Note: Descriptions are shown in the official language in which they were submitted.
CA 02684930 2009-10-29
-1-
WO 2008/138433 PCT/EP2008/002640
METHOD FOR FINDING THE BEARING OF A SOUND-EMITTING
TARGET
The invention relates to a method for finding the
bearing of a sound-emitting target by means of an
elongated underwater antenna which has a multiplicity
of electroacoustic transducers, as claimed in the
precharacterizing clause of claim 1.
In a known direction-finding method for passive
location of a sound-emitting, that is to say sound-
producing, sound-reflecting or sound-scattering target
or object (EP 01 308 745 Bl), a so-called linear
antenna is used having a multiplicity of
electroacoustic transducers arranged in one or more
rows. Linear antennas such as these are, for example,
so-called towed arrays or flank arrays arranged on the
vessel hull or boat body. The linear antenna covers a
reception sector within which incident sound which is
emitted from the target or object and propagates in the
water is received by the transducers. In order to
determine the direction of the incident sound, a fan of
directional characteristics or beams which covers the
reception sector is generated by means of a so-called
beamformer, using the signals received by the
electroacoustic transducers. Every directional
characteristic, which can be scanned electronically in
the horizontal direction with respect to a reference
direction, has a relatively small horizontal beam angle
and a vertical beam angle of greater or lesser size
depending on the number of vertically arranged
electroacoustic transducers, as well as a main
direction with maximum reception sensitivity. The
CA 02684930 2009-10-29
WO 2008/138433 - 2 - PCT/EP2008/002640
horizontal direction at right angles to the underwater
antenna, also referred to as the lateral direction, is
normally chosen as the reference direction. In order to
produce the various directional characteristics or
beams, the signals received by the transducers are
delayed in time, to be precise such that they are in-
phase for the respective scan angle of the main
direction of the directional characteristic, and the
in-phase received signals are added to form so-called
array signals, which form the directional
characteristic. The delay times for the signals
received by the individual transducers are calculated
on the basis of the speed of sound measured at the
antenna location, the position of the transducers
within the underwater antenna, and the scan angle of
the respective directional characteristic. When sound
is received, the fan of directional characteristics is
searched for that directional characteristic in which
there is a sound reception maximum. This is determined
by detecting the level maximum of the array signals
which form the directional characteristics. The scan
angle of the main direction of the directional
characteristic is emitted as the target bearing. The
target bearing is presented in numerical or graphic
form on a display.
In linear antennas such as these, both the horizontal
and the vertical beam angle of the directional
characteristics or beams vary with the scan angle of
the main direction of the directional characteristic or
the magnitude of the bearing angle. In the case of a
bearing angle of 0 laterally with respect to the
underwater antenna, the horizontal beam angle is at its
narrowest, and the vertical beam angle is at its
widest. As the bearing angles increase with respect to
the reference direction in the forward or astern
directions, the vertical and horizontal beam angles of
the directional characteristics approach one another.
CA 02684930 2009-10-29
WO 2008/138433 - 3 - PCT/EP2008/002640
By way of example Figure 1 shows the -3dB contour line
of a directional characteristic or beam of a linear
antenna for finding a target bearing, for three
different scan angles of its main direction, that is to
say for three different target bearings or bearing
angles (3. The bearing angle (3 is plotted on the
abscissa, and the vertical sound incidence angle y is
plotted on the ordinate. While the directional
characteristic is symmetrical in the vicinity of the
lateral direction ((3=0 ), the -3dB contour line for a
target whose bearing is astern or ahead bulges "like a
banana". If the sound comes exclusively from the
horizontal direction, then the maximum of the
directional characteristic always lies on the (3
coordinate, precisely at the indicated target bearing.
