Three-Dimensional Missile Guidance Laws Design Using Fuzzy Schemes

Elhalwagy Y.Z. and Tarbouchi M.

Electrical and Computer Engineering Department
Royal Military College, Kingston, Ontario, CANADA K7K7B4

Abstract: - This paper proposes three-dimensional fuzzy missile guidance laws based on line-of-sight,
proportional navigation, and mixed strategy guidance. The results are promising and clearly demonstrate
the potential of fuzzy guidance schemes against non-maneuvering and maneuvering targets. The miss
distance and commanded acceleration profile are used in the performance evaluation of the proposed
fuzzy guidance laws. A complete six-degrees-of-freedom flight simulation model for anti-aircraft
command guided missile system is developed for verification.

Keywords: - Fuzzy Logic, Neural Network, Missile Guidance and control

NOMENCLATURE cl, cm, cn = Aerodynamic moment coefficients that
Abbreviations describe the missile airframe.
CoG = Center of Gravity d, l , m, s = Missile diameter, length, mass, and
CLOS-STT = Command Line Of Sight Skid To Turn 2
FLOSCG = Fuzzy Line Of Sight Command Guidance reference cross-sectional area (π d / 4) respectively.
FPNCG = Fuzzy Proportional Navigation Command ρ, Q = Air density and atmospheric dynamic pressure
Guidance ( 1 2 ρVm2 ).
FSG = Fuzzy Supervisory Guidance g = Acceleration due to gravity.
LOS = Line-of-Sight εm, βm, εt, βt = Missile and target LOS angles
LATAX = Lateral Acceleration rm, rt, rmt = Missile, target slant ranges and relative
MSG = Mixed Strategy Guidance
PN/APN = Proportional Navigation / Augmented PN distance respectively
6DOF = Six-Degrees-of-Freedom xr, yr, zr = Missile-Target relative distance
Symbols vc , atp. aty = closing velocity, target acceleration in
Ixx, Iyy, Izz = Moments of inertia about airframe axes. pitch and yaw respectively.
Prox, Proy, Proz = Products of inertia.
U, v, w, Vm = Velocity components along the 1 Introduction
missile’s axes (X, Y, Z), and Total missile velocity, The control hierarchy of a guided weapon consists
respectively. of an outer guidance loop and an autopilot. As seen in
p, q, r = Missile angular rates (roll, pitch, and yaw), Fig. 1, the guidance loop first measures the difference
respectively. between the target and weapon trajectories and then
Tx = Engine thrust force (component in the missile produces guidance inputs to the autopilot, which
longitudinal axis). controls the flight of the weapon through, for example,
Fx, Fy, Fz = Aerodynamic forces. moving fins or other control surfaces and thrust vector
Mx, My, Mz = Aerodynamic moments. control. The autopilot and its control elements are
onboard the weapon and accepts guidance inputs
c m α , c m δ , c mq = Pitching moment derivatives. through telemetry.
c n β , c n δ , c mr = Yawing moment derivatives. The aim of the guidance loop is to generate inputs
so that the weapon will intercept the target in minimum
c βy , c δy = Side force derivatives. time. This is a nonlinear time-varying control problem
and while classical design methods have produced
c αz , c δz = Z-axis force derivatives. reliable systems, recently there have been numerous
α, β = Angle-of-Attack and sideslip angle. studies [1-10] into the application of intelligent control
φ, θ, ψ = Euler’s angles (roll, pitch attitude, and theory to guidance loop design. This is due in part to
heading angles), respectively. the inadequacy of classical techniques in dealing with
cx, cy, cz = Aerodynamic force coefficients that highly maneuverable targets.
describe the missile airframe. Common guidance laws in use are LOS guidance
(either CLOS or beam-rider system) [11-14], LOS rate
guidance [12], and other advanced guidance strategies,

