Chinese

Journal of
Chinese Journal of Aeronautics 21(2008) 162-168
Aeronautics
www.elsevier.com/locate/cja

Integrated Guidance and Control of Homing Missiles Against

Ground Fixed Targets
Hou Mingzhe*, Duan Guangren
Center for Control Theory and Guidance Technology, Harbin Institute of Technology, Harbin 150001, China

Received 19 October 2007; accepted 18 December 2007

Abstract

This paper presents a scheme of integrated guidance and autopilot design for homing missiles against ground fixed targets. An in-
tegrated guidance and control model in the pitch plane is formulated and further changed into a normal form by nonlinear coordinate
transformation. By adopting the sliding mode control approach, an adaptive nonlinear control law of the system is designed so that the
missile can hit the target accurately with a desired impact attitude angle. The stability analysis of the closed-loop system is also con-
ducted. The numerical simulation has confirmed the usefulness of the proposed design scheme.

Keywords: integrated guidance and control; pitch plane; ground fixed target; nonlinear; adaptive

1 Introduction1 siles; Palumbo et al[2] used the state-dependent Ric-

cati equation technique to deal with a more com-
The traditional way to design a missile guid-
prehensive model characterized by nonlinear motion
ance and control system is to form subsystems
in three dimensions; Shkolnikov et al[3] and Shima
separately followed by integrating them. Although
et al[4-5] developed an integrated guidance and
this approach proves successful so far in building up
control design by using sliding mode control; Choi
a number of missile guidance and control systems
and Chwa[6-7] proposed an adaptive nonlinear
with outstanding performance, it is arguable that the
guidance law considering the target uncertainties
overall system performance is constrained because
and control loop dynamics by way of sliding mode
synergistic relationships between the interacting
control approach; Sharma and Richards[8] intro-
subsystems are not fully exploited.
duced an integrated guidance and control design for
In order to improve the whole system perfor-
homing missiles based on the backstepping strategy.
mance, a new approach known as the integrated
Still, there were suggested other interesting and ef-
guidance and control (IGC) system design method is
fective methods such as subspace stabilization
developed by incorporating various control theories.
method[9] and T -D method[10]. Therefore, it stands
Menon and Ohlmeyer[1] used the feedback lineari-
to reason that the integrated guidance and control
zation method associated with the linear quadratic
system design will be one of the most noticed
regulator technique to formulate a nonlinear inte-
subjects in missile technology.
grated guidance and control laws for homing mis-
This work focuses on developing integrated
guidance and control for homing missiles against
*Corresponding author. Tel.: +86-10-86418024.
E-mail address: houlechuan@126.com ground fixed targets, which usually requires the
Foundation items: Major Program of National Natural Science Founda-
tion of China (60710002); Program for Changjiang minimum miss-distance and the desired impact atti-
Scholars and Innovative Research Team in University
 Hou Mingzhe et al. / Chinese Journal of Aeronautics 21(2008) 162-168 · 163 ·

tude angle to enable the warheads of the missile to spectively.

acquire better killing effects[11]. To achieve this, Zha
et al[12] gave out the guidance law design by taking
into account the dynamics of the missile control
system. Different from this idea, this paper firstly
establishes the model of the integrated guidance and
control loop in the pitch plane, and then directly
gives the fin deflection command by adopting the
sliding mode control approach. Hence, the proposed
method deserves more to be called an integrated
guidance and control system design method.
Fig.1 Two-dimensional engagement geometry.

2 Model Derivation By differentiating the Eq.(2b) and substituting

Eqs.(2a)-(2b) into the derivative, it is easy to obtain
This section presents the model derivation of that
the integrated guidance and control system in the ª V 2 R º R 
pitch plane. q «  » q  T (3)
¬V R ¼ R
Linearized missile dynamics in the pitch plane
As shown in Fig.2, the flight path angle is re-
is described by
lated with the pitch angle and the attack angle as
57.3QScDy  P
½ T - D (4)
D Zz  D  'D
°
mV ° or equivalently
2 Zz
QSL mz D
57.3QSLmz °° T -  D (5)
Z z Zz  D ¾ (1)
J zV Jz ° where - is the pitch angle, which satisfies
57.3QSLmz Gz ° - Z z
G z  'Z °
Jz °¿

where D is the attack angle, Z z the angular pitch

rate, G z the deflection angle for pitch control, Q
the dynamic pressure, S the aerodynamic refer-
ence area, J z the moment of inertia about z-axis,
m the missile mass, P the thrust of the missile, L the
reference length, V the velocity of the missile, cDy
the lift force derivative with respect to the attack
Fig.2 Schematic of angular relationship.
angle, mDz , mZz and mGz represent the pitch
z z

