I am getting this error when running my code >>>>Dot indexing is not supported for variables of this type. Error in rl.util.ex​pstruct2ti​meserstruc​t (line 7) observation

The below code is the one I am running

Create Simulink Environment and Train Agent

This example shows how to convert the PI controller in the watertank Simulink® model to a reinforcement learning deep deterministic policy gradient (DDPG) agent. For an example that trains a DDPG agent in MATLAB®, see Train DDPG Agent to Balance Double Integrator Environment.

Water Tank Model

The original model for this example is the water tank model. The goal is to control the level of the water in the tank. For more information about the water tank model, see watertank Simulink Model.

Modify the original model by making the following changes:

1. Delete the PID Controller.
2. Insert the RL Agent block.
3. Connect the observation vector , where h is the height of the water in the tank, , and r is the reference height.
4. Set up the reward .
5. Configure the termination signal such that the simulation stops if or .

The resulting model is rlwatertank.slx. For more information on this model and the changes, see Create Simulink Environment for Reinforcement Learning.

open_system("RLFinal_PhD_Model_DroopPQ1")

Create the Environment

Creating an environment model includes defining the following:

• Action and observation signals that the agent uses to interact with the environment. For more information, see rlNumericSpec and rlFiniteSetSpec.
• Reward signal that the agent uses to measure its success. For more information, see Define Reward Signals.

Define the observation specification obsInfo and action specification actInfo.

% Observation info
obsInfo = rlNumericSpec([3 1],...
    LowerLimit=[-inf -inf 0  ]',...
    UpperLimit=[ inf  inf inf]');

% Name and description are optional and not used by the software
obsInfo.Name = "observations";
obsInfo.Description = "integrated error, error, and measured height";

% Action info
actInfo = rlNumericSpec([1 1]);
actInfo.Name = "flow";

Create the environment object.

env = rlSimulinkEnv("RLFinal_PhD_Model_DroopPQ1","RLFinal_PhD_Model_DroopPQ1/RL Agent1",...
    obsInfo,actInfo);

Set a custom reset function that randomizes the reference values for the model.

env.ResetFcn = @(in)localResetFcn(in);

Specify the simulation time Tf and the agent sample time Ts in seconds.

Ts = 1.0;
Tf = 200;

Fix the random generator seed for reproducibility.

rng(0)

Create the Critic

DDPG agents use a parametrized Q-value function approximator to estimate the value of the policy. A Q-value function critic takes the current observation and an action as inputs and returns a single scalar as output (the estimated discounted cumulative long-term reward for which receives the action from the state corresponding to the current observation, and following the policy thereafter).

To model the parametrized Q-value function within the critic, use a neural network with two input layers (one for the observation channel, as specified by obsInfo, and the other for the action channel, as specified by actInfo) and one output layer (which returns the scalar value).

Define each network path as an array of layer objects. Assign names to the input and output layers of each path. These names allow you to connect the paths and then later explicitly associate the network input and output layers with the appropriate environment channel. Obtain the dimension of the observation and action spaces from the obsInfo and actInfo specifications.

% Observation path
obsPath = [
    featureInputLayer(obsInfo.Dimension(1),Name="obsInLyr")
    fullyConnectedLayer(50)
    reluLayer
    fullyConnectedLayer(25,Name="obsPathOutLyr")
    ];

% Action path
actPath = [
    featureInputLayer(actInfo.Dimension(1),Name="actInLyr")
    fullyConnectedLayer(25,Name="actPathOutLyr")
    ];

% Common path
commonPath = [
    additionLayer(2,Name="add")
    reluLayer
    fullyConnectedLayer(1,Name="QValue")
    ];

% Create the network object and add the layers
criticNet = dlnetwork();
criticNet = addLayers(criticNet,obsPath);
criticNet = addLayers(criticNet,actPath);
criticNet = addLayers(criticNet,commonPath);

% Connect the layers
criticNet = connectLayers(criticNet, ...
    "obsPathOutLyr","add/in1");
criticNet = connectLayers(criticNet, ...
    "actPathOutLyr","add/in2");

View the critic network configuration.

figure
plot(criticNet)

Initialize the dlnetwork object and summarize its properties.

criticNet = initialize(criticNet);
summary(criticNet)

Create the critic approximator object using the specified deep neural network, the environment specification objects, and the names if the network inputs to be associated with the observation and action channels.

critic = rlQValueFunction(criticNet, ...
    obsInfo,actInfo, ...
    ObservationInputNames="obsInLyr", ...
    ActionInputNames="actInLyr");

For more information on Q-value function objects, see rlQValueFunction.

