矩阵变换器的间接模型预测控制策略

Indirect Model Predictive Control Strategy of Matrix Converter

为解决矩阵变换器(MC)的直接模型预测控制(MPC)算法计算量大的问题,基于MC的等效间接调制策略,将MC的预测控制等效为虚拟整流环节和虚拟逆变环节的预测控制。在虚拟整流环节采用预测控制实现电网侧单位功率因数运行;利用预测控制实现虚拟逆变环节对负载电流的控制,并采用无差拍的预测控制方法预选开关状态,从而进一步减少计算量。最后,将虚拟整流环节和虚拟逆变环节控制中得到的最佳开关状态等效合成为MC的开关状态。与传统直接MPC方法相比,间接MPC的算法计算量明显降低,减少了算法执行时间。在此针对基于MC的间接MPC策略,建立完整的理论分析和推导,然后在实验平台上进行验证。实验结果表明,所提基于MC的间接MPC对负载电流和网侧单位功率因数具有良好控制效果。

In order to alleviate the intensive computational burden in the model predictive control of the matrix converter(MC),the model predictive control(MPC)of the MC is divided into the predictive control of the virtual rectifier and virtual inverter based on the equivalent indirect modulation of the MC.Firstly,predictive control is used in the virtual rectifier to achieve a unity power factor.Then,the predictive control is used in the virtual inverter to control the load current and the deadbeat control technique is adopted to pre-select the switch states,thus further reducing the computational burden.Finally,the switch states obtained in the virtual rectifier and inverter are combined to produce the equivalent switch states of the MC.Compared with the traditional direct MPC,the computational burden and execution time of the proposed indirect MPC strategy is reduced.A complete theoretical analysis and derivation are established for the indirect MPC of MC,and then the proposed strategy is verified on an experimental prototype.The experimental results show that the load current and input power factor can be regulated effectively by the proposed indirect MPC.