Study On Innovative Design Of Product Platform Driven By Uncertain User Information

Under the production mode of mass customization, the key to enterprises successdepends on their ability to seize the uncertain demands of customers. It is to say thatcustomers play an important role in business decisions of enterprises. With technologicaladvancements and increasingly fierce market competition, enterprises have urgent demandsfor product design theory and its practice based on product platform. Aiming at productplatform design driven by uncertain customer in mass customization environment, thedissertation fastens on the following issues:(1) After obtaining uncertain demands of customers, they are analyzed and processed.Customer demands have fuzzy semantics, missing information and redundant information.First, fuzzy sets are used to quantify fuzzy semantic information. Then fuzzy membershipdegree is used to discrete them. Secondly, rough set is used to granulate demandinformation to filling missing information. The attribute reduction of rough set is used toeliminate redundant information. This will prepare for the next mapping of product designparameters.(2) TRIZ theory is used to expand customer demands. The product features model isconstructed. Case library of injection machine is established. The generalized euclideandistance of NN is used to compare the information of demand and case to select closermultiple cases. Then F-AHP evaluation model is established to evaluate them to select themost satisfactory products for customers.(3) If design parameters are continuous values, the mapping model of SVR isestablished. If design parameters are discrete values, the mapping model of SVC isestablished. TRIZ is used to solve conflicts of mapping process. Product design parametersare adjusted according to schemes of conflict resolution. Then LSSVC is used to domodular design for products.According to the relationship between design parameters andmodules, modules of similar case are modified, adjusted and done variant design.(4) Take injection machine as an example. By processing uncertain customer demands,configuration design of injection machine is done based on-NNand evaluated based onF-AHP. If customers are not satisfied with configure product，custom design of injectionmachine is done based on rough set and support vector machine. Research results ofconfiguration and custom design for uncertain customer demands based on product platform in this paper are verified by injection machine instance.