JP6677516B2 - Autonomous traveling device - Google Patents

Description

  The present invention relates to an autonomous traveling device, and more particularly to an autonomous traveling device having a function of measuring a distance to an obstacle and a function of avoiding a collision with the obstacle.

2. Description of the Related Art Today, autonomous traveling devices that move autonomously, such as a transport robot that transports luggage and a monitoring robot that monitors conditions in and around a building or a predetermined site, are used.
Such a conventional autonomous traveling device stores map information and travel route information of an area to be traveled in advance, and uses an information acquired from a camera, a range image sensor, and a GPS (Global Positioning System) to detect an obstacle. Drive on a predetermined route while avoiding the road.

Further, in a conventional autonomous traveling device (hereinafter, also simply referred to as a vehicle), when an obstacle is found on an autonomously traveling route, the vehicle travels while decelerating, changes the route, or collides with the obstacle. Is stopped.
In order to detect an obstacle, for example, a camera or a distance sensor that emits laser light and detects reflected light from an object is used.
The distance sensor detects the distance to the object by measuring the time until the reflected light returns, and detects the position of the object by changing the emission direction of the laser.
For example, the distance sensor scans the left and right sides of the vehicle in the horizontal direction at an angle of 90 degrees with respect to the front in the traveling direction of the vehicle. Is detected.

Further, in order to drive safely, speed control is performed in accordance with the measured distance from the distance sensor to the obstacle.
For example, when the distance to the measurement point at the position where the reflected light is detected is relatively long, such as 10 m or more, the vehicle travels at a high speed normal speed, and when the distance to the measurement point is within 3 m, deceleration is started. Further, when the distance to the measurement point is within 1 m, speed control such as stopping is performed.

That is, in an area where an obstacle is detected, weighting for speed control is set in accordance with the distance from the distance sensor to the measurement point. For example, the weighting value is set highest in an area where the distance to the measurement point where the stop processing is to be performed is relatively short, and the distance to the measurement point where the deceleration processing is to be performed is within a predetermined distance range. The speed control based on the measured distance to the obstacle has been performed by setting the weighting value to the next higher value and setting the weighting value to the lowest value in a relatively far area where deceleration is not performed.
Further, for other vehicles (obstacle candidate objects) recognized as obstacles, the direction and position of the relative speed vector with respect to the own vehicle are obtained, and the weighting function calculated from the relative speed vector and the position is used to determine the position of the vehicle. A recognizing device that performs high-speed and accurate recognizing determination of an object serving as an obstacle to a vehicle has also been proposed (see Patent Document 1).

JP 2002-225656 A

However, in the speed control in the conventional autonomous traveling device, only the deceleration and the stop processing are performed based on the distance from the vehicle to the obstacle, and the deceleration and the stop processing are performed based on the traveling direction of the vehicle. Did not.
For example, if a fan-shaped area where the position of the distance sensor is the center position of the key of the fan is set as the obstacle detection area, if the vehicle enters a fan-shaped area of a predetermined distance L1 relatively short from the center position, the vehicle is moved. The process of stopping is performed.
If the fan-shaped area is, for example, a fan shape opened at 180 degrees, the vehicle is stopped when an obstacle is detected within a distance L1 ahead in the traveling direction of the vehicle. Alternatively, the vehicle is also stopped when an obstacle is detected not within the forward direction of the vehicle but within a distance L1 of a fan-shaped end portion in the right direction with respect to the traveling direction.
That is, when an obstacle exists in a region within the distance L1 between the fan-shaped region in the right direction and the fan-shaped region in the left direction with respect to the traveling direction, a process of stopping the vehicle has been performed.

If an obstacle is detected immediately in front of the traveling direction, it is meaningful to stop for safe driving. However, if it is detected that an obstacle exists at a position on the right side of the traveling direction and within a relatively short distance L1, but does not interfere with traveling forward in the traveling direction, the problem on safe traveling is raised. Stopping the vehicle when there is no speed control is not appropriate speed control from the viewpoint of rapid autonomous driving.
When a sector-shaped region having a distance L2 longer than the distance L1 from the center position of the distance sensor is set as the region for performing the vehicle deceleration processing, the right-side end of the sector-shaped region and the distance L2 If an obstacle is detected in the area within the range, even if the position of the obstacle does not hinder the traveling of the vehicle in the traveling direction, the vehicle is decelerated, and the deceleration process is performed. Is not an appropriate speed control. That is, although it is possible to continue running without decelerating, deceleration processing is performed, so that it can be determined that this corresponds to erroneous detection of an obstacle.

Further, for example, when the course is changed slightly rightward with respect to the current traveling direction, or when the vehicle turns right, a portion of the sectoral area in the rightward direction with respect to the traveling direction in the sectoral area for detecting an obstacle. Since the vehicle is going to travel from now on, it is important information for safe driving whether or not an obstacle exists in the fan-shaped area in the right direction.
However, when the vehicle turns to the right, the portion of the sector in the left direction with respect to the traveling direction moves away from the vehicle.Therefore, whether there is an obstacle in the sector in the left direction depends on safe driving. It is considered that the information is not important information, and if an obstacle is detected in this fan-shaped area in the left direction, there may be no problem in safe driving even if the existence of the obstacle is almost ignored.
That is, when speed control is performed by detecting an obstacle existing within a certain distance based on information on a measured distance from a distance sensor without considering a future course that the vehicle is about to travel, Although it is possible to continue traveling without deceleration for safe traveling, unnecessary deceleration processing and stoppage may be performed.

  Therefore, the present invention has been made in consideration of the above circumstances, and performs an obstacle detection process in consideration of the traveling direction of the autonomous traveling device in order to secure the safety of the autonomous traveling, thereby performing traveling. It is an object of the present invention to provide an autonomous traveling apparatus that effectively detects an obstacle having a high possibility of interfering and continues traveling quickly and safely without decelerating or stopping as much as possible.

  The present invention provides a vehicle, a traveling control unit that controls a driving member that causes the vehicle to travel, and a predetermined light emitted to a predetermined obstacle determination area including a front space in a traveling direction, and the predetermined light is emitted into the obstacle determination area. A distance detection unit that receives light reflected by an existing object and detects a distance to the object; and an inertial measurement that detects a current traveling direction of the vehicle body by measuring a displacement caused by traveling of the vehicle body. Section, route information in which the traveling route of the vehicle body is determined in advance, determination area information in which the position and size of each divided area obtained by dividing the obstacle determination area into a plurality are set, and setting is performed based on the traveling state of the vehicle body. A storage unit that stores an area weight value indicating the degree of importance of detecting an obstacle in each divided area, and a control unit, wherein the control unit detects the current traveling direction detected by the inertial measurement unit, Or Autonomous driving, wherein, based on a traveling direction to be traveled in a traveling route defined in the route information, an area weight value is set for each divided area in consideration of a subsequent traveling direction of the vehicle body. An apparatus is provided.

Further, a comparatively large area weight value is set for the divided area A including the traveling direction of the vehicle body and some divided areas near the divided area A.
Further, when it is determined that an object whose distance has been detected exists in the divided region in which the relatively large region weight value is set, a deceleration process or a stop process of the vehicle body is performed in accordance with the region weight value. It is characterized by the following.
According to this, a relatively large area weight value is set in the divided area including the traveling direction of the vehicle body, so that, when an object exists in the traveling direction of the vehicle body, the vehicle body is surely decelerated or stopped. Can be.

Further, when it is determined that no object exists in the divided area in which the relatively large area weight value is set, the deceleration processing and the stop processing of the vehicle body are not performed.
Further, in a divided region other than the divided region in which the relatively large region weight value is set, when it is determined that the object whose distance is detected is present, the vehicle deceleration process and the stop process are not performed. I do.
According to this, when it is determined that an object is present in a divided area other than the divided area in which the relatively large area weight value is set, the deceleration processing and the stop processing of the vehicle body are not performed. Unlike the above, when an object exists at a position where there is little possibility of collision with the vehicle body, traveling can be continued quickly and safely without unnecessary deceleration or stop.

Further, the divided areas to which the relatively large area weight values are set include a divided area A including the traveling direction of the vehicle body and both divided areas of the divided area A. .
Further, a largest area weight value is set for a divided area including the traveling direction of the vehicle body, and a smaller area weight value is set for other divided areas as the distance from the traveling direction increases.

Further, when the vehicle body is going to turn to the left or right in a stationary state, a relatively large area of the obstacle determination area is divided into a plurality of divided areas existing in a direction in which the vehicle body travels by the turn. An area weight value is set, and a comparatively small area weight value is set for another divided area existing in a direction opposite to a direction in which the vehicle body advances by turning.
According to this, when the vehicle body is going to turn, a relatively large area weight value is set in a plurality of divided regions existing in the direction in which the vehicle body travels by turning, so that the object existing in the traveling direction at the time of turning can be reliably determined. And appropriate processing such as deceleration can be performed.
Also, a relatively small area weight value is set for other divided areas existing in the direction opposite to the direction in which the vehicle body travels by turning, so that even if an object exists in the direction opposite to the traveling direction, unnecessary Traveling can be continued without deceleration or stopping.

Further, when the inertia measuring unit detects that the vehicle body skids in a specific direction, the obstacle determination area compares the obstacle determination area with a plurality of divided areas existing in a direction in which the vehicle body advances due to the skid. A relatively small area weight value is set for a plurality of divided areas existing in a direction opposite to the direction in which the vehicle body advances due to the skid.
According to this, when the vehicle body skids, a relatively large area weight value is set in a plurality of divided regions existing in the direction in which the vehicle body advances due to the skid, so that the object existing in the direction advancing due to the skid is surely detected. Upon detection, appropriate processing such as deceleration can be performed.
In addition, since a relatively small area weight value is set for a plurality of divided areas existing in the direction opposite to the direction in which skidding proceeds, unnecessary deceleration is performed even if an object exists in the direction opposite to the direction in which skidding proceeds. And do not stop.

  Further, with the light emission position as a center, the light is scanned in a direction horizontal to the ground, and a horizontal fan-shaped area that can be reached by the light is defined as the obstacle determination area. , Wherein the obstacle determination area is an area divided into a plurality of sector shapes having a predetermined central angle.

Further, with the light emission position as a center, the light is scanned in a direction perpendicular to the ground, and a vertical fan-shaped area to which the light can reach is defined as an obstacle determination area. The obstacle determination area is a rectangular area divided into a plurality of pieces in the vertical direction, and the area weight value is changed according to the height of each divided area from the ground. It is characterized in that a region weight value smaller than the region is set.
According to this, a smaller area weight value is set in the divided area closest to the ground than in the other divided areas. For example, when the rear wheel of the vehicle rides on a step and the front of the vehicle leans downward, Detection as an obstacle can be reduced.

Further, the distance detection unit includes a light-emitting unit that emits light to the obstacle determination area in a traveling direction, a light-receiving unit that receives light, and the light-emitting unit that emits light toward a plurality of measurement points in the obstacle determination area. And a scanning control unit that scans the light emission direction so that the light is emitted. The distance detection unit determines when the light is emitted from the light emitting unit and the light reflected by the object is the light receiving unit. Calculating a distance between the light emitting unit and the plurality of measurement points by using a time difference from a time when it is confirmed that the object is received, and detecting a distance to the object from the calculated distance. It is characterized by.
Further, the distance detection unit may use a LIDAR that emits a laser to a two-dimensional space or a three-dimensional space in a predetermined obstacle determination area and measures distances at a plurality of measurement points in the obstacle determination area. Features.

