Inference Algorithm Independent SVD Fuzzy Rule Base Complexity Reduction
Péter Baranyi*, Yeung Yamb**, Chi Tin Yang**, Péter Várlakic***, and Pél Michelberger***
*,***Research Group for Mechanics of the Hungarian Academy of Science
*Dept. Telecommunication and Telematics, Budapest University of Technology and Economics Budapest, Sztoczek u.2., Budapest, H-1111, Hungary
**Dept. Mechanical and Automation Engineering, Chinese University of Hong Kong Shatin, N.T. Hong Kong
This paper is motivated by the fact that fuzzy technique is popular engineering tool, however, its use is restricted by its exponential complexity. One of the complexity reduction techniques is the SVD based reduction method which deals with arbitrary shaped fuzzy sets. Despite its advantages it is applicable to rule bases only which are based on product-sum-gravity (PSG) inference algorithm. This paper introduces the extension of the PSG based SVD-reduction method to arbitrary inference based fuzzy algorithms. An example with the control of an automatic guided system is presented to demonstrate the reduction process.