Considering that the terminal attack angle is

moment derivatives with respect to the attack angle,
very small in the proposed design scheme, it is easy
the nondimensional angular pitch rate and the de-
to obtain from Eq.(4) that
flection angle for pitch control, respectively, 'D T d | -d
and 'Z are unknown bounded uncertainties.
where T d and -d are the desired impact flight
As shown in Fig.1, dynamics of the missile to
and the desired impact attitude angles, respectively.
the ground fixed target range and the line-of-sight
Assume that qd is the desired impact LOS
(LOS) angle are governed by
angle at the engagement time, the following equa-
R Vcos(q  T ) (2a)
tion and inequality hold true
Rq Vsin(q  T ) (2b) ­ Rq Vsin(qd  T d ) 0
°
where R, q and T are the relative distance, the LOS ® ʌ
°¯ qd  T d  2
angle and the flight path angle of the missile, re-
· 164 · Hou Mingzhe et al. / Chinese Journal of Aeronautics 21(2008) 162-168

Thus Let x1 V and x2 V , then it is easy to obtain

qd T d x1 x2
and hence and
qd | -d x2 a22V  a23D  '2 x3  '2
Let where x3 a22V  a23D and '2 '2 .
V q  qd Hence
then differentiation yields 1 1
D a23 x3  a23 a22 x2 (8)
V q (6)
Differentiating the term x3 yields
According to the above analysis, the integrated x3 a22 x2  a23D  a22V  a23D a22 x2 
guidance and control model of the homing missile a23D  a22 x3  a23 (a33D  Z z )=(a22 
in the pitch plane can be written into 1 1 
a23 a22 a23  a22 a33 )x2  (a22  a23 a23 
ª V º ª0 1 0 0 ºªV º ª 0 º ª 0 º½
« V » «0 a » « » « » « ° a33 )x3  a23Z z  a22 '2  a23'3 x4  '3
a23 0 » « V » « 0 » '»
« » « 22
 u  « 2 »° where
« D » «0 0 a33 1 » « D » « 0 » « '3 » °°
« » « »« » « » « » ¾ (7) x4 1
(a22  a23 a22 a23  a22 a33 )x2 
¬Z z ¼ ¬0 0 a43 a44 ¼ ¬Z z ¼ ¬b4 ¼ ¬ '4 ¼ °
1 
ªV º ° (a  a a  a )x  a Z
22 23 23 33 3 23 z
y « » °
¬V ¼ °¿ and

V 2R  R '3 a22 '2  a23 '3
where a22  , a23  (57.3QScDy  P) ,
V R mVR hence
1 1 2  1
1 57.3QSL D Zz a23 x4  (a23 a22  a23 a23  a23 a33 )x3 
a33  (57.3QScDy  P ), a43 mz , a44 (9)
mV Jz 1  2
(a a  a a a  a a a )x  1
23 22 23 22 23 23 22 33 2
QSL2 Z 57.3QSL G R Similarly,
mz , b4
z
mz , u G z , '2
z
'D ,
J zV Jz R x4 a2 x2  a3 x3  a4 x4  b4 u  '4
'3 'D , '4 'Z . where
The object of this paper is to design an appro- 1
a2 a22  a23 1  
a22 a23  a23 2
a22 a23  2a23 2
a22 a23 
priate control law for the integrated guidance and 1  
a22 a33  a22 a33  a23 1 
a22 a23  a23 a23 a22 a33 
control system Eq.(7), which not only guarantees a  1 
a a a a a a a a a a a
44 22 23 22 23 44 22 33 44 22 43
stable resulted closed-loop system but also acquires 2  2 1  1 
a3 2a22  2a23 a23  a23 a23  a33  a23 a23 a22  a43 
a zero miss-distance together with a desired impact
2a 1a a  a a  a a  a 1a a  a a
attitude angle. In the missile engagement, the null 23 23 33 22 33 44 22 23 44 23 44 33
1 
rate of rotation of the LOS q presents the ideal a4 2a23 a23  a22  a33  a44
condition for zero miss-distance, and the terminal b4 a23b4
1 1 
impact attitude angular constraint can be satisfied '4 (a22  a23 a22 a23  a22 a33 )'2  (a22  a23 a23 
by the null LOS angle error V q  qd . a33 )'3  a23 '4
Let x [ x1 x2 x3 x4 ]T
3 Adaptive Nonlinear Control Law Design then
This section presents the process of designing ª0 1 0 0 º ª0º ª 0 º½
«0 0 1 0 » «0» « »°
the control law for the integrated guidance and con-
x « » x  « » u  « '2 » °
trol system described by Eq.(7). «0 0 0 1 » «0» « '3 » °°
« » « » « »¾ (10)
Because the system Eq.(7) is a time-varying ¬ 0 a2 a3 a4 ¼ ¬b4 ¼ ¬ '4 ¼ °
ª1 0 0 0 º °
system, based on the idea in Ref.[6], it is changed y «0 1 0 0 » x °
into a normal form with the help of nonlinear coor- ¬ ¼ ¿°
dinate transformation and the sliding mode control Clearly, in order to attain the predetermined
approach is used to design its feedback control law. objective, it is only needed to design an appropriate
 Hou Mingzhe et al. / Chinese Journal of Aeronautics 21(2008) 162-168 · 165 ·