  According to the present invention, the obstacle determination area is divided into a plurality of divided areas, and the area weight value is set for each of the divided areas in consideration of the subsequent traveling direction of the vehicle body. By performing the obstacle detection using, the traveling can be quickly and safely continued without unnecessary deceleration or stop.

BRIEF DESCRIPTION OF THE DRAWINGS It is an external view of one Example of the autonomous traveling apparatus of this invention. FIG. 2 is an explanatory diagram of a configuration related to traveling of the autonomous traveling device of the present invention. 1 is a configuration block diagram of an embodiment of the autonomous traveling device of the present invention. FIG. 4 is a schematic explanatory diagram of one embodiment of a distance detection unit of the present invention. FIG. 3 is a schematic explanatory diagram of a scanning direction of a laser emitted from a distance detection unit according to the present invention. It is the schematic explanatory drawing which looked at the irradiation field of the laser of this invention from the upper direction and the back. FIG. 4 is an explanatory diagram of one embodiment of information stored in a storage unit of the present invention. FIG. 4 is an explanatory diagram of one embodiment of information stored in a storage unit of the present invention. FIG. 3 is an explanatory diagram of an embodiment of an obstacle determination area and a divided area according to the present invention. FIG. 3 is an explanatory diagram of one embodiment of a divided area according to the present invention. FIG. 7 is an explanatory diagram of an example of setting a region weight value when traveling straight ahead and an embodiment of obstacle detection. FIG. 7 is an explanatory diagram of an example of setting a region weight value at the start of a right turn and an example of obstacle detection. FIG. 7 is an explanatory diagram of an example of setting a region weight value at the start of rightward travel and an example of obstacle detection. FIG. 7 is an explanatory diagram of an example of setting a region weight value when detecting a right skid and an embodiment of obstacle detection. FIG. 9 is an explanatory diagram of an example of obstacle detection in a case where detection points of measurement points are different at the start of a right turn. FIG. 8 is an explanatory diagram of an example of setting a region weight value in a vertical direction.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is not limited by the description of the following embodiments.
<Configuration of autonomous traveling device>
FIG. 1 shows an external view of an embodiment of the autonomous traveling apparatus according to the present invention.
In FIG. 1, an autonomous traveling device 1 of the present invention is a vehicle having a function of autonomously moving while avoiding an obstacle based on predetermined route information.
In addition, the autonomous traveling device 1 may include various functions such as a transport function, a monitoring function, a cleaning function, a guidance function, and a notification function in addition to the movement function.
In the following embodiments, an autonomous traveling apparatus that can autonomously travel in a predetermined outdoor monitoring area or passage and monitor and transport the monitoring area and the like will be mainly described.

In the external view of FIG. 1, the autonomous traveling device 1 (hereinafter, also referred to as a vehicle) mainly includes a vehicle body 10, four wheels (21, 22), a monitoring unit 2, and a control unit 3.
The monitoring unit 2 is a part having a function of checking a state of a moving area or a road surface and a function of monitoring a monitoring target. For example, a distance detecting unit 51 for checking a state of a moving front space, a camera (imaging unit) 55, a position information acquisition unit 58 for acquiring information on the current position where the vehicle is traveling.
The control unit 3 is a part that executes a traveling function and a monitoring function of the autonomous traveling apparatus of the present invention, and includes, for example, a control unit 50, an image recognition unit 56, an obstacle detection unit 57, a communication unit 54, It comprises a storage unit 70 and the like.

The autonomous traveling device 1 of the present invention uses the camera 55, the distance detection unit 51, the obstacle detection unit 57, and the like to travel while checking the state of the vehicle body 10 in the forward direction in the traveling direction. For example, when it is detected that an obstacle or a step is present ahead, the course is changed by performing operations such as stopping, rotating, retreating, and moving forward in order to prevent collision with the obstacle. Then, when an obstacle is recognized by image recognition or when a contact is detected, a predetermined function such as a stopping operation of the vehicle body is executed.
However, if an obstacle is detected within the specified obstacle detection area, but it is possible to run without colliding with the obstacle or if the obstacle does not hinder running, autonomous driving will continue without deceleration. Let it.

FIG. 2 is an explanatory diagram of a configuration related to traveling of the autonomous traveling device of the present invention.
FIG. 2A is a right side view of the vehicle 1, and the right front wheel 21 and the rear wheel 22 are indicated by phantom lines. FIG. 2B is a cross-sectional view taken along line BB of FIG. 2A, and sprockets 21b, 22b, 31b, and 32b described later are indicated by phantom lines. Front wheels (21, 31) are arranged on the front surface 13 of the vehicle body 10, and rear wheels (22, 32) are arranged on the rear surface 14.
A band-shaped cover 18 is provided on each of the side surfaces 12R and 12L of the vehicle body 10, and extends along the front-back direction of the vehicle body 10. Axles 21a, 31a and axles 22a, 32a for rotatably supporting the front wheels 21, 31 and the rear wheels 22, 32 are provided below the cover 18. Each axle 21a, 31a, 22a, 32a is independently rotatable when not coupled by a power transmission member.

  Belts 23 and 33, which are power transmission members, are provided on the pair of left and right front wheels (21, 31) and rear wheels (22, 32). Specifically, a sprocket 21b is provided on the axle 21a of the right front wheel 21, and a sprocket 22b is provided on the axle 22a of the rear wheel 22. Further, between the front wheel sprocket 21b and the rear wheel sprocket 22b, for example, a belt 23 provided on the inner surface side with a projection meshing with the sprocket is wound. Similarly, a sprocket 31b is provided on the axle 31a of the left front wheel 31, and a sprocket 32b is provided on the axle 32a of the rear wheel 32. A sprocket 32b is provided between the front wheel sprocket 31b and the rear wheel sprocket 32b. , A belt 33 having the same structure as the belt 23 is wound.

Therefore, the pair of left and right front wheels and rear wheels (21 and 22, 31 and 32) are connected and driven by the belts (23 and 33), so that only one of the wheels needs to be driven. For example, the front wheels (21, 31) may be driven. When one of the wheels is a driving wheel, the other wheel functions as a driven wheel driven without slipping by a belt as a power transmission member.
As a power transmission member for connecting and driving the pair of left and right front wheels and the rear wheel, a sprocket and a belt provided with a projection meshing with the sprocket are used.For example, a sprocket and a chain meshing with the sprocket are used. Is also good. Further, when slippage is allowable, a pulley and a belt having large friction may be used as the power transmission member. However, the power transmission member is configured such that the rotation speeds of the drive wheel and the driven wheel are the same.
In FIG. 2, the front wheels (21, 31) correspond to driving wheels, and the rear wheels (22, 32) correspond to driven wheels.

  Two motors, an electric motor 41R for driving the right and left front wheels 21 and 22 and an electric motor 41L for driving the left and right front wheels 31 and 32, are provided on the front wheel side of the bottom surface 15 of the vehicle body 10. ing. A gearbox 43R is provided as a power transmission mechanism between the motor shaft 42R of the right electric motor 41R and the axle 21a of the right front wheel 21. Similarly, a gear box 43L is provided as a power transmission mechanism between the motor shaft 42L of the left electric motor 41L and the axle 31a of the left front wheel 31. Here, the two electric motors 41R and 41L are arranged in parallel so as to be symmetrical with respect to the center line in the traveling direction of the vehicle body. Has been established.

  The gear boxes 43R and 43L are composed of a plurality of gears and shafts, and are assembly parts that change the torque, the number of revolutions, and the direction of rotation to transmit the power from the electric motor to the axle that is the output shaft. A clutch for switching the cutoff may be included. The left and right rear wheels 22 and 32 are supported by bearings 44R and 44L, respectively, and the bearings 44R and 44L are disposed close to the right side surface 12R and the left side surface 12L of the bottom surface 15 of the vehicle body 10, respectively. .

With the above configuration, the pair of front and rear wheels 21 and 22 on the right side in the traveling direction and the pair of front and rear wheels 31 and 32 on the left side can be driven independently. That is, the power of the right electric motor 41R is transmitted to the gearbox 43R via the motor shaft 42R, and the rotation speed, torque or rotation direction is changed by the gearbox 43R and transmitted to the axle 21a. The rotation of the axle 21a causes the wheels 21 to rotate, and the rotation of the axle 21a is transmitted to the axle 22a via the sprocket 21b, the belt 23, and the sprocket 22b, thereby rotating the rear wheel 22. The transmission of power from the left electric motor 41L to the front wheel 31 and the rear wheel 32 is the same as the above-described right operation.

When the rotational speeds of the two electric motors 41R and 41L are the same, the autonomous traveling device 1 will move forward or backward if the gear ratios (reduction ratios) of the gearboxes 43R and 43L are the same. When changing the speed of the autonomous traveling device 1, the gear ratio of each of the gear boxes 43R and 43L may be changed while maintaining the same value.
When the traveling direction is changed, the gear ratio of each gear box 43R, 43L is changed to change the rotation speed of the right front wheel 21 and the rear wheel 22 and the rotation speed of the left front wheel 31 and the rear wheel 32. What is necessary is just to make a difference. Further, by changing the rotation direction of the output from each of the gear boxes 43R and 43L, the rotation directions of the left and right wheels are reversed, so that the stationary turning around the center of the vehicle body becomes possible.

  In the case where the autonomous traveling device 1 is turned at a fixed position, since a steering mechanism for changing the angle of the front and rear wheels is not provided, the larger the distance between the front and rear wheels (wheel base), the greater the resistance applied to the wheels. A large driving torque is required for turning. However, since the gear ratios in the gear boxes 43R and 43L are variable, a large torque can be applied to the wheels only by lowering the rotation speed of the wheels during turning.

  For example, as the gear ratio in the gearbox 43R, when the number of teeth of the gear on the motor shaft 42R side is 10, the number of teeth of the intermediate gear is 20, and the number of teeth of the gear on the axle 21b is 40, the rotation speed of the axle 21b Is 1/4 the number of rotations of the motor shaft 42R, but a torque four times as large is obtained. Further, by selecting a gear ratio that further reduces the number of revolutions, a larger torque can be obtained, so that it is possible to turn even on a road surface with a large resistance to wheels such as uneven terrain and sandy ground. .

  Further, since the gearboxes 43R and 43L are provided between the motor shafts 42R and 42L and the axles 21a and 31a, vibrations from the wheels 21 and 31 are not directly transmitted to the motor shafts. Further, a clutch for transmitting and disconnecting (disconnecting) power is provided in the gear boxes 43R and 43L, and when the electric motors 41R and 41L are not energized, the electric motors 41R and 41L and the axles 21a and 31a serving as drive shafts are provided. It is desirable to shut off the power transmission during. Thus, even if a force is applied to the vehicle body 10 at the time of stop and the wheels rotate, the rotation is not transmitted to the electric motors 41R and 41L, so that no back electromotive force is generated in the electric motors 41R and 41L, and the electric motor There is no risk of damaging the circuits 41R and 41L.

As described above, since the left and right pair of front and rear front wheels and the rear wheel are connected by the power transmission member and the four wheels are driven by the two electric motors arranged on the front wheel side, the rear wheels are exclusively used. It is not necessary to provide a gearbox dedicated to the rear wheel required between the electric motor and the electric motor and the rear wheel, and the installation space for the electric motor dedicated to the rear wheel and the gearbox can be reduced. .
As described above, the two electric motors 41R, 41L are disposed on the left and right sides in the traveling direction on the front wheels 21, 31 on the bottom surface 15 of the vehicle body 10, and the gearbox 43R is provided on the left and right sides of each of the electric motors 41R, 41L. , 43L, but only the bearings 44R, 44L are disposed on the rear wheels 22, 32 side of the bottom surface 15, so that the bottom surface 15 of the vehicle body 10 A wide accommodation space 16 can be secured up to the end.

