Errata List For

Nonlinear Systems – Third Edition

Updated on September 16, 2010
Please e-mail error reports to [email protected]

Preface
1. Page xiv, Line 5: Change “books” to “book”
Chapter 1
1. Page 24, Second line of Exercise 1.4: Change “++” to “+”
2. Page 27, Exercise 1.13(c): Change “Section 1.3” to “Section 1.2.3”
3. Page 27, Exercise 1.14: Change “v the” to“v is the”

Chapter 2
1. Page 52, Last line: change “3” to “4”
2. Page 81, Exercise 2.14: Change “ ÿ ” to “ mÿ ”
3. Page 82, Exercise 2.18: Change “zg(z) > 0” to “zg(z) > 0 for z 6= 0”
Chapter 3
1. Page 105, Exercise 3.2(5): Change “Section 1.2.5” to “Section 1.2.6”

Chapter 4
1. Page 125, Figure 4.5: At the bottom right corner change “x2 = x1 ” to “x2 = −x1 ”
2. Page 128, Line 21: At the end of the sentence “When V (x) is positive definite, Ωc is bounded for
sufficiently small c.” insert the footnote “Ωc may have more than one component. This sentence refers
to the bounded component that contains the origin.”
3. Page 145, Lemma 4.3: Replace the last two sentences “If D = Rn . . . . . . class K∞ .” by “If D = Rn
and V (x) is radially unbounded, then there exist class K∞ functions α1 and α2 such that the foregoing
inequality holds for all x ∈ Rn .”
4. Page 167, Line 2: Change “Theorem 4.15” to “Theorem 4.14”

5. Page 176, Line 17: Change “consequences” to “consequence”

6. Page 176, Line 3 of the proof of Lemma 4.6: Change “satisfies (4.10) through (4.12)” to “satisfies the
inequalities of the theorem”
7. Page 182, Exercise 4.6: change “for all z ∈ R” to “for all z 6= 0”
8. Page 184, Exercise 4.19: change “roots” to “root”
9. Page 185, Exercise 4.21, part (b): Take D = Rn
10. Page 186, Line 4: Change “nonsingular” to “positive definite”
11. Page 188, Exercise 4.35, change“∀ r1 , r2 ∈ [0, a)” to “∀ r1 , r2 ∈ [0, a/2)”
Chapter 5
1. Page 198, Line 6: Change “uτ ” to “u”
2. Page 206, Line 16: Change “x0 ∈ R” to “x0 ∈ R2 ”

1
 3. Page 209, four lines before Section 5.3: Change “W3 (x) = −(1 − θ)kxk42 ” to “W3 (x) = (1 − θ)kxk42 ”
4. Page 218, Line 2 from the bottom: Change “competed” to “completed”

5. Page 219, Example 5.13: Change “e2 = ψ(t, y2 )” to “y2 = ψ(t, e2 )”

6. Page 224, Exercise 5.16, part (a): remove “finite-gain”
Chapter 6

1. Page 228, Line 10 of Section 6.1: Change “uy = Gy 2 ” to “uy = Gu2 ”

2. Page 232, Line 10 from the bottom: Change “Ly” to “Lu”
3. Page 238, Line 8: Change “of of” to “of”
4. Page 238, Lemma 6.1: The condition “limω→∞ ω 2 M T [G(jω) + GT (−jω)]M is positive definite” in the
third bullet is only necessary; it is not sufficient. A correct condition is given in the paper: Corless,
M. and Shorten, R., “On the Characterization of Strict Positive Realness for General Matrix Transfer
Functions”, IEEE Transactions on Automatic Control, Vol. 55, No. 8, pp. 1899–1904, 2010. The
paper shows the mistake in the proof of Lemma 6.1.
5. Page 239, Line 7 from the bottom: Change the (1,1) element of the matrix1 from “s” to “s + 1”
6. Page 239, Line 3 from the bottom: Change the (1,1) element of the matrix from “ω 2 ” to “ω 2 + 1”