Check the critic with a random input observation and action.

getValue(critic, ...
    {rand(obsInfo.Dimension)}, ...
    {rand(actInfo.Dimension)})

For more information on creating critics, see Create Policies and Value Functions.

Create the Actor

DDPG agents use a parametrized deterministic policy over continuous action spaces, which is learned by a continuous deterministic actor.

A continuous deterministic actor implements a parametrized deterministic policy for a continuous action space. This actor takes the current observation as input and returns as output an action that is a deterministic function of the observation.

To model the parametrized policy within the actor, use a neural network with one input layer (which receives the content of the environment observation channel, as specified by obsInfo) and one output layer (which returns the action to the environment action channel, as specified by actInfo).

Define the network as an array of layer objects.

actorNet = [
    featureInputLayer(obsInfo.Dimension(1))
    fullyConnectedLayer(3)
    tanhLayer
    fullyConnectedLayer(actInfo.Dimension(1))
    ];

Convert the network to a dlnetwork object and summarize its properties.

actorNet = dlnetwork(actorNet);
summary(actorNet)

Create the actor approximator object using the specified deep neural network, the environment specification objects, and the name if the network input to be associated with the observation channel.

actor = rlContinuousDeterministicActor(actorNet,obsInfo,actInfo);

For more information, see rlContinuousDeterministicActor.

Check the actor with a random input observation.

getAction(actor,{rand(obsInfo.Dimension)})

For more information on creating critics, see Create Policies and Value Functions.

Create the DDPG Agent

Create the DDPG agent using the specified actor and critic approximator objects.

agent = rlDDPGAgent(actor,critic);

For more information, see rlDDPGAgent.

Specify options for the agent, the actor, and the critic using dot notation.

agent.SampleTime = Ts;

agent.AgentOptions.TargetSmoothFactor = 1e-3;
agent.AgentOptions.DiscountFactor = 1.0;
agent.AgentOptions.MiniBatchSize = 64;
agent.AgentOptions.ExperienceBufferLength = 1e6; 

agent.AgentOptions.NoiseOptions.Variance = 0.3;
agent.AgentOptions.NoiseOptions.VarianceDecayRate = 1e-5;

agent.AgentOptions.CriticOptimizerOptions.LearnRate = 1e-03;
agent.AgentOptions.CriticOptimizerOptions.GradientThreshold = 1;
agent.AgentOptions.ActorOptimizerOptions.LearnRate = 1e-04;
agent.AgentOptions.ActorOptimizerOptions.GradientThreshold = 1;

Alternatively, you can specify the agent options using an rlDDPGAgentOptions object.

Check the agent with a random input observation.

getAction(agent,{rand(obsInfo.Dimension)})

Train Agent

To train the agent, first specify the training options. For this example, use the following options:

• Run each training for at most 5000 episodes. Specify that each episode lasts for at most ceil(Tf/Ts) (that is 200) time steps.
• Display the training progress in the Episode Manager dialog box (set the Plots option) and disable the command line display (set the Verbose option to false).
• Stop training when the agent receives an average cumulative reward greater than 800 over 20 consecutive episodes. At this point, the agent can control the level of water in the tank.

For more information, see rlTrainingOptions.

trainOpts = rlTrainingOptions(...
    MaxEpisodes=5000, ...
    MaxStepsPerEpisode=ceil(Tf/Ts), ...
    ScoreAveragingWindowLength=20, ...
    Verbose=false, ...
    Plots="training-progress",...
    StopTrainingCriteria="AverageReward",...
    StopTrainingValue=800);

Train the agent using the train function. Training is a computationally intensive process that takes several minutes to complete. To save time while running this example, load a pretrained agent by setting doTraining to false. To train the agent yourself, set doTraining to true.

doTraining = true;

if doTraining
    % Train the agent.
    trainingStats = train(agent,env,trainOpts);
else
    % Load the pretrained agent for the example.
    load("WaterTankDDPG.mat","agent")
end

Validate Trained Agent

Validate the learned agent against the model by simulation. Since the reset function randomizes the reference values, fix the random generator seed to ensure simulation reproducibility.

rng(1)

Simulate the agent within the environment, and return the experiences as output.

simOpts = rlSimulationOptions(MaxSteps=ceil(Tf/Ts),StopOnError="on");
experiences = sim(env,agent,simOpts);

Local Reset Function

function in = localResetFcn(in)

% Randomize reference signal
blk = sprintf("RLFinal_PhD_Model_DroopPQ1/Droop/Voutref");
h = 3*randn + 0.5;
while h <= 0 || h >= 400
    h = 3*randn + 200;
end
in = setBlockParameter(in,blk,Value=num2str(h));

% Randomize initial height
h1 = 3*randn + 200;
while h1 <= 0 || h1 >= 1
    h1 = 3*randn + 0.5;
end
blk = "RLFinal_PhD_Model_DroopPQ1/Gain";
in = setBlockParameter(in,blk,Gain=num2str(h1));

end

I am getting the following results without rewards at all

Zero rewards

When I stop the trainging I see this error::::

Dot indexing is not supported for variables of this type.

