purpose of this project is
to design an integrated sensor which detects detergent and soil level in the
water. At the end of this project, it is expected to get an integrated sensor
which is compact as much as possible. Profile: - Electronics board design and
testing, embedded software development. - You must have knowledge about
digital signal processing. - You may need to conduct some experiments in the
laboratories which are located on Arcelik Cayirova Washing Machine Plant. It
will be required to allow time for these experiments. If you are interested,
please do not hesitate to contact Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
|
||||||
|
Precise Detection of the Amount and
the Position of the Unbalanced Load in the Laundry Using Accelerometers |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2016-2017/1 |
|||||
|
Description |
: Determining the unbalanced loading conditions of a
drum of a washer or washer-dryer is beneficial to protect the system. The
width of a washer’s body is fixed in a commonly used appliance and the depth
of the body varies according to the declared capacity of the washing machine
and the drum volume. The market trends increasingly lead to implement more
capacity declaration in a standard body, which forces the producers to fit
bigger size drum and tub in the fixed standard body. Consequently, the
distance between the tub and body is reduced day by day. For this reason, the
mechanical vibrations of the mechanical system should be controlled in a more
precise way to prevent the touch of the tub to the body. According to the
determined unbalanced load, the electronic control card may decide the
maximum allowable spinning speed of the drum to prevent any damage to the
system. Moreover, before the machine switching over to a very high spinning
speed, the imbalance of the drum must be estimated very well to keep the machine
secure. In the spinning cycle, the imbalance of the drum can result in too
much vibration which can move the machine or enforcing the mechanical
endurance of the rotating part. The purpose of this senior design project is
to determine the imbalance of the drum using accelerometer(s). At the end of
the project, it is expected to get a highly reliable systemdetermines the
amount and position of the unbalanced load in the drum. Profile: - You must
be interested in embedded systems software. - You must have knowledge about
digital signal processing. - You may need to conduct some experiments in the
laboratories which are located on Arcelik Cayirova Washing Machine Plant. It
will be required to allow time for these experiments. If you are interested,
please do not hesitate to contact Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
|
||||||
|
Developing Algorithms for Determining
the Loading Conditions of a Laundry Total Load and Unbalanced Load
Discrimination of a Laundry in a Washing Machine Drum |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2016-2017/1 |
|||||
|
Description |
: Determining the loading conditions of a drum of a
washer or washer-dryer is beneficial in order to optimize the washing or
drying procedure(s). According to the determined unbalanced load, the
electronic control card may decide the maximum allowable spinning speed of
the drum to prevent any damage to the system. If the total load amount is
known, water and detergent amount for washing can be modified or drying
duration can be optimized. The knowledge of total load and unbalanced load
makes it possible to obtain optimal washing results in terms of
cleaning/drying and energy. Moreover, before the machine switches over to a
very high spinning speed, the imbalance of the drum must be estimated very
well to keep the machine secure. In the spinning cycle, the imbalance of the
drum can result in too much vibration which can move the machine or enforcing
the mechanical endurance of the rotating part. Last but not least, in case of
high imbalance in the drum, the consumed current and power changes
dramatically over the spinning cycle and this may shorten the life duration
of the electronic parts. The purpose of this senior design project is to
develop and test the algorithms to determine the total load and the
unbalanced load concentration of the laundryin the drum.At the end of the
project, it is expected to get a highly reliable algorithm that determines
the total load and the unbalanced load. Profile: - You must be interested in
embedded systems software - You must have knowledge about control systems
theory and digital signal processing - You may need to simulate the loading
conditions in MATLAB. Thus, familiarity with MATLAB is required. - You may
need to conduct some experiments in the laboratories which are located on
Arcelik Cayirova Washing Machine Plant. It will be required to allow time for
these experiments. If you are interested, please do not hesitate to contact
Prof. Mehmet Akar. Phone: +90 212 359 6854 |
|||||
|
|
||||||
|
Hardware Development for Electronic
Stethoscope |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2015-2016/2 |
|||||
|
Description |
: Pulmonary sounds are traditionally diagnosed with
stethoscopes which have limited frequency response. Depending on the model
and manufacturer of the stethoscope, the response is limited to a frequency
range between 100 Hz to 200 Hz and usually falls off at a rate of 30 dB per
octave. On the other hand, a computerized pulmonary sound acquisition system
developed in our laboratory, LAL, the frequency repsonse extends up to
2000/4000 Hz with a flat response. The aim of this project is to build two
separate sound acquisition systems and compare the output both audibly and on
a display regarding the presence of adventitious sounds such as crackles and
wheezes. |
|||||
|
|
||||||
|
Software Development for Electronic
Stethoscope |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2015-2016/2 |
|||||
|
Description |
: Pulmonary sounds are traditionally diagnosed with
stethoscopes which have limited frequency response. Depending on the model
and manufacturer of the stethoscope, the response is limited to a frequency
range between 100 Hz to 200 Hz and usually falls off at a rate of 30 dB per
octave. On the other hand, a computerized pulmonary sound acquisition system
developed in our laboratory, LAL, the frequency repsonse extends up to
2000/4000 Hz with a flat response. The aim of this project is to develop a
MATLAB program that will simulate a traditional stethoscope and compare the
recorded sounds using the two systems in terms of the lung sound content and
adventitious lung sound (such as crackles and wheezes) content. |
|||||
|
|
||||||
|
Li-Po Battery Monitor and Control
Circuit for Minik Robots |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2015-2016/1 |
|||||
|
Description |
: Li-Po Battery Monitor and Control Circuit for Minik
Robots |
|||||
|
|
||||||
|
Dimmable LED Driver Design and
Implementation |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2015-2016/1 |
|||||
|
Description |
: Dimmable LED Driver Design and Implementation |
|||||
|
|
||||||
|
Two Dimensional ElectroChemical
Micropatterning using Electrode Array |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2014-2015/2 |
|||||
|
Description |
: Electrochemical deposition inside a microfluidic
channel will be simulated using a finite element analysis software (COMSOL)
for specific diffusion and reaction boundary conditions. Based on this
knowledge, a PCB with multiple electrodes will be used to pattern desired
shapes by electrochemical deposition. |
|||||
|
|
||||||
|
Energy Harvesting Using Thermoelectric
Cooler or Heater |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2014-2015/2 |
|||||
|
Description |
: Energy harvesting capacity of a commercial
thermoelectric cooler or heater will be tested and characterized. The
generated energy will be regulated using a simple analog circuit to supply
power to a commercially available RF transmitter. |
|||||
|
|
||||||
|
Location Tracker |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2014-2015/2 |
|||||
|
Description |
: The purpose of this project is to design and
implement hardware that can be used to track the position and possibly
orientation of a robot. The tracking will be based on sending messages to a
central collector device. One student will be responsible for the hardware
(design, building and programming) while the second student will be
responsible for processing the transmitted data appropriately and letting
authorized people to track the robot. The first part requires hardware skills
while the second part will require programming skills. |
|||||
|
|
||||||
|
Comparison of Object Recognition
Approaches |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2014-2015/2 |
|||||
|
Description |
: The purpose of this project is to apply several
state-of-the-art object recognition approaches in classifying and detecting
objects in a given scene or set of scenes and compare their performance. The
classification problems seeks answer to the question "Does the given
image contain any instances of a particular objects?". The detection
problem seeks answer to the question "Where are the instances (if any)
of an object in this image?". The approaches are the winner approaches
in the Pascal competition that was held upto 2012. The students need to have
intermediate to advanced level of programming skills (C++) as well as some
experience with image processing. |
|||||
|
|
||||||
|
Autonomous Referee for Small Size
Football League |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2014-2015/1 |
|||||
|
Description |
: The purpose of this project is to develop a
vision-based program so that the need for presence of an active human referee
decreases or vanishes during matches. The project is being supported and
proposed by Technical and Executive Committees of SSL league of Robocup
Federation. The designed software should use SSL shared vision to detect all
objects, do tracking and decide accordingly. Through the project the students
will learn about many topics such as vision-based programming, software
development, OpenCV, OpenGL, QT. Note that this project is a long term (2
semester) project and it is expected that students communicate with other
research teams (in other international institutes) working on this topic. |
|||||
|
|
||||||
|
Design and Programming of Low Level
Controller of an Autonomous Robot |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2014-2015/1 |
|||||
|
Description |
: The aim of this project is to program a Digital
Signal Controller (DSC) so that it can act as a low level controller of an
autonomous robot. The robot is a small size football player which is designed
in our department. On the robot, the DSC is responsible for controlling the
speed of five BLDC motors, IR sensors, ball kicking and communication with
high level controller. Through this project, the student(s) will learn about
principles of microprocessors and controllers, C programming language,
controller design, fundamental of electrical motors. |
|||||
|
|
||||||
|
Detection and Localization of
Small-Sized Soccer Player Robots Using Kinect |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: The very first aim of this project is to test the
ability of detection, localization and controlling multiple autonomous soccer
player robots using Kinect motion sensors. Then, a preliminary software
should be designed for detection and localization of robots and the ball.
