Research On Optimal Design For Complex Multiple Response Problems Of Considering The Quality Loss

The optimization design of complex multiple response problems has always been a difficulty in the field of optimal design, its complexity are often shown as:(1) if there is a correlation between each response;(2) each response has different dimension or different measure standard 、 it’s hard to unified measure;(3) the balance of designer’s subjectively intention and objectively importance of each response as set weight about each response;(4)each response may exists different characteristic requirements such as “larger the better”、“smaller the better” or “definite purpose”; etc. The existing optimal design methods research on complex multiple response problems about products, mostly focuses on the optimal characteristics of multiple response and rarely consider the invisible quality loss due to deviate from design target of each response may caused by multiple optimization design method itself, and then appear safety concerns of product quality due to the quality loss existed in design stage, even appear customer complaints、products recalls and other quality problems of product in the product using stage. For this, put forward a systemic optimization design method of complex multiple response problems considering the quality loss about product design. The main research works are as follows:Firstly, in view of the traditional optimization design of complex multiple response problems based on the Response Surface Methodology(RSM), doesn’t considering the correlation between each response in dealing with multiple response problems as using the traditional satisfaction function, the traditional composite satisfaction function was improved,make it can be used the situation not only product each response independent but have correlation between each response;Secondly, scientifically gives optimal weights value to each response using subjective and objective weight theory, by putting subjective weight method and objective weight method fusion organically, can express decision maker’s intention, also can effectively reflect objective practical importance of each response, better reflect the important degree of each response, avoid the shortage situation of credibility about set weight values to each response caused by unilateral using subjective empowerment theory or objective empowerment theory;Thirdly, in order to overcome the insufficient of the existing optimization design of multiple responses based on satisfaction function method only considers the compound satisfaction as a single evaluation index, from the point of view of the optimization design method of complex multiple response may causes product quality loss, put forward tocalculation quality loss of different situation as set different weight to each response through the multivariate quality loss function, makes the evaluation index of optimization design of multiple response from single the compound satisfaction index into two evaluation index as the compound satisfaction and quality loss, provide a more comprehensive theoretical basis for comprehensive evaluation the difference between each optimization design scheme of multiple response problems;Finally, use a micro precision positioning stage as an example, according to its technical parameters requirements, the design of multiple response was optimized as considering the quality loss of design, the feasibility and effectiveness of this method was verified.