利用模糊关联规则挖掘和遗传算法的工业产品设计优化方法

打开文本图片集

中图分类号：TP393 文献标识码：A

Optimization Method for Industrial Product Design Using Fuzzy Association Rule Mining and Genetic Algorithm

ZHANG Qing， LI Cong， GAO Guangyin TaizhouInstituteof Science&Technoogy，Nanjing UniversityofScienceandTechnology，Taizhou Jiangsu2253Oo，China .bstract： In the initial stage of the industrial product development process， handling a significant amoun of multidimensional industrial data is required.However，the complexity and uncertainty inherent in this process often lead to the occurrence of the Fuzzy Front End （FFE） problem，increasing the difficulty of product design. To address this issue，this study proposes a multi-layered artificial inteligence product design approach aimed at preventing defects in the design process. The framework integrates various artificial intelligence technologies，including big data analytics，a fuzzy inference system based on recursive association rules （RAFIS），and a Mamdani fuzzy inference system. The proposed method incorporates Fuzzy Association Rule Mining（FARM） and Genetic Algorithm （GA） into RAFIS to bridge the gap between customer attributes and design parameters. Firstly，in FFE stage，data collction and management were organized，and the dataset was then inputed into FARM and GA to obtain optimal fuzzy rules and membership functions. Subsequently，these results were utilized to establish a Mamdani fuzzy inference system with customizing product design features. By optimizing parameters within the Mamdani inference system， including the type，partition，and range of membership functions，the method achieved customized product design. The experimental application analysis focused on electric scooters，and a fuzzy comprehensive evaluation method was employed to assess the proposed design solutions. The results indicate that two design approaches obtained high levels of satisfaction，validating the effectiveness and feasibility of the method.

Key words：artificial intelligence；product design； fuzzy association rule mining；genetic algorithm；big data analysis

第四次工业革命，也被称为工业4.0（I4.O），在工艺研究领域引起了极大关注。（剩余11355字）