1
such as PN guidance [11-14], APN [13], and optimal The nonlinear differential equations that describe the
guidance laws [13]. LOS guidance is designated as a missile dynamics in the space are given to show the
three-point guidance system. A beam-riding missile nonlinearities in the system kinematics and dynamics
generates its own commands internally whereas CLOS along with the environmental changes. In section 3 the
missile receives its commands from a remote station. In investigations of designing missile guidance laws
CLOS guidance strategy, the missile approaches the based on the fuzzy logic theory are presented.
target along the line joining the control point and the Evaluations of the three-dimensional missile-target
target. The homing guidance system, which contrasts engagement scenarios are given in section 4. Finally,
with the LOS guidance, is designated as a two-point this paper ends with the conclusions.
guidance system and is implemented mostly as LOS
rate guidance. PN guidance law issues acceleration 2 Missile Guidance and Control Model
commands, perpendicular to the instantaneous missile- The missile simulation that was used to generate all
target LOS, which are proportional to the LOS rate and results in this paper is a 6-DOF nonlinear dynamic
closing velocity. Basically, this guidance law tries to model of a guided missile system. The missile is
nullify the LOS change, placing the missile and the aerodynamically controlled via two pairs of rear
target on a collision trajectory. Simply, APN is a PN control fins. It has two identical control channels, each
guidance law with an extra term to account for the channel has lateral acceleration autopilot loop that
maneuvering target. Classical guidance laws different control the missile lateral acceleration to be very close
from these guidance laws were discussed in [12,13], to the target at the end of engagement. The autopilot
where the performance of various guidance laws was consists of a pneumatic fin servo, one accelerometer,
extensively compared. one rate gyro, and the conditioning electronic circuits.
The main advantage of intelligent over classical In addition, a roll position control loop is utilized to
control is that the former can provide robust systems keep the missile attitude fixed throughout the flight.
when there are model and environmental uncertainties. The equations for the missile’s CoG kinematical
Neural networks and fuzzy logic [15-17], by giving and dynamical motion, kinematical and dynamical
control laws based on input-output relationships, avoid rotation of the missile body around its CoG, and the
the need for accurate knowledge of system dynamics, on-board measuring and control devices are examined.
and are thus insensitive to their changes. Examples of Environmental parameter changes such as air density,
application of intelligent control to missile autopilot velocity of sound as a function of altitude, and wind all
design are in [1-6]. However, only a few [7-10] deal effect the plant model. The motion of the missile in
with the design of a guidance law. space is described by means of 6-differential equations.
Hopfield neural network architecture was developed Referring to Fig. 2, the missile equations of motion are
to solve the optimal control problem for homing guided expressed in the body coordinate system as [14]:
missile [7]. As an alternative approach, a fuzzy-logic-
based closed loop optimal law for homing missile
T x − Fx − mg sin θ = m U (
& + qw − rv )
guidance was investigated [8]. Both of these studies are Fy + mg cos θ sin ϕ = m (v& − pw + rU )
based on the well-known PN guidance method. It has Fz + mg cos θ cos ϕ = m (w & − qU + pv )
been shown in [9] the superiority of two-fuzzy-logic ( )
M x = I xx p& + qr I zz − I yy + Prox
(2.1)
based homing guidance schemes over the traditional
M y = I yy q& + pr (I xx − I zz ) + Proy
PN or APN guidance methods. However, all the efforts
listed above were limited to examining the single plane ( )
M z = I zz &r + pq I yy − I xx + Proz
motion under certain parameter constraints for the The aerodynamic coefficients are computed at
homing guidance systems. Very recently, a three several operating points and a linear interpolation
dimensional differential game missile guidance law procedure computes their values at any intermediate
using neural networks has been presented [10]. The point. The aerodynamic coefficients, considered to be
results showed the great advantage of neural network one of the major uncertainties in the model dynamics,
based guidance law over the PN guidance law. have in general nonlinear dependence on the Mach
In this paper, fuzzy logic is used to design two pure number and incidence angles. The aerodynamic forces
guidance strategies for CLOS guided missile then a and moments are given by:
MSG [18] one. (1) Fuzzy logic based-LOS guidance F x , y , z = c x , y , z sQ , M x , y , z = c l , m , n sQ l (2.2)
law, (2) Fuzzy logic based APN command guidance
laws, and (3) Fuzzy Supervisory controller based on The aerodynamic force and moment coefficients that
MSG. These guidance algorithms are employed for describe the missile airframe are given by:
guiding a missile to pursue and intercept a moving and c x = c xo + c αx α 2 , c y = c βy β + c δy δ r ,
very often accelerating target which is considered a
c z = c αz α + c δz δ e , c l = c lp p , (2.3)
highly nonlinear time-varying system.
The paper is organized as follows. In section 2, a 6- cm = c m α α + c m δ δ + c mq q , c n = c n β β + c n δ δ + c mr r .
DOF-missile guidance and control model is presented.

