面向无人机近距离空战问题的高模仿性强化学习控制方法

本算法在Harfang3D Dog-Fight Sandbox 强化学习环境中表现优秀，整合后的代码放在paper分支下，本分支为老版本

依赖

安装Harfang3D sandbox的Release版本或源代码，推荐安装源代码版本，这样可以自行更改环境的端口
安装本代码所需依赖，可以在官网requirements.txt的基础上补充安装，也可以通过以下命令直接安装
```
pip install -r requirements.txt
```

飞行规则自定

Name		Name	Last commit message	Last commit date
Latest commit History 44 Commits
__pycache__		__pycache__
pictures		pictures
.gitignore		.gitignore
BCNeuralNetwork.py		BCNeuralNetwork.py
Constants.py		Constants.py
DemoEnv.py		DemoEnv.py
DemoEnvAi.py		DemoEnvAi.py
DemoEnvAi_new.py		DemoEnvAi_new.py
HarfangEnv_GYM.py		HarfangEnv_GYM.py
NNs.zip		NNs.zip
NeuralNetwork.py		NeuralNetwork.py
README.md		README.md
ROTNeuralNetwork.py		ROTNeuralNetwork.py
ROTNeuralNetwork2.py		ROTNeuralNetwork2.py
ROTNeuralNetwork3.py		ROTNeuralNetwork3.py
ReplayMemory.py		ReplayMemory.py
Train.py		Train.py
TrainBC.py		TrainBC.py
TrainROT.py		TrainROT.py
TrainROT2.py		TrainROT2.py
TrainROT3.py		TrainROT3.py
actor.py		actor.py
demo.py		demo.py
demo_AI.py		demo_AI.py
dogfight_client.py		dogfight_client.py
expert_data_ai.csv		expert_data_ai.csv
plot.py		plot.py
read_data.py		read_data.py
requirements.txt		requirements.txt
socket_lib.py		socket_lib.py
test.py		test.py