However, if the sound comes from a vertical direction,
as is indicated by way of example by the dashed
straight line running parallel to the (3 coordinate,
then the maximum of the directional characteristic runs
along the dashed-dotted line of the maximum reception
sensitivity indicated in Figure 1 to greater horizontal
bearing angles (3, which leads to an offset in the
actual bearing, that is to say an excessively large
bearing angle (3 is in principle indicated. The bearing
error that results in this case is symbolized in
Figure 1 by the double-headed arrow on the (3
coordinate. As the magnitude of the bearing increases,
that is to say as the bearing angle increases, this
offset increases. In consequence, in the case of
vertical sound incidence, systematic bearing errors
occur as a function of the bearing angle (3 in all
linear underwater antennas, or underwater antennas
which are similar to a linear form, because of their
large vertical beam angle 20_3dB of, for example, 75 (
37.5 upwards and downwards), as is illustrated in the
diagram in Figure 2.
The invention is based on the object of specifying a
method for finding the bearing of targets of the type
CA 02684930 2009-10-29
WO 2008/138433 - 4 - PCT/EP2008/002640
mentioned initially, which is subject to less error and
therefore produces more accurate bearings.
According to the invention, the object is achieved by
the features of claim 1.
The method according to the invention has the advantage
that the very large bearing errors which occur in the
so-called endfire region of the underwater antenna,
particularly for bearing angles which are well away
from the lateral direction, are virtually largely
eliminated, and equally reliable bearings are therefore
obtained in all bearing directions of the linear
underwater antenna. At large water depths, in which the
sound propagation model calculates a sound ray profile
which does not change in an angle range around the
measured target bearing, a single correction, derived
from the vertical sound incidence angle, to the speed
of sound included in the time delays of the received
signals is sufficient to obtain a very accurate
bearing, even in the "endfire" region, by adaptation of
the directional characteristic. At shallow water
depths, in which the bottom profile of the water
channel can normally change significantly in different
bearing directions, and the sound propagation can
therefore change significantly, an improved target
bearing is obtained by iterative determination of the
vertical sound incidence angle and repeated correction
of the speed of sound, continually, with this target
bearing approaching a convergence value, which
indicates the minimized-error bearing to the target,
after a small number of iterations.
Expedient embodiments of the method according to the
invention together with advantageous developments and
refinements of the invention will become evident from
the further claims.
CA 02684930 2009-10-29
WO 2008/138433 - 5 - PCT/EP2008/002640
According to one advantageous embodiment of the
invention, the reciprocal of the cosine of the
determined vertical sound incidence angle is used as
the correction factor for the multiplication by the
sound propagation speed.
The invention will be described in more detail in the
following text with reference to one exemplary
embodiment, which is illustrated in the drawing, in
which:
Figure 1 shows, by way of example, a -3dB contour
line, plotted against the bearing or the
bearing angle R and the vertical sound
incidence angle y, of a directional
characteristic of a linear underwater
antenna for three different bearing
angles,
Figure 2 shows a schematic illustration of the
bearing error of the linear antenna as a
function of the bearing R,
Figure 3 shows a block diagram in order to
explain the direction-finding method,
and
Figure 4 shows a schematic illustration of a
sound ray profile calculated by means of
a sound propagation model within a
vertical beam angle 20_3dB of the linear
antenna for a bearing angle of R=45 .
The method for finding the bearing of a sound-emitting
target by means of an elongated underwater antenna will
be described in more detail in the following text using
the block diagram illustrated in Figure 3. In this
case, a sound-emitting target means an object which is
at a location remote from the antenna and produces
CA 02684930 2009-10-29
WO 2008/138433 - 6 - PCT/EP2008/002640
sound, reflects sound or scatters sound back, which
sound propagates in the water and is received by the
underwater antenna. Examples of an elongated or linear
underwater antenna, or an underwater antenna which is
similar to a linear form, are so-called towed arrays,
which are towed by a watercraft, or flank arrays which
are arranged on the vessel hull or boat body of
underwater vehicles, or else PRS (Passive Ranging
Sonar) antennas which are fitted to a watercraft.
