EMD-LS多重分形去趋势波动分析在GIS机械缺陷诊断中的应用

Application of EMD-LS Multifractal Detrended Fluctuation Analysis in GIS Mechanical Defect Diagnosis

气体绝缘组合电器GIS作为110 kV以上电网核心电力设备,检测和诊断机械缺陷对于保障其可靠运行具有重要意义。基于经验模态分解EMD方法与最小二乘法LS提取GIS振动信号中的趋势项,把每个子序列的局部趋势项代入协方差函数而后求取波动函数,采用多重分形去趋势波动分析方法对GIS不同工作状态下的振动信号进行多重分形谱分析,根据多重分形指标来分析GIS的故障状态。该算法通过ZF27-252型弹簧机构的GIS的3种典型机械故障试验数据得到验证,数据表明该方法能有效识别诊断GIS机械故障。

As gas insulated switchgear(GIS) is the core power equipment of the power grid above 110kV,the detection and diagnosis of mechanical defects is of great significance to ensure its reliable operation.The empirical mode decomposition(EMD) method and the Least Squares(LS) are applied to extract the trend terms in GIS vibration signals,and the local trend term of each subsequence is substituted into the covariance function and then the wave function is obtained.The multifractal detrended fluctuation analysis method is used to analyze the multifractal spectrum of GIS vibration signals under different working conditions,and the multifractal index is used to analyze the fault state of GIS.The proposed algorithm is verified by GIS experimental data of ZF27-252 spring mechanism,the data shows that this method is available to effectively identify and diagnose GIS mechanical faults.