You are on page 1of 39

SLIDING-MODE IMPACT TIME GUIDANCE LAW

DESIGN FOR INTERCEPTING VARIOUS TARGET


MOTIONS

A MINI PROJECT REPORT

submitted by

VISHNU V S
TVE18EEGN18

to

the A P J Abdul Kalam Kerala Technological University

in partial fulfillment of the requirements for the award of the Degree

of

Master of Technology
In
Guidance and Navigational Control

Department of Electrical Engineering


College of Engineering Trivandrum

MAY 2019
DEPARTMENT OF ELECTRICAL ENGINEERING
COLLEGE OF ENGINEERING TRIVANDRUM
THIRUVANANTHAPURAM - 695 016

CERTIFICATE

This is to certify that this report entitled “Sliding-Mode Impact Time Guid-
ance Law Design for Intercepting Various Target Motions” submitted by
VISHNU V S (Roll No. TVE18EEGN18) to A P J Abdul Kalam Kerala Techno-
logical University towards partial fulfilment of requirements for the award of degree
of Master of Technology in Guidance and Navigational Control is a bonafide record
of the mini project work carried out by him under our guidance and supervision.

Dr. Sreeja. S Dr. P. Sreejaya


Project Coordinator P G Coordinator
Assistant Professor Professor
Dept. of Electrical Engineering Dept. of Electrical Engineering
College of Engineering Trivandrum College of Engineering Trivandrum

Dr. S. Ushakumari
Professor & Head
Dept. of Electrical Engineering
College of Engineering Trivandrum

i
Acknowledgements
A lot of efforts and hard work has been put into the completion of this project work.
However, it would not have been possible without the kind support and help of
many individuals and other sources. I would like to extend my sincere thanks to all
of them. I take this opportunity to express my deep sense of gratitude and sincere
thanks to all who helped me to complete this seminar report successfully.
I thank God Almighty for guiding me throughout the project. I thank Dr. Jiji C.V.,
Principal, College of Engineering Trivandrum for his support and cooperation. I ex-
press my sincere gratitude to Dr. S. Ushakumari, Head of the Department, Depart-
ment of Electrical Engineering and Dr.P. Sreejaya, P G Coordinator, Department
of Electrical Engineering for their inspiring guidance and unstained support.
I am greatly thankful to Prof. R.Harikumar, Stream Head, Guidance and Navi-
gational Control, Department of Electrical Engineering, for his supervision, assis-
tance and helpful suggestions. I deeply indebted to Dr.Arun Kishor W C, Pro-
fessor, Department of Electrical Engineering and Dr.Sreeja. S, Assistant Profes-
sor,Department of Electrical Engineering, for their excellent guidance, positive crit-
icism and valuable comments. Finally I thank to my parents and friends near and
dear ones who directly and indirectly contributed to the successful completion of the
project.

VISHNU V S

ii
Abstract
The Proportional Navigation Guidance (PNG) law has been widely used the missile
guidance to achieve minimal miss distance, simple structure, and effectiveness. How-
ever, advanced homing guidance laws with impact angle or impact time constraint
aiming to raise kill probability have received considerable attention recently. In
this project, a sliding-mode guidance laws with impact time constraint are proposed
for intercepting various target motions, including nonmaneuvering and maneuvering
targets. To achieve the desired impact time, the line-of-sight (LOS) dynamics is con-
trolled to track a novel time-varying LOS profile created by the LOS rate-shaping
approach. To nullify the LOS tracking error, the terminal sliding-mode technique
is used to obtain a finite time convergent guidance law against nonmaneuvering
targets, and it is combined with the inertial delay control to derive an ultimately
bounded guidance law against maneuvering targets. Furthermore, fast online de-
termination of the single tuning parameter involved in the LOS profile is provided
via a systematic parameter selection methodology. Numerical simulations under
time-varying velocities and various target maneuvers are conducted to validate the
effectiveness and robustness of the proposed guidance law.

iii
Contents

Certificate i

Acknowledgements ii

Abstract iii

Contents iv

List of Figures vi

List of Tables vii

Abbreviations viii

Symbols ix

1 Introduction 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 LOS Rate Profile Shaping 4


2.1 2D Missile-Target Engagement Geometry . . . . . . . . . . . . . . . . 4
2.2 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 LOS Rate Profile Shaping . . . . . . . . . . . . . . . . . . . . . . . . 6
2.4 Parameter Selection of δ1 : Impact Time Achievement . . . . . . . . . 9

3 Impact Time Guidance Law for Various Targets 11


3.1 Impact time Guidance Law for Non-Maneuvering Targets . . . . . . . 11
3.1.1 Finite Time Convergence of the ITG Law . . . . . . . . . . . . 13
3.2 Impact time Guidance Law for Non-Maneuvering Targets . . . . . . . 14

4 Numerical Simulations and Results 18


4.1 Optimal δ1 under different impact times . . . . . . . . . . . . . . . . 18
4.2 Various Impact Times Against a Constant-Velocity Moving Target . . 20

iv
Contents v

4.3 Various Impact Times Against a Maneuvering Target . . . . . . . . . 21

5 Conclusion 24

A Appendix A 25
A.1 Theorem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

B Appendix B 27
B.1 Algorithm 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Bibliography 28
List of Figures

