The main interest of our research group is in model
predictive control, as
one of the tools in optimization
problem solving, and in machine learning
techniques, as a support mechanism in
different decision making processes. Ultimately,
our group aims at developing an optimization
framework in which these two approaches are
combined, in order to get more efficient and
more effective solutions in more complex and
uncertain scenarios.
Some of our past work includes model predictive
control in unit
commitment problem solving
on an array of multiple hydropower plants, as
well as control of specific processes in thermo and solar power plants.
Other examples include medical applications,
like feedback control of the glucose level in
humans. Another
very specific area of interest is
mobile vehicle motion planning. As part of my
PhD research I used model predictive control to
address the problem of mobile ground vehicle
motion planning on rough terrains. The
research resulted in a novel MPC-based
optimization framework capable of navigating
the vehicle through both known and unknown
rough terrains. In
order to use
the MPC optimization framework on terrains of a
larger scale, an appropriate cost map was
developed forming terminal costs within the
objective function. The initial work was
performed in cooperation with the Politecnico
di Milano DEIB Merlin research group and
the
robotics group at JPL/NASA.
Currently, an extensions to the algorithm is
being developed and
is performed in collaboration with the Merlin
research group. Another novel coverage
planner for multi-vehicle navigation
planning, intended to cover a given region
of interest in an efficient and coordinated
manner, was developed as part of a
research stay at Imperial
College London.
Our group now
adapted this algorithm in constrained
coverage problems by using model predictive
control.
When it comes to machine learning
applications, some of the techniques were used
in addressing problems of extracting brain
signals with a low signal to noise ratio. As an
example, an approach based on Principal
component analysis combined
with Independent component analysis was
developed in cooperation with Airlab
at Politecnico di Milano to improve
a P300 speller device. Other experience
involves industrial applications in working
with Bayes classifiers, logistic
regression, neural networks and other
classification and regression tools, while
solving some decision making and control
problems. Neural networks in cruise
control were implemented and the results
validated on an Audi
A8 in
Ingolstadt in cooperation with the
Friedrich-Alexander-Universität Erlangen-Nürnberg.
Currently, our research group is working on
ideas in the field of nonlinear control
theory, in combining model predictive
control and some available approximate
nonlinear control techniques such as SDRE
control. Other ideas involve the use of model
predictive control supported by sample based
techniques such as RRT (rapidly exploring
random trees), a well known tool in the robot
motion planning community. Ultimately, our
research group is aiming at developing a novel chance
constrained model predictive control framework
by using classification techniques.
Current research involves as well work on the
development of a fault tolerant based planner for
a quadcopter, to solve the underwater
coverage problem by using an underwater vehicle.
The focus thereby is to find an appropriate
cooperative control framework to guide the
aerial vehicle while it physically interacts
with the underwater vehicle. The work is done as
part of the
MORUS project.
CURRENT COLLABORATORS
|
PAST COLLABORATORS
|
- Mehmed Brkić (PhD student at ETF
Sarajevo))
Model predictive control for coverage
problem
Teaching assistant |
- Aida Čaršimamović (Telecom BH,
BiH)
MPC in hydropower plants -
master thesis |
-
Nedim Osmić (ETF Sarajevo and PhD student at FER))
Fault tolerant control |
-
Naida Škaljić (Petrolinvest, BiH)
MPC in biomedical systems - master thesis |
- Dinko Osmanković, PhD (ETF
Sarajevo)
Development of a MPC solver |
- Goran Huskić (PhD student at
University of Tübingen, Germany)
Flatness-based control - master
thesis |
-
Almir Salihbegović (PhD student at ETF Sarajevo))
Robust Control |
- Amel Selimović (Javno preduzeće za
prostorno planiranje i uređenje grada
Zenica, BiH)
Terrain classification for mobile
vehicle planning on rough terrains -
master thesis |
-
Nadir Kapetanović (PhD student at FER Zagreb)
MPC motion planning on rough terrains - master thesis
Cost map construction on rough
terains and underwater environments |
- Selma Musić (PhD student at TUM,
Germany)
Automated Insertion of Geophones using
Light-Weight Robots - master
thesis |
-
Armin Dajić (master student at ETF Sarajevo))
Cooperative control |
- Faris Delić (Visoko, BiH)
MPC in thermal power plants -
master thesis
|
-
Samir Džudzadanović (KV TEAM)
Nonliear MPC - master thesis
Model predictive control using state
dependent Riccatti equation |
- Benjamin Seferagić (Kachel
GmbH, Germany)
Constrained coverage problem with
multiple vehicles -
master thesis |
-
Mina Ferizbegović (master student at ETF Sarajevo))
A novel sample based planner (RRT)
and chance constrained MPC - master thesis |
- Bostan Aldin (Systech, BiH)
Cooperative control of multiple vehicles
- master thesis |
-
Faris Janjoš (master student at ETF Sarajevo))
Nonlinear control with a sample based
planner combined with SDRE - master thesis |
- Šalaka Edin (Systech, BiH)
Chance constrained MPC - master
thesis |
-
Zlatan Tucaković (master student at ETF Sarajevo))
Explicit MPC combined with SDRE - master thesis |
|
MPC in solar power plants
teaching assistant |
|
-
Hakija Agić (ENERGOINVEST)
teaching assistant |
|
-
Mladen Milenković (ENERGOINVEST)
teaching assistant |
|
|
|