一种面向维修资源配送调度的遗传-烟花混合算法

PDF下载HTML

Title:

A genetic-firework hybrid algorithm oriented to maintenance resource distribution scheduling

作者:

李猛¹, 和伟辉², 毛攀登¹, 齐小刚³, 刘立芳¹

1. 西安电子科技大学 计算机学院, 陕西 西安 710071;
2. 西安卫星测控中心, 陕西 西安 710049;
3. 西安电子科技大学 数学与统计学院, 陕西 西安 710071

Author(s):

LI Meng¹, HE Weihui², MAO Pandeng¹, QI Xiaogang³, LIU Lifang¹

1. School of Computer Science and Technology, Xidian University, Xi’an 710071, China;
2. Xi’an Satellite Control Center, Xi’an 710049, China;
3. School of Mathematics and Statistics, Xidian University, Xi’an 710071, China

关键词:

维修资源; 配送调度; 遗传算法; 烟花算法; 车辆路径; 多配送中心; 资源调度; 时间窗口

Keywords:

maintenance resources; distribution scheduling; genetic algorithm; fireworks algorithm; VRP; multi-distribution center; resource scheduling; time window

分类号:

TP391

DOI:

10.11992/tis.202108027

摘要:

为减轻资源供应不及时对维修活动顺利开展的影响，本文针对配送式供应保障，基于带时间窗的多配送中心车辆路径规划问题提出了一种半开放式的协同配送调度模型，使得多个资源库存中心之间达成了协同合作与互相保障，从而减少了资源的供应时长和调度成本，提高了全局调度效率。为高效地求解该模型，本文提出了一种遗传-烟花混合算法，混合算法在经典遗传算法的基础上引入了烟花算法的爆炸算子以增加种群优秀个体的数量，丰富种群基因的多样性，从而提高算法的寻优能力。通过仿真实验对比，证明了爆炸算子对遗传算法容易“早熟”的缺点有所改善，且混合算法具有更高的求解效率。

Abstract:

In order to alleviate the impact of belated supply of maintenance resources on smooth maintenance activities, in this paper, a semi-open collaborative resource distribution scheduling model is established based on the multiple distribution centers’ vehicle routing problem (VRP) with time-windows to achieve collaboration and mutual guarantee among multiple depots. In this way, the delivery time and the scheduling cost are reduced and the overall resource supply efficiency is improved. In order to efficiently solve the model, this paper proposes a genetic-firework hybrid algorithm, which introduces the explosion operator of the firework algorithm into the classical genetic algorithm to increase the amount of good individuals and the diversity of the population, so as to improve its optimization capacity. And through comparison of simulative experiments, it is proved that the hybrid algorithm has higher solving efficiency and that the introduction of explosion operators can solve the problem that genetic algorithm is easy to converge.