Note: Descriptions are shown in the official language in which they were submitted.
CA 02222124 1997-11-2
DESCRIPTION
DEVICE FOR DETECTING MOVING BODY DEVIATING FROM COURSE
TECHNICAL FIELD
The present invention relates to a device for detecting
that a moving body, such as an automated vehicle, is deviating
from a predetermined course, and more particularly to an
arrangement of marks for detecting deviation and sensors for
detecting the marks.
BACKGROUND ART
Up to now, the method for guiding an automated vehicle
along a predetermined traveling path to a destination was
called dead reckoning of estimating the current position of
the vehicle with a direction detector and a device for
measuring length of travel, and automatically steering the
vehicle through predetermined passing points on the
predetermined path, instructed in advance. For example, the
method to express the predetermined traveling path with a
series of coordinate points, in the case of effecting steering
by dead reckoning, was disclosed in Japanese Patent
Application No. 60-120275 and is already known.
The disadvantage of dead reckoning is that slippage of
the vehicle and irregularities of the road surface result in
errors of the estimated position of the vehicle and the
vehicle cannot correctly pass through the predetermined
passing points.
CA 02222124 1997-11-2~
Here, in the case of traveling outdoors, it is possible
to correct the shift position intermittently with GPS (global
positioning system) or radiolocation, etc. This can resolve
the problem of being unable to correctly pass through the
predetermined passing point.
However, the disadvantage of GPS, etc., is that these
systems cannot be used outdoors or underground. Consequently,
there is need for a system which can make it possible for a
vehicle to correctly pass through the predetermined passing
points even indoors.
Therefore, with the object of resolving such issues at
low cost, the inventors have invented various methods, and the
applicant has applied for various patents, for methods wherein
guidance marks (referred to as "marks" below), having line
segments in a predetermined geometrical form, are established
intermittently on the predetermined traveling path and provide
means to correctly guide a vehicle along the predetermined
traveling path.
For example, in Japanese Patent Application No. 59-
213991, three line segments comprising metal plates, etc., are
placed in the form of a Z and a plurality of these marks is
established intermittently so that all of the line segments of
these Z-shaped marks lie across a predetermined traveling
path; the three line segments of the marks are detected in
sequence by sensors placed on an automated vehicle during
travel; the course deviation of the automated vehicle (the
deviation of the actual mark passing position from the
predetermined mark passing position) is found by a calculation
CA 02222124 1997-11-2~
based on the detection timing and the geometrical relationship
of the Z; and the estimated position of the automated vehicle
is intermittently calibrated based on the course deviation
found.
However, in this method where marks are arranged along a
predetermined traveling path, there is no margin for selecting
a traveling route and the automated vehicle can only travel
along a single traveling route.
Therefore, in Japanese Patent Application No. 60-213916,
in order to operate the automated vehicle on one traveling
route selected from among a plurality of traveling routes, as
shown in Figure 15, a plurality of Z-shaped marks are arranged
in a state where the marks adjacent to one another vertically
and horizontally are inclined at the same angle and the marks
adjacent to one another in oblique directions are rotated by
90~, so as to be able to select one traveling route from among
a plurality of predetermined traveling paths.
The way the marks are arranged, shown in Figure 15, it is
possible to establish a plurality of predetermined traveling
paths inside the lot wherein the marks are arranged, but a
large number of predetermined traveling paths cannot be
established.
For example, the path A' in Figure 15 passes through the
marks, but the estimated position cannot be corrected using
the marks through which the path passes because the path A'
does not cross all line segments of the marks. Also, path B'
and path C' do not cross all of the line segments of the
marks, and so the estimated position cannot be corrected in
CA 02222124 1997-11-2
these cases either.
Therefore, with the conventional method, it is clear that
for a large number of predetermined traveling paths, the
estimated position cannot be corrected using marks.
Specifically, the prior art cannot respond with flexibility to
changes in the traveling route, because it is not possible to
establish arbitrary predetermined traveling paths within the
predetermined traveling area.
DISCLOSURE OF THE INVENTION
The present invention was made in view of the foregoing
situation; it is an object of the present invention to make
possible the establishment of arbitrary predetermined
traveling paths throughout an entire traveling area and make
possible a flexible response to changes in the traveling
route.
This object is attained as follows.
