基于BP神经网络的木制民居个性化定制

Personalization of wooden houses based on BP neural network

为解决木制民居定制过程复杂、难以融入用户感性需求的问题,提出一种基于BP神经网络及模块化设计的个性化定制方法。利用语义量表获取用户的感性意象,基于模块化设计思想对木制民居进行解构并编码,通过BP神经网络构建用户感性意象与造型模块之间的映射关系,根据设计目标计算并查找相应的设计编码组合,指导木制民居设计。该方法可提高木制民居个性化定制的效率,实现木制民居设计中用户感性意象的融入和应用。以苗族传统木制民居为实例,验证了该方法的可行性。

To solve the problem that the customization process of wooden dwellings is complex and difficult to integrate into users’emotional needs,a personalized customization method based on BP neural network and modular design was proposed.The kansei image of users was obtained by using the semantic scale,the wooden dwellings were deconstructed and coded based on the modular design idea,the mapping relationship between the perceptual image of users and the modeling module was constructed via BP neural network,and the design code combination was calculated and searched according to the design target and guide the design of wooden dwellings.This process improved the efficiency of individualized customization of wooden dwelling houses,and the application of users’kansei image in the design of wooden dwelling houses was realized.With the example of the traditional wooden dwelling houses of Miao nationality,the feasibility of this method is verified.