基于改进HHT算法的GIS机械振动信号时频分析

Time-frequency Analysis of GIS Mechanical Vibration Signals Based on Improved HHT Algorithm

为了研究可现场带电作业的GIS振动信号检测方法,掌握不同机械故障类型下的振动信号特性,建立GIS机械故障诊断数据库,利用110 k V单相分箱GIS设备建立了一套GIS机械故障检测平台。通过构建螺丝松动和互感器振荡两种常见的GIS机械故障,并利用传感器测量GIS正常运行振动信号和两种故障下GIS振动信号,利用总体集合经验模态分解提取振动信号的特征量,再将通过EEMD预处理得到的信号进行希尔伯特变换。实测信号的分析结果说明,采用经EEMD算法的预处理后再进行HT变换的方法对GIS机械振动信号进行时频分析可以有效地处理GIS振动信号,从而建立GIS机械故障诊断数据库,为实现现场带电检测GIS机械故障提供理论依据。

A GIS mechanical fault diagnosis database is established to study on-spot hot-line test methods for GIS vibration signals and to grasp characteristics of the vibration signal in case of various mechanical faults. A 110 kV single-phase GIS equipment is used to establish a GIS mechanical fault detection platform. Vibration signals for normal GIS operation as well as GIS vibration signals under the two faults are measured with sensors. Characteristic quantity of the vibration signals are extracted through the Ensemble Empirical Mode Decomposition, and the Hilbert transform is made on the signals obtained through EEMD preprocessing. Analysis results of the measured signals show that time-frequency analysis of GIS mechanical vibration signal in the above described method can effectively process GIS vibration signal, and the GIS mechanical fault diagnosis database can be established, thus providing a theoretical basis for on-site hot-line detection of GIS mechanical faults.