By

Saravanan Elanchezhian
 The name "guided missile" implies that the missile is controlled either by an
internal guidance system or by commands transmitted to the missile by radio from
the ground or launching vehicle.
 A navigation system is one that automatically determines the position of the
vehicle with respect to some reference frame, for example, the earth, and displays
this to an operator. If the vehicle is off course, it is up to the operator to make the
necessary correction.
 A guidance system, on the other hand, automatically makes the necessary
correction to keep the vehicle on course by sending the proper signal to the
control system or autopilot. The guidance system then performs all the functions of
a navigation system plus generating the required correction signal to be sent to the
control system.
 The control system controls the direction of the motion of the vehicle or simply
the orientation of the velocity vector.
 The type of guidance system used depends upon the type and mission of the
missile being controlled, and they can vary in complexity from an inertial guidance
system for long-range surface-to-surface or air-to-surface winged missiles to a
simple system where the operator visually observes the missile and sends
guidance commands via a radio link.
 Irrespective of the type of guidance system used, the guidance command serves as
the input to the missile control system.
 The command may be in the form of a heading or attitude command, a pitching or
turning rate command, or a pitch or yaw acceleration command, depending upon
the type of guidance scheme used.
 Missile Types:
 Flown in the same manner as manned aircraft, missiles that are banked to turn,
such as cruise missiles and remotely piloted vehicles. Control system design is
similar to that of fixed wing aircrafts.
 Other aerodynamic missiles which use aerodynamic lift to control the direction
of flight, such as the Sidewinder, Patriot, etc., and "ballistic missiles" (which are
guided during powered flight by deflecting the thrust vector and become free-
falling bodies after engine cutoff’).
 One feature of these missiles is that they are roll stabilized; thus there is no
coupling between the longitudinal and the lateral modes, which simplifies the
analysis.
 For analyzing the guidance system for ballistic missiles, the rotating earth cannot
be assumed as an inertial reference.
 Roll stabilization can be accomplished by different means, depending on the type
of missile.
 For aerodynamic missiles, the required rolling moment is achieved by differential
movement of the control surfaces.
 For ballistic missiles, the rolling moment can be obtained by differential swiveling
of small rockets mounted on the side of the missile, as is done on the Atlas, or by
differential swiveling of the two main rocket engines if more than one engine is
used.
 The rolling motion need to be detected so that it can be controlled, and the
reduction of roll rate to zero or maintenance of the roll angle equal to some
specified reference.
 The use of a roll rate gyro would not be satisfactory unless it was desirable only to
reduce the roll rate considerably.
 The use of the roll rate gyro would result in a Type 0 system, which would further
result in a steady-state error in roll rate in the presence of a constant disturbing
rolling moment.
 To maintain a desired roll angle, some form of an attitude reference must be used.
This can be a vertical gyro for air-to-air or surface-ta-air missiles.
 In either case the feedback would be a signal proportional to the roll angle about
the longitudinal axis of the missile. Still another possible method is the use of an
integrating gyro with its input axis along the longitudinal axis of the missile.
 where 𝐶𝑙𝛿𝑎 , is the rolling moment
generated by the aerodynamic
General block diagram of a roll stabilization system controls or the reaction controls
divided by 𝑆𝑞𝑑 , and 𝑑 is the
diameter of the missile.
The servo might be represented by a first-order time lag or a The reference area, 𝑆 , is usually
second-order system. taken as the cross-sectional area of
the missile. 𝐶𝑙𝑝 is the change in the
rolling moment due to a rolling
velocity, arises from the change in
the angle of attack on the wings
The transfer function of the missile for 𝛿𝑎 input to roll angle output caused by a rolling velocity. 𝐶𝑙𝑝 , is
would be the same as the one-degree-of-freedom rolling mode due mostly to aerodynamic friction
and thus is negligible.
derived for aircraft.
The requirement for the lead
network is demonstrated by
drawing the root locus of the system
for the worst condition, that is, 𝐶𝑙𝑝 =
0 (roll rate feedback could also be
used for stabilization).

