论文部分内容阅读
对脉冲ＴＩＧ焊工艺中熄接电流一正面熔宽动态过程提出了一种自学习模糊神经网络控制方法，模糊控制器采用误差、误差变化及误差加速度的加权和的解析描述形式，利用人工神经网络直接对过程的建模，实现对模糊加权因子的自学习优化调整。将上述方法用于控制仿真和实际脉冲ＴＩＧ焊过程控制试验，结果表明提出的自学习模糊神经网络控制方案有效。
A self-learning fuzzy neural network control method is proposed for the dynamic process of pulse quenching current-positive melting in pulse TIG welding. The fuzzy controller adopts the analytic description form of error sum, error variation and weighted sum of error acceleration. By using artificial neural network Directly model the process and realize self-learning optimal adjustment of fuzzy weighting factors. The above method is used to control the simulation of the actual pulse TIG welding process control test. The results show that the proposed self-learning fuzzy neural network control scheme is effective.