An effective population-based iterated greedy algorithm for solving the multi-AGV scheduling problem with unloading safety detection

Abstract

With the increasing prevalence of automated guided vehicles (AGVs), the multi-AGV scheduling problem has become a hot research topic in recent years. However, little attention has been devoted to unloading safety detection for multi-AGV scheduling. This paper investigates a multi-AGV scheduling problem with unloading safety detection (MAGVSUSD) in a matrix manufacturing workshop. The aim is to minimize the total cost composed of travel cost, penalty cost and cost of AGV used. To address the MAGVSUSD, a mixed-integer linear programming model and a population-based iterated greedy (PIG) algorithm are proposed. In the PIG, a hyper-heuristic based on neighborhood operators and a population-based initialization method are proposed to obtain multiple high-quality initial solutions. The two-stage destruction strategy is designed to avoid falling into the local optimal solution and to enhance its ability to explore a larger solution space. The two-choice reconstruction strategy is developed to obtain better neighbor solutions by switching the search landscape. The local search based on the winning neighbor operators is proposed to dynamically adapt to each distribution situation in the workshop. A large number of experimental results demonstrate the proposed algorithm's superiority over existing algorithms in addressing the considered problem.