Heuristic constraints enforcement for training of and rule extraction froma fuzzy/neural architecture - Part II: Implementation and application

This paper is the second of two companion papers. The foundations of the pr oposed method of heuristic constraint enforcement on membership functions f or knowledge extraction from a fuzzy/neural architecture was given in Part I. Part II develops methods for forming constraint sets using the constrain ts and techniques for finding acceptable solutions that conform to all avai lable a priori information. Moreover, methods of integration of enforcement methods into the training of the fuzzy-neural architecture are discussed. The proposed technique is illustrated on a fuzzy-AND classification problem and a motor fault detection problem. The results indicate that heuristic c onstraint enforcement on membership functions leads to extraction of heuris tically acceptable membership functions in the input and output spaces. Alt hough the method is described on a specific fuzzy/neural architecture, it i s applicable to any realization of a fuzzy inference system, including adap tive and/or static fuzzy inference systems.