Block diagram of roll stabilization system for
root locus analysis
The two poles at the origin
would move directly into the
right half plane; the effect of the
lead circuit is evident.
If 𝐶𝑙𝑝 , is not zero, the transfer
function of the missile consists
of a pole at the origin and a
1
pole at 𝑠 = − 𝜏 = −ሺ𝑆𝑞𝑑 2 /

Root locus for the roll

stabilization system
 An alternate control system could consist of an integrating gyro with its input axis
parallel to the longitudinal axis of the missile. This would provide the same quality
of control as the attitude reference system as long as the drift rate of the gyro were
low enough.
 The root locus for this system would be the same as shown in previous slide.
 Some aerodynamic missiles
such as the Sidewinder use
control surfaces referred to as
rollerons to reduce the roll rate.
This technique is illustrated in
figure which shows a small
 As illustrated in Figure, the H vector of the gyro wheel is to
gyro wheel attached to the
the right, and if the missile is rolling to the right, the 𝜔 vector
movable control surface on the
is toward the nose of the missile, resulting in a gyroscopic
missile tail fin.
output torque 𝑀𝑔 coming out of the page.
The gyro wheels can be spun
 This torque will cause the control surface to be deflected to up using exhaust gases vented
the right, thus slowing down the roll rate. from the rocket engine or by a
 H is the angular momentum, or moment of momentum, of a short lanyard wrapped around
revolving body. the shaft of the gyro wheel and
attached to the missile
launcher.
 Aerodynamic missile and axis system

 By using a body axis system the product of inertia term Jxz is zero
and Iz = Iy. Thus for P = 0 there is no coupling between the
longitudinal and lateral equations. The aerodynamic data for
these missiles may be given
 Control can be accomplished either by conventional control for either of the configurations.
surfaces with the canards stationary or absent, or by use of the
canards with no control surfaces on the main lifting surfaces. A majority of the data is
obtained from wind tunnel test
 Missiles such as the one shown in Figure are often referred to as using the + configuration.
cruciform missiles due to the Y and Z symmetry of the lifting
surfaces, and are usually flown with a roll angle of 45°. To convert the stability
derivatives and/or force and
 As these missiles use direct side force to tum, they are referred to moment coefficients from the +
as skid-ta-tum missiles, and by flying with a roll angle of 45° to the x configuration it is only
(referred to as the x configuration) all four lifting surfaces are used necessary to multiply the
for pitch and yaw control.
values by 1.414 (2cos45°).
 If the missile is flown with zero roll angle as shown in Figure, it is
said to be in the + configuration.
 As a result of the Y and Z symmetry, the pitch axis short-period approximation
transfer functions can be used for both the pitch and the yaw axis analysis;
however, the analysis need only be accomplished about one of the axes, due to the
symmetry just mentioned.
 Unlike the calculation of the transfer functions for aircraft, the calculation of the
missile transfer functions is complicated by the need to determine a consistent set
of physical properties and flight conditions.
 The physical properties of interest are the mass, moment of inertia, and cg location,
which are functions of the fuel used.
 The flight conditions (altitude and velocity) determine the Mach number, which is
needed to calculate the missile stability derivatives; the latter can also be functions
of angle of attack and control surface deflection.
 During the early design phase, flyouts can be run at different launch conditions by
developing and running a point mass missile simulation, using the design thrust
profile with the corresponding mass and the estimated drag coefficient variation
with Mach number.
 Using the data from selected points from the point mass flyouts, a preliminary
dynamic analysis can be performed.
 Using the results from the initial dynamic analysis, a full missile simulation can be
developed to obtain data for the final autopilot and guidance system design.
 This was the procedure followed in the development of this autopilot, and the
analysis using data from the final missile simulation will be presented.
 From the missile flyout described in Section 8·2 the missile parameters listed in
Table 7·1 for six points in the trajectory were obtained and used to calculate the
corresponding missile transfer functions.
 Missile Parameters for Dynamic Analysis

