A Study on Applying Feature-based Modeling and Neural Network to Shape Generation

• Main Authors: Chung-Chih Lin, 林忠志
• Other Authors: Shih-Wen Hsiao
• Format: Others
• Language: zh-TW
• Published: 2003
• Online Access: http://ndltd.ncl.edu.tw/handle/73817607739353958996

碩士 === 國立成功大學 === 工業設計學系碩博士班 === 91 === Recently, due to the keen competition of market, how to push a successful product fitting consumers’ desires has become the motivation for enterprises to keep their healthy growths. However, the key of a new successful product depends on industrial designers’ abilities. In the past, many designers relied on their experiences to design a new product, but they couldn’t grasp consumers’ desires effectively in this way. And it’s easy to turn into a risk for enterprises or designers. A feature-based modeling method for shape design is proposed in this article. In this study, we take mobile phone form design as an example to show the design procedures. Firstly, we use feature-based modeling to construct a feature framework and build 3D models. Then, semantic differential method (SD) is used to quantify the relationships between consumers’ image words and morphing prototypes. Furthermore, we apply back-propagation neural network (BPN) to establish the relationships between product form parameters and adjective image words. Secondly, a database between CAD software and consumers’ image words is built to provide designers for grasping consumers’ desires quickly. To penetrate this database, designers can start CAD software to modify the feature parameters to generate new shapes. Though mobile phone form design is taken as an example in this study, this method can also be used to develop other similar products.