基于BP神经网络的旗袍造型设计方法与应用

Modeling Design Method and Application of Cheongsam Based on BP Neural Network

为解决旗袍造型设计过程复杂、消费需求难以满足的问题,文章基于感性工学及模块化设计理念对旗袍造型进行解构与编码,从领型、襟型、袖型、开衩及摆型进行解构,并筛选了10对感性评价词。通过调查问卷量化用户的感性意象,使用因子分析、BP神经网络等探究用户感性意向与旗袍造型之间的数量关系,根据预测目标模拟出相应的设计编码,应用于针对性的旗袍造型设计。结果表明,该方法有效提高了旗袍造型的设计效率,实现旗袍设计中用户感性意象的融入与应用,并减少了测量用户感性意象的样本量。以该方法完成了高贵风格旗袍的设计方案,验证了方法的可行性。

In order to solve the problem that the design process of cheongsam is complex and the consumer demand is difficult to meet,this paper deconstructs and codes the cheongsam shape based on Kanesei engineering and modular design concept,deconstructing it from the collar type,skirt type,sleeve type,slit and pendulum type,and screening 10 pairs of perceptual evaluation words.The questionnaire was used to quantify the user’s perceptual image,factor analysis and BP neural network were used to explore the quantitative relationship between the user’s perceptual intention and the chi-pao shape,and the corresponding design code was simulated according to the prediction target and applied to the targeted cheongsam shape design.The results show that this method can effectively improve the design efficiency of cheongsam modeling,realize the integration and application of user’s perceptual image in cheongsam design,and reduce the sample size of measuring user’s perceptual image.This method is used to complete the design of the noble style cheongsam,and the feasibility of the method is verified.