Oriented Multi-objective Robust Design Optimization To Improve Product Quality

With the development of technology and industrial production, product quality is increasingly becoming a key factor in the success of enterprises. As an important support tool in Continuous Quality Improvement, robust design optimization can be used in product design stage to improve product or process quality. Previous studies were aimed at single quality characteristic, but as the complexity of the problem, we must be settled that was to optimze the robust design of production which has multiple quality characteristics. That was robust design optimization of multiple quality characteristics, and they are increasingly displaying their important theoretical value and application background.In this dissertation, the methodologies for robust design of the products or processes with multiple quality characteristics are developed. To solve the question of multi-objective robust design optimization from the ground up, three stages must be gotten over. Firstly was to choose an appropriate method of robust design. Secondly was to determine a right multi-objective optimization technique. Thirdly was to transforming general optimization to robust optimization. So in this paper, dual response surface method was selected to be the method of robust design, then desirability function of a multi-objective optimization technique is improved to establishment the multi-objective optimization model. Thus the question of multi-objective robust design optimization could be realized. The research includes the following aspects.First, the author analyse the idea of robust design, and mainly introduce two basic robust design methods:Taguchi Parameter Robust Design and Dual Response Surface Model.Secondly, this paper studies several traditional multi-objective optimization methods:quality loss function, Mahalanobis distance and desirability function, then an example is given where the three methodologies are applied to the data, the results show that the better apploache is the one based on the desirability function.At last, the author builds a multi-objective robust design optimization model based on dual response surface model and desirability function, and robustness at the optimality. Finally, the plementation and effectiveness of the proposed method are illustrated by an example, which gives a satisfactory result.