优化开关序列的PWM整流器模型预测控制策略

Model Predictive Control Strategy for PWM Rectifier with Optimized Switching Sequence

有限集模型预测控制(FCS-MPC)策略通过在一个采样周期内对有限开关状态遍历寻优,得到当前周期的最优开关状态作用于系统。但由于开关频率不固定,造成系统谐波频谱分散,增加了滤波器设计难度。该文以三相电压源型PWM整流器为研究对象,在分析模型预测控制原理的基础上,提出一种基于多矢量合成的优化开关序列模型预测控制(OSS-MPC)策略,在保证系统控制精度的同时,针对有限集模型预测控制策略系统计算量大、开关频率不固定的问题提出改进,通过对有限控制集优化,减小系统计算量;以多矢量合成为基础,通过开关序列重新排列解决开关频率不固定造成的网侧电流频谱分布不均问题。为验证所提出方法的可行性,进行仿真实验并搭建测试样机,在与有限集模型预测控制策略对比的基础上,增加与现有改进策略空间矢量调制模型预测控制(SVM-MPC)的对比实验,进一步验证该文所提策略的有效性与优越性。

FCS-MPC predicts the optimal switching state by traversing the limited switching state at the current cycle.However,because of the irregular switching frequency,the harmonic spectrum of the system is dispersed,which further increases the difficulty of filter design.In this paper,taking the three-phase PWM voltage source rectifier as the research object,by analyzing the principle of model predictive control,an optimized switch sequence model predictive control(OSS-MPC)strategy based on multi-vector synthesis is proposed.It solves the problems of FCS-MPC,such as large system calculation and unstable switching frequency,while ensuring the control precision of the system.By optimizing the finite control set,the system calculation is reduced.Through the rearrangement of switch sequence,the uneven distribution of network side current spectrum caused by unstable switching frequency is solved.A simulation experiment is carried out and a test prototype is built.Compared with the finite set model predictive control strategy and SVM-MPC strategy,the superiority and effectiveness of the proposed method have been verified.