基于混合逻辑动态模型的三相逆变电路有限控制集模型预测控制策略

FCS-MPC Strategy for Inverters Based on MLD Model

逆变电路传统开关函数模型只能描述电路的控制变迁而忽略了电路的条件变迁,为此,建立了三相逆变电路混合逻辑动态(mixed logical dynamical,MLD)模型。在此基础上,将其作为预测模型,提出了电路的有限控制集模型预测控制(finite control set model predictive control,FCS-MPC)策略。FCS-MPC充分考虑了电路的离散特性,选择有限控制集中使目标函数值最小的开关状态作为电路开关管的控制信号,从而控制电路的输出电压,无需任何调制器,可简化MPC的优化问题。此外,基于全维状态观测器设计了电路负载电流观测器,增强了控制器的鲁棒性。仿真和实验结果验证了所提方法的有效性。

Since the conventional switch function model of inverters can only describe the controlled transition, while ignore the conditional transition of inverters, this study builds a mixed logical dynamical （MLD） model for inverters and treats it as a predictive model to work out a finite control set model predictive control （FCS-MPC） strategy for inverters. The FCS-MPC strategy takes full account of the discreteness of inverters, and selects the switch state that minimizes the objective function as the control signal for inverters to control the output voltage, which simplifies the optimization of MPC without any modulators. In addition, a load current observer is designed based on the full state observer, which enhances the robust of the controller. The simulation and test results validate the proposed method.