Abstract: In the actual industrial production process,due to the limited resources of Automate Guided Vehicles (AGVs), the integrated problem FJSP-AGV comsidering the constraint of alimited number of AGVs in the Flexible Job-shop Scheduling Problem (FJSP) has significant research value. Traditional evolutionary algorithms are easy to fall into local optimum and are not suitable for solving this scheduling problem with high complexity.In light of the aforementioned challenges, it initially established a mathematical model for FJSP-AGV and subsequently proposed an improved genetic algorithm guided by heuristic rules. The algorithm utilized various crossover and mutation methods to evolve the population for different coding segments.Simultaneously,it adjusted parameters adaptively during the evolutionary process and guided mutations through heuristic rules for local search,thereby enhancing the algorithm′s capability to escape local optima and consequently minimize the maximum completion time of the system. Comparison and analysis with other advanced algorithms on two small and medium-sized datasets demonstrated that the algorithm proposed yielded the most comprehensive solving effect.

Key words: Flexible Job-shop Scheduling Problem (FJSP), Automated Guided Vehicles (AGV), vehicle scheduling, Genetic Algorithms (GA), heuristic rules