2.1 Homing guidance geometry of missile–target engagement. . . . . . . . 4


2.2 Desired LOS rate profiles under various δ1 . . . . . . . . . . . . . . . 8

4.1 Optimal δ1 under different impact times for various target motions. . 19
4.2 Fight trajectories and LOS rate variation of the missile and target . . 21
4.3 Range histories & Missile acceleration trends . . . . . . . . . . . . . . 21
4.4 Fight trajectories and LOS rate variation of the missile and target . . 22
4.5 Range histories & Missile acceleration trends . . . . . . . . . . . . . . 23

vi
List of Tables

4.1 Optimal δ1 under different impact times . . . . . . . . . . . . . . . . 19


4.2 Various δ1 and impact time errors against a constant-velocity target . 20
4.3 Various δ1 and impact time errors against a maneuvering target target 22

vii
Abbreviations

PNG Proportional Navigational Guidance


LOS Line-Of-Sight
IAG Impact Angle Guidance
ITG Impact Time Guidance
PIP Predictive Intercept Point
ST Stationary Target
CT Constant-Velocity Target
MT Maneuvering Target

viii
Symbols

M Missile
T Target
Vm Velocity of Missile in m/s
Am Missile Acceleration in m/s2
Vt Velocity of Target in m/s
At Target Acceleration in m/s2
r Relative Range
φm Look angle of Missile in deg.
λ LOS angle of Missile in deg.
γ Path angle of Missile deg.
γm Path angle of Target deg.
rf Miss Distance in m.
timp Actual Impact Time in sec.
δi , (i = 1, 2, 3) Constant Parameters
λd Desired LOS profile of Missile in deg/sec.
te Time elapsed after the launch in sec.
Td Desired impact time
e LOS tracking error
s, σ Desired sliding surface
k, α, η,  Design parameters
h Uncertain term
τ Time constant in sec.

ix
Chapter 1

Introduction

1.1 Introduction

The well-known proportional navigation guidance (PNG) law has been widely used
and studied for decades owing to its achievement of the minimal miss distance, sim-
ple structure, and effectiveness[9][7].However, advanced homing guidance laws with
impact angle or impact time constraint aiming to raise kill probability have received
considerable attention recently. The impact angle guidance (IAG) law is to send a
missile to the target with a specific incident angle on the collision, which depicts a
one-to-one engagement scenario[8]. However, the impact time guidance (ITG) law
can lead several missiles to attack the target simultaneously, which corresponds to
a many-to-one engagement scenario (or the so-called salvo attack). Obviously, the
kill probability in the latter scenario is much higher than that in the former case.
Other advantages of the ITG law are that the missile can capture a target or pass
through a certain way point at a desired impact time.

The reason that most ITG laws are developed to intercept nonmaneuvering targets
lies on the assumption that the target’s maneuver is small to be neglected, for
instance, while intercepting a warship. However, for a maneuvering target (even

1
Introduction 2

if its maneuver is small) in the practical case, the ITG law without considering
the target maneuver may need to make a trade off between the zero miss distance
and the impact time constraint, especially when the target acceleration is unknown.
Thus, the design of an ITG law for intercepting a maneuvering target encountered
in practice is necessary and will be a more challenging and complex task as a result
of the nonlinear engagement kinematics with unknown target accelerations, system
uncertainties, and physical limitations of actuators, even without the information of
time-to-go and PIP.

In this project, the technique proposed can achieve the desired impact time against
various target motions including nonmaneuvering and maneuvering targets with-
out the knowledge of target maneuvers, time-to-go information, and PIP for the
first time. Specifically, the desired impact time is achieved for various target mo-
tions by tracking a time-varying LOS profile that is a cubic polynomial in time
created by the LOS rate-shaping approach. The terminal sliding-mode and inertial-
delay control are used to derive a finite time convergent ITG law and an ultimately
bounded ITG law for intercepting nonmaneuvering targets and maneuvering targets,
respectively[1]. The obtained guidance laws lead to only one tuning parameter whose
optimal value is obtained for meeting the desired impact time via an optimization
routine based on an estimated safe range of the parameter, which facilitates fast
online parameter determination with a small number of iterations.

1.2 Literature Review

With rising interest over recent years, a number of advanced guidance laws have
been proposed to deal with the impact time constraint. The existing ITG laws may
be roughly classified into two main categories, depending on whether the estimation
of time-to-go is required or not. The first category includes those studies involving
time-to-go estimation procedures. Specifically, one early ITG law was presented in
Introduction 3

[3] via combining the PNG law with an additional feedback term relating to the
difference between the desired and estimated time-to-go values.

