永磁同步电机伺服系统的RBF神经网络PID控制

摘要

永磁同步电机(PMSM)交流伺服系统中含有未知复杂干扰,属于一类复杂的非线性、不确知系统,会影响PID控制算法的控制质量,无法满足伺服系统的高精度指标要求,利用神经网络对未知信息数据的自学习和自适应能力,设计出一种基于RBF神经网络与传统PID控制相结合的智能PID控制器.仿真结果表明,应用了RBF神经网络PID控制的伺服系统,不但结构简单,而且能适应环境变化,干扰被有效抑制,控制精度明显提高,有较强的鲁棒性,达到了理想的控制效果.%Permanent magnet synchronous motor (PMSM) AC servo system contains unknown complicated interference,which belongs to a complicated nonlinear,unknown system and will have undesired effect on the control quality of PID control algorithm.This will fail to meet the high accuracy demand of servo system.Thus,by using self learning and adaptive ability of neural network to the unknown information data,we design a PID intelligent controller,combining traditional PID with RBF neural network.Simulation shows that the application of the RBF network control servo system not only is simple in structure,but also can adapt to the environment changes.The interference is effectively controlled; control accuracy is obviously improved; and it has strong robustness to achieve the ideal control effect.