A Simple Prediction Method of Ballistic Missile

Trajectory to Designate Search Direction and its

Verification Using a Testbench

Dong Gwan Lee, Kil Seok Cho and Jin Hwa Shin
Agency for Defense Development
Republic of Korea
E-mail: [email protected], [email protected], [email protected]

Abstract— A recent air defense missile system is required to constant velocity relative to a star[1]. The high-fidelity
have a capability to intercept a tactical ballistic missile(TBMs) by kinematic models of TBM target which are highly nonlinear
performing engagement control efficiently. The missile defense and needed for complex numerical integration are crucial in
system has to predict TBM trajectory accurately with cueing order to propagate precisely TBM’s states over a long
information received from an early warning system in order to unobserved time period. Generally, TBM target flying on the
designate search direction and volume to detect/track TBM as earth forms an elliptical trajectory with a given focus at the
fast as it can and also generates necessary engagement center of the earth. If the TBM’s position and velocity are
information. In this paper, we proposed a method to predict given precisely at a moment, the whole elliptical trajectory of
TBM trajectory based on the Kepler's law for the missile defense
the TBM can be obtained mathematically. In this paper, we
system to detect and track TBM using the cueing information
investigate an efficient and effective trajectory prediction
received and analyzed the method of the proposed method in
terms of the bias and standard deviation of predicted position
method based on the Keplerian trajectory model in which the
and velocity errors with respect to the transmission period of elliptical information is well described[5, 6]. We derived the
cueing data between the missile defense system and the early elliptical trajectory of the TBM by solving the two-body
warning system. Also we introduce the test bench system and use problem[3] and also used this model to predict future position
it to verify the proposed method in simulated real-time and velocity of the TBM with small computation load. We
environments. analyzed the performance in terms of the mean and standard
deviation of the predicted position and velocity errors and the
Keywords— tactical ballistic missile; ballistic trajectory sensitiveness as a result of changing tracking errors. After
prediction; testbench; target designation; Kepler’s Law; constructing and simulating the proposed ballistic trajectory
prediction method, we applied it to a practical test bench
I. INTRODUCTION system to verify the method by implementing as a module in
the mission software of the Fire Control Station(FCS). As a
Recently, threatening of tactical ballistic missiles, cruise result, we confirm that its performance might be suitable to use
missiles(CM), and unmanned aerial vehicles(UAV) which in the FCS.
have a capability to carry weapons of mass destruction(WMD)
has gradually increased. Because TBM is the extremely In this paper, Section II describes the derivation of the
threatening, it is essential to track TBM trajectories accurately orbital motion of TBM target and Section III presents the
and kill warheads of TBM. Ballistic missile and projectile prediction method of TBM target trajectory by using cued
tracking uses dynamic models which are governed by forces. target data from the higher echelon(HE). Section IV introduces
In general, the motion of a space object including TBM is the test bench system to verify the TBM target trajectory
influenced by forces due to gravity, atmospheric drag, thrust, prediction. The simulation and verification results are
and sophisticated force models[1]. Under consideration of the presented in Section V and the concluding remarks are
earth gravity and multiple perturbing forces in the complex followed in Section VI.
environment including third body gravity, atmospheric drag,
orbital dynamics of TBM are significantly nonlinear and II. DERIVATION OF ORBITAL MOTION EQUATION
complicated because of many unknown characteristics of TBM
and the space environment[2]. The motion of ballistic missile A ballistic trajectory of a TBM target can be derived
and projectile is governed by the second Newton’s law. We mathematically from the 1st and 2nd Kepler’s law. The 2nd
refer to these equations of motion involving forces as dynamic Kepler’s law can be defined from the 2nd Newton’s law. This
models. The second Newton’s law states that the rate of change means that the angular momentum is conservative. The
of momentum is equal to the total external force acting on the trajectory equation of the TBM target can be obtained by
object in an inertial coordinate system. The inertial coordinate applying two-body problem between the earth and the TBM
system is a coordinate system which is fixed or moving with a target based on two following assumptions.