To improve the estimation accuracy of time-to-go, two kinds of ITG laws based on
the modified PNG law were presented in [2]. With the concept of [3], the time-
to-go estimation error was reduced in [6] for meeting the impact time constraint
via the predictive trajectory information and a time varying gain, respectively. In,
sliding-mode ITG laws were derived by involving the impact time error in the slid-
ing surface. Using the same estimate of time-to-go in [3], a Lyapunov-based ITG
law was proposed in [4].There have been several studies on the ITG laws without
relying on such information, which leads to the second category. To be specific, the
impact time constraint was achieved in [8] via the polynomial shaping of the relative
range. By using a polynomial function of the guidance command, an augmented
guidance law with impact time constraint was proposed in [5] with three unknown
coefficients. Based on the time-varying sliding-mode technique and trajectory re-
shaping approach, the ITG law with two unknown tunable coefficients was designed
in [10]. The target is assumed stationary or moving with a constant velocity and
without lateral maneuvers in the design of most ITG laws, which is not always satis-
fied in reality and may limit the feasibility of the guidance laws. The only previously
published works dealing with impact time constraint against maneuvering targetsis
presented in [10]. The ITG laws presented in [10] are derived by still assuming a
stationary target, and the interception of a maneuvering target is achieved via the
notion of predictive intercept point (PIP) that highly depends on the prior knowl-
edge of time-to-go and target acceleration. Such error-prone estimate of PIP is also
necessary in many other ITG laws for intercepting constant moving targets.
Chapter 2

LOS Rate Profile Shaping

2.1 2D Missile-Target Engagement Geometry

Figure 2.1: Homing guidance geometry of missile–target engagement.

A planar homing geometry is considered for a missile M and a maneuvering target T,


as shown in Fig.2.1. The missile is assumed to travel with a constant speed VM , which
is perpendicular to the missile acceleration AM . The target travels with a constant
speed VT , which is perpendicular to the target acceleration AT . The relative range is
denoted as r. The angles φM and λ are the look angle and LOS angle of the missile,

4
LOS Rate Profile Shaping 5

respectively. The angles γ and γT denote the path angle of the missile and target,
respectively. All the angles are defined as positive in the counterclockwise direction.
(xM , yM ) and (xT , yT ) denote the locations of the missile and target, respectively.

With the geometry in Fig.2.1, the equations of motion for the missile and target can
be derived as follows:

ẋm = Vm cos γ




 ẏm = Vm sin γ
 γ̇ = Am /Vm


ẋt = Vt cos γt (2.1)
 t = Vt sin γt



 γ̇t = At /Vt



γ̇ = Vt cos (γt − λ) − Vm cos(γ − λ)

˙ = Vt sin (γt − λ) − Vm sin(γ − λ)


rλ (2.2)

2.2 Problem Formulation

Differentiating Eq.(3.2) with respect to time, the LOS dynamics can be obtained as

2ṙλ̇ Am cos φm At cos φt


λ̈ = − − + (2.3)
r r r

where φm = γ − λ, φt = γt − λ.
It can be seen from Eq. (3.3) that AM takes effect for all |φm | =
6 π/2.But |φm | = π/2
is not a stable equilibrium because φ̇m 6= 0 when |φm | = π/2, which means that the
value of φm can only stay on the condition |φm | = π/2 instantly. Therefore, it
is applicable to derive the guidance law using Eq. (3.3). For the impact-time-
constrained guidance problem, the guidance mission is described as the following
objectives: 
rf → 0
(2.4)
t imp → Td

where rf denotes the miss distance, timp is the actual impact time, and Td represents
the desired impact time.
LOS Rate Profile Shaping 6

2.3 LOS Rate Profile Shaping

The main idea of this approach is to find a proper LOS rate profile that can shape
the missile trajectory by increasing or decreasing the LOS rates such that the missile
can hit the target at a desired impact time. Naturally, the designer may come up
with various candidate functions that meet this desired behavior. For simplicity, a
quadratic polynomial in time is adopted for the LOS rate profile and is given by

t2e te
λ̇d = δ1 2
+ δ2 2 + δ3 (2.5)
Td Td

where λ̇d depicts the desired LOS rate profile, te is the time elapsed after the launch,
te denotes the desired impact time, and δi (i = 1, 2, 3) are constant parameters.
To determine the values of δi (i = 1, 2, 3), the boundary conditions (including the
initial and final conditions) for the LOS rates are presented as follows:

λ̇d (t0 ) = λ̇ (t0 )
(2.6)
λ̇d (Td ) = λ̇ (Td ) = 0

where t0 is the initial time. For simplicity, t0 is set to zero, and λ̇(t0 ) is abbreviated
to λ˙0 .
Substituting Eq. (4.2) into Eq. (4.1) yields

δ2 = −δ1 Td − λ̇0 Td
(2.7)
δ3 = λ̇0

Substituting Eq. (4.3) into Eq. (4.1) yields

t2e  t
e
λ̇d = δ1 2 − δ1 + λ̇0 + λ̇0 (2.8)
Td Td

Then, the desired LOS profile can be obtained by integrating Eq. (4.4) as follows:
 
δ1 t3e δ1 + λ̇0 t2e
λd = − + λ̇0 te + δ4 (2.9)
3Td2 2Td
LOS Rate Profile Shaping 7

where λd denotes the desired LOS profile, and δ4 is a constant.


Note that the LOS profile presented in Eq. (4.5) always meets the conditions in
Eq. (4.2), regardless of the constant value of δ4 . Select δ4 = λ0 corresponding
to a selected boundary condition λd (t0 ) = λ(t0 ) = λ0 . Thus, the LOS profiles
presented in Eqs. (4.4) and (4.5) can be adjusted with only one tuning parameter
δ1 . For better visual presentation of the proposed LOS rate profile, a series of
examples under various cases are shown in Fig. 2.2. The values of δ1 are selected as
±0.05, ±0.1, ±0.15, ±0.2, ±0.25; the desired impact time varies with 45, 50, 55, 60,
and 65 s; and the initial LOS rates are considered as λ̇0 = 0, ±0.03 rad/s. It can be
seen that the LOS rate profile keeps negative when δ1 > 0 and λ̇0 ≤ 0, whereas it
keeps positive when δ1 > 0 and λ̇0 ≤ 0, which means that the LOS angle decreases
or increases monotonously under these two cases.

