基于长短期记忆网络和LightGBM组合模型的短期负荷预测

Short-term Load Prediction Based on Combined Model of Long Short-term Memory Network and Light Gradient Boosting Machine

短期负荷预测是电网安全调度与平稳运行的基础,为进一步提升负荷预测的精度,提出了基于长短期记忆(LSTM)网络和轻梯度提升机(LightGBM)的组合预测模型。首先,根据LSTM网络和LightGBM模型的输入结构,将预处理后的负荷数据、温度数据、日期数据以及节假日信息分别输入2个模型中,通过训练得出各自的预测结果。然后,采用最优加权组合法确定权重系数,并得出组合模型的预测值。最后,采用实际负荷数据进行算例分析,结果表明所提方法能够有效结合2种模型的优点,在保留对时序数据整体感知的同时兼顾非连续特征的有效信息,与其他模型相比具有更高的预测精度。

Short-term load prediction is the basis for safe dispatch and smooth operation of the power grid.To further improve the accuracy of load prediction,a combined prediction model based on long short-term memory(LSTM)network and light gradient boosting machine(LightGBM)is proposed.Firstly,according to the input structure of the LSTM network and LightGBM model,the pre-processed load data,temperature data,date data and holiday information are input into the two models,and the respective prediction results are obtained after training.Then,the optimal weighted combination method is used to determine the weight coefficients,and the prediction value of the combined model is obtained.Finally,taking actual load data as examples for analysis,the results show that the proposed method can effectively combine the advantages of the two models.It takes the effective information of discontinuous features into account while preserving the overall perception of time-series data.Compared with other models,the proposed method has higher prediction accuracy.