978-1-4799-7862-5/15/$31.00 © 2015 IEEE

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.
 Assumption 1. Each body is spherically symmetric (Mass is whichG is fixed to space and called as the orbital plane. Crossing (2)
concentrated at its center). into h leads toward a form that can be integrated, then elliptical
Assumption 2. Only the gravitational forces act along the trajectory equation is obtained as
line joining the centers of the two bodies.
According to the above two assumptions, the two-body p
r= (5)
problem cannot be applied to the boost phase in which the 1+ ε cosν
energy is cumulated. It can be only applied to the ballistic
phase in which just gravitational forces act on the body of the where p is orbital parameter, ε is eccentricity. Equation (5) is
TBM target. With the two-body problem between the earth
expressed in the polar coordinates with the origin located at a
and the TBM target, the trajectory equation is obtained as in G
(1). focus and where the polar angle, ν is the angle between r and
the point on the conic nearest to the focus such as perigee [3].
The time of flight(TOF) of a TBM target from one point to
rG = − G ( M + m) rG another point in its orbit is derived as a function of eccentric
(1) anomaly based on the 2nd Kepler’s law as depicted in Fig. 1.
r3
This figure shows an ellipsoid and a circle including the
G ellipsoid with a diameter which is the same as the length of the
where r is the distance vector between the centers of the earth semi-major axis of the ellipsoid[3]. The flight time begins at the
and the TBM target, G is universal gravitational constant, M perigee. In Fig. 1, a is length of semi-major axis, u is the
and m are mass of the earth and the TBM target, respectively. eccentric anomaly, and ν is the true anomaly in (1). r is the
If M >> m , then G ( M + m) ≈ GM ≡ μ , gravitational distance between focus F1 and point P , R is the projected
parameter and (1) is derived as point of point P on the major axis, P' is the cross point of the
line P R and circle.
rG + μ rG = 0 . (2) Relationship between r and u is derived as
r3

Some useful information about the nature of orbital motion r = a (1 − ε cos u ) (6)
and trajectory equation can be obtained from the (2). After dot
G
multiplying (2) by r and some vector algebraic manipulation, and the relationship between u and ν is described as
the specific mechanical energy( E ) normalized to the mass is
given as
ε + cosν
cos u = . (7)
1 + ε cosν
r 2 § μ · r 2
E= + ¨C − ¸ = + Veff . (3) Using (4), (6), and (7), we can get the relationship between
2 © r¹ 2
the TOF( t ) and eccentric anomaly( u ) as follows
The first term of E is kinetic energy per unit mass of the
TBM target and the second term ( Veff ) is the effective potential
a3 τ
energy per unit mass. C in Veff is a constant of integration and t= (u − ε sin u ) = (u − ε sin u ) (8)
μ 2π
can be any arbitrary constant. In convenience, if we assume that
C is set to zero which is equivalent to a zero reference for where τ is the period of elliptical trajectory.
potential energy at infinity, the potential energy of the TBM
G
target is always negative. By doing cross multiply (2) by r and
some vector algebraicG manipulation, (2) is derived as the specific
angular momentum( h ) that remains constant along with its orbit.
G
The expression for h is given as

G G G
h = r ×v (4)
G G
where v is the target velocity vector. Since h is the vector cross
G G
product of r and v , it has to be perpendicular to the plane
G G G
containing r and v . Because h is a constant vector, the two
G G
vectors r and v always remain in the same plane. This means Fig. 1. Definition of eccentric anomaly and true anomaly
that the motion of the TBM target must be confined to a plane

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.
 III. PREDICTION OF TBM TARGET TRAJECTORY 