It should be emphasized that the condition λ˙d (Td ) = 0 is only a necessary but
not sufficient condition for guaranteeing a zero miss distance at the desired impact
time according to the parallel approaching principle. Moreover, from the feasibility
perspective, large values of δ1 may lead to large LOS rates, which may saturate the
missile accelerations and fail to intercept the target. Therefore, to meet the zero
miss distance and desired impact time simultaneously, the following requirements
should be satisfied.

1. The missile can precisely track the desired time-varying LOS profile presented
in Eq. (4.5) such that the first guidance objective in Eq. (3.4), r(Td ) = 0, is
met.

2. Determine a safe range of δ1 to ensure that the zero miss distance is practically
achievable.

3. Select a proper LOS profile with a specified δ1 from its safe range to satisfy the
desired impact time, i.e., the second guidance objective in Eq. (3.4), timp → Td .
LOS Rate Profile Shaping 8

Desired LOS rate profiles under various 1 and .0 =0 ,Td =40 Desired LOS rate profiles under various 1 and .0 =0 ,Td =45 Desired LOS rate profiles under various 1 and .0 =0 ,Td =50
0.2 0.15 0.08

0.15 0.06
0.1

0.1 0.04

0.05
0.05 0.02
d

d
0 0 0
.

.
-0.05 -0.02
-0.05

-0.1 -0.04

-0.1
-0.15 -0.06

-0.2 -0.15 -0.08


0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec.) Time (sec.) Time (sec.)

. . .
Desired LOS rate profiles under various 1 and 0
=0 ,Td =55 Desired LOS rate profiles under various 1 and 0
=0 ,Td =60 Desired LOS rate profiles under various 1 and 0
=0 ,Td =65
0.08 0.08 0.08

0.06 0.06 0.06

0.04 0.04 0.04

0.02 0.02 0.02


d

d
0 0 0
.

.
-0.02 -0.02 -0.02

-0.04 -0.04 -0.04

-0.06 -0.06 -0.06

-0.08 -0.08 -0.08


0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec.) Time (sec.) Time (sec.)

. . .
Desired LOS rate profiles under various 1 and 0
=0.03 ,Td =40 Desired LOS rate profiles under various 1 and 0
=0.03 ,Td =45 Desired LOS rate profiles under various 1 and 0
=0.03 ,Td =50
0.2 0.15 0.08

0.15 0.06
0.1
0.1
0.04

0.05 0.05
0.02
0
d

d
0 0
.

.
-0.05
-0.02
-0.1 -0.05

-0.04
-0.15
-0.1
-0.2 -0.06

-0.25 -0.15 -0.08


0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec.) Time (sec.) Time (sec.)

. . .
Desired LOS rate profiles under various 1 and 0
=0.03 ,Td =55 Desired LOS rate profiles under various 1 and 0
=0.03 ,Td =60 Desired LOS rate profiles under various 1 and 0
=0.03 ,Td =65
0.08 0.08 0.08

0.06 0.06 0.06

0.04 0.04 0.04

0.02 0.02 0.02


d

d
.

0 0 0

-0.02 -0.02 -0.02

-0.04 -0.04 -0.04

-0.06 -0.06 -0.06


0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec.) Time (sec.) Time (sec.)

Desired LOS rate profiles under various 1 and .0 =-0.03 ,Td =40 Desired LOS rate profiles under various 1 and .0 =-0.03 ,Td =45 Desired LOS rate profiles under various 1 and .0 =-0.03 ,Td =50
0.25 0.15 0.08

0.2 0.06
0.1
0.15
0.04

0.1 0.05
0.02
0.05
d

0 0
.

0
-0.02
-0.05 -0.05

-0.04
-0.1
-0.1
-0.15 -0.06

-0.2 -0.15 -0.08


0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec.) Time (sec.) Time (sec.)

Desired LOS rate profiles under various 1 and .0 =-0.03 ,Td =55 Desired LOS rate profiles under various 1 and .0 =-0.03 ,Td =60 Desired LOS rate profiles under various 1 and .0 =-0.03 ,Td =65
0.06 0.06 0.06

0.04 0.04 0.04

0.02 0.02 0.02

0 0 0
d

d
.

-0.02 -0.02 -0.02

-0.04 -0.04 -0.04

-0.06 -0.06 -0.06

-0.08 -0.08 -0.08


0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
Time (sec.) Time (sec.) Time (sec.)

Figure 2.2: Desired LOS rate profiles under various δ1


LOS Rate Profile Shaping 9

To accomplish these requirements, the sliding-mode technique is employed to nullify


the LOS tracking error, and an optimization routine based on the safe range of δ1
is used to find the optimal δ1 with fast online computation.