2
 The aerodynamic force and moment coefficients FPNCG are presented in the next subsections with a
that are presented previously are usually defined as a suggested FSG scheme to cope with the problems for
function of α, β, and other parameters. Therefore, it is each individually.
desirable to show the relationship between the velocity
components and these angles. These relations are 3.1 Fuzzy-LOS Command Guidance
defined as: The CLOS guidance is classified as a three-point
guidance law. In this guidance strategy, the missile
α=tan-1(W/U), β=tan(V/U), θ = V 2 + W 2 U 2 . (2.4)
T maneuvers so as to be on LOS between the target
where θT is the angle between the velocity vector and tracker and the target. A guidance computer is utilized
the missile longitudinal axis and it is referred to as the at the ground-based station and produces the
resultant angle of incidence. The orientation of these acceleration commands, which are sent via radio link to
variables in the airframe coordinate system with all the autopilot. The main objective of the CLOS
these conventions are shown in Fig. 2. guidance is to constrain the missile to lie as nearly as
A simplified block diagram of the missile guidance possible on the LOS. If the missile is always on the
loop and the location of the autopilot loop are shown in tracker-target LOS, then the missile will surely hit the
Fig. 1. In the autopilot loop, the difference between the target. Geometry of Missile-Target interception is
desired and actual accelerations is sensed and used to shown in Fig. 4. The guidance strategy adopted in this
drive the control surface actuator. A simplified block section is given by:
diagram of a typical autopilot composed of a control εm = εt , βm = βt (3.1.1)
fin driver and measuring instruments is shown in Fig. Thus, the method of control is proportional to the
3. The control fin driver converts the input signal into lateral displacement of the missile from the target LOS
mechanical deflections of the fins for the missile that is given by
guidance. The accelerometer measures the missile's EP = rm(εt-εm), EY = rm (βm - βt) cos εm. (3.1.2)
acceleration. This is modeled by the second order
with the sign indicating the direction of missile
transfer function of the form:
[ ]
Out.[volt] Inp. m / s 2 = − k a τ a2 s 2 + 2τ aζ a s + 1 (2.5)
movement that is required to nullify the error. The
control signals depend not only on the error signal but
In order to isolate the accelerometer pendulum from also on its derivative with the aim of increasing the
missile oscillations, it must be placed as near as stability and improving the transients in the guidance
possible to the missile's CoG. The damping gyro has system.
two degrees of freedom and is utilized in the autopilot The FLOSCG scheme presented here based on the
to damp the oscillations of the missile around its CoG. three-point guidance law is shown in Fig. 5. The fuzzy
Its transfer function is: guidance law has four inputs and two outputs for
Out.[volt] Inp.[deg / s ] = k g τ g2 s 2 + 2τ g ζ g s + 1 .(2.6) generating the demanded acceleration to steer the
missile in space. The inputs are the errors EP and EY
The STT steering policy requires that the roll along the pitch and yaw axes and the change in errors.
autopilot performs attitude stabilization in the This scheme uses 11 uniformly distributed triangular
maneuver plane. A roll position controller is utilized to membership functions for each of its input universe of
keep an adequate roll damping. discourse and the minimum to represent the premise
and implication. Whenever the input is high, the
3 Fuzzy Logic Based Guidance laws saturation of the left most and the right most
In general a fuzzy logic controller contains four membership functions are considered. For illustration,
main components; fuzzification, rule-base, inference sample rules of the fuzzy scheme take the following
mechanism, and defuzifiaction. The fuzzification form for pitch and yaw guidance:
interface simply modifies the inputs so that they can be & is J then A is R (1)
If EP is I1 and EP 1 pc p
interpreted and compared to the rules in the rule base.
& is N then A
If EY is M1 and E
The rule base holds the knowledge in the form of a set y 1 yc is R y(1)
of rules of how best to control the system. The Variables, Ii, Ji, Mi, Ni, Rp(i), and Ry(i), take the
inference mechanism or the decision making logic
linguistic values expressed by linguistic sets such as
evaluates which control rules are relevant at the current
LN and LP that are interpreted as large negative and
time and then decides what is the input to the plant
large positive respectively.
should be. The defuzzification interface converts the
Since the steering command signals must be crisp,
conclusions reached by the inference mechanism into
the center of gravity defuzzification method [15] is
the inputs to the plant to be controlled.
used to calculate the crisp control action. Tuning via
From the literature, beam-riding guidance can be
scaling universes of discourse is applied. A great effort
significantly improved by taking the beam motion into
has been made to choose the proper scaling gains (gpe,
account (CLOS). This is analogous to homing guidance
in which PN performance is improved by tacking target gpc, gye, gyc, gpu, gyu) shown in Fig. 5. More
maneuvering into account (APN). So a FLOSCG and