A. Reference Frame and Elements of orbital motion

To predict trajectory of a TBM target, we construct the
elliptical trajectory in the orbital plane. However, valid
information about the TBM target can be obtained in the
Cartesian coordinates such as ECF(Earth-Centered-Earth-Fixed)
or ENU(East-North-Up) coordinates. Assuming that the rotation
of the earth is ignored during the trajectory propagation, the
ECI(Earth-Centered-Inertial) coordinates is proper to represent
the orbital plane. Thus, the reference frames employed in this
paper are ECI, ECF, and ENU coordinates. In Fig. 2, the ECI
denoted by x I y I z I is a fixed inertial frame at any arbitrary
moment where its origin is at the center of the earth and is a right- Fig. 2. Geometry for the reference frames
handed system. The axis x I points in the direction of the line
where the longitude crossing the GMT and the equator meet. The B. Coordinate Transformation and Prediction
axis z I points in the direction of the north pole. The ECF
Given TBM target information represented in the ECF
denoted by x F y F z F has a rotational motion with the Earth as its coordinate system, the coordinates must be transformed into
origin is in the same position as the ECI. The ENU denoted by the ECI in which the rotation of the earth is compensated.
x N y N z N is fixed on the surface of the earth where a sensor is After that, we rotate it from an initial ECI coordinate system
placed. The axis z N of the ENU is normal to the earth’s surface into x' y ' z ' coordinate system through a rotation angle θ 3
while the axis x N points the east and the axis y N towards to the about the z I = z‫ ڂ‬axis. Then we rotate about the x ' axis into
north. In Fig. 2, Ω E is the angular speed of earth’s rotation, t is the x′′y ′′z ′′ coordinate system through an angle θ1 , the x′′
rotation time from a reference moment. φ , λ , h are longitude, axis remains coincident with the x' axis. Finally, we rotate
latitude, and height of sensor, respectively. about the y ′′ = y axis through an angle θ 2 into the XYZ
coordinate system. This XYZ coordinate system contains the
The elements used in (5)~(8) are as follows. TBM target’s orbital plane and the TBM target’s orbital plane
remains in the X-Y plane of XYZ coordinate system. Fig. 3
illustrates the propagation of a ballistic trajectory using an
Rt2Vtng
2
p= (9) elliptical trajectory in the orbital plane after the coordinates are
μ transformed from ECF into the orbital plane. In general, tactical
data link systems transmit TBM target information represented
in the ECF coordinates. The TBM target’s orbital plane is re-

ε = 1+
(V t
2
)
− 2μ / Rt p
(10)
calculated and updated at every moment when cued target data
are received. After transforming from the Cartesian coordinates
μ into the orbital plane, we note that the angle for rotation matrix
must be calculated to satisfy that the velocity component normal
to the orbital plane becomes zero. The velocity vector only
p consists of X and Y components and the velocity of Y axis is VR
a= (11)
1− ε 2 and the velocity of Z axis is equal to zero.

(
τ = 2πμ 2μ / Rt − Vt 2 )−1.5
(12)

where Vt and Vtng are magnitude of target velocity and tangential

G G G
component of velocity vector in the orbital plane ( Vt = VR + Vtng ),
2
respectively. Here Vt = Vx2 + V y2 + Vz2 , Vtng = Vt 2 − VR2 ,
( )
VR = xVx + yV y + zVz / Rt . x, y, z,V x ,V y ,Vz are represented in
the ECI coordinate system. The Rt is the distance between the
centers of the earth and the TBM target