7. Page 239, Last line: Change the (1,1) element of the matrix from “s/(s + 1)” to “(s + 2)/(s + 1)”
8. Page 240, Line 4: Change the (1,1) element of the matrix from “2ω 2 /(1 + ω 2 )” to “2(2 + ω 2 )/(1 + ω 2 )”
9. Page 241: First line of proof of Lemma 6.4: Change “V (s)” to “V (x)”
10. Page 244, Example 6.5: Change “f is locally Lipschitz, G and h are continuous” to “f and G are
locally Lipschitz, h is continuous”

11. Page 244, Line 6 from the bottom: Change “−ky T y” to “ky T y”
12. Page 245, Section 6.5: In describing the feedback connection of Figure 6.11 it should be noted that u1 ,
y1 , u2 , and y2 could be vectors of the same dimension.
13. Page 257, Line 4 to 6 from the bottom: Change “the condition
· ¸
1 + jωa 1 + (a − 1)ω 2
Re = > 0, ∀ ω ∈ R
1 − ω 2 + jω (1 − ω 2 )2 + ω 2

if a ≥ 1. Thus, choosing a ≥ 1, we can apply Lemmas 6.3 and 6.4 to conclude that” to “the conditions
· ¸
1 + jωa 1 + (a − 1)ω 2
Re = > 0, ∀ ω ∈ R
1 − ω 2 + jω (1 − ω 2 )2 + ω 2

and · ¸
1 + jωa
lim ω 2 Re =a−1>0
ω→∞ 1 − ω 2 + jω
if a > 1. Thus, choosing a > 1, we can apply Lemmas 6.1 and 6.4 to conclude that”
14. Page 260, Exercise 6.9, Line 4: Change V to V̇
Chapter 7
1. Page 299, Line 2: Change “with G(s)” to “with output e, zero input, and G(s)”
1 This G(s) and the one given at the end of the page need to be corrected because G(0) + GT (0) is singular.

2
Chapter 8
1. Page 311, Last line: Change “−kyk42 ” to “− 12 kyk42 ”

2. Page 312, Line 2: Change to “V̇ ≤ − 14 kyk42 , for kyk2 < 1

4k ”

3. Page 313, Line 2: Change “xpr in Suppose” to “xpr in Figure 8.1. Suppose”
4. Page 313, Line 8 from the bottom: Change “find tc by Figure 8.1. integrating” to “find tc by integrat-
ing”
Chapter 9
1. Page 342, Line 11 from the bottom: Change “V̇ (x)” to “V̇ (t, x)”
Chapter 11
1. Page 434, Equation (11.21) and Page 439, Equation (11.24): Change “ŷ(t/ε)” to “ŷ((t − t0 )/ε)”
Chapter 12
£ ¤ £ ¤
1. Page 482, Line 5: Change “K = K1 K2 ” to “K = K1 + K3 C K2 ”
2. Page 482, Line 15: Change “We note . . . K3 = 0” to “We note that in the stabilization of (xss , uss ) we
can take K3 = 0”
3. Page 483, Line before Equation (12.24): Change “ρ3 ” to “ρ2 ”
4. Page 490, Line 10: Change “v = α” to “ρ = α”
5. Page 492, Line 22: Change “from ψ to u” to “from ψ to u − M3 (α)e”
6. Page 495, second line after (12.57): Change “initited” to ”initiated”
Chapter 13
1. Page 517, Line 5 from the bottom: Change “reduces” to “reduces to”
2. Page 527, Line 8 from the bottom: Change “span(g, adf g, ad2f )” to “span(g, adf g, ad2f g)”

