Robot path planning is the basis for robot to perform various tasks successfully. Dynamic window approach (DWA) is a flexible way to solve this task, however its evaluation functions are not comprehensive enough and the weight of these functions will not be adjusted with the change of environment in real time, which makes the robot tend to circumnavigate the outside of the dense obstacle area, and it is easy to fall into the failure of the evaluation function when it meets the combination of “C” shaped obstacles. In this paper, an improved DWA based on Q-learning is proposed. Firstly, we add a new evaluation function to prevent the robot from circling in place. Then, on the basis of defining the state space, action space and reward function in the robot motion planning, Q-learning is used to adjust the weight of the evaluation function adaptively, so that it can adapt to the changing environment. Finally, the simulation results show that the new evaluation function not only solves the above problems, but also improves the overall motion efficiency of mobile robots.
Q-Learning Based Robot Path Planning with Improved Dynamic Window Approach
Lect. Notes Electrical Eng.
International Conference on Autonomous Unmanned Systems ; 2021 ; Changsha, China September 24, 2021 - September 26, 2021
Proceedings of 2021 International Conference on Autonomous Unmanned Systems (ICAUS 2021) ; Kapitel : 234 ; 2364-2375
18.03.2022
12 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
A Fuzzy-Based Improved Dynamic Window Approach for Path Planning of Mobile Robot
Springer Verlag | 2023
|Mobile Robot Path Planning Based on Improved Ant Colony Fusion Dynamic Window Approach
British Library Conference Proceedings | 2021
|Global Dynamic Path Planning Based on Fusion of Improved A^* Algorithm and Dynamic Window Approach
British Library Online Contents | 2017
|