Generating And Optimizing Of Fuzzy Rule Bases

This paper research is how to improve the method of generating andoptimizing fuzzy rule bases based on FCM. By collecting information of aboutit, we improved the FCM and optimized the fuzzy rule bases by optimize thefuzzy reason system based on GA.Firstly this paper improved the FCM method. To the weakness, such as, toinitialize a sensitive, slower speed of training rules method there are twoaspects improvement: first, proposed a new clustering identify guidelines toimprove clustering effect. Second, through the combination of HCM and FCM,with HCM to find clustering center, and then use FCM to Increase clusteringspeed. The paper proposed, and generated a new FCM to the generate initialfuzzy rule bases. We adjust parameters by RBF fuzzy neuro network togenerate complete the fuzzy rule bases.Secondly this paper proposed a new optimizing method by using GA basedon Homaifar optimization method, and provide the corresponding optimizationalgorithms. This method in the following ways to improve the defect ofHomaifar method: genetic codes, selection function, adaptation function, therealization algorithms. The new method optimize fuzzy rule bases byoptimizing fuzzy neuro network model that got by optimizing systemparameters and the structure and formulate reasonable rules and codes.Finally by simulation to representative and universal example in Matlab,and by comparison to results, the effectiveness and feasibility of methodsreferred to can proved...