 From the missile flyout, the

missile parameters listed in Table
for six points in the trajectory
were obtained and used to
calculate the corresponding
missile transfer functions.
 The moment stability derivatives
 For the aerodynamic analysis of missiles, body axis rather than 𝐶𝑚𝛼 and 𝐶𝑚𝛿 are taken with
stability 𝑡
respect to the reference cg and
 axis stability derivatives are used; thus from Table 𝐶𝑧𝛼 = −𝐶𝑁𝛼 , where must be corrected for the location
𝜕𝐶
𝐶𝑁𝛼 = 𝜕𝛼𝑁 and 𝐶𝑁 is the normal force coefficient (perpendicular to the of the actual cg relative to the
reference cg.
X axis and positive up).
 For drag the axial force coefficient 𝐶𝐴 (positive along the negative X
axis) is used rather than 𝐶𝐷 .
 The distances shown measured from the nose of the missile are
considered negative (measured in the negative X direction).
 The aerodynamic forces acting on the missile for tail control 𝛿𝑐 = 0 are
𝐹𝑧 , acting at the center of pressure (cp), and 𝐹𝑡 , assumed to be acting at
the quarter chord point of the mean aerodynamic chord (MAC).
 For the case of canard control with 𝛿𝑐 = 0.
 Such a configuration is
 From Missile Parameters for Dynamic Analysis Table it can be assumed for the missile being
seen that 𝐶𝑚𝑞 and 𝐶𝑚𝛼ሶ . were neglected, as is often the case with studied here. However, if either
missiles. 𝐶𝑚𝑞 or 𝐶𝑚𝛼ሶ is nonzero, then the
 Most missiles will have some pitch damping from the main wing reference values of these
if they are similar to the one shown in Figure without the canards. stability derivatives must be
 If the missile has a fixed wing plus a tail surface, then 𝐶𝑚𝑞 will be corrected for the actual cg
larger. location.

 As 𝐶𝑚𝛼ሶ results from the time required for the downwash

generated by the wing to reach the tail, unless the missile has a
wing and tail there will be no downwash effect.
 This would be true of the missile sketched in the Figure without
the canards.
From the missile flyout, the
missile parameters listed in
earlier Table for six points in
the trajectory were obtained
and used to calculate the
corresponding missile
transfer functions.
Figure shows the block diagram of a typical missile autopilot found in present day air-to-air and
surface-ta-air missiles. As the missile is symmetrical with respect to the Y and Z axes, the pitch and
yaw autopilots are the same. The pitch autopilot is shown in Figure, which employs rate feedback to
damp the short-period oscillations.
 Block diagram of missile autopilot

There are two paths shown

for the rate feedback signal:
one is for the boost phase
(B), the other for the
sustainer and coast phases
(S).
 As the missile is not guided during the boost phase, the integration in the
boost phase loop attempts to zero out the missile pitch rate, thus
providing pitch control during boost (no control required in yaw).
 At the termination of boost the acceleration loop and alternate rate
feedback loop are engaged (switch in S position).
 Accelerations sensed by accelerometer located forward of missile cg

As the accelerometer is not

located at the missile center
of gravity (cg), the ac-
celerometer will sense both
the normal acceleration of
the missile cg 𝑎𝑧 and the
 To understand the choice of signs at the summer in the accelerometer tangential acceleration due
block it is necessary to examine above Figure which shows the to a pitch angular
accelerations sensed by an accelerometer located ahead of the acceleration 𝜃ሷ .
missile cg.
 In the figure 𝑥𝑐𝑔 and 𝑥 𝑎𝑐𝑐𝑒𝑙 are the distances from the nose of the
missile to the missile cg and accelerometer respectively, with both
distances taken as negative.
 For this example 𝑥 𝑎𝑐𝑐𝑒𝑙 = - 2.5 m. Then for the pitch loop, a positive pitch rate (positive angle of attack)
resulting from a negative pitch tail deflection will generate a negative 𝑎𝑧 . This fin deflection will also
generate a positive pitch angular acceleration, resulting in a negative output from the accelerometer. The
sign of 𝑥𝑐𝑔 + 2.5 will be negative, as the magnitude of 𝑥𝑐𝑔 is greater than 2.5. As both acceleration terms have
the same sign, they are summed positively at the summer.
 This is not true for the yaw channel, where both a positive yaw rate (negative sideslip angle) and yaw
angular acceleration result in a positive accelerometer output. But as the sign of 𝑥𝑐𝑔 + 2.5 is negative, the
angular acceleration term must be subtracted from 𝑎𝑦 to obtain the proper accelerometer output.
 Since the cg of the missile moves forward as the rocket fuel is burned, the distance
between the missile cg and the accelerometer decreases, thus reducing the
magnitude of the tangential acceleration sensed by the accelerometer.
 The equations for the outputs of the two accelerometers after substituting for the
accelerometer location are,
 Block diagram of missile autopilot