Fig. 3. Ballistic trajectory propagation using elliptical trajectory

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.
 The total transformation matrix from the ECI to the XYZ is TABLE I. PREDICTION OF POSITION AND VELOCITY USING ELLIPTICAL
TRAJECTORY OF TBM TARGET
given as
཰ Calculation of Δν during the ΔT pred
XYZ
CECI = Rot y′′ (θ‫) ڍ‬Rot x′ (θ‫) ڌ‬Rot z I (θ‫) ڎ‬ (13) a. Using (5), calculate the true anomaly from the perigee at
current time (ν cur )
z' § − V z′′ ·
where θ1 = tan −1 , θ 2 = tan −1 ¨¨ ¸¸ , and b. Inserting ν cur into (7), then calculate the eccentric
2
x' + y ′ 2
© V x′′ ¹ anomaly at current time ( ucur )
§ xI
·
θ 3 = tan −1 ¨¨¸¸ . The TOF and the true anomaly at current c. Inserting ucur into (8), then calculate the TOF at current
© yI
¹ time ( t cur )
time, and prediction interval ( ΔT pred ) are needed to predict a
d. Predicted TOF ( t pred ) from the perigee as the following.
TBM trajectory. The process of trajectory prediction using
elliptical trajectory is summarized in Table I. In Table I, t pred = t cur + ΔT pred
XYZ
coordinate transformation matrix C orbital plane is given by
e. Calculate u pred at the predicted TOF by inserting t pred
into (8) and solving (8). Newton-Raphson method is
plane = RotZ (Δν ) Rot X (π ) .
XYZ
Corbital (14) applied to solve this equation.
Practically, we have to transform the position and velocity in f. Transforming the u pred to ν pred .
the ECI to the ECF or the ENU Cartesian coordinates.
g. Calculate Δν as the following.
IV. TEST BENCH FOR VERIFICATION Δν = ν pred −ν cur
Air & missile defence C2 systems are required to manage
air tracks including TBM tracks and their related tactical ཱ Calculate the predicted range at t pred as follows.
information propagated by other systems such as higher
echelon systems through tactical data links. In order to develop p
RTD =
the track management software including the proposed method 1 + ε cos(ν pred )
in this paper, we need a system-level test bench that consists of
air & TBM track generation tools based on tactical scenarios ི Calculate the predicted velocity as follows.
and track reporting simulators over the tactical data links. In
this section, we introduce a test bench composed of simulation μp
control modules, object generation modules, system simulators Vtng =
RTD
interconnected with HLA/RTI(High Level Architecture/Run-
Time Infrastructure) simulation network, and a shared-memory 2 2
network which is used for rapid data transmission. Fig. 4 VR = VTD − Vtng
depicts the test bench configuration to verify track
management software containing the proposed method. In Fig. 2μ (R t − RTD )
4, the control manager generates various kinds of offensive and VTD = Vt 2 +
Rt RTD
defensive scenarios that an operator wants to make and
subscribes the scenarios to participating simulators.
VTD is the magnitude of the velocity at time t pred and is
calculated according to the law of energy conservation.
ཱི Transformation from orbital plane to the ECI
ªsin(Δν ) º
PECI = (C )XYZ T
ECI
«cos(Δν )» R
« » TD
«¬ ‫ڋ‬ »¼

ªVtng º
VECI = ( )
XYZ T XYZ «
CECI Corbital plane «VR »
»
«‫» ڋ‬
¬ ¼
Fig. 4. Test bench configuration

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.
 In addition, the control manager can start or stop 9
x 10
4

1500

simulations and controls other modules remotely. The threat 8

simulator and the own missile simulator model air target, TBM 7

target and own missile dynamics and generate trajectories of

6

Total velocity [m/sec]

1000
5

target and missile objects, respectively. The SAM battery

Height [m]
4

consists of the FCS, the radar simulator, and the launcher 3

500

simulator. The battery operates cooperatively in coordination 2

with the HE system. The FCS integrates and manages track

1

data received from both the radar and the HE to track tactical -1
0 50 100 150 200 250 300 350
0
0 50 100 150 200 250 300 350

ballistic missiles accurately. When the FCS receives cueing time [sec] time [sec]

Fig. 5. eight and total velocity profiles of 300km SRBM

data from the HE, it makes target designation information such
as the predicted position and velocity of the target in the future 16
x 10
4

2500

and then requests the radar to detect and track the designated 14

target. After that the radar tracks the target, the FCS can 12 2000

engage the TBM target. Therefore, the radar simulator

10

Total velocity [m/sec]

