Finshed DQN-2D Holonomic Arquitecture
Index
Test model
Training plot:
A program has been created that simulates the scenario 100 times and calculates an approximate average success rate. Each time the parking is successfully performed, the episode is terminated.
After several tests it has been determined that the trained model has a reliability between 80-90%, 33 random positions/orientations are used.
This post is licensed under CC BY 4.0 by the author.