Fuzzy system optimization based on heuristic search algorithm

ABC algorithm is an up-to-date heuristic optimization method based on principles of honey bees behaviors in finding/gathering nectar from food sources in the nature. In this study, performance of the ABC algorithm is investigated on the problem of nonlinear dynamic systems modeling/ identification using ANFIS (Adaptive Neuro-Fuzzy Inference Systems) fuzzy inference model. Fort this investigation, five benchmark problems selected from the literature have been used.Each benchmark problem is identified /modeled with ANFIS which of parameters tuned by ABC. Obtained results for each benchmark have been comparatively examined with the results of PSO and DE popular and commonly used algorithms. According to the obtained results in this study, it is determined that performance of ABC algorithm is very close to performance of DE algorithm. However, it is concluded that performance of ABC in terms of run time per generation was two times better according to that of DE. Based on this conclusion, it is evaluated that operating of ABC algorithm in the application of real time embedded systems would be more efficient.Key words: Artificial Bee Colony, ANFIS, Fuzzy Modeling, Dynamic System.