8 1500

simulates searching, detecting, and tracking of the target based

Height [m]
6

on the target designation request of the FCS. The launcher 4 1000

simulator simulates the launch sequence of own intercept 2

missiles and their operational status. The HE and the FCS are
0 500

-2

connected through Link-16 tactical data link and exchange J- -4

0 50 100 150 200 250 300 350 400 450
0
0 50 100 150 200 250 300 350 400 450

messages related to the TBM engagement mission each other. time [sec] time [sec]

This test bench can simulate the simultaneous engagement Fig. 6. Height and total velocity profiles of 500km SRBM
against several TBM targets.
around 80km(slant range) from the FCS with respect to various
TOs(12sec, 4sec and 2sec) in case of 300km SRBM and
V. SIMULATION RESULTS
500km SRBM, respectively. Normal time interval between
In this section, we present two kinds of methods which are TOs in the Link-16 network is 12sec, but it is possible to
the Monte Carlo simulation and the simulation using the test reduce the time interval between TOs by using the DUR (Data
bench to verify the performance of the proposed method. Update Request). It is assumed that the tracking errors in polar
Firstly, we are taking into account the performance of the coordinates and radial velocity are white Gaussian with
proposed method in terms of the mean and standard standard deviation such as 50m, 1.5mrad, 30m/sec in the HE
deviation(STD) of the predicted position (range, azimuth, system which generates TBM target states in the ECF every
elevation) and the total velocity in the FCS point of view 100msec. In Table II and III, we show that the performance of
through 100 Monte Carlo simulations. Also we compare the the proposed method is enhanced by reducing the time interval
performance with respect to some different noise sets. between TOs. Even though the TO is 12 second, we confirm
Secondly, the proposed method is tested and verified using the that these position and velocity errors are suitable to designate
test bench introduced in Section IV. We assume the following direction of search and detection for the employed local radar
conditions in the simulation study. based on the cueing data received from the HE through the
Link-16 data link. These kinds of trends can be shown in case
• The HE transmits the most recently updated TBM data
of other types of TBM trajectories such as lofted and depressed
at each transmission opportunity(TO). which are not mentioned in this paper. Furthermore, we
• No transmission delay between the HE and the FCS. investigate the sensitiveness of the proposed method by
changing the standard deviation of tracking errors in the HE.
• No loss due to any coordinate transformation (i.e. no Because the accuracy of velocity affects the total
coordinate alignment error). transformation matrix in (13) dominantly, we used the tracking
error parameters in which the position errors are fixed, but the
• The earth model is WGS-84. radial velocity error is varied. Table IV illustrates the
We considered two short-range ballistic missile (SRBM) simulation results obtained by 100 Monte Carlo runs. The
trajectories (300km and 500km) [4, 7] to simulate the proposed tracking errors in Table IV are denoted that the first row is the
method. Fig. 5 and Fig. 6 show the height and total velocity STD of range error, the second is that of azimuth and elevation
profiles versus the flight time of 300km and 500km SRBM in angles, and the third is that of velocity error regarding 500km
the ENU coordinate system at the radar site, respectively. The SRBM. As shown in Table IV, the performance results of the
location of the radar site is the same as that of the FCS which is proposed method demonstrate that the predicted position and
located on the ground impact point. The trajectory is originally velocity errors have somewhat biases, and the less tracking
generated in the ENU coordinate system positioned at the errors, the more accurate prediction performance in terms of
launch site and then transformed into the radar site. Table II the standard deviation especially in case of range, elevation,
and Table III summarize the mean and the STD of the and velocity. The biases are due to the implementation
predicted position (range, azimuth angle, elevation angle) and approach of the Newton-Raphson method used to find the
total velocity errors at the point that each SRBM is deployed at predicted eccentric anomaly ( u pred ) which produces a large
position error even if its value is very small.

