摘要
针对消费者对配色结果表述相对模糊的问题,提出一种适用于求解复杂高维优化问题的分层蜂群优化算法,并以女性服饰为例,将该算法用于色彩设计模型进行计算机辅助配色。首先采用BP神经网络构建了女性裙装色彩设计多目标评价模型,将调查问卷结果的量化值同服装色彩选配值一起作为BP神经网络的输入,将服装色彩设计的诸多评价目标作为BP神经网络的输出,经训练建立了服装色彩设计与女性多目标评价对应的评价模型。采用多目标多蜂群优化算法对该模型进行优化,以获得诸多评价结果较好的配色组合。通过设计实例证明了所提算法对配色的实用性。
To solve the problem of relative fuzziness of color matching results, a hierarchical bee colony optimization algorithm was proposed, which was suitable for solving complex high dimensional optimization problems. By taking the women dress as an example, the algorithm was used for clothing color design model. The multi-objective evalua- tion model of women on dress color design was constructed with BP neural network, the quantitative values of ques- tionnaire results were combined with the clothing color matching value as input to BP neural network, and many e- valuation objects of clothing color design were used as the output of BP neural network. After training, the evalua- tion model of clothing color design and female multi-target evaluation was established. To obtain a good combination of color sets of many evaluation results, the multi-objective hybrid artificial bee colony optimization algorithm was used to optimize the model. The proposed method was also proved to have practicability on color design and aid de- signer effectively.
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2018年第2期381-389,共9页
Computer Integrated Manufacturing Systems
基金
国家自然科学基金资助项目(61703306)
天津市自然科学基金资助项目(16JCQNJC00600)
天津师范大学博士基金资助项目(52XB1002)~~
关键词
智能计算
蜂群算法
神经网络
计算机辅助配色
产品设计
intelligent algorithm
bee colony algorithm
neural network
computer aided color design
product design