3
emphasis should be put on finding out the optimum The simulation results in section 4 showed that for
values. the existing autopilot the FLOSCG and FPNCG are not
working well individually. However this problem can
3.2 Fuzzy-PN Command Guidance be resolved in two ways. The first is to redesign the
One of the most widely used homing guidance laws autopilot (the hardest choice) while the second is to
for a few decades is PN guidance law. In this type of combine the two based on MSG. In the next section the
guidance system the missile seeker provides the suggested solution is investigated.
information required for guidance process. However, a
missile seeker is not required in command guidance. 3.3 Fuzzy Supervisory Guidance
External missile/target trackers transmit and receive Since aircrafts become smarter and smarter, no single
radar signals. i.e. we can assume that εm, βm, rm, εt, guidance strategy seems to be adequate to have
βt, and rt are available as well as target maneuvers. In satisfactory performance. The MSG approach suggests
to design two or more pure guidance strategies each
order to implement PN in the command guidance has adequate performance against some of the set of all
system, λ and λ& should be available from the possible target behaviors [18].
measurement information [3]. The yaw and pitch A new missile guidance strategy which combines
components of the line of sight angle can be computed the proposed FLOSCG and FPNCG with certain
using missile and target position vectors in the inertial weights depending on both the relative distance
frame as: between the missile and the target and the closing
λy = tan-1 (yr/xr) and λz = tan -1 (zr/xr) (3.2.1) velocity as well. The proposed fuzzy-supervisory
The yaw and pitch components of LOS rates and the guidance scheme is shown in Fig. 7. The fuzzy rule
closing velocity can be computed as base block has two inputs and two outputs for tuning
(
λ& y = (x r y& r − y r x& r ) x r2 + y r2 , ) the weights (W1, W2). These weights are used to
combine between the two proposed pure guidance
λ& p = (x r z& r − z r x& r ) (x 2
r + z r2 ) (3.2.2) strategies based on the relative distance between the
missile and the target. The inputs are the relative range
v c = −r&mt = −(x r x& r + y r y& r + z r z& r ) rmt between missile and target and the range rate. This
The pitch and yaw acceleration commands are given by simple fuzzy structure also needs a less complicated
Apc = N vc λ
& + 0.5a N + gravity bias.
z tp
rule base.