2.4 Parameter Selection of δ1: Impact Time Achieve-


ment

A safe range of δ1 should be determined to ensure that the zero miss distance is
practically achievable before meeting the desired impact time. For simplicity, only
δ1 > 0 is considered in this work because the similar analysis can be used for a nega-
tive δ1 . It has been shown in Fig. 2 that larger LOS rates are created by greater δ1 ,
which may cause large look angles |φm | to cross π/2. Note that this situation can
lead to a large miss distance and fail the interception because the proposed guidance
law may saturate the missile acceleration with frequent switches when |φm | crosses
π/2. Thus, the constraint max (|φm |) < π/2 is imposed to determine the safe range
of δ1 . However, it is difficult to derive an analytical expression of φm by integrat-
ing the differential equations with the desired LOS rate profile because the path
angle variation and the target motion are unknown, which makes it hard to find a
straightforward solution of maximum δ1 (denoted as δu ) satisfying max (|φm |) < π/2.
A partial-covering numerical routine is employed here and presented as a pseudocode
in Algorithm 1 to solve this problem. The main idea for this routine is that the value
of δu for stationary targets can provide a good reference for nonstationary targets
and is easy to determine by integrating the derivative of φm numerically without the
target information. To start with, the kinematic equations for a stationary target
are given by 
 γ̇ = Am /Vm
ṙ = −Vm cos φm (2.10)
 r = −V sin φ /λ̇
m m
LOS Rate Profile Shaping 10

From Eq. (4.6), the derivative of φm for a stationary target can be obtained as

λ̈
φ̇m = γ̇ − λ̇ = λ̇ + tan φm (2.11)
λ̇

Note that λ̇ = λ˙d when the missile precisely follows the desired LOS rate profile
from the beginning; then, Eq. (4.7) becomes

λ̈d
φ̇m = λ̇d + tan φm (2.12)
λ̇d

The look-angle profile can be obtained by integrating the right hand side of Eq. (4.8)
with numerical methods such as the Euler or Runge–Kutta method. Moreover, δu
can be determined by evaluating the condition max (|φm |) ≥ π/2, as presented in
Algorithm 1. It is worth noting that only φm , λ0 , Td are required to determine δu .
Another benefit brought by Algorithm 1 is that it provides a way to investigate the
changing histories of δu under various impact times.
Chapter 3

Impact Time Guidance Law for


Various Targets

3.1 Impact time Guidance Law for Non-Maneuvering


Targets

A terminal sliding-mode technique is applied to precisely track the desired LOS


profile for intercepting nonmaneuvering targets. For this case, the LOS dynamics
presented in Eq. (3.3) reduces to

2ṙλ̇ Am cos φm
λ̈ = − − (3.1)
r r

The LOS tracking error is selected as e = λ − λd , and the sliding surface is defined
as
s = ė + k|e|α sign(e) (3.2)

11
ITG Law for Various Targets 12

where k > 0 and 0 < α < 1 are design parameters, and sign(·) is the signum
function. Taking the time derivative of Eq. (5.2) leads to

ṡ = λ̈ − λ̈d + kα|e|α−1 ė (3.3)

where ė = λ̇ − λ̇d , and ~λd can be obtained by differentiating Eq. (4.4):


 
te δ1 + λ̇0
λ̈d = 2δ1 − (3.4)
Td2 Td

Substituting Eq. (5.1) into Eq. (5.3) yields

2ṙλ̇ Am cos φM
ṡ = − − − λ̈d + kα|e|α−1 ė (3.5)
r r

To guarantee s=0, the missile acceleration AM is designed by combining an equiv-


alent controller and a discontinuous controller (denoted as Aeq disc
m and Am , respec-
tively) in the form of
Am = Aeq disc
m + Am (3.6)

Let ṡ = 0; Aeq
m can be obtained as

!
r 2ṙλ̇
Aeq
m = − − λ̈d + kα|e|α−1 ė (3.7)
cos φm r

The discontinuous controller is designed as

η
Adisc
m = sign(s) (3.8)
cos φm

where η > 0 is a positive parameter to be designed. Then, substituting Eqs. (5.4),


(5.7), and (5.8) into Eq. (5.6) results in

   
r  2ṙλ̇ te δ1 + λ̇0 η
Am = − − 2δ1 2 + + kα|e|α−1 ė + sign(s) (3.9)
cos φm r Td Td r
ITG Law for Various Targets 13

3.1.1 Finite Time Convergence of the ITG Law

It is worth noting that the LOS tracking error can converge to zero in finite time
under the guidance law in Eq. (5.9). A Candidate Lyapunov function is selected as

1
V = s2 (3.10)
2

Taking the time derivative of Eq. (5.10) results in V̇ = sṡ, and substituting Eq.
(5.9) into Eq. (5.5), V̇ becomes

η η
V̇ = −s sign(s) = − |s| < 0, for s 6= 0 (3.11)
r r

It can be seen from Eq. (5.11) that V̇ is negative-definite for all s 6= 0. Under this
condition, the sliding mode can always occur. Then, Eq. (5.11) is rewritten as

η 2η 1/2
V̇ = − |s| = − V (3.12)
r r

Consider that 0 < r < r0 , where r0 denotes the initial relative range between the
missile and target. Thus, √
2η 1/2
V̇ ≤ − V (3.13)
r0
According to Theorem 1 (see the Appendix A), the convergence time for the reaching
phase can be obtained as √
2r0 1/2
τ1 ≤ V0 (3.14)
η
After the reaching phase, s = 0 is guaranteed; then, Eq. (5.2) reduces to

ė = −k|e|α sign(e) (3.15)

