MATLAB Answers

How to set the reinforcement learning block in Simulink to output 9 actions

5 ビュー (過去 30 日間)
Aaron Amusan
Aaron Amusan 2021 年 5 月 15 日
I am trying to tune 3 PID controllers in matlab/simulink via reinforcement learning thus the reinforcemt learning toolbox. I am trying to follow "Tune PI Controller using Reinforcement Learning" (https://www.mathworks.com/help/reinforcement-learning/ug/tune-pi-controller-using-td3.html) as best I can and extrapolate it to create three PID controllers controlled by a DDPG agent. i can't really figure out how to make it account for three PID controllers, hence nine gain values.
I can't really share much about the model since I am bound to a NDA by my university, but I feel as though my question is nonspecifc enough to both not get me in trouble, but also allow potential helpers to get what I am trying to do. Please let me know if more info is needed.

回答 (2 件)

Yikai
Yikai 2021 年 5 月 17 日
the number of the actions is defined at actionInfo of your environment

Emmanouil Tzorakoleftherakis
Emmanouil Tzorakoleftherakis 2021 年 5 月 17 日
Hello,
the example you are referring to does not output 3 values for the pid gains. The PID gains are "integrated" into the neural network architecture and the policy output is still the same as PID. If you want to follow the same setup, the output of the policy in your case would output 3 values and the neural network weights would match the weights of the PID controllers.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by