22-Mec-A3 System Analysis and Control · December 2014
Question 1 of 6: Root Locus and Range of Stabilising Gain
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper format. National Exams, December 2014 —
07-Mec-A3 System Analysis and Control. Three hours, closed book
(Casio or Sharp approved calculator; semi-logarithmic graph paper is the only permitted aid).
Six questions are printed; the rubric states that any four questions constitute a complete
paper and that all questions are of equal value, so each carries 25 marks.
A two-page table of Laplace transform pairs is appended to the paper.
All six questions are solved here.
Reference texts. K. Ogata, Modern Control Engineering, 5th ed.
(Prentice Hall); N. S. Nise, Control Systems Engineering, 8th ed. (Wiley);
R. C. Dorf & R. H. Bishop, Modern Control Systems, 13th ed. (Pearson);
G. F. Franklin, J. D. Powell & A. Emami-Naeini, Feedback Control of Dynamic
Systems, 8th ed. (Pearson). Section references in the concept panels below follow
Ogata 5th ed. and Nise 8th ed.
Convention used throughout. Angles are quoted in
degrees for phasor results and in radians inside time functions; a lagging phase is written
as a negative angle. "Time constant" always means the reciprocal of the magnitude of the
real part of the governing closed-loop pole.
Question 1: Root Locus and Range of Stabilising Gain (25 marks)
The open loop therefore has $n=2$ poles at $s=0$ and $s=-2$, and $m=2$ finite zeros at
$s=\pm j2$ (the imaginary roots of $s^{2}+4=0$).
Find. An accurate root-locus sketch, and the set of positive gains $K$
for which every closed-loop pole lies strictly in the left half-plane.
[Figure not reproduced: Figure 1 — the unity-feedback configuration as printed on page 2 of the examination paper. See the official exam paper.]
Approach. Form the closed-loop characteristic polynomial, apply the
Routh–Hurwitz coefficient test to settle stability for all $K$ at once, then construct
the locus from the standard rules (real-axis segments, breakaway points from
$dK/ds=0$, and the terminal behaviour as $K\to\infty$); because $n=m=2$ the locus turns out
to be an exact circle, which is what makes an "accurate" sketch possible.
Write the characteristic equation. Closed-loop poles satisfy
$1+G(s)H(s)=0$, i.e. $s(s+2)+K(s^{2}+4)=0$. Collecting powers of $s$,
$$(1+K)\,s^{2} \;+\; 2\,s \;+\; 4K \;=\; 0 .$$
The loop is second order for every finite $K$, and the coefficient of $s^{2}$ grows with $K$
— the tell-tale signature of a plant with as many zeros as poles.
Apply the Routh–Hurwitz test. For a quadratic
$a_{2}s^{2}+a_{1}s+a_{0}$ the necessary and sufficient condition for both roots to lie in
the left half-plane is simply that $a_{2},a_{1},a_{0}$ all share the same sign. Here
$a_{2}=1+K$, $a_{1}=2$ and $a_{0}=4K$. Since $a_{1}=2>0$ is fixed, the requirement is
$1+K>0$ and $4K>0$, both of which hold for every $K>0$. Hence
$$\boxed{\;0 < K < \infty \;\Longrightarrow\; \text{the closed loop is stable for
every positive gain.}\;}$$
There is no finite gain margin to compute: the locus never crosses into the right half-plane.
Confirm this directly from the pole locations. Solving the quadratic,
$$s \;=\; \frac{-2 \pm \sqrt{4-16K(1+K)}}{2(1+K)} .$$
Whenever the roots are complex their real part is
$\sigma=-\dfrac{1}{1+K}$, which is negative for all $K>0$ and tends to $0^{-}$ as
$K\to\infty$. Whenever the roots are real, both the sum $-2/(1+K)$ and the product
$4K/(1+K)$ are respectively negative and positive, which forces both roots to be negative.
Stability for all positive gain is therefore confirmed twice over.
Locate the real-axis segments. A point on the real axis belongs to the
$K>0$ locus when the total number of real poles and real zeros to its right is odd. The only
real critical points are the poles $s=0$ and $s=-2$; the zeros sit on the imaginary axis and
do not count. The segment $-2 < \sigma < 0$ has exactly one critical point ($s=0$) to
its right and so lies on the locus, while $\sigma>0$ and $\sigma<-2$ do not.
Find the breakaway and break-in points. Solving the characteristic
equation for the gain gives $K=-\dfrac{s(s+2)}{s^{2}+4}$. Setting $dK/ds=0$,
$$(2s+2)(s^{2}+4)-2s\,(s^{2}+2s)=0 \;\Longrightarrow\; -2s^{2}+8s+8=0
\;\Longrightarrow\; s^{2}-4s-4=0 ,$$
whose roots are $s=2\pm2\sqrt{2}$, that is $s=-0.8284$ and $s=+4.8284$. Only
$s=-0.8284$ lies on the $K>0$ real-axis segment, so it is the breakaway point; the
other root is a break-in belonging to the negative-gain (complementary) locus. Substituting
back,
$$\boxed{\;s_{b}=2-2\sqrt{2}=-0.8284,\qquad
K_{b}=\frac{\sqrt{2}-1}{2}=0.2071 . \;}$$
The same gain follows from setting the discriminant $4-16K(1+K)$ to zero, which is the
independent check that the two branches meet exactly there.
Show that the complex branches form a circle. Writing a complex root as
$s=\sigma+j\omega$, we have $\sigma=-1/(1+K)$ and
$\sigma^{2}+\omega^{2}=|s|^{2}=4K/(1+K)$. Eliminating $K$ between the two relations gives
$\sigma^{2}+\omega^{2}=4+4\sigma$, i.e.
$$\boxed{\;(\sigma-2)^{2}+\omega^{2}=8\;}$$
— a circle centred at $s=+2$ with radius $2\sqrt{2}=2.8284$. It cuts the real axis
exactly at the breakaway and break-in points $2\mp2\sqrt{2}$ and passes through the zeros
$\pm j2$, since $(0-2)^{2}+2^{2}=8$. Only the left-hand arc ($\sigma\le 0$) is traced by
positive gain.
Assemble the sketch. For $K$ increasing from $0$ the two poles move
along the real axis towards each other, meet at $-0.8284$ when $K=0.2071$, split into a
complex-conjugate pair that runs along the circular arc, and finally converge on the finite
zeros $\pm j2$ as $K\to\infty$. The loop is therefore stable for every positive gain but
becomes marginally stable only in the limit, where the poles reach the imaginary
axis and the response degenerates into an undamped 2 rad/s oscillation.
Root locus of
$K(s^{2}+4)/[s(s+2)]$: real-axis segment from $-2$ to $0$, breakaway at $-0.828$
($K=0.207$), then the circular arc of radius $2\sqrt2$ about $s=+2$ carrying the poles onto
the zeros at $\pm j2$ as $K\to\infty$. The dashed circle is the construction circle; only
its left arc belongs to the positive-gain locus.
Quantity
Result
Open-loop poles / zeros
$s=0,\,-2$ / $s=\pm j2$
Real-axis locus segment
$-2 \le \sigma \le 0$
Breakaway point and gain
$s=-0.8284$, $K_{b}=0.2071$
Locus of the complex branches
circle $(\sigma-2)^{2}+\omega^{2}=8$
(centre $+2$, radius $2.8284$)