Also, it should interpret the position data for each team's high level
controller. C and C++ are the main programming languages in this project
along with OpenCV and OpenNI libraries. |
|||||
|
|
||||||
|
Hand-Controlled Omni-Directional
Autonomous |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: The purpose of this project is to control an
autonomous robot using a bare hand where the hand movement is being detected
by a camera and its movements are interpreted by an interface program for the
robot so that it can move or perform different tasks. Developing feasible
methods for hand detection in diverse positions, controlling autonomous
robot, implementing a Kalman filter for delay compensation, etc. are matters
of interest. The programming languages to be used are C++, Visual C# .Net and
Matlab. Additionally, OpenCV / OpenNI will be required for vision related
programming. The expected outcome of the project is that two persons play air
hockey using their both hands and four autonomous robots with the ability of
shooting a ball. |
|||||
|
|
||||||
|
Squash Playing Robots |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: The purpose of this project is to design a
multi-agent robotic system which is capable of playing a Squash game.
Consider two (or more) robots that can play Squash with each other by
shooting a ball to a wall. The robots should be able to move freely in 2-D
space. Path planning, motion control and prediction, localization, obstacle
avoidance, strategy planing and intelligence play are matters of interest. A
simulator software should be designed in the first step with the intention of
simulating the robot's behavior. In the next step, the simulator will be
turned to a high level controller and strategy planner for the robots. Motion
control and prediction methods, Kalman filter and path planning algorithms
are primary tools which need to be implemented. Visual C# .NET and MATLAB
will be adopted for programming, simulation and software design. The
algorithms are expected to be implemented on SSL autonomous robots that are
originally designed to play football. |
|||||
|
|
||||||
|
Comparison of the Split-Step method
for Various Power Spectral Models |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: In this project, atmospheric turbulence effects will
be modeled using a split-step approach to numerically solve the parabolic
wave equation. Atmospheric turbulence effects are modeled by a number of
phase screens. These phase screens are generated on a numerical grid of
finite size, which corresponds to a narrow spatial area of atmospheric
turbulence. In previous works, the Modified Kolmogorov spectrum has been used
in the modeling. In this project, different spectral models will be used and
the effect of the model on the results will be investigated. Simulation
results will be compared with other simulation models. |
|||||
|
|
||||||
|
Regenerating images distorted by
Atmospheric Turbulence |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: Effects on images deformed by atmospheric turbulence
usually show themselves as geometric distortion or space and time-varying
blurring in addition to the noise. In this project, these two effects are
mitigated through two separate image processing procedures. Considering video
frames as distinct samples of images, an optical flow estimation is used to
discard the geometric distortion with non-local regularization. Then a
spatial invariant blind deconvolution is applied to eliminate the blurring
kernel. The invariancy is obtained through an overlapping patch division and
construction of a diffraction-limited blurred image. In addition to these
operations, a sparse regularization is applied to denoise the input images.
The result image will be free from the diffraction-limited blur with
geometric deformations discarded in the first step. |
|||||
|
|
||||||
|
A single-channel pulmonary sound
signal transmission to computer over Bluetooth |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: Communicating a single-channel pulmonary sound
signal to the computer over Bluetooth (BLE113 and WT12 (both from Bluegiga)
modules and their development kits are available) |
|||||
|
|
||||||
|
Flow inspiration and expiration
estimation |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: Flow (inspiration and expiration) estimation, on the
level of phase estimation (waveform itself is not pursued, but only the
timings of inspiration and expiration are to be found). |
|||||
|
|
||||||
|
Extracting reference spectrum Using pulmonary
sound signals |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/2 |
|||||
|
Description |
: Extracting a reference spectrum for healthy subjects
and comparing it with those of different pathological cases, using the
pulmonary sound signals acquired at 4 locations on the chest wall (the
healthy and pathological data bases are readily available, therefore no
additional data acquisition will be required for this project) |
|||||
|
|
||||||
|
Wireless and Batteryless Motion
Sensing Systems |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/1 |
|||||
|
Description |
: Purpose of this project is to realize a wireless and
batteryless smart sensor system that can run and communicate by generating
energy from frequent daily motions. For this purpose, electromotive force,
which is induced on a coil by a neodymium magnet inside a toroid that has a
winded coil on is going to be used. Instant speed value will be deduced from
the change of the induced power with respect to position. Speed value will be
transmitted with an RF signal at 868 MHz after digitized with a low power
electronic circuit that will be developed for this purpose. For transmission
purposes, a commercial transmitter circuit will be used, which can work with
an electromagnetically generated power and can send its own identification
and sensor information. A wireless and batteryless system that can both
harvest energy and measure instant velocity from the harvested energy has not
been demonstrated before. This way, speed information that comes with daily
motions can be tracked by a computer with existing software at present and
necessary statistical studies can be done in the future. |
|||||
|
|
||||||
|
Microwave Resonator Based Biosensor |
||||||
|
Principal Investigator (PI) |
: Hamdi Torun - hamdi.torun@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/1 |
|||||
|
Description |
: The aim of this project is to develop a biosensor
using a microwave resonator. Selective adsorption of biomolecules on the
surface of the resonator induces changes in electrical characteristics of the
resonator. The change is measured to assess the concentration of
biomolecules. Successful implementation of the project requires design of the
resonator and simulating resonator parameters. The devices will be printed on
a PCB and proof of principle experiments will be conducted using
predetermined molecules already available for the project. |
|||||
|
|
||||||
|
Electromagnetic Manipulator for
Magnetic Microparticles |
||||||
|
Principal Investigator (PI) |
: Hamdi Torun - hamdi.torun@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/1 |
|||||
|
Description |
: The aim of this project is to develop an
electromagnetic manipulator. The manipulator comprises electromagnets that
allow manipulating magnetic microparticles such as microbeads and nanowires
in liquid. Successful implementation of the project requires design of the
electromagnets, having them manufactured and implementing control strategies
to manipulate microparticles. Magnetic microparticles are available for the
project. |
|||||
|
|
||||||
|
Computer Interface for a Voltammetry
or Amperometry Device |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2013-2014/1 |
|||||
|
Description |
: Voltammetry and amperometry are electrochemical
methods to detect and measure small concentrations of substances in solution.
