模型预测控制在光伏并网逆变器中的应用

Photovoltaic Grid Inverter Using a Model Predictive Control Method

针对光伏并网逆变器中电流环采用模型预测控制计算量大导致的延时问题,以三相光伏并网逆变器的离散化数学模型为预测模型,在dq坐标系下,对并网电流进行两步预测,当前时刻选定的电压矢量为上一时刻的第二步预测值,根据代价函数进行最优化求解,具有计算量小的特点。仿真结果表明,采用改进的模型预测控制与传统模型预测控制相比,能有效提高光伏系统电流环的动态性能,并显著减小并网电流谐波。

This study aims at effectively solving the time delay problem caused by large amount of calculation of current loop in photovoltaic(PV)grid-connected inverter by adopting the model predictive control.A prediction model was constructed according to the discrete of three-phase PV grid inverter and using two-step-prediction method for grid under the dq coordinates system.In order to compensate the delay,sampling current state and the new switch state are using a moment of last step.The simulation result shows that the improved predictive model can effectively improve the dynamic performance of PV system and reduce the grid current harmonic compared with the traditional one.