GTAW焊接质量神经网络模糊控制系统的研究

Research of the Artificial Neural Network Fuzzy Control System for GTAW Welding Quality

研究了一种应用于GTAW焊接质量控制中，利用神经网络进行模糊推理的控制器，采用两个BP网络实现了参数自调整模糊与积分混合控制，采用对输入进行“编码”的方式，减少了BP网络的训练时间和增强了控制实时性，仿真试验和工艺试验都表明了该控制方法的合理性和有效性。

We studied an artificial neural network fuzzy controller for GTAW welding quality, We take two BP network in order to realize the parameter self-tuning fuzzy and integral control. We take the coding method for input to reduce the training time of BP network and speed up the artificial neural network learning and system real-time controlling. The simulation and technological experiment reveals that the control method is reasonable and efficient.