Ayc = N vc λ & + 0.5a N

y ty (3.2.3) 4 Simulation Results
The effective navigation ratio, N, is kept constant In section 3.1 and 3.2, the design methodology of
during the flight time. The magnitude of the missile the fuzzy guidance laws to generate the steering
demanded (pitch, yaw) acceleration (Apc, Ayc) is commands are proposed. This section presents
numerical simulation results for the FLOSCG and
restricted especially during the initial phase of
FPNCG to demonstrate the performance of the
guidance to not let the missile control surfaces reach
proposed guidance laws. Miss distance and
saturation. It is a well-known fact that larger effective
commanded acceleration profile are used for the
navigation ratio enables the missile to remove the
performance evaluation. Table 1 includes different
initial heading error more rapidly, thus causing a much
target scenarios with non-maneuvering and
tighter trajectory. However, this results in larger
maneuvering; approaching and receding targets with
missile acceleration at the beginning of flight and
different speeds. The cost function is defined as
lower at the end [13]. Then, highly time varying tf
navigation ratio can prevent the commands from LATAX = ∫ A 2PC + A 2YC dt (4.1)
reaching saturation in the beginning of flight and to
compensate for the target maneuverability. A three-dimensional missile-target engagement
A simple fuzzy rule structure is implemented to simulation was set up using the presented mathematical
compute the proper N and provide it to the model in section 2. A complete 6-DOF-flight model for
conventional PN/APN guidance law. These constitute CLOS-STT missile system is developed. A computer
in somehow what is called a hybrid fuzzy scheme. The code that solves the model is carried out with
proposed FPNCG scheme is shown in Fig. 6. The fuzzy BORLANDC. Modular concept is considered in the
rule base block has two inputs and one output for code development. A simplified block diagram of the
generating the time varying effective navigation gain. model is shown in Fig. 8.
The inputs are the relative range between missile and The model is broken down into the following major
target and the range rate. This simple fuzzy structure parts: missile-target geometry, guidance, autopilot,
needs a less complicated rule base. . For illustration, a airframe, and kinematics. In the missile-target
sample rule is given as: geometry module, the missile position relative to the
If rmt is Large and r&mt is large then N is small target is calculated. The guidance parameters, which

4
are the deviation errors between the ideal and actual clearly demonstrate the potential of this fuzzy guidance
position of the missile measured by guidance radar, are scheme against non-maneuvering and maneuvering
then calculated. The guidance module receives the targets with different speeds.
guidance parameters and generates the guidance
steering command signals through different fuzzy
guidance schemes. The guidance signals are supplied 5 Conclusion
to the autopilot to steer the missile in space. For Fuzzy approaches to CLOS-STT missile guidance
completeness, the actuator dynamics is considered. In have been presented. The obtained results show the
general, each fin actuator has a finite bandwidth, for superiority of the proposed FPNCG over FLOSCG.
simplicity, the effectiveness of the fin deflection angles The results are for three-dimensional engagement. The
is modeled by a first order system with surface position FSG scheme has been suggested as a simpler solution.
saturation and rate saturation. In the airframe module The use of combined guidance law (homing and
various forces and moments are calculated. They command) allows achieving a great accuracy of fire
involve aerodynamic, weight, thrust, and control forces even in case of rapid and/or maneuvering; approaching
and moments. The aerodynamics forces and moments and/or receding targets. Fuzzy approach is promising in
