基于决策树和聚类算法的智能电表误差估计与故障检测

针对智能电能表故障检测问题,研究了低压电能系统模型,提出了一种电能表误差估计方法。该方法采用决策树对异常数据进行过滤,并对不同损失水平的数据进行分类,然后对数据进行聚类,得到不同使用行为的数据集。此外,建立了仪表数据矩阵,采用递推算法求解矩阵方程,对仪表误差进行估计。采用递推算法求解方程,并对仪表误差进行估计。最终通过算例验证了所提方法的有效性。

Error Estimation and Fault Detection of Smart Meter Based on Decision Tree and Clustering Algorithm

Aiming at the problem of intelligent watt hour meter fault detection, the low-voltage power system model is studied, and an error estimation method of watt hour meter is proposed.The method uses the decision tree to filter the abnormal data, classifies the data with different loss levels, and then clusters the data to obtain the data sets with different usage behaviors. In addition, the instrument data matrix is established, and the instrument error is calculated by solving the matrix equation. Finally, the effectiveness of the proposed method is verified by an example.