基于BP神经网络的配网系统光伏输出功率平滑控制分析

Analysis of pv output power smoothing control in distribution system based on BP neural network

为充分降低配网系统光伏输出功率神经网络输入变量,基于BP神经网络的配网系统光伏输出功率平滑控制。选择BP神经网络来训练数据,建立基于光伏预测模型及储能配合的光伏功率平滑控制。以MATLAB构建了神经网络预测模型实例验证得到:在12:00.14:00时段内光伏输出功率明显偏大,引起配网节点电压越限的结果,以沃兹低通滤波器对光伏输出功率进行了平滑优化处理后光伏实际输出功率变化得更加平缓,在9:00.15:00时段内光伏输出功率明显偏大,最大时段有明显的增加。可以利用构建的模型完成光伏输出功率的精确预测。

In order to fully reduce the input variables of the photovoltaic output power neural network of the distribution network system, the photovoltaic output power smoothing control of the distribution network system based on BP neural network is proposed.BP neural network is selected to train the data, and a photovoltaic power smoothing control based on photovoltaic prediction model and energy storage coordination is established. With MATLAB neural network prediction model is constructed to verify and the results get: photovoltaic (pv) power output during 12:00-14: 00 obviously slants big, which cause the limit of the distribution network node voltage as a result, in watts low-pass filter on the photovoltaic power output smooth optimization photovoltaic (pv) of the actual output power changes more gently, during 9:00-15:00 photovoltaic (pv) power output is significantly larger, there has been a marked increase maximum period of time.The constructed model can be used to accurately predict the photovoltaic output power.