This project involves designing a computer interface for a previously made
voltammetry/amperometry device. There is no time limitation for the project,
and junior/senior students can start right away. Tasks: a) Establish a
connection between the analog outputs of the existing device and a PC by
using a data-acquisition card. b) Program the data-acquisition card
(preferably in Matlab) for obtaining readings from the device. c) Process the
readings by filtering and statistical analyses. d) Design a user panel
screen. e) Plot results from a chemical titration, generate and print a
report. Contact: Prof. Dr. Yasemin Kahya (EE Dept., kahya@boun.edu.tr, tel:
212 3596851) Doç. Dr. Burak Güçlü (Biomedical Eng. Inst.,
burak.guclu@boun.edu.tr, tel: 216 5163467) |
|||||
|
|
||||||
|
CRACKLE OR WHEEZE DETECTION Software |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2012-2013/2 |
|||||
|
Description |
: Motivation - Crackles and wheezes are adventitious
sound components superimposed on background pulmonary sounds (explosive and
musical in character, respectively), and are associated with certain
pulmonary diseases (e.g., crackles with pneumonia and wheezes with asthma).
Their spectral (e.g., pitch for crackles, main frequency component for
wheezes) and temporal (e.g., count and place of occurences in one flow cycle)
characteristics play important role in diagnosis. Aim - To be able to count
them or extract their spectral and temporal characteristics, the first step,
therefore the aim of this project, is to detect crackles or wheezes
individually. For this purpose, suitable time-frequency transformations,
spectral estimation methods, and decision algorithms will be used in the
scope of the project. Either crackle or wheeze detection problem can be
chosen by the student (suitable for the signal processing option) |
|||||
|
|
||||||
|
FLOW PHASE DETECTION Hardware and
Software |
||||||
|
Principal Investigator (PI) |
: Yasemin Kahya - kahya@boun.edu.tr |
|||||
|
Semester |
: 2012-2013/2 |
|||||
|
Description |
: Motivation - An inspiration phase followed by an
expiration phase is called a flow cycle and is measured to synchoronize the
pulmonary sound data acquired on the chest wall. It is measured at the mouth
via a differential pressure transducer, however it adds to the pysical load
of the whole system. To simplify the system, a hardware and/or software
solution to detect inspiration and expiration phases from the pulmonary sound
data is aimed to be designed. Aim - The hardware part will include a single
channel (or two-channel) pulmonary sound acquisition circuitry with a
processor to detect inspiration and expiration phases; and the software part
will include the algorithm design and code implementation for this purpose.
The project can either be assigned to a group of two students, or be divided
and assigned as two independent projects (suitable for both electronics and
signal processing options). |
|||||
|
|
||||||
|
Development of a Wi-Fi Automated
Performance Test Package |
||||||
|
Principal Investigator (PI) |
: Mutlu Koca ve Mehmet Akar |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: Automated performance test package that performs
wireless tests to established values for packet error rate, average
throughput etc. using iPerf, or another software such as the Ruckus
evaluation tool. This project will be conducted in collaboration with
Airties. |
|||||
|
|
||||||
|
Development of a WiFi Coverage Map |
||||||
|
Principal Investigator (PI) |
: Mutlu Koca ve Mehmet Akar |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: A Heat Map application that shows coverage maps of
multiple access points laid out on a floor plan. The application allows the
user to input a floor plan and wall information to generate a floor map. Then
the user inputs AP locations on the floor map. The application receives RSSI
information from multiple access points, calculates attenuation and draws a
coverage map. This project will be conducted in collaboration with Airties. |
|||||
|
|
||||||
|
Development of a WiFi Diagnostic
Platform Tool |
||||||
|
Principal Investigator (PI) |
: Mutlu Koca ve Mehmet Akar |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: A real time Wi-Fi A Diagnostic Platform Tool for
real-time WiFi channel occupancy statistics gathering. The tool generates a
history of channel access, scans multiple channels simultaneously using
channel hopping. This project will be conducted in collaboration with
Airties. |
|||||
|
|
||||||
|
Evaluation of Energy Harvesting
Wireless Switches and Sensors |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: Wireless and batteryless switches and sensors that
receive energy from their surroundings will be investigated. Mechanical,
solar and thermal energy sources will be studied. A commercially available
wireless and batteryless energy harvesting hardware will be evaluated. |
|||||
|
|
||||||
|
Visualization of Electromagnetic
Fields Using A Ferromagnetic Fluid |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: A ferromagnetic fluid containing magnetic micro /
nano particles will be used to visualize electromagnetic fields around coils.
Coils will also be electrically driven to generate different 3D shapes on the
fluid. |
|||||
|
|
||||||
|
Generating Lighting Effects Using 3D
LED Cube |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: LEDs will be placed in three dimensions to form a
cube. Each LED will have a x,y and z coordinate. By turning on and off LEDs
in a predetermined fashing, a three dimentional lighting effects will be
created. |
|||||
|
|
||||||
|
Realization of a Stroboscope Using a
Flashing Power LED |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2012-2013/1 |
|||||
|
Description |
: A stroboscope is an instrument used to make a
periodically moving object appear to be stationary. It can be used for the
study of rotating or vibrating objects. In this project, this instrument will
be realized using a periodically flashing power LED. The flashing frequency
should be adjustable and displayed. |
|||||
|
|
||||||
|
On-Off Keying Modulator Design |
||||||
|
Principal Investigator (PI) |
: Ahmet Öncü - ahmet.oncu@boun.edu.tr |
|||||
|
Semester |
: 2011-2012/2 |
|||||
|
Description |
: The purpose of this project is to design a special
kind of on-off keying modulator for life detection systems used in rescue and
biomedical instrumentation applications. This on-off keying modulator is just
a slightly modified version of a normal on-off keying modulator. The main
idea is the same, the signal is passed or suppressed, but in this special
design, we do not have data. We will have a square wave as an input which has
a period around 1ms. We need the modulator to be in the ON state just after
the rising edge of the input for a duration, namely D. For the rest of the
cycle, the modulator should be in the OFF state. The duration D will be
around 2ns, but we should be able to control this duration. The carrier
frequency will be around 4GHz, but we want the modulator to work without
problems for 2GHz to 7GHz. |
|||||
|
|
||||||
|
Fully Differential Biquad Filter
Design |
||||||
|
Principal Investigator (PI) |
: Ahmet Öncü - ahmet.oncu@boun.edu.tr |
|||||
|
Semester |
: 2011-2012/2 |
|||||
|
Description |
: The purpose of this project is to design a fully
differential biquad filter for life detection systems used in rescue and
biomedical instrumentation applications. The filter will have two corner
frequencies. The lower corner frequency should be around a few mHz and it
should suppress the zero-frequency noise at least 70dB. The upper corner
frequency should be tunable between 10Hz and 100Hz. Main concerns for the
designer will be achieving high zero-frequency unwanted signal suppression
and a low noise design with high CMRR and PSRR. |
|||||
|
|
||||||
|
Sonar Signal Processing |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2011-2012/2 |
|||||
|
Description |
: The purpose of this project is to develop a
sophisticated processing algorithms and tools for sonar signal processing.
