Legged Robots Simulation using Deep Reinforcement Learning

Date 

The project focuses on two methods for quadruped locomotion. In the first part, different gaits with Central Pattern Generators are implemented. In the second one, the Markov Decision Process (MDP) is designed to be used by state-of-the-art deep reinforcement learning algorithms, for instance PPO and SAC, for the robot to learn to walk on different terrains by itself.