Multi-Objective Optimization of the Housing’s Fillet Angle of the Universal Rolling Mills Using Improved Genetic Algorithm Combined with FEM

In this paper, the standard genetic algorithm is modified based on the deduction of Rolle theory. A kind of genetic algorithms is set up. Combined the modified genetic algorithms with finite element method, a new kind engineering optimization method is established. This optimal method takes fully advantage of the unique feature of genetic algorithms that the optimal solution is achieved by a rapid global search and takes advantage of the unique features of FEM that the computation is accuracy. This method can solve multi-objective problems with more complex constraints in the engineering. The procedure is developed by VC++. Housing’s fillet angle of the universal rolling mills is optimized using this method. Compared with the zero order method and first order method companied with ANSYS, the result of this optimizing method is better.