Computer Science · Computer Vision and Pattern Recognition
A Fusion Algorithm Integrating Hybrid a-Star and Dynamic Window Approach for Enhanced Mobile Robot Path Planning
Zhewei Guo, Yunfeng Yang, Shi-Lu Dai
See on the mapAbstract
In this paper, we present a comprehensive path planning framework that integrates an advanced Hybrid A-star algorithm with the dynamic window approach (DWA). This integration aims to enhance the efficiency and real-time obstacle avoidance capabilities of mobile robots in complex environments. The enhanced Hybrid A-star algorithm reduces unnecessary nodes and improves path smoothness by leveraging historical node information and dynamically adjusting heuristic weights. Meanwhile, the DWA algorithm's obstacle avoidance is strengthened in environments with unknown obstacles through real-time fuzzy logic control of the evaluation function weights. Simulation experiments reveal improved performance in computation time and path node count while maintaining comparable path lengths for the advanced Hybrid A-star algorithm. The fusion algorithm effectively leverages global path information and dynamically adapts to unknown obstacles, outperforming single-planning algorithms. Physical experiments further demonstrate the feasibility and effectiveness of the proposed approach.