Drone auto-navigation stimulation developed in Pybullet OpenAI Gym Environment. Trust Region Policy Optimization (TRPO) and Proximal Policy Optimization (PPO) is implemented.

• You can run & see the real time successful rate by:

python run.py --model PPO --version 0  

• To train the model, you can run:

python PPO_trainer.py

or

python TRPO_trainer.py

• You can modify the GUI in drone_envs/config "display": p.GUI,

Thanks Medium series on creating OpenAI Gym Environments with PyBullet.