Research on springback control based on GRNN-ES
Springback is a kind of formation defect in punch machining. It can be decreased by adjusting the processing parameters. In this paper it put forward to optimize the parameters with response surface method. Firstly it built an artificial neural network for mapping the relation between the parameters and springback. Then it optimized the neural network model with evolution strategy. At last it would get the optimum of parameters after optimization. In order to get more precise parameters it modified the ANN and ES in this paper. Then it took a trial to certify the method. The results showed that the method proposed in this paper can decrease the springback effectively.
snringbock control ANN ES optimization
Zhaohu Deng Yanqin Zhang
Mechanical and Electrical Engineering Department Guangdong Polytechnic College Guangzhou, China Department of Electromechanical Engineering Dongguan Polytechnic College Dongguan, China
国际会议
太原
英文
139-143
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）