For the fault recognition of rolling bearings,a GA-BP neural network model based on genetic algorithm (GA) and BP neural network was proposed.In this model,the output error of the training data was used as the objective function,and genetic algorithm was used to optimize the initial weight and the threshold of BP neural network.During the process of fault diagnosis,the feature which mixed by the empirical mode decomposition (EMD) energy ratio and time-domain features was used as the input of the neural network to recognize the fault of rolling bearings under different conditions.The results of the fault diagnosis show that this model is better than the traditional BP neural network model in convergence accuracy,convergence speed and recognition rate.