Using the missile transfer function

for case 1, the block diagram for the
boost autopilot is shown in Figure.

Block diagram for the boost autopilot

The closed loop transfer function is

 Root locus for the
boost autopilot

The root locus is shown in

Figure with the closed loop
poles indicated for 𝑆𝑏 𝑆ሺ𝑟𝑔) = 1.

 From the transfer function it can be determined that the damping ratio
for the complex closed loop poles is 0.41 with a damped natural
frequency of 81.07 rad/sec; thus the real pole at - 0.284 is dominant.
 The boost autopilot proved quite adequate, as the missile pitch angle
decreased by only 0.5° from the launch angle of 19.8°.
 Block diagram of missile autopilot

The block diagram for the

inner loop of the complete
autopilot for case 4 (maximum
dynamic pressure) is shown in
Figure.

Block diagram of the rate loop for the complete missile autopilot

The closed loop transfer function for 𝑆ሺ𝑟𝑔) = 0.15 voIt/(deg/sec) is

Block diagram of missile autopilot

Using the closed loop transfer

function from previous slide
𝜃ሶ 𝑠
, the block diagram for
𝑒𝑞 ሺ𝑠)
the acceleration feedback loop
from can be drawn and is shown
in Figure.

Block diagram of acceleration loop of the complete missile autopilot

The 𝑎𝑧 transfer
function for case 4 is

Then using the 𝜃ሶ transfer function for case 4 from Table (transfer function)
𝑎𝑧 ሺ𝑠)
the 𝜃ሺ𝑠)
ሶ transfer function is

𝑎𝑧 ሺ𝑠)
The ሶ transfer function is the
𝜃ሺ𝑠)
𝑎𝑧 ሺ𝑠)
ratio of the transfer
𝛿𝑡 ሺ𝑠)
𝜃ሶ 𝑠
function to the 𝛿 ሺ𝑠) transfer
𝑡
function.

From Figure From flyout Table 𝑋𝑐𝑔 = −3.16 𝑚;

and 𝑔 = 9.81𝑚/𝑠 2
Simplified block diagram of the accelerometer loop for root locus analysis
𝑎𝑧 ሺ𝑠)
Substituting Eqs. ሶ
𝜃ሺ𝑠)
and into yields

The closed loop transfer function is

Figure shows the zero-angle
root locus for the acceleration
loop and the Bode root locus is
also shown.
The locations of the closed loop
poles for Sa = 1.5 are indicated
in both figures.
From Bode root locus the phase
margin is 85°, and the gain
margin is 3.16. The gain margin
frequency can be obtained from
either figure and is 73.56
rad/sec.
 There is another type of control
 Multiplying above two equations yields the final closed loop
frequently employed for
transfer function for an acceleration input command to a missile
aerodynamic missiles, and that
output acceleration:
is the use of hinge moment
commands to the canards.

 An application of the final value theorem to last equation for a unit step
input shows that the steady-state value of az is 2g.
 This results from the gain of 0.5 in the feedback loop and the integration
in the forward loop resulting in a Type 1 system.
 Blakelock, JH, “Automatic Control of Aircraft and Missiles”, 2nd edition, John Wiley
& Sons, 1991.