How to create an custom Reinforcement Learning Environment + DDPG agent from ROS environment

Hi,

I´m working on an Agent for a control problem with wam-v (wam-v is a catamaran boat with sensors and cotroller that allow me to make it autonomous). I already wrote a code to control the wam-v in ROS GAZEBO environment which simulate all the forces, waves and mass for the wam-v. Currently, I want to use the matlab RL toolbox to build a DDPG agent to create the autonomy. How can I use the functions that I already wrote as input and output to build a custom matlab environment to train the agent?

[Asvin Kumar]

I don't know these toolboxes well enough to answer your question.

If you haven't seen these pages, they might give you a start.

https://www.mathworks.com/help/reinforcement-learning/matlab-environments.html

https://www.mathworks.com/help/reinforcement-learning/simulink-environments.html