基于神经网络和模糊控制的电解加工间隙控制

Interelectrode Gap Control in Electrochemical Machining Based on Neural Network and Fuzzy Control

利用阴极面上承受的来自电解液的压力来反映电解加工间隙的大小.用BP神经网络实现间隙向力信号的非线性映射,把间隙的误差转化为力的误差及误差的变化信号,以此作为模糊控制器的输入;以加工电压的增量作为模糊控制器的输出,实现对间隙的控制.根据间隙控制的理论公式,建立了仿真模型,在Matlab的Simulink模块中通过对神经网络、模糊控制器和电解加工系统联合组成的智能控制系统的仿真试验来整定模糊控制器的3个增益参数.结果表明,所提出的控制方法对间隙的控制具有很好的效果,无稳态误差,且速度快、鲁棒性好.

In this paper, the force resulting from the electrolyte and loaded on the cathode surface is used to express the size of the interelectrode gap in ECM （ Electrochemical Machining）. The nonlinearity mapping between the force and the interelectrode gap is first implemented by a BP neural network, and a fuzzy controller is then designed to control the gap, with the voltage increment as the output and the force error transferred from the error of the interelectrode gap and the change of force error as the inputs. Moreover, a simulation model is set up based on the theoretical expression of the interelectrode gap. An intelligent control system consisting of the BP neural network, the fuzzy controller and the ECM system is finally tested based on the Simulink module of Matlab to adjust the three increment parameters of the fuzzy controller. Simulated results show that the interelectrode gap can be well controlled by the proposed method, with high speed, good robustness and with no stable error.