A Two-Level Path Planning Approach Combining Improved ACO and DWA for Mobile Robots in Warehousing Environments

Abstract

With the continuous improvement of the accuracy and efficiency requirements for robot path planning in intelligent warehousing systems, enhancing the ability of path planning in complex dynamic environments has become a key issue. Therefore, a two-level path planning model integrating an improved Ant Colony Optimization algorithm with a Dynamic Window Approach is proposed. At the global level, a multi-objective fitness function that considers path length, corner smoothness, and obstacle clearance is introduced to guide the search, and elite solutions are strengthened in each iteration by increasing the pheromone reinforcement on high-quality paths. At the local level, a velocity-based DWA module dynamically samples feasible trajectories in velocity space and evaluates them using weighted functions of heading, speed, and obstacle distance to generate optimal control actions. Experimental results show that the proposed model improves accuracy to 88.4%, reduces root mean square error to 0.21, achieves a smoothness score of 0.94, and completes tasks with a success rate of 97.3%. The model effectively enhances global optimality and real-time local obstacle avoidance, making it suitable for complex warehousing and logistics scenarios.