摘要
为了优化 GIS局部放电状态检测效果,提高检测率,引入多特征融合方法原理,开展了基于多特征融合的GIS局部放电状态检测方法研究.首先,需要采集 GIS设备的局部放电数据并对其进行预处理,为后续放电状态检测提供有力的数据支持.其次,提取 GIS局部放电特征,包括时域特征与频域特征.在此基础上,融合提取到的时域特征和频域特征,根据融合处理后的 GIS局部放电特征,检测 GIS局部放电状态.实验结果表明,提出的方法应用后,4 种不同类型的局部放电绝缘故障模型的放电状态检测率均达到了 98%以上,能够更好地捕捉到 GIS设备的局部放电特征,从而更准确地判断其放电状态.
Abstract
In order to optimize the detection effect of GIS partial discharge state and improve the detection rate,the princi-ple of multi feature fusion was introduced,and a study detection method based on multi feature fusion was carried out.First partial discharge data from GIS equipment was collected and preprocessed to provide strong data support for subse-quent discharge state detection.Second GIS partial discharge features including time-domain feature and frequency-domain feature were obtained by feature extraction.On this basis,the extracted time-domain and frequency-domain features were fused,and the GIS partial discharge state was detected by using the fused GIS partial discharge characteristics.The appli-cation of the proposed method was proved capable of achieving detection rates of discharge state higher than 98%with re-spect to four different types of partial discharge insulation fault models,and better capturing partial discharge characteris-tics of GIS equipment,and facilitating more accurate determination of GIS discharge state.
NETL
NSTL
万方数据