基于CARIMA模型的MGPC算法在异步电动机中的应用

摘要

提出一种对角受控自回归积分滑动平均模型（CARIMA）的多变量广义预测控制（MGPC）算法,该算法主要把CARIMA模型中的A与C矩阵构造成对角多项式的形式,仿真结果表明,该控制结构和方法保持了MGPC方法的优秀控制性能,实现了对转矩变化的有效控制,使低速变化时具有很好的平缓控制效果,提高了系统的稳态性能。%This paper proposed a kind of multivariable generalized predicative control （MGPC） algorithm based on controlled angle return integral moving average （CARIMA） model. The algorithm was mainly to construct matrix A and matrix C of CARIMA into diagonal form. The simulation result shows that the control structure remains excellent control performance of MGPC method, realizing effective control for torque changes and making very good smooth control effect at low speed, which raises stability performance of the system.