基于ANFIS-PI的同步磁阻电机矢量控制

Vector Control of Synchronous Reluctance Motor Based on ANFIS-PI

针对同步磁阻电机在正常运行过程中磁路饱和导致的电感变化问题,提出一种采用自适应神经模糊推理系统(ANFIS)的PI参数整定方法,以提高电机控制的响应速度。先采用电流调节系统对样机有限元模型进行仿真,得到电流电感数据并拟合电感表达式,进而计算最大转矩电流比的电流角,然后搭建了基于ANFIS-PI的同步磁阻电机矢量控制模型,将经过神经网络训练后的控制模型与有限元软件进行联合仿真。结果表明,与常规PI控制相比,采用ANFIS-PI法可更准确地调整PI参数,进而提高电机动态响应速度。

Aiming at the problem of inductance change caused by magnetic circuit saturation during the normal operation of synchronous reluctance motors,this paper proposes a PI parameter tuning method using adaptive-network-based fuzzy inference system(ANFIS)to improve the response speed of motor control.First,the prototype finite element model was simulated using a current regulation system to obtain the current and inductance data and fit the inductance expression.And then the current angle of the maximum torque per ampere was calculated.Finally,a vector control model of synchronous reluctance motor was built based on ANFIS-PI,and the control model trained by neural network was jointly simulated with finite element software.Compared with the traditional PI control method,the results show that the ANFIS-PI method can more accurately adjust the PI parameters,thereby improving the dynamic response speed of the motor.