  Each of the electric motors 41R and 41L uses a battery (rechargeable battery) 40 such as a lithium ion battery as a power source, and installs the battery 40 in the housing space 16. Specifically, the battery 40 has a rectangular parallelepiped outer shape, for example, and can be placed at a substantially central position on the bottom surface 15 as shown in FIG. Further, it is desirable that the rear surface 14 of the vehicle body 10 be configured to be openable and closable with respect to, for example, the top surface or the bottom surface 15 so that the battery 40 can be easily inserted into and removed from the storage space 16. As a result, a large-capacity battery 40 for realizing long-time running can be mounted in the housing space 16 of the vehicle body 10, and operations such as replacement, charging, and inspection of the battery 40 can be easily performed from the rear surface 14. Will be possible. Furthermore, since the battery 40 can be disposed on the bottom surface 15, an electric vehicle that has a low center of gravity of the vehicle body 10 and can run stably can be obtained.

FIG. 3 is a block diagram showing the configuration of an embodiment of the autonomous traveling apparatus according to the present invention.
3, the autonomous traveling apparatus 1 of the present invention mainly includes a control unit 50, a distance detection unit 51, a traveling control unit 52, wheels 53, a communication unit 54, a camera 55, an image recognition unit 56, an obstacle detection unit 57, A position information acquisition unit 58, a rechargeable battery 59, a speed detection unit 60, a monitoring information acquisition unit 61, an inertial measurement unit 62, an area weight calculation unit 63, and a storage unit 70 are provided.
In addition, the autonomous traveling device 1 is connected to the management server 5 via the network 6, autonomously travels based on instruction information and the like sent from the management server 5, and transmits acquired monitoring information and the like to the management server 5.
As the network 6, any network currently used can be used, but since the autonomous mobile device 1 is a mobile device, a network capable of wireless communication (for example, a wireless LAN) can be used. preferable.

  As a network for wireless communication, the Internet or the like open to the public may be used, or a wireless network of a dedicated line in which connectable devices are limited may be used. As a wireless transmission method in a wireless communication path, various wireless LAN (Local Area Network) (whether or not WiFi (registered trademark) authentication is performed), ZigBee (registered trademark), Bluetooth (registered trademark) LE (Low Energy ) May be used in consideration of the wireless reach distance and the transmission band. For example, a mobile phone network may be used.

The management server 5 mainly includes a communication unit 91, a monitoring control unit 92, and a storage unit 93.
The communication unit 91 is a part that communicates with the autonomous traveling device 1 via the network 6, and preferably has a wireless communication function.
The monitoring control unit 92 is a part that executes a movement control for the autonomous traveling device 1, an information collection function and a monitoring function of the autonomous traveling device 1, and the like.
The storage unit 93 stores information for instructing the autonomous traveling device 1 to move, monitoring information (reception monitoring information 93a) sent from the autonomous traveling device 1, a program for monitoring control, and the like. It is.

The control unit 50 of the autonomous traveling device 1 is a part that controls the operation of each component such as the traveling control unit 52, and is realized mainly by a microcomputer including a CPU, a ROM, a RAM, an I / O controller, a timer, and the like. You.
The CPU performs the running function, the image recognition function, the obstacle detection function, and the like of the present invention by organically operating various hardware based on a control program stored in the ROM or the like in advance.

The distance detection unit 51 is a unit that detects a distance from a current position of the vehicle to an object existing in a predetermined space (for example, an obstacle determination area) including a space ahead in the traveling direction and a road surface. The distance detection unit 51 is arranged near the center of the front surface of the vehicle body. Here, when the vehicle travels outdoors, the object means, for example, a building, a pillar, a wall, a protrusion, or the like.
The distance detection unit 51 emits predetermined light to an obstacle determination area in the forward space in the traveling direction, and then receives reflected light reflected by an object and a road surface existing in the forward space to determine a distance to the object and the road surface. Is detected. Specifically, the distance detecting unit 51 mainly includes a light emitting unit 51a that emits light to an obstacle determination area in a predetermined forward space in a traveling direction, a light receiving unit 51b that receives light reflected by an object, And a scanning control unit 51c that changes the emission direction of the image two-dimensionally or three-dimensionally.

FIG. 4 is an explanatory diagram of one embodiment of the distance detecting section 51 of the present invention.
Here, it is shown that the laser 51d emitted from the light emitting unit 51a is reflected by the object 100, and a part of the laser that has reciprocated by the light receiving distance L0 and returned is received by the light receiving unit 51b.

As the emitted light, laser, infrared light, visible light, ultrasonic wave, electromagnetic wave, or the like can be used, but a laser is preferably used because distance measurement must be sufficiently possible even at night.
Further, LIDAR (Light Detection and Ranging or Laser Imaging Detection and Ranging: lidar) is used as a distance detection sensor today, but this may be used as the distance detection unit 51.
LIDAR is a device that emits a laser to a two-dimensional space or a three-dimensional space in a predetermined obstacle determination area and measures distances at a plurality of measurement points in the obstacle determination area.
In addition, the LIDAR detects the reflected light reflected by the object with the light receiving unit 51b after emitting the laser from the light emitting unit 51a, and calculates the light receiving distance L0 from, for example, the time difference between the emission time and the light receiving time. This light receiving distance L0 corresponds to measured distance information 72 described later.

Assuming that the laser emitted from the light emitting unit 51a hits an immovable object separated by a distance L0, the laser travels a distance (2L0) corresponding to twice the distance L0 from the tip of the light emitting unit 51a to the object surface, The light is received by the light receiving unit 51b.
The time when the laser is emitted and the time when the laser is received are shifted by the time T0 required for the laser to travel the distance (2L0). That is, a time difference occurs. The light receiving distance L0 can be calculated by using the time difference T0 and the speed of light.
Further, the distance to the object (obstacle) is detected from the calculated light receiving distance L0.

FIG. 4 shows a case where the distance detection unit 51 is not moved, and shows a case where the laser emitted from the light emitting unit 51a travels on the same optical path.
Therefore, when the light reflected on one point of the object 100 is received, only the distance between the tip of the light emitting unit 51a and one point of the object is calculated.

The scanning control unit 51c is a unit that scans the light emission direction so that the light is emitted toward a plurality of predetermined measurement points in the obstacle determination area in the space ahead of the traveling direction. Is gradually changed at regular intervals, thereby gradually moving the optical path along which the emitted laser travels.
The LIDAR 51 calculates the distance to the object (horizontal two-dimensional scanning) by changing the laser emission direction by a predetermined scanning pitch within a predetermined horizontal two-dimensional space. When the distance is calculated three-dimensionally, the emission direction of the laser is changed by a predetermined scanning pitch in the vertical direction, and the above-described two-dimensional scanning in the horizontal direction is performed to calculate the distance.

FIG. 5 is a schematic explanatory diagram of the scanning direction of the laser beam emitted from the distance detection unit (LIDAR) 51.
FIG. 6 shows a view of the irradiation area of the laser emitted from the distance detection unit (LIDAR) 51 when viewed from above (FIG. 6A) and a view when viewed from behind (FIG. 6B). Show.
In FIG. 5, one point indicates a position (hereinafter, referred to as a measuring point) where the laser hits on a two-dimensional plane (vertical plane) in the vertical direction at a position separated by a predetermined distance.

For example, if the direction of the distance detection unit 51 is changed so that the emission direction of the laser beam emitted from the light emitting unit 51a of the distance detection unit 51 moves to the right by a predetermined scanning pitch in the horizontal direction, the laser moves in the horizontal direction. Of the next position (measurement point) shifted to the right by the scanning pitch.
If an object exists at the position of this vertical plane, a part of the reflected light of the laser reflected at each measurement point is received by the light receiving unit 51b.
As described above, when the irradiation direction of the laser is sequentially shifted by a predetermined scanning pitch in the horizontal direction, the laser is irradiated to a predetermined number of measurement points. For each of the plurality of measurement points irradiated with the laser, the presence or absence of reception of the reflected light is checked to calculate the distance.

FIG. 6A is an explanatory diagram illustrating an example in which the laser irradiation direction is shifted in the horizontal direction by a scanning pitch and the laser is scanned in the left-right direction (that is, the horizontal direction).
For example, as shown in FIG. 6A, when a laser beam is irradiated in the rightmost direction, if an object exists in that direction, the light receiving distance L0 is calculated by receiving the reflected light from the object. You.

As shown in FIG. 5, when the laser scanning direction is the vertical direction, for example, when the laser emitting direction is shifted upward by a predetermined scanning pitch in the vertical direction, the laser The vertical plane of the next position (measurement point) shifted by the scanning pitch in the upward direction in the direction.
After shifting the emission direction of the laser upward by one scanning pitch in the vertical direction, as shown in FIG. The measurement point at a position shifted by the scanning pitch is irradiated with a laser.
As described above, by sequentially performing the horizontal laser scanning and the vertical laser scanning, a predetermined three-dimensional space is irradiated with the laser, and if an object exists in the three-dimensional measurement space, the laser is irradiated. The distance to the object is calculated.

In addition, when light (laser) emitted toward a plurality of measurement points is reflected by an object, if it is confirmed that the light reflected by the object is received by the light receiving unit, the distance is calculated. It is determined that a part of the object exists at the position of the measured point.
Furthermore, the object is present in an area including a plurality of measurement points determined to include a part of the object, and the information of the area including the plurality of measurement points is used to characterize the shape of the object or the posture of the human body. Get detection information.
The detection information is information characterizing some object, but may be obtained by the distance detection unit 51 or may be obtained from image data of the object taken by the camera 55.

In the two-dimensional scanning, the direction in which the laser is scanned is described as the horizontal direction. However, the direction is not limited to this, and the direction in which the laser is emitted may be changed in the vertical direction.
When irradiating a laser to a three-dimensional measurement space, the same two-dimensional scanning in the vertical direction may be performed sequentially after shifting two-dimensionally in the vertical direction by a predetermined scanning pitch in the horizontal direction.

FIG. 6B is a schematic explanatory diagram of the measurement points of the laser irradiated on the three-dimensional space when the laser is scanned in the horizontal direction and the vertical direction.
If there is no object in the direction of one measuring point from which the laser is emitted, the laser proceeds on the optical path as it is, no reflected light is received, and the distance cannot be measured.
Conversely, when reflected light is received with respect to the laser emitted to a certain measurement point, the distance is calculated, and it is recognized that an object exists at a position separated by the calculated distance.
FIG. 6B shows that reflected light has been detected at the six measurement points in the lower right part, and an object (for example, a human body or an obstacle) is included in an area including the six measurement points. Is recognized to exist.

Further, when a predetermined number or more (for example, 10 or more) of the measurement points whose distances are measured are detected in the detection area of the object including the plurality of measurement points, the area within the measurement point is detected. It is determined that there is an object in.
However, if the number of measuring points whose distance has been measured is smaller than a predetermined number, or if the distance cannot be measured at a measuring point around one measuring point whose distance has been measured, the vicinity of that measuring point will not be measured. Has a high possibility that no object exists, and the distance measurement at the measurement point is determined to be erroneous.

