基于自适应模型预测控制的光储虚拟同步机平滑并网策略

Smooth Grid-connected Strategy for Photovoltaic-storage Virtual Synchronous Generator Based on Adaptive Model Predictive Control

基于虚拟同步发电机(virtual synchronous generator,VSG)控制的光伏-储能单元可工作在孤岛及并网两种模式,孤岛转并网运行时,由于传统控制精度差而存在电压电流冲击问题,且因为频率和相角间的相互耦合,在预同步过程中会产生较大频率的波动。为此,提出了一种针对光储虚拟同步发电机单元在上述模式切换过程中的自适应模型预测控制(adaptive-modelpredictivecontrol,A-MPC)方法,在采用更为精确的模型预测控制(model predictive control,MPC)策略的基础上,引入切换过程中频率变化量对MPC的加权系数进行自适应调整,较好地改善了基于VSG控制的换流器在预同步过程和运行模式切换过程中的动态响应。最后,通过搭建控制器级硬件在环仿真实验,验证了所提控制策略可以更好地减小孤岛转并网过程中的电压电流冲击。

The photovoltaic-storage unit based on virtual synchronous generator(VSG)control can operate in islanding and grid-connected modes.When the islanding mode is transferred to the grid-connected mode,there is a voltage and current inrush problem due to the poor traditional control accuracy.For the reason of the mutual coupling between the frequency and the phase angle,a large frequency fluctuation will occur during the pre-synchronization process.Therefore,this paper proposes an adaptive-model predictive control(A-MPC)method for the photovoltaic-storage virtual synchronous generator unit in the above-mentioned mode switching process.Since adopting a more accurate model predictive control(MPC)strategy,the frequency variation during the switching process is introduced to adjust the weighting coefficient of the MPC adaptively.The proposed A-MPC can improve the dynamic response of VSG-controlled converters in the pre-synchronization process and the mode switching process.Finally,a controller-level hardware-in-the-loop simulation experiment platform is built,and the results show that the surge voltage and current during the modes switching process can be limited when adopting the proposed control strategy.