基于经验模态分解与偏差校正的配电网短期负荷预测方法  

Short term load forecast method of distribution network based on empirical mode decomposition and deviation correction

作　　者：肖明伟 舒晓欣 汪涵 陈彦斌 刘于良 Xiao Mingwei;Shu Xiaoxin;Wang Han;Chen Yanbin;Liu Yuliang(State Grid Anhui Electric Power Co.Ltd.Wuhu Power Supply Company.Wuhu,241027 China)

机构地区：[1]国网安徽电力有限公司芜湖供电公司,芜湖241027

出　　处：《现代科学仪器》2024年第4期190-195,共6页Modern Scientific Instruments

摘　　要：受到温度、法定节假日、甚至突发事件等不确定因素的影响,配电网短期负荷预测准确性不高,针对上述问题,提出一种基于经验模态分解与偏差校正的配电网短期负荷预测方法。对配电网负荷数据进行预处理,包括缺失数据填补、异常值处理以及归一化;利用经验模态分解,将预处理好的配电网负荷时间序列分解为若干独立有限个本征模函数分量,以这分量为输入,利用深度置信网络分析配电网短期负荷值。引入模糊控制方法,将温度变及法定节假日这两种常见的不确定因素考虑在内,优化深度置信网络计算得出的基本短期负荷预测结果,实现偏差校正。结果表明:该方法提高了配电网短期负荷预测精度。Under the influence of uncertain factors such as temperature,rest days and even emergencies,the accuracy of distribution network short-term load forecast is not high.Aiming at the above problems,a distribution network short-term load forecast method based on empirical mode decomposition and deviation correction is proposed.Preprocessing the load data of distribution network,including missing data compensation,abnormal value processing and normalization;Using empirical mode decomposition,the preprocessed distribution network load time series is decomposed into several independent Intrinsic Mode Function components.Taking these components as inputs,the short-term load value of distribution network is analyzed by deep confidence network.The fuzzy control method is introduced to take the two common uncertain factors of temperature change and rest day into account,optimize the basic short-term load forecast results calculated by the depth confidence network,and realize the deviation correction.The results show that the accuracy of distribution network short-term load forecast is improved by the research.

关 键 词：经验模态分解 偏差校正 配电网短期负荷 深度置信网络 

分 类 号：TP33.3[自动化与计算机技术—计算机系统结构]