基于IMFO-RBF的变压器绝缘老化评估策略  

Evaluation Method of Transformer Insulation Aging Based on IMFO-RBF

作　　者：赵春明 冷俊 翟冠强 王昕[2] ZHAO Chun-ming;LENG Jun;ZHAI Guan-qiang;WANG Xin(Electrie Power Research Institute of Jilin Province,Changchun 130021,China;Center of Electrical&Eleetronie Technology,Shanghai Jiao Tong University,Shanghai 200240,China)

机构地区：[1]国网吉林省电力有限公司电力科学研究院,吉林长春130021 [2]上海交通大学电工与电子技术中心,上海200240

出　　处：《水电能源科学》2022年第8期203-207,共5页Water Resources and Power

基　　金：国网吉林省电力有限公司科技项目(2019-24)。

摘　　要：为更好地评估变压器绝缘老化状态,先深入研究了变压器绝缘老化程度与极化去极化电流曲线的联系,分析了油中糠醛含量检测法,并提取了两者的老化特征量;其次建立绝缘老化状态的RBF神经网络评估模型,通过改进飞蛾优化算法优化RBF神经网络的径向基函数中心点位置,提高评估模型的评估精度;最后基于IMFO-RBF评估模型,结合PDC电流数据及糠醛含量数据,建立变压器绝缘老化健康值,分析了变压器绝缘老化状态。仿真结果表明,IMFO-RBF能够通过PDC数据及油中糠醛含量进行融合评估,且评估结果与实际情况贴近,具有一定的参考性与实用性。In order to better evaluate the aging state of transformer insulation,firstly,the relationship between the degree of transformer insulation aging and the polarization depolarization current curve was deeply studied.The furfural content detection method in the oil was also analyzed,and the aging characteristics of the two were extracted.Secondly,establish the RBF neural network evaluation model of the insulation aging state was established,and the position of the radial basis function center point of the RBF neural network was optimized by improved the moth optimization algorithm for improving the evaluation accuracy of the model.Finally,based on the IMFO-RBF evaluation model,combined with PDC current data and furfural content data,the transformer insulation aging health value was established to analyze the insulation aging state of the transformer.The simulation results show that the IMFO-RBF can conduct fusion evaluation based on PDC data and furfural content in oil,and the evaluation result is close to the actual situation,which has certain reference and practicability.

关 键 词：变压器绝缘 极化去极化电流法 油中糠醛含量 改进飞蛾优化算法 变压器绝缘老化健康值 

分 类 号：TM855[电气工程—高电压与绝缘技术]