Computer aided selective assembly model for multi-dimension chains

Studies of selective assembly problem have been criticized for two reasons: most models were confined to single dimension chain; quality loss could be used only on symmetrical tolerance. To deal with these problems, a new selective assembly model for multi-dimension chains was proposed. Based on Taguchi method, evaluation rules for quality loss oriented to asymmetrical tolerance was established. According to Genetic Algorithm (GA), a new multi-objective GA was designed to remain population diversity and weighted Pareto method was used to represent preference information. This model was applied in enterprise information system, and its optimization efficiency was proved by statistics method. Simulation results revealed that the proposed model was able to find much better spread of solutions and better convergence near the true Pareto-optimal front.