Wang, Xiaoming
(Electric Power Research Institute of Guangxi Power Grid,Nanning,China,530023)
,
Lin, Xiangyu
(Electric Power Research Institute of Guangxi Power Grid,Nanning,China,530023)
,
Zhou, Ke
(Electric Power Research Institute of Guangxi Power Grid,Nanning,China,530023)
,
Lu, Yufeng
(Electric Power Research Institute of Guangxi Power Grid,Nanning,China,530023)
High voltage circuit breaker is a critical equipment of power system. It is very important to ensure the circuit breaker to operate in a normal state. According to statistics, most defect and fault of high voltage circuit breaker is caused by mechanical faults. In this research, the sound and curren...
High voltage circuit breaker is a critical equipment of power system. It is very important to ensure the circuit breaker to operate in a normal state. According to statistics, most defect and fault of high voltage circuit breaker is caused by mechanical faults. In this research, the sound and current signals were collected in the simulation experiment of typical mechanical faults, namely iron core jam, two kinds of tripping mechanism faults, and spring fatigue. Then the signals were down sampled, flipped and stacked to fit deep learning model. A convolution neural network (CNN) model consisting eight layers was developed to extract features and categorize faults from the pre-processed signals. The results indicate that the mechanical fault diagnosis accuracy rate is up to 94%, higher than conventional methods using sound or current signal.
High voltage circuit breaker is a critical equipment of power system. It is very important to ensure the circuit breaker to operate in a normal state. According to statistics, most defect and fault of high voltage circuit breaker is caused by mechanical faults. In this research, the sound and current signals were collected in the simulation experiment of typical mechanical faults, namely iron core jam, two kinds of tripping mechanism faults, and spring fatigue. Then the signals were down sampled, flipped and stacked to fit deep learning model. A convolution neural network (CNN) model consisting eight layers was developed to extract features and categorize faults from the pre-processed signals. The results indicate that the mechanical fault diagnosis accuracy rate is up to 94%, higher than conventional methods using sound or current signal.
이 논문을 인용한 문헌
관련 콘텐츠
*원문 PDF 파일 및 링크정보가 존재하지 않을 경우 KISTI DDS 시스템에서 제공하는 원문복사서비스를 사용할 수 있습니다.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.