A* Algorithm-Based AGVs Path Planning for Intelligent Warehousing

With the advancement of the Internet of Things (IoT) and artificial intelligence (AI), manufacturing enterprises are rapidly transitioning to intelligent production. Automated Guided Vehicles (AGVs), characterized by flexible deployment, as well as autonomous navigation and loading, have been widely adopted in intelligent, logistics, healthcare, and aviation. Numerous methods, such as the Dijkstra algorithm, A* algorithm, and their variants, have been applied to enable the autonomous navigation of AGVs. However, most state-of-the-art AGV path planning approaches struggle to efficiently find an optimal and smooth route for an AGV. In this study, we propose a modified A* algorithm for AGV path planning in warehousing environments. Specifically, we integrate an adaptive heuristic reward strategy and the path smoothing technique into the classical A* algorithm to derive an optimal and feasible path for an AGV. Furthermore, we introduce a dual-neighborhood dynamic expansion strategy to improve the search efficiency and accelerate the identification of the optimal navigation path. Extensive simulation results demonstrate that the improved A* algorithm significantly reduces both the number of expanded nodes and path lengths. In intelligent warehousing environments, compared to the 16-neighborhood search A* algorithm, the proposed method reduces the number of expanded nodes by 34% and achieves approximately 13.5% shorter path lengths.