基于Q学习的光伏储能微网VSG与APF协调优化研究

Coordinated control strategy of photovoltaic-energy storage micro-grid with VSG and APF based on Q-learning algorithm

针对虚拟同步发电机(Virtual Synchronous Generator,VSG)的功频调节性能及VSG与合并有源滤波(Active Power Filter,APF)的光伏电源无功配比协调控制问题,文章提出了基于Q学习的光伏储能微网VSG与APF协调优化方法。在虚拟环境下,通过对VSG中虚拟惯性、阻尼以及微网中VSG无功分配比例进行训练,得到更优的控制性能。在MATLAB/Simulink环境中,仿真验证了所提出的基于Q学习算法改善控制性能的有效性与可行性。

Aiming at the coordinated control of active power-frequency regulation performance of VSG and reactive power distribution between the VSG and the photovoltaic power with active power filter(APF),in this paper,the coordinated optimization method of photovoltaic-energy storage micro-grid with virtual synchronous generator and active power filter based on Q-learning algorithm is proposed.The virtual inertia,damping of the VSG,and the reactive power distribution ratio of VSG are trained in the virtual environment.Then,the more optimal control performance is obtained.At last,the tradeoff solution is obtained to satisfy the three objectives.The results are simulated in the MATLAB/Simulink,the effectiveness and feasibility of the proposed method after the training based on the Q-learning algorithm is verified.