Article ID: | iaor1994243 |
Country: | Germany |
Volume: | 15 |
Start Page Number: | 31 |
End Page Number: | 42 |
Publication Date: | May 1993 |
Journal: | OR Spektrum |
Authors: | Rommelfanger H. |
Keywords: | artificial intelligence: expert systems |
In this paper the basic concepts of fuzzy control are explained and discussed. The relatively simple constructed fuzzy controllers prove to be superior to controllers based on classical mathematics, because they are suitable for systems with time-variant parameters and show a high robustness. Even when one of the control rules is neglected, the fuzzy controlled systems work, however less stable. By means of high-speed fuzzy controller hardware it is easy to control nonlinear and/or timevariant systems too. Furthermore it is analysed, if and how the fuzzy logic algorithm can be transferred to non-technical expert systems. Where unique decisions are required it is essential to be more careful by formulating the control rules and their agregations.