基于检定数据自相关性的低压智能电能表局部异常点检测方法

A Local Anomaly Detection Method for Low-voltage Intelligent Electricity Meters Based on Autocorrelation of Test Data

低压电能表在运行过程中,若不能及时找出局部异常点故障,会直接影响智能电网的电力传输质量。为保证低压电能表的安全运行,提出基于检定数据自相关性的低压智能电能表局部异常点检测方法。该方法首先对电能表内部结构展开具体分析,获取电能表运行影响因素,并基于分析结构,结合检定自相关函数实施电能表数据的去噪处理;通过对低压智能电能表状态特征值的提取,利用BP AdaBoost复合神经网络建立电能表的局部异常点检测模型,将获取的特征值输入模型中,根据模型输出实现低压智能电能表局部异常点的精准检测。实验结果表明,利用该方法开展低压智能电能表局部异常点检测时,可靠性高,检测效果好。

During the operation of the low-voltage electricity meter,if the local abnormal point fault cannot be found in time,it will directly affect the power transmission quality of the smart grid.In order to ensure the safe operation of low-voltage electricity meter,a local anomaly detection method of low-voltage smart electricity meter based on the autocorrelation of verification data is proposed.The method firstly analyzes the internal structure of the electricity meter,obtains the influencing factors of the electricity meter operation,and implements the denoising processing of the electricity meter data based on the analysis structure and the verification autocorrelation function.The BP-AdaBoost composite neural network is used to establish the local abnormal point detection model of the electricity meter,and the obtained eigenvalues are input into the model,and the accurate detection of the local abnormal point of the low-voltage intelligent electricity meter is realized according to the model output.The experimental results show that when using this method to detect local abnormal points of low-voltage intelligent electricity meter,the reliability is high and the detection effect is good.