摘要
基于已有电能计量自动化系统的实时采集的电量、负荷、报警及线损数据,分析现有用电检查所得窃电现象的样本,通过构建指标体系,建立一种自适应的防窃漏电诊断模型。基于模糊评价法与神经网络方法,结合了两者的优点,能够反映窃漏电现象的特征。经实证研究,能够实时监测计量自动化系统的运行数据,发现诊断窃漏电现象,具有较强的识别能力。
Based on the data collected on a real-time basis by the existing automatic electric power metering system about quantity of electricity, load,alarm and line loss,this paper analyzes electricity stealing samples obtained in the current inspection method,and by creating an index system,establishes a self-adaptive diagnosis model for electricity stealing and leakage.Combining fuzzy evaluation method and neural network method,this model can monitor on a real-time basis the operational data of the automatic metering system to detect electricity stealing and leakage and has a strong capability of identification.
出处
《电气自动化》
2014年第2期60-62,共3页
Electrical Automation
基金
广东电网公司重点科技项目
K-GD2012-341
大规模智能用电系统海量数据处理与数据挖掘技术研究及应用
关键词
窃漏电
计量自动化
模糊神经网络
自适应
负荷梯度
electricity stealing and leakage
automatic metering
fuzzy neural network
self adaptation
load gradient