基于大数据分析的电网窃电行为特征自动提取模型

Automatic extraction model of power grid stealing behavior characteristics based on big data analysis

目前研究的电网窃电行为特征提取方法数据挖掘能力较差,导致反窃电甄别效果不理想。为了解决上述问题,提出了基于大数据分析的电网窃电行为特征自动提取模型。建立大维随机矩阵,对数据进行预处理,实现反窃电甄别,并完成输出反馈控制。通过线损识别、台区识别、窃电嫌疑判断实现特征识别,利用滑动窗口提取视在功率、三相电压不平衡率、三相电流不平衡率和功率因数等特征参数,分析是否存在异常。实验结果表明,基于大数据分析的电网窃电行为特征自动提取模型能够有效确定特征值平均谱,精准地实现发窃电信息甄别。

The current research on the power grid electricity stealing behavior feature extraction method has poor data mining ability,resulting in an unsatisfactory anti-electricity stealing screening effect.In order to solve the above problems,an automatic extraction model of grid electricity stealing behavior features based on big data analysis is proposed.Establish a large-dimensional random matrix,preprocess the data,realize anti-stealing screening,and complete output feedback control.Realize feature recognition through line loss recognition,station area recognition,and power theft suspect judgment.Use sliding windows to extract characteristic parameters such as apparent power,three-phase voltage unbalance rate,three-phase current unbalance rate and power factor,and analyze whether there is an abnormality.The experimental results show that the automatic extraction model of power theft behavior characteristics of power grid based on big data analysis can effectively determine the average spectrum of characteristic values,and accurately realize the identification of power theft information.