NPM IRID FOPID (correct plant) compared with IOPID with anti-windup, static error too large

[cnpcshangbo]

Neighborhood phase margin method for tuning fractional order PID controller in Simscape environment

Video

https://drive.google.com/file/d/1gkdWI1xhcpaHrljClASuOK9nrwt1UG8r/view

Plant model

[cnpcshangbo]

Docs

https://www.icloud.com/pages/0mgClyIkPIeL3LbBYER3EVAdw#extended_abstract

NPM Result

% k_p=x(1); k_i=x(2); k_d=x(3); lambda=x(4); mu=x(5); 2.6000 0 0.9588 0 1.5000

[cnpcshangbo]

IRID code

gam=0.5; Ts=0.01; dfod=irid_fod(gam,Ts,5);

IRID result

s^0.5=>Discrete transfer function

num=[1,-3.16336631308587,3.71959536925849,-1.96568707629962,0.436885459917815,-0.0273811687803932];

den=[0.0935414346693485,-0.239486864955964,0.209356682991156,-0.0676440492684792,0.00352282535674651,0.000822432073709469];

[cnpcshangbo]

FOPID controller in Simscape/Simulink

Simulink model

[cnpcshangbo]

Result

[cnpcshangbo]

To-do

• Change x0 to the GNCC parameters;
• Change mu=1.5;
• Try put the whole controller to discrete transfer function;
• Check if my FOPID is adding one more integrator.

[cnpcshangbo]

New issue: Verifying if put the whole controller to discrete transfer function is the same as IRID for FO derivator.