摘要
在电力系统中,对电力用户的用电行为实施准确识别和分析对于提高电力系统的运行效率、实现需求响应和能效分析等方面具有重要意义。传统的用电行为识别方法通常采用侵入式采集方式,这会对用户的正常用电活动造成一定干扰。因此,提出一种基于非侵入式负荷辨识的电力用户用电行为识别方法。设计基于PSOSC-MDTW的非侵入式负荷辨识方法,实现电力用户的非侵入式负荷辨识。基于功率分解的特性,设计非侵入式负荷监测模型,模型采用一维U型全卷积神经网络,通过学习总功率与各用电设备功率之间的非线性关系,对输入的总功率序列实施分解,从而得到各个用电设备的功率。提出一种结合XGBoost、LOF、BSMOTE、CAE改进模型的电力用户用电行为识别模型,实现电力用户用电行为识别。测试结果表明,该方法在混合采样参数的取值为0.3时,F1-score与AUC值达到最大。
In the power system,accurately identifying and analyzing the electricity consumption behavior of power users is of great significance for improving the operational efficiency of the power system,achieving demand response,and energy efficiency analysis.Traditional methods for identifying electricity consumption behavior typically use intrusive collection methods,which can cause certain interference to users' normal electricity consumption activities.Therefore,a non-invasive load identification based method for identifying the electricity consumption behavior of power users is proposed.Design a non-invasive load identification method based on PSOSC-MDTW to achieve non-invasive load identification for power users.Propose a power user behavior recognition model that combines XGBoost,LOF,BSMOTE,and CAE improved models to achieve power user behavior recognition.The test results show that the F1 score and AUC values reach their maximum when the mixed sampling parameter value is 0.3.
作者
江御龙
张涛
刘伟
刘永春
张高山
张辉辉
JIANG Yulong;ZHANG Tao;LIU Wei;LIU Yongchun;ZHANG Gaoshan;ZHANG Huihui(Nari Technology Development Limited Company,Nanjing 211106,China)
出处
《自动化与仪器仪表》
2024年第11期280-283,共4页
Automation & Instrumentation
基金
国电南瑞科技项目:工商业用户非介入式负荷辨识技术研究(524608220037)。
关键词
非侵入式负荷辨识
非侵入式负荷分解
电力用户
用电行为识别
non-invasive load identification
non intrusive load decomposition
electricity users
identification of electricity consumption behavior
作者简介
江御龙(1989-),男,江苏盐城人,硕士,工程师,研究方向为用电信息采集,E-mail:nrjylnj@163.com。
