Poles of transfer function and stability

Click For Summary
SUMMARY

This discussion focuses on the analysis of transfer functions and system stability through the linearization of differential equations. The key equations derived include the state-space representation and the final transfer function, which reveals poles at 7.11 × 10-5, -0.909, and 0. The presence of a pole at the origin indicates that the system is an integrating system and will not settle to a steady state following a step input change. The conversation also addresses challenges in performing partial fraction decomposition due to an unusual term in the denominator.

PREREQUISITES
  • Understanding of linear differential equations
  • Familiarity with state-space representation
  • Knowledge of Laplace transforms
  • Experience with transfer function analysis
NEXT STEPS
  • Study the implications of poles in control systems stability
  • Learn about the Routh-Hurwitz criterion for stability analysis
  • Explore methods for partial fraction decomposition in transfer functions
  • Investigate the effects of disturbance inputs on system response
USEFUL FOR

Control engineers, system analysts, and students studying control theory who are interested in transfer function analysis and system stability.

gfd43tg
Gold Member
Messages
947
Reaction score
48

Homework Statement


upload_2015-9-24_17-51-1.png

upload_2015-9-24_17-51-23.png


Homework Equations

The Attempt at a Solution


part a[/B]
$$ \vec {x} = \left [ \begin{array}{ccc} X-X_{s} \\ S-S_{s} \end{array}
\right ] $$

$$ \vec {u} = \left [ \begin{array}{ccc} S_{0}-S_{0s} \\ \frac {F_{0}}{V}-
\big (\frac {F_{0}}{V} \big )_{s} \end{array} \right ] $$
$$ \vec {y} = \left [ \begin{array}{ccc} X-X_{s} \end{array} \right ] $$
part b
Manipulated input
$$ u_{M} = S_{0}-S_{0s}$$

Disturbance input
$$ u_{D} = \frac {F_{0}}{V}- \big (\frac {F_{0}}{V} \big )_{s} $$
part c
The differential equations
$$ \frac {dX}{dt} = \mu (S) X - \frac {F_{0}}{V}X = f_{1} $$
$$ \frac {dS}{dt} = - \mu(S) \frac {X}{Y} + \frac {F_{0}}{V}(S_{0}-S) =f_{2} $$
are linearized into the form
$$ \dot{\vec{x}} = A \vec {x} + B \vec {u} $$
Where ##A_{ij} = \frac {\partial f_{i}}{\partial x_{j}}## and ##B_{ij} = \frac {\partial f_{i}}{\partial u_{j}}## For example, ##A_{11} = \frac {\partial f_{1}}{\partial X} = \mu(S)-\frac {F_{0}}{V}## and ##B_{11} = \frac {\partial f_{1}}{\partial S_{0}}=0##

The matrices ##A## and ##B## are solved symbolically and the differential equations are linearized below

$$ \left [ \begin{array}{ccc} \frac {d(X-X_{s})}{dt} \\ \frac {d(S-S_{s})}{dt} \end{array} \right ] = \left [ \begin{array}{ccc} \frac
{\mu_{m}S_{s}}{K_{s}+S_{s}}- \big ( \frac {F_{0}}{V} \big )_{s} & \frac {\mu_{m}K_{s}}{(K_{s}+S_{s})^{2}}X_{s} \\ - \frac {\mu_{m}S_{s}}{K_{s}+S_{s}} \cdot \frac {1}{Y} & - \frac {\mu_{m}K_{s}}{(K_{s}+S_{s})^{2}} \frac {X_{s}}{Y} - \big ( \frac {F_{0}}{V} \big )_{s} \end{array} \right ] \left [ \begin{array}{ccc} X-X_{s} \\ S-S_{s} \end{array} \right ] + \left [ \begin{array}{ccc} 0 & -X_{s} \\ \big (\frac {F_{0}}{V} \big )_{s} & S_{0s}-S_{s} \end{array} \right ] \left [ \begin{array}{ccc} S_{0}-S_{0s} \\ \big ( \frac {F_{0}}{V} \big )
- \big ( \frac {F_{0}}{V} \big )_{s} \end{array} \right ] $$

