Neural networks in control?

Abstract

In recent years, neural network research bas rapidly gained interest.Application research, in which the application possibility of neural networks in diverse disciplines is studied, is the largest branch of neural network research.Industrial systems control is one of the disciplines in which neural networks possibly can be applied.So far, the studies presented in literature focus on specific control problems like robot controL Until now, more genetic approaches have received little attention.The roots of my personal interest in neuraJ networks were laid by my education as a cognitive psychoiogist I supplemented this education, in which intelligence and artificial intelligence played a major role, with courses on the biological background of intelligence.The courage of Koos Rooda, to appoint a psychoiogist as an A.I.O. at the Eindhoven University of Technology, allowed me to start this research.The combination of the industriaJ systems control research performed at the group I started in, and my personaJ interest in neural networks served as a starting point of this research in the application of neural networks in industrial systems controL A genetic approach to the application of neural networks in control, taking into consideration control ranging from management level to machine level.Next, I would like to thank all people that have contributed to this work in one way or another.My colleagues for their discussions and their ability to make a psychoiogist feel at home at the Department of Mechanica!Engineering.I was lucky to have Joep Vaes as my roonunate for four years.I want to thank him for the discussions, the fun, the invaluable programming support, and the conunents on drafis of this thesis.His rapid eyes spot every error.