Abstract:Aiming at the roadway congestion problem of multiple AGVs (automatic guided vehicles) in the automated storage and retrieval system, a system optimization strategy to avoid congestion was proposed, and the generated AGV picking path was used as a constraint to generate subsequent AGV running trajectories. Adjacent nodes of the warehouse was assigned time links to build a space-time network map, and a path optimization model was established based on discrete space-time networks in this environment, which considered congestion. A global optimization algorithm ST-SA(space time simulated annealing) was designed that combined space-time networks and SA (simulated annealing) to solve the model, and through simulation experiments, the effectiveness of the model and algorithm was proved. The experimental results show that the system optimization strategy can control and optimize the AGV path planning process. ST-SA can quickly search for a reasonable and efficient AGV picking path plan, reduce the AGV’s working time in the roadway and avoid multiple AGVs collision and congestion in the automated storage and retrieval system.