부산대학교 도서관

The product robust design which based on evolutionary algorithm could improve the general degree of the products greatly, achieve maximum product function, and meet the diverse needs of users. The product characteristic matrix and the product's binary tree structure gene are formed by semantic analysis and hierarchical cluster analysis. Through crossover and mutated of the robust factor, robust population and using population, using evolutionary operator such as similarity, correlation, fitness equation to get the global optimal product design solution, which can ensure that product has no sensitivity and small fluctuations while being used in different environment, facing different user with different usage, and then the product could achieve strong robustness. The example of the high-heeled shoes shows that this method has better adaptability and innovativeness in product design.