基于EMD-LSTM的光伏发电预测模型

随着能源消费结构的改变,可再生能源发电的消纳比例逐渐上升。文中以光伏发电功率为研究对象,分析了不同天气状态下的发电功率曲线特性及不同气象因素与光伏发电出力的相关性,进而提出了一种经验模态分解-长短期记忆神经网络(EMD-LSTM)方法融合的光伏发电功率预测模型。首先对预处理后的光伏发电功率历史序列进行重构,并对重构后的出力序列进行EMD分解,针对分解得到的各子序列分别建立长短期记忆神经网络模型,最后将各子序列预测模型得到的结果叠加得到光伏发电功率预测值。采用国内某地区光伏发电的实际出力数据对模型进行了检验,与滑动平均自回归模型(ARIMA)、支持向量机模型(SVM)、LSTM等预测模型相比,文中所提出的模型预测误差小,能有效提高光伏发电功率的预测精度。

Photovoltaic power generation prediction model based on EMD-LSTM

With the change of energy consumption, the consumption proportion of renewable energy power generation rise gradually. This paper chose photovoltaic power as the research object, studied the power curve characteristics un-der different weather conditions, analyzed the correlation of the efforts of various meteorological factors and pho-tovoltaic power generation and proposed a power forecasting model suitable for photovoltaic power generation which provides advice for operation power grid. Firstly, the historical sequence of photovoltaic power generation reconstructed by removing the points whose output value is 0. Then, the reconstructed historical time series is decomposed in EMD model, and the decomposed sub-sequences are respectively predicted in the LSTM network. Finally, the results of each sub-sequence are superimposed to obtain the predicted result of photovoltaic power generation. The actual output data of photovoltaic power generation in a certain region in China were used to test the model. Compared with the SVM model, the prediction error of the model proposed in this paper was reduced by 14.3%, and the results showed that the model could effectively improve the prediction accuracy of photovoltaic power generation