ISS0023 Intelligent Control Systems

Course description

News
Syllabus

Instructors: 

Semester: 

  • Autumn

Year: 

2017

Awarded ECTS points: 

5.0

Description: 

The course gives an overview of :

  • complex systems modeling and control methods and their applications in design of reliable control systems. 
  • artificial intelligence methods (artificial neural networks, fuzzy systems, genetic algorithms) based systems identification and control techniques and their applications in development of intelligent control systems. 
  • artificial intelligence methods based classification and recognition techniques and their applications.
 
Topics include:
  • Nonlinear systems, Principes of nonlinear systems identification and control; 
  • Adaptive control systems; 
  • Artificial neural networks. Structures of artificial neural networks and training algorithms; 
  • Artificial neural networks based identification of nonlinear systems; 
  • Artificial neural networks based control of nonlinear systems; 
  • Self-learning neural networks; 
  • Artificial neural networks based image recognition and pattern classification;
  • Fractional Order Modelling and Control; 
  • Dynamic feedback linearization based control of nonlinear systems; 
  • Genetic algorithms and their applications for identification and control of nonlinear systems.

Policies: 

Lab reports:

6 labs = 6 reports
Each report gives up to 1 point.
Each report has to be presented during 2 weeks after the lab!
Later presented reports (before December 18) – multiplied by coefficient 0.8
After December 18 – coefficient 0.6
5 best report will give up to 5 points.
 
Exam prerequisites:
Course ISS0023 is declared (included into Your semester plan),
Laboratory trainings are performed,
Reports are presented and accepted (written report on each laboratory training).
 
Estimation criterion:
Grades are based on the report of the final small practical project (design of a control system).
 
Exam  - up to 72 hours
Small practical project – design of a control system according to given control criteria;
Simulation of the control system;
Analysis of results and writing a report.
2 tasks – each one gives maximum 5 points.
 
Average of 2 exam tasks and labs = YOUR COURSE GRADE
 
Grades:
„5“ - student's knowledge is excellent, his/her answers are clear and complete, accurate in details,
deliberative and individual. Independent thinking. Student is capable of analyzing the problem and possible solutions, proposing his own solution and prooving its efficiency.
„4“ - student's knowledge is very good, his/her answers are clear and complete, accurate in details,
but less individual views. Student is capable of analyzing the problem, finding a suitable solution proving its efficiency.
„3“ - student's knowledge is good, his/her answers are clear, but there are a few errors in the
discussion and lack of personality and individual point of view. Student is capable of analyzing the problem and applying standard methods for is solution.
„2“ - student's knowledge is satisfactory, but all the answers are not clear and student makes more
errors (but not errors in basic concepts) and lack of personality and individual point of view.
„1“ - Student responses are weak, there are major mistakes in arguments (not knowing the content,
mistakes in concept), and student needs constant help and guidance in formulating responses to
questions. 
Calendar

November 2017

Mon Tue Wed Thu Fri Sat Sun
30
31
1
2
3
4
5
 
 
 
 
 
 
 
6
7
8
9
10
11
12
 
 
 
 
 
 
 
13
14
15
16
17
18
19
 
 
 
 
 
 
 
20
21
22
23
24
25
26
 
 
 
 
 
 
 
27
28
29
30
1
2
3
 
 
 
 
 
 
 
Materials

Lectures

Titlesort ascending Published Short description Files
NN-structures in control 16.11.2017 Lecture slides on NN-structures, dynamic feedback linearization, NN-ANARX and application of Genetic Algorithms NN_structures2016.pdf Genetic2017.pdf
NN-based identification and control - lecture slides 20.10.2017 Neural Networks based identification and control of nonlinear systems, supevised and unsupervised learning NN2017_part3.pdf
NN supervised learning 20.10.2017 Training of NNs - lecture slides NN2017_part2.pdf
Lecture 1 08.09.2017 05.09.2017 Introductory lecture ISS0023_introduction_2017.pptx ISS0023_introduction_2017.pdf
Introduction to artificial neural networks 21.09.2017 Introduction to neural networks, neural network structure and mathematical model - lecture slides NN2017_part1.ppt NN2017_part1.pdf
Fuzzy Logic based Modeling and Control (S. Astapov) 10.10.2017 Lecture slides on fuzzy logic based modeling and control Sergei_Astapov_Fuzzy_Control_lecture_slides.pdf
Adaptive control 05.09.2017 Slides - part 1 adaptive_control_1_2017.ppt adaptive_control_1_2017.pdf
Adaptive control 05.09.2017 Slides - part 2 adaptive_control_2_2017.ppt adaptive_control_2_2017.pdf

Exercises

Titlesort ascending Published Short description Files
Adaptive control 12.09.2017 Nonlinear Adaptive Control adaptju1.mdl
Adaptive control 12.09.2017 Model Reference Control aw2d.m aw_dsim2.mdl aw_dsim44.mdl

Laboratory works

Titlesort ascending Published Short description Files
Neural Networks based control 25.10.2017 Example of NN-based control system design in MATLAB/Simulink NN_control.zip lab_NN2_2017.ppt lab_NN2_2017.pdf
Introduction to Neural Networks 27.09.2017 NN training data - two inputs, one output + function used to generate the data (you may use this file to validate the result) nn_data.mat answer.m
Image recognition lab 08.11.2017 Example: Character recognition CharacterRecognition.zip
Fuzzy Logic Control 11.10.2017 Files and instructions for the fuzz logic laboratory work Sergei_Astapov_Fuzzy_Control_Lab.pdf fuzzy lab.zip

Literature

Titlesort ascending Published Short description Files
Fuzzy control 05.09.2017 Book by K. Passino FCbook.pdf
E. Rüstern Adaptiivjuhtimine (In Estonian, Eesti keeles) 05.09.2017 Prof. Ennu Rüsterni materjal - ülevaade adaptiivjuhtimidest (Eesti keeles) ISS0022_ülevaade_adaptiivsüsteemidest.pdf ISS0022_ülevaade_adaptiivsüsteemidest_2.pdf
Assignments
Titlesort ascending Published Short description Files
Home work 4 (pattern recognition) 08.11.2017 Design supervised and unsupervised NN based systems for recognition of numbers given in file all_numbers.m. Proof experimentally that the system correctly recognizes all numbers with 18% of random noise. REPORT DEADLINE: November 24, 2017 all_numbers.m
Home work 3 (NN based control) 25.10.2017 Design Neural Network based nonadaptive and adaptive control systems for the black box nonlinear system given in the file. See HW3.pdf for a detailed task. DEADLINE: November 10, 2017 nonlinear_system.slx HW3.pdf
Home work 2 (Fuzzy Systems) 20.10.2017 Solve all tasks from the attached file. Explain solutions, analyze results. Present a report. DEADLINE - October 27, 2017 Sergei_Astapov_Fuzzy_Control_Lab.pdf
Home work 1 (NN training) 20.10.2017 Solve all the tasks from the attached pdf file. DEADLINE - October 27, 2017 HW1_2017.pdf homework1.mat