JPH10253372A - Recognition apparatus for position of moving body - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain a recognition apparatus by which the present position of a moving body can be recognized simply by a constitution wherein the running distance of a wheel and the swiveling angle of the moving body are found, the position of a swiveling center is decided by using them and the position of the moving body is found on the basis of the change portion of its coordinates. SOLUTION: Rotary encoders 4 which are attached to respective wheels measure the running distance Li of the wheels, and a gyro sensor 6 measures a swiveling angle Δθ. A division circuit 16 finds a distance (ri) up to a swiveling center from the respective wheels on the basis of the running distance Li and the swiveling angle Δθ. An intersection computing part 18 finds a swiveling center on the basis of the radius (ri) from the respective wheels on as to send its coordinates to a position computing part 20. The position computing part 20 finds the change portion of the coordinates of a running vehicle on the basis of the swiveling angle Δθ and the swiveling center R, and it finds new coordinates by adding previous coordinates P. When the running vehicle performs a linear motion and when the swiveling angle from the gyro sensor 6 is constant, it is regarded that the running vehicle performs the linear motion by the running distance Li, and the new coordinates P are found. A swiveling motion and the linear motion are changed over by switches S1 , S2 .

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for recognizing the position of a moving object such as an unmanned vehicle.

[0002]

2. Description of the Related Art The traveling of a moving object such as an unmanned traveling vehicle requires recognition of a current position. The traveling path of the moving object can be approximated to a straight line or a circle in a short time range, and it is easy to obtain the current position of the moving object performing a linear motion. The problem is the position recognition at the time of circular motion. In the conventional method, the direction to the turning center is obtained from the steering angle of each wheel of the moving body. For example, the turning center is at the intersection of the lines drawn from four wheels to the turning center. And If the turning center is known, the distance from the moving body to the turning center (turning radius) is also known, and the current position is advanced along the circumference of the known radius from the turning center by the traveling distance calculated from the number of rotations of the wheels and the like. It will be determined by the point.

However, in this method, the condition for position recognition is that the steering angle can be accurately measured. For the measurement of the steering angle, it is necessary to make a precise adjustment for each wheel so that a change in the direction of the wheel can be reliably measured. This adjustment requires time and a high technique, and it is difficult to make a precise adjustment in practice.

[0004]

SUMMARY OF THE INVENTION An object of the present invention is to make it possible to recognize the position of a moving body without adjusting the steering angle.

[0005]

The present invention relates to an apparatus for recognizing a current position of a moving body by integrating a change in the position of the moving body, wherein the moving body has a plurality of wheels, and at least two of the wheels have rotation speeds. Means for obtaining the turning angle of the moving body, and means for obtaining the distance from each wheel to the turning center of the moving body using the rotation speed of each wheel and the turning angle. Means for determining the position of the turning center from the intersection of a plurality of circles passing through the turning center about each wheel, and a change in the position when the moving body turns by the turning angle about the obtained turning center. Means for obtaining a new position of the moving body from the minute and the coordinates before the turn is provided.

Preferably, the moving body is an unmanned vehicle that travels along a predetermined traveling route on a trackless basis, and a mark is provided at a predetermined position on the traveling route. A means for calibrating is provided.

Preferably, the means for determining the turning angle is constituted by an optical gyro sensor.

[0008]

According to the present invention, at least two wheels of the moving body are provided with means for detecting the number of revolutions such as a rotary encoder, and the traveling distance of each wheel is obtained. In addition, a means for obtaining a turning angle such as a gyro sensor is provided, and a distance from each wheel to the turning center of the moving body is obtained from the number of rotations and the turning angle of the wheel. For example, the distance from the turning center to each wheel is r
If the travel distance of each wheel is Li and the turning angle is △ θ, it is considered that the wheel has traveled an arc of an angle △ θ on the circumference of the radius ri, so the relation of ri = Li / △ θ There is. Then, the distance from each wheel to the turning center is determined.

Since at least two wheels of the moving body are provided with means for detecting the number of revolutions, two or more circles passing through the turning center with the wheel as the center can be obtained. Therefore, the position of the turning center is determined from the intersection of these circles. When the position of the turning center is determined, it is considered that the moving object has turned around this point by the turning angle △ θ. This is the change in the coordinates of the moving object. Therefore, the current position of the moving object can be recognized by adding a change in the coordinates to the coordinates of the previous moving object.

According to the present invention, the steering angle of the wheels is not required for recognizing the current position of the moving body. Therefore, according to the present invention, the current position of the moving object can be easily recognized.