Specifically, in a first invention of the present
invention there is provided a device for detecting a moving
body deviating from a course, in which predetermined passing
marks through which a moving body passes are arranged
intermittently on a predetermined traveling path, so that all
line segments constituting marks comprising at least a first
and a second line segments which are not parallel to each
other, cross the predetermined traveling path; an actual
position on the predetermined passing marks is detected by
detecting, by means of a line segment crossing detector
mounted on the moving body, the crossing of all line segments
CA 02222124 1997-11-2~
constituting the predetermined passing marks during the
traveling of the moving body; and a course deviation from the
predetermined traveling path of the moving body is detected
based on a predetermined passing position on the predetermined
passing marks and the actual position on the predetermined
passing marks which was detected, characterized in that:
a plurality of marks are respectively inclined and
arranged in advance within a predetermined traveling area,
so that, in the case where an arbitrary predetermined
traveling path in the predetermined traveling area, through
which the moving body must travel, is selected and the moving
body is caused to travel along the arbitrary predetermined
traveling path, the line segment crossing detector mounted on
the moving body can detect the crossing of all line segments
constituting the predetermined passing marks corresponding to
the arbitrary predetermined traveling path.
With the constitution of the aforementioned first
invention, marks are arranged in advance as shown in Figure 7,
within a predetermined traveling area 20 through which the
moving body must travel, in a state where the marks adjacent
to one another vertically and horizontally are rotated by 90~
and the marks adjacent to one another in oblique directions
are inclined at the same angle, for example. For this reason,
when the arbitrary predetermined traveling paths A and B
within the predetermined traveling area 20 are selected and
the moving body travels along the pertinent arbitrary
predetermined traveling paths A and B, the crossing of all
line segments constituting the predetermined passing marks
CA 02222124 1997-11-2~
corresponding to the pertinent, arbitrary predetermined
traveling paths A and B is detected with the line segment
crossing detectors placed on the moving body. In this way,
course deviation can be detected for an arbitrary
predetermined traveling path.
Also, in a second invention of the present invention, a
distance between two or more line segment crossing detectors
is established so that at least one line segment crossing
detector of the two or more line segment crossing detectors
can detect the crossing of all line segments constituting the
predetermined passing marks.
With the constitution of the aforementioned second
invention, the course deviation can be detected with certainty
for an arbitrary predetermined traveling path, because the
crossing of all line segments constituting the predetermined
passing marks is detected by at least one line segment
crossing detector (paths A and B), even if the crossing of all
line segments constituting the predetermined passing marks is
not detected by a number of line segment crossing detectors
from among two or more line segment crossing detectors (paths
C and D).
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a view from above of the traveling area of an
embodiment;
Figure 2 is a block diagram to show the constitution of
the control system of an embodiment;
Figure 3 is a diagram to show the constitution of a side
CA 02222124 1997-11-2
wheel drive vehicle of an embodiment;
Figure 4 is a diagram to show the constitution of a front
wheel steered vehicle of an embodiment;
Figure 5 is a diagram to show the locus of a vehicle of
an embodiment;
Figure 6 is an inclined view to show the constitution of
the mark body of an embodiment;
Figure 7 is a view from above to show the locus through
which a sensor passes inside the traveling area;
Figure 8 is a graph to show the relationship between the
interval between a main sensor and auxiliary sensor, the
detection interval, and the error count;
Figure 9 is a view from above to show loci of sensors
when a vehicle is traveling at an angle through a traveling
area;
Figure 10 is a view from above to show loci of sensors
when a vehicle is traveling at an angle through a traveling
area;
Figure 11 is a flow chart to show the processing
procedure through the approach of a vehicle to a traveling
area shown in Figure 1;
Figure 12 is a view from above to show the situation
where the cumulative error, due to the dead reckoning method,
increases along with the progress of the vehicle;
Figures 13(a)-13(e) are figures illustrating geometrical
forms of the line segments constituting the mark body;
Figures 14(a)-14(c) are figures illustrating geometrical
forms of the line segments constituting the mark body; and
CA 02222124 1997-11-2~
Figure 15 is a view from above to show the conventional
arrangement of mark bodies.
BEST MODE FOR CARRYING OUT THE INVENTION
Below, the mode for carrying out the device for detecting
moving body deviating from course, relating to the present
invention, is explained with reference to the figures.
Figure 1 is a view from above of the predetermined
traveling area 20 through which an automated vehicle 1
(referred to as "vehicle" below), which is a moving body, is
to travel, this traveling area 20 is covered with mark bodies
24, discussed below.
The mark bodies 24 (referred to below, appropriately, as
"marks") are laid out in such a manner that the marks adjacent
to one another vertically and horizontally are rotated by 90~
and the marks adjacent to one another in oblique directions
are inclined at the same angle.
