Path planning for orchard mobile robots based on an improved ant colony algorithm and the dynamic window approach

Abstract

To address the challenges of insufficient navigation accuracy, low real-time performance, non-smooth paths, and excessive turning points in orchard mobile robots, this study proposes an integrated path planning method combining an improved Ant Colony Optimization (ACO) algorithm and the Dynamic Window Approach (DWA), supported by LiDAR-based Simultaneous Localization and Mapping (SLAM) for map construction. Firstly, the heuristic function, pheromone update strategy, and redundant node removal are optimized to enhance the efficiency of global path planning. Subsequently, the improved ACO is integrated with DWA to achieve local dynamic obstacle avoidance and further improve path smoothness. Finally, comparative experiments are conducted in both static and dynamic orchard environments among the Sparrow Search Algorithm (SSA), the A* algorithm, conventional ACO, and the proposed approach. Simulation results demonstrate that the proposed method reduces the number of turning points and path length by up to 4.29%, and decreases runtime by up to 12.95%.