基于马尔可夫状态转移的变压器剩余寿命预测  

Residual Life Prediction of Transformers Based on Markov State Transition

作　　者：辛雨 熊炜[1] 袁旭峰[1] 陆之洋[1] 罗云鹤 郭丹 XIN Yu;XIONG Wei;YUAN Xufeng;LU Zhiyang;LUO Yunhe;GUO Dan(School of Electrical Engineering,Guizhou University,Guiyang 550025,Guizhou,China)

机构地区：[1]贵州大学电气工程学院,贵州贵阳550025

出　　处：《电网与清洁能源》2023年第5期20-29,共10页Power System and Clean Energy

基　　金：国家自然科学基金项目(52067004);贵州省科学技术基金项目([2019]1058)。

摘　　要：提出了一种基于马尔可夫状态转移矩阵的变压器缺陷转移率预测方法。利用变压器历史缺陷记录,建立威布尔分布模型拟合得到变压器缺陷转移率,将其应用于马尔可夫状态转移矩阵中,得到变压器不同状态之间的转移率,并引入了回退因子进行修正;基于该缺陷转移率,通过构建比例故障模型对变压器剩余寿命进行预测,结果表明,所提方法预测出的剩余寿命价值更加贴合变压器实际运行状况,验证了其合理性和可靠性。This paper proposes a method based on the Markov state transition matrix for predicting the transformer defect transfer rate.The historical defect records of transformers are used to establish a Weibull distribution model fitting to obtain the transformer defect transfer rate,and apply it to the Markov state transition to obtain the transfer rate between different states of the transformer.In addition,a back-off factor is introduced for correction.Based on the defect transfer rate,the residual life of transformers is predicted by building a proportional hazard model.Finally,the effectiveness and feasibility of the proposed method is verified through an example,which has important guiding significance for the optimization of the transformer maintenance strategy.

关 键 词：变压器 马尔可夫状态转移矩阵 缺陷转移率预测 比例故障模型 剩余寿命预测 

分 类 号：TM411[电气工程—电器]