Multi-response Robust Optimization Considering Quality Loss And Tolerance

Research on robust optimization design for multi-response problems has been the focus and difficulty in the field of optimization design.At the present stage,the composition of multi-response products is usually complex,and its main characteristic is that it has multiple quality characteristics.Although the traditional multi-response optimization method is suitable for the optimization of multi-response products,it only consider the optimality of optimization results while ignoring the influence of uncontrollable factors and controllable factors’ tolerance to the feasibility and robustness of the optimization results,eventually lead to large fluctuations in product performance,high manufacturing costs and even produce the problem such as non-conforming product.Although the traditional robust optimization design method can solve the problem of traditional multi-response optimization method to a certain extent,the optimal objects are mainly focused on products with single quality characteristic which has been difficult to meet the needs of multi-response products.Therefore,how to combine thought of robustness to multi-response optimization methods to optimize the products with multiple quality characteristics have become main problems which need to resolve in the optimization phase of multi-response product.To the end,a new systemic model of multi-response robust optimization considering quality loss and tolerance is proposed to solve these problems.The concrete research content is as follows:First of all,according to the problem that the traditional multi-response optimization method only consider the optimality and ignore the influence of uncontrollable factors on the robustness of the optimization result in parameter design phase of products,by combining the dual response surface method and multivariate loss function,a new method of multi-response robust optimization based on improved quality loss function is proposed to solve this problem.Finally,the example proves that the new method can effectively improve the robustness of multi-response problem.Secondly,in order to solve the problem that the traditional multi-response optimization method and the traditional robust optimization design method ignore the disturbance of controllable factors’ tolerance in parameter design phase of products,a new model of multi-response robust optimization considering tolerance of factors is proposed by considering the disturbance of controllable factors’ tolerance and the method of multi-response robust optimization based on improved quality loss function which is mentioned above.Finally,the example proves that using the new method can get the feasibility robustness.Eventually,on the basis of getting the feasible and robust parameter combination of multi-quality products,tolerance design is usually needed to further control the robustness and cost of multi-quality products.However,the traditional tolerance design optimization method is usually only for the product with a single quality characteristic,so a new method of robust tolerance design based on tolerance-quality loss model is proposed.The method by constructing the multi-response tolerance quality loss model with load in order to taking into account the quality level and manufacturing cost of the multi-quality products,thus completing the tolerance design of multi-quality products.Finally,an example is given to show that the proposed method can effectively control the robustness and cost of the product.