On the near side of the traveling area 20, the marks 40
and 41 are arranged in such a manner that a line extending
from the reference axis of measurement 42 (in effect, the
center line of the marks; the axis crossing all line segments
constituting the marks at right angles) of marks 40 and 41,
having the same constitution as marks 24, crosses the
traveling area 20. Furthermore, the standby position 19 of
the vehicle 1 is located on the near side of these marks 40
and 41. The vehicle 1 is halted facing the already
aforementioned reference axis of measurement 42.
The vehicle 1 comprises a direction detector and a device
CA 02222124 1997-11-2~
for measuring length of travel, calculates its current
position using these detectors, and is automatically steered
to travel on the predetermined traveling path. The technology
for this automatic steering is the known technology relating
to a previous application (Japanese Patent Application No. 60-
120275) made by the applicant. A detailed explanation is not
included since this technology is not directly related to the
gist of the present invention.
Figure 2 is a diagram of the entire constitution of the
control system of the vehicle 1.
Here, the vehicle 1 may be a side wheel drive vehicle as
shown in Figure 3 or the front wheel steered vehicle as shown
in Figure 4. This embodiment is explained presupposing the
use of the side wheel drive vehicle as shown in Figure 3.
This side wheel drive vehicle 1 comprises a cylindrical
car body 15, for example, as shown in Figure 3, on the side of
that car body 15, side drive wheels 2 and 3 are placed on the
same axis as the diameter of the circle. Also, castors 16F
and 16R are placed on the front and rear of the car body 15.
A main sensor 5, to detect the aforementioned mark bodies 24,
40, and 41 as discussed below, is mounted on the lower surface
at the center point of the line segment connecting wheels 2
and 3, being the lower surface of the car body 15. Likewise,
an auxiliary sensor 6 is placed on the line segment connecting
the wheels 2 and 3, on the lower surface of the car body 15,
as being separated from the main sensor 5 just by a distance
d, discussed below.
As shown in Figure 2, encoders 2E and 3E are mounted on
CA 02222124 1997-11-2~
the side wheels 2 and 3 of the vehicle 1; the encoders 2E and
3E measure the length of travel of the vehicle 1 by detecting
the rotary position of the wheels 2 and 3. Specifically, the
encoders 2E and 3E output pulse signals showing the length of
travel; the counter 7 takes up the pulse signals as data for
each side separately; and the count results of the counter 7
are output to both the mark detection section 8 and the dead
recko~; ng calculation section 9.
Signals output from the main sensor 5 and auxiliary
sensor 6 which show the crossing of line segments 23a, 23b,
and 23c of the marks 24 (See Figure 6) are input to the mark
detection section 8; meanwhile, the counter measurement value,
showing the length of travel and which is output from the
encoders 2E and 3E, is input to the mark detection section 8.
As a result, the travel distance, from the crossing of a line
segment of the mark 24 (for example, line segment 23a) to the
crossing of the next line segment 24b, and the travel
distance, from the crossing of the line segment 24b to the
crossing of the next line segment (for example 23c), are
attained in the mark detection section 8. The difference
between the actual passing position on the mark 24 and the
central position (predetermined passing position) of the mark
24, in effect the course deviation for the predetermined
traveling path, is detected based on these two travel
distances and the geometrical form (Z) of the line segments of
the mark 24.
Moreover, the calculation procedure applied can be that
which was explained in detail in Japanese Patent Application
CA 02222124 1997-11-2~
No. 60-287439. In this way, the actual passing position in
each mark 24 is output from the mark detection section 8 in
sequence for each passage through the marks 24.
In the dead reckoning calculation section 9, the serial
current position of the vehicle 1 is estimated and calculated
based on the revolutions of the side wheels 2 and 3 output
from the counter 7; meanwhile the sequential current direction
of the vehicle 1 is estimated and calculated based on the
difference between the revolutions of the side wheels 2 and 3
output from the counter 7. The estimated and calculated
values of these serial current positions and current
directions of the vehicle 1 are output to the position
calibration section 10.
In the position calibration section 10, the mark 24, for
the correction of the currently estimated position, is
selected based on these serial current positions and current
directions input by the dead reckoning calculation section 9
and the stored contents of the mark arrangement information
storage section 11. Then the angle at which to enter the
traveling area with respect to the reference axis of
measurement of the selected mark 24 is found from the
information of the current direction; meanwhile, it is
determined whether the correction of the estimated position,
using the selected mark 24, can be trusted. As a result, in
the case where it is judged that the correction of the
estimated position using the selected mark 24 can be trusted,
the position correction signal, which will correct the current
estimated position, is output to the dead reckoning
CA 02222124 1997-11-2~
calculation section 9 according to the actual position on the
mark 24 output from the mark detection section 8, at the time
when the crossing of the last line segment to be crossed in
the mark 24 is detected.
