Optimized Path Planning for Autonomous Guided Vehicle Through Fusion of Improved A* and Dynamic Window Approach

This paper addresses the issues of slow speed, safety concerns, and non-smooth paths in traditional A* algorithm-based path planning for Autonomous Guided Vehicle (AGV) in complex environments. The study proposes an improved fusion algorithm combining the A* and Dynamic Window Approach (DWA). The enhanced A* algorithm first plans the global path, and then the DWA algorithm handles local path planning under the guidance of the global path. First, by introducing the concept of obstacle density, the study integrates environmental information, optimizes the evaluation function, refines the child node selection process, and applies bidirectional smoothness optimization to the paths. These improvements enhance search efficiency, ensure adequate safety distances from obstacles, and strengthen the safety margin and rationality of the paths, making them more aligned with the robot’s needs. Next, the integration of the DWA algorithm allows for dynamic obstacle identification and local path regeneration for avoidance, resulting in more realistic paths for dynamic obstacle navigation. Finally, key nodes from the global path planning of the A* algorithm are extracted to serve as references for local planning in the DWA algorithm, ensuring that local paths are more closely aligned with global paths. Simulations and practical experiments confirm the effectiveness of the proposed algorithm, providing a more efficient and feasible solution to the path planning problem.