are calculated in the velocity coordinate system. the realm of designing missile guidance laws.
However the thrust and weight forces are computed in
the board and reference coordinate systems References:
respectively. Thus, the solution of the dynamical [1] Lightbody, G. and Irwin, G. W., “Neural Model
problem necessitates a reliable means for coordinate Reference Adaptive Control and Application to a BTT-
transformations between these systems. The CLOS Guidance System”, IEEE Int. Conf. on Neural
transformations between these coordinate systems Networks (Orlando, Fla.), 1994, pp. 2429-2435.
achieved by the Euler’s angle method. Finally, the [2] McDowell, D. M., et. al., “Hybrid Neural Adaptive
kinematics module solves the force and moment Control for Bank-to-Turn Missiles”, IEEE Trans. on
equations and produces the missile flight parameters, Control Systems Technology, Vol. 5, No. 3, 1997, pp.
which are the instantaneous acceleration, velocity, and 297-308.
position data of the missile. The flight path variables [3] Lin, C. M., and Maa, J. H., “Flight Control System
are, then derived from the airframe module. Design by Self-Organizing Fuzzy Logic Controller”,
Table 1 shows that the missile succeeds in AIAA J. of GC&D, Vol. 20, No. 1, 1996, pp. 189-190.
interception for all cases. It is clear the significant [4] Geng, Z. J., and McCullough, C. L., “Missile
lower miss distance using FPNCG over the FLOSCG. Control Using Fuzzy Cerebellar Model Arithmetic
Regarding the LATAX, FPNCG was superior in 8 Computer Neural Networks”, AIAA J. of GC&D, Vol.
cases (T2-T9) and inferior in only one case (T1). 20, No. 3, 1997, pp. 557-565.
Figures 9 through 11 show the missile-target [5] Kim, B. S., and Calise, A. J., “Nonlinear Flight
engagement scenario (T3-Table1), control fin Control Using Neural Networks”, AIAA J. of GC&D,
deflection as a result of the guidance commands, and Vol. 20, No. 1, 1997, pp. 26-33.
histories of the resultant missile angle of attack and [6] McFarland, M. B., and Calise, A. J., “Neural
sideslip angle. In this case it is apparent from Fig. 9 Networks and Adaptive Nonlinear Control of Agile
that FPNCG scheme with the advantage of time Antiair Missiles”, AIAA J. of GC&D, Vol. 23, No. 3,
varying navigation ratio results in much tighter 2000, pp. 547-553.
trajectory than the FLOSCG scheme. However in the [7] J. E. Steckt and S. N. Balakrishna, “Use of Hopfield
case of FLOSCG, the existing autopilot is able to track Neural Networks in Optimal Guidance”, IEEE Trans.
the error trajectory and achieve the demanded on Aerospace and Electronic systems, Vol. 30, No.1,
acceleration smoothly. On the other hand, as the 1994, pp. 287-293.
relative distance between the missile and the target is [8] N. Rahbar and M. B. Menhaj, “Fuzzy-Logic-Based
large the proportional navigation can handle the Closed-Loop Optimal Law for Homing Missile
heading error rapidly with a moderate acceleration Guidance”, AIAA J. of GC&D, Vol. 23, No. 3, 2000,
commands. While the relative range becomes small pp. 573-574.
(near the interception) the acceleration command [9] Mishra, S. K., Sarma, I. G., and Swamy, K. N.,
becomes very large and rapid which leads to more “Performance Evaluation of Two Fuzzy-Logic-Based
challenge on the existing autopilot as shown in Fig.s 10 Homing Guidance Schemes”, AIAA J. of GC&D, Vol.
and 11. 17, No. 6, 1994, pp. 1389-1391.
The same principle criteria for performance [10] Choi, H.-L., et. al, “A three-dimensional
evaluation have been used to evaluate the suggested Differential Game Missile Guidance Law Using Neural
FSG presented in section 3.3. The results are tabulated Networks”, AIAA GN&C Conference, Montreal,
in Table 1 and shown in Fig.s 12 and 13. The miss Canada, 2001, A01-37172.
distance is small enough for the missile to hit the [11] Locke, A., “Principles of Guided Missile Design”,
target. The results are found to be encouraging and D. Van Nostrand Co., Princeton, NJ, U.S.A., 1955.