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.
 TABLE II. MEAN AND STD OF PREDICTD POSITION AND VELOCITY
ERRORS AT EACH TOABOUT 300KM SRBM @80KM FROM THE FCS

Time interval between TOs

Prediction Error 12 sec 4 sec 2 sec
range Mean 269.78 21.67 71.75
[m] STD 376.79 212.59 224.33
azimuth Mean 0.1691 0.1289 0.2446
[deg] STD 0.1269 0.1352 0.1470
elevation Mean 0.2033 -0.0152 0.0146
[deg] STD 0.0952 0.1043 0.0842
velocity Mean -2.3652 -1.3378 -0.9907
[m/sec] STD 32.5101 27.8925 28.4578

TABLE III. MEAN AND STD OF PREDICTD POSITION AND VELOCITY

ERRORS AT EACH TOABOUT 500KM SRBM @80KM FROM THE FCS

Time interval between TOs

Prediction Error 12 sec 4 sec 2 sec
range Mean -431.58 -327.81 -323.67
[m] STD 140.01 167.5 189.1
azimuth Mean 0.1575 0.1453 0.1597
[deg] STD 0.1020 0.1157 0.1152
elevation Mean -0.3368 -0.2783 -0.2722
[deg] STD 0.0943 0.0863 0.1028
velocity Mean 7.1581 4.3063 3.7548
[m/sec] STD 27.4772 27.3573 31.8217 Fig. 7. Comparison of predicted and true profiles of position and total
velocity of the 500km SRBM with 12 second TO
TABLE IV. SENSITIVENESS BY CHANGING THE STANDARD DEVIATION OF
TRACKING ERRORS ABOUT 500KM SRBM @80KM FROM THE FCS

Tracking error 0m 20 m 20 m 20 m
0 mrad 1.5 mrad 1.5 mrad 1.5 mrad
PredictionError 0 m/s 0 m/s 50 m/s 100 m/s
range Mean 82.79 80.56 60.75 -55.91
[m] STD 0 56.92 582.5 1141
azimuth Mean 0.1645 0.1686 0.1635 0.1672
[deg] STD 0 0.022 0.021 0.025
elevation Mean -0.013 -0.012 -0.012 -0.011
[deg] STD 0 0.035 0.047 0.061
velocity Mean -0.337 -0.347 1.637 11.94
[m/sec] STD 0 0.08 48.97 98.25

Fig. 7 and Fig. 8 depict the predicted(black circle) and

true(blue square) profiles of position and total velocity of the
500km SRBM in case of the 12 and 4 second TO, respectively.
These results are the mean values of position and total velocity
in the ENU coordinate system located at the center of FCS
obtained by 100 Monte Carlo runs. The slant range of SRBM
viewed on the FCS is decreased approximately from 128km to
75km. The prediction in the FCS is performed periodically (in
every time interval of 1 sec between TO) and terminated at a Fig. 8. Comparison of predicted and true profiles of position and total
velocity of the 500km SRBM with 4 second TO
slant range of around 75km. Even though the TO is 12 second,
the observed trends of predicted mean position and total
velocity are similar to those that are obtained from the case of 4
second TO. Fig. 9 shows the standard deviation of predicted
range errors which is increased between TO. The circles(red)
indicate the moment when the cueing data is received from the
HE at every TO(12sec and 4sec, respectively) and the
arrows(black) indicate the moment that the trajectory prediction
is terminated at the point that the SRBM target is deployed at
around 75km from the FCS. As shown in Fig. 9, the level of
increment of range prediction error depends on time length of
the TO and the time of prediction termination. Also the
precision of prediction may be affected by the accuracy of the Fig. 9. Standard deviation of range errors at each TO about 500km SRBM
cued TBM state estimates that the HE provides at each TO.

