Evaluation And Optimization Of Product Design Based On Data Fusion

The product design is a very complex process as "design evaluation-re-design", and it contains a larger number of indeterminate and creative reasoning process. The purpose of evaluation and optimization of product design is making the function and structure of product meet the customers'demand by continuous evaluating and optimizing the product design parameters, and improving the product design scheme. According to statistics, more than 80% cost of product in the life cycle depends on product design. So the evaluation and optimization of product design is becoming more and more important for the new product whether can be successful, and it have been paid more and more attention. Comparing with the traditional product evaluation and optimization methods, in this paper, a model using data fusion technology is put forward to deal with a lot of indeterminate and creative reasoning process in the evaluation and optimization of product design. And then the evaluation and optimization of product design can is more scientific, objective and reasonable by using much theory such as cloud model, Kano model, evidence theory and quantitative theory type I. At last, the model is used in the process of stroller design evaluation and optimization.The data fusion technology which can recognize the object through multiple data sources is introduced into the process of product design evaluation and optimization, and the product evaluation of many different customers is disposed by data fusion technology. Design variables index set which highly meeting the customers'demand is figured out by fusing among some design variables index set, and the product design evaluation and optimization model is constructed in order to get the more scientific, objective and reasonable evaluation by dealing with the customers'feedback, and optimize the product design based on the evaluation.In the cloud theory, the Expected value, Entropy and Hyper Entropy are used to describe the digital characteristics of cloud. So the fuzziness of qualitative concept could be fully combined with the randomness of membership grade, the mapping between qualitative concept and Quantitative numerical. In this paper, the cloud theory is integrated into the product design evaluation and optimization model, and then the X conditions cloud generator is set up for dealing with the uncertainty description of qualitative and quantitative index. The Kano model is used to divide the qualitative description of the product by customers into five levels, and regulate the qualitative evaluation of the product. In this paper, the initial weight of each evaluation criterion is figured out through entropy weight theory, and the adjustment coefficient is determined by using Kano model, and then the final weight of each evaluation criterion is figured out by fusing the initial weight and the adjustment coefficient.The quantitative theory type I is a branch of multivariate statistics, and it's used to research the relationship among some random variables and reveal the data contained in the law. In this paper, the quantitative theory type I is used to select the variables index with the highest customer satisfaction from the limited design variables index set, and then the variables index with the highest customer satisfaction is fused in order to optimize the product design.Now, the 80 generation is growing up, and many of them have been the parents. So they are the main customer who buying the stroller. Due to they are used to buy things online, so we can collect the feedback through the internet. In this paper, the product design evaluation and optimization system based on data fusion technology is constructed by comprehensive using the theory and acknowledge which is mentioned above, in order to achieve the objective and scientific evaluation of the stroller, and continuous improve the product design scheme, and perfect the design parameters of stroller, and at last figure out a product design scheme with the highest customer satisfaction.