How to export trained RL agent to TwinCat machine learning inference

6 ビュー (過去 30 日間)
chetan thakur
chetan thakur 2022 年 12 月 1 日
コメント済み: yu ZHANG 2022 年 12 月 6 日
I trained a agent using reinforcement learning in MATLAB. I wish to deploy this to real hardware.
I am new to deploying trained agents and do not have clue here.
I plan to use Beckhoff IPC system.
Can you please guide me on this.
It is okay if it can be done without Beckhoff IPC.
Thanks in advance.


Dimitri MANKOV
Dimitri MANKOV 2022 年 12 月 1 日
Hi Chetan,
In case you'd like to deploy your algorithm on a real-time target machine that specifically uses Simulink Real-Time in R2021a, you will need Speedgoat hardware. Amongst other protocols, Speedgoat offer EtherCAT Master / Slave modules to exchange data with external devices.
If you don't plan to use Simulink Real-Time, you will need to find hardware compatible with the toolchain that your project will be based on.
I hope this is helpful!
  1 件のコメント
yu ZHANG 2022 年 12 月 6 日
Can i use rl_agent to generate HDL coder?
i meet some error.but i cant write this block;


その他の回答 (0 件)


Find more on Simulink Real-Time in Help Center and File Exchange

Community Treasure Hunt

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

Start Hunting!

Translated by