面向维修资源分配调度的遗传-长鼻浣熊混合优化算法

PDF下载HTML

Title:

Hybrid genetic long-nosed raccoon optimization algorithm for maintenance resource allocation and scheduling

作者:

秦敏敏¹, 刘立芳¹, 齐小刚^2,3

1. 西安电子科技大学 计算机科学与技术学院, 陕西 西安 710071;
2. 西安电子科技大学 数学与统计学院, 陕西 西安 710071;
3. 西安市网络建模与资源调度重点实验室, 陕西 西安 710071

Author(s):

QIN Minmin¹, LIU Lifang¹, QI Xiaogang^2,3

1. School of Computer Science and Technology, Xidian University, Xi’an 710071, China;
2. School of Mathematics and Statistics, Xidian University, Xi’an 710071, China;
3. Xi’an Key Laboratory of network modeling and resource scheduling, Xi’an 710071, China

关键词:

资源受限; 项目调度; 多模式; 资源配置; 分配调度; 动态发布; 多维修中心

Keywords:

resource constraints; project scheduling; multimode; resource allocation; allocation scheduling; dynamic publishing; multiple maintenance centers

分类号:

TP391

DOI:

10.11992/tis.202303035

摘要:

鉴于传统的资源受限的项目调度问题（resource-constrained project scheduling problem，RCPSP）已经难以满足当下实际需求，对资源受限的项目调度问题进行扩展已是大势所趋，所以本文结合设备动态发布维修任务的特性，对原RCPSP问题进行抽象，加入了与设备相关的多模式的资源配置问题，从而建立了面向多维修中心的多模式的动态资源分配调度模型。为了更好地求解所提出的模型，本文提出了一种遗传-长鼻浣熊混合优化算法，该算法是在原长鼻浣熊优化算法的基础之上加入了遗传算法的选择、交叉以及变异算子，主要用于扩大搜索范围，从而跳出局部最优；为了进一步提高候选解的质量，还加入了贪婪算子的操作。通过对仿真实验结果的对比分析，发现不论是从收敛速度还是求解质量等方面，新提出的遗传-长鼻浣熊混合优化算法均以绝对的优势优于其他算法。

Abstract:

Traditional resource-constrained project scheduling problems (RCPSPs) can hardly meet current practical needs. Hence, expanding the RCPSP is an inevitable trend. Therefore, this paper abstracts the original RCPSP problem based on the characteristics of dynamically releasing maintenance tasks for equipment, adding multimodal resource allocation issues related to equipment. Thus, a multimode dynamic resource allocation and scheduling model for multiple maintenance centers is established. This paper proposes a hybrid optimization algorithm combining genetic and long-nosed raccoon algorithms to solve the proposed model effectively. The algorithm is added with the selection, crossover, and mutation operators of the genetic algorithm based on the original coati optimization algorithm, which are mainly used to expand the search range, thereby jumping out of local optimization. Furthermore, greedy operator functions are added to further improve the quality of candidate solutions. Comparative analysis of the results of simulative experiments revealed that the newly proposed genetic long-nosed raccoon hybrid optimization algorithm is superior to other algorithms in terms of convergence speed and solution quality.