Parameter optimization in WEDG of engineering ceramics based on the integration of genetic algorithm and fuzzy neural network
                                                                                     WANG He1, YANG Yong2
（1.Mechanical Engineering Department, Henan University of Engineering, Zhengzhou 451191, China;2.Shenyang Machine Tool（Dongwan）Intelligent Equipment Co., Ltd., Dongwan 523808, China）

Abstract: Aiming at the problem that it was difficult for the processors to select proper process parameters in the practical operation of wire electrode discharge grinding (WEDG) of engineering ceramics, the process parameters affecting the processing effect were summarized. A prediction model was established to predict the change of the processing effect of WEDG of engineering ceramics with process parameters by BP fuzzy neural network (BPFNN). The redundancy was eliminated by rough sets theory, that is, to reduce the attributes and rules of the network model. Genetic algorithm(GA) was integrated to optimizes the electrical parameters to obtain the best processing effect. According to numerical simulation and experimental analysis, it was verified that the best processing effect could be obtained by using optimized processing electrical parameters to machine boron carbide, and the simulation results were very close to the experiment results. The results indicate that the GAFNN model can realize the optimization of electrical parameters of WEDG of engineering ceramics, which has a certain reference value for practical production.

Key words: engineering ceramic; wire electrical discharge grinding（WEDG）; genetic algorithm(GA); fuzzy neural network; parameter optimization