An interactive generative design technology for appearance diversity – Taking mouse design as an example

Abstract

As one of the core competitiveness of the product, the appearance should be designed to cater to the various personalized needs of consumers and adapt to the rapid iterations of the product. To improve the design efficiency, reduce the design difficulty, and ensure the diversity of appearance design results, an interactive generative design technology based on mapping from features to parameters (IGDT-MFP) is proposed. The mouse was selected as the application object of this technology. Firstly, the local appearance features of the mouse were defined. Through Gray Relation Analysis (GRA), the correlation mechanism between local features and geometric parameters was established, which realized the design space's dimensionality reduction and the design's Simplification. Secondly, by improving the sampling method of the design space, the Latin Hypercube Sampling (LHS) method in the multi-parameter dimension, called LHSNM, was proposed, which can efficiently explore the design space and achieve the diversity of design solutions. Then the rule for interactive design was developed. By exploring the design space with a purposeful approach, the invalid sampling process was avoided, and the design efficiency was greatly improved. Lastly, an interactive design interface was built so that even non-professionals with no design experience can complete generative design of the diversity of mouse appearances and get a satisfactory solution. After testing and analysis, IGDT-MFP, as an interactive generative design method, shows high user-friendliness and design efficiency and can generate various satisfactory solutions for designers.