Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
Cece ~21
: ,
B~CKGRCU~TD 0~ TE~ I~JE~TION
.. 1~ Field of the inventlon
~:! The present invention relates to surve~ s~stems
for determinin~ the shape of a fixed path and
. 5 particlllarly for determinin~ such shape l.~hen the
fixed path defines the profile of an ext.ended
surface. ~he invention is particularly, though
not excl.usively, suitable for use in an under-
ground mining installation, wherein coal is won
. , .
from a lon~wall coalface.
.~ Such an installation includes a coal mining
machine adapted to traverse to and fro along an
armoured face conve~or comprising a plurality of
SeGtionS or pans extending along the longwall
~: 15 coalface, a pluralit~ of self advancing roof
.~ supports arranged along the armoured face con-
. veyor on a side awa~ from the coalface and double
actin~ hydraulic rams correcting -the roof supports
: to the conve70r. As coal is won~ the coalface
equipmerlt i~ advanced further into the coal seam
towards remaining coal, and the mine roof over
; worked out parts of the coal seam behind the roof
supports is allowed to col1apse. Advance of the
conveyor 15 effected in a snake-like manner,
whereb~ after passage of the mining machine on a
traverse, the sections are pushed up to the newly
: exposed coalface b~ the double actin~, rams, After
the sections have been pushed to the coalface the
~ roof supports are one b~ one lowered pulled up
to the conve~or b~ the rams and t,hen reset to the
;~ -2-
~.
roof.
; During normal working of the coal mining
installation, it is desirable that the coalface
is maintained substantiall~ straight, since face
curvat~re tends to increase stresses upon and
wear rate of the mining e~nipment, thereby
causing a greater frequency of breakdown of the
mining equi~ment. ~uch breakdowns are costl~ in
terms of lost production. Additionally, since
the roof which collapses behind the roof support
` tends to fracture along straight lines, curvature
of the face tends to cause falling of the roof
in advance of intended points resulting in dis-
ruption of normal working and requiring manual
` 15 shoring up.
It is also desirable that the direction of
advance of the coalface further into the seam is
consistent (the direction usuall~ being normal to
the coalface) since otherwise the conveyor tends
to move in its entiret~ towards one end of the
face. ~uch movement of the c~nve~or necessitates
time consl~ing transfer of c~nveyor sections from
one end of the face to the other. ~he movement -
; of the conveyor becomes an even more serious
problem in an inclined sea~, where gravit~ tendsto enhance such move~ent. Moreover, in an inclined
seam, the coalface is frequently deliberately
angled with respect to a line substantially
orthogonal to the coal seam in order to reduce
the incline and the movement.
--3~
5;3
Maintenance of a particul.ar direction of
advance of the coal ace also ensures that the
total face length Le.between parallel erld tunnels
or 'gates' does not vary significantl.y so that no
changes in the number of conve.yor secti.ons o.r roof
supports is necessar~.
2. Description of the Prior Art
In the kind of mining installation described
it has been usual for an operator to advance the
conveyor towards the newl~ exposed coalface in
order to attempt to keep the coalface straight
and to keep the direction of advance of the coal-
face substantiall~ constant. However, as it is
difficult to determine relative positions of
conveyor sections throughout the conve~or length
which extends alon~ a coalface of t~picall~ two
` hundred metres, the operator has either to guess
~ when he feels an7 conveyor section is advanced
.; sufficiently or e].se advance the section as far
- 20 as possible towards the coalface. The mine roof
.` supports are then advanced. up to the conve~or.
Consequentl~, if lumps of coal or other n-on
uniformities prevent either the conve~or section
coming into abutment with the coalface or the full
advance of the roof support, then the coalface
becomes misaligned on the current or subsequent
.; traverses of the coal winning machine.
It is known for lasers to be used to measure
the direction of advance of the coalface and also
for lasers to be used to provids a straight line
-4_
4~;3
along which rock tunnelling maehines can steer.
SUMMARY OF THE INVENTION
` According to the principles of the present invention
a survey system is provided whereby the profile of the coalface
may be determined even where there may be no line of sight from
end to end of the coalface. A mobile receiver on the mining
machine is provided which receives optical signals from points
along the coalface and deduces the coalface profile from the
:~ signals together with a signal indicating the machine's position
the deduction being effected with processing circuitry. The
` detector periodically scans the sources and the processing cir-
euitry solves equations to determine their position.
