In order to address the integrated scheduling problem of AGVs (automated guided vehicles) and machines with considering path conlict in manufacturing system, an improved discrete whale optimization algorithm based on time window and Dijkstra algorithm was proposed. First, with the goal of minimizing the maximum completion time, a mathematical model of AGV-and-machine integrated scheduling was established. Then, a three-stage coding was used to realize the integrated coding of AGVs and machine, and a continuous space and discrete space were established. Second, in order to ensure the quality and diversity of the initial population, an extended GLR population initialization method combining chaotic mapping and opposition learning was designed. Then, the Levy flight operator and threshold restart operation were used to further improve the algorithm's global search capability. Finally, in order to improve the local search ability of the algorithm, a variable neighborhood search algorithm combined with the features of the problem was introduced. Standard simulation experiments and flexible simulation experiments have proved the feasibility and superiority of the proposed algorithm to solve the problem of AGV-and-machine integrated scheduling.