Abstract:Internal high pressure forming technology is a kind of advanced manufacturing technology that using liquid as a transmission medium,and controlling the internal pressure and axial thrust to achieve the goal of forming hollow parts.It is widely applied in lightweight fields such as aviation,aerospace,and auto.Tube internal high pressure forming process is very complex.The forming results relate to many factors.Among them,the internal pressure,the axial feed loading paths,and their matching relationship are especially significant on the forming results.How to find out the influence law of various factors affecting internal high pressure forming and to make reasonable optimization are important problems faced by internal high pressure forming technology.Uniform design method was used to design BP neural network training samples and testing samples.BP neural network and genetic algorithm were analyzed and combined.Algorithm program of BP neural network and genetic algorithm program were written based on MATLAB language.Loading path and processing parameter of hollow double throw crankshaft were optimized and the optimum forming parameters were gotten.In addition,the accuracy of the simulation result was verified with the software DynaForm.Thus,parameter optimization of loading path of tube internal high pressure forming is finished and forming quality of tube internal high pressure forming is further improved.