5
 [12] Lin C-F, "Modern Navigation, Guidance, and -Z
Control Processing", Prentice Hall, Englewood cliffs,
NJ, USA, 1991. Vt
[13] Zarchan P., “Tactical and Strategic Missiles Vm
Guidance”, 2nd Ed., AIAA, Washington D.C., U.S.A., rmt Target
1994.
Missile
[14] Garnell, P., and East, D., “Guided Weapon rt
Control Systems”, 2nd Ed., Pergaman press, New York, rm λz
1980. εt
[15] Passsino, K. M., and Yurkovich, S., “Fuzzy
Control”, Addison Wesly Longman, Inc., 1998. X
[16] Mehrato, K., Mohan, C. K., and Ranka S., βt
εm
"Elements of Artefical Neural Networks", MIT Press, βm
2nd Ed., 2000.
[17] Jang, J.-S., Sun, C.-T., and Mizutani, E., "Neuro- Y
Fuzzy and Soft Computing", Prentice Hall, 1997.
[18] Shinar, J., et al, “Mixed Strategy Guidance: A Fig. 4 Geometry of Missile-Target interception
New High-Performance Missile Guidance Law”,
AIAA J. of GC&D, Vol. 17, No. 1, 1994, pp. 129-135.

Measuring
demanded
missile Actual
acceleration Autopilot loop
missile
Angular acceleration
LOS target Guidance limiter
channel Airframe
direction receiver computer

Guidance loop

LOS missile
Space kinematics
direction

Fig. 1 Block diagram of the guidance-autopilot loop

EP gpe
gp Pitch gpu
&P
E steering Apc
Guidance
EY gye commands Ayc
&Y
E gyc Yaw gyu
Guidance
FLOSCG

Fig. 5 FLOSCG scheme

Fig. 2 Missile airframe reference axes
Correction Rate
Circuit Gyro Achieved Body Rate

Control Surfaces
Demanded + Amplifier
- δc δ
deflection
& Control Surface Missile Airframe
+ Actuator Dynamics
Acceleration - Limitier

Feedback Achieved Acceleration

Correction Accelerometer
Circuit

Fig. 3 Simplified block diagram of pitch autopilot.

6
 rmt
Fuzzy
g-bias r&mt Rule Base

rmt N
Fuzzy
Rule Base PN Guidance Apc, Ayc FPNCG
r& W1
Law
-1
vc Missile-target
λ& λ& relative parameters
p, y Apc, Ayc
FLOSCG W2 +

Fig. 6 FPNCG scheme Fig. 7 Supervisory Fuzzy Guidance Scheme

Target Missile-Target guidance fin Airframe

steering
Guidance Autopilot
parameters Geometry parameters commands deflection

fuzzy guidance schemes in Figs 5, 6, 7.

Missile parameters Kinematics

Fig. 8 Simplified block diagram of the main simulation model modules.

Three Dimensional Missile-Target Scenario Control Fin defelection

20

10
8000 Impact
del p

6000 -10
(dash) FLOSCG (sold) FPNCG
Altitude (m)

To -20
4000 0 5 10 15 20 25 30

FPNCG FLOSCG
2000 20

10
0
10000
del y

0
2.5
2 -10
5000 1.5
1 4
x 10
0.5 -20
Cross Range (m) 0 0 0 5 10 15 20 25 30
Down Range (m)
Time (sec)

Fig. 9 Missile-target engagement scenario (T3-Table1) Fig. 10 Pitch and yaw control fin deflection

7
 Missile Angle of Attack and Side-Slip Angle Three Dimensional Missile-Target Scenario
6

2 8000 Impact
α (ο)

0
6000
-2

Altitude (m)
T
-4 o
0 5 10 15 20 25 30 4000 FLOSCG 340 m/sec
FPNCG 2 g Maneuver
4 2000 FSGL (25km, 6Km, 3Km)
(dash) FLOSCG (sold) FPNCG
2
0
β (ο)

10000
0
2.5
2
-2 5000 1.5
1 4
x 10
-4 0.5
0 5 10 15 20 25 30 Cross Range (m) 0 0
Down Range (m)
Time

Fig. 11 Missile angle of attack and sideslip angle. Fig. 12 Missile-target Engagement Scenario (T3)

20 20

10 10
del P

0 del Y 0

-10 -10

-20 -20
0 10 20 30 0 10 20 30
Time (sec) Time (sec)

6 1.5

1
4
Angle of Attack (o)

Sideslip Angle (o)

0.5

2 0

-0.5
0
-1

-2 -1.5
0 10 20 30 0 10 20 30
Time Time

Fig. 13 Pitch and yaw control fin deflection and corresponding α and β in case of employing FSG.

TABLE 1
Target Initial Parameters Miss Distance [m] LATAX
XT, YT, ZT VT AT FLOSCG FPNCG FSG FLOSCG FPNCG FSG
[Km] [m/sec] [g] *e+004 *e+004 *e+004
T1 0 5.35 0.487 0.354 3.1755 4.6043 2.3478
T2 25, 6, 3 340 1 7.59 0.864 0.475 5.3086 5.0548 3.2331
T3 Approaching Target 2 9.56 1.429 1.128 8.4554 6.5736 4.7226
T4 0 6.78 0.680 0.708 3.6902 4.7420 2.3012
T5 430 1 9.86 1.064 1.125 5.7532 5.0165 3.3331
T6 2 12.45 1.641 1.847 8.8391 6.3892 4.7191
T7 0 3.45 0.513 0.255 4.4709 3.8431 3.0257
T8 5, 6, 3 330 1 6.23 1.105 0.985 5.6481 5.2074 2.9300
T9 Receding Target 2 8.52 1.378 1.012 7.6442 6.6642 4.6614