control law for the system Eq.(10) so that the re- Ș ȌȘ  Ȟ

sulted closed-loop system is stable with the outputs if Ȍ is a Hurwitz matrix and Ȟ converges to zero
converging to zero asymptotically. asymptotically, then Ș also converges to zero as-
The adaptive nonlinear control law is designed ymptotically.
under the following assumption. Propositions 1 and 2 can be easily confirmed
Assumption 'i d Ti (i=2,3,4), '2 d N 2 , from the conclusions in Ref.[13], so the proofs are
'3 d N3 and '2 d W2 , where Ti, N2, N3 and W2 omitted. From Proposition 2, the following proposi-
tion can be obtained.
are all unknown constants.
Proposition 3 If the coefficients in the ordi-
Let M 2 T2 , M 3 N 2  T3 , M 4 W2  N3  T4 ,
nary differential equation
then based on the sliding mode control theory, the
e1[  e2[  "  en[ ( n 1)  [ ( n ) X
proposed adaptive nonlinear control law can be ex-
satisfy the Hurwitz polynomial
pressed as
g ( U ) e1  e2 U  "  en U n 1  U n
u b41[(c1  a2 ) x2  (c2  a3 ) x3  (c3  a4 ) x4 
and X converges to zero asymptotically, then [ ( m )
ks  (c Mˆ  c Mˆ  Mˆ )sgn( s )]
2 2 3 3 4 (11)
(m=0, " ,ní1) converge to zero asymptotically, too.
with the adaptation law described by Next, the Proof of Theorem is given.
  
M̂ 2 c2 O2 s , M̂ 3 c3O3 s , M̂ 4 O4 s Proof of Theorem Let
where k ! 0 , O2 ! 0 , O3 ! 0 , O4 ! 0 are design M i M i  Mˆ i
parameters, Mˆ i is an estimate of M i (i=2,3,4), then
 
and M i  Mˆ i
s c1 x1  c2 x1  c3 
x1  
x1 where i=2,3,4.
with the design parameters c j , where j 1,2,3 , Define the Lyapunov function as
satisfying the Hurwitz polynomial 1 2 1 2 1 2 1 2
E (s  M 2  M 3  M 4 )
g ( U ) c1  c2 U  c3 U 2  U 3 2 O2 O3 O4
Remark In fact, it is easy to check that the and take a time derivative to have
inequality b4 z 0 always holds true, so Eq.(11) is 1  1  1 
E ss  M 2 M 2  M 3 M 3  M 4 M 4
always nonsingular. O2 O3 O4
Then the following theorem can be deducted. s[c1 x2  c2 ( x3  '2 )  c3 ( x4  '2  '3 )  a2 x2 
Theorem The system described by Eq.(10) 1  
a3 x3  a4 x4  b4u  '2  '3  '4 ]  M 2M 2 
and Eq.(11) is stable in the sense that its states are O2
1   1
all bounded and its outputs converge to zero asymp- M 3 M 3  M 4 M 4 s[c2 '2  c3 ('2  '3 ) 
O3 O4
totically under Assumption.
Before conducting the proof, some useful lem- ('2  '3  '4 )  (c2 Mˆ 2  c3 Mˆ 3  Mˆ 4 )sgn( s)] 
mas and propositions are introduced as follows. s (c M  c M  M )  ks 2 d  ks 2  s [c ˜
2 2 3 3 4 2
Lemma (Babalat’s lemma) If f (t ) is a ( M 2  Mˆ 2 )  c3 ( M 3  Mˆ 3 )  ( M 4  Mˆ 4 )] 
t
uniformly continuous function and lim ³ f (W )dW s (c M  c M  M )  ks 2
2 2 3 3 4
t of 0