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.
 Furthermore, we implemented this proposed method as the TABLE VI. COMPARISON OF EXECUTION TIME PER TO
(PROPAGATION TIME: 12SEC)
FCS software and tested it in the test bench. The radar simulator
determines whether the difference between the designated TBM Method Execution Time/TO (msec)
target position generated by the proposed method and the Proposed 0.419
detected position by the radar simulator is within a 10 D ×10 D Single EKF:propagation 4.4
window. The radar simulator is also equipped with a tool for ȕ-VSIMM:propagaion 10.80
analyzing results of simulation. Fig. 10 shows the logged data
to be analyzed in the radar simulator. VI. CONCLUSIONS
Table V shows an example of the test results which are In this paper, we investigated the method based on the
achieved in the radar simulator according to the target Kepler orbit equation to predict the trajectories of TBM targets
designation request by the FCS. The coordinates are denoted in using the track information cued from remote systems. The
the ENU coordinate system. These results are obtained during simulation results show that the proposed method is a
the simultaneous engagement of four TBM targets. If the radar promising prediction method of TBM target trajectories for
simulator is successful in detecting the TBM target by using engagement of lower-tier missile defense systems whose
the designated position, the GUI panel of the FCS displays the detection range is short compared with early warning radars
detected TBM target in the screen. We confirm that the FCS because of the system limitations. Moreover, we were known
can engage the TBM target detected by its local radar based on that this approach used to implement the proposed method
the cueing data received from the HE. might cause an unexpected bias.

Table VI shows comparison results of the execution time For future research, we will examine more deeply in
per TO (propagation interval time is 12sec). In short, the practical points of view how sensor/impact point positions and
proposed method is much faster than those of using numerical various shapes of TBM trajectories influence on their
integration of the TBM dynamics for propagation such as the simulation results. In addition, the test bench will be upgraded
single EKF (approximately 10 times) or ȕ-VSIMM[8] and implemented with an automatic analysis function of the
(approximately 25 times). logged data.

REFERENCES
[1] Mahendra Mallick, Steve Rubin, and Ba-Ngu Vo, “An Introduction to
Force and Measurement Modeling for Space Object Tracking,” in 16th
International Conference on Information Fusion, Fusion 2013, Istanbul,
Turkey, July 2013.
[2] K. DeMars, M. Jah, D. Giza, and T. Kelecy, “Orbit determination
performance improvements for high area-to-mass ratio space object
tracking using an adaptive Guassian mixtures estimation algorithm,” in
21st International Symposium on Space Flight Dynamics, 2009.
[3] Roger R. Bate et al, Fundamental of Astrodynamics. Dover Publications,
Inc., 1971.
[4] Francis J. Hale, Introduction to Space Flight, Prentice Hall, 1994.
[5] Pei, R. Y., “Instantaneous Impact Point Analysis,” Journal of the
Institute of Navigation, Vol. 14, No. 2, 1967, pp. 195-204.
[6] Baker, Jr. R. M. L., Mucha T. J., Darby D. R., Jacoby N. H., Johnson, Jr.
Fig. 10. Tool for analyzing logged data
A. W. and Ryan R. E., “Range-Safety Debris-Pattern Analysis,” The
Journal of Astronautical Science, Vol. 23, No. 4, 1975, pp.287-323.
[7] Zarchan, P., Tactical and Strategic Missile Guidance, 3rd Edition, AIAA, 1997.
TABLE V. EXAMPLES OF THE RESULTS OF TARGET DESIGNATION
REQUEST OF THE FCS (ENU) [8] Tae Han Kim, Kyung Rok Moon, and Taek Lyul Song, “Variable-
Structured Interacting Multiple Model Algorithm for the Ballistic
X Y Z Coefficient Estimation of a Re-entry Ballistic Target,” International
Journal of Control, Automation and Systems, Vol. 11, Issue 6, 2013,
Designated position
-7,188m 93,458m 78,419m pp.1204-1213.
TD by the FCS
Result #1 Detected position
-7,986m 91,664m 78,398m
by radar simulator
Designated position
134m 93,456m 80,851m
TD by the FCS
Result #2 Detected position
-45m 91,588m 79,524m
by radar simulator

Authorized licensed use limited to: INSTITUTO MILITAR DE ENGENHARIA. Downloaded on September 01,2022 at 13:02:52 UTC from IEEE Xplore. Restrictions apply.