An effective multi-AGVs dispatching method applied to matrix manufacturing workshop

Xu-jin Zhang,Hong-yan Sang,Jun-qing Li,Yu-yan Han,Peng Duan

DOI: https://doi.org/10.1016/j.cie.2021.107791

2022-01-01

Abstract:This paper studies the problem of multiple automatic guided vehicles (multi-AGVs) dispatching in a matrix manufacturing workshop. The goal is to minimize the transportation cost that includes the cost of travelling distance, the cost of penalty time and the cost of AGVs. For the purpose, a mixed integer linear programming model is set up and an improved iterated greedy (IIG) algorithm is proposed. In the algorithm, an AGV route merging strategy and a workshop partition strategy are designed to reduce the cost of AGVs and travelling distance. Two rules are designed to quickly identify infeasible solutions to save the operation time. A nearest neighbor heuristic is used to generate an initial solution with high quality. In the local search stage, four effective operators are used to improve the quality of the solution. A repair strategy is proposed to avoid the algorithm falling into local optima. Finally, we use 110 real instances to test the IIG and the other six algorithms in the literature. The comparative experiments show that the proposed algorithm and strategies have much better performance for solving this problem.

computer science, interdisciplinary applications,engineering, industrial