In "frestimat​e(modelnam​e,op,io,in​)", it is required for each of the linearization input I/O points that the sample time of the Simulink signal where it is located is one of the following:

Hello

Getting the following error when tried linearzing power facotor model

In "frestimate(modelname,op,io,in)", it is required for each of the linearization input I/O points that the sample time of the Simulink signal where it is located is one of the following: 1. It is either continuous or fixed in minor time step. 2. Its sample rate is equal to the sample rate of the input signal "in". The linearization I/O io(1) does not satisfy this requirement. Consider either changing the sample time of Simulink signal where io(1) is located to continuous sample time or changing the sample time of input signal "in".

Input pertuberation : Duty Cycle

Open Loop Output : Inductor current

Any idea how to approach the problem ?

Answer 1

[Antonino Riccobono]

Hi, both input linearization point and output measurement point are required to be equally sampled and must match the perturbation signal's sample time. Please, check the sample time in your model by going to the Debug tab -> Information Overlays -> Colors in the section Sample Time.

If you there are different sample times, consider to either fix your model sample time(s) or using Rate Transition blocks to set the smple time(s) as appropriate.

Hope this helps!

[Sreeraj A]

Hello Antonio !

Thank you for the reply.

I have already checked the sample time and both input and output are running at the same sample time.

I was trying to follow the method from

https://in.mathworks.com/videos/active-power-factor-correction-1546869199547.html and getting the above error when tried to plot the Bode diagram.

[Antonino Riccobono]

I see that you are modeling the PFC rectifier with Simscape Electrical - Specialized Power Systems components and you are using a discrete simulation type in the Powergui.

Can you set the simulation type to continuous and give it a try?

[Sreeraj A]

Thank you Antonino for the suggetion.

When changed to contionus simulation. Bode estimation seems to run infinitly ! Verified model simulation is workin fine only when tried to run the Bode estiamtion it goes endless !

[Antonino Riccobono]

Can you please share a screenshot of the settings of your input perturbation signal? I assume you are using Sinestream, correct?

[Sreeraj A]

yes, I have tried using fiexed sample time Sinstream with sample time.

Also I would like to know is it not possbil to use Bode Plot with Fixed-step solver with Discrete Power Gui ? I am aiming for code generation as well from the model that's the reason using fixed-step solver.

[Antonino Riccobono]

It seems like that you are doing things correctly. There could be many reasons why your estimation seems to be running endless - my biggest suspicion is on solver settings and small sample times.

Let me give you some hints to begin with, but feel free to share your model with me so I can have a closer look while you are also working on it.

Solver settings: Configuration Parameters -> variable-step, ode23tb

Sample times: if your switching frequency is fsw (switching period Tsw=1/fsw), your PWM sample time should be Ts_PWM=Tsw/100 to resolve 1% duty cycle (smaller PWM sample times slow simulation/estimation down); control sample time can be chosen as Ts_control = Tsw.

If this works, then you can use the PID Tuner app to tune your controller (one loop at time as explained in the video).

Finally, you can use discrete simulation type in the Powergui for your code generation task.

[Sreeraj A]

Sorry for the delayed responce.

To rule out the sample time problem I tried taking the same model as in the demo and tried the Bode Plot and even after more than 2hour simulation still cannot plot !I can see the simulation has completed from the scope but the simulation is still running. I belive I have got resonalble configuration in my laptop(12GB RAM, i5 Processor 10th Gen)

https://www.mathworks.com/matlabcentral/fileexchange/69653-active-power-factor-correction

Any help would be much appreicated.

[Sreeraj A]

Attached the model and m-file.

[Antonino Riccobono]

Cool! I'll take a look and come back to you asap.

[Sreeraj A]

Thank you Antonino for the support.

[Antonino Riccobono]

Hi,

I made a few changes in the model to speed up estimation time. Please, see attachments. Based on these changes, use a sample time = 4e-7 s when creating a fixed sample time sinestream object.

Hope this helps,

Antonino

[Sreeraj A]

Thank you very much for you support. I will try this.

[nada zz]

Thank you so much ... This solution works for me

[Jakub Siedliski]

I tried Antonio's model and unfortunately the simulation goes on forever and matlab will not respond to clicks (although it does not say it's not responding), it just says it's still simulating the model. I'm running a ryzen 5 3600 with 16GB DDR4 RAM. Is this normal?

[Prashant Mishra]

I'm getting this error in the same model. Can you help please?