ECE486: Control Systems

I Lecture 6C Routh–Hurwitz stability criterion

Goal: formulate and learn how to apply the Routh–Hurwtiz

stability criterion.
Checking for Stability?

Consider a general transfer function:

q(s)
H(s) =
p(s)

where q and p are polynomials, and deg(q) ≤ deg(p).

We need tools for checking stability: whether or not all roots of
p(s) = 0 lie in OLHP.
For simple polynomials, can just factor them “by inspection”
and find roots.
Now, this is hard to do for high-degree polynomials — it’s
computationally intensive, especially symbolically.
But: often we don’t need to know precise pole locations, just
need to know that they are strictly stable.
Checking for Stability

Problem: given an nth-degree polynomial

p(s) = sn + a1 sn−1 + a2 sn−2 + . . . + an−1 s + an

with real coefficients, check that the roots of the equation

p(s) = 0 are strictly stable (i.e., have negative real parts).

Terminology: we often say that the polynomial p is (strictly)

stable if all of its roots are.
A Necessary Condition for Stability

Terminology: we say that A is a necessary condition for B if

A is false =⇒ B is false

Important!! Even if A is true, B may still be false.

Necessary condition for stability: a polynomial p is strictly

stable only if all of its coefficients are strictly positive.

Proof: suppose that p has roots at r1 , r2 , . . . , rn with Re(ri ) < 0

for all i. Then

p(s) = (s − r1 )(s − r2 ) . . . (s − rn )

— multiply this out and check that all coefficients are positive.
Routh–Hurwitz Criterion
Necessary & Sufficient Condition for Stability

Terminology: we say that A is a sufficient condition for B if

A is true =⇒ B is true

Thus, A is a necessary and sufficient condition for B if

A is true ⇐⇒ B is true

— we also say that A is true if and only if (iff) B is true.

We will now introduce a necessary and sufficient condition for

stability: the Routh–Hurwitz Criterion.
Routh–Hurwitz Criterion: A Bit of History
J.C. Maxwell, “On governors,” Proc. Royal
Society, no. 100, 1868
... [Stability of the governor] is mathematically
equivalent to the condition that all the possible
roots, and all the possible parts of the impossible
roots, of a certain equation shall be negative. ...
I have not been able completely to determine
these conditions for equations of a higher degree
than the third; but I hope that the subject will
obtain the attention of mathematicians.

In 1877, Maxwell was one of the judges for the Adams Prize, a
biennial competition for best essay on a scientific topic. The
topic that year was stability of motion. The prize went to
Edward John Routh, who solved the problem posed by Maxwell
in 1868.
In 1893, Adolf Hurwitz solved the same problem, using a
different method, independently of Routh.
Edward John Routh, 1831–1907 Adolf Hurwitz, 1859–1919
Routh’s Test

Problem: check whether the polynomial

p(s) = sn + a1 sn−1 + a2 sn−2 + . . . + an−1 s + an

is strictly stable.

We begin by forming the Routh array using the coefficients of p:

sn : 1 a2 a4 a6 . . . (if necessary, add zeros in the

sn−1 : a1 a3 a5 a7 . . . second row to match lengths)

Note that the very first entry is always 1, and also note the
order in which the coefficients are filled in.
Routh’s Test

sn : 1 a2 a4 a6 ...
sn−1 : a1 a3 a5 a7 ...
sn−2 : b1 b2 b3 . . .
Next, we form the third row marked by sn−2 :

sn−2 : b1 b2 b3 ...
 
1 1 a2 1
where b1 = − det =− (a3 − a1 a2 )
a1 a1 a3 a1
 
1 1 a4 1
b2 = − det =− (a5 − a1 a4 )
a1 a1 a5 a1
 
1 1 a6 1
b3 = − det =− (a7 − a1 a6 ) and so on ...
a1 a1 a7 a1

Note: the new row is 1 element shorter than the one above it
Routh’s Test, continued

sn : 1 a2 a4 a6 . . .
sn−1 : a1 a3 a5 a7 . . .
sn−2 : b1 b2 b3 ...
sn−3 : c1 c2 . . .
Next, we form the fourth row marked by sn−3 :

sn−3 : c1 c2 . . .
 
