Artificial neural networks and fuzzy logic have been the two main parallel methodological developments relevant to Intelligent Control. Artificial neural networks were original developed to emulate the human brain’s neuronal-synaptic mechanisms that store, learn and retrieve information on a purely experimental basis, whereas fuzzy logic was developed to emulate human reasoning, using linguistic expressions. Neuro-controllers can automatically learn by interacting with their environments with little a priori knowledge so as to cope with ill-defined, complex dynamics, be robust to minor faults and disturbances, and be able to deal with nonlinear relationships over wide operating envelopes.

一.內容
1. Introduction to Neuro-Control Systems
2. Neural Networks and Learning Algorithms
2. The Stone-Weierstrass Theorem and its Application to Neural Networks
3. Identification and Control of Dynamical Systems Using Neural Networks
4. Associative Memory Networks
5. Instantaneous Learning Algorithms
6. The CMAC Algorithm
7. The modeling Capabilities of The Binary CMAC
8. Adaptive B-spline Networks


二.教科書
1. Neurofuzzy Adaptive Modelling and Control: Martin Brown and Chris Harris, Prentice Hall (民全)
2. Neuro-Control Systems: theory and applications, edited by Madan M.Gupta and Dandina H. Rao IEEE Press,1994

三.成績評量方式
期末口頭報告50% + 期末書面報告50%

四.預修課程
控制系統