The student undertaking the project will work with a sonar signal processing
hardware and software system that has been designed and developed in ISL. |
|||||
|
|
||||||
|
Multirobot Exploration Using Minik
Robots |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2011-2012/2 |
|||||
|
Description |
: The purpose of this project is to integrate the
Minik robots with sensors and use the whole team in a multi-robot scene
acquisition task. The acquisition will be made with cameras mainly, and the exact
robot localization will be found by using a ground truth which will be
developed by the student. This acquisition is expected to end in a dataset
formed by Minik robots. The students need to have intermediate to advanced
level of programming skills as well as some experience with sensors. |
|||||
|
|
||||||
|
Multirobot Simulation Server |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2011-2012/2 |
|||||
|
Description |
: The purpose of this project is to develop control
and analysis software for a previously designed multirobot simulation
environment that also has a web-based interface. The first part of the
project will involve getting the simulators running. The second part of the
project will involve implementing some of these algorithms on the EDAR IIG
robots. |
|||||
|
|
||||||
|
Synchronisation of Flashing Electronic
Oscillators |
||||||
|
Principal Investigator (PI) |
: Yağmur Denizhan - denizhan@boun.edu.tr |
|||||
|
Semester |
: 2011-2012/2 |
|||||
|
Description |
: The purpose of this project is to build a system of
flashing electronic oscillators suitable for studying the synchronisation of
coupled oscillators. The project is inspired by various examples of
synchronisation observed in the nature, including that of the flashing rhythm
of fireflies. |
|||||
|
|
||||||
|
DESIGN OF AN IR OPTICAL RECEIVER
SYSTEM USING DISCRETE COMPONENTS |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The aim of this project is to design an IR optical
receiver system using discrete components. This receiver system will be used
in an optical transmission system which is a part of an MRI catheter
localization architecture. Following tasks will be performed by the student:
(1) Literature study about photodiodes and photodiode amplifiers (2)
Characterization of IR photodiodes and extraction of equivalent circuits (3)
Design of the photodiode amplifier using ADS simulation tool (4) PCB
implementation of the photodiode amplifier and realization of the optical
receiver system |
|||||
|
|
||||||
|
DESIGN OF AN IR LASER DRIVER IN 0.18UM
TRIPLE-WELL TECHNOLOGY |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The purpose of this project will be designing an
Infra-Red laser driver in 0.18um triple-well CMOS technology. The driver will
be used in an optical transmission system which is a part of an MRI catheter
localization architecture. Following tasks will be performed by the
student(s): (1) Literature study about lasers and CMOS laser drivers (2)
Characterization of IR Lasers and extraction of equivalent circuits (3)
Design of CMOS laser driver using ADS simulation tool (4) Drawing the layout
of the circuit (5) Verification of the design by post-layout simulations |
|||||
|
|
||||||
|
DESIGN OF A SUB 1-DB NF ULTRA
LOW-NOISE CMOS AMPLIFIER AT 123 MHZ IN 0.18UM TRIPLE-WELL TECHNOLOGY |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: This project will deal with design and
implementation of an ultra low noise amplifier with sub 1-dB noise figure in
0.18um triple-well CMOS technology. The amplifier will be used to process the
magnetic resonance imaging system operating at 3 Tesla. It is important to
design both the amplifier and the receive antenna which play efficient roles
in increasing the signal-to-noise ratio of the overall receiver block.
Following tasks will be performed by the student(s): (1) Literature study
about low-noise amplifiers and basic MRI operation (2) Design of the ULNA
using an RF circuit simulation tool (3) Drawing the layout of the circuit (4)
Verification of the design by post-layout simulations |
|||||
|
|
||||||
|
Electrochemical Etching of Steel Using
Nano Second Pulses for Nano Patterning |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: Electrochemical etching can be very precise if 10V
pulses of around 10 ns duration is applied between the steel substrate and
the counter electrode that are separated by a sub-micron gap. In this
project, an experimental setup will be used to implement this method. Results
will be analyzed under light and scanning electron microscopes. |
|||||
|
|
||||||
|
Photodiode Characteristics of LEDs and
Their Applications |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: Light Emitting Diodes(LEDs)are diodes that emit
light. At the same time they work very inefficient photodiodes. In this
project photodiode properties of LEDs will be measured using semiconductor
parameter analyzer. Light sensitivity of the LEDs will be determined. Based
on these properties, smart circuits will be built using only LEDs that will
be turned on automatically when the ambient light becomes dark. |
|||||
|
|
||||||
|
Texture Defect Detection by Dual-Tree
Complex Wavelet Transformation and Gaussian Mixture Model |
||||||
|
Principal Investigator (PI) |
: Ayşın Ertüzün - ertuz@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The purpose of this project is to apply complex
wavelets to the defect detection problem. The texture fueatures will be
extracted using dual-tree Complex Wavelet transformation which will be
followed by Gaussina-mixture –model based classifier which labels textile
images as defective or non-defective. The results of dual-tree complex
wavelts will be compared with those of real wavelets. |
|||||
|
|
||||||
|
Development of Algorithms for Decision
Fusion |
||||||
|
Principal Investigator (PI) |
: Ayşın Ertüzün - ertuz@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: Once the distributed detection algorithms have been
developed and analyzed, the decision fusion will be concentrated on. Most
approaches in multiple-sensor based detection in current use rely on a
centralized scheme, where the individual sensors transmit the recorded data
to a central detector, which then makes a decision. This is particularly true
for current systems for the detection of earthquakes. An alternative to
centralized detection is a decentralized scheme, where a set of local
detectors is used to compile individual decisions. The decisions from all
systems are then forwarded to a fusion center, which compiles the overall
decision. The objective of this project is to investigate the use of Bayesian
inference for decision fusion. This motivates the use of particle filters for
posterior probability based fusion.. In this project we will concentrate on
decision fusion algorithms which will reduce the false alarm and the
misdetection rate in the context of early detection of earthquake. |
|||||
|
|
||||||
|
Distributed Detection Using Particle
Filters |
||||||
|
Principal Investigator (PI) |
: Ayşın Ertüzün - ertuz@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: Distributed detection in impulsive noise
environments remains an open problem. The presence of impulsive
(heavy-tailed) noise can cause conventional techniques to break down. Hence
robust detectors are required. These techniques should perform near optimally
in the Gaussian case and should not degrade significantly if the underlying
noise distribution changes. Most current detection algorithms assume the
Gaussian distribution. The objective of this project is to consider
distributed detection in a non-Gaussian environment and develop suitable
detectors at the nodes of the sensor network. We intend to solve the
distributed detection problem using Bayesian inference and the applicability
of sequential Monte Carlo estimation, namely, the relevance of particle
filters to this problem will be analyzed. |
|||||
|
|
||||||
|
Multi-Robot Exploration Using Minik
Robots |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The purpose of this project is to integrate the
Minik robots with sensors and use the whole team in a multi-robot scene
exploration task. The students need to have intermediate to advanced level of
programming skills as well as some experience with sensors. |
|||||
|
|
||||||
|
Computer Based LED Control |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The purpose of this project is to design and develop
the electronics and software for computer based control of a LED based
lighting system. The student should have intermediate to advanced electronic
circuit design and debugging capabilities as well as some experience with
microprocessor/ FPGA programming skills. |
|||||
|
|
||||||
|
2D Realization Server |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The purpose of this project is to develop control
and analysis software for a previously designed for automated inspection of
PCBs in order to expand the capabilities of the system – including
multi-platform image acquisition and learning. The student needs to have
intermediate to advanced C/C++ programming skill |
|||||
|
|
||||||
|
Automated PCB Inspection |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2010-2011/1 |
|||||
|
Description |
: The purpose of this project is to develop control
and analysis software for a previously designed for automated inspection of
PCBs in order to expand the capabilities of the system – including
multi-platform image acquisition and learning. The software will have two
parts: 1- Hardware control, image acquisition and processing. 2- Graphical
User Interface. The students need to have intermediate to advanced C/C++
programming skills. |
|||||
|
|
||||||
|
MICROFABRICATED FIBER OPTIC
TRANSMISSION PLATFORM |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2009-2010/1 |
|||||
|
Description |
: This project aims to integrate the fiber cables of
125 micron diameter with a microfabricated Silicon platform for optical
communication and power transmission. Such a system is a crucial ingredient
of the optically powered receiver to be used in magnetic resonance imaging
system. Since metals are heated in the presence of RF magnetic fields, the reciever
which is held in the MRI dequipment has to communicate with external
equipments using optical (i.e. conductor free) paths. Following tasks will be
performed by the student: * Litterature study about fiber optic alignment,
photodiodes and LEDs * Optical design of the proposed system * Design of
lithography masks for KOH etch process * KOH etch of Silicon substrates * PCB
Base for wire bonding and CMOS IC / Fiber platfor integration * Optical
measurments for power and signal transmission |
|||||
|
|
||||||
|
DESIGN OF A SUB 1-DB NF ULTRA
LOW-NOISE AMPLIFIER AT 123 MHZ |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2009-2010/1 |
|||||
|
Description |
: This project will deal with design and
implementation of an ultra low noise amplifier with sub 1-dB noise figure.
