基于声纹及振动的变压器故障诊断技术研究

Transformer Fault Diagnosis Technique Based on Voiceprint and Vibration

噪声和振动包含了变压器机械状态的大量信息。基于噪声声纹成像和振动信号的分析方法是变压器机械状况带电检测的有效手段,分析声纹成像及振动信号测试结果,对变压器故障诊断具有重要意义。文中基于变压器铁心及绕组噪声及振动机理,建立振动噪声测量系统,并对1台110 kV变压器进行测量。根据声纹成像噪声源定位和振动信号特征值测试分析结果,判定该主变C相高压套管升高座存在故障,验证了该诊断技术的有效性。

Noise and vibration signals contain rich information about transformer’s mechanical state. Combination of noise voiceprint imaging technology and vibration signal analysis is an effective method to detect the mechanical states of transformer winding and iron core. The analyses of transformer’s noise voiceprint and vibration signal results are of great importance to transformer fault diagnosis. In this study,a noise and vibration measurement system was established based on the analyses of the noise and vibration mechanisms of transformer’s iron core and winding,and a 110 kV transformer was measured. According to the analysis results of voiceprint imaging based noise source location and vibration signal’s characteristic value,the transformer was judged to have a fault in its C-phase high-voltage level bushing,verifying the effectiveness of the proposed diagnostic technique.