Two past position correction signals output from the
position calibration section 10 are stored in the angle
correction section 12; the corrected position, corrected
according to those two past position correction signals, and
the estimated position immediately previous to the reception
of the two past position correction signals are stored in the
angle correction section 12.
Specifically, an error in the direction traveled by the
vehicle 1 therebetween can be confirmed by comparing the
change in two coordinate positions which were properly
corrected and the change in two estimated coordinate
positions. Then, the angle correction section 12 corrects the
serial estimated direction of the vehicle 1 by outputting the
directional error attained in this way to the dead reckoning
calculation section 9.
In this way, in the dead reckoning calculation section 9,
the current estimated position is corrected by the actual
position on the mark 24 based on the position correction
signals which were input, while the current estimated
direction is corrected based on the directional error which
was input.
The corrected position and corrected direction output
from the dead reckoning calculation section 9 are output to
the steering controller section 13. In the steering
12
CA 02222124 1997-11-2~
controller section 13, steering control is effected based on a
known point tracking technology. This point tracking
technology is a steering control method which changes the
target site on the predetermined traveling path sequentially
as the vehicle 1 advances; steering control is effected as
follows: the coordinate position data of the target site is
read in sequence from the predetermined traveling sequence of
points data storage section 14, wherein coordinate position
data of each site on the predetermined traveling path to be
traveled is stored; and orders for reaching this coordinate
position of the target site, in effect the direction of
progress (target steering angle) and the speed of progress
(target speed), are output to the motor controller 4.
Moreover, if the vehicle 1 does not operate at a constant
speed and is supposed to vary the operating speed according to
the location, it is possible to have and to store a
correspondence to information relating to the speed of the
vehicle 1 in the case of aiming at a target site, for each
target site in the predetermined traveling sequence of points
data storage section 14.
The motor controller 4 outputs the revolution speed order
to the motors 2M and 3M which independently drive the left and
right wheels 2 and 3 respectively. Here, in the independent
driving vehicle, the average of the rotational speeds of the
side motors 2M and 3M determines the speed of the vehicle l;
and the difference between the rotational speeds of the side
motors 2M and 3M determines the steering angle. Then, based
on this type of relationship, the motor controller 4 outputs
13
CA 02222124 1997-11-2~
orders for rotational speed corresponding to the target speed
and target steering angle of the vehicle 1 to the side motors
2M and 3M.
Next, the structure of the mark body 24 constituting the
travel area 20 is explained below.
Figure 6 is an inclined view to show the arrangement of
the mark body 24. A mark pattern 22 is mounted on the upper
surface of the base plate 21, so that the center 21a of the
square base plate 21, with the length of one side being a,
agrees with the center 22a of the mark pattern 22. The
lateral width of this mark pattern 22 is a; the depth
corresponding to the direction of progress of the vehicle 1
has a dimension of a or less; such was already disclosed in
Japanese Patent Application 59-213991.
When the mark pattern 22 is mounted on the base plate 21,
the upper edge 21b of the square base plate 21 is made
parallel to the line segments 23a and 23c (two parallel line
segments of the Z) of the mark pattern 22. Moreover, in the
case of preparing the mark body 24, the line segments 23a-
23c, which constitute the Z, may also be applied directly on
the base plate 21, without the plate-shaped mark pattern 22
being mounted on the base plate 21. Moreover, line segments
23a-23c are constituted of materials (for example, metal if
the sensors 5 and 6 are metal detectors) which can enable the
detection of the pertinent line segments by sensors 5 and 60
As shown in Figure 7, the mark bodies 24 are arranged on
the traveling area 20 in a specific pattern with the mark
bodies 24 having two orientations.
14
CA 02222124 1997-11-2~
Measurement of the actual position on the mark body 24 is
possible if sensors 5 and 6 can have loci (below referred to
as "sensor loci") so as to cross the three line segments 23a-
23c of the mark body 24. This aspect is already disclosed in
Japanese Patent Application No. 59-213991. As already
discussed for Figure 1, the reference axis of measurement is
defined as the axis which becomes the reference of measurement
for sensor loci (of course, this also crosses line segment
23b) passing through the center point of the mark body 24 and
crossing two line segments 23a and 23c at right angles.
