An arc contacts life assessment method based on dynamic resistance measurement and bp neural network

Arc contacts ablation of high voltage SF6 circuit breaker will cause potential problems for safe operation of the grid. In view of the difficulty of predicting arc contacts life by existing methods, an arc contacts life assessment method based on dynamic contact resistance and BP neural network is proposed. The dynamic contact resistance-travel curve is obtained from the high-current ablation experiments, from which the data characterizing the arc contacts state is extracted as samples to train the neural model. The model is optimized by k-fold cross-validation method, which is finally validated by experimental results. The results show that the arc contacts life prediction value agrees well with the experimental value, which means that the BP neural network can effectively predict the arc contacts life and provide a basis for decision making for arc contacts overhaul and replacement.