LSTM-RF的中长期风电功率组合预测方法

Wind Power Mid-long Term Load Forecasting Based on LSTM RF Combination Forecasting Method

风电功率的准确预测是减少风电并网对电网造成冲击的有效手段之一。利用深度学习算法中的长短期记忆网络(LSTM)对中期风电功率出力进行了预测,综合考虑功率数据、气象数据等多维特征,采用LSTM算法和随机森林(RF)算法搭建预测模型,预测风电场1~7日的风电功率出力。基于某风电场2014年1月到2016年12月的实际发电数据,通过实验对比BP神经网络、支持向量机(SVM)和自回归积分滑动平均模型(ARIMA)等算法可知,提出的预测方法在较为突变的天气状况下仍能保持较高的预测精度,能为风电并网和电网调度提供辅助支撑。

The accurate prediction of wind power is one of the effective means to reduce the impact of wind power grid on the power grid.The article uses the long and short-term memory neural network(LSTM)in the deep learning algorithm to predict the mid-term wind power output.The model comprehensively considers multi-dimensional features such as power data and meteorological data.The article uses the LSTM algorithm and the random forest algorithm to build a prediction model to predict the wind power output of the wind farm in 1~7 days.Based on the actual power data from January 2014 to December 2016,comparing the BP neural network,SVM and ARIMA models,the method proposed in this paper obtains a better result in the special weather condition,so it can provide support for the wind power integration and the network dispatching.