变压器油中溶解气体浓度的GM(1,n)预测  

Prediction of Dissolved Gas in Transformer Through GM(1,n) Model

作　　者：李艳[1] 程宏波[2,3] 辛建波[2] 康琛[2] 

机构地区：[1]国网江西省电力公司赣州供电分公司,江西赣州341000 [2]国网江西省电力公司电力科学研究院,江西南昌330096 [3]华东交通大学电气与自动化工程学院,江西南昌330013

出　　处：《华东交通大学学报》2017年第3期131-136,共6页Journal of East China Jiaotong University

基　　金：江西省重点研发计划项目(20161BBH80033);江西省博士后择优资助项目(2016KY36)

摘　　要：电力变压器是电力系统的关键设备,其状态的发展演变是一个随时间累积的过程,变压器油中溶解气体的含量会随着时间变化逐步发展,对变压器油中溶解气体的浓度进行预测可以提前发现变压器存在的故障隐患,为预防性维修提供依据。在分析变压器油中溶解气体产生机理的基础上,采用灰色模型,利用在线监测系统获取的溶解气体浓度数据,应用GM(1,n)模型进行预测的结果表明,该模型能够预测出精度较高的油中溶解气体浓度,由此可对变压器未来的状态进行预警,为预防性维修提供依据。Power transformer is the key equipment of power system, whose state is a cumulative process over time. The content of dissolved gases in transformer oil changes over time. By predicting the concentration of dis- solved gases in transformer oil, we can find the possible faults of the transformer in advance, which may provide some reference for preventive maintenance. Based on the analysis of dissolved gas generation mechanism in transformer oil, this study adopted gray model and obtained the concentration data of dissolved gases by online monitoring system. Through the application of GM （1 ,n） model ,the results showed that this model is able to ob- tain the concentration content of dissolved gases with high accuracy. It thus could offer the pre-warning for the future state of transformer and provide preventive maintenance accordingly.

关 键 词：电力变压器 灰色模型 GM(1 N)模型 故障预测 

分 类 号：TM762[电气工程—电力系统及自动化]