Multi-stage Interactive Genetic Algorithm for Collaborative Product Customization

Runliang Dou,Chao Zong,Guofang Nan

DOI: https://doi.org/10.1016/j.knosys.2015.10.013

IF: 8.139

2015-01-01

Knowledge-Based Systems

Abstract:Products are becoming increasingly more complex and intelligent, which requires users to participate in the design process in order to meet customer demands and enhance market competition. Interactive genetic algorithm (IGA) can effectively solve the optimization problem. However, the challenge still remains for IGA to ameliorate user fatigue and reduce the noise in the process of evolution. To address the issue, a multi-stage interactive genetic algorithm (MS-IGA) is proposed, which divides the large population of the traditional interactive genetic algorithm (TIGA) into several stages according to different functional requirements. The proposed MS-IGA is then applied to the car console conceptual design system, to better capture the knowledge of users’ personalized requirements and accomplish the product design. This is especially important in the field of complex product configuration design, such as in cars, personal computers, smart phones and the like. Through the users’ graphic interface, customers separately evaluate product design at every different stage of its evolution, which makes the proposed algorithm more directional than the TIGA. We also introduce genetic sense units, which represent different functional modules, in order to realize the customers’ collaborative design. The extensive experimental results are provided to demonstrate that our proposed algorithm is correct and efficient according to the efficiency test, convergence analysis and fatigue test for application of the product design system, including car interior and other modular product.