基于雾计算和随机森林算法的变压器温度监测及故障预测方法研究

Research on Transformer Temperature Monitoring and Fault Prediction Method Based on Fog Calculation and Random Forest Algorithm

本文中作者提出了基于雾计算和随机森林算法的变压器温度多源监测和故障预测方法,通过多种传感装置实现了变压器温度数据的实时监测,通过边缘物联装置构建雾计算节点实现了数据本地汇集和处理,采用数据融合和随机森林算法实现了变压器过热故障的预测预警。试验结果表明,文章提出的方法有效实现了变压器过热故障的实时监测,具备较好的可行性和实用性。

A multi-source data monitoring and fault prediction method for transformer based on fog calculation and random forest algorithm are proposed in this paper by the authors.This method realizes real-time monitoring of transformer temperature data through a variety of sensor devices,and realizes local collection and processing of data by deploying fog computing nodes through edge devices,and this method uses data fusion and random forest algorithm to realize the prediction and alarm of transformer overheating fault.The experimental results show that the proposed method effectively realizes the real-time monitoring of transformer overheating fault,and has good feasibility and practicality.