Although the type of the moving body is arbitrary, it is preferably an unmanned traveling vehicle traveling on a predetermined limited traveling route without a track, and a mark is provided at a predetermined position on the traveling route. The moving body is provided with a means for determining the position of the moving body with respect to the mark, and the result of the position recognition is calibrated. As a means for determining the turning angle, a gimbal type gyro sensor or the like may be used, but preferably an optical gyro sensor, particularly preferably an optical fiber gyro sensor is used. The optical gyro sensor is inexpensive and has a long service life as compared with a gimbal type mechanical gyro sensor. A mechanical gyro sensor needs to warm up for a long time until the rotation of the gimbal is stabilized, but an optical gyro sensor can start measurement in a few minutes.

[0012]

1 to 7 show an embodiment. FIG. 1 shows the configuration of a position recognition device. Reference numeral 1 denotes an unmanned traveling vehicle for unmannedly transporting articles in factories, warehouses, and the like, and 2R denotes a right front wheel and 2L.
Is a left front wheel, 3R is a right rear wheel, and 3L is a left rear wheel. A rotary encoder 4 is attached to each wheel to measure the number of rotations of the wheel. The rotary encoder 4 may be attached to at least two wheels, and any sensor can be used as long as it can detect the number of rotations of the wheels. Reference numeral 6 denotes an optical fiber gyro sensor, 8 denotes a position recognition unit, and 10 denotes a CCD camera, which may be another camera. Reference numeral 12 denotes a calibration unit for calibrating the current position based on the image of the mark taken by the CCD camera 10. Reference numeral 14 denotes an antenna for receiving a travel command from a host computer or the like.

FIG. 2 shows the processing of signals from the respective sensors. Reference numeral 16 denotes a division unit, and the gyro sensor 6 calculates the travel distance Li (i is 1 to 4 and indicates a wheel number) obtained by the rotary encoder 4. The distance ri (i is 1 to 4) from each wheel to the turning center is obtained by dividing by the obtained turning angle △ θ. Reference numeral 18 denotes an intersection calculation unit that obtains an intersection when a circle having a radius ri is drawn from each wheel, and uses this point as the turning center of the unmanned vehicle 1. The coordinates R of the turning center are sent to the position calculation unit 20, and it is assumed that the unmanned traveling vehicle 1 makes a turning motion around the turning center R by the turning angle △ θ, and a change ΔP in the coordinates of the unmanned traveling vehicle 1 during this period.
Ask for. Then, new coordinates are obtained by adding ΔP to the previous coordinates P.

The above process is a process when the unmanned traveling vehicle 1 is turning, and the gyro sensor 6 reports the change of θ to the position calculating unit 20. If the vehicle is moving, a new position P is obtained assuming that the unmanned traveling vehicle 1 has moved linearly in the same direction as before by the traveling distance Li of each wheel. Switches S1 and S2 are used to switch the processing between the turning motion and the linear motion. During the linear motion in which θ is constant, the switch S1 is closed, and the dividing unit 16 and the intersection calculating unit 18 are not used. During the turning motion, the switch S2 is closed, and a new position P is obtained as described above.

The calibration unit 12 includes a displacement calculation unit 2
2 and a map 23 storing the positions of the marks, the coordinates P of the unmanned traveling vehicle 1 input from the position calculator 20 determine which mark is being observed, and output the mark coordinates Mp. I do. On the other hand, the displacement calculation unit 22 captures, for example, two points of the mark with the CCD camera 10 and calculates the displacement of the CCD camera 10 with respect to the mark according to the principle of triangulation. Assuming that this displacement is Q in two-dimensional coordinates, Mp
The coordinates of the unmanned vehicle 1 are determined by + Q, and the coordinates are fed back to the position calculator 20 to calibrate the current position.

FIG. 3 shows the configuration of the optical fiber gyro sensor. Reference numeral 24 denotes a light source such as a laser diode, and reference numeral 25 denotes a detector such as a photodiode. 26 is a coupler,
The light from the light source 24 is split and the optical fibers 27 are split from both sides.
And returns the light from the optical fiber 27 to the detector 25.
Send to When the unmanned traveling vehicle 1 is turning, a phase difference is generated between the light rotating clockwise through the ring-shaped optical fiber 27 and the light rotating counterclockwise, and this phase difference is detected by the detector 25.
To detect. The phase difference is proportional to the angular velocity (rate of change of θ) of the unmanned vehicle 1, and the turning angle △ θ is determined from this. The optical fiber gyro sensor 6 has a long life because there is no mechanical movable part such as a gimbal. In addition, since it is not necessary to wait for the rotation to stabilize unlike a gimbal, the warm-up time is extremely short, about several minutes, and precise adjustment is unnecessary, so that the cost is low.

