具有建模误差补偿的三相逆变器模型预测控制算法

摘要

对于三相逆变器模型预测控制系统,传统的有限控制集模型预测控制(FCS-MPC)算法未考虑建模误差对系统控制的影响,导致所选择的最优控制量在实施控制时已失去最优性.对此设计了一种具有建模误差补偿的三相逆变器有限控制集模型预测控制(FCS-MPCMEC)算法,该算法以前一时刻三相逆变器的建模误差补偿当前时刻三相逆变器预测模型的输出,从而实现对三相逆变器预测模型建模误差的在线补偿.通过对比实验分析了基于FCS-MPC算法及FCS-MPCMEC算法的三相逆变器在空载、带阻感性负载、带非线性负载及负载投入等工况下的控制性能.实验结果表明:在上述所有工况下,FCS-MPCMEC算法对三相逆变器的控制效果均优于传统FCS-MPC算法.%Because the traditional FCS-MPC(Finite Control Set Model Predictive Control) scheme for threephase inverter neglects the influence of modeling errors on system control,the selected optimal control action may not be optimal.An FCS-MPCMEC(Finite Control Set Model Predictive Control with Modeling Error Compensation) scheme is proposed,which realizes the online compensation of predictive model by modifying its output for current moment according to its modeling error of last moment.The control performance of three-phase inverter is compared between FCS-MPC and FCS-MPCMEC schemes for different operating conditions:without load,with resistive-inductive load,with nonlinear load,and load switching-on.The experimental results demonstrate that,the control performance of FCS-MPCMEC scheme is better than that of traditional FCS-MPC scheme under all operating conditions.