Hierarchical game joint optimization for product family-driven modular design

Abstract

Product family design takes advantage of modularity to enable product variety while maintaining mass production efficiency. Focusing on a set of similar product variants, product family modularity (PFM) is achieved by reusing common components and minimizing fulfillment costs throughout the product realization process. On the other hand, traditional modular design emphasizes technical system modularity (TSM) that focuses on a single product and is geared towards product decomposition in light of technical feasibility. While it is appealing to incorporate product family considerations into the prevailing modularization theories and methods, the key challenge lies in that TSM and PFM are essentially associated with different goals and decision criteria. This leads to a dilemma that TSM and PFM are competing in decision making for identification of modules by grouping similar components. Realizing the importance of game-theoretic decision making underlying product family-driven modular design, this paper proposes to leverage TSM and PFM within a coherent framework of joint optimization. A hierarchical game joint optimization model is developed in line with bilevel programming. A two-dimension evaluation criteria taxonomy is presented for TSM and PFM criteria measure. A bilevel nested genetic algorithm is put forward for efficient solution of the non-linear hierarchical joint optimization model. A case study of robotic vacuum cleaner modular design is reported to gain insight into joint optimization of TSM and PFM. Results and analyses demonstrate that the proposed hierarchical joint optimization model is robust and can empower modular design in cohesion with product family concerns.