Parameter Forecast and Parameter Optimization of High Inner-pressure Hydro-forming Tees Based on Genetic Algorithms and Back Propagation Algorithms

摘要

In allusion to the non-linear relation in key parameters(internal pressure, axial extrusion force, radial back force) of Three-way tube hydroforming process, forecast and optimization models are established based on genetic algorithms and back propagation algorithms. By using the capacity of non-linear mappings of Artifical Neural Network and the capacity of global optimization of Genetic Algorithms, and using dynaform Finite Element simulation software to get samples and check results, the models can forecast and optimize parameters of T tube hydroforming process. The original tube could be in any bore, length, and wall thickness of a certain range. The results show that by using these models, parameters' matching test numbers could be reduced, and the quality of T tube is excellent.