Based on the forming theory of planetary cycloidal gear

using simulative method of BP neural network

the tooth profile of planetary cycloidal gear is accurately designed. The nonlinear model is established between the gear parameter and profile by BP neural networks in the method. The simulating experiment proves that adopt the neural network s nonlinear simulation characteristic to design tooth profile of cycloidal gear can implement more precise and shorten the design period; The method apply profile design for the non-involute gear.