摘要
在输电线路变压器运行绝缘老化问题处理过程中,识别老化状态时过于依赖概率分布假设条件,无法准确识别复杂非线性数据。因此,提出输电线路变压器运行绝缘老化状态识别方法。对传感器实时采集的混合信号进行分离处理,获取能够反映绝缘老化状态的局部放电信号。将局部放电信号划分为正、负半周期两部分,计算半周期信号的偏斜度、陡峭度及放电量不对称度,捕获局部放电统计特征参量。将特征参量进行分类学习,识别出绝缘老化状态。实验结果表明,该方法识别结果F_(1)值保持在0.85以上,实现了准确识别。
In the process of addressing insulation aging issues in power transmission line transformers,the identification of aging conditions overly relies on probabilistic distribution assumptions,making it difficult to accurately detect complex nonlinear data.Therefore,this paper proposes a method for identifying insulation aging states in power transmission line transformers.It involves separating and processing mixed signals collected in real time by sensors to obtain partial discharge signals that reflect the insulation aging state.The partial discharge signals are divided into positive and negative half-cycles,and the skewness,steepness,and discharge asymmetry of the half-cycle signals are calculated to capture statistical parameters of partial discharge.These feature parameters are then classified and learned to identify the insulation aging state.Experimental results show that the F_(1) score of this method remains above 0.85,achieving accurate identification.
作者
云凤现
岳莉
于世成
徐浩洋
武凤鹏
YUN Fengxian;YUE Li;YU Shicheng;XU Haoyang;WU Fengpeng(Shandong Institute for Product Quality Inspection,Jinan 250000,China)
出处
《通信电源技术》
2025年第22期46-48,共3页
Telecom Power Technology
关键词
输电线路
变压器
绝缘系统
老化状态
局部放电特征
transmission lines
transformers
insulation systems
aging status
partial discharge characteristics