FIG. 4 shows the distance ri from the wheel to the turning center.
Is shown. When the movement of the unmanned traveling vehicle 1 in a short time involves turning, it can be approximated to an arc-shaped movement centered on the turning center. The length L of the arc can be obtained from the rotation speed of the wheel by the rotary encoder 4. The angle △ θ of the arc Li of this arc can be obtained by the gyro sensor 6. There is a relationship Li = ri / iθ between the distance ri from the turning center to the wheel and the arc length Li. Thus, the distance ri can be easily obtained from the traveling distance Li obtained by the rotary encoder 4 and the turning angle △ θ obtained by the gyro sensor.

Here, if the direction of the wheel itself is known precisely, the coordinates of the turning center can be found immediately from FIG.
However, for this purpose, it is necessary to precisely adjust the direction of the axle of each wheel and accurately measure the change in the direction of the axle as the steering angle. And this measurement is extremely difficult. Thus, in the present invention, the distance ri to the turning center is obtained for a plurality of wheels, and the coordinates R of the turning center are obtained from these.
Ask for.

FIG. 5 shows this mechanism.
For R and the left rear wheel 3L, the distance to the turning center is determined. Then the actual turning center is the right front wheel 2R
And a circle having a predetermined radius from the left rear wheel 3L. In the embodiment, since the rotary encoder 4 is provided for each of the four wheels, the number of intersections obtained by calculation further increases, and these can be averaged to more accurately determine the coordinates R of the turning center.

When the coordinates R of the turning center are determined, the change ΔP in the position of the unmanned traveling vehicle 1 during this time is determined at a point where the unmanned vehicle 1 has moved by the turning angle Δθ about the turning center R. The original coordinates of each wheel are determined based on the original coordinates of the unmanned traveling vehicle 1, and the coordinates of the turning center R are determined based on the original coordinates. Next, the change ΔP in the position of the unmanned traveling vehicle 1 is determined by the determined turning center coordinate R and the turning angle Δθ. Then, when ΔP is added to the original coordinates P, new coordinates of the unmanned traveling vehicle 1 are obtained. By repeating this step, the coordinates of the unmanned traveling vehicle 1 during the turning motion can be obtained almost continuously. In the case of the linear motion, the traveling direction of the unmanned traveling vehicle 1 during the linear motion is determined from the integrated value of the turning angle △ θ before that. Here, the traveling distance of each wheel is constant in the linear motion, and new coordinates are determined assuming that the traveling distance of each wheel advances in the direction of the movement direction of the unmanned traveling vehicle 1.

FIG. 6 shows calibration for position recognition. The unmanned traveling vehicle 1 travels on a predetermined traveling route. The traveling route has a branch, a loop, and the like, and various positions actually travel within the width of the permitted traveling route. Marks 30 are provided at a plurality of positions along the traveling route. As shown in FIG. 6, the mark 30 is composed of line segments and their intersections and end points, and can be geometrically processed as a graph. Now, when the mark 30 is photographed by the CCD camera 10, the camera 1 with respect to the mark 30 is formed based on the principle of triangulation using, for example, two points on the mark.
The position of 0 can be determined. For example, in the case of FIG. 6, the position of the camera 10 is obtained using the two points 31 and 32 of the mark 30.

As shown in FIG. 2, the calibration section 12 has a map 23 in which the coordinates of each mark on the traveling route are recorded. The coordinates of the unmanned traveling vehicle 1 before calibration are obtained by the position calculation unit 20, and it is possible to know which mark the user is looking at from this. In order to prevent the marks from being confused, the pattern may be changed for each mark, and the pattern for each mark may be recorded in the map 23. Now, according to the image taken by the CCD camera 10,
The coordinates Q of the unmanned traveling vehicle 1 based on the mark 30 are determined, and the coordinates of the mark 30 are known as Mp in the map 23. Therefore, Mp + Q becomes the actual coordinates of the unmanned vehicle 1, and the position calculator 20 is calibrated using this.