Select a Candidate Lyapunov function as

1
W = e2 (3.16)
2
ITG Law for Various Targets 14

Taking the time derivative of Eq. (5.16) results in Ẇ = eė, and substituting Eq.
(5.15) into it, Ẇ becomes


Ẇ = −k|e|α+1 = −( 2)α+1 kW (α+1)/2 (3.17)

for all 0 < (α + 1)/2 < 1 As the above case, convergence time for the sliding phase
is obtained as √
( 2)1−α (1−α)/2
τ2 ≤ W (3.18)
k(1 − α) 0
where W0 is the initial value of W. Therefore, the convergence time for the guidance
system is given by
√ √
2r0 1/2 ( 2)1−α (1−α)/2
τf = τ1 + τ2 ≤ V0 + W (3.19)
η k(1 − α) 0

With the preceding stability analysis, it can be known that the desired LOS profile
can be precisely tracked by the ITG law in Eq. (5.9) in finite time while intercepting
a nonmaneuvering target. It should be noted that, for the impact-time-constrained
guidance problem, the convergence time must be less than the desired impact time.
Otherwise, the LOS profile cannot be tracked before the collision.Fortunately, τf = 0
holds in the proposed approach thanks to the initial conditions e(0) = s(0) = 0,
which leads to V0 = W0 = 0. Although there may exist small initial measurement
errors on e(0), s(0), τf can be still less than Td by selecting appropriate guidance
parameters. Thus, the guidance law in Eq. (5.9) always guarantees that the LOS
profile can be precisely tracked before the collision.

3.2 Impact time Guidance Law for Non-Maneuvering


Targets

In this subsection, an ITG law is designed for intercepting a maneuvering target


with unknown target acceleration using the terminal sliding-mode and inertial-delay
ITG Law for Various Targets 15

control methods, based on the LOS rate-shaping approach. The inertial-delay con-
trol is used to estimate the target acceleration via a first-order transfer function.
For a maneuvering target, the LOS dynamics in Eq. (3.3) is recalled here

2ṙλ̇ Am cos φm At cos φt


λ̈ = − − + (3.20)
r r r

Similar to Eq. (5.2), the sliding surface in this section is defined as

σ = ė + k|e|α sign(e) (3.21)

where k > 0 and 0 < α < 1 are the same as those in Eq. (5.2), and e = λ − λd still
denotes the LOS tracking error. Taking the time derivative of Eq. (6.2) leads to

2ṙλ̇ Am cos φm h
σ̇ = − − − λ̈d + + kα|e|α−1 ė (3.22)
r r r

where h = At cos φt is considered as an uncertain term. Consider that the target


acceleration is limited in practice; then, it is assumed that the uncertainty term
h = At cos φt is continuous, slowly varying, and its first-order derivative is unknown
but bounded by |ḣ| ≤ H, where H is a small positive number.
The missile acceleration Am is designed by combining an equivalent controller (which
f
is chosen to be same as Aeq
m in Eq. (5.7)) and a feedback controller (denoted as AM ,

which compensates the unknown target maneuvers) in the form of

Am = Aeq f
m + Am (3.23)

The feedback controller is designed as

ĥ + εrσ
Afm = (3.24)
cos φm

where  is a positive parameter to be designed.The ĥ denotes the estimate of h


and can be designed by the inertial-delay control with a first-order transfer function
ITG Law for Various Targets 16

expressed in the time domain as follows:

˙
h = τ ĥ + ĥ (3.25)

where τ is a time constant to be designed.


To obtain the explicit expression of ĥ, Eqs. (5.7) and (6.5) are substituted into Eq.
(6.4), which results in
!
r 2ṙλ̇ ĥ + εrσ
Am = − − λ̈d + kα|eα−1 ė + (3.26)
cos φm r r

Substituting Eq. (6.7) into Eq. (6.3) leads to

rσ̇ = h − ĥ − εrσ (3.27)

but,
d(rσ)
= rσ̇ + σ ṙ (3.28)
dt
Substitute Eq. (6.8) in Eq. (6.9),

d(rσ)
= h − ĥ − εrσ + σ ṙ (3.29)
dt

Which gives h as,


d(rσ)
h= + ĥ + εrσ − ṙσ (3.30)
dt
From Eqs. (6.6) and (6.11), it follows that

ˆ d(rσ)
τ ĥ = + εrσ − ṙσ (3.31)
dt

Then, the estimation of the uncertain term can be given by


 Z 
1
ĥ = rσ + (εrσ − ṙσ)dt (3.32)
τ
ITG Law for Various Targets 17

Substituting Eq. (6.13) into Eq. (6.7) yields


! Z
r 2ṙλ̇ σ 1
Am = − − λ̈d + kα|e|α−1 ė + εσ + + (εrσ − ṙσ)dt
cos φm r τ τ cos φm
(3.33)

It should be emphasized in Eq. (6.14) that large guidance command may be created
because of the small value of τ . However, it should be noted that e(0) = ė(0) = 0
holds thanks to the initial boundary conditions, which leads to σ(0) = 0. Moreover,
it is shown that σ, e, ė can be ultimately bounded with small boundaries under the
guidance law in Eq. (6.14). Thus, the guidance law in Eq. (6.14) will not lead to
large commands during the whole guidance process.
Chapter 4

Numerical Simulations and


Results

The performance of the proposed ITG law is validated through a series of numerical
simulations for intercepting various target motions such as the stationary, constant-
velocity moving, and maneuvering targets. In all the simulation cases, the speed of
the missile is assumed 250 m/s, whereas the various target motions are presented
in each subsection. Initial positions of the missile and target are assumed (0, 0 km)
and (10, 0 km), respectively. For all the cases, the design parameters are selected
as k = 1, α = 0.8, η = 500, ζ = 0.1, ε = 1, τ = 0.01, whereas the values of δ1
under various cases are given in the following subsections. The values of δ2 , δ3 , δ4
are selected according to the LOS initial conditions given in Chapter 2.