In principle, in all antenna systems, the target
bearing, that is to say the bearing angle, which the
bearing direction to the target includes with a
reference direction, for example the horizontal
perpendicular to the underwater antenna, is determined
from the output signals from the electroacoustic
transducers, by means of signal processing that is
carried out in a bearing apparatus. To do this, it is
necessary to enter the current value of the speed of
sound at the location of the underwater antenna in the
bearing apparatus, with this value generally being
measured in advance in the vicinity of the underwater
antenna. The bearing apparatus is designed differently
depending on the underwater antenna being used and,
when using a linear antenna with a multiplicity of
transducers arranged in one or more rows alongside one
another, is fundamentally different than when using a
so-called PRS antenna, which comprises three reception
bases, which are arranged at a relatively long distance
apart from one another and are composed of
electroacoustic transducers which are arranged aligned
with one another. The term linear antenna subsumes the
abovementioned flank arrays and towed arrays.
Figure 3 schematically illustrates a linear antenna 10
in the form of a flank array and with a multiplicity of
electroacoustic transducers 11 arranged in one or more
rows alongside one another. In general, the
electroacoustic transducers 11 are arranged on an
CA 02684930 2009-10-29
WO 2008/138433 - 7 - PCT/EP2008/002640
antenna mount 12 at a constant distance d from one
another. In the case of a flank array, each of the
transducers 11 which are arranged alongside one another
on the horizontal, also referred to as a stave,
normally comprises a plurality of transducer elements
arranged on the vertical, although this is not
illustrated any further here. The number of vertical
transducer elements, which are preferably located at
equal distances from one another, is considerably less
than the number of transducers 11 arranged in one or
more rows horizontally alongside one another for beam
forming. The output signals from the transducer
elements which are arranged vertically one above the
other are added and normalized and form the signals
received by the transducers 11 or staves, referred to
in the following text as received signals. In the case
of so-called towed arrays, there are no transducer
elements arranged vertically one above the other so
that, in this case, the output signals from the
transducers 11 directly represent the received signals.
The bearing apparatus 17 which is connected to the
outputs of the transducers 11 or staves comprises a
beamformer 18, a reception level measurement device 19
and a level maximum detector 20.
In order to find the bearing of a target, a fan 13 of
directional functions or directional characteristics
14, also referred to as a beam fan, is formed in the
bearing apparatus 17 by means of appropriate signal
processing using the signals received from the
electroacoustic transducers 11 or staves, with each
directional characteristic 14 (also referred to as a
beam) having a narrow horizontal beam angle and a
relatively wide vertical beam angle, which is dependent
on the number of transducer elements arranged
vertically one above the other in each transducer 11.
The axes 15 of the directional characteristics 14 or
beams represent the main direction of the directional
characteristic in which the maximum reception
CA 02684930 2009-10-29
WO 2008/138433 - 8 - PCT/EP2008/002640
sensitivity occurs. These are defined and are produced
electronically by a horizontal scan angle or beam angle
(3 with respect to the common reference direction 16.
In order to form the directional characteristic 14, the
signals received from the transducers 11 or staves are
delayed in time or phase in the beamformer 18, to be
precise such that they are in-phase for one specific
reception direction or bearing direction (3j. The delay
times ii,j which are` calculated using:
2i, j= i- d- sin Rj = Cset (1)
are for this purpose stored for all reception
directions j(j=1...m) and all transducers 11 or staves
i(i=0...n) in the beamformer 18. d is the horizontal
transducer separation on the antenna mount 12, and cset
is the value of the speed of sound in water as entered
in the beamformer 18. The value of the speed of sound
cmeas as measured at the antenna location before the
start of the direction-finding process is assumed as
the start value for eset. The in-phase received signals
which are obtained in each reception direction in this
case are added to form so-called array signals. The
levels of the array signals are measured in the
reception level measurement device 19 and are stored,
associated with the reception directions j or the
bearing angles (3j. A level maximum detector 20
determines the level maximum and emits the bearing
angle, which is associated with the array signal with
the level maximum, as the target bearing (3zk where
k=1,2...K. This array signal with the maximum level
represents the directional characteristic 14 in which a
sound reception maximum occurs.
