CAJ | 학술논문

为提高生鲜农产品物流配送效率,提出了具有普遍通用性的生鲜农产品三层配送网络结构,建立了配送成本最小和顾客满意度最大的多目标数学模型,引入惩罚函数处理约束,并对遗传算法改进,在选择之前根据非劣解水平进行排序,并利用拥挤程度对同级的不同个体排序,然后利用精英保留策略、最大保留交叉等操作获得Pareto最优解。为验证算法的性能,首先与标准遗传算法的结果对比,同时对相关参数进行设置,得到的结果变化误差较小。结果表明：建立的模型与算法是有效的,可为实际应用提供指导。
위제고생선농산품물류배송효솔,제출료구유보편통용성적생선농산품삼층배송망락결구,건립료배송성본최소화고객만의도최대적다목표수학모형,인입징벌함수처리약속,병대유전산법개진,재선택지전근거비렬해수평진행배서,병이용옹제정도대동급적불동개체배서,연후이용정영보류책략、최대보류교차등조작획득Pareto최우해。위험증산법적성능,수선여표준유전산법적결과대비,동시대상관삼수진행설치,득도적결과변화오차교소。결과표명：건립적모형여산법시유효적,가위실제응용제공지도。
To improve the efficiency of fresh agricultural product logistics distribution, a general three-layer fresh agricultural products distribution network is proposed, and a multi-objective mathematical model with the distribution cost minimum and customer satisfaction maximum established. Penalty function is introduced to deal with constraints and GA put forward to solve this question, according to the Pareto solution level and the level of congestion, sorting the same level of different individuals before choosing operation. Then the elite reserved strategy and maximum preserved crossover operation are used to obtain the Pareto optimal solutions. To verify the performance of algorithm, firstly, it is compared with the results of the standard genetic algorithm. At the same time, it sets different parameters and the result error is reduced. The results show that the established model and algorithm are effective and can provide guidance for practical application.