DTU
Uddannelse
Previous page | Current version Archive 2001/2002 
 
31360 Fuzzy, Neural, and Adaptive Control
Danish title: Fuzzy, neural og adaptiv regulering
Language:  English    ECTS-creditpoints:  10, External examination.   
Type:  , course at phd level, open university
Program:  Electrical Engineering
Class schedule:   E2
Exam schedule:   E2-A (dec 12 2001), F2-A (maj 29 2002)
Both days of E2 are used for the oral examination
Recommended semester:  4th -7th semester
Scope and form:  Survey lectures, team work, programming exercises, laboratory work, e-learning
Evaluation:  Oral presentation and approval of coursework
Examination:  13-scale
Prerequisites:  36230 / 50300 / 72131 / 72231 / 31300 Knowledge of control theory corresponding to an introductory control course.
Participant limitation:  Max. 30
Aim:  General objectives:
- to teach the basics
- to show applications
- to teach controller design
- to teach differences and similarities of fuzzy, neural, and adaptive control
Contents:  The course concerns nonlinear systems, operator's rules, and self-learning controllers.
Fuzzy control appears in consumer products such as washing machines, video cameras, and cars, as well as in industry for controlling for instance cement kilns, underground trains, and robots. Fuzzy control is 'control with rules', and it is based on fuzzy logic which roughly speaking is 'computing with words'. There is a connection to artificial neural networks, and the course explains what neuro-fuzzy data processing is, and it shows some applications.
Objects with partly unknown dynamics may possibly be controlled by adaptive controllers, and the course describes adaptive model following control systems and parametric model identification. The course goes through part of the theory for stability in nonlinear systems. The course demonstrates practical implications of nonlinearities, disturbances, and the choice of test signals. Differences and similarities between various adaptive configurations will be shown. A few topics will be taught through journal articles.
In the last part of the course the participants will test algorithms on an advanced, practically relevant problem in the laboratory (Matlab, Simulink, lab equipment interface).
For further details, see the home page of the course.
Remarks:  The course topics are related to courses 50320 and 04342, but there is only little overlap. Those courses are recommended before or after this course if you wish further specialisation.
Contact:  Jan Jantzen, building 326, (+45) 4525 3561, jj@oersted.dtu.dk
Ole Ravn, building 326, (+45) 4525 3560, or@oersted.dtu.dk
Department: 031 Ørsted DTU
Course URL:  http://www.oersted.dtu.dk/31360
Keywords:  control, regulation, computer, Matlab, E-learning
Signup:  Via CampusNet/Campusnet
Updated:  04-09-2001