The amplifier will be used to process the magnetic resonance imaging system
operating at 3 Tesla. It is important to design both the amplifier and the
receive antenna which play efficient roles in increasing the signal-to-noise
ratio of the overall receiver block. Following tasks will be performed by the
student: * Litterature study about low-noise amplifiers and basic MRI
operation * Design of the ULNA using ADS simulation tool * PCB implementation
of the circuit * Measurement of the characteristics of the amplifier using
Vector Network Analyzer and Noise Figure Meter |
|||||
|
|
||||||
|
DESIGN AND FABRICATION OF
GLASS/POLYMER MICROSCANNERS |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2009-2010/1 |
|||||
|
Description |
: Scanned light beams are used to produce display
images for a wide variety of applications, including such applications as automotive
head-up displays and head-worn displays. A micromirror producing angular
motion is used to deflect a modulated light beam on an image plane to create
a display. In order to construct a two-dimensional rasterlike rectangular
image, the mirror is rotated about two orthogonal axes at different
frequencies. In this project, we aim to produce a 2D laser scanner composed
of 2 different materials, namely glass and FR4. The design will aim a
slow-scan (horizontal-scan) frequency of 60 Hz (polymer part) and a fast-scan
(vertical-scan) frequency of 8 kHz and above (glass part). Following tasks
will be completed: * A litterature search on glass wet eching/micromachining
techniques * Process characterization of glass substrates * 2D scanner design
and simulations * Layout design of the photomasks for fabrication *
Fabrication of glass MEMS structure and integration with polymer substrate *
Characterization of the device |
|||||
|
|
||||||
|
Digital Rights Protection Project:
Measuring the Performance of Existing Fingerprinting Methods |
||||||
|
Principal Investigator (PI) |
: F. Kerem Harmancı - harmanci@boun.edu.tr |
|||||
|
Semester |
: 2009-2010/1 |
|||||
|
Description |
: This project is about digital rights management of
images. The objective is to test the performance of existing fingerprinting
techniques for collusion resistance and detection performance under different
geometric attacks such as rotation, scaling, cropping and translation.
Fingerprinting enforces digital rights whereby unique labels, known as
digital fingerprints are inserted (watermarked) into the image prior to
distribution. |
|||||
|
|
||||||
|
Collaborative Control Applications in
Mobile Robot Networks |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2009-2010/1 |
|||||
|
Description |
: The objective of this project is to advance the
current state of the BRocks team which competed in Robocup 2009 that was held
in Austria. RoboCup is a competition platform whose ultimate goal is to
develop a team of fully autonomous humanoid robots that can win against the
human world champion team in soccer by the year 2050. Competitions are held
in various categories, including SSL in which the competitors build their own
robots. We are looking for motivated students who are willing to carry out
their 491/492 projects in one of the following areas: 1) PCB design and
development (electronics or controls option) 2) Image processing
(communications or controls options) 3) brushless DC motor control and higher
level strategy development. |
|||||
|
|
||||||
|
Coordinated Motion of 3D Particles
Robot Flocks and Protein Folding |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/2 |
|||||
|
Description |
: The purpose of this project is to study the motion
of particles in 3-dimensional space with two applications in mind: motion
planning of a flock of robots and protein folding. The student(s) will be participating
in an ongoing MS project which has been studying the problem for some time
and using/will be doing code development on a Java-based simulator designed
and built with this purpose. |
|||||
|
|
||||||
|
Automated PCB Inspection |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/2 |
|||||
|
Description |
: The purpose of this project is to work on a system
that has been previously designed for automated inspection of PCBs in order
to expand the capabilities of the system – including multi-platform image
acquisition and learning. |
|||||
|
|
||||||
|
Tracked EDAR-G3 Robot |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/2 |
|||||
|
Description |
: The purpose of this project is to work on the
systems development of our third generation tracked robot EDAR G3. This will
be a coordinated effort with a group of ME students.The student(s) will be
responsible for getting the EDAR electronics (main CPU card, driver cards, dc
motors for the linear and rotational motion) into working status. |
|||||
|
|
||||||
|
Computer Based LED Control |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/2 |
|||||
|
Description |
: The purpose of this project is to design and deveop
the electronics and software for computer based control of a LED based
lighting system. |
|||||
|
|
||||||
|
Power-line based Remote Device Access |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/2 |
|||||
|
Description |
: The purpose of this project is to design and deveop
the electronics and software for power-line based remote device access via
X10 protocol. |
|||||
|
|
||||||
|
Design and Realization of Low-Power
Low-Noise Amplifiers |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The low noise amplifier (LNA) is a special type of
electronic amplifier used in communication and medical systems to amplify
very weak signals. The low power LNA is one of the main components of medical
systems. The use of discrete components in LNA design is problematic due to
various parasitic effects caused by interconnections between components. On
the other hand, when a specific LNA design is needed, I:C design can be time
consuming and expensive. In this project a low power LNA will be designed by
using discrete circuit components and the design will be compared with IC
counterparts. The aim of the project is investigating low power LNA design
and their possible use in medical systems. During the project, student needs
to complete the following tasks: - Literature search about low power LNA
design - Determination of the frequency and power characteristics of the LNA
- Design and SPICE simulations of various LNA topologies (incorporating the
circuit parasitics) - Circuit realization and test of designed LNAs - IC
Layout of the designed LNA and post layout simulation and comparisons |
|||||
|
|
||||||
|
Design and Realization of Laser-based
Optical Energy Scavenging Systems |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Energy scavenging is the process by which energy is
captured and stored. A variety of different sources exist for scavenging
energy, such as solar power, thermal energy, wind energy, salinity gradients
and kinetic energy. Solar energy scavenging is a common and well known
optical energy scavenging method. On the other hand, laser based optical
energy scavenging systems, which can be used in various portable applications
or/and medical applications, are rare in literature. In this project, an
electrical circuit which is powered by a laser-based optical energy
scavenging system will be designed. In the project lasers will be used as the
power source. The optical power will be converted to the electrical power by
the photodiodes in order to be used in an electric circuit. The aim of the
project is investigating the laser-based optical energy scavenging methods
and stating their advantages or disadvantages over the other energy
scavenging systems, such as capacitive energy scavenging. During the project,
student needs to complete the following tasks: - Literature survey of energy
scavenging systems - Design and SPICE simulations of various voltage and
current multiplier configurations - Circuit realization and test of designed
voltage and current multipliers - Design and SPICE simulations of an optical
energy scavenging system (incorporates the voltage (or current) multiplier) -
Circuit realization and test of designed optical energy scavenging system -
Design of a test circuit which powered by realized scavenging system. |
|||||
|
|
||||||
|
Design and Fabrication of a PolymerLED
integrated miniaturized MEMS Display |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Light emitting Diodes in which organic compounds are
used as the emitting layer material are known as an Organic Light Emitting
Diode (OLED). To function as a semiconductor, the organic emitting material
must have conjugated pi bonds. The emitting material can be a small organic
molecule in a crystalline phase, or a polymer. Polymer materials can be