When counting the number of measuring points whose distance has been measured, in principle, one measuring point is counted as one. For example, in a predetermined detection area, the distance is detected at ten measuring points. In this case, the number of measurement points in the area is counted as 10.
However, as will be described later, if the distance is detected at N measurement points in the weighted area in the determination area for detecting the object, the number of measurement points is not counted as N. , And the number considering the weight.
For example, if the weight is set to twice the actual number of measurement points, and if the number of measurement points whose distance is detected in the weighted area is 10, then the number of measurement points in that area is: Count as 10 × 2 = 20.

  In particular, in the present invention, the weight of a part of the object detection area to which the traveling direction of the vehicle belongs (a divided area to be described later) is set to be larger than that of the other areas, which may have a great effect on the running of the vehicle. Make sure that objects that exist in a certain traveling direction are detected.Conversely, even if an object exists in a low-weighted area that is considered to have little effect on the traveling of the vehicle, normal traveling To be continued.

When the laser 51d is incident on the light receiving section 51b of the distance detecting section 51, an electric signal corresponding to the received light intensity of the laser is output.
The control unit 50 checks the electric signal output from the light receiving unit 51b, and determines that the laser has been received, for example, when an electric signal having an intensity equal to or higher than a predetermined threshold is detected.
A conventionally used laser light emitting element is used for the light emitting section 51a, and a laser light receiving element for detecting a laser is used for the light receiving section 51b.

Further, the control unit 50 uses the time difference T0 between the emission time of the laser emitted from the light emitting unit 51a and the light receiving time at which it has been confirmed that the reflected light has been received by the light receiving unit 51b. A light receiving distance L0 that is a distance between a plurality of measurement points is calculated.
The control unit 50 obtains the current time using, for example, a timer, calculates the time difference T0 between the laser emission time and the light reception time at which the laser light reception is confirmed, and calculates the time difference T0 between the two times and the laser time. Is used to calculate the light receiving distance L0.

The traveling control unit 52 is a part that controls a driving member that causes the vehicle body to travel. The traveling control unit 52 mainly controls the rotation of the wheels 53 corresponding to the driving member to perform a straight traveling, a rotating operation, and the like. To run. The driving member includes a wheel, a caterpillar (registered trademark), and the like.
The wheels 53 correspond to four wheels (21, 22, 31, 32) as shown in FIGS.
Further, as described above, of the wheels, the left and right front wheels (21, 31) may be used as drive wheels, and the left and right rear wheels (22, 32) may be driven wheels that do not perform rotation control.
Further, encoders (not shown) are provided for the left and right wheels of the drive wheels (21, 31), respectively, and the traveling distance and the like of the vehicle are measured based on the number of rotations, the rotation direction, the rotation position, and the rotation speed of the wheels, and the traveling is controlled. You may. The encoder corresponds to a speed detection unit 60 described later.

The communication unit 54 is a part that transmits and receives data to and from the management server 5 via the network 6. As described above, it is preferable to have a function of connecting to the network 6 by wireless communication and communicating with the management server 5.
For example, when the abnormal state occurs and the notification process is executed, the communication unit 54 places the notification information including the occurrence of the abnormal state, the date and time of occurrence of the abnormal state, and the occurrence position in a position different from the autonomous traveling device. To the management server 5.
Further, the notification information may be transmitted to a terminal owned by a person in charge at a position different from the autonomous traveling device, or may be transmitted to at least one or both of the management server and the terminal.
Although it is necessary to set in advance where the transmission destination is, it may be possible to change or add the transmission destination based on the content of the abnormal state or the like in accordance with the operation mode of the vehicle. .

The camera 55 mainly captures an image of a predetermined space including a space in front of the vehicle in the traveling direction, and the captured image may be a still image or a moving image. The captured image is stored in the storage unit 70 as input image data 71, and is transferred to the management server 5 in response to a request from the management server 5.
Further, not only one camera 55 but also a plurality of cameras 55 may be provided. For example, four cameras may be fixedly installed so as to shoot the front, left, right, and rear of the vehicle body, respectively, and the shooting direction of each camera may be changed. May be provided.
When the vehicle travels outdoors and the weather is good and the shooting area is sufficiently bright, an image taken by the camera is analyzed to detect a human body, an obstacle, a state of a road surface, and the like.

The image recognition unit 56 is a unit that recognizes an object included in image data (input image data 71) captured by the camera 55. For example, an object included in image data is extracted, and when the extracted object is an object having predetermined characteristics of a human body, the object is recognized as a human body. Further, the image data (human body image) of the recognized human body part is compared with the person registration information stored in advance in the storage unit 70 to determine whether or not the human body image can match a previously registered person. . The image recognition processing may use an existing image recognition technology.
The object to be recognized is not limited to the human body, and may recognize an obstacle such as a wall, a pillar, a step, an animal, or a narrow passage.

The obstacle detection unit 57 is a unit that mainly detects an object (obstacle) using information acquired from the distance detection unit 51. In particular, the position of the object whose distance has been detected by the distance detection unit 51 in the obstacle determination region and the direction of the position where the object exists with respect to the traveling direction are detected.
For example, the distance detection unit 51 detects that an obstacle exists at the position of the measurement point where the reflected light is received and the distance is calculated.
Further, as described above, since the distances to the plurality of measurement points are calculated, the size, position, shape, and distance to the obstacle are obtained from the position information of the calculated measurement points. You.

Further, assuming that the scanning direction of the laser at the center of the laser scanning direction shown in FIG. 6A is the traveling direction of the vehicle, the position of the measuring point where the obstacle is detected and the scanning direction of the laser are determined. From the relationship, it can be determined whether the detected obstacle is on the right side, the left side, or just on the traveling direction with respect to the traveling direction.
Further, as a reference for determining the direction in which the obstacle exists, the traveling direction of the vehicle may be set to zero degree, and the angle of the position where the obstacle exists may be calculated. That is, the direction in which the obstacle exists with respect to the traveling direction can be detected.

Further, for each measurement point at a different position in the laser scanning area, an interval between the plurality of measurement points (inter-point distance Ws) is calculated in addition to the position, the distance, and the direction with respect to the traveling direction.
Information about these detected measuring points (for example, measuring point number, distance to the measuring point, position, direction, distance between the measuring points) is stored in the storage unit 70 as obstacle information 77 as described later. Is done.
However, since the obstacle information 77 is always obtained even while the vehicle is running, it is updated every predetermined time. In addition, information (a size, a shape, a position, a distance, a direction, and the like) of an obstacle existing at the positions of the measurement points is generated from the information of the plurality of measurement points.

  By using the distance to each measurement point detected as described above, the direction of the measurement point or obstacle with respect to the traveling direction, the number of measurement points in a predetermined area, and the like, the change of the traveling direction of the vehicle and the speed control are performed. Is performed, and if it is determined that there is no need to decelerate even if an obstacle is found in the traveling direction, the vehicle continues traveling without decelerating.

The position information acquisition unit 58 is a part that acquires information (latitude, longitude, and the like) indicating the current position of the vehicle. For example, the position information acquisition unit 58 may acquire the current position information 73 using a GPS (Global Position System). .
While comparing the acquired current position information 73 with the route information 74 stored in the storage unit 70 in advance, a direction in which the vehicle should travel is determined, and the vehicle is made to travel autonomously.
In order for the vehicle to travel autonomously, it is preferable to use information obtained from all of the above-described distance detection unit 51, camera 55, obstacle detection unit 57, and position information acquisition unit 58, or from at least one of them. Autonomous driving may be performed using the obtained information.

  Further, as the position information acquisition unit 58, similarly to the GPS, another satellite positioning system currently used may be used. For example, using the Quasi-Zenith Satellite System (QZSS) in Japan, GLONASS (Global Navigation Satellite System) in Russia, Galileo in EU, Hokuto in China, IRNSS (Indian Regional Navigational Satellite System) in India, etc. Is also good.

The rechargeable battery 59 is a part that supplies electric power to each functional element of the vehicle 1, and mainly supplies electric power for performing a traveling function, a distance detection function, an image recognition function, a collision detection function, and a communication function. It is.
For example, rechargeable batteries such as lithium ion batteries, nickel hydride batteries, Ni-Cd batteries, lead batteries, and various fuel cells are used.
In addition, a battery remaining amount detection unit (not shown) is provided to detect the remaining capacity (battery remaining amount) of the rechargeable battery and determine whether to return to a predetermined charging facility based on the detected remaining battery amount. Is determined, and when the remaining battery level becomes lower than the predetermined remaining level, the control unit may automatically return to the charging facility.

  The speed detector 60 is a part that detects the speed of the vehicle 1 during traveling, and mainly corresponds to an encoder attached to the wheels 53. The CPU acquires speed information from the speed detection unit 60 in real time, and controls the travel control unit 52 to accelerate or decelerate according to the situation.

  The monitoring information acquisition unit 61 is a unit that acquires information on a predetermined monitoring target. For example, the monitoring information acquisition unit 61 acquires information collected by a vehicle autonomously traveling in a predetermined area and information on a traveling state of the vehicle, and stores the information in a storage unit 70. Is stored as transmission monitoring information 75. As a device corresponding to the monitoring information acquisition unit 61, for example, a thermometer, a hygrometer, a microphone, a gas detection device, and the like may be provided.

The inertia measurement unit 62 is a part that measures the acceleration and angular velocity with respect to the three-dimensional space when the vehicle travels. This is a part for detecting the position, the inclination, the vibration, the current traveling direction of the vehicle body, and the like.
The inertia measurement unit 62 mainly measures acceleration and angular velocity with respect to three axes of the vehicle in a vertical direction, a horizontal direction, and a front-rear direction. If, inertial measurement units containing any one or more sensors and a magnetic sensor that measures the magnitude and direction of geomagnetism: a (IMU inertial measurement unit).
By using the magnitude and direction of the measured acceleration and angular velocity, the traveling direction of the vehicle can be detected, and when the vehicle slides in a direction different from the traveling direction during traveling, the vehicle has slipped. The direction can be detected.

The region weight calculation unit 63 sets the weight (also referred to as a region weight value or a weight value) of a determination region for detecting an obstacle based on information on a traveling state of the vehicle 1 and a traveling direction in which the vehicle 1 is going to travel. This is the part to do.
For example, of all the determination areas for detecting an obstacle, a part of an area having a predetermined width in the traveling direction in which the vehicle is about to travel will be used as a criterion for performing obstacle detection. High weight (for example, double weighting) is given.
As will be described later, the set weight is assigned to each divided region obtained by dividing the obstacle determination region into several regions.
Also, the area weight value assigned to each divided area is given a different value depending on the traveling state of the vehicle, for example, when traveling straight, when turning, when detecting side skid, or when changing the course to the left or right, etc. A different value is assigned to each divided area in advance.

What region weight value is assigned to each divided region in accordance with the running state of the vehicle may be stored in the storage unit 70 in advance. One example of the region weight value 79 is shown in FIGS. 7 and 8 described later.
For example, when the vehicle is going to travel straight ahead, the area weight value 79 set in advance as the weight when traveling straight ahead may be read from the storage unit 70 and set as the current weight of each divided area.
When the inertial measuring unit 62 detects that the vehicle has skid to the right, the weight value 79 of the right skid stored in the storage unit 70 is read in advance, and the current weighting of each divided area is performed. It should just be set as.
According to the present invention, the level of the weight is set not based on the distance from the front of the vehicle, but based on the traveling direction of the vehicle.
Although the area weight value 79 of each divided area may be fixedly stored in advance, an appropriate weight using a predetermined calculation formula based on the actual running speed, acceleration, angular velocity, etc. of the vehicle is used. The value may be set each time.

