摘要
为提高食谱设计质量与效率,提出一种基于交互式进化计算的食谱智能优化方法。根据用户评价值和食谱菜品优化模型确定食谱样本隐式指标与显式指标;基于NSGA-Ⅱ算法获得食谱样本Pareto优化解。为使Pareto优化解更好兼顾用户偏好与营养均衡,引入指标均衡度指导进化。当指标均衡度较低时,采用遗传算法模块对显式指标单独进化,提高指标均衡性。用遗传算法最优解替换原优化结果中指标均衡度最低个体,改造Pareto前沿。以传统交互式遗传算法和食物交换法为对比方法,验证了该方法在个性化适应性、有效性和可用性等3个方面均优于对比方法。该方法为个性化食谱设计提供了一种智能化新思路。
To improve the design quality and the efficiency of personalized recipe,an intelligent optimization method based on interactive evolutionary computation(IEC)was proposed.The qualitative and quantitative indexes were confirmed by user evaluation and recipe optimization model.The Pareto optimal solutions for recipes were obtained based on NSGA-Ⅱ.To make Pareto optimal solution consider both user preference and balanced nutrition,an index balance degree was introduced to guide evolution.The quantitative index was optimized by GA when the population index balance degree was low,and the individual with the lowest index balance degree was replaced by the optimal solution of GA.Compared with traditional interactive genetic algorithm(IGA)and food exchange method,the results indicate that this method is superior to comparison method in adaptability,validity and usability.This method provides a new intelligent way for personalized recipe design.
作者
郭广颂
席俊杰
文振华
GUO Guang-song;XI Jun-jie;WEN Zhen-hua(School of Intelligent Engineering,Zhengzhou University of Aeronautics,Zhengzhou 450046,China;School of Aeronautical Engineering,Zhengzhou University of Aeronautics,Zhengzhou 450046,China)
出处
《计算机工程与设计》
北大核心
2021年第4期1143-1150,共8页
Computer Engineering and Design
基金
国家自然科学基金项目(51975539)
河南省科技开放合作计划基金项目(182106000019)
河南省重点研发与推广专项基金项目(212102210491)。
关键词
进化算法
食谱
交互
混合指标
优化
evolutionary computation(EC)
recipe
interactive
hybrid index
optimization