In this way, it is possible to compensate for a decrease in the accuracy of detecting the coordinates of the turning center R due to unevenness or the like in the traveling route. That is, in FIGS. 4 and 5, it is assumed that the unmanned traveling vehicle 1 travels on a flat, non-slip plane. However, if the traveling route has irregularities, the traveling distance of each wheel obtained by the rotary encoder 4 and The actual traveling distance is different, and an error corresponding to the actual traveling distance occurs. Such errors gradually increase when integrated. Therefore, by providing the mark 30 on the traveling route, accumulation of errors can be prevented.

FIG. 7 shows an algorithm of position recognition in the embodiment. It is assumed that the unmanned traveling vehicle 1 warms up the gyro sensor 6, for example, once a day, captures the mark with the CCD camera 10 near the position where the mark exists, and calibrates the initial position. Thus, the initial value P of the position is obtained. The initial value of the turning angle △ θ is 0, and the initial motion direction is the direction of the angle θ in the 0 direction. When the unmanned traveling vehicle 1 starts traveling, it is determined whether or not the angle θ changes, whether it is a linear motion or a turning motion. If it is a linear motion, the traveling distance Li of each wheel is used as the traveling distance of the unmanned traveling vehicle 1 as it is. , The traveling direction is determined using the angle θ, and a new coordinate P is obtained.

When a turning motion is involved, the distance ri to the turning center is determined for each wheel as shown in FIG. Next, as shown in FIG. 5, the intersection of the circle with the radius ri is obtained from each wheel.
The intersection is defined as the turning center and the coordinates are set as R. Assuming that the distance from the turning center to the center of gravity of the unmanned traveling vehicle 1 is ρ, ρ is determined by the difference between the original center-of-gravity coordinates of the unmanned traveling vehicle 1 and the coordinates of the turning center R, and the change in position due to the turning {P is ρ} It is determined by θ. Then, new coordinates are determined by correcting the coordinates of the unmanned traveling vehicle 1 by ΔP.

When the unmanned vehicle passes near the mark 30, two points on the mark are observed using the CCD camera 10, and a displacement Q from the mark is determined from the triangulation principle. Then, the coordinates Mp of the mark are obtained from the map 23,
Calibrate the coordinate position as P = Mp + Q.

As a result, the current position of the unmanned vehicle 1 can be recognized without measuring the steering angle, the current position can be calibrated by the mark 30 and the CCD camera 10, and the gyro sensor is inexpensive, has a long life and is easy to warm up. The optical fiber gyro sensor 6 can be used.

[Brief description of the drawings]

FIG. 1 is a diagram showing an arrangement of a moving object position recognition apparatus according to an embodiment.

FIG. 2 is a block diagram of a position recognition device for a moving object according to the embodiment;

FIG. 3 is a diagram showing a configuration of an optical fiber gyro sensor used in the embodiment.

FIG. 4 is a characteristic diagram showing calculation of a distance from a wheel to a turning center in the embodiment.

FIG. 5 is a characteristic diagram showing determination of a turning center and calculation of a change in vehicle coordinates in the embodiment.

FIG. 6 is a characteristic diagram showing coordinate calibration using a camera and a mark in the embodiment.

FIG. 7 is a flowchart illustrating a position recognition algorithm of a moving object in the embodiment.

[Explanation of symbols]

 Reference Signs List 1 unmanned traveling vehicle 2R, 2L front wheel 3R, 3L rear wheel 4 rotary encoder 6 optical fiber gyro sensor 8 position recognition unit 10 CCD camera 12 calibration unit 14 antenna 16 division unit 18 intersection calculation unit 20 position calculation unit 22 displacement calculation unit 23 Map 24 Light source 25 Detector 26 Coupler 27 Optical fiber 30 Mark 31, 32 Point on mark S1, S2 Switch

Claims (3)

[Claims] 1. A device for recognizing a current position of a moving body by integrating a change in position of the moving body, wherein the moving body has a plurality of wheels, and at least two of the wheels have rotation number detecting means. Along with providing, means for determining the turning angle of the moving body, means for determining the distance from each wheel to the center of turning of the moving body using the rotation speed of each wheel and the turning angle, Means for determining the position of the turning center from the intersection of a plurality of circles passing through the turning center as a center, and a change in position when the moving body turns by the turning angle about the obtained turning center and before turning. And a means for determining a new position of the moving object from the coordinates of the moving object. 2. The vehicle according to claim 1, wherein the moving body is an unmanned vehicle that travels along a predetermined traveling route on an untracked path, and a mark is provided at a predetermined position on the traveling route. 2. The apparatus according to claim 1, further comprising means for calibrating the position. 3. The moving object position recognition apparatus according to claim 1, wherein the means for obtaining the turning angle is constituted by an optical gyro sensor.