flexible; such LEDs are known as PLEDs or FLEDs. Compared with regular LEDs,
OLEDs are lighter, and polymer LEDs can have the added benefit of being
flexible. Today, OLEDs have been used to produce visual displays for portable
electronic devices such as cellphones, digital cameras, and MP3 players. In
this project, In this project, a miniaturized MEMS display will be designed
and fabricated. The design will aim a slow-scan (horizontal-scan) frequency
of 60 Hz and will consider the integration of OLED with MEMS actuator. During
the project, student needs to complete the following tasks: - Design of the
MEMS actuator analytically - Finite Element MEthod (FEM) simulations of the
device - Layout design of the photomasks for fabrication - OLED layout design
and fabrication - OLED and MEMS actuator integration |
|||||
|
|
||||||
|
Analog and Digital LED Driving &
Modulation Methods |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: LEDs are widely used as indicator lights on
electronic devices and increasingly in higher power applications such as
flashlights and area lighting. In this project LEDs are integrated on a MEMS
actuator and to create together a novel type of minituarized display. MEMS
actuator functions as 1D scanner in vertical axis at 60HZ of scanning
frequency. A LED array integrated on this actuator will be modulated
approximately at 8kHz to create an horizontal scan axis in order to be able
to produce a 2D MEMS miniaturized display. There will be two distinct parts
of this project. LED dies will be driven and modulated using both analog and
digital methods. At the end of the project, these methods will be compared
regarding their performance and fabrication convenience. As for the digital
driving and modulation, the student will do the followings: - MAX6946 LED
driver development board tests; - Learn MAX6946 working principles and
programming techniques. - Making a PCB using MAX6946 and SMD LEDs (or LED
dies,wire bonding included!); - Power calculations and measurements. -
Current vs Luminance measurements using integrating sphere. - Contrast ratio
calculations and measurements. For the anolog part: - Designing an analog LED
driver/modulator; - Simulations and calculations - Making a PCB using
discrete active and passive components, SMD LEDs(or LED dies,wire bonding
included!); - Power calculations and measurements. - Current vs Luminance
measurements using integrating sphere. - Contrast ratio calculations and
measurements. |
|||||
|
|
||||||
|
Simulated and Experimental Study of
Boost Converter Using Sliding Mode Control |
||||||
|
Principal Investigator (PI) |
: Okyay Kaynak - kaynak@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: A boost converter (step-up converter) is a power
converter with an output DC voltage greater than its input DC voltage. The
purpose of this project is to design a sliding mode controller for a Boost
Converter and implement it on hardware by using a dSPACE card. The project
will consist of two main stages: 1. Simulation studies of the overall (open
loop and closed loop) system and open loop hardware design and its
implementation, 2. Closed loop control of the hardware by using a dSPACE
card. |
|||||
|
|
||||||
|
Efficient Wireless Video Transmission
for home entertainment systems |
||||||
|
Principal Investigator (PI) |
: Emin Anarım - anarim@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Project PIs: Emin Anarim, Tolga Kurt (Airties),
Mutlu Koca, Mehmet Akar. Wireless video distribution in home networks is one
of the upcoming trends of home entertainment systems. One of the major
problems in wireless video distribution over Wi-Fi networks is interference.
Wireless video unlike data have very low tolerance to jitter and delay.
Therefore, interference should be avoided as much as possible. Among the
available channels, the best one should be selected automatically without
affecting the existing streaming video transmission. This selection and
channel change procedure should be effectively handled in multi-hop mesh
networks utilizing IEEE 802.11n standard. Airties Wireless Networks and
Boğaziçi University will be investigating these issues in a joint
project. The research and development activities related to the project.
Specific research topics can be listed as follows: · Interference avoidance
in single-hop wireless video distribution · Interference avoidance in
multi-hop wireless video distribution · Optimizing transmission modes for
wireless video · Analyzing further issues with wireless video distribution ·
Generating sub-products to the wireless video roadmap The project is made up
of three phases: the preliminary investigation in the first two months, the
test and analysis in the second two months, followed by the development in
the final stage. |
|||||
|
|
||||||
|
DESIGN OF STEEL MICROMIRRORS |
||||||
|
Principal Investigator (PI) |
: Arda Yalçınkaya - arda.yalcinkaya@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Scanned light beams are used to produce display
images for a wide variety of applications, including such applications as
automotive head-up displays and head-worn displays. A micromirror producing
angular motion is used to deflect a modulated light beam on an image plane to
create a display. In order to construct a two-dimensional rasterlike
rectangular image, the mirror is rotated about two orthogonal axes at
different frequencies. In this project, a micro-mirror to be fabricated using
a steel machining process, will be designed. The design will aim a slow-scan
(horizontal-scan) frequency of 60 Hz and a fast-scan (vertical-scan)
frequency of 4 to 8 kHz. In this project, student needs to complete the
following tasks: - Design of the mirror scanner analytically - Finite Element
MEthod (FEM) simulations of the device - Layout design of the photomasks for
fabrication |
|||||
|
|
||||||
|
Portable Laser Projector |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: In this project an already micromachined steel micro
mirror will be used to form a 2D image using a laser-diode and the necessary
optics. The steel micro mirror is actuated electromagnetically. The mirror is
designed so that it can scan both in x and y directions depending on the
frequency of the applied electrical signal. As the mirror is actuated, it
steers the laser beam to form a 2D image on a wall. Circuits needed for the
actuation of the micro mirror and circuits needed to drive the laser have
already been implemented. In this project, students need to do the following:
-To form a USB link between a computer and the portable laser projector. -To
prepare a GUI (Graphical User Interface) on the computer. The pictures or
texts displayed on this GUI will be projected on a wall by the laser
projection system. -To design a compact package that houses the micro mirror,
laser and the necessary lens. |
|||||
|
|
||||||
|
Design of a Small Electric Monopole or
a Magnetic Dipole Antenna |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop an
electric monopole or a magnetic dipole small antenna using a simple design.
An electrically small antenna would also contain an inductor to reduce the
operating frequency without enlarging the size of the antenna. Similarly the
operating frequency of a magnetic small antenna can be controlled with the
use of capacitors. The antennas can be made of pure copper or they can be
processed on the Roger 5880 Duroid plate which is covered by copper with the
thickness of 17ƒİm. During the design process, Ansoft HFSS and Ansoft
Designer softwares will be used extensively for the optimization of the
design at the desired resonant frequency of the antenna. Ansoft HFSS is a
software which allows the user to draw three dimensional antenna samples and
then run their radiation simulations. The simulation results should contain
information about the 2D and 3D radiation patterns of the structure, the
S-Parameters in a logarithmic scale, and the radiation efficiency or the
directivity of the antenna. The Ansoft Designer needs to be used for the
simulations when the antenna structure contains lumped circuit elements such
as resistors, capacitors or inductors. |
|||||
|
|
||||||
|
Analyzing Turbulence in a One Meter
Water Tube For Free Space Optical Communication Applications |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Direct line-of-sight optical communication links,
which are commonly called “optical wireless” systems or free space optical
(FSO) communication links are becoming increasingly popular. Such links can
provide virtually unlimited bandwidth at a relatively low cost and high
performance communication over short distances up to a few kilometers.
Unfortunately, the atmosphere is not an ideal communication channel.