In addition to the above components, the vehicle 1 may include a collision detection unit that detects that the vehicle 1 has collided with, contacted, or approached an obstacle while traveling. For example, a contact sensor or a non-contact sensor including a pressure-sensitive switch, a micro switch, an ultrasonic sensor, and an infrared distance measuring sensor is used, and is disposed, for example, on a bumper of a vehicle body.
Although one collision detection unit may be provided, it is preferable to provide a plurality of collision detection units at predetermined positions in front, side, and rear of the vehicle body in order to detect a collision from the front, side, and rear of the vehicle body.
For example, it is preferable that a plurality of collision detection units be mounted between the elastic member constituting the bumper and the vehicle body at a predetermined distance from each other.
If the collision detection unit detects a collision with an obstacle after the traveling control unit 52 executes the vehicle deceleration process, the traveling control unit 52 stops the vehicle in order to prevent the vehicle body from being damaged. Execute the processing to be performed. Further, the CPU may recognize the position where the obstacle exists based on the signal output from the collision detection unit. Further, based on the recognized position information of the obstacle, the CPU stops the vehicle body or determines a direction in which the vehicle should travel next while avoiding the obstacle.

  The storage unit 70 is a unit that stores information and programs necessary for executing each function of the autonomous traveling device 1, and includes semiconductor storage elements such as ROM, RAM, and flash memory, storage devices such as HDD and SSD, and others. Is used. In the storage unit 70, for example, input image data 71, measured distance information 72, current position information 73, route information 74, transmission monitoring information 75, set speed information 76, obstacle information 77, determination area information 78, area weight value 79 and the like are stored.

  The input image data 71 is image data captured by the camera 55. When there are a plurality of cameras, the information is stored for each camera. The image data may be either a still image or a moving image. The image data is used for detection of a suspicious person, detection of an abnormal state, determination of a course of a vehicle, and the like, and is transmitted to the management server 5 as one piece of transmission monitoring information 75.

The measurement distance information 72 is the light receiving distance L0 calculated based on the information acquired from the distance detection unit 51 as described above. One light receiving distance L0 means a distance measured at one measurement point in a predetermined distance measurement area.
The information 72 is stored for each measurement point belonging to a predetermined distance measurement area, and is stored in association with the position information of each measurement point. For example, when there are m measurement points in the horizontal direction and n measurement points in the vertical direction, the light receiving distances L0 respectively corresponding to a total of m × n measurement points are stored.

  If there is an object (obstacle, road surface, pillar, etc.) that reflects the laser in the direction of each measurement point and the reflected light from the object can be received, the light receiving distance L0 to the object is stored. You. However, when no object is present in the measurement point direction, the reflected light is not received. Therefore, instead of the light reception distance L0, for example, information indicating that measurement could not be performed may be stored as the measurement point distance information 72. .

  The current position information 73 is information indicating the current position of the vehicle acquired by the position information acquisition unit 58. For example, the information is information including latitude and longitude acquired using GPS. This information is used, for example, to determine the course of the vehicle.

  The route information 74 is information in which a traveling route of the vehicle body is determined in advance, and stores a map of a route to be traveled in advance. For example, when a moving route or area is fixedly determined in advance, the information is stored as fixed information from the beginning. However, when it is necessary to change the route, information transmitted from the management server 5 via the network 6 may be stored as new route information 74.

  The transmission monitoring information 75 is information of various monitoring targets acquired by the monitoring information acquisition unit 61 during traveling and while stopped, and is information transmitted to the management server 5 via the network 6. This information includes, for example, input image data 71 taken by the camera 55, travel distance, travel route, environmental data (temperature, humidity, radiation, gas, rainfall, voice, ultraviolet rays, etc.), terrain data, obstacle data, It includes road surface information, warning information, and the like.

The set speed information 76 corresponds to the running speed of the vehicle, and the current running speed is set based on this information.
For example, when traveling on a route without obstacles, a relatively high speed is set in the set speed information 76, and the traveling control unit 52 sets the wheels so that the vehicle travels at the speed set in the set speed information 76. 53 is controlled.
If it is determined that the vehicle needs to be decelerated by detecting an obstacle or a step, a relatively slow speed is set in the set speed information 76.

  The obstacle information 77 is information on each measurement point and the detected obstacle, and is based on the distance from the current position to the obstacle, the shape, position, direction, size, color, height difference, inclination angle, and the like of the obstacle. Information. For example, the shape and size of the obstacle and the distance to the obstacle are acquired by the distance detection unit 51 and stored as a part of the obstacle information 77. Information for identifying an obstacle is also obtained by the camera 55, the image recognition unit 56, and the obstacle detection unit 57.

FIG. 7A shows an example of information on each measurement point in the obstacle information 77.
Here, for each measuring point, a measuring point number, a distance to the measuring point, a position, a direction, a distance Ws between the measuring points, and the like are stored. These pieces of information are acquired while traveling autonomously, and thus change every moment.
The position of the measurement point is stored, for example, in two-dimensional or three-dimensional spatial coordinates, and the direction may be stored as an angle with respect to the traveling direction, or as information such as left or right with respect to the traveling direction. Is also good.
The distance Ws between the measurement points means a distance between the measurement points at which the distances are measured. Since the coordinates of each measurement point are known in advance, they can be obtained by a predetermined distance calculation. The distance between adjacent measurement points is constant and relatively short.
For a plurality of measurement points whose distances have been measured, only relatively short measurement point distances Ws have been calculated, and the plurality of measurement points are concentrated in an area having a predetermined spread. It is determined that an object exists in the area.

However, for a plurality of measurement points whose distances have been measured, there are a number of relatively long measurement point distances Ws, and the plurality of measurement points do not exist densely in one region, If the measurement points are dispersed at a predetermined distance or more, there is little possibility that an object is present at each measurement point, and there is a possibility of erroneous detection, In addition, it is considered that there is an object that does not cause any problem even when traveling on it.
Therefore, as described later, when a plurality of measurement points are concentrated within a predetermined range, there is a high possibility that an object exists within the range, and the weight of the divided region including those measurement points is weighted. Set high.
Conversely, when a plurality of measurement points are dispersed at a predetermined distance or more, it is considered that the possibility of erroneous detection is high, and the weight of the divided region including those measurement points is set to be low. .
By setting such weights, the accuracy of object detection can be improved and unnecessary deceleration processing can be reduced.

The determination area information 78 means information of an area for detecting an object such as an obstacle (hereinafter, referred to as an obstacle determination area or a determination area), and the position of each divided area obtained by dividing the obstacle determination area into a plurality of pieces. And information on the size. The information is mainly about an area where an object can be detected by LIDAR corresponding to the distance detection unit 51.
For example, as shown in FIGS. 5 and 6 described above, by scanning the laser two-dimensionally or three-dimensionally, the three-dimensional space ahead in the traveling direction becomes a determination area for detecting an obstacle.

However, there is a limit to the distance over which the reflected light can be detected depending on the performance of the emitted laser, so a fan-shaped area within a predetermined detection distance L is the determination area.
When the light is scanned in the horizontal direction with respect to the ground around the light emission position, that is, the position of the light emitting unit 51a, and a horizontal fan-shaped area in which the light can reach is defined as an obstacle determination area, The divided area is an area obtained by dividing the obstacle determination area into a plurality of sector shapes having a predetermined central angle.
In addition, as the determination area, a fan-shaped area within 180 degrees in front of the traveling direction with the traveling direction as the center may be set in advance, but in consideration of stationary turning, direction change, and the like, a portion behind the traveling direction is considered. A region exceeding 180 degrees including the region may be set as the determination region.

FIG. 9 shows an embodiment of the determination area.
FIG. 9A shows a determination area 78a having a fan shape centered on the position of the LIDAR arranged on the front of the vehicle 1 and opened at an angle of 270 degrees.
The determination area 78a has a fan shape opened by an angle of 135 degrees left and right with respect to the traveling direction of the vehicle 1, and an area opened 45 degrees behind the side of the vehicle 1 is also a determination area. .
In FIG. 9A, the determination area 78a is divided into nine areas (A1 to A9).
Each of the divided regions is a sector-shaped region having a central angle (also referred to as a detection angle) of 30 degrees, and a length corresponding to a radius of each sector corresponds to a detection distance L.
When an object exists in any of the nine divided regions, the distance to the object and the position are detected.

Also, in principle, as in the conventional case, if the distance to the detected object is shorter than a predetermined stop determination distance, the vehicle is stopped, and if shorter than the predetermined deceleration determination distance, the vehicle speed is reduced. .
However, according to the present invention, a weight (importance) is given to each divided region of the determination region based on the traveling direction of the vehicle, the current traveling state, the position of the detected object, and the like, and the object is detected. The weight of the divided area to which the measurement point belongs is added to the number of measurement points, the importance of the detected object is determined, and the speed control of the vehicle is performed.
For example, even if an object is detected in the determination area, if the divided area determined to have the object does not hinder the progress of the vehicle and is a low-weight area, even if the object is Even if it is within the deceleration determination distance, the vehicle continues traveling without performing deceleration processing.

  Further, for example, in FIG. 9A, when an object is detected in the divided region A2 when the direction of the divided region A5 of the determination region is the traveling direction of the vehicle, the vehicle temporarily stops and turns rightward. In this case, the weights of the divided areas A1, A2, and A3 in the right direction of the vehicle are set to be larger than those of the other divided areas, and the distance to the object existing in the divided area A2 is set to the divided area A2. Even if it is longer than the deceleration determination distance in the above, deceleration processing is performed.

FIG. 7B shows an embodiment of the determination area information 78.
Here, information on the nine divided areas (A1 to A9) shown in FIG. 9 is shown.
Information such as a position and coordinates for specifying the divided area is stored in advance for each divided area.
For example, for a divided area having an area number of A1, a determination width W1 for specifying the sector shape, a detection angle α1, and an outline of a sector shape whose origin is the position of the light emitting unit 51a of the distance measurement unit 51 are shown. XY coordinate values and the like are stored in advance.
The determination width W1 corresponds to the length of the fan-shaped lateral width as shown in FIG. 10C described later, and a value obtained by adding a predetermined margin width ma to the vehicle width W0 is set in advance. .
Similarly, information for specifying the other divided areas (A2 to A9) is set in advance.

Each divided area included in the determination area 78a may be fixedly set to the same size and shape in advance, but is not limited thereto.
The size of each divided region may be changed in real time while taking into consideration the traveling speed, traveling direction, possible stopping distance, and the like of the vehicle.
Further, the number of divided areas is not limited to nine, and the angle of the determination area 78a is not limited to 270 degrees.
However, in the following embodiments, the size and shape of each divided area are assumed to be the same, and the number of divided areas is also set to 9 for easy explanation.

The area weight value 79 means a weight value of each divided area included in the determination area for detecting the presence or absence of an obstacle, and indicates a degree of importance of detecting an obstacle of each divided area set based on the traveling state of the vehicle body. Show. The area weight value of each divided area is set according to the traveling state of the vehicle.
Based on the current traveling direction detected by the inertia measuring unit 62 or the traveling direction to be traveled on the traveling route defined in the route information 74, the control unit 50 controls the vehicle body for each divided area based on the traveling direction. An area weight value is set in consideration of the subsequent traveling direction.
For example, as shown in FIG. 11, which will be described later, a relatively large area weight value is set for a divided area A5 including the traveling direction of the vehicle body and some divided areas near the divided area A5.

