Abstract
A modular structure recommendation method and component reducing method is proposed for the product family design. A customer requirement and attributes of products are represented as fuzzy set and fuzzy variable based on not only point-base approach but also range-base approach. Based on the fuzzy-based representation of the product and the customer requirement, candidates of product family are calculated by combination of components and consideration of the range. An evaluation method also proposed for clustering of the solutions of product family. A proto-type system is developed and applied for a design of a handy-drill as an example. The design results indicates that the proposed method can reduce the number of components from twenty-three to fourteen, and fifteen percent of a production cost. Based on the clustering of the generated candidates of the product family plan, a modular structure is found as common characteristics in the high-evaluated plans. The designer can get the modular structure as the highly-balanced modules that can satisfy the customer requirements and low-cost. The design example concludes that the proposed method effectively works for finding of the good modular structure in the product family.