摘要
电力系统负荷聚类和特性分析对电网的安全与经济调度、运行具有重要意义,是提升调度人员对电网感知能力的重要技术手段。为了解决传统负荷聚类方法需要人工设定负荷特征指标和无法考虑负荷时序特性等问题,提出了一种由长短期记忆(LSTM)自动编码器构成的负荷聚类方法。利用LSTM的时序记忆能力和自动编码器的非线性特征提取能力,实现了考虑负荷时序特性的自动特征提取和非线性降维。然后,基于提取的负荷特征采用k-means聚类算法进行电力负荷聚类分析。最后,采用实际供电区域的负荷数据进行验证,并对负荷特性进行详细的分析。结果表明所提方法与其他负荷特征提取方法相比,有较好的负荷聚类效果。
Load clustering and characteristic analysis of the power system are of great significance for the safe,economic dispatching and operation of the power grid,which is an important way to improve the perception ability of regulators.In order to solve the problems that traditional load clustering methods require manual setting of characteristics indices for power load and cannot consider sequential characteristics of loads,a load clustering method is proposed which is made up of long-short-term memory(LSTM)and auto-encoder.The sequential memory capability of LSTM and extraction capability of non-linear characteristics for the auto-encoder are used to achieve automatic extraction of characteristics and non-linear dimensionality reduction considering sequential characteristics of loads.Then,based on the extracted load characteristics,k-means clustering algorithm is used for clustering analysis of power loads.Finally,the load data in an actual power supply area is used for verification,and load characteristics are analyzed in detail.The results show that compared with other extraction methods of load characteristics,the proposed method has better efficiency for load clustering.
作者
庞传军
余建明
冯长有
刘艳
江叶峰
PANG Chuanjun;YU Jianming;FENG Changyou;LIU Yan;JIAN Yefeng(NARI Group Corporation(State Grid Electric Power Research Institute),Nanjing 211106,China;Beijing Kedong Electric Power Control System Co.,Ltd.,Beijing 100192,China;National Electric Power Dispatching and Control Center,State Grid Corporation of China,Beijing 100031,China;Stale Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 210024,China)
出处
《电力系统自动化》
EI
CSCD
北大核心
2020年第23期57-63,共7页
Automation of Electric Power Systems
基金
国家电网公司科技项目(5100-201940013A-0-0-00)。
关键词
负荷聚类
负荷特征
长短期记忆
自动编码器
load clustering
load characteristic
long-short-term memory(LSTM)
auto-encoder
