Reinforcement learning/Experiecne buffer/Simulink

I am trying to create a experience buffer for my DDPG algorithm in Simulink. However, I can't find anyway or blocks to help me create a experience buffer to store (state,action,reward,next state) in simulink.

I have tried to create a experience buffer by rl.util.ExperienceBuffer in Matlab function but here is my code and the error

I hope you can help me deal with this problem. Thank you verry much!

Answer 1

[Emmanouil Tzorakoleftherakis]

Why do you want to create your own buffer? If you are using the built-in DDPG agent, the buffer is created automatically for you. In any case, in 22a we added a feature that allows you to create your own experience buffer (see here). You can potentially use this to manually modify the experience buffer of a built-in agent.

[hieu nguyen]

Actually, I want create my own DDPG agent with my own neural network structure and my own optimizer algorithm in Simulink. I have tried to do that with built-in DDPG agent and custom agent but it is quite difficult so I decide to built everything by myself in Simulink with Matlab function.Now, the only problem I have is creating a experience buffer to store and batch data.