Comparison between LQR and DQN for Cartpole
Comparison between LQR and DQN for Cartpole
The Project objective is to apply the topic of the Advanced Control System I (MEN573) course and study Reinforcement Learning.
The Project objective is to apply the topic of the Advanced Control System I (MEN573) course and study Reinforcement Learning.
This project is implemented based on what was learned in the course MEN573 Advanced Control System I. (Course Grade: A+) / This Project is implemented in the UNIST Robotics and Mobility Lab.
This project is implemented based on what was learned in the course MEN573 Advanced Control System I. (Course Grade: A+) / This Project is implemented in the UNIST Robotics and Mobility Lab.
The Basic Concept of LQR Controller (State Feedback Controller)
The method of design LQR Controller
The parameter of LQR Controller
First, I implemented the LQR Controller for Cartpole using python.
First, I implemented the LQR Controller for Cartpole using python.
The system output without controller
The system output with LQR Controller
The Basic Concept of DQN
Parameter of DQN
Second, I implemented the DQN for Cartpole using python.
Second, I implemented the DQN for Cartpole using python.
The reward according to the episode
The system output after learning
If you want to get some material, then you should access the google drive URL.
If you want to get some material, then you should access the google drive URL.
KAIST
291, Daehak-ro, Yuseong-gu, Daejeon, Republic of Korea