An integrated cognitive computing approach for systematic conceptual design

• Published in: Journal of Zhejiang University. A. Science Vol. 17; no. 4; pp. 286 - 294
• Main Authors: Zheng, Hao, Feng, Yi-xiong, Tan, Jian-rong, Zhang, Zhi-feng, Zhang, Zi-xian
• Format: Journal Article
• Language: English
• Published: Hangzhou Zhejiang University Press 2016; Springer Nature B.V
• Subjects: Associative memory; Case studies; Civil Engineering; Classical and Continuum Physics; Cognitive ability; Cognitive psychology; Computation; Computer applications; Computer memory; Conceptual design; Decision making; Design; Design engineering; Engineering; Genetic algorithms; Industrial Chemistry/Chemical Engineering; Integration; Life cycle engineering; Life cycles; Mechanical Engineering; Neural networks; Product life cycle; Similarity; Software; Technical services; 优劣解距离法; Genetic algorithm (GA); 遗传算法; Cognitive computing; Technique for order preference by similarity to ideal solution (TOPSIS); 概念设计; 认知计算; Conceptual design; TH122
• ISSN: 1673-565X; 1862-1775
• DOI: 10.1631/jzus.A1500161

Conceptual design plays an important role in product life cycle, which requires engineers to use sound design theory, cross-disciplinary knowledge, and complex technical support to acquire design concepts. However, the lack of sufficient computational tools makes it difficult for designers to fully explore in the wide design solution spaces. Therefore, this paper proposes an integrated cognitive computing approach to formalize the cognitive activities of conceptual design. A cognitive computing model composed of concept associative memory, concept generation, and decision-making process is established based on the integration of cognitive psychology and engineering design. First of all, the Hopfield neural network is used to acquire similar concept solutions for specific subfunctions from a knowledge base. Then, morphological matrix and genetic algorithm are introduced to produce a set of feasible candidate solutions in the concept generation process. Furthermore, a technique for order preference by similarity to an ideal solution is applied to evaluate the generated concept solutions and obtain the optimal solution automatically. Finally, a case study is given to demonstrate the effectiveness and efficiency of the proposed approach.