Further, in the present invention, the presence of an object is determined based on the magnitude of the area weight value set for each divided area, and subsequent deceleration processing or stop processing of the vehicle body is performed.
Mainly, when it is determined that an object whose distance is detected is present in a divided area in which a relatively large area weight value is set, a deceleration process or a stop process of the vehicle body is performed in accordance with the set area weight value. .
On the other hand, when it is determined that no object exists in the divided area in which the relatively large area weight value is set, the deceleration processing and the stop processing of the vehicle body are not performed.
Also, when it is determined that an object whose distance has been detected exists in a divided region other than the divided region in which the relatively large region weight value is set, the deceleration process and the stop process of the vehicle body are not performed.
Thus, for example, even if an object exists near the vehicle body, if the object exists in a divided area where a relatively large area weight is not set, the vehicle continues traveling without performing deceleration processing or the like of the vehicle body. be able to.

The above-mentioned divided regions to which relatively large region weight values are set include a divided region A including the traveling direction of the vehicle body and a divided region which is both of the divided regions A when the vehicle is going straight ahead. Is preferred.
Alternatively, the largest region weight value may be set for a divided region including the traveling direction of the vehicle body, and a smaller region weight value may be set for other divided regions as the distance from the traveling direction increases.
Thus, when an object exists in the traveling direction, deceleration processing or the like can be reliably performed. When an object exists in a divided region that does not hinder the advancement of the vehicle body, unnecessary deceleration or stopping is performed. , It is possible to continue running quickly.

FIGS. 7C and 8A show an example of the set value of the area weight value.
Here, as shown in FIG. 9A, it is assumed that the determination region 78a is divided into nine, and the weight value Mn is one of three levels (M1, M2, M3) for each divided region An. Is set.
The three weight values Mn have the highest weight of M1 and the lowest weight of M3, and satisfy M1>M2> M3.
As a specific example of the weight value Mn, a magnification by which the number of measuring points whose distances are detected in the divided area may be employed. For example, M1 means 3 times, and M2 means 2 times. 2, 0.5 which means 0.5 times may be set in advance as M3.

For example, if the weight value of a certain divided area A3 is M1, and there are ten measuring points whose distances are detected among the measuring points in the divided area A3, the weight after weighting is applied to the divided area A3. The number of measurement points is 10 × 3 = 30.
Further, when the weight value of the divided area A6 is M3 and the number of the measuring points whose distance is detected in the divided area A6 is 8, the number of the measuring points after weighting in the divided area A6 is 8 × 0.5 = 4.
It is determined that the greater the number of measurement points after weighting, the higher the possibility of the presence of the object, and the smaller the number of measurement points after weighting, the lower the possibility of the presence of the object.
However, the presence of an object is actually determined comprehensively using not only the number of measurement points, but also the distance from the vehicle, the direction to the traveling direction, the position of the measurement points, camera images, ultrasonic sensors, etc. Is done.

FIG. 7 (c) shows an example of the weight value Mn when the vehicle is going straight ahead and the weight value Mn when the vehicle is going to turn right and left when turning straight ahead. I have.
FIG. 8A shows a case where the vehicle is going to proceed in the right licking direction from the straight traveling direction, a case where the vehicle is going to proceed in the left licking direction, and a case where the vehicle has slipped to the right by the inertia measuring unit 62. FIG. 6 shows an example of the weight value Mn in the case of detecting the vehicle slippage and in the case of detecting that the vehicle has slipped to the left.

FIG. 10 is an explanatory diagram of one embodiment of one divided area of the determination area shown in FIG.
FIG. 10A shows a divided area A5 in the traveling direction of the vehicle 1 when traveling straight.
Here, the divided area A5 is a fan-shaped area where the detection distance of the obstacle is L and the detection angle is α.
Further, the detection angle α of the divided area A5 is an angle opened by α / 2 in the left and right directions with the traveling direction as the center.
The detection distance L is a distance at which the reflected light can be detected, and is determined by the intensity of the laser light emitted from the light emitting unit 51a, and is, for example, about 25 m.
Further, assuming that the determination area 78a including nine divided areas (A1 to A9) has an angle of 270 degrees, the detection angle α of one divided area is 30 degrees.

FIG. 10B shows an example in which one divided area A5 is classified into three according to the distance from the vehicle. The range within the shortest distance L1 from the front of the vehicle is defined as the stop area a1, the area outside the stop area a1 and within the next short distance L2 (> L1) is defined as the deceleration area a2, and the area outside the deceleration area a2. Therefore, the range within the detection distance L (> L2) is referred to as a normal area a3.
It is preferable that these distances L1 and L2 be set to predetermined values in advance, but the setting may be changed each time in relation to the running speed of the vehicle. For example, when traveling at a relatively low speed, the distances L1 and L2 may be set relatively short.

The stop area a1 is an area where the vehicle stops when an obstacle is detected in this area, for example, L1 is about 1 m.
The deceleration area a2 is an area for reducing the traveling speed of the vehicle when an obstacle is detected in this area, and is, for example, about L2 = 5 m.
The amount of deceleration to be performed is determined by calculating the maximum braking distance of the vehicle body on a frozen road surface, for example, the stopping distance when running at 5 km / h. Will be about 3m, so decelerate to about 2.5km / h from a distance of 5m in front of it. When the speed is reduced to 2.5km / h, the stopping distance is about 1m. The normal area a3 is an area in which normal traveling is continued even when an obstacle is detected in this area, and travels at the same speed as the traveling speed immediately before detecting the obstacle in principle.

Also, the three areas (a1, a2, a3) shown in FIG. 10B show fixed areas when weighting is not performed. The size of each area may be changed correspondingly.
For example, when the largest weight value M1 is set for the divided area, the distance L1 defining the stop area a1 and the distance L2 defining the deceleration area a2 are changed to longer values, and the sizes of the stop area a1 and the deceleration area a2 are changed. And the size of the normal area a3 may be reduced.
Thus, when an obstacle is detected in the divided area in which the large weight value M1 is set, the vehicle can be more safely and reliably decelerated and stopped.

Conversely, when the smallest weight value M3 is set in the divided area, the distances L1 and L2 are changed to shorter values, the sizes of the stop area a1 and the deceleration area a2 are reduced, and the size of the normal area a3 is reduced. May be increased.
As a result, even when an obstacle is detected in the divided area, it is possible to increase the possibility that the vehicle can run at the normal speed by increasing the range of the normal area a3. Unless an object is detected, unnecessary deceleration processing and stop processing can be prevented.

FIG. 10C shows another embodiment of the divided area.
9 and 10 (a) show the case where the determination area 78a is equally divided into a plurality of divided areas, but here, the size of one divided area is determined in consideration of the width of the vehicle. An example is shown.
In FIG. 10C, the width of the vehicle is set to W0, the detection distance is set to L, and both are set to predetermined constant values.
Further, a predetermined margin width ma is set in the left-right direction of the vehicle 1, and the width W1 of the divided region in a direction perpendicular to the traveling direction of the vehicle is determined.
Here, it is assumed that W1 = W0 + 2ma.
When the width W1 of the divided area is determined, the detection angle α is also determined by a predetermined calculation formula.

Further, the margin width ma is the length of the vehicle body width W0, the magnitude of the estimated swing width in the case where the vehicle is likely to swing in the left-right direction while traveling straight in the traveling direction, the magnitude of the traveling road surface. The expected magnitude of the sliding width when the wheel may slip in the left and right direction depending on the condition, and the expected movement in the left and right direction when the vehicle may move in the left and right direction when the vehicle suddenly stops. It is set in advance in consideration of the size of the distance, the traveling speed of the vehicle, the manner of turning, and the like. Alternatively, the margin width ma may be changed each time based on the actual driving situation, road surface condition, weather, road surface inclination, and the like.
For example, when the vehicle width W0 is 0.8 m, the margin width ma may be set to about 0.3 m, and the width W1 of the divided area A5 in the direction in which the vehicle body goes straight may be set to about 1.4 m.

Hereinafter, a description will be given of a setting example of the area weight value set corresponding to the traveling state and some embodiments of the obstacle detection.
(Example 1)
Here, an example of an area weight value set for a divided area when traveling straight and obstacle detection will be described.
FIG. 11 illustrates the weight values Mn at the time of straight traveling shown in FIG. 7C for nine divided regions (A1 to A9).
Assuming that the traveling direction of the vehicle 1 when traveling straight is the direction of the divided area A5, when an obstacle is found in the area A5 when traveling straight, it is most necessary to avoid a collision. Is the highest M1.

Also, when traveling straight, the weight value of two divided areas (A4, A6) adjacent to the divided area A5 is set to M2, which is a lower value than the area A5.
This is because, when a moving object is detected in the divided areas A4 and A6 when traveling straight, there is a high possibility that the vehicle will collide with the vehicle.
Further, when the vehicle is traveling straight in the direction of the area A5, it is considered that the possibility of collision is low even if an obstacle exists in the other divided areas (A1, A2, A3, A7, A8, A9). . Therefore, the weight value of these divided areas (A1, A2, A3, A7, A8, A9) is set to the lowest value M3.
By setting such weight values, when an obstacle is present in a divided region having a high weight value, the number of measurement points whose distances have been detected is multiplied by a high magnification. Its presence will be more reliably detected.

For example, as shown in FIG. 9B, when the vehicle is traveling straight in the direction of the area A5, if the distance is detected for the measurement points of a part of the areas A5 and A6, the measurement points of the areas A5 and A6 Are multiplied by the weight values (M1, M2) to calculate the number of measurement points.
Since the weight value M1 of the area A5 is the highest and the weight value M2 of the area A6 is the next highest, the number of measurement points larger than the actual number of measurement points is calculated, and the calculated number of measurement points is determined to be an obstacle. If the number of measurement points exceeds a predetermined judgment measurement number, it is more reliably determined that the obstacle OB1 exists near the boundary between the areas A5 and A6.
Further, since the distance L0 to the obstacle OB1 is also calculated by the distance detection unit 51, if the distance L0 is shorter than a predetermined deceleration determination distance, a predetermined deceleration process is performed.

On the other hand, as shown in FIG. 11B, it is assumed that an obstacle OB1 is detected in the divided area A8 in which the area weight value is set to M3 when traveling straight in the area A5 direction. In this case, a fixed number of measuring points whose distance has been detected are detected in the area A8. If the weight value M3 is 0.5, the number of measuring points considering the weight value is equal to the actual number of measuring points. Calculated as half.
Therefore, when the number of measurement points calculated in the area A8 is smaller than the number of determination points at which an obstacle is determined to exist, even if the actual number of measurement points in the area A8 is larger than the number of determination points, It is determined that the obstacle OB1 does not exist in A8.
That is, in the case of FIG. 11B, the obstacle OB1 exists in the left direction with respect to the traveling direction of the vehicle. By judging that the obstacle OB1 does not exist, normal traveling can be continued without decelerating or stopping.

(Example 2)
Here, an example of an area weight value during a right turn and obstacle detection will be described.
If the vehicle is stationary and tries to turn to the left or right, it is likely to collide with objects in the direction of the turn, and may collide with objects in the direction opposite to the direction of the turn. Is considered low.
Therefore, when turning to the left or right, a relatively large area weight value is set for a plurality of divided areas existing in the direction in which the vehicle body advances due to the turning in the obstacle determination area, and the turning is performed. A relatively small area weight value is set for the other divided areas existing in the direction opposite to the direction in which the vehicle body travels.