1 a1 a3 1
where c1 = − det = − (a1 b2 − a3 b1 )
b1 b1 b2 b1
 
1 a1 a5 1
c2 = − det = − (a1 b3 − a5 b1 )
b1 b1 b3 b1
and so on ...
Routh’s Test, continued

Eventually, we complete the array like this:

sn : 1 a2 a4 a6 . . .
sn−1 : a1 a3 a5 a7 . . .
sn−2 : b1 b2 b3 . . . (as long as we don’t get stuck with
sn−3 : c1 c2 . . .
.. division by zero: more on this later)
.
s1 : ∗ ∗
s0 : ∗
After the process terminates, we will have n + 1 entries in the
first column.
The Routh–Hurwitz Criterion
Consider degree-n polynomial
p(s) = sn + a1 sn−1 + . . . + an−1 s + an
and form the Routh array:
sn : 1 a2 a4 a6 ...
n−1
s : a1 a3 a5 a7 ...
sn−2 : b1 b2 b3 ...
sn−3 : c1 c2 ...
..
.
s1 : ∗ ∗
s0 : ∗

The Routh–Hurwitz criterion: Assume that the necessary

condition for stability holds, i.e., a1 , . . . , an > 0. Then:
I p is stable if and only if all entries in the first column are
positive;
I otherwise, #(RHP poles) = #(sign changes in 1st column)
Example
Check stability of

p(s) = s4 + 4s3 + s2 + 2s + 3

All coefficients strictly positive: necessary condition checks out.

s4 : 1 1 3
s3 : 4 2 0
s2 : 1/2 3
s1 : −22 0
s0 : 3

Answer: p is unstable — it has 2 RHP poles (2 sign changes in

1st column)
Low-Order Cases (n = 2, 3)
n=2 p(s) = s2 + a1 s + a2
s2 : 1 a2
s1 : a1 0  
1 a2
s0 : b1 b1 = − a11 det = a2
a1 0

— p is stable iff a1 , a2 > 0 (necessary and sufficient).

n=3 p(s) = s3 + a1 s2 + a2 s + a3
s3 : 1 a2
s2 : a1 a3  
1 a2
s1 : b1 0 b1 = − a11
det = a1 aa21−a3
a1 a3 
1 a1 a3
s0 : c1 c1 = − b1 det = a3
b1 0

— p is stable iff a1 , a2 , a3 > 0 (necc. cond.) and a1 a2 > a3

Stability Conditions for Low-Order Polynomials

The upshot:
I A 2nd-degree polynomial p(s) = s2 + a1 s + a2 is stable if
and only if a1 > 0 and a2 > 0
I A 3rd-degree polynomial p(s) = s3 + a1 s2 + a2 s + a3 is
stable if and only if a1 , a2 , a3 > 0 and a1 a2 > a3

I These conditions were already obtained by Maxwell in

1868.
I In both cases, the computations were purely symbolic: this
can make a lot of difference in design, as opposed to
analysis.
Routh–Hurwitz as a Design Tool
Parametric stability range

We can use the Routh test to determine parameter ranges for

stability.
Example: consider the unity feedback configuration

+ s+1
R K s3 + 2s2 s Y
controller
plant

Note that the plant is unstable (the denominator has a negative

coefficient and a zero coefficient).
Problem: determine the range of values the scalar gain K can
take, for which the closed-loop system is stable.
Example, continued
+ s+1
R K s3 + 2s2 s Y

Problem: determine the range of values the scalar gain K can

take, for which the closed-loop system is stable.
Let’s write down the transfer function from R to Y :
Y forward gain
=
R 1 + loop gain
s+1
K · s3 +2s2 −s K(s + 1)
= s+1 = 3 2
1 + K · s3 +2s2 −s s + 2s − s + K(s + 1)
Ks + K
= 3
s + 2s2 + (K − 1)s + K

Now we need to test stability of p(s) = s3 + 2s2 + (K − 1)s + K.

Example, continued
Test stability of

p(s) = s3 + 2s2 + (K − 1)s + K

using the Routh test.

Form the Routh array:

s3 : 1 K −1
s2 : 2 K
K
s1 : 2 −1 0
s0 : K

For p to be stable, all entries in the 1st column must be positive:

K>2 and K > 0 (already covered by K > 1)

Note: The necessary condition requires K > 1, but now we

actually know that we must have K > 2 for stability.
Some Comments on the Routh Test

I The result (#(RHP roots)) is not affected if we multiply or

divide any row of the Routh array by an arbitrary positive
number.
I If we get a zero element in the 1st column, we can’t
continue. In that case, we can replace the 0 by a small
number ε and apply Routh test to that. When we are done
with the array, take the limit as ε → 0. (see Ex. 3.33 in
FPE)
I For an entire row of zeros, the procedure is a more
complicated (see Example 3.34 in FPE) – we will not worry
about this too much.