Here, the mark body 24-1 within the traveling area 20 is
"placed normal" so that the reference axis of measurement 25
is vertically oriented. In this case, the measurement of the
mark body 24 with the sensors 5 and 6 becomes possible if
sensor loci are within at least a 45~ horizontal range with
respect to the reference axis of measurement 25. The way of
arranging these two line segments 23a and 23c in a vertical
position is called the "Z" arrangement. Meanwhilè, the mark
body 24-2 positioned above the mark body 24-1 is the Z
arrangement rotated 90~; the reference axis of measurement 26
is oriented in a sideways direction. This arrangement with a
90~ rotation is called an "N" arrangement.
Ultimately, as shown in Figure 7, the marks 24 within the
traveling area 20 adjacent to a Z type mark 24 vertically and
horizontally are all N type and the marks 24 adjacent to a Z
type mark 24 in oblique directions are all the Z type.
Likewise, the marks 24 adjacent to an N type mark 24 in the
traveling area 20 vertically and horizontally are all Z type
CA 02222124 1997-11-2~
and the marks 24 adjacent to an N type mark 24 in oblique
directions are all the N type.
The marks 24 cover such a traveling area 20 in a uniform
pattern wherein the marks 24 adjacent vertically and
horizontally are rotated 90~ and the marks 24 adjacent in
oblique directions are inclined at the same angle. Below,
this pattern of arranging the marks 24 is called
"interleaving."
In Figure 7, in the case where the sensor loci have the
horizontal orientation called A, the actual position on the
marks 24 can be measured at every other mark because the marks
24 in the N-type arrangement are arranged at every other
position in a horizontal direction. Moreover, in the case
where sensor loci pass vertically, measurement at every other
position becomes possible in the same way due to the marks 23
in the Z-type arrangement placed at every other position in a
vertical direction.
Also, in the case of the 45~ inclined sensor locus called
B, the actual position on the marks 24 can be measured
continuously because the marks 24 in the Z-type arrangement
are arranged consecutively in the direction of a 45~ angle.
Moreover, in the case where sensor loci pass over the marks 24
in the N-type arrangement, arranged consecutively in an angled
direction, continuous measurement becomes possible in the same
way due to these marks 24 in the N-type arrangement.
As above, the predetermined traveling path of the vehicle
1 may be arbitrary paths, including paths A', B', and C' (See
Figure 15) which could not be used before now.
16
CA 02222124 1997-11-2~
However, measurement becomes impossible if the sensor
locus A agrees with the reference axis of measurement of the
marks 24 in the N-type arrangement and does not cross all
three line segments 23a-23c of the marks 24.
Consequently, in the case of sensor loci like the sensor
locus C in Figure 7 which deviates widely from the reference
axis of measurement, but is parallel to the sensor locus A,
and passes through the seam between marks 24 located above and
below, sensors 5 and 6 can only detect the end portions of the
line segments 23a-23c because the sensor locus grazes the
upper and lower ends of the N of the marks 24. Therefore, the
signal output by the sensors 5 and 6 becomes weak and
measurement becomes impossible. Moreover, measurement becsmes
impossible in the same way, in the case where sensor loci pass
along the seam between the marks 24 to the right and left.
Also, measurement bPCom~ impossible if, as the sensor
loci B, the sensor loci form a 45~ angle (45~ or less) with
the reference axis of measurement of the marks 24 in the Z-
type arrangement, pass through the centers of the marks 24,
and do not cross all three line segments 23a-23c of the marks
24.
Consequently, in the case of sensor loci like the sensor
loci D which deviate widely from the center points of the
marks 24, but are parallel to the sensor loci B, the sensor
loci cannot cross all three of the line segments 23a-23c of
the marks 24 (can only cross one or two of the line segments)
and measurement becomes impossible.
Therefore, it is necessary that measurement definitely be
CA 02222124 1997-11-2~
possible with one sensor (sensor loci A, B), even though
measurement with the other sensor, among the two sensors 5 and
6, becomes impossible (sensor loci C, D). In this case, the
establishment of the interval d between the two sensors 5 and
6 becomes important.
The turning center of the side wheel drive vehicle 1 of
the embodiment is on the axle connecting the left and right
driving wheels 2 and 3 as shown in Figure 3. Moreover, the
turning center is external to the car body in the case of the
vehicle turning and describing a large turning radius.
