Personalized Product Model Dynamic Evolution For Customer Multi-domain Requirements

Under the background of economic globalization, the market environment is becoming complicated and changeful, in order to be more flexibility to adapt to the dynamic customer requirement, reduced the waste of resources caused by mass production, mass customization production mode becomes the hotspot to solve the problem. Dynamic multi-domain customer requirement, the author puts forward the concept of flow node, it is planned that flow node module granularity, it is constructed dynamic evolution model faced the personalized needs of customers which made customers more deeply involved in the design of the product, the concrete content is as follows:(1) The characteristics and the classification of the customer requirement and the way of gaining customer requirement are analyzed, customer requirement is represented formally by using the theory of matter-element, and the fuzziness of the customer requirements is transformed into the product reliable design parameters.(2) One of the key issues of personalized product customization technology is to solve module granularity in the product planning, granularity planning based on complex network community detection module, the complex networks theory is introduced into the product module partition of the field, the concentration correlation matrix is constructed on the basis of the physical correlation matrix, the structure correlation matrix, the function correlation matrix, using the theory of fuzzy sets to module partition of the matrix related to the concentration data, the F statistic method was applied to division results to get the preselection, the modularity in complex network is treated as the index of module partition.(3) The concept of flow node is proposed, the similar module set is treated as flow node instance set, flow nodes constraint relationship between multi-domain model is set up, the trigger value and jump value acted as constraints conditions of flow within and between flow, rules extraction strategy is deduced based on rough set theory.(4) Aiming at customer requirements expressed incompletely and flow nodes cannot be instantiated, flow node instances could be optimized and ranked in the condition of unconstrained. In this paper, it takes the flow node example optimization method which based on fuzzy multiple attribute evaluation, firstly, instance attribute weight vector is obtained by using the analytic hierarchy process, then making evaluation in the condition of unconstrained based on the fuzzy multiple attribute decision, enterprises and designers to customers comprehensive benefit biggest instance, solved the problem of the customer requirements expressed incompletely.(5)The evolutionary operator is built, the valid instance and virtual instance are defined, providing the four basic evolution behavior for product evolution, realization of virtual instances to instantiate again through customer requirements for product customization system, so as to drive the entire product model evolution.