The area weight value 79 in FIG. 7C shows the weight value Mn at the time of right turn and at the time of left turn. At the time of right turn, the weight value of the three divided areas (A1, A2, A3) is shown. Is the highest M1, the weight value of the two divided areas (A4, A5) is the next highest M2, and the weight value of the other divided areas (A6, A7, A8, A9) is the lowest M3.
FIG. 12A illustrates a setting example of the region weight values illustrated in FIG. 7C which are set when the vehicle starts turning right.
When the vehicle turns right at the current position where the vehicle is stationary, there is a possibility that the vehicle may proceed toward the right divided region with respect to the current position. Therefore, the right divided region (A1, A2, A3) Set the weight value to the highest.
Thus, after starting the right turn, an obstacle existing in the three divided areas (A1, A2, A3) that are likely to hinder the progress of the vehicle can be detected more reliably. Become.

FIGS. 12B and 12C show an embodiment in which an obstacle is detected after starting a right turn.
FIG. 12B shows a case where the obstacle OB1 is detected in the divided area A7 in which the lowest weight value M3 is set.
Here, assuming that the weight value M3 is 0.5, by performing the weighting, the number of measurement points after weighting is calculated as a numerical value smaller than the actual number of measurement points at which the distance is detected. If the calculated number of measurement points is smaller than the predetermined number of measurement points, it is determined that the obstacle OB1 does not exist.
That is, even if the obstacle OB1 exists in the divided area A7 after the start of the right turn, it is determined that the obstacle OB1 is not present because the obstacle OB1 does not significantly affect the running of the vehicle. By doing so, it is possible to continue the right turn and the subsequent traveling without unnecessary deceleration or stop.

On the other hand, FIG. 12C shows a case where the obstacle OB2 is detected in the divided area A2 in which the highest weight value M1 is set.
In this case, it is considered that the obstacle OB2 has a high possibility of hindering the right turn and the subsequent traveling of the vehicle.
Here, the number of measurement points at which the distance is detected in the area A2 is calculated as a numerical value larger than the actual number of measurement points by weighting.
When the calculated number of measurement points is larger than a predetermined judgment measurement number, it is more reliably determined that the obstacle OB2 exists in the area A2.
Therefore, by considering the distance L0 to the obstacle OB2, a process of decelerating or stopping the subsequent traveling of the vehicle is performed.

(Example 3)
Here, an example of the area weight value when starting to proceed in the right licking direction and obstacle detection will be described.
FIG. 13 illustrates the weight value Mn when trying to start traveling in the right licking direction shown in FIG. 8A.
As shown in FIG. 13A, it is assumed that the vehicle 1 starts to proceed in the direction of the division area A4 in the right tanning direction among the nine division areas.
In this case, the weight value of the divided area A4 in the traveling direction is set to the highest M1, and the weight value of the adjacent divided areas (A3, A5) of the area A4 is set to the next highest M2.
This is because, when there is an obstacle in the divided area A4, which is the traveling direction, the obstacle is reliably detected and the deceleration processing and the stop processing are performed. Also, it is for surely detecting an obstacle that straddles both the divided areas A4 and A5.
In the other areas (A1, A2, A6 to A9), even if there is an obstacle, there is a high possibility that the obstacle will not hinder the running of the vehicle, so the lowest weight value M3 is assigned.

FIGS. 13B and 13C show an embodiment in which an obstacle is detected after the vehicle starts traveling in the right licking direction.
FIG. 13B shows a case where the obstacle OB1 is detected in the divided area A7 in which the lowest weight value M3 is set.
In the divided area A7, the number of measurement points whose distances have been detected is calculated as a numerical value smaller than the actual number of measurement points by weighting the weight value M3.
If the calculated number of measurement points is smaller than a predetermined judgment measurement number, it is determined that the obstacle OB1 does not exist.
That is, even if the obstacle OB1 exists in the area A7 which is the left direction of the vehicle after the vehicle starts traveling in the right licking direction, the obstacle OB1 does not hinder the traveling of the vehicle in the right licking direction. Since it is not considered to give, it is possible to prevent unnecessary deceleration and stop by judging that there is no obstacle OB1, and it is possible to continue running in the right licking direction.

On the other hand, FIG. 13C illustrates a case where the obstacle OB2 is detected in a portion that straddles both the divided regions A3 and A4 for which a high weight value is set.
Since the divided area A4 is in the traveling direction, it is considered that the obstacle OB2 is likely to hinder subsequent traveling of the vehicle.
Here, in the areas A3 and A4, the number of measurement points whose distances have been detected is calculated as a numerical value larger than the actual number of measurement points by weighting the weight values M1 and M2, respectively.
If the calculated number of measurement points is larger than the predetermined determination measurement number, it is determined that the obstacle OB2 exists in both the areas A3 and A4.
Therefore, the vehicle deceleration process or the stop process is performed by considering the distance L0 to the obstacle OB2.
In FIG. 13, the case of proceeding in the right licking direction has been described. However, in the case of proceeding in the left licking direction, the direction only changes from right to left. Thus, obstacle detection and speed control are performed.

(Example 4)
Here, an example of an area weight value when right skidding is detected and an obstacle detection will be described.
If the road surface is wet with rain or is frozen, it may slip in a direction different from the actual traveling direction.
In other words, when slipping in an unintended direction, if there is an object in the slipping direction, it is necessary to decelerate or stop, but even if there is an object in the direction opposite to the slipping direction, deceleration must be performed. It is thought that there is no.
Therefore, when the inertia measuring unit 62 detects that the vehicle body skids in a specific direction, a relatively large area weight is assigned to a plurality of divided areas existing in the direction in which the vehicle body advances due to skidding in the obstacle determination area. A relatively small area weight value is set for a plurality of divided areas existing in the direction opposite to the direction in which the vehicle body advances due to skidding.

FIG. 14 illustrates the weight value Mn in the case of the right skid shown in FIG. 8A.
As shown in FIG. 14A, it is assumed that the vehicle 1 skids to the right while facing the direction of the divided area A5 which is the front direction of the vehicle.
At this time, since the traveling direction of the vehicle is the right direction, the weight value of the three divided regions (A1, A2, A3) in the right direction of the vehicle is set to the highest M1, and the weight of the region A4 adjacent to the divided region A3 is set. The value is set to the next highest weight value M2, and the weight values of the other areas (A5 to A9) are set to the lowest weight value M3.
By performing such weighting, it is possible to reliably detect an obstacle existing in the direction of the right skid, and an obstacle existing in the left direction opposite to the direction of the right skid can be detected by the right skid even if it exists. Collisions are unlikely to occur, so do not detect them.

FIGS. 14B and 14C show an embodiment in which an obstacle is detected when it is detected that the vehicle has skid to the right.
FIG. 14B shows a case where the obstacle OB1 is detected in the divided area A7 in which the lowest weight value M3 is set.
In the divided area A7, the number of measuring points at which the distance is detected is calculated as a numerical value smaller than the actual number of measuring points by weighting the weight value M3, and the calculated number of measuring points is smaller than the predetermined judgment measuring number. If so, it is determined that the obstacle OB1 does not exist.
Since the direction of the divided area A7 where the obstacle OB1 exists is significantly different from the direction in which the vehicle skids to the right, it is considered that the vehicle will not collide with the obstacle OB1.
Therefore, even if it is determined that the obstacle OB1 does not exist, the vehicle 1 is not hindered, and it is not necessary to perform the deceleration process or the process of stopping the skid.

On the other hand, FIG. 14C shows a case where the obstacle OB2 is detected in the divided area A2 in which the highest weight value M1 is set.
In the divided area A2, the number of measuring points whose distances have been detected is calculated as a numerical value larger than the actual number of measuring points by weighting the weight value M1.
If the calculated number of measurement points is larger than a predetermined judgment measurement number, it is determined that the obstacle OB2 exists in the area A2.
Therefore, it is considered that there is a high possibility that the obstacle OB2 will collide due to the skid of the vehicle. Therefore, a process of preventing the skid of the vehicle is performed by considering the distance to the obstacle OB2. To prevent sideslip, for example, the vehicle travels in the direction opposite to the obstacle or travels away from the obstacle.
In FIG. 14, a description has been given of the case of skidding to the right. However, in the case of skidding to the left, only the direction of skidding changes from right to left. Thus, obstacle detection and speed control are performed.

(Example 5)
Here, an embodiment will be described in which the area weight value is changed according to the positions of a plurality of measurement points whose distances have been detected.
FIG. 15 shows another example of setting the area weight value at the start of the right turn.
Before starting a right turn, execute the obstacle detection process for the entire determination area, and reset the area weight value in consideration of the arrangement and density of the detected obstacles. explain.
FIG. 15A shows an example of the same setting as the region weight value at the start of the right turn shown in FIGS. 7C and 12A.

When the vehicle attempts to make a right turn, the vehicle is once stopped, and before actually making a right turn, the laser beam is two-dimensionally and three-dimensionally applied to all the determination regions 78a (divided regions A1 to A9). Light is scanned to perform an obstacle detection process.
At this time, the position of the measurement point at which the distance has been detected is stored.

FIGS. 15B and 15C show examples of obstacle detection at the start of a right turn.
FIG. 15B shows that, in the divided areas A2 and A3, the distance between the plurality of measuring points whose distances are detected is considerably short, and the plurality of measuring points whose distances are detected are within a predetermined range. Shows a case where they exist densely.
In FIG. 15B, only four measurement points are shown for ease of explanation. However, if distances are detected at all the measurement points existing in the area AS surrounding the four points, the four points are detected. It is considered that there is a high possibility that the obstacle OB1 exists in the area AS surrounding the point.

Therefore, the distance between the measuring points is calculated for each of the plurality of measuring points whose distances are detected, and the variance of the calculated distance between the measuring points is not large. If it is included in the range of the fixed value, it is determined that there is a high possibility that the obstacle OB1 exists in the area AS, and the weight value M1 of the divided areas A2 and A3 including the area AS is determined. Is changed to a higher value M11 (M11> M1).
As a result, the weight values of the divided areas A2 and A3 are set higher, so that obstacles existing in the areas (A2, A3) can be more reliably detected.

On the other hand, FIG. 15C shows a case in which a plurality of measurement points whose distances are detected exist in the divided areas A2 and A3, but the measurement points are located far apart.
In FIG. 15C, four measuring points are shown as the measuring points at which the distances are detected. However, the distance between the measuring points of these four measuring points is considerably large, and the four measuring points are calculated as shown in FIG. Suppose that it could not be included in the predetermined area AS as shown in b).
For example, if the distance between each measurement point is 20 cm or more, and there is no other measurement point whose distance is detected between the measurement points, each measurement point is a dispersed isolated point. Therefore, it cannot be determined that one obstacle exists in the area including the four measurement points, and different obstacles may exist at the positions of the four measurement points.
Alternatively, if there is no other measurement point whose distance is detected around the four measurement points, for example, an object that cannot be an obstacle, such as grass, may have been detected, or a false detection may have occurred. obtain.
Therefore, as shown in FIG. 15C, when a plurality of measuring points whose distances are detected are detected in the divided area, the plurality of measuring points are detected in a dispersed state separated by a predetermined distance or more. If so, the weight value is not changed.