Therefore, in the case where the two sensors 5 and 6 are
established on the axle connecting the left and right driving
wheels 2 and 3 and separated just by the distance d, these two
sensors 5 and 6 are separated just by distance d usually in a
direction at right angles to the direction of progress of the
vehicle 1 (See Figure 5).
Also, in the case of using the front wheel steered
vehicle shown in Figure 4 as the vehicle 1, the two sensors 5
and 6 are established on the axle connecting the side rear
wheels 17L and 17R which do not steer and are separated just
by the distance d. In this case as well, the two sensors 5
and 6 are separated just by distance d usually in a direction
at right angles to the direction of progress of the vehicle 1
(See Figure 5) because, when the vehicle 1 progress while it
is steered, it is turned with the center being the turning
center present on the axle connecting the side rear wheels 17L
and 17R which do not steer.
Figure 8 is a graph showing the results of a calculation
CA 02222124 1997-11-2~
of the capacity for the main sensor 5 and auxiliary sensor 6
to detect the mark body 24. The horizontal axis of the graph
is the interval d between the main sensor 5 and the auxiliary
sensor 6; the length of one side of the square mark body 24 is
100%.
The left vertical axis of the graph is the interval
(travel distance) wherein the main sensor 5 or auxiliary
sensor 6 detects the mark body 24 concurrent with the progress
of the vehicle 1; a maximum distance, a minimum distance, an
average distance, and a standard deviation in a range of less
than 1000% (in effect, the length corresponding to ten times
the length of one side of the mark 24; the length of one side
of the square mark body is 100%) are shown.
The right vertical axis of the graph shows a number (this
is called the error count) in the case where the interval
wherein the main sensor 5 or auxiliary sensor 6 detects the
mark body 24 exceeds 1000%.
As clear from this graph, there is a bilateral symmetry
relationship between the cases where the distance d between
the main sensor 5 and the auxiliary sensor 6 is positive and
is negative. It is understood that the error count reaches
its m;n;mum when the distance d is +35%, +106%, and +177%.
Figure 9 is a figure to explain the reason why the error
count reaches the minimum when the distance d is these
specific values. All the sensor loci in Figure 9 are inclined
only to 45~.
The sensor locus 30 shown with the solid line deviates
greatly from the center point of the mark 24 in the same way
19
CA 02222124 1997-11-2~
as the sensor locus D in Figure 7 (passing through the center
point of the lower side of each mark body 24); therefore, the
sensor locus cannot cross all three line segments 23a-23c of
each mark 24 (can only cross one or two of the line segments)
and as a result, the sensors cannot detect the marks 24.
Meanwhile, the sensor locus 31 or sensor locus 32 shown
with the dotted line in the figure deviates only by 50~ of the
distance (half the length of one side of the mark 24) from the
sensor locus 30. Because the sensor locus 31 or sensor locus
32 passes through the center points of all the marks 24, the
actual position on the mark 24 can be detected at all the
marks 24.
Ultimately, the distance d between all the sensor loci
(31, 32, ...) shown with the dotted lines in Figure 9 and the
sensor locus 30 shown with the solid line can be expressed as
follows:
d=(N+1/2) / ~ x100% (N is an integer) ... (1)
According to the aforementioned formula (1), d-about 35
when N=0, d=about 106~ when N=1, and d=about 176% when N=2;
this coincides with the results of the calculation shown ln
Figure 8.
Therefore, the mark 24 can definitely be measured with
the other sensor in the case where the distance d between the
main sensor 5 and the auxiliary sensor 6 bec~ -s the specific
distance shown in the aforementioned formula (1) (having the
relationship between the solid and dotted lines shown in
CA 02222124 1997-11-2
Figure 9).
Figure 10 shows the situation wherein the main sensor 5
and auxiliary sensor 6 alternately measure the marks 24 in the
case where the vehicle 1 travels at an angle. In Figure 10,
the locus of the main sensor 5 is shown with the solid line 33
and the locus of the auxiliary sensor 6 is shown with the
dotted line 34.
As shown in this figure, the main sensor 5 crosses the
three line segments 23a-23c of the mark 24-3; at the time when
passed through the last line segment 23a, the coordinate
position of that site 35 is found through calculation. Next,
the auxiliary sensor 6 intersects the three line segments 23a-
23c of the mark 24-4; at the time when passed through the last
line segment 23a, the coordinate position of that site 36 is
found through calculation. In this way, the main sensor 5 and
the auxiliary sensor 6 alternately measure the coordinate
positions on the mark 24.