exists, then f (t ) converges to zero asymptotically. From the above computation, it can be concluded
Proposition 1 For the system that E  Lf and
z ĭz  d f 2 1
³ 0 s dt d k [ E (0)  E (f)]  f
where ĭ is a Hurwitz matrix, if there exists such a
positive scalar N1 that d d N1 , then there exists that is s 2  L2 .
such a positive scalar N 2 that z d N 2 . From E  Lf , it can be obtained that M 2  Lf ,
M 3  Lf , M 4  Lf and s  Lf , and accordingly,
Proposition 2 For the system
· 166 · Hou Mingzhe et al. / Chinese Journal of Aeronautics 21(2008) 162-168

M̂ 2  Lf , M̂ 3  Lf and M̂ 4  Lf . In addition, a The simulation step is set to be 0.001 s, the de-

simple calculation gives
s d k s  c2 ( M 2  Mˆ 2 )  c3 ( M 3  Mˆ 3 )  ( M 4  Mˆ 4 )
so s  Lf . Thus, ds 2 / dt  Lf . This means that s 2
is uniformly continuous. Combining this with the
L2-property of s 2 , Lemma can be used to conclude
that s 2 converges to zero asymptotically. Hence,
s converges to zero asymptotically. Using Proposi-
tion 3, it can be concluded that x1 and x2 con-
verge to zero asymptotically, that is, the outputs of
the system described by Eqs.(10)-(11) converge to
zero asymptotically.
In addition, the closed-loop system described
by Eqs.(10)-(11) can be rewritten into
sired impact pitch angle í90°, the position coordi-
nate of the target (2 000, 0) m,; the initial velocity
of the missile Ma=0.6, the initial body rotational
rate of the missile 0 rad/s, the initial flight path an-
gle of the missile 0°, the initial attitude angle of the
missile 0°, and the initial position coordinate of the
missile (0, 2 500) m. Design parameters of control
law Eq.(11) are set to be O2 O3 O4 2.7 u 105 ,
c1 9, c2 11, c3 8 and k 0.105 .
In addition, the deflection angle and the de-
flection angle rate for pitch control used in all
simulations are assumed to satisfy G z d 30°,
G d 3 (°)/s.
z

x Ac x  F (12) For the nominal system, Fig.3 shows the tra-

where jectory of the missile; Fig.4 the curve of the relative
ª 0 1 0 0 º distance between the missile and the target with the
« 0 0 1 0 »
« » terminal relative distance only 0.039 m; Fig.5 the
Ac
« 0 0 0 1 » curve of the pitch angle with the desired impact
« »
¬ kc1 (c1  kc2 ) (c2  kc3 ) (c3  k ) ¼ pitch angle í89.948°; Fig.6 the small attack angle at
F [0 f2 f3 f 4 ]T the engagement time as mentioned in Section 2, and
f2 '2 , f3 '3 Fig.7 the curve of the deflection angle for pitch con-
f4 (c2 Mˆ 2  c3 Mˆ 3  Mˆ 4 )sgn( s)  trol.
k[c '  ('  ' )]  '
3 2 2 3 4

From Assumption and the above description,

F  Lf holds true, which means that there exists
such a positive number F that F d F . In addi-
tion, under the conditions that k ! 0 and that the
polynomial
g ( U ) c1  c2 U  c3 U 2  U 3
is Hurwitzian, it can be concluded that Ac is a
Hurwitz matrix. Therefore, Proposition 1 can be
Fig.3 The trajectory of the missile.
used to conclude that the states of the system Eq.(12)
are all bounded, that is, the states of the system de-
scribed by Eqs.(10)-(11) are all bounded. Here the
proof is completed.

