Study On Parameter Optimization For Multi-response Using Hesitant Fuzzy Decision Making Methodology

As the first toolgate to control products' quality,design of processing parameter has a decisive influence on the quality of products.With the increasingly intelligent industrial production and manufacturing,more and more noise factors appear in the production process,which seriously affects the products' quality.In the face of numerous noise's interference,the Taguchi Method design a large number of tests containing noise factor to obtain robust parameters.This mesure may greatly increase the production cost and it is contrary to the principle of production and manufacturing economy.In addition,the orthogonal test table need to expend with the number of noise factors increasing exponentially.This will generate too mach data which may overflow test table.It will be difficult to parameter design.In order to solve these problems,this study proposes three multi-response parameter optimization methods,which aim at designing the parameter optimization combination that can resist noise interference and reduce production cost.The the idea that all the elements in hesitant fuzzy sets tolerate more than one membership degree,is applyed to the process parameters optimization,which is the main feature of this study.It can obviously improve the robustness of the multi-response system.The main contents of research are as follows:(1)In view of the fact that there are many noise disturbances in the production process and their levels are difficult to measure,a parameter optimization method based on hesitant fuzzy decision is proposed.By constructing a hesitant fuzzy decision matrix firstly,it uses the positive ideal point method simplify the multi-response system.Next,the parameter optimization results are obtained through the main effect analysis,which not only improves the product quality,but also reduces the production cost.(2)For the case that the weight of comprehensive optimization index is difficult to be determined,this study improves it through fuzzy logic reasoning.It make full use of the mechanism relationship between responses to obtain a more objective optimization index,which makes the parameter optimization design more reliable.In the stage of parametic design,it uses BP neural network to improve the parameter optimization results obtained by the main effect method,which increases the robustness of the parametic design.(3)Making allowances for that the simplified results of the multi-response system have low resolution,this study fully considers the relationship between the scheme.Then it combines the two-way projection decision with the artificial neural network tosolve the problem of low resolution.Applying the three parameter optimization methods to actual cases,the results after calculating show that the proposed parameter optimization design methods can effectively improve and improve product quality.In summary,this study,combining hesitant fuzzy sets and intelligent technology with parameter optimization,conducts in-depth exploration and researches on the optimization design of robust parameters.It can solve the problem that the existing quality technologies are difficult to reach the higher quality requirements of enterprises in the complex industrial process.At the same time,it enriches the theoretical research results in the field of quality management.