An approach to configuring product family using rough set theory

Configuration design is one of the most important elements in the design for product family, with which designers could customise products for customers according to their specific requirements. The traditional approach to configuring products is based on configuration rules which reflect the mapping relations between customer requirements and design variables. But how to acquire the configuration rules is becoming the bottleneck. A kind of general way is to depend on design experts themselves to create such rules. It would take plenty of time for design experts to establish rules database. As such, this paper presents a framework of configuration design for product family through giving the variant components with different attribute values, in which the acquisition of configuration rules is connected with the idea of data mining using the rough set theory from large numbers of historical design data instead of the design expert's experience. It can help not only design experts but also novice designers analyse the dependence between customer requirements and design parameters, and acquire the configuration rules hidden in the historical data easily.