This is a simple demo using the DQN reinforcement learning algorithm for semi-active suspension control.
In the future, we will make improvements in the following places:
(1) Representation method of suspension state
(2) Reward function design
(3) Algorithm design
(4) Action design
