Coordinated optimisation of platform-driven product line planning by bilevel programming. Issue 13 (3rd July 2017)

• Record Type: Journal Article
• Authors: Miao, Chenlu; Du, Gang; Jiao, Roger J.; Zhang, Tiebin
• Is Part Of: International journal of production research. Volume 55:Issue 13(2017)
• Page Start: 3808
• Page End: 3831
• Publication Date: 2017-07-03
• Subjects: product line -- product platform -- design methods -- decision-making -- coordinated optimisation -- bilevel programming; Factory management -- Periodicals; 658.57
• DOI: 10.1080/00207543.2017.1294770
• Languages: English
• ISSNs: 0020-7543

Abstract

Abstract : Product line planning (PLP) aims at an optimal combination of product feature offerings, suggesting itself to be a determinant decision for a company to satisfy diverse customer needs and gain competitive advantages. Fulfilment of planned product lines must make trade-offs between product variety and production costs. To balance the costs of product lines, manufacturers often adopt a product platform configuration (PPC) approach to redesign product and process platforms by adding new modules to the legacy platforms. The PPC is an effective means of providing product variety while controlling the manufacturing costs. The PLP and PPC problems have traditionally been investigated separately in the marketing research and engineering design fields. It is important to coordinate PLP and PPC decisions within a coherent optimisation framework. This paper proposes a bilevel mixed 0–1 nonlinear programming model to formulate coordinated optimisation for platform-driven product line planning. The upper level deals with the PLP problem by maximising the profit of an entire product line, whilst the lower level copes with the multiple product platforms optimisation for the optimal PPC in accordance with the upper level decisions of product line structure. To solve this bilevel programming model, a bilevel genetic algorithm is developed to find the optimal solution. A case study of coordinated optimisation between an automobile line and its product platforms is presented to demonstrate the feasibility and effectiveness of the proposed bilevel programming in comparison with a typical 'all-in-one' approach and a non-joint optimisation programming.