基于递归模糊神经网络的风电平滑控制策略

Wind Power Smoothing Control Strategy Based on Recursive Fuzzy Neural Network

为实现风电平滑并网,在满足风电波动要求的基础上,以减少风电平滑输出时间延时和降低混合储能容量为目的,设计了一个基于递归模糊神经网络与规则控制相结合的风电平滑并网策略。首先,通过递归模糊神经网络对风电原始出力进行平抑,获得符合国家1 min/10 min风电波动要求的并网功率。其次,依据混合储能系统各自的约束条件(如功率、荷电状态等)建立基于规则控制的功率分配算法,实现了功率在混合储能系统之间的合理分配。最后,以新疆地区的风电历史数据为依据,在Matlab/Simulink中进行仿真,结果表明该平滑策略是合理和有效的。

To implement smooth grid-connection of wind power,on the basis of satisfying the demand on the fluctuation of wind power,taking decreasing the time delay of wind power smooth output and reducing the capacity of hybrid energy storage as the objective,a wind power smooth grid-connection strategy based on the recursive fuzzy neural network combined with rule control was designed.Firstly,by means of recursive fuzzy neural network the original output of wind power was leveled to obtain the grid-connected power met the national requirement on wind power fluctuation 1 min/10 min.Secondly,according to respective constraint conditions of hybrid energy storat such as power,state of charge,etc.,a rule control-based power distribution algorithm was established to realize the rational distribution of power among hybrid energy storage systems.Finally,taking historical data of wind power in Xinjiang as basis,simulation calculation was performed by Matlab/Simulink.Simulation results show that the proposed smooth strategy is reasonable and efficient.