In accordance with the present invention there is
provided a survey system comprising a mobile optical receiver
.- adapted to travel in a fixed path, a series of optical sources
~ adapted to be positioned in a line generally along said path,
sa~d
the receiver being responsive to light projected by ~ optical
sources visible to the reeeiver to determine the relative angul-
ar positions of said sourees, the system ineluding control cir-
cuitry for effecting such souree angle determinations periodieal-
ly, eaeh time in respeet of a predetermined minimum group of
said sourees and at sueh intervals in relation to the speed of
the mobile receiver that a plurality of angle determinations are
made on each group of sources as they become visible to the re-
ceiver, the system further ineluding processing circuitry adap-
ted to correlate the different angular measurements involving
the same sourees and to derive therefrom information identifying
the shape of said fixed path.
~RIBF DESCRIPTION OF THE DRAWINGS
Figure 1 is a diagrammatic plan view of a section
through the coalface showing the face profile in an exaggerated
form;
.' -/~
~ -5-
'~ 5~
Figure 2 shows a plan view of one en~ of khe coalface
with coal cutting machine and survey system in operation;
Figure 3 is a diagrammatic perspective view along the
coalface showing the pit props in position;
Figure 4(a) and (b) are plan and elevation of an
optical source/.receiver module;
Figure 5 is a block diagram of the signal processiny
circuitry of the survey systems.
Figure 6 shows a plan view of part of a
,,
;'
-5a-
~?~4~3
coalface installation including a second
embodiment of surve,v s~s~em; and
Figure 7 shows a block diagr?Am of a signal
source, detector and processing circuitry of the
embodiment of Figure 6.
DESCRIPTION OF THE PREFERRED EMBO~IMENTS
Referring to Figure, 1 there is shown a
coalface 1 extending across a coal seam 6 for,
t~pically, a few hundred metres. The coalface
is serviced b~ mine roadways 2 and 3 known as
the main gate and the return gate. It is ~ '
means of these roadwa~s that men and equipment
~, can be transported to and from the coalface and
won coal transported therefrom.
Where the coal seam is horizontal from side
to side the desired direction of advance is
parallel to the 'gates' 2 and 3. An ideal profile
for the coalface is therefore normal to this
direction and parallel to straight lines such as
4 and 5. In general it is satisfactor~ if the
c~alface profile lies for the whole of its length
between the parallel lines 41 5 spaced at about
one metre apart for a coalface length of, sa~
three hundred metres.
Reference now to Figure 2 shows a mining
' machine 10 which traverses to and fro along the
coalface with a cutting drum 11 winning coal
during the traverse to a depth into the coalface
of about sixt~ centimetrss, The mining machine
is electricall~ powered and runs on rail portions
. .
4~53
.
of a conveyor which comprises a plurality of
sections movable towards the coalface in a snake-
like manner as previously explained. The general
line of the conveyor, however1 follows tbe profile
of the coalface.
~ urning now also to Figure 3 retro-
reflectors 12 9 some of whic~ are a]so shown in
; Fig~re 2 can be seen positioned on vertical
sections (~e.legs) of roof supports 17 at fixed
intervals,e.g.every fifth supp~rt. Figure 3 also
shows alternative positîons for the reflectors
on horizontal arms or canopies 18 and on o-ther
oblique portions of the supports but these can
`~ provide difficulties and a loss of accuracy. In
any case however, retro-reflectors are placed in
identical positions on the selected props. An
optical source and detector unit 16, which co-
oper~tes with the retro-reflectors is mounted on
the machine. ~he unit 16 is also shown in Figure
4a and 4b and is described subsequentl~.
Spacing of the retro-reflectors is affected
by a number of fsctors. A sin~le set of obser-
vation involves simultaneous or substantially
simultaneous responses from several reflectors.
The gre~ter the range disparity between the
nearest and the farthest the more accurate are
the results. However, this must be balanced
against other factors~ ~he greater the beam
path the greater the likelihood of obstruction
~0 b~ variations in the face and by dust etc.,in
':'
~.''
--7--
s~
the air. In addition~ the greater the range
of the optical system the greater has ~he power
of the source to be. Safet~ requirements put
severe restrictions on volta~es than can be used
so that power requirements of the equipment
generally are of major importance.
In the light of the above factors a
reflector spacing of five metres is found sat-
isfactor~, giving a maximum reflector range of
about 20 metres, which does of course impl~ an
optical path le~gth of 40 met~es.
The reflectors themselves are of the single-
corner cube type of prism form giving internal
reflections from externall~ mirrored surfaces.
It will be explained subsequently that the optical
source and detector are adjacent but not co-
incident so that the apex area of the prism, which
merel~ returns a beam directl~ to the source, is
not emploved, The prism can therefore be flattened
at its apex to re~uce its dimensions.
