基于振动信号的变电站变压器运行状态在线监测

Online Monitoring of Transformer Operation Status in Substations Based on VibrationSignals

由于传统方法主要通过感知内部元件丰富的状态信息来监测变压器的运行状态,导致超高压(Ultra-High Voltage,UHV)变压器带电运行下的故障识别结果存在较大误差,因此提出基于振动信号识别的超高压变电站变压器运行状态在线监测方法。采集500 kV超高压变电站变压器的振动信号,并填充处理采集信号中的缺失值,运用BP神经网络识别故障振动信号进行变压器运行状态的评估,实现在线监测。实验结果表明,设计方法识别变压器故障振动信号的准确率较高,证实了该方法可以高精度地监测变压器运行状态。

Because the traditional method mainly monitors the operating status of transformers by sensing the rich status information of internal components,there is a large error in the fault identification results of Ultra-High Voltage(UHV)transformers under live operation.An online monitoring method of transformer operating status of UHV substations based on vibration signal recognition is proposed.The vibration signal of 500kV ultra-high voltage substation transformer is collected,and the missing value in the collected signal is filled and processed.BP neural network is used to identify the fault vibration signal for the evaluation of transformer operation status,and online monitoring is realized.The experimental results show that the accuracy of the design method is high,which proves that the method can monitor the transformer operation state with high precision.