Keywords.

Modern guidance law, Augmented proportional navigation,

Zero-effort-miss

6.3 MODERN GUIDANCE LAWS

Essentially the modern guidance laws attempt to take into account target
accelerations (i.e., target maneuvers), the time delay in the missile autopilot,
and the effect of noise and uncertainty in measurement and estimation, as
these are responsible for a missile’s failure to perform well. Theoretically,
many of them succeed in achieving their goals, but implementation of such
laws become difficult due to their inherent complexity. Much of the appli-
cation research on missile guidance is focused on the problem of efficient
implementation of these laws or incorporation of some of their desirable fea-
tures in classical guidance laws. Below we shall discuss two such guidance
laws : Augmented Proportional Navigation (APN) and the Modern Guidance
Scheme (MGS). Both can be considered to be extensions of the classical PN
law in a linearized geometry. Indeed, we shall develop these guidance laws as
logical extensions to the PN law, though their actual design could be done
more rigorously through optimal control theory.

Augmented Proportional Navigation

In Fig.6.7 we show a missile-target geometry with a non-maneuvering

target, and with a small angle assumption, that is, we assume the LOS angle
θ to be small. This assumption allows the system to be treated as a linear
one. Note that y is the vertical separation between the missile and the target.
The time for intercept is tf and so,

tgo = tf − t

The commanded missile latax for the classical PN law is given by

110
 Figure 6.7: Small angle missile-target engagement geometry

aM = N Vc θ̇ = N Vc dθ
dt

Note that tgo can be written as,

tgo = R/(−Ṙ) = R/vc

and so,

R = Vc tgo

111
Since the angle θ is small,

θ∼
= sinθ = Y
R
= Y
Vc tgo

Then we have

 
d y d
aM = N Vc =N (yt−1 ) (6.1)
dt Vc tgo dt go

= N {ẏt−1 −2
go + y(−1)tgo (−1)} (6.2)

= N {ẏt−1 −2
go + ytgo } (6.3)

N
= (ẏtgo + y) (6.4)
t2go

Now, let us analyze the expression within brackets in the above equation.
It represents the (vertical) miss-distance that will occur at the end of time
tgo (i.e., at final time t), provided that the missile does not apply any latax.
Thus, this quantity is called the Zero-Effort-Miss (ZEM). Hence, the com-
manded missile latax for the classical PN law, in linearized geometry, can be
written as,

N
aM = (ZEM ) (6.5)
t2go

So, proportional navigation (PN) in the small angle case turns out to be
a guidance law which generates a latax command proportional to the ZEM
and inversely proportional to the square of the time-to-go.

112
 We shall now extend this idea further to obtain the APN guidance law.
Suppose, in the same small angle case, we have a maneuvering target such
that it causes a non-zero Ÿ . Note that in the previous case (PN) we had
assumed a non-maneuvering target and so we had only a non-zero Ẏ which
was caused by the target velocity. Then, from simple kinematics, the ZEM
is given by,

ZEM = y + ẏtgo + (1/2)ÿt2go

= y + ẏtgo + (1/2)aT n t2 go

where, ÿ =aT n = target acceleration normal to the LOS. Then, the com-
manded latax is given by,

N
aM = {Y + Ẏ tgo + (1/2)aT n t2go } (6.6)
t2go

This is the Augmented Proportional Navigation (APN) guidance law. How-

ever, the above equation represents the guidance law in a linearized geometry.
The guidance law in the general non-linear geometry is given by,

aM = N Vc θ̇ + (1/2)N aT n (6.7)

This can be easily deduced from the above equation. Note that in addition
to the LOS rate and the closing velocity, the APN guidance law also requires
the target acceleration normal to the LOS for its implementation. In fact
this requirement is its main drawback since measurement (or estimation) of
target acceleration is prone to noise.

113
 The advantage of the APN guidance law, as compared to PN, is that
the commanded latax is initially high but falls as the missile approaches the
maneuvering target. This is shown in Fig.6.8 below. Though the APN law
takes into account the target acceleration, it does not account for the time

Figure 6.8: Command latax for PN and APN for different N and for maneu-
vering target

delay in the lateral autopilot of the missile which causes a difference between
the commanded latax and the achieved latax. The modern guidance scheme
described below takes this delay into account.

Modern Guidance Scheme

The modern guidance scheme is derived using the theory of optimal con-
trol. The latax that the missile pulls as it maneuvers induces a drag which
affects its velocity. In an attempt to minimize this maneuver induced drag,
the MGS guidance law is designed in such a way that it minimizes the fol-

114
lowing quantity (which is a measure of the maneuver induced drag) :
 tf
a2M dt (6.8)
0

under the condition that the terminal miss-distance is zero. In the linearized
geometry (that is, small angle assumption) it means that y(tf )=0. Solving
this problem in a linearized setting, using optimal control theory, and assum-
ing the autopilot to be a first order dynamical system, we obtain the MGS
guidance law as,
  
N 1 2 (e−T + T − 1)
aM c = 2 y + ytgo + ( )aT n tgo − aM a (6.9)
tgo 2 w2

where, aM c is the commanded latex and aM a is the currently achieved latex

of the missile. Also,
T = wtgo (6.10)

w=Bandwidth of the guidance system (autopilot) dynamics which is ex-

pressed as a first order time lag system

1
= τ

where τ is the time constant in the first order system which models the lateral
autopilot. Here, the navigation ”constant” N is no longer a constant but is
a time-varying quantity denoted by N’. It is given by,

6T 2 (e−T − 1 + T )
N = (6.11)
2T 3 + 3 + 6T − 6T 2 − 12T e−T − 3e−2T
The expression within brackets in the expression for aM c above represents the
ZEM. Note that the first part of the guidance law is identical to the APN
law. In fact, as the time delay t → 0, we have T → ∞, we get a perfect
autopilot. Also,

Lim N =3
(6.12)
T →∞

115
Hence, for a perfect autopilot, the guidance command reduces to the APN
guidance law with the optimal value of the navigation constant as 3. This jus-
tifies the choice of N as 3 to 5 in the PN and APN guidance laws. Translated
to the non-linear settings the MGS law is expressed as,
 −T 
 1 e +T −1
aM c = N Vc θ̇ + N aT n − aM a (6.13)
2 T2
The MGS law has all the advantages of the APN guidance law in addition to
the advantage that it takes care of the autopilot time delay. It suffers from
the drawback that it requires the estimation or measurement of aT n , aM a and
tgo which are difficult to measure accurately.

PROBLEMS AND EXERCISES

1. Consider the missile target engagement shown below. Answer the

Figure 6.9:

following questions:

116
 (I) what should be the angle θM in order that the condition for LOS
guidance is met?

(II) Suppose the missile is guided by the pure pursuit guidance law then
what should be θM in order to satisfy the condition of velocity pursuit and
of attitude pursuit? The angle of attack is 2 degrees and the velocity vector
lags the missile longitudinal axis.

(III) Answer (II) when the missile is guided by a deviated pursuit guidance
law with the angel of deviation=2.5 degrees.

2. Consider the missile-target engagement geometry given below at some

instant in time t. Answer the following questions:

(I) What is the commanded latex if the missile uses (a) PN with N=4
(b) APN with N=3 (c) MGS with the current achieved latex same as the
commanded latex in (b) and the time delay t=0.1 sec?

(II) What is the estimated time-to-go? Is the actual time-to-go for the
three guidance laws less, more or the same as the estimated value? Assume
that the target employs a constant measure level, i.e., constant aT throughout
the engagement.

3. Obtain the value of N’ when T → ∞.

117
Figure 6.10:

118