The size of the prism is dictated b~ the
size of the detector window and the spacing
between the centres oi the source and detector
lenses, It ma~ be seen that the reflected patch
of ligh~, ignorin~ blurrin~ is twice the size
of the reflector, irrespective of range. ~he
reflected patch of light has to encompass the
detector window between the centre and the
peripher~ of the patch so that the reflector must
be of comparable size to the detector window,
53
--
which, as will be seen, is in effect the
objective lens of the receiver optical system.
There are various alternative reflectors
which may possibly be used but which are, in
general, markedly inferior to the single-corner
cube prism reflector. The nearest is perhaps the
corner cube array which has an advantage in
reduced depth but it still provides a light patch
of only twice the si~e of one corner element
(acting, in fact, like a concave mirror with the
source near the centre of curvature) and produces
a net reflected beam much weaker than the single-
corne~ prism. Other alternatives include 'cats-
eyes' as used on roads, and light-scattering
surfaces of the kind employing a multitude of
; very small glass beads adhering to a sticky tape.
In addition to the reflector angle information
that is derived, a distance transducer (shown in
Figures 5 and 7) is mounted on the cutting
machine to indicate the distance travelled by the
receiver. This distance is, of course, measured
along the conveyor profile. The distance transducer
typically comprises pulse counter odometer means
which senses movement of a mechanical component
d~ on the machine and derives electrical pulses in
proportion to said movement.
Referring again to Figure 2 in addition to
the series of retro reflectors 12, mounted on the
props, a single reflector 13 is positioned at the
9 _
,"~,,.";p.
,~ ,
extreme end of the coalface and directed also
towards the receiver mounted on the cutting
machine 10~ This reflector 13 is on the optical
axis of -the receiver when the minin~ m~chine is
correctl~ positioned both in distance from the
coalface and in heading across it. I-t ma~ be, of
course 9 that the desired heading is not normal to
the side gates but is slewed from this direction
to counteract the effect of a non-horizontal seam,
~,to bias the advancing direction of t'ne roof
supports slightl~ up hill. In an~ event, when
the mining ~achine is correctly located, the
~ angle of the beam reflected from the reflector
13, with respect to the receiver optical axis,
is zero~ ~his facilit~ provides a direct
indication of the mining machine heading and
: also facilitates initial set up, but it suffers
from the basic disadvantages of a long beam path
: system in that gross errors ma~ cause interruption
of the line of sight and the effects of dust are
accentuated.
.~ ~he reflector 1~, is a retro-reflector in the
horizontal plane onl,y, consisting of a long
triangular section prism the axis of which is
, 25 vertical. A beam will therefore be returned.
irrespective of mining machine heading and the
height of tne reflector will take account of
- vertical undulatlons of t,he mine floor in the
vicinity of the coalface.
Referring now to Figure 4(a) and (b), there
-10-
..
i3
is shown the layout of the source/receiver unit
16 mounted on the mining machine. Figure 4(b)
shows the module as seen from the rear of the
mining machine. The optical source 21 is mounted
closely beneath the receiver objective lens 22,
The source 21 is a gallium-arsenide ]ight-
emittin~-diode (LED) which can therefore be
operated at low voltage and is robust and 9 Ui ted
to the envirQnment. The power limitations of
; 10 this source are overcome to a large extent b~
pulsing it at a low duty ratio. It is arranged
; to have a beam spread sufficient to embrace a
minimum group of reflectors 12 within a range of
twent~ metres and ma~ typically provide a 25
horizontal fan beam by means of a lens now shown.
~he number of reflectors in a group is determined
by various factors as will be explained. In an
alternative arrangement two source beams are
; produced, one o~ fairly wide angle for the near
reflectors of the group, and one narrow angle for
the more distant ones of the group. The narrow
angle then compensates for the long path.
A significant advantage of the LED as a
source is that its output is in the near infra-
red region at about 0.9~ . The ~pectralefficiency of silicon photo-diodes used in the
detector at this wavelength is improved sub-
stantially as compared with the output of a
tungsten lamp (which although of high power
has other disadvantages in addition). In addition
.
,:
:
.
53
the detector can use a filter to detect this
., ;
radiation amongst substantial visible and other
'noise' radiation in the environment. ~he latter
may be caused by deliberate illumination of the
coalface, ~iners' lamps, reflections off odd
surfaces etc.
Referring now to ~igure 4(a), a detector
arra~ 23 is placed approximatel~ a' the focal
length of the ob~ective lens 22. A cylindrical
lens 24 is positioned, with its axis horizontal,
between the lens 22 and the detector array 23~
to produce a vertical spread of the field of view
and so ensure that any vertical und~lations in
the positions of the retro-reflectors 12, or
elsewhere, do not cause the image to miss the
detector array 23. The resulting vertical
extent of the image is several times the horizontal
extent.
