dynamic_sac_path

When arbitrary starting point and goal point are given, TD3 agent computes the optimal collision-free path.

Learning is done using Tensorflow and testing (path generation and manipulator operation) is done in Matlab.

When arbitrary starting point and goal point are given, TD3 agent computes the optimal collision-free path.

Learning is done using Tensorflow and testing (path generation and manipulator operation) is done in Gazebo in ROS and the

open manipulator by Robotis is used for experiment. It is verified that the proposed method indeed generates

collision-free path for the real manipulator.