基于多特征融合的GIS局部放电状态检测方法

Multi-feature Fusion-based Detection of Partial Discharge State of GIS

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。

In order to optimize the detection effect of GIS partial discharge state and improve the detection rate,the principle 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 subsequent 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 application of the proposed method was proved capable of achieving detection rates of discharge state higher than 98%with respect to four different types of partial discharge insulation fault models,and better capturing partial discharge characteristics of GIS equipment,and facilitating more accurate determination of GIS discharge state.