In order to reduce the initially described bearing
errors of the bearing apparatus 17, which are
considerable in particular for relatively large bearing
angles (3, the sound ray profile for a sound propagation
CA 02684930 2009-10-29
WO 2008/138433 - 9 - PCT/EP2008/002640
direction which is predetermined by the target bearing
pzx obtained is generated in the block 21 using a sound
propagation model. Acoustic sound propagation models
such as these are widely known. A listing can be found
in: Heinz G. Urban "Handbuch der Wasserschalltechnik"
[Manual of Water Sound Technology] STN ATLAS Elektronik
GmbH, 2000, pages 305 and 306.
By way of example, Figure 4 schematically illustrates
the calculated profile of the sound rays within a
vertical beam angle of the linear antenna 10 of about
75 for a sound propagation direction which is
predetermined by a target bearing (3Z1=45 , assumed by
way of example. Depth intervals are plotted on the
ordinate, and range intervals on the abscissa. The
shaded area represents the seabed. A vertical sound
incidence angle y at the antenna location is determined
from this sound ray profile for an estimated target
range and an estimated target depth. In the
illustration in Figure 4, the antenna is located in the
vicinity of the origin of the coordinate system. The
target range is either known or is estimated by means
of other sensors or methods. For example, the target
range can be taken from a method, which is also carried
out for data support purposes, for passive
determination of target data, for example as described
in DE 101 29 726 Al. However, it can also be made more
precisely from previous target bearings. In the diagram
in Figure 4, the target range of the target that is
annotated Z is assumed to be 47 000 yards. The target
depth is likewise estimated, for example in this case
to be 10 m.
From the sound ray diagram, that sound ray originating
from the target Z and having the least attenuation is
determined. In the example in Figure 4, two sound rays,
inter alia, propagate from the target Z which is
located at a depth of 10 m and arrive at the antenna
location at a vertical sound incidence angle of 3.75
CA 02684930 2009-10-29
WO 2008/138433 - 10 - PCT/EP2008/002640
and 22.5 . The high attenuations during the sound
propagation are caused by reflections on the water
surface and on the seabed. While the sound ray which
originates from the target Z and arrives at the antenna
at a vertical incidence angle of 3.75 experiences
multiple reflections on the water surface, the sound
ray which originates from the target Z and arrives at
the antenna at a vertical incidence angle of 22.5 for
example is attenuated only by propagation losses in the
water, and therefore to a considerably lesser extent,
because of the lack of bottom reflection and surface
reflection. This incidence angle Yk of 22.5 is
therefore determined as the vertical sound incidence
angle yl associated with the target bearing (3Z1.
A correction factor is calculated in the correction
block 22 from the vertical sound incidence angle Yk at
the antenna location obtained by means of the sound
propagation model in the block 21, as the reciprocal of
the cosine of the vertical sound incidence angle yk.
This correction factor is used to correct the measured
value of the speed of sound cmeas used to form the
directional characteristics 14 and the array signals at
the start of the direction-finding process. For this
purpose, the speed of sound Cmeas is multiplied by the
correction factor in the speed of sound correction
element 23, using:
cset = cmeas = 1 where k = 1, 2...K (2).
cos yk
This new value of the speed of sound is now set in the
beamformer 18. A direction-finding process is now
carried out once again in the manner as described above
using the delay times ii,j that have been modified in
this way from the stored delay time record, thus
resulting in a better target bearing (3zk which, in the
abovementioned example with a first target bearing of
CA 02684930 2009-10-29
WO 2008/138433 - 11 - PCT/EP2008/002640
(3Z1=45 , now leads to a better target bearing of
(3z2=41. 3 .
At great water depths, the bottom profile does not
change over a relatively large area around the bearing
direction, which means that the sound ray profile as
calculated by means of the sound propagation model in
the block 21, and as illustrated in Figure 4, is still
valid for the better target bearing (3Z2 and therefore
produces an identical vertical sound incidence angle
of, for example, 72=22.5 . The direction-finding process
can therefore be ended, and the better target bearing
(3Z2 as determined by the bearing apparatus 17 is output
and indicated as the bearing of the target (3z.
