0
\$\begingroup\$

I have calculated the "active power" and the "the reactive" power of a system in dq axis in nonlinear form and I am doing system identification on this "nonlinear" plant as a MIMO system.

The problem I have is that the system identification works with "normalized models" and this causes that either there is a steady state error or no correct model is identified.

Is there any way to normalize a nonlinear plant (input/output) other than using the per unit parameters?

\$\endgroup\$
0
\$\begingroup\$

For best results, I suggest you need;

  • to convert to linear control e.g. 1/x, log x or at least piecewise linear control, then
  • have tuning weights for MIMO signals all of order unity and not higher order terms. e.g. position feedback for position control, current or acceleration feedback for likewise control.
  • use per unit variables for grid control
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.