Research On Mechanical Multi-Objective Optimization Design Based On Improved Genetic Algorithm

Optimization of design is one of advanced and modern design methods at present time, which combines optimization theory with computer technique and provides perfect and reasonable design schemes for engineering design. It is very difficult to find global optimum result using traditional optimization methods to solve large-scale, complicated, nonlinear and discontinuous optimization problems in practices. Genetic Algorithm(GA), a fresh optimization algorithm, developed in recent year can undertake the responsibilities. GA is an iterative, adaptive, heuristic and random searching method. The characteristics of global optimization and connotative parallel are its most important advantages.In the paper, modifications for GA based on the ordinary one have been put forward and implemented. Probability of getting global optimization result increased greatly and convergence speed up obviously. The improved GA was used to get an effective solution set for multi-objective optimization problems and similarity priority ratio method of fuzzy and grey clustering method of grey system theory were used respectively to rank priorities of effective solutions in the set. The resolving processes of multi-objective optimization problems in mechanical industry (optimization of belt and worm transmission) were explained in detail, which demonstrated that the jobs have been done were valuable for engineering application.