How do I remove this error? its is in a charge controller connected to a P & O algorithm of a 1.57 MW solar plant.

[Jan]

If the value of the derivative gets infinite, an integration cannot work for numerical reasons.

[Joel Van Sickel]

Hello,

there are many ways this type of interaction could be caused that is model dependent. Are you able to share the model that causes this?

Regards,

Joel

Answer 1

[Dinesh]

Hi Phodiso!

This message could be caused by a problem in your system that makes it difficult for the Simulink solver to accurately determine step size due to large differences in state values. Also, Integration cannot work if the derivative gets infinity. To check this, you need to follow these steps:

Step 1

• In the model editor, click on Simulation and then Configuration Parameters.
• Select Data Import/Export and check the 'States' option in the 'Save to Workspace' field.
• This will allow you to log the states as output.

Step 2

• Run the simulation.

Step 3

Plot the results to determine if there are any large differences in the state values of the system.

• plot(tout,xout)

To identify if the issue is caused by a singularity in your system, you can check if any of the states are changing significantly more than the others. In case you find such a situation, you can take a couple of steps to fix this. Firstly, you could rescale the model to ensure that the states are comparable in magnitude. Additionally, setting specific error tolerances for your integrator blocks could also help resolve the issue.

To learn more about setting absolute tolerance for your block's outputs, you can reference the setting the absolute tolerance section MATLAB’s integrator block documentation.

https://www.mathworks.com/help/simulink/slref/integrator.html

Hope this helps,

Thank you!!