In shallow water regions, when the water depth is
small, the bottom profile changes in different sound
propagation directions, as a result of which a
different sound ray profile is calculated for the new
target bearing in the sound propagation model, and this
results in a different vertical sound incidence angle
Yk. For the better target bearing that is obtained in
the example of (3z.,=41.3 , the sound ray profile
calculated once again in block 21 using the sound
propagation model results in a different vertical sound
incidence angle yz for the same target range and target
depth. The better target bearing (3Z2 of (3z2=41 . 3 in the
example that is obtained therefore still contains a
bearing error, although this bearing error is reduced.
In order to eliminate this as well, the better target
bearing (3zk (in the example (3Z2=41. 3 ) is subjected to
the same procedure as the first target bearing (3z(k_1)
obtained (in the example (3Z1=45 ). The sound ray profile
for the target Z, which is assumed to be at the same
range and depth, is calculated once again for the
better target bearing (3zk (in the example (3Z2=41 . 3 ) by
means of the sound propagation model in the block 21. A
new vertical sound incidence angle yk of, for example,
y3=33.75 is now determined from this sound ray profile,
CA 02684930 2009-10-29
WO 2008/138433 - 12 - PCT/EP2008/002640
and a new correction factor is calculated in the
correction block 22 by calculation of the reciprocal of
the cosine of this new vertical sound incidence angle
yk. Using the new sound incidence angle, in the example
73=33.75 , the new value of the speed of sound cset to be
set in the beamformer 18 is once again calculated in
the speed of sound correction element 23 using equation
(2). The bearing of the target is once again found
using the set of time delays that has been changed as a
consequence of the new cset in the beamformer 18,
resulting in a new better target bearing, for example
(3Z3=37.84 . This described process is repeated until the
vertical sound incidence angle yk determined using the
most recently obtained, and once again improved, target
bearing (3zk from the sound propagation model no longer
changes within predetermined limits. When this is the
case, then the target bearing (3zk which has been
determined most recently by the bearing apparatus 17
and has once again been improved is output and
indicated as the bearing of the target (3z. If the
vertical sound incidence angle y3=33.75 in the last
indicated example were no longer to change, then the
bearing of the target would be (3z=37.84 .
The equality in the vertical sound incidence angles 7k-1
and yk determined in successive runs k using the sound
propagation model in the block 21 can be determined in
a simple manner by forming the difference between the
most recently obtained sound incidence angle yk using
the target bearing (3zk which has once again been
improved, and the vertical sound incidence angle yk-1
which was obtained previously using the better target
bearing (3z(k-1) . If this difference (yg - 7k-1) is less
than a preset value S, then equality has been reached
and the most recently obtained, better target bearing
(3zk is output as the bearing of the target (3z. For this
purpose, by way of example, the previously obtained
vertical sound incidence angle yk_1 is passed via a
memory or a shift register 23 to a subtractor 24, and
CA 02684930 2009-10-29
WO 2008/138433 - 13 - PCT/EP2008/002640
the subsequently obtained sound incidence angle yk is
passed directly to the subtractor 24, and the
difference is compared with the preset value S in a
comparator 25. If this preset value S has been
undershot, then a gate 26 connected downstream from the
bearing apparatus 17 is opened, as a result of which
the most recently obtained, once again improved, target
bearing Rzk is passed to the display 27, where it is
indicated as the bearing of the target RZ.
The signal processing in a bearing apparatus which is
connected to a so-called PRS antenna comprising three
reception bases, which are arranged a long distance
apart from one another on an alignment line, with
electroacoustic transducers, is described by way of
example in US 4 910 719, which is referred to expressly
here. In this bearing apparatus as well, the
respectively set value of the speed of sound in water
is corrected as described above in the blocks 21, 22
and 28 and is input to the bearing apparatus as
cset(new). The described iterative process until the
bearing of the target RZ is achieved is also the same.
