Answered
I want to convert the PI controller in this instance to a PID controller, so can my network be changed to something like this?
That should work - you are essentially adding one more weight to the PI layer for the D term. Potential issues may arise when yo...

2 years ago | 1

| accepted

Answered
Deep reinforcement learning and TD3 algorithm in the PID control
Have you seen this example?

2 years ago | 0

| accepted

Answered
How do I stop the reinforcement learning training and ensure agent is saved when the Episode Manager Crashes. (Stop button on episode manager not accessible)
If you use the stop button, the agent you will get will have the parameters learned to that point. You can then continue trainin...

2 years ago | 0

Answered
How could I realize from the graph of "reinforcement learning episode manager" in DDPG or rlTD3Agent that the network is learning well?
You should be seeing the average reward curve going upwards (not monotonically) which is not the case here.

2 years ago | 0

| accepted

Answered
How do i define and extract the MPC constraints?
The getconstraint function is supposed to return mixed input/output constraints. One use case would be that someone creates the ...

2 years ago | 0

Answered
RL Agent Action Limits doesn't working.
Please take a look at the DDPG algorithm and specifically step 1 here. DDPG promotes exploration by adding noise on top of the a...

2 years ago | 0

Answered
I received this error in DDPG "Model input sizes must match the dimensions specified in the corresponding observation and action info specifications."
The easiest way to discover your error yourself is to use the default agent feature have use the network architecture that's aut...

2 years ago | 0

Answered
Error with using loggedSignals in the reset function while creatin RL environment using custom functions
The very first line in our reset function is LoggedSignals.currentTrial = LoggedSignals.currentTrial + 1; How would the functi...

2 years ago | 0

| accepted

Answered
Understanding Entropy Loss for PPO Agents Exploration
Hi, In PPO, the goal of training is to strike a balance between the entropy term and fine tuning the probabilities for all avai...

2 years ago | 0

Answered
How to reduce noise from SAC RL-agent?
Hi, There is an agent option that achieves exactly what you want and outputs the mean output values after training: sacagent.U...

2 years ago | 0

Answered
Reinforcement Learning - SAC with hybrid action spaces
Hi, Unfortunately, hybrid action spaces are not currently supported out of the box. One potential workaround is to use multi-ag...

2 years ago | 0

| accepted

Answered
How to set a variable learning rate in Reinforcement Learning Toolbox?
Hi, This is not supported as of R2023b but the developmet team is aware and will look into supporting this capability in the fu...

2 years ago | 0

| accepted

Answered
How can i set the constraints on states rather than input and output in mpc?
For linear MPC, you can add constraints on inputs and outputs only. You can either set the desired states to also be outputs (if...

2 years ago | 0

| accepted

Answered
How do we know that the PI controller can be modeled using a single neuron?
The network used to model the PI controller is exactly this one actorNet = [ featureInputLayer(numObs) fullyConnected...

2 years ago | 0

Answered
Using a Simulink dynamic motorcycle in a driving scenario trajectory.
Here is a video that shows how to do that using Model Predictive Control Toolbox: https://www.mathworks.com/videos/understandin...

2 years ago | 0

Answered
question about external action of DDPG
The loss function does not change. What happens is that the experience buffer is populated with the action from the external sig...

2 years ago | 0

| accepted

Answered
use a linear state space model in NLMPC
Is there a reason you want to convert to nonlinear MPC? If you get a good solution with linear MPC, going to nonlinear will only...

2 years ago | 0

Answered
Integral MPC in Simulink
Why don't you use the MVRate constraint instead of adding the term in the cost function? https://www.mathworks.com/help/mpc/ug...

2 years ago | 0

Answered
Why reinforcement learning has different results of action between sim() and getAction()?
Hi, Which release are you using? We tried in R2023a and R2023b with UseExplorationPolicy =0 and getAction and sim provide the s...

2 years ago | 0

Answered
Epsilon greedy policy for DQN
You can use the formula here to calculate the epsilon value

2 years ago | 1

| accepted

Answered
Reinforcement learning agent for mixed action space.
Reinforcement Learning Toolbox does not support agents with both continuous and discrete actions. Can you share some more detail...

2 years ago | 0

Answered
can i decide the RL agents actions
It seems like the paper you saw uses some logic to implement the behavior you mention. You could do the same with an if statemen...

2 years ago | 0

Answered
The agent can learn the policy through the external action port in the RL Agent so that the agent mimics the output of the reference signal
It seems the agent started learning how to imitate the existing controller but needs more time. What does the Episode Manager lo...

2 years ago | 0

Answered
How to import a model built by comsol in the reinforcement learning designer
I haven't used Comson before but it seems that you may be able to co-simulate your model with Simulink. In that case, the proces...

2 years ago | 0

Answered
Multi-Agent Reinforcement learning
As of R2023b, you can do multi-agent reinforcement learning using MATLAB environments. Please take a look at this example and R2...

2 years ago | 0

Answered
How can i scale the action of DDPG agent in Reinforcement Learning?
DDPG training works by adding noise on top of the actor output to promote exploration. In that case you may see constraint viola...

2 years ago | 0

Answered
How I can define eight discrete actions in RL section
The implementation shown here is one option. Hope that helps

2 years ago | 0

Answered
In SImulink, during DDPG training to regulate the CO2 concentration the environment is not simulating. I can see only the variables specified in function is updating.
Hello, I would start by taking a look at the output of the agent. If the agent output does not make sense, the environment will...

2 years ago | 0

Answered
5G Handover with Reinforcement Learning, mismatch of input channels and observations in reinforcement learning representation
I suspect you did not set up your critic network properly. If you share that code snippet we can take a closer look. An alternat...

2 years ago | 0

Answered
Implementing mpctools package (from Rawlings group) in Simulink
I cannot comments on mpctools, but if your objective is to use IPOPT in Simulink, Model Predictive Control Toolbox allows you to...

2 years ago | 0