Conventional methods of transformer winding fault diagnosis directly determine transformer winding faults without feature extraction of vibration signals,resulting in low diagnosis accuracy.Thereby the present study proposed a multisensor-information-fusion-based method to diagnose transformer winding faults.The method obtains winding vibration characteristics by extracting from winding vibration signals measured before performing diagnosis,and utilizes the extracted vibration characteristics and employs the multisensor information fusion algorithm to determine winding deformation faults.The proposed method was proved by comparative experiments to have superior diagnosis accuracy under different data preprocessing compared with conventional methods,and thereby,a certain research value.
multisensor information fusiontransformerwinding deformationfault diagnosis methods
NETL
NSTL
万方数据
