Study On Information Feedback System Of Costume Supply Chain Based On Sensory Evaluation

With the demands of clients becoming more and more diverse and individual, the design of products is combined tightly to the market demands. Clients pay more attention to the comfort, the appearance, the function, and the health of the fabric. So, how to quickly respond to the market changing, how to satisfy the customer's demand, and how to decline the cost, as so to earn the better profit are a series of important conundrums in the front of the enterprises.In this thesis, taking the practical demands of supply chain management as the point of departure, we study and design the information feed system prototype by combining the sensory evaluation and intelligent technique. The system consists of three parts, as the feedback of the product quality of the providers, the feedback of clients demands, and the enterprise decision making support system.Firstly, we get the objective evaluation data by measure the samples and process the data by dimension reduce and clustering analysis. As principal component analysis mainly use the linear correlation of the data, we propose a nonlinear principal component analysis method, by combining the mercer kernel function with it. While processing data with clustering analysis, the K-means clustering converges to the local minima. In order to overcome the problem, we propose a novel genetic algorithms-based mercer kernel clustering method. Simulation experiments verify its classification effectiveness.Next, we apply sensory evaluation to analysis the consumption psychology of customers, and propose an analysis model of questionnaire data. When get the questionnaire data from market research, we do factor analysis to study the questions relationship, and get the new score data. At last, we use fuzzy c-means clustering analysis method to process the data and interpret the results.Then, we study decision support system mainly on the relation of the objective and subjective evaluation of the samples. When get the feedback information from providers and clients, enterprise organizes sensory evaluation experts to evaluate the fabric samples. In order to process the data easily, we use a 2-tuple fuzzy linguistic representation model to transform the nonnumerical value to numerical value and aggregate the data. And we use adaptive-network-based fuzzy inference system to study the relationship of the objective and subjective evaluation.Finally, on .Net software platform, we develop the information feedback system of costume supply chain based on B/S structure. The system consists of the client tier, the middle service tier, and the database tier. The client tier mainly designs web pages to help user interact with the system. And the middle service tier has two functions. One is to encapsulate the business rules of the enterprise, and the other is to encapsulate the intelligent technology. Last, the essence function of the database tier is to provide data for the middle service tier.The work in this thesis puts focus on the practical demands of the enterprise, and the practical applications, and has a certain creativity and practical influence for the sensory evaluation applied to costume supply chain management. Based on the developed prototype system for costume supply chain management, we can do further study and development, and wish to provide a better platform for the supply chain management applied to the enterprise.