Error in rl.util.expstruct2timeserstruct (line 7)

observation = {experiences.Observation};

Error in rl.env.AbstractEnv/sim (line 138)

s = rl.util.expstruct2timeserstruct(exp,time,oinfo,ainfo);

[Image Analyst]

experiences does not have a field, property, or method called Observation. Are you sure you spelled everything right, including the vapitalization?

Just before that line, put one of these:

fn = fieldnames(experiences)
props = properties(experiences)
m = methods(experiences)

and then:

observation = {experiences.Observation};

Does any of those lines print out anything in the command window that is not an error message?

[Walter Roberson]

How are you stopping the training?

I suspect that you are encountering a glitch due to stopping the training.

[Mxolisi]

Hi Walter Roberson, I am not sure which country you based in. I would like to kindly ask to have a Teams meeting with you. I am based in South Africa but whatever timezone you on I can organize the meeting.

[Mxolisi]

In the meantime this is how I stop the traning:

[Mxolisi]

@Image Analyst

I am getting no output on the command window with the following from the functions you suggested above, how possible is it to have a chat with you on Teams.

[Mxolisi]

Hi again, it looks like my problem is on the scope below which actually returns some of the data to the workspace but I do not know how to actually create it....see below (the small blue scope on the right)

[Mxolisi]

Oh no I am wrong, these are just viewers.

[Image Analyst]

@Mxolisi we don't do private consulting here. If we can't solve it here, you can get a maintenance contract and call Mathworks tech support and they will help you. It would be cheaper than what I'd charge you. We try to solve things here so others can benefit from the solution in the future.

Since you said that none of the commands I gave you produced anything in the command window, it seems that experiences does not have any fields, methods, or properties at all. So you should not try to reference the Observation object of experiences because it does not exist. This seems to be corroborated by the error message you got. "Dot indexing is not supported for variables of this type." Perhaps you somehow, somewhere overwrite experiences with just a simple double variable?

OK let's try something different. Just before the line that threw the error, put these lines:

workspace
experiences
whos experiences

There definitely will be something printed to the command window when you do that. Again, make sure it's before where the error occurs, not after. What do you see?

If you open up the variable in the workspace (double click on experiences there) what do you see in the new window that pops up?

[Mxolisi]

@Image Analyst, thank you very much for your response. There is still NO output on the Command Window, I also cant see "experiences" under workspace.

[Mxolisi]

@Image Analyst, @Walter Roberson I really appreciate your responses!! I will try and get hold of Mathwork to see if I can get some help.

[Image Analyst]

There definitely should be something in the command window. I don't have Simulink so I'm not sure what you're seeing. Do you have a command window showing, and do you have the editor window open to the line causing the error (line 7 of rl.util.expstruct2timeserstruct.m)?

Go to this line, line 7:

observation = {experiences.Observation};

and set a breakpoint there. When the debugger stops there, you should see some variables listed in the workspace panel. Post a screenshot of that.

Then, in the command window type this:

>> experiences

>> whos experiences

Even if experiences does not exist, you would at least see this error message on the first one saying that it does not exist, and the whos command wouldn't say anything. Just like this:

>> experiences

Unrecognized function or variable 'experiences'.

>> whos experiences

>>

Post a screenshot of your editor window showing you've stopped at that line, plus the command window showing something.

[Walter Roberson]

@Mxolisi

Sorry, I reserve online meetings for my paying customers (and you cannot afford me.)

[Mxolisi]

This is what I see after stopping the training:

This is what I get on the Command window:

I hope this is what you need for diagnostics

This is whst I see in the workspace:

[Walter Roberson]

You do not seem to have put a breakpoint on line 7 of rl.util.expstruct2timeserstruct.m

[Mxolisi]

I am failing to understand, how must it look like after I have put breakpoint?

[Walter Roberson]

dbstop in rl.util.expstruct2timeserstruct.m 7

Then run the code. It will (presumably) stop just before executing line 7, and will show the debugging prompt

K>>

At that point you can explore the then-current variable experiences