3. Page 533, five lines before Example 13.17: Change “ξ˙ = f0 (η, ξ)” to “η̇ = f0 (η, ξ)”
4. Page 535, Line 13: Change “Chapter 10” to “Chapter 9”
Chapter 14
1. Page 553, Line 5: It should be noted that u = −β(x)sgn(s) is used only for s 6= 0 since in ideal sliding
mode control u is not defined on the sliding surface s = 0. Alternatively, we can write u = −β(x)sgn(s)
for all s if sgn(s) is not defined at s = 0. The same remark applies throughout the chapter to ideal
sliding mode control.
√ √
2. Page 553, Line 10: Change “W = V ” to “W = 2V ”
3. Page 553, Line 15: Change “V̇ ≤ −2g0 β0 |s|” to “V̇ ≤ −g0 β0 |s|”
4. Page 568, Line 3 of Theorem 14.2: Change “stale” to “stable”
5. Page 573, last line and Page 574, Lines 4, 7, and 9 from the bottom: Change “rρ ” to “r(ρ) ”
6. Page 578: To show the inequality satisfied by V̇0 , we need the additional condition:
¯ ¯
¯ ∆(x , v , w, r) − ∆(x , v , w, r) ¯
¯ ss 1 ss 2 ¯
¯ ¯ ≤ `|v1 − v2 |, 0≤`<1
¯ Lg Lfρ−1 h(xss , w) ¯

for all (v1 , v2 , w, r) ∈ R × R × Dw × Dr .

3
 7. Page 587, Line 9: Change “k0 =” to “k0 ≥” and “k =” to “k ≥”
8. Page 611, Line 14: Change “designing C” to “designing H”

9. Page 622, Theorem 14.6: In the fourth bullet of the theorem add the requirement that f (X ) is contin-
uously differentiable in some neighborhood of X = 0.
10. Page 625, Line 18: Change “eρ ” to “êρ ”

11. Page 644, Exercise 14.56(d): Change the units of I from “Kg/m2 ” to “Kg m2 ”
Appendix

1. Page 662, Line 14: Change “If D = Rn ,” to “If D = Rn and V (x) is radially unbounded,”
2. Page 662, Line 18: Change “If V (x)” to “Because V (x)”
3. Page 693: Replace lines 18 to 24 (starting with “This expression . . . ” and ending with “. . . parameterized
in η.”) by the following:2
This expression is valid for any t ∈ R. The limit of the integral term as t → −∞ is
Z 0
exp(−Bs)G(π(s; y(τ ), η), η(π(s; y(τ ), η))) ds (C.60)
−∞

which is well defined because η is bounded, G is globally bounded in π, and B is Hurwitz. Let us
rewrite the expression (C.60) with y(τ ) replaced by y and denote it by (P η)(y).
Z 0
(P η)(y) = exp(−Bs)G(π(s; y, η), η(π(s; y, η))) ds (C.61)
−∞

With this definition, we can write

exp [B(t − τ )] [z(τ ) − (P η)(y(τ ))] =

Z t−τ
z(t) − exp [−B(s − t + τ )]G(π(s; y(τ ), η), η(π(s; y(τ ), η))) ds
−∞

Substituting ξ = s − t + τ in the integral and using π(ξ + t − τ ; y(τ ), η) = π(ξ; y(t), η), we obtain

exp [B(t − τ )] [z(τ ) − (P η)(y(τ ))] = z(t) − (P η)(y(t))

which shows that if z(τ ) = (P η)(y(τ )), then z(t) = (P η)(y(t)) for all t ∈ R. Hence, z = (P η)(y)
defines an invariant manifold for (C.58)–(C.59) parameterized in η.
4. Page 695, Line 2: Change “, η((π(t; y, η)))” to “, η(π(t; y, η)))”
5. Page 698, Equation (C.66): Change “A1 +” to “A1 y +”

6. Page 716, Lines 17 and 18: Change “U (x)” to “U (X )”

Bibliography
1. Page 725, Ref.[9]: Change “Anstaklis” to “Antsaklis”
2. Page 731, Ref.[93]: Change “1995” to “1994”

3. Page 732, Ref.[109]: Change “Trans. Amer. Math. Soc., 24:19–17” to “Amer. Math. Soc. Transl.,
Ser. 2, 24:19–77”
2 The need for this change in the proof was pointed out by Professor H. Shim. The original proof implicitly used
limt→−∞ exp [−B(t − τ )]z(t) = 0, which cannot be guaranteed at this point in the proof because we cannot guarantee that z(t)
is bounded as t → −∞.

4
Index
1. Page 744, Feedback Passivation: Change “607” to “606”

In the third print of the book, a new typo appeared on page 553, line 13: Change “dotV ” to “V̇ ”