4.1 Optimal δ1 under different impact times

The optimal values of δ1 under different impact time constraints (varied from 40
to 65 s) for various target motions and the values of δu obtained from Algorithm 1
(see Appendix B.1) are presented in Fig. 4.1 and Table 4.1. It can be seen that the

18
Numerical Simulations and Results 19

0.11

0.1

0.09

0.08
Optimal 1

0.07

0.06

0.05

0.04

0.03 U
CT
MT
0.02

40 45 50 55 60 65
Desired impact time (sec)

Figure 4.1: Optimal δ1 under different impact times for various target motions.

longer the desired impact time is, the greater that δu is. Meanwhile, all the optimal
δ1 are within the profile of δu . Another observation is that the optimal δ1 changes
more slowly as the impact time becomes longer.

Table 4.1: Optimal δ1 under different impact times

Desired Impact Time (sec.) Optimal δ1


40 0.095
45 0.098
50 0.101
55 0.104
60 0.107
65 0.110
Numerical Simulations and Results 20

4.2 Various Impact Times Against a Constant-


Velocity Moving Target

The first set of simulations is conducted for the case when the target is moving with
a constant speed of 15 m/s without lateral maneuvers. Various desired impact times
(selected as 50, 55, 60, and 65 s) are considered for verifying the effectiveness of the
ITG law in Eq. (7.3). The values of δ1 and the impact time errors resulting from
the autopilot lag are presented in Table 4.2. The simulation results are shown in

Table 4.2: Various δ1 and impact time errors against a constant-velocity target

Desired Impact Time (sec.) δ1


50 0.067
55 0.077
60 0.083
65 0.091

Fig. 4.2 and 4.3. It can be observed from Fig. 8.2.1 that the flight trajectory of
the missile is more curved if longer desired impact time is required. In Fig. 8.2.2,
the magnitudes of the missile accelerations are much smaller than the maximum
allowable acceleration. Meanwhile, the missile acceleration increases with the growth
of the desired impact time. From Figs. 8.3.1 and 8.3.2, all the LOS rates and the
relative ranges converge to zero at the desired impact times. Additionally, it can be
observed from Table 4.2 that impact time errors due to the autopilot lag are within
an acceptable small range, which shows the superior performance of the proposed
ITG law.
Numerical Simulations and Results 21

1400 0.02
Target
65s
1200 0
60s
55s
50s
1000 -0.02

LOS rates (rad/sec)


Altitude (m.)

800 -0.04

600 -0.06

400 -0.08
50s
55s
200 -0.1
60s
65s

0 -0.12
0 2000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70
Downrange (m.) Time (sec.)

8.2.1 Flight trajectories 8.2.2 LOS rates variations

Figure 4.2: Fight trajectories and LOS rate variation of the missile and target

10000 0
65s
9000 60s
55s -200
8000 50s

7000
Acceleration (m/sec2)
Relative Range (m.)

-400
6000

5000 -600

4000
-800
3000

2000 65s
-1000 60s
55s
1000
50s

0 -1200
0 10 20 30 40 50 60 70 0 5 10 15 20 25 30 35 40 45
Time (sec.) Time (sec.)

8.2.3 Range Histories 8.2.4 Acceleration Trends

Figure 4.3: Range histories & Missile acceleration trends

4.3 Various Impact Times Against a Maneuvering


Target

The proposed ITG law in Eq. (7.4) is used to intercept a maneuvering warship that
sails with a speed of 15 m/s and maneuvers with At = 50 cos(10t)m/s2 . The other
simulation settings are chosen as the same as the previous settings. The values of δ1
and the impact time errors resulting from the autopilot lag are presented in Table
4.3. It can be seen that the impact time errors are still small in spite of intercepting
Numerical Simulations and Results 22

a maneuvering target. The simulation results are presented in Fig. 4.4 and 4.5. In
Fig. 8.4.1, all the missiles can intercept the maneuvering target with various flight
trajectories under different desired impact times. It can be observed from Fig. 8.4.2
that the missile acceleration increases when the desired impact time grows. From
Fig. 8.5.1, it can be seen that zero LOS rates are achieved at the desired impact
times. It is clearly shown in Fig. 8.5.2 (the thinner solid lines depict the estimated
uncertainties) that the uncertainties due to the unknown target maneuvers are pre-
cisely estimated via Eq. (6.13).

Table 4.3: Various δ1 and impact time errors against a maneuvering target
target

Desired Impact Time (sec.) δ1


50 0.0716
55 0.0878
60 0.0967
65 0.1022

Missile-Target Trajectory
1400 0.02
Target
65s
1200 0
60s
55s
50s
1000 -0.02
LOS rates (rad/sec)
Altitude (m.)

