This is a gym reinforcement learning environment. The game environment I build is from CUMC2020 problem B, with the DQN arthimetic we apply on that when triing to solove the problem. I finished it on Windows10.
Before you take it into use, please ensure you have what you need such as keras, gym and some basal labs. And please take this into register in gym.
You should make lots of changes to run the environment and don't have lot's of hints these days but I will ensure you can run it easily in two weeks. So this is a initial version which I used on CUMC2020