By taking the diamond granite processing of milling cutter as the study object, the key factor of milling force of the processing efficiency directly affected by milling cutter breakage was studied. Aimed at the limitation of granite processing 〖JP+1〗milling force theoretic formula in practical application, a method was proposed which combined the experiment and the neural network forecast. By 〖JP〗using experiment to get the model samples, testing samples and the tool of fracture limit predicted by milling force of neural networks, and using RBF neural network on granite processing milling force to have the modeling and prediction, the accuracy of the model was verified by the experimental data, to obtain the optimal processing parameters by the prediction model which was applied in the actual process. The results show that this is an effective method to improve the processing efficiency of tool.