air_combat

• Project Description: A series of air combat game environments packaged according to the gym interface for reinforcement learning.

• Project Instruction: You can use any file(airwar.py、airwar_3D.py) as a standard gym environment, each with integrated class methods such as step(), reset(), and render(). In order to facilitate the training of the algorithm, the environment also provides functions such as get_actor_observation(), get_critic_observation(), and get_avail_agent_actions(), so you do not need to process additional environment information.

• If you want to see the effect of the environment directly, you can directly run the test.py file, which can basically show all the action effects.

If you have any questions or suggestions, you can leave a message to my email - 597396936@qq.com.