中文摘要：

由于五相电机调速系统包含25=32个有效电压矢量,当采用模型预测控制（MPC）方法进行系统控制时,存在滚动优化次数过多、耗时严重等问题,造成电流内环跟踪品质的大幅降低.研究了一种基于片上系统（SOC）的电机模型预测控制系统,在建立五相电机调速系统离散预测模型的基础上,结合SOC并行处理的特点,进行MPC价值函数优化设计.样机实验结果表明SOC应用于五相电机MPC系统的可行性和有效性.对比顺序执行的数字处理系统可知,SOC有效提升了系统执行速度和控制精度,为价值函数中加入更多复杂优化指标提供了可能.

英文摘要：

As five-phase motor system contains 25 =32 effective voltage vectors,when we use model predictive control （MPC） method to control the system,the computational burden for this control scheme is very high and the tracking quality of the current loop greatly reduced.To solve this problem,the five-phase motor model predictive control system was studied.The discrete model of five-phase motor speed control system was established,and MPC value function was optimally designed based on the parallel processing characteristics of SOC.The experimental results show that SOC efficiently improve the execution speed and control precision of the system,and it provide the possibility to add more complex optimization index into the value function.