4 Numerical Simulations
In this section, the feasibility and applicability
of the proposed integrated guidance and control
logic is verified by the numerical simulations for
some passive homing missile’s nonlinear dynamic Fig.4 The curve of the relative distance.
model in the pitch plane.
 Hou Mingzhe et al. / Chinese Journal of Aeronautics 21(2008) 162-168 · 167 ·

guidance and autopilot design for homing missiles

against ground fixed targets to improve the per-
formance of the missile guidance and control system.
An integrated model of guidance and control loop in
the pitch plane is first formulated, and then the
adaptive nonlinear control law is designed by
adopting the sliding mode control approach. Nu-
merical simulation results have confirmed a simul-
Fig.5 The curve of the pitch angle.
taneous attainment of small miss-distance and a de-
sired impact attitude angle, which demonstrates the
usefulness of the proposed design scheme.

References
[1] Menon P K, Ohlmeyer E J. Nonlinear integrated guidance-control

laws for homing missiles. AIAA-2001-4160, 2001.

[2] Palumbo N F, Jackson T D. Integrated missile guidance and con-

trol: a state dependent Riccati differential equation approach. Pro-

ceedings of the 1999 IEEE International Conference on Control

Fig.6 The curve of the attack angle.
Applications. 1999; 243-248.

[3] Shkolnikov I, Shtessel Y. Integrated guidance-control system of a

homing interceptor: sliding mode approach. AIAA-2001-4218,

2001.

[4] Shima T, Idan M, Golan O M. Sliding mode control for integrated

missile autopilot-guidance. AIAA-2004-4884, 2004.

[5] Idan M, Shima T, Golan O M. Integrated sliding mode autopi-

lot-guidance for dual control missiles. AIAA-2005- 6455, 2005.

[6] Choi J Y, Chwa D Y. Adaptive nonlinear guidance considering

Fig.7 The curve of the deflection angle. target uncertainties and control loop dynamics. IEEE Transactions

In order to show the robust characteristics of on Aerospace and Electronic Systems 2003; 39(4): 1139-1143.

the proposed method, two cases are selected as fol- [7] Chwa D Y, Choi J Y. Observer-based adaptive guidance law con-

lows: sidering target uncertainties and control loop dynamics. IEEE

(1) All of the aerodynamic parameters are as- Transactions on Control Systems Technology 2006; 14(1): 112-

sumed to increase by 20% relative to the nominal 123.

values. The terminal relative distance is 0.024 m [8] Sharma M, Richards N. Adaptive, integrated guidance and control

and the terminal impact pitch angle í88.272°. for missile interceptors. AIAA-2004-4880, 2001.

(2) All of the aerodynamic parameters are all [9] Tournes C, Wilkerson P. Integrated terminal guidance and auto-

assumed to decrease by 20% relative to the nominal matic pilot using subspace stabilization. AIAA-2001-4275, 2001.

values. The terminal relative distance is 0.248 m [10] Xin M, Balakrishnan S N, Ohlmeyer E J. Integrated guidance and

and the terminal impact pitch angle í91.893°. control of missiles with T -D method. IEEE Transactions on Con-

All of the simulation results testify to the fea- trol Systems Technology 2006; 14(6): 981-992.

sibility and practicability of the proposed method. [11] Song J M, Zhang T Q. Passive homing misile’s variable structure

proportional navigation with terminal angular constraint. Chinese

5 Conclusions Journal of Aeronautics 2001; 14(2): 83-87.

[12] Zha X, Cui P Y, Chang B J. An integrated approach to guidance

This paper presents a scheme of integrated
· 168 · Hou Mingzhe et al. / Chinese Journal of Aeronautics 21(2008) 162-168

and control for aircraft applying to attack the ground fixed targets.
Journal of Astronautics 2005; 26(1): 14-18. [in Chinese]
[13] Khalil H K. Nonlinear systems. 3rd ed. New Jersey: Prentice Hall,
1996.
Biographies:
Hou Mingzhe Born in 1981, he is now working toward the
Ph. D. degree in the Center for Control Theory and Guidance
Technology at Harbin Institute of Technology. His main
research interests include control of nonlinear systems and
integrated guidance and control for aircrafts.
E-mail: houlechuan@126.com
Duan Guangren Born in 1962, he received Ph. D. degree
from Harbin Institute of Technology in 1989, and became a
professor in 1991. In August 2000, he was elected to be spe-
cially-hired professor at Harbin Institute of Technology
sponsored by the Cheung Kong Scholars Program of the
Chinese government. His main research interests include
robust control, linear systems, control of aircrafts and mag-
netic bearing control.
E-mail: g.r.duan@hit.edu.cn