The arrangement in which the sensors 5 and 6 are mounted
on is known; therefore, the serial position of the vehicle 1
can be calculated by determining the coordinate position on
the mark 24 as noted above, in effect the coordinate positions
of the sensors 5 and 6.
Figure 11 shows the procedure for the vehicle 1 standing
by as shown in Figure 1, from the application of electric
power to the approach to the traveling area 20.
Initially, the vehicle 1 is stopped at the standby
position 19 (Step 101); power is applied to the vehicle 1 in
this state (Step 102). At this stage, the vehicle l still
CA 02222124 1997-11-2~
does not precisely confirm its own position and direction.
Next, the vehicle 1 starts to move slowly forward without
changing its direction from the initial state (the direction
of the reference axis of measurement 42 of the marks 40 and
41) (Step 103).
Next, the vehicle 1 passes the mark 40 directly before
it. Because the orientation and central coordinate of the
mark 40 are known, the vehicle 1 can measure the coordinate
position of passing points on the mark 40 by passing the mark
40. In effect, at the time when the vehicle 1 passes the
final line segment 23a of the mark 40, the current position
P40 (x40, y40) of the vehicle 1 can be measured (Step 104).
However, at this time, the precise value of the direction of
travel is still unclear.
The vehicle 1 continues to move straight forward without
any changes; the passing coordinate position P41 (x41, y41) on
the mark 41 can be measured at the time when the vehicle 1
passes the next mark 41. Because the passing coordinate
position is determined at two points in this way, the
direction of the vehicle 1 between these passing points can be
found as the direction of a line connecting the two points.
The precise direction, in addition to its position, is
determined in this way (Step 105). Consequently, after this
stage, where the initial position and direction are
determined, it becomes possible to calculate its serial
position and direction with known dead reckoning technology.
Even after the vehicle 1 enters the traveling area 20 (Step
106), the serial position and direction within the traveling
CA 02222124 1997-11-2
area 20 can be estimated and calculated.
The vehicle 1 travels along a predetermined traveling
path L within the traveling area 20. At this time, the actual
position of each of the predetermined passing marks 24-5, 24-
6, 24-7 ..., on the predetermined traveling path L over which
the vehicle 1 passes, are measured in sequence in the same way
as the marks 40 and 41.
As a result, the estimated and calculated positions are
corrected serially according to the actual positions on the
mark measured in sequence and the vehicle 1 can travel
precisely along the predetermined traveling path L.
Moreover, this embodiment is explained with the
presupposition that the standby position 19 and the marks 40
and 41 used for the initial settings are external to the
traveling area 20, but the standby position may also be
established at a known position within the traveling area 20.
Also, the position and direction of the vehicle 1 remain
the same as before the power was turned off in the case where
the vehicle 1 is driven to the standby point, stopped
temporarily and the power is cut, then the vehicle 1 is let
stand at that position, the power is turned on once more, then
the vehicle 1 is started. In such a case, the data from
directly before the power is cut off can be used without any
processing as the data for the initial position and direction
when the power is turned on.
Figure 12 is a diagram to show the situation where the
error in its estimated position increases with the progress of
the vehicle 1, which estimates and calculates its position
CA 02222124 1997-11-2
with the dead reckoning technology.
Specifically, even if the position of the vehicle 1 is
accurately measured at the initial position, the range of
error reaches a size corresponding to 18a when the vehicle 1
advances just by distance a. In effect, it is known that the
vehicle 1 is within that error range 18a, but the precise
position within the error range 18a cannot be specified. In
the same way, the error range grows to 18b, 18c, 18d, and 18f
when the traveling distance of the vehicle 1 extends to b, c,
d, and f.
When the traveling distance is b, the error range 18b is
sufficiently less than the area 24a of the mark body 24.
Therefore it is at least determined that the vehicle 1 is
present on the mark body 24. Because the orientation and
central position of the mark body 24 are known, the precise
position on the mark 24 can be measured by the vehicle 1
passing the mark body 24 and detecting that all line segments
were crossed; this being the case, the vehicle 1 can update
its precise position through correction at that time.
However, the vehicle 1 fails to detect the mark 24 at any
of distances a, b, c, and d and could just barely detect the
mark 24 at distance f.
In this case, the error range 18f at distance f becomes
much greater than an individual mark 24a; the mark detected
here cannot ultimately be determined to be a particular mark.
