PI controller tuning in hall sensor FOC example

Hi, I am using the Simulink example of field-oriented control with a hall sensor (link at the end) for my master's thesis. I just wanted to know what tuning method was used for the two PI controllers (one speed and two current). Was it tuned using the FOC autotuner or another method ?

The PID tuner does not work as it says that it is a non-linear plant. I just wanted the information to add to my report

I would be extremely grateful to receive an answer

https://uk.mathworks.com/help/mcb/gs/foc-pmsm-using-hall-sensor-example.html

Answer 1

[Sam Chak]

Hi @Rohan

In the Thesis, you can probably explain like that

"When the 'Tune' button is clicked, the Simulink PID Tuner does not function properly because the plant cannot be linearized or linearizes to zero. Specifically, the permanent magnet synchronous motor with a Hall sensor may exhibit special nonlinear characteristics that cannot be mathematically linearized under the selected or any operating conditions. PID tuning in Simulink is a model-based optimization approach aimed at achieving a good balance between performance and robustness as specified by the PID control practitioner. The tuning algorithm requires a linear mathematical model, whether in state-space representation or as a transfer function. If the linear model is unavailable, the algorithm will attempt to compute a linear model of the plant within the Simulink model. From the perspective of the PID Tuner, the combination of all blocks between the PID controller output and input is considered the plant. Thus, if these blocks contain nonlinearities, they will be linearized to the specified operating conditions. However, not all types of nonlinearities can be linearized; for example, the discontinuity at zero in a signum function cannot be linearized. When this occurs, one option is to design a PID controller based on simulated frequency-response data. The steps to design a PID controller from the plant's frequency-response data are described below: blah, blah, blah..."

https://www.mathworks.com/help/slcontrol/ug/designing-pid-controller-in-simulink-with-estimated-frequency-response.html

[Rohan]

Wow, Thank you so much for taking your time and giving a detailed response !

This is better than anything I could have hoped for haha, much appreciated

Regards

[Sam Chak]

You are welcome, @Rohan. If you find the reponse helpful, please consider clicking 'Accept' ✔ on the answer and voting 👍 for it. Your support is greatly appreciated!

[Sam Chak]

Hi @Rohan

I have made a correction:

... the discontinuity at zero in a signum function cannot be linearized.