基于改进长短期记忆神经网络的短期负荷预测

Short-term load forecasting based on improved long short-term memory neural network

电力系统负荷预测是系统规划、设计和运行的有力支撑和重要保障。实际应用中,存在由于数据采集设备故障、系统突发事件导致相关数据资料不准确,使得短期负荷预测的精度不高。文中提出基于小波变换的长短期记忆神经网络WT-LSTM(Wavelet Transform-Long Short-Term Memory)负荷短期负荷方法,利用小波变换的时频特性对负荷数据的伸缩变换进行细化,实现高频系数量化处理;结合长短期记忆神经网络的梯度计算,从而提高负荷预测结果的精度。通过变电站负荷数据以及区域办公楼实验实际负荷进行实验分析,仿真结果表明文中提出的负荷预测方法能够有效处理负荷原始数据中的噪声,针对不同负荷类型能够有效提高负荷预测精度和预测方法的鲁棒性。

Accurate load forecasting is a powerful support and important guarantee for planning,design and operation of power system.In practical application,there are some problems,such as data acquisition equipment failure and system emergency,which lead to inaccurate data and affect the accuracy of short-term load forecasting results.In this paper,short-term load forecasting method based on wavelet transform-long short-term memory(WT-LSTM)is proposed,which utilizes the time-frequency characteristics of wavelet transform to refine the scaling transformation of load data,and the quantification of high-frequency coefficients is realized.The gradient calculation of long short-term memory neural network is combined to improve the accuracy and reliability of load forecasting result.Through the experimental analysis o load data of substation and the actual load of regional office building,the simulation results show that the proposed method can effectively deal with the noise in the original load data,so as to improve the robustness and accuracy of load forecasting results of different load types.