基于变压器绝缘劣化降级时间预测的绝缘老化评估方法  

Predicted Insulation Aging Assessment Method Based on Insulation Deterioration Degradation Time of Transformer

作　　者：刘琦 付楚珺 于鲜莉[1] 李昂[1] 李波[1] 郑思齐 LIU Qi;FU Chujun;YU Xianli;LI Ang;LI Bo;ZHENG Siqi(Inner Mongolia Power Research Institute,Hohhot 0l0020,China)

机构地区：[1]内蒙古电力科学研究院,呼和浩特010020

出　　处：《高压电器》2023年第3期116-122,共7页High Voltage Apparatus

基　　金：内蒙古电力(集团)有限责任公司科技项目(内蒙古电网运行变压器油远程监控系统及充油设备绝缘老化检测技术研究与应用)。

摘　　要：电力变压器绝缘老化评估为电网的长期安全稳定运行提供了技术支持,在保证非破坏性试验的基础上,对绝缘状态进行诊断也是一直以来的研究热点。文中基于广义维纳过程对变压器建立老化模型,通过相关试验数据得到老化诊断指标,从而衡量变压器绝缘老化状态。对变压器健康状态进行分级,采用上述模型预测最短降级时间,相比于直接进行寿命预测,短期预测可以将未来人为操作信息纳入模型,最大化解决信息挖掘不充分的问题。最后对老化评估模型的相关分布参数进行优化求解,采用内蒙古电网设备真实历史数据迭代训练,得到适用于该地区的变压器老化状态评估模型,验证了该方法可以准确的评估变压器的老化状态。Insulation aging assessment of power transformer has provided technical support for the long-term safe and stable operation of power grid.The diagnosis of the insulation condition based on assuring non-destructive tests has always been the hot spot of such research.In this paper,the aging model for transformer based on the generalized Wiener process is set up.The aging diagnosis index is obtained through related test data,thus measuring the aging condition of transformer.Through health status of transformer is classified and the above model is used to predict the minimum degradation time.Short-term prediction,compared to the direct life prediction,can add future human operate information into the model,and effectively solve the problem of inadequate information mining.Finally,the related distribution parameters of the aging assessment model is subject to optimal solution.The real historical data of power grid equipment in Inner Mongolia power grid is used to perform iterative training to get the transformer aging state assessment model of the region.Moreover,it is proved that the method can assess accurately the aging condition of the transformer.

关 键 词：变压器 绝缘老化评估 广义维纳过程 优化求解 

分 类 号：TM855[电气工程—高电压与绝缘技术] TM411