800 -0.04

600 -0.06

400 -0.08
50s
55s
200 -0.1
60s
65s

0 -0.12
0 2000 4000 6000 8000 10000 12000 0 10 20 30 40 50 60 70
Downrange (m.) Time (sec.)

8.4.1 Flight trajectories 8.4.2 LOS rates variations

Figure 4.4: Fight trajectories and LOS rate variation of the missile and target
Numerical Simulations and Results 23

10000 400
65s
9000 60s
200
55s
8000 50s

7000 0

Acceleration (m/sec2)
Relative Range (m.)

6000
-200
5000
-400
4000

3000 -600

2000 65s
60s
-800
55s
1000
50s

0 -1000
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30 35 40 45
Time (sec.) Time (sec.)

8.5.1 Flight trajectories 8.5.2 Closer to the point of interception

Figure 4.5: Range histories & Missile acceleration trends


Chapter 5

Conclusion

In this project, the problem of impact time guidance law design against nonma-
neuvering and maneuvering targets is solved by tracking a novel time-varying line-
of-sight (LOS) profile, which can be curved to meet the desired impact time with
only one tuning parameter. The terminal sliding-mode technique is used to obtain
the guidance law against nonmaneuvering targets, whereas it is combined with the
inertial-delay control to derive the guidance command against maneuvering targets.
The unknown target maneuvers are precisely estimated via the inertial-delay con-
trol based estimation procedure. It is shown that the uncertainty estimation error
and the LOS tracking error are ultimately bounded within small boundaries. The
only one tuning parameter δ1 selected from an estimated safe range for meeting the
desired impact time is determined via an optimization routine with a small number
of iterations, which facilitates fast online computation of this guidance parameter.
In addition,

24
Appendix A

Appendix A

A.1 Theorem 1

Suppose there exists a continuous function V : D → R such that the following


conditions hold:

1. V is positive definite.

2. There exist real numbers c > 0 and α ∈ (0, 1) and an open neighborhood V ⊆ D
of the origin such that

V̇ (x) + c(V (x))α ≤ 0, x ∈ V\{0} (A.1)

Then the origin is a finite-time-stable equilibrium of the system

ẏ(t) = f (y(t)) (A.2)

Moreover, if N ⊆ D and T is the settling-time function, then

1
T (x) ≤ V (x)1−α , x ∈ N (A.3)
c(1 − α)

25
Appendix B 26

and T is continuous on N . If in addition D = Rn , V is proper, and V̇ takes negative


values on Rn \{0}, then the origin is a globally finite-time-stable equilibrium of Eq.
(A.2).
Appendix B

Appendix B

B.1 Algorithm 1

Assign φM 0 , λ0 , Td , q1 = 0, t = 0, ∆h = 0.001;
Step 0 : Assign Q = [q1 , q2 , . . . , qm ] ; // The elements in Q increase with
sufficiently small steps
for n = 1 to m
Step 1 :
Assign δ1 = Q(n)
while t ≤ Td
t = t + ∆h
Step 2 : Calculate λd and λ̈d from Eq. (4.4) and Eq. (5.1), respectively;
Calculate φM by integrating Eq. (4.8)
end while
if max (|φM |) ≥ π/2
δU = δ1
Step 3 :
end if
end for

27
Bibliography

[1] Q. Hu, T. Han, and M. Xin. Sliding-mode impact time guidance law de-
sign for various target motions. Journal of Guidance, Control, and Dynamics,
42(1):136–148, 2018.

[2] I.-S. Jeon and J.-I. Lee. Optimality of proportional navigation based on non-
linear formulation. IEEE Transactions on Aerospace and Electronic Systems,
46(4):2051–2055, 2010.

[3] I.-S. Jeon, J.-I. Lee, and M.-J. Tahk. Impact-time-control guidance law for anti-
ship missiles. IEEE Transactions on control systems technology, 14(2):260–266,
2006.

[4] M. Kim, B. Jung, B. Han, S. Lee, and Y. Kim. Lyapunov-based impact


time control guidance laws against stationary targets. IEEE Transactions on
Aerospace and Electronic Systems, 51(2):1111–1122, 2015.

[5] T.-H. Kim, C.-H. Lee, I.-S. Jeon, and M.-J. Tahk. Augmented polynomial guid-
ance with impact time and angle constraints. IEEE Transactions on Aerospace
and Electronic Systems, 49(4):2806–2817, 2013.

[6] J.-I. Lee, I.-S. Jeon, and M.-J. Tahk. Guidance law to control impact time and
angle. IEEE Transactions on Aerospace and Electronic Systems, 43(1):301–310,
2007.

[7] G. M. Siouris. Missile guidance and control systems. Springer Science & Busi-
ness Media, 2004.
28
Bibliography 29

[8] R. Tekin and F. Holzapfel. Impact angle control based on feedback linearization.
In AIAA Guidance, Navigation, and Control Conference, page 1509, 2017.

[9] P. Zarchan. Tactical and strategic missile guidance. American Institute of


Aeronautics and Astronautics, Inc., 2012.

[10] Y. Zhao, Y. Sheng, and X. Liu. Analytical impact time and angle guidance via
time-varying sliding mode technique. ISA transactions, 62:164–176, 2016.

You might also like