基于遗传算法的模糊控制送丝系统

Fuzzy Logic Control Welding-wire Feeder Based on Genetic Algorithm

高质量的焊接在很大程度上决定于送丝机构的性能 ,模糊控制相对于传统的 PID控制 ,具有更高的系统鲁棒性和稳定性 .而模糊控制的三个因子对控制器的性能有很大的影响 ,常规的模糊控制都是采用人工的多次现场调试来确定 ,不仅费时 ,而且也不能得到最优的控制性能 .采用遗传算法来寻优整定模糊控制器的参数 ,以送丝系统的跟踪误差为目标函数 ,采用二值编码的遗传寻优整定 ,不仅得到了最优性能的模糊控制送丝系统 ,而且使其设计过程大大简化 .最后通过仿真比较 。

Welding quality depends on the performance of welding wire feeder on a great extent. Fuzzy logic control(FLC) has better robustness and stability than conventional PID control. The parameters of FLC have great effect on the performance of the controller. In general, they are determined by lots of trial on site, which wastes time and is difficult to obtain optimal parameters. In this paper, genetic algorithn (GA) is used in the determination of the parameters of FLC. And the method was applied in the control of a welding wire feeder. The tracking error of the feeder was as the objective function, and the binary GA was used. Because GA is a global and high effecttive optimization method, the FLC welding wire feeder with GA has the optimal performance, and its designing process is greatly simplified. The simulation results show that the performance of FLC feeder with GA is better than that of conventional FLC feeder.on sit