2nd Order Control system PD controller

AI Thread Summary
The discussion revolves around finding appropriate values for Kp and Kd in a PD control system to achieve a 5% overshoot and a settling time of 3 seconds. The user derived a transfer function but noted it didn't fit the standard second-order form, leading to confusion in calculating Kp and Kd. They calculated Kp as 15.14 but were given answers of 17.93 for Kp and 12.12 for Kd. Other participants suggested running simulations in MATLAB to verify the results and emphasized that multiple solutions may exist for such control problems. The conversation highlights the importance of plotting and simulating to confirm design criteria in control systems.
DC2R
Messages
2
Reaction score
0
Hi,

This question on PD control is from a practice quiz.

1. Homework Statement
upload_2016-4-28_10-3-23.jpeg

If you can't see it- the question asks to find values for Kp and Kd such that the system achieves 5% OS and has a settling time Ts of 3s.
Cs = 3
Cd = 2
m = 5

Homework Equations


ω_n^2/(s^2 + 2ζw_n + ω_n^2) - 2nd order form

Overall Transfer Function I found:
Y/R = CG/1+CG = [ (Kp + Kds)/m ] / [ s^2 + ((Cd + Kd)/m) + ((Cs + Kp)/m)]

Ts = 4/σ (2% of final value); where σ = ζω_n

The Attempt at a Solution


From the transfer function I found above i noticed that it is not exactly in the 2nd order form. However I was told that we still set

ω_n = SQRT[ ((Cs + Kp)/m) ] {1}
because when we add a derivative controller it only adds a zero which effects the numerator, not the poles on the denominator.

firstly we know for 5%OS → ζ≈0.7
from Ts = 3 = 4/ζω_n
→ ω_n = 4/3ζ = 40/21

from eqn {1} - solving for Kp
Kp = (ω_n^2)m - Cs = (5)(40/21)^2 - 3 = 15.14

At this point I stopped because their answers were Kp = 17.93 and Kd = 12.12.
Could you explain if the process i used is incorrect because I can't understand what I am doing wrong.

Thank you
(btw I just joined Physics Forums like an hour ago)
 

Attachments

  • problem.JPG
    problem.JPG
    32.2 KB · Views: 1,210
Physics news on Phys.org
There are often many ways to solve control problems. Plot your solution. Does it meet the design criteria?
 
hi, can someone please have a look at my working to see if I am on the right track, I spent many hours on it and I am not even sure if this is the correct way.

donpacino, I am a beginner in Matlab, I am not sure how a plot would verify my results.
 
DC2R said:
hi, can someone please have a look at my working to see if I am on the right track, I spent many hours on it and I am not even sure if this is the correct way.

donpacino, I am a beginner in Matlab, I am not sure how a plot would verify my results.
run a simulation to see if your KP and KI values meet the criteria for rise time and settling time.

There are often multiple solutions that will work for problems like this
 
Back
Top