Atmospheric turbulence can cause fluctuations in the received signal level,
which increase the bit errors in a digital communication link. In order to
quantify the performance limitations, a better understanding of the effect of
the intensity fluctuations on the received signal at all turbulence levels is
needed. The purpose of this project is to develop an experimental setup to
analyze turbulence and aperture averaging of a He-Ne Laser Beam propagating
through a heated 1-meter water tube emulating 1km of free space. The setup
will give the convenience of controlling the turbulence level in a lab
setting through adjusting the water temperature instead of relying on the height
above the ground or the air temperature in a free-space setting. The
turbulence index parameter Cn2 should be very different for water than it
would be for air, but the effect of the receiver size on power variance
“Aperture Averaging”, should exhibit similarities in both scenarios. |
|||||
|
|
||||||
|
An Optical Multi-Touch Sensor
Implementation |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop a novel
multi-point optical touch sensor using a simple design. A touch sensor can be
developed using LEDs, plastic optical fibers, photo-detectors, and a
micro-controller. The sensor should use the micro bending concept of optical
fibers in a two dimensional mesh orientation where the bends of the fibers
are comparable to its radius so that energy coupling and radiation occurs
into the cladding at the bending regions. As a result bends act as “sensing
points” for human touch since the power absorbed by the human skin at these
points will decrease the power received at the end of the fiber. |
|||||
|
|
||||||
|
A Geometrical Optics Approach for
Modeling Aperture Averaging in Free Space Optical Communication Applications |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Intensity scintillation and beam wander caused by
atmospheric turbulence are two significant phenomena that affect free space
optical (FSO) communication links. Intensity fluctuations at the receiver
lead to a received power variance that depends on the size of the receiver
aperture. Increasing the size of the receiver aperture reduces the power
variance. This effect of the receiver size on power variance is called
aperture averaging. If there were no aperture size limitation at the
receiver, then there would be no turbulence-induced scintillation. In
practice, there is always a tradeoff between aperture size, transceiver weight,
and potential transceiver agility for pointing, acquisition and tracking
(PAT) of FSO communication links. The purpose of this project is to develop a
geometrical simulation model to predict the aperture averaging factor and
beam wander of a Gaussian beam propagating through a simulated turbulent
medium. |
|||||
|
|
||||||
|
Image Analysis of Intensity
Scintillation Patterns on a Free Space Optical Communication Link |
||||||
|
Principal Investigator (PI) |
: Heba Yüksel - heba.yuksel@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Even after several decades of study, inconsistencies
remain in the application of atmospheric turbulence theories to experimental
systems, and the demonstration of acceptable correlations with experimental
results. An imaging system for measuring the effects of atmospheric
turbulence and obscuration on a free space optical (FSO) communication link
has been constructed. A He-Ne laser beam propagates over a range of 863
meters in atmospheric turbulence conditions that vary diurnally and
seasonally from weak to strong. A high performance digital camera with a
frame-grabbing computer interface is used to capture received laser intensity
distributions at rates up to 30 frames per second and various short shutter
speeds, down to 1/16,000s per frame. In this project, a LabVIEW code should
be developed to analyze the captured image frames from the given FSO link to
evaluate the turbulence index parameter Cn2, temporal and spatial intensity
variances, and aperture averaging. Such analysis should be compared with
existing theories for the development of new theories. |
|||||
|
|
||||||
|
Bayesian Texture Defect Detection |
||||||
|
Principal Investigator (PI) |
: Ayşın Ertüzün - ertuz@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to apply Bayesian
concepts to automated quality inspection more specifically to defect
detection problem in textile images. |
|||||
|
|
||||||
|
Texture Defect Detection by Dual-Tree Complex
Wavelet Transformation |
||||||
|
Principal Investigator (PI) |
: Ayşın Ertüzün - ertuz@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to apply complex
wavelets to the defect detection problem. Complex wavelets will be compared
with real wavelets. They both will be used for extracting features from
images and for image classification to label defective images. |
|||||
|
|
||||||
|
CCA Based Texture Defect Detection |
||||||
|
Principal Investigator (PI) |
: Ayşın Ertüzün - ertuz@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to apply Canonical
Correlation Analysis (CCA) concepts to automated quality inspection more
specifically to defect detection problem in textile images. |
|||||
|
|
||||||
|
Measuring Rotation Rate and
Accleration of an Object Using Gyroscopes and Accelerometers |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Gyroscopes and acceloremeters are part of inertial
sensors. Gyroscopes measure rotation rates. Accelerometers measure
acclerations. In this project, using commercially available MEMS gryoscopes
and accelerometers, rotation rate and accleration of an object will be
measured. Electronic readout circuit for these inertial sensors will be made
on a PCB. Measurements made with the circuit will be transfered to a
computer. A grafical user interface (GUI) will be implemented to display
these measurements real time on a computer. For example if you rotate the
object, on which the sensors are, a virtual object on the computer will
simultaneously rotate at the same rotation rate as the real object. |
|||||
|
|
||||||
|
Inductively Coupled Wirelessly Powered
Light Emitting Diodes and Sensors |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop an
inductively coupled wireless system. The first simple system will power a
light emitting diode wirelessly that is couple centimeters away. The second
system should be able to power up a wireless circuit that has no battery and
should be able to read the digital or the sensor data stored in that circuit.
This system is similar to RFID systems. |
|||||
|
|
||||||
|
Speech recognition driven reading
tutor |
||||||
|
Principal Investigator (PI) |
: Murat Saraçlar - murat.saraclar@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop a reading
tutor that uses automatic speech recognition to verify the user inputs. The
tutor should provide feedback, encouraging students, and pointing out
mistakes. The underlying speech recognizer should be accurate enough to judge
the correctness of the inputs. |
|||||
|
|
||||||
|
Digital Oscilloscope Adapter Unit for
a PC |
||||||
|
Principal Investigator (PI) |
: Avni Morgül - morgul@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to produce a digital
storage oscilloscope to operate as an adapter to any PC. The project consists
of three tasks: i) Design and construct a 9-bits or 10
bits Analog-to-Digital converter block with high input impedance and high
speed. ii) Design and construct a USB
interface between the A/D converter hardware and PC and writing the necessary
software modul for the interface. iii) Writing the high level software
on the PC to display the waveforms and do the necessary measurements and
calculations (Frequency, amplitude, RMS, FFT etc.) |
|||||
|
|
||||||
|
Design and Implementation of
Log-Domain Analog Filter Structures |
||||||
|
Principal Investigator (PI) |
: Oğuzhan Çiçekoğlu - cicekogl@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Log-domain filters are a sub-class of externally
linear internally non-linear (ELIN) filters. In this class of filters the
signal is processed in a nonlinear way, usually in compressed form to avoid
distortion due to large signal peaks. However the input output transfer
function is a linear transfer function. In this type of filters it is
possible to build analog filters employing only bipolar transistors and
capacitors. Therefore very simple filter structures are possible. Mainly two
techniques are well-known in the literature. The purpose of the project is to
design several biquadratic filters in the log domain using the well-known
techniques. The designed filters will be tested and improved if necessary.
