摘要：针对永磁直线同步电机的跟踪性能易受推力波动等干扰影响的问题，以及BP神经网络收敛速度慢和易于陷入极小值的问题，提出了基于遗传优化神经网络的控制方法。该算法在复合前馈PID控制算法的基础上，将遗传算法全局寻优和BP神经网络局部寻优相结合，利用神经网络实现了对永磁直线同步电机的干扰的快速，准确的在线补偿。实验结果表明，与复合前馈PID控制方法和神经网络控制方法相比，基于遗传优化神经网络的控制方法有效的提高了系统的跟踪性和鲁棒性，并能有效的消除干扰对系统的影响。

关键词：永磁直线同步电机；复合前馈PID；遗传优化神经网络；干扰抑制
Abstract: For the tracking performance of permanent magnet linear synchronous motor influenced by force ripple and other disturbances, and the BP neural network converges slowly and easily gets in the local minimum, a method of control based on neutral network optimized by genetic algorithm was proposed This scheme combined the general optimization of the genetic algorithm together with the local optimization of BP neural network on the basis of feed-forward plus PID control A neutral network was used to estimate the disturbances The experimental results show the efficiency of the proposed method, compared with combined feed-forward plus PID control and neural network The proposed method can improve tracking precision and robustness, and reduce the influence of disturbances

Key words: permanent magnet linear synchronous motor; feed-forward plus PID control; neutral network optimized by genetic algorithm; reduce disturbances