The detector array 23 consists of a line of
~ 20 256 photo-diode elements arranged, in a basic
i situation, horizontàll~ through the optical axis
of the receiver. The image of a retro-reflector
12 projected on to the array will illuminate one
or two photo-diode elements at a lateral position
;~ 5 corres?onding to the angle of the retro-reflector
off the optical a~is. In the situation described,
; the reflectors 12 are above the height of the
receiver and consequently the image of a line of
reflectors is inclined. The array 23 is therefore
inclined correspondingly,
., .
-12-
, ,,
...
Several images of retro-reflectors will be present simultaneously
and these will sweep slowly over the array as the coal cutting machine trav~
erses the coalface.
Figure 5 shows the basic arrangement for identifying the image posi-
tions on the array and computing the output values accordingly. The diode
array 23 is coupled to a shift register 26. A '1' state is shifted through
the register 26 by a clock pulse train, producing an interrogating pulse for
; each diode element of the array in passing. Output current pulses are thus
produced serially from those diodes illuminated. The current pulses are con-
verted to video output pulses for further conversion to digital format in
known manner. The data processing is then performed in two stages by process-
ing circuitry 27, comprising firstly a pre-processor 28 which extracts merely
the relative angular positions of the nearest group of retro-reflectors and
secondly a further microprocessor 29, which performs the remainder of the
;~ computation. The further microprocessor 29 also receives a distance signal
from the previously mentioned distance transducer which is now shown at 30.
. A display 31 i.s connected to the output of the further microprocessor. The
pre-processor 2~ may be a microprocessor of the type sold under the trade
mark Intersil/Harris 6100. The microprocessor 29 may be of the type sold
2~ by Marconi Space and Defence Systems Limited under the trade mark MC 1~00.
The process of computing the face profile has to take account of a
number of indeterminate factors. Thus, the reflectors are assumed to be
:.
- 13 -
randoml~ offset fro~. tneir nominal position. The
line of the reflec~ors is not therefore a true
co?v of the face profil.e. The heading of the
cutting machine i.e.the direction of the optic~l
axis is i~ genera~. not known in view of local
variations in the convevor path. In the worst
case the d.istance tr~velled alon~ the fixed
path of the conve~or b.y the cutting machine and
the receiver is not known. The principle of the
computatlon is therefore to measure the angular
:~: relationship of a group of the reflectors G
number of times ideally a ~reat number of times
throughout the period in which they are visible
to the receiver and by correlating the sets of
- 15 values for the same group of reflectors taken at
;~ different positions of the receiver the relative
positions of the reflectors of this group can be
determined in terms of x and. y co-ordinates of a
pre-determined reference. Each set of angle
~ 20 values for a group in effect provides an eauation
relating the positions of the reflectors of that
group to each other and to the receiver position .
at the instant of taking the angle values. A
sufficient number of sets of values will therefore
.~
enable the equations to be solved for x and y. A
surplus of such sets of values can be assessed on
a statistical basis to provide the best values for
x and y As the receiver travels along the face
one reflector will pass from view and another
will come into view. ~ach reflector will there-
.,
:
-14-
i:
;~'
,, .
fore feature in a number of successive ~,roups
and multip]e angle determinations will be made
in respect of each group that a reflector fe~tures
in. ~he statistic`al data are thus increased
accordingl~. The mathematical processes involved
in the extraction of x and y in a situation of
this kind are well known an~ emplo~ the Kalman
filtering process. ~his operates by iterative
adjustment of a mathematical model of the reflector
dispositions to minimise the difference between
the observed angular position of a reflector as
presented by the microprocessor and an estimated
value of that angular position, with respect to
the variables: y displacement of detector arra~
from x axis; x position of each particular
reflector; y position of each particular reflector;
distance in x moved b~ the detector array; and
the skew angle of the receiver optical axis to
the x axis.
On a first run the mathem~tical model relies
on data fed in : nominal interval between reflectors;
~umber--of reflectors; mean distance between
receiver and line of reflectors etc. However,
after the first run, each estimated profile and
reflector disposition can be used for the next
run, since each profile will be fairly closel~
related to the next.
During a run, for each set of observations,
:...
or 'scan', the estimated values of the reflector
,, .
angle off the optical axis are calculated for
-15-
.,~
,
.':
~l~2~ 3~3
each reflector, using the previously estimated
reflector positions and the assumed new position
of the mining machine and receiver. The differences
between these estimated values and ~he currently
observed angles are used to adjust the estimated
machine position and the reflector positions so as
to minimise the mean square error between estimated
and measured angles over all scans.
