电力系统及其自动化学报2023，Vol.35Issue(12)：40-49,10.DOI:10.19635/j.cnki.csu-epsa.001217
基于TASSA-Mg LSTM的配电网线损预测方法
Prediction Method for Distribution Network Line Loss Based on TASSA-Mg LSTM
摘要
Abstract
To predict the line loss of distribution network more accurately,a method based on transient adaptive spar-row search algorithm(TASSA)to optimize Mogrifier long short-term memory(Mg LSTM)is proposed.First,the corre-lation degrees between 12 electrical characteristic parameters and line loss are obtained through the slope gray correla-tion analysis method,and the optimal electrical characteristic parameter system is determined through the prediction and verification of distribution network data.Then,TASSA is used to optimize the four important parameters of the Mg LSTM neural network and determine the best Mg LSTM network structure.On this basis,a neural network line loss pre-diction model based on TASSA-Mg LSTM is built.Finally,a case study of one distribution network in Gansu Province shows that the proposed method has a high prediction accuracy.关键词
配电网/线损预测/斜率灰色相关性分析/麻雀搜索算法/长短期神经网络/深度学习Key words
distribution network/line loss prediction/slope grey correlation analysis/sparrow search algorithm(SSA)/long short-term neural network/deep learning分类
信息技术与安全科学引用本文复制引用
吴丽珍,秦文彬,赵一凡,陈伟..基于TASSA-Mg LSTM的配电网线损预测方法[J].电力系统及其自动化学报,2023,35(12):40-49,10.基金项目
国家自然科学基金资助项目(62063016) （62063016）
