MATLAB Answers

0

Control of LLC Converter

Asked by Martin Wattenberg on 23 Jul 2018
Latest activity Commented on by Arkadiy Turevskiy on 22 Aug 2018
Hello,
I am currently trying to develop a control model for an LLC converter.
This is a topology in power electronics that relies heavily on the resonant behavior of an LC tank.
Please see the image attached for a simplified version
The output of the system is the amplitude of Vout. The input to the system is:
  1. the amplitude of Vin,
  2. the frequency of Vin.
Both parameters can be changed individually. The controller should later control the frequency of Vin to maintain Vout, while changes in amplitude of Vin are considered external disturbances.
My problem is how to model the frequency behavior in MATLAB? I have calculated the bode plot but how can I continue with the controller design?
R = 10;
L = 10e-6;
C = 10e-6;
G = tf([L*L*R*C,L*R*R*C,0,0],[L*L*R*C,L*L+L*R*R*C,2*R*L,R*R])
bode(G)
Thank you very much for your support!

  0 Comments

Sign in to comment.

2 Answers

Answer by Arkadiy Turevskiy on 21 Aug 2018

What is G - is it transfer function from Fs(frequency) to Vout? Probably it is transfer function from Vin to Vout right?
If you know the transfer function Fs to Vout, then you can design your controller using Control System Designer or PID Tuner in Control System Toolbox.
But in general, as I understand, you cannot determine that transfer function and you have to measure it - see t his application note from TI with proposed approach.
You also might want to consider simulating the system in Simscape Power Systems and designing your controller in simulation, as shown here .

  0 Comments

Sign in to comment.


Answer by Martin Wattenberg on 22 Aug 2018

Dear Arkadiy Turevskiy,
thank you for taking the time and replaying to my question.
The simscape power system in combination with the system identification toolbox could be a possibility. This way I could realize the proposed approach by TI.
However, the converter is a white box to me. I was looking for a more straightforward (and numerically more efficient) solution to my problem.
I am still interested in controlling the frequency to maintain Vout.

  1 Comment

If it is white box, does it mean you can analytically derive a transfer function from Fs to Vout? If you can, then the problem becomes straightforward.

Sign in to comment.