Reduce overshoot with PI

[pattiane]

Hi all,

I am struggling with PI controller to control a curretn controller.The control is in discrete time.

Basically, i have a set-up with an inverter and LCL filter.
The idea is to control, using dSPACE, the current in d and q axis measured in the output of the LCL filter.

What I'm currently doing is to use a model of LCL filter to tune my PI regulator in Simulink.The reference current is a step.
The model of LCL filter is good, there is no problem.
My simulation model is also right, I checked with an other sample of LCL parameters.

But with my set-up of LCL filter, I cannt reduce the resulted overshoot .

I don't know if you get my point...
Basically, I need to know why in certain systems we cannot reduce the overshoot, is it because of a wrong compensation of poles?...

Thanks

 

[pattiane]

Hello skogsgurra,

About Rd in not working system,it is high.
Actually, this damping resistor is calculated theoritically by taking into account resonance frequency and capacitor:
Rd= 1/3 * 1/ (OMEGA_res.C)
But it is true that in real, it should be very small about mOhm. But I wanted to check firstly with the theoritical value.

 

[Skogsgurra]

Suddenly very quiet in this thread. How's it going?

Gunnar Englund

http://www.gke.org

 

[pattiane]

Hi All,

Yes, it is Skogsgurra.
Ive resolved my problem yersteday...

The source of problem was the used design of PI controller.
I didn't use a normal controller, the design was slight different.
Once tuned the PI using SISOTOOL(Simulink GUI), Ive taken the transfer function of PI from that, with
Kp (z-zp)/(z-1) format.
Since the discretization of the controller is done by Tustin method in SISOTOOL, zp= (Tsw-2Ti)/(Tsw+2Ti), I can find out Ti knowing the switching period Tsw.

"A PI controller can only be used to place one pole since the integrator has its own pole."PNachtwey
I forgot to mention also that the capacitor C of LCL filter can be negligible.
why? Because, C is DELTA wired in 3-phases, then in Y wired configuration, the current passing through C is equal to :
Ic = Vphase to neutral / Zc
with Zc= 9.6 uF * (2*pi*50Hz)
Ic is about 0.7 A, so the capacitor can be negligible, then
the transfert function of the plant (LCL) becomes:
H(s)= 1 / (Li+Lg)s
1 pole instead of 3 poles, you can check it works, the control is the same.

Thanks anyway for your help and questions.

LP

 

[Skogsgurra]

You got me confused there. And I still am. Difference equations (z) are perhaps more common in pure DSP programming. Control people are used to s domain.



Gunnar Englund

http://www.gke.org

 

[pattiane]

Skogsgurra,

Ok, let's say that I was able to tune the controller using SISOtool, which enables to find pole and zero of the controller.
In PI case:
Kp (1+ Ki/s) = K( s + Ki) / s with Ki=1/Ti
Ki is the zero of the plant.

So I was to find Ki and Kp via SISOtool.That's all.

 

[Skogsgurra]

So, there wasn't any problem at all?

Gunnar Englund

http://www.gke.org

 

[pattiane]

The problem was from the design of PI with Simulink blocks.
But I still don't understand why this design works with the first set-up and not with the second set-up, which is mine.

In fact,using PI as a transfer function has resolved the problem, in my global system.