A convenient reference line at least for the
first run is taken to be the line through the
first two reflectors, this then bein~ the x axis
referred to above.
There is a minimum number of reflectors that
can feature in a ~roup,i.~.of simultaneous angle
determinations~ for the extraction of the co-
ordinates of the reflectors and the receiver to be
possible. If there are absolutel~ no other ref-
erences than the angular displacements of the
reflectors in a group, then four reflectors is a
minimum group. If, however, a further reference
is supplied, in the form of the distance travelled
by the receiver, for example, or the range of each
reflector, then the minimum number can be reduced.
~he greater the number in a group however, the
greater the amount of statistical information
available and the more accurate the results.
Since the positions of the reflectors
transverse to the coal face cannot be relied
upon,~for reasons explained above~ the final
~0 layout of these reflectors is of secondary
-16~
.
,4~i;3
importance compared to the path of the receiver~
which, as mentioned, moves in and out transversely
in synchronism with the coalface. The reflectors
therefore provide a set of temDorary reference
points for the location of this fixed path, the
lack of prior knowledge of the positions of these
reference points being compensated by the multi-
plicity of angle determinations taken.
~he micro processor determining the
observed reflector angles is mounted on the
cutting machine. ~he main processor, which
analyses the observations and calculates the
estimated dispositions, is mounted, together
with the display 31 ln the gate.
~he output is then displayed at the machine
in the following form
(a) A small number (e.g.4) indicator lamps
indicating such things as: c~mputer
running; self-check O~Ko; input data
validity; power supply levels correctO
(b) A digital (LED) display showing, on
demand, any of the data words normally
output on the data link. Thus for
example, with the machine stationary,
the angles of the nearest retro-
reflectors can be displayed in turn
and checked against ~easurements
taken by surveying methods.
(c) An analog display of face profile.
~his is of most interest to the miners
.,
~ 17-
, '
5;3
working at the face. A row of (parallel)
linear LED displa,ys gives a long rect-
angular dot matrix on wbich the current
or previous profile is graphically
displayed.
~he term "optical" as used in this
specification applies to any visible or invisible
radiation having optical properties and thus
includes, for example~ infra red and ultra-violet
light.
Various modifications of the system are
possible. Thus the reflectors fitted to the roof
~ supports may be replaced by lamps or other sources
; so obviatin~ the primary source on the receiver
module. The reflector arr~ngement does however
have the advantage that the active devices are
all kept in one place rather than distributed
along -the coalface with the attendant disadvantages
of suppl~ cables.
In Figures ~ and 7 which should now be
referred to the same reference numerals are used
where appropriate. ~he previously mentioned
. .
armoured face conveyor is now shown at 32 in
~ Flgure 6. A primary source 33 emits a laser beam
,~ 25 of fan shape extended in the vertical direction
and swept horizontally, broadside as it were, so
; as to intercept the reflectors sequentially. I~
Figure 6 for example, the laser beam intercepts
~ retro-reflector 12a~ then 12b and then 12c. ~he
`- 30 vertical extension ensures that undulations from
-18-
, ~
:"
the horizon-tal do not prevent the reflectors
from seeing the laser beam. A receiver s`~own at
34 is a localised wide angle optical detector,
instead of array as described in the previous
embodiment, the angular displacement of each
reflector from the last being indicated by the
time interval between reception of the reflected
beams. The processing circuitry 27 is then
adapt;ed to measure these time intervals by
utilisation of a modified pre-processor 3~.
Instead of a linear array of optical
detectors~ a rectangular array may be used to
give an equivalent vertical sp~ead to the field
of view in the absence of the cy]indrical lens.
Charge cnupled devices can also be used as
` combined optical detectors and analog shift
registers in place of the standard optical
elements proposed.
The profile display may be provided by an
electroluminescent panel or liquid crystal panel.
In an alternative mounting arr ngement for
the reflectors they could be positioned on a
spill plate portion of the conveyor, so that,
being fixed with respect to the conveyor the~
.~ 25 define the face profile exactly. Although this
would simplify the calculations by the removal
of a variable, it would mean that the reflectors
were low down and vulnerable to damage of dis-
; placement.
Z0 It will be clear t;hat, although tha syste~
;,
'
-19-
;;3
described is particularly suited to a mining
operation, the inver.tion is also applic~ble to
the survey of any extended surface or path to
which a mobile receiver can be constrained,
whether above or be]ow ground.
. .
.,
. ~ .
fii
~, .
'':''~
~,~
.,
'~,'' .
,, .
,,
-20-
,,
