Abstract: The high voltage circuit breaker is the key control equipment in the power grid. Its normal operation can ensure the stability of the system. The fault diagnosis of high voltage circuit breakers can detect problems in the early stage of equipment failure and avoid accidents. It analyzes the characteristics of the displacement signal of the high voltage circuit breaker. Four parameters, such as average speed are selected as fault diagnosis features. The fault diagnosis method for high voltage circuit breaker based on Spark platform is proposed. The method principle and parameter selection process are introduced. The proposed method is validated with real data, and the classification accuracy can reach 93%. Finally, the accuracy and time consumption of this method and several traditional classification models are compared and analyzed, which verify the feasibility of this method. The research results provide reference for the fault diagnosis of high voltage circuit breakers.