In effect, the surrounding marks 24b, 24c, 24d, 24e, and 24f,
in addition to the mark 24a, are all within the error range
18f. Therefore, so long as sensors 5 and 6 normally pass the
24
; CA 02222124 1997-11-2~
mark 24 and detect all three line segments, the precise
position on that mark 24 can be calculated, but it is not
determined which mark is passed. Consequently, in this case,
it cannot be updated to the precise coordinate position of the
vehicle 1; the vehicle 1 fails to reduce the cumulative error
due to the dead reckoning technology.
Specifically, in the case where the vehicle 1 travels on
the interleaved marks 24, it is necessary to detect the next
mark 24 correctly while the error range of the cumulative
error due to the dead reckoning technology is sufficiently
less than the area of the mark bodies 24. Therefore,
consideration is given whether there is any problem regarding
this issue in practical application.
For example, when the side wheel drive vehicle 1, wherein
the space between the wheels 2 and 3 on the right and left is
60 cm and the wheel diameter is 200 mm, travels directly
forward for 3 m on a fairly irregular office floor, it has
produced a cumulative position error of 1 cm. This means that
the vehicle can travel for a distance of ten of the mark
bodies 24 with a side length of 30 cm. Consequently, if it is
possible to correctly detect the next mark body 24 during the
travel for a distance of ten times the side length of the
marks 24, since the cumulative error range at that site is
sufficiently less than the dimensions of the mark bodies 24,
it becomes clear whether a position was measured by passing
over a mark body 24 laid in a particular position and
direction. According to the above, it is thought that there
is no problem in practical application.
CA 02222124 1997-11-2~
Moreover, in the embodiment, it is presupposed that the
vehicle 1 has wheels on the sides, but it may also be a
vehicle with a single wheel. The method for finding the
current position and direction by changing the length of
travel of one wheel and the steering angle of that wheel is
already known according to Japanese Patent Application No. 61-
151421.
Also, in the embodiment, the position and direction of
the vehicle 1 is estimated based on the travel length of the
wheels, but a known inertial navigation method to find the
current position based on an accelerometer and gyro signal may
also be applied.
Also, in the embodiment, the explanation presupposes the
case of using two sensors, but a larger number of sensors may
also be used. In this case, depending also on the method of
arrangement, the traveling distance of the detection interval
of the marks 24 can generally be shortened.
Also, in the embodiment, sensors 5 and 6 are placed on
the axis connecting the side wheels, but placement in other
locations is naturally also possible. In effect, in the case
where the car body is rigid and does not change form, the
sensor loci and locus of the center of the car body can be
made to correspond by the addition of an operation for
geometrical coordinate transformation.
Also, in the embodiment, as shown in Figure 1, square
marks 24 are prepared as the elements; the traveling area 20
is formed with these square marks 24 providing coverage in
sequence so as to leave no gaps. However, a square unit 27 of
26
CA 02222124 1997-11-2~
a combination of four square marks 24 may be prepared, the
traveling area 20 may also be formed with this square unit 27
providing coverage in sequence so as to leave no gaps.
Likewise, a square unit 28 which is a combination of nine
square marks 24 or a square unit ... which is a combination of
sixteen marks may be prepared; the traveling area 20 may also
be formed with these square units 28 ... providing coverage in
sequence so as to leave no gaps.
Also, the traveling area 20 may be formed by directly
laying down the line segments 23a-23c, rather than laying the
square mark bodies 24 to form the traveling area 20.
Also, in the embodiment, it is presupposed that the marks
24 comprise line segments in a Z (N) form, but the line
segments may be in other geometrical forms as well. For
example, as shown in Figures 13(a)-13(e) and Figures 14(a)-
14(c), marks 50-57 having line segments in various geometrical
forms, which already became known in Japanese Patent
Application No. 60-108792 (Japanese Patent Publication No. 7-
3339), may also be used. Essentially, an arbitrary
configuration can be used if it is a geometrical form at least
having two line segments which are not parallel to each other
and with which the actual position on a mark can be detected
with a line segment crossing detector.
Also, in the embodiment, the course deviation (deviation
of predetermined passing position on a mark from the actual
passing position) attained through the detection of marks is
used to correct the position estimated by dead reckoning, but
the present invention is not limited to this. The use of the
CA 02222124 1997-11-2
course deviation attained is arbitrary.
As explained above, the present invention attains the
marked effects that the predetermined traveling path can be
established arbitrarily across the entire traveling area and
it is possible to respond flexibly to changes in the traveling
route.
INDUSTRIAL APPLICABILITY
The present invention may also be applied to manned
vehicles as well as automated vehicles; additionally, an
arbitrary structure may be applied as the structure of the
moving body.
28
