A fault inversion approach based on random forest is used for the inversion of interturn short circuit faults of the motor, which is applied to help high voltage circuit breaker open and close. Firstly, this paper builds a Simulink model, according to the parmeters of the motor, which is used in high voltage circuit breaker, to obtain motor current data in different conditions. Seven working conditions of the mechanism are selected for simulation and the simulation model obtains 2520 fault samples in total. Then, this paper uses wavelet packet transform to handle the current signal and obtains the fault features, including the energy entropy. Finally, the extracted fault features are divided into two sets, containing of training and test sets. The training sets are input into the random forest model to help build the final fault inversion model. The test sets are input into the model to acquire fault inversion results, which indicating that this method can invert the interturn short circuit fault of the motor used in the high voltage circuit breaker. Compared with KNN and XGBoost algorithms, this method has higher inversion accuracy under certain stability conditions.