Path planning in mobile robots with Temporal Difference methods, Q Learning and SARSA.
The following video shows the agent iterating in the environment until it converges onto the optimal path https://user-images.githubusercontent.com/79185485/197855234-d59dae3b-bb08-4aee-8397-43119bb09253.mp4
The Optimal Path
The Q-Learning script converges in 14 minutes, SARSA script converges in 63 minutes