基于模糊神经网络PID的开关磁阻电机控制系统研究

Research on switched reluctance motor control system based on fuzzy neural network PID

针对开关磁阻电机存在的转矩脉动大、噪声大、速度不稳定等问题,对开关磁阻电机的启动、运行、调速等方面进行了研究,提出了一种基于模糊神经网络PID的控制方法,将模糊控制理论与BP神经网络相结合,构成了模糊BP神经网络,根据系统误差,误差的变化,以及误差变化的变化实时调整PID控制参数,使电机在整个转速范围内获得了最优的PID参数。实验采用DSP作为控制核心,不对称逆变桥作为功率变换器,驱动一台2 k W的开关磁阻电机运行。研究结果表明,该方法大大改善了开关磁阻电机控制系统的动、静态性能,控制精度高、转矩脉动小,对干扰有较高的鲁棒性。

Aiming at the problem of torque ripple,large noise and speed instability of switched reluctance motor,the starting,running and adjusting speed of switched reluctance motor were studied,and a new approach was proposed based on fuzzy neural network PID control.Fuzzy control theory and BP neural network was combined to form a fuzzy BP neural network,which could adjust the parameters of PID according to the system error E,the change of the error of EC,and the change of the change of the error ECC. DSP was used as a control core and the asymmetric inverter bridge was used as a power converter,which could drive a 2 k W switched reluctance motor. The results indicate that this method can greatly improve the dynamic and static performance of switched reluctance motor control system. It has high control precision,small torque ripple,and high robustness to interference.