基于改进虚拟同步机的风储微网综合控制

微电网中负荷变化与风电等分布式电源出力不确定性给整个微网稳定带来很大困难。针对风电出力波动问题,采用虚拟同步发电机控制结合深度强化学习对电池储能系统输出进行控制：首先搭建包含风电、电池储能、负荷、外部电网的微网模型,其次利用深度确定策略梯度算法对虚拟调速器进行设计,结合奖励函数通过反复学习训练生成调速器实现对虚拟同步发电机的改进。最后,在Matlab/Simulink软件中搭建对应的仿真模型,与下垂控制、传统虚拟同步发电机控制进行对比,仿真验证了并网到离网切换场景与孤岛运行场景下,所提出的控制方法对微网频率与电压有良好的稳定效果,可以实现对负荷有功功率与无功功率的快速追踪。

Research on Integrated Control of Wind Storage Microgrid Based on Improved Virtual Synchronous Machine

Load changes in the microgrid and the uncertainty of the output of distributed power sources such as wind power have brought great difficulties to the stability of the entire microgrid. Aiming at the problem of wind power output fluctuations, the use of virtual synchronous generator control combined with deep reinforcement learning to control the output of the battery energy storage system: firstly build a microgrid model including wind power, battery energy storage, load, and external grid, and secondly use the depth to determine the strategy gradient algorithm The virtual governor is designed, combined with the reward function to generate the governor through repeated learning and training to realize the improvement of the virtual synchronous generator. Finally, the corresponding simulation model was built in Matlab/Simulink software to compare with droop control and traditional virtual synchronous generator control. The network frequency and voltage have a good stabilizing effect, which can realize fast tracking of load active power and reactive power.