基于大数据的配电网运行线损异常诊断模型构建

Construction of Abnormal Diagnosis Model of Distribution Network Operation Line Loss Based on Big Data

该文针对当前配电网运行线路损耗异常无法甄选异常候选集且诊断不精准的问题,构建配电网运行线损异常诊断模型。采用了一种新的自适应函数,获取最小碰集的线损电量。通过长短时记忆网络方法预处理线损电量数据,获取日线损序列。结合最小二乘支持向量机将序列转变为优化求解问题,通过求解参数构建线损异常诊断模型。计算测试结果与实际结果之间线损误差,确定线损严重程度。由算例仿真结果可知,该模型线损电量统计结果与实际数据存在最大为0.1 MW的误差,说明使用该模型诊断结果精准。

Aiming at the problem that abnormal candidate sets cannot be selected and the diagnosis is inaccurate due to abnormal line loss in current distribution network operation,a diagnosis model of abnormal line loss in distribution network operation is constructed. A new adaptive function is adopted to obtain the line-loss power of the minimum collision set. The line loss power data is preprocessed by the long-short-term memory network method,and the daily line loss sequence is obtained. Combined with the least squares support vector machine,the sequence is transformed into an optimization solution problem,and a line loss anomaly diagnosis model is constructed by solving the parameters. Calculate the line loss error between the test result and the actual result to determine the severity of the line loss. From the simulation results of the example,it can be seen that there is a maximum error of 0.1 MW between the statistical results of the line loss and the actual data of the model,indicating that the diagnostic results of the model are accurate.