Institute of Information Theory and Automation

You are here

AS

BSc. Topic: Sequential Monte-Carlo Estimation Seen as Learning on Variable Grid (Kárný)

Type of Work: 
bachelor
Affiliation/Phone: 
ÚTIA AV ČR, v.v.i., oddělení AS, 266052274
Supervisor: 
Kárný

Recursive estimation of model parameters is a key part of adaptive systems predicting or influencing their complex random environment. Most models do not allow us to use the desired exact Bayesian estimation. Therefore it is necessary to implement them approximately. Monte Carlo procedures allow this, but their efficiency is not great.

Tasks: 

1. Learn about Bayesian parameter estimation.
2. Familiarize yourself with the concept of recursive Bayesian estimation.
3. Learn about assigning a priori probability to hypotheses.
4. Design an estimation algorithm that at each step: i) generates a new sample of parameters; ii) assigns a priori probability to all samples; iii) correct those probabilities with the Bayes relation; iv) excludes the least suitable parameter sample.
5. Program the result for a simple useful model and compare the quality of your algorithm with a suitable standard.

Bibliography: 

Recommended literature (parts selected after agreement with the supervisor)

1. V. Peterka, Bayesian System Identification, in P. Eykhoff "Trends and Progress in System Identification", Pergamon Press, Oxford, 239-304, 1981.
2. A. Doucet, V.B. Tadic, Parameter estimation in general state-space models using particle methods, Annals of the institute of Statistical Mathematics,55(2),409-422,2003.
3. A. Doucet, M. Johansen, A tutorial on particle filtering and smoothing: 15 years later, In: Handbook of Nonlinear Filtering, Oxford Univ. Press, UK, 2011.
4. M. Kárný, On Assigning Probabilities to New Hypotheses, Pattern Recognition Letters, 150(1), 170-175, 2021.

2022-09-15 10:16

BSc./Mgr. Topic: Numerical methods in the design of control of industrial robots (Belda)

Type of Work: 
bachelor
diploma
Affiliation/Phone: 
UTIA CAS, dept. of AS, 26605 2310
Supervisor: 
Belda
Keywords: 
Model predictive control, numerical integration methods, industrial robots, nonlinear dynamic models

The topic of the bachelor/diploma thesis is focused on the selection and implementation of a suitable numerical method for the predictive control algorithm, which uses a default physical nonlinear model describing the robot's dynamics. Numerical methods should be used in the construction of prediction equations that express the dependence of future planned outputs on unknown calculated inputs (control actions) and also for control using internal simulation in its design. Algorithms will be created in the MATLAB language with a connection to the C language.

Bibliography: 

[1] Rektorys, K. et al.: Survey of Applied Mathematics, Available Edition.
[2] Belda, K.: Nonlinear Model Predictive Control Algorithms for Industrial Articulated Robots. Informatics in Control, Automation and Robotics. Lecture Notes in Electrical Engin., 613. Springer, 2020, pp. 230-251.
[3] Belda, K., Záda, V.: Predictive Control for Offset-Free Motion of Industrial Articulated Robots. Proc. 22nd IEEE Int. Conf. Methods and Models in Automation and Robotics. West Pomeranian University of Technology, Szczecin, Poland, 2017, pp. 688-693.
[4] Other literature according to the specific focus of the thesis.

Note: 
The topic is suitable for FNSPE CTU, by agreement for FEE, FME CTU and other universities.
2022-09-15 10:17

BSc./Mgr. Topic: Mathematical modelling of motion of industrial robots (Belda)

Type of Work: 
bachelor
diploma
Affiliation/Phone: 
UTIA CAS, dept. of AS, 26605 2310
Supervisor: 
Belda
Keywords: 
Parametric models, analytic geometry, time parameterisation, industrial robots, kinematic quantities

The theme of the bachelor/diploma thesis is focused on the motion modelling of industrial articulated robots. Modelling will deal with the investigation of parametric models of both planar and spatial curves containing the so-called geometric parameter. This parameter determining the position on the given curve will be used in the design of a suitable time parameterization of the robot's motion. Its output will be the time dependencies of partial coordinates of a specific coordinate system and their respective derivatives.

Bibliography: 

[1] Rektorys, K. et al.: Survey of Applied Mathematics, Available Edition.
[2] Belda, K.: Smoothing and Time Parametrization of Motion Trajectories for Industrial Machining and Motion Control. Proc. of the 16th Int. Conf. on Informatics in Control, Automation and Robotics. ICINCO 2019. Prague, The Czech Republic 2019, Vol. 2, pp. 229-236.
[3] Záda, V., Belda, K.: Structure Design and Solution of Kinematics of Robot Manipulator for 3D Concrete Printing. IEEE Trans. Automation Science and Engineering. IEEE, New York, 2022, 1-12 pp.
[4] Other literature according to the specific focus of the thesis.

Note: 
The topic is suitable for FNSPE CTU, by agreement for FEE, FME CTU and other universities.
2022-09-15 10:17

Seminar Applications of Modeling and Optimization: Mapping Out the Opportunity Space

The speaker will be Rudolf Kulhavý, seminar will be held on Thursday 28. 7. 2022 at 13:00 in the AS-meeting room 474.

The world of practical applications of modeling and optimization is rich, colorful, and hard to overlook. The seminar will attempt to yield more insight by structuring the space of opportunities around three dimensions of every economic activity, namely Product (design) Process (improvement) Customer (management) A range of real-life applications in multiple industries will be used to illustrate the classification.

2022-09-13 14:43

Seminar Discounted Fully Probabilistic Design of Decision Strategy

The seminar is postponed.

The speaker will be Soňa Molnárová, the seminar will be held xxx in the AS-meeting room 474.

2022-11-25 09:29

Petr Nedoma

20.7.2022 po těžké nemoci odešel pan Petr Nedoma.

Byl to velice přátelský a dobrý člověk, spolehlivý kolega a věrný kamarád.

Vzpomínáme na něj..

2022-07-26 17:45

Ing. Adam Jedlička

Position: 
Ph.D. student
Kontakty
Room: 
Mail: 
Research interests: 
transfer learning, exploration in reinforcement learning, health monitoring and fault detection
Podrobnosti o doktorském studiu
External Supervisor: 
Prof. Ing. Luděk Müller, Ph.D.
Faculty: 
Fakulta aplikovaných věd ZČU
Thema of Study (CS): 
Kybernetika
Beginning of Study: 
2023-04-04
2024-03-06 08:54

Mgr. Mikhajlo Sokolov

This person is no longer active at UTIA.
Position: 
Doktorand
2022-07-11 10:24

Seminar On Assigning of Prior Probability to a New Hypothesis

The speaker will be Miroslav Kárný, the seminar will be held on Wednesday 25.5.2022 at 11:00 in the AS-meeting room 474. 

2022-09-13 14:42

Seminar User‘s feedback in Preference elicitation

The speaker will be Tereza Siváková, the seminar will be held on Monday 16. 5. 2022 at 11 am in the AS-meeting room 474

2022-09-13 14:41

Pages

Subscribe to RSS - AS