The designs will be compared with those in the literature, discussed and
improvement possibilities will be studied. ELIN filters attract increased
attention since electronic components are themselves nonlinear and this
non-linearity is directly used in the designs. The design philosophy is
therefore completely different from the classical approach that was dominant
in the last few decades. |
|||||
|
|
||||||
|
Improved Design of Chaotic Circuits |
||||||
|
Principal Investigator (PI) |
: Oğuzhan Çiçekoğlu - cicekogl@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Description of the project: The purpose of this
project is to investigate the operation of various chaos-generating circuits
in the literature. Chaotic signal generation circuits find application in
secure communication instrumentation. Chua circuit, piecewise linear Rossler
circuit, hysteretic circuit, piecewise linear Duffing, 2 attractor circuit
and others will be built in the laboratory and tested. Improvement and
simplification possibility of these circuits will be taken into
consideration. The possibilities to modify some circuit parameters in an
analog manner will be investigated. Next step is to design and propose
alternative circuits to those well known. Comparison in terms of various
performance parameters is to be made. |
|||||
|
|
||||||
|
Solutions for Signal Limitations in
Current Conveyor Based Filters |
||||||
|
Principal Investigator (PI) |
: Oğuzhan Çiçekoğlu - cicekogl@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Description of the project: Current conveyor based analog
filters are attracting increasing attention due to their inherent wide
bandwidth and greater linearity. They are accepted to be alternatives to the
well-known opamps. Many current conveyor based analog filters are presented
in the literature and still newer circuits are proposed. Due to the advances
in the technology toward low voltage low power systems signal limitation
problem is today more important than in the past. Decrease in the supply
voltages reduce the peak-to-peak voltage swing possibility whereas the noise
floor remains nearly constant. The dynamic range is reduced and this in turn
reduces the signal processing quality of the filters. |
|||||
|
|
||||||
|
Nonlinear Systems Analysis and Control
Tool Kit |
||||||
|
Principal Investigator (PI) |
: Yağmur Denizhan - denizhan@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop a series
of computer programs that allow the simulation of nonlinear systems,
accomplish some typical tasks (like estimation of local dynamics, detection
of equilibrium points and periodic orbits, computation of Lyapunov exponents
etc.) needed for the analysis of nonlinear systems and allow the application
of some typical nonlinear control methods. The resulting tool kit will be
used in graduate courses on analysis and control of nonlinear systems (like
SCO 591 and EE 591), as well as a simulation frame in MS and PhD theses on
related areas. |
|||||
|
|
||||||
|
Design and implementation of GSM
blocking circuit |
||||||
|
Principal Investigator (PI) |
: Avni Morgül - morgul@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to design and realize
an electronic circuit to jam and block the operation of GSM portable phones
in classrooms, conference halls, theaters etc.. |
|||||
|
|
||||||
|
Automatic Speech Recognition of
Turkish Broadcast News using Open Source Software |
||||||
|
Principal Investigator (PI) |
: Murat Saraçlar - murat.saraclar@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop an
automatic speech recognition system using open source software. The system
will be trained and tested using the Turkish Broadcast News corpus being
collected at BUEE. The performance of the system will be compared with our
best system trained on the same corpus. |
|||||
|
|
||||||
|
Background wideband noise measurement
and evaluation for UWB applications |
||||||
|
Principal Investigator (PI) |
: Mutlu Koca - mutlu.koca@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to measure and
calculate long time fluctuations and average of the background noise of
different frequency bands and different environments. |
|||||
|
|
||||||
|
EM Field measurement in primary
schools which are very close to transmitters and GSM Base stations |
||||||
|
Principal Investigator (PI) |
: Avni Morgül - morgul@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to measure long time
fluctuatins and average of the electromagnetic pollution at a number of
primary schools located near the Base stations and TV transmitters. |
|||||
|
|
||||||
|
Spinning Display (Spindis) |
||||||
|
Principal Investigator (PI) |
: Şenol Mutlu - senol.mutlu@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to produce a flat
display with a simple method. The method is to attach a small Light Emitting
Diode (LED) display on a rotating plate. As the plate rotates, the LEDs are
turned on and off in a calculated manner to display a TV quality image.
According to the text and/or image information, a micro processor calculates
which LEDs will be on and off as the display panel rotates with a constant
angular velocity of w. The rotation is achieved with a motor. |
|||||
|
|
||||||
|
Smart House |
||||||
|
Principal Investigator (PI) |
: Levent Arslan - arslanle@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to build the
electrical and software infrastructure for a smart house. Possible pieces may
include a centrally controllable lighting system, microphone array design for
voice recognition and speech synthesis to interact with the appliances,
entertainment system which allows individual configurations in different
rooms. |
|||||
|
|
||||||
|
A Simulation Platform for Robot
Football |
||||||
|
Principal Investigator (PI) |
: Mehmet Akar - mehmet.akar@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop a software
package that allows the simulation of two robot teams playing football
against eachother. The simulator should allow modules for control,
communication and vision information to be transmitted to and from the
robots. |
|||||
|
|
||||||
|
Brain Computer Interface |
||||||
|
Principal Investigator (PI) |
: Bülent Sankur - bulent.sankur@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to develop electronic
devices and/or signal processing tools for understanding the functioning of
the brain during cognitive and emotional tasks. The first goal is “affective
computing”, where the computer understands and track the state of the mind of
the patient and where the patient may communicate some information to the
computer. In addition, clinical problems, such as attention deficit patients
or onset of Alzheimer’s disease can be investigated. The student can
potentially pursue his/her d-research in MS and PhD theses. |
|||||
|
|
||||||
|
Coordinated Motion of 3D Particles
Robot Flocks and Protein Folding |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to study the motion
of particles in 3-dimensional space with two applications in mind: motion
planning of a flock of robots and protein folding. The student(s) will be
participating in an ongoing MS project which has been studying the problem
for some time and using/will be doing code development on a Java-based
simulator designed and built with this purpose as well as a design on Webots
-- a robot simulator environment. |
|||||
|
|
||||||
|
Robot Sensory Feedback |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project endow our second
generation robot EDAR G2 with various sensory feedback capabilities –
including limit switchs for the limitation of coordinate extensions, voltage
level detection of the batteries and cameras for visual feedback. |
|||||
|
|
||||||
|
Acoustic Condition and Speaker
Segmentation in Broadcast News |
||||||
|
Principal Investigator (PI) |
: Murat Saraçlar - murat.saraclar@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: The purpose of this project is to segment broadcast
news programs. Ideally each segment should include a single speaker and be
marked by the speaker identity. In addition each segment should have similar
acoustic conditions (i.e channel, background, style). This will enable us to
adapt our baseline speech recognition system to individual speakers/acoustic
conditions. |
|||||
|
|
||||||
|
Automated Inspection and Sorting |
||||||
|
Principal Investigator (PI) |
: H. Işıl Bozma - bozma@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Thepurpose of this project to develop a real-time
vision-based automated inspection system using color cameras and Janome
robot. The visual data coming from cameras will be processed in order to
either i) learn new objects or ii) recognize an object based on the already
learned set of classes. |
|||||
|
|
||||||
|
Realization of a 200 Watt LLC Resonant
Switch Mode Power Supply for LCD Television Sets. |
||||||
|
Principal Investigator (PI) |
: Ömer Cerid - cerid@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Description of the project: This power supply may be
used by BEKO if successful. The power supply will be composed of three
sections: a) Power factor correction (PFC) b) LLC resonant main supply c)
Standy-by auxiliary flyback type supply |
|||||
|
|
||||||
|
8-bit Microcontroller Controlled
Monochrome 10x56 Dot Matrix LED Display with Graphic Capabilities and RS-485
Multidrop Communication. |
||||||
|
Principal Investigator (PI) |
: Ömer Cerid - cerid@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Not available. Contact the PI. |
|||||
|
|
||||||
|
Microcontroller Controlled Chip Card
Secure Door Entry System for BETA Laboratory. |
||||||
|
Principal Investigator (PI) |
: Ömer Cerid - cerid@boun.edu.tr |
|||||
|
Semester |
: 2008-2009/1 |
|||||
|
Description |
: Not available. Contact the PI. |
|||||
|
|
||||||