Using the steady state values and parameters given, the equations are linearized and tabulated in
the matrices ##A## and ##B##

$$ \left [ \begin{array}{ccc} \frac {d(X-X_{s})}{dt} \\ \frac
{d(S-S_{s})}{dt} \end{array} \right ] = \left [ \begin{array}{ccc} 7.11 \times 10^{-5} & 0.267 \\
-0.303 & -0.909 \end{array} \right ] \left [ \begin{array}{ccc} X - 1.877 \\
S-0.563 \end{array} \right ] + \left [ \begin{array}{ccc} 0 & -1.877 \\
0.1 & 5.687 \end{array} \right ] \left [ \begin{array}{ccc} S_{0}-6.25 \\ \frac {F_{0}}{V} - 0.1 \end {array} \right ] $$
part d
Recasting the matrices as two differential equations, where ##X = x_{1}## and ##S = x_{2}##,

$$ \frac {dx_{1}}{dt} = 7.11 \times 10^{-5}x_{1} + 0.267 x_{2} -
1.877u_{D} \hspace {2 in} (1)$$

$$ \frac {dx_{2}}{dt} = -0.303x_{1} - 0.909x_{2} + 0.1u_{M}+5.687u_{D} \hspace {1.7 in} (2) $$

Taking the Laplace transform of both differential equations,

$$ sX_{1}(s)= 7.11 \times 10^{-5}X_{1}(s) + 0.267 X_{2}(s) -
1.877U_{D}(s) \hspace {1.25 in} (3)$$

$$ sX_{2}(s) = -0.303X_{1}(s) - 0.909X_{2}(s) + 0.1U_{M}(s)+5.687U_{D}(s) \hspace {0.80 in} (4)$$

Using eq. (3) to solve for##X_{2}(s)= f(X_{1}(s),U_{D}(s))##, we get

$$ X_{2}(s)= 3.745X_{1}(s)(s-7.11 \times 10^{-5})+7.030U_{D}(s) $$

Plug in the expression for ##X_{2}(s)## into equation (4) to eliminate ##X_{2}(s)##. We isolate ##X_{1}(s)##, and the final expression for ##X_{1}(s)= Y(s) = f(U_{D}(s),U_{M}(s))## is

$$ Y(s)= \frac {-7.030(s+0.909)+5.687}{12.360(s-7.11 \times
10^{-5})(s+0.909)+1} U_{D}(s)+ \frac {0.1}{12.360(s-7.11 \times
10^{-5})(s+0.909)+1} U_{M}(s) $$
part e
If the disturbance input, ##\frac {F_{0}}{V}##, were to have a step change of magnitude 0.5, then the transfer function derived in part (d) would have an extra term for ##U_{D}(s)##, which is ##\frac {0.5}{s}##.

$$ Y(s)= \frac {-7.030(s+0.909)+5.687}{12.360(s-7.11 \times
10^{-5})(s+0.909)+1} \cdot \frac {0.5}{s} + \frac {0.1}{12.360(s-7.11 \times
10^{-5})(s+0.909)+1} U_{M}(s) $$

The poles are ##7.11 \times 10^{-5}##, ##-0.909##, and ##0##. With a pole at the origin, and one barely positive, the system is an integrating system and never settles to a steady state when a step input change is made.
part f

Working with only the disturbance input,

$$ Y(s)= \frac {-3.515(s+0.909)+2.844}{12.360s(s-7.11 \times 10^{-5})(s+0.909)+s} $$

I am stuck because my transfer function looks funky, it has the + constant term in the denominator, and I don't think it is right. There should be some (s + a) kind of term, this + s all by itself is something I haven't seen before. The substitution is really ugly for this problem, but I don't think I did it incorrectly. I can't do partial fraction decomposition with that term sticking out there. When I try and solve for ##s = -0.909## or ##s = 7.11 \times 10^{-5}##, I can't get rid of these terms from my transfer function, and then the answer is trivially zero for the constants.

 
Last edited:
Physics news on Phys.org
Maylis said:
I am stuck because my transfer function looks funky
I don't quite understand why you get stucked.

Just multiply

12.360s(s−7.11×10−5)(s+0.909) = As3 + Bs2 + Cs + D

then add the last s, so that you will get ( C+1 ) instead of C.

Now find the roots in

As3 + Bs2 + (C+1)s + D

rewrite the denominator ( with the roots changed ) and do the partial fraction decomposition.
 
Yes, I realized it later. I think I was too tired and hungry and had a headache so I didn't notice
 

Similar threads

  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
Replies
3
Views
3K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 9 ·
Replies
9
Views
2K
  • · Replies 11 ·
Replies
11
Views
2K