(Example 6)
FIG. 16 is an explanatory diagram of an embodiment in a case where weighting is performed on the determination region in the height direction.
When scanning the LIDAR in the vertical direction, it is preferable to divide the obstacle determination area into a plurality of areas in the height direction (vertical direction) so as not to erroneously detect the ground as an obstacle.
Therefore, when the light is scanned in the direction perpendicular to the ground with the light emission position as the center and the vertical fan-shaped area to which the light can reach is set as the obstacle determination area, each divided area is determined as the obstacle determination area. The region is a rectangular region divided into a plurality of regions in the vertical direction, and the region weight value is made different according to the height of each divided region from the ground.
For example, an area weight smaller than the other divided areas is set to the divided area closest to the ground.

When an obstacle is detected by operating the LIDER of the distance detection unit 51 three-dimensionally, the laser light emitted from the light emitting unit 51a is scanned not only in the horizontal direction but also in the vertical direction.
When the laser light is scanned downward, the presence of the ground at a position several meters away from the vehicle is detected by reflecting the laser light on the ground.
If the road surface on which the vehicle travels is a road surface that is always flat with no unevenness, it will always detect the ground ahead from a position several meters away when scanning downward, so it is included in such a detection area. The distance information acquired from the measurement point to be measured may be recognized as the ground and may be determined as not an obstacle.

However, the road surface actually traveling outdoors has few flat roads, has irregularities and steps, and has uphill and downhill slopes.
Therefore, for example, when only the front wheels of the vehicle ride on a road surface having a protrusion, the laser emission direction is directed upward as compared to a flat road, and the road surface may not be detected.
Also, when only the rear wheel of the vehicle rides on a bumpy road surface, the emission direction of the laser will be directed downward compared to a flat road, the laser light hitting the ground will increase, and the distance Will be detected.
That is, even if the measurement point is not detected when traveling on a flat road, if the emission direction is directed downward, the distance will be detected, and it will be recognized as ground when traveling on a flat road. For a missing station, it may be determined that an obstacle exists at the location of the station, even though the ground is detected.

In this case, even though the rear wheel has just climbed on the road with the protrusion temporarily, unnecessary deceleration processing may be performed by recognizing the ground as an obstacle.
Therefore, in this embodiment, the determination area in the vertical direction is also weighted, and the weight setting is changed in accordance with the inclination of the vehicle in the vertical direction, and the detected ground is erroneously determined to be an obstacle. And reducing unnecessary deceleration processing.

FIG. 16A shows an example of weighting the determination area in the vertical direction when the vehicle is traveling on flat ground.
A LIDAR, which is a distance detection unit 51, is attached to the upper front part of the vehicle 1, and a laser beam is emitted from the light emitting unit 51a to the front in the traveling direction, and as shown in FIG. It is assumed that laser light is scanned three-dimensionally.
FIG. 16A is a cross-sectional view of the vehicle 1 as viewed from the lateral direction. When the laser beam is scanned in the vertical direction, for example, when the scanning angle θ in the vertical direction is 50 degrees, the laser beam is emitted downward. When the laser light hits the ground and is emitted upward, the laser light is irradiated toward the three-dimensional space above the vehicle.
The ground is not an obstacle, so when traveling on flat ground, even if the distance is detected by receiving the reflected light from the measurement point corresponding to the position of the ground, for those measurement points, Exclude from obstacle detection.

Therefore, in FIG. 16A, in order to prevent the ground from being detected as an obstacle, the presence or absence of an obstacle is detected in a rectangular three-dimensional space parallel to the ground within the laser irradiation range in the vertical direction. This is a determination area.
The determination area shown in FIG. 16A is composed of three divided areas (A11, A12, A13) divided in the height direction, and weight values of the respective divided areas are (W11, W12, W13). I do.
The magnitude relation of the weight values is W11>W12> W13, and the weight value of the divided region at the highest position is the largest value.
The divided area A11 at the highest position is a rectangular three-dimensional space parallel to the ground and having a height of H1. If a distance is detected at a measurement point in the area A11, the weight value W11 is set to the largest numerical value. Therefore, the weighted number of measurement points is calculated as a numerical value larger than the number of measurement points at which the distance is actually detected.

The divided area A12 at the second highest position is a three-dimensional space parallel to the ground and having a height of H2. In this area A12, the number of measuring points whose distance has been detected is weighted by a weight value W12 smaller than W11.
The divided area A13 at the lowest position is slightly higher than the ground, and is a three-dimensional space parallel to the ground and having a height of H3.
Since the weight value W13 of this area A13 is the smallest, the number of measuring points whose distance is detected in this area is weighted by the low weight value W13 and calculated as a numerical value smaller than the actual number of measuring points, for example.
The smallest weight value W13 is set in the lowest divided area A13 because, for example, there is a high possibility that the ground is erroneously detected as an obstacle, or an object that cannot be an obstacle such as grass is detected. Yes, by reducing the weight value, the possibility of erroneous detection or the like is reduced.

When the vertical scanning angle θ is 50 degrees, the height of each divided region is, for example, about H1 = 20 cm, H2 = 30 cm, and H3 = 50 cm.
However, the height of each divided area is not fixed to this numerical value.For example, based on the traveling speed, the road surface condition, the height of grass and the undulation of the road surface, other values are fixedly set or the traveling state May be changed in correspondence with.
Further, the length of each divided region in the traveling direction (the length in the left-right direction of the paper surface) may be set based on, for example, a supposed maximum amount of tilt of the vehicle body.

As shown in FIG. 16A, when the vehicle is traveling on flat ground, if there is a measuring point whose distance is detected only in the lowest divided area A13, the number of measuring points in that area is reduced. Is given a low weight, it is difficult to determine that an obstacle exists.
As a result, it is possible to reduce the determination of the ground as an obstacle, or to make it difficult to determine a small object such as grass as an obstacle, and to reduce unnecessary deceleration processing and stopping.
On the other hand, if the distance is detected at many measurement points from the low position to the relatively high division area, the number of the measurement points at the high position is counted as a large number of measurement points by a large weight value, so that the measurement is more reliable. Then, it is determined that an obstacle exists in the determination area.
As described above, in the vertical direction, the obstacle determination area is divided into a plurality of divided areas, a low weight area is weighted with a small weight, and a high weight area is weighted with a large weight. Stopping can be reduced, and obstacle detection can be performed more reliably.

FIGS. 16 (b), (c), and (d) are explanatory diagrams of an embodiment of weighting in the vertical direction when the vehicle body is tilted downward.
FIGS. 16B, 16C and 16D show a state where the rear wheel of the vehicle body 1 rides on the convex surface G of the ground. At this time, assuming that the vertical range of the laser beam emitted from the LIDAR 51 is within the range of the scanning angle θ as in FIG. , The determination area also leans downward.
The fact that the front surface of the vehicle body 1 is tilted downward can be detected by an acceleration sensor, a gyro sensor, or the like of the inertial measurement unit 62, and the angle of the vehicle body tilted downward can also be calculated.
When the determination area is a three-dimensional space inclined downward, the ground is detected at a portion of the determination area located below in the determination area.
If the ground is detected, the detected ground part may be judged as an obstacle and unnecessary deceleration processing may be performed.Therefore, measurement of the part of the determination area where the ground may be detected due to the inclination of the vehicle body may be performed. At this point, it is preferable to invalidate the distance measurement or to reduce the weight of such a determination region.

FIG. 16B shows an embodiment in which the method of dividing the divided region belonging to the inclined determination region is changed. Here, the determination area is divided into three divided areas (A21, A22, A23), and the boundary of each divided area is set to be parallel to the ground.
For example, the heights of the left ends of the three divided areas (A21, A22, A23) are respectively defined as the heights (H1, H2, H3) of the three divided areas (A11, A12, A13) in FIG. The height of the right end of each divided region is the same as the height of the left end from the ground, regardless of the inclination of the divided region.
In this case, the height of the right end of the highest position of the divided area A21 becomes shorter, so that the space of the divided area A21 becomes smaller.
On the other hand, since the height of the right end of the divided area A23 at the lowest position is longer, the space of the divided area A23 is wider than the divided area A13 of FIG. 16A, but this area also includes the ground. .
Therefore, in order to reduce erroneous detection of the ground as an obstacle, the weight value W23 of the divided area A23 is set to a value smaller than the weight value W13 of the divided area A13 in FIG. W13).

Alternatively, as shown in FIG. 16B, it is considered that the rear wheel rides on the convex surface G and inclines, and it is considered that the rear wheel often returns immediately to traveling on a flat road. The value may be set to zero, the distance measurement at the measurement point included in the divided area A23 may be invalidated, and the obstacle detection in the divided area A23 may not be performed.
For the upper two divided areas (A21, A22), for example, the same weight values (W11, W12) as those of the divided areas (A11, A12) in FIG. What is necessary is just to detect an object.
As described above, when the front surface of the vehicle body is tilted downward, the weight value of the divided region A23 at the lowest position among the determination regions is set to be small, and the obstacle detection of the divided region A23 is temporarily performed. Unnecessary deceleration processing and stoppage can be prevented.

FIG. 16C shows an embodiment in which the size and shape of the three divided regions (A11, A12, A13) of the inclined determination region are the same as those in FIG. 16A. Here, the three divided areas (A11, A12, A13) also incline similarly to the inclination of the vehicle body, so that the ground is detected in the lowest divided area A13.
As described above, when the angle of inclination of the vehicle body is calculated by the inertial measurement unit 62, the inclination angle of the divided area is also known, so that it is also recognized that the ground may be detected in the divided area A13.
Therefore, in order to reduce erroneous detection of obstacles due to the ground, the weight value W23 of the divided area A13 at the lowest position is set to a value lower than the weight value W13 in FIG. 16A (W23 <W13). ).
Alternatively, as described with reference to FIG. 16B, the weight value of the divided area A13 may be set to zero and the obstacle detection of the divided area A13 may be temporarily disabled.

FIG. 16D shows another embodiment in which the method of dividing the divided region belonging to the inclined determination region is changed.
In FIG. 16B, the divided areas are set such that the boundaries of the three divided areas (A21, A22, A23) are parallel to the ground. In FIG. 16D, the three divided areas belonging to the determination area are set. The boundary of (A31, A32, A33) is set to have a slope corresponding to the slope of the determination area.
In addition, a point P where the right end of the determination area intersects the ground is generated due to the inclination of the determination area. With this point P as a base point, a boundary between the lowest divided area A33 and the upper divided area A32 is set. .

The boundary between the divided areas A32 and A33 and the boundary between the divided areas A31 and A32 are set to be parallel as shown in FIG. As a result, the lowermost divided area A33 includes an area below the point P for detecting the ground.
When the inclination of the vehicle body is determined by the inertia measuring unit 62, the position at which the determination area intersects with the ground is calculated based on the inclination below the scanning angle θ and the inclination of the lowermost side and the right side of the determination area. The position of P is determined.

When the point P is obtained, the divided area A33 is set as shown in FIG.
Further, the boundary between the upper divided regions A31 and A32 may be set, for example, such that both divided regions have the same height, or based on the magnitude of the inclination angle of the determination region. You may decide.
As for the weight values of the two divided regions (A31, A32), for example, the same weight values (W11, W12) as those of the divided regions (A11, A12) in FIG.
On the other hand, the lowest divided area A33 includes a portion below the point P where the ground is detected. Therefore, as the weight value W33 of this area A33, the above-described weight value W13 or It is preferable to set a numerical value lower than W23 (W11>W12> W33, W13>W23> W33).

In the divided area A33, the ground is detected in an area below the point P. Therefore, the weight value of the divided area A33 is set to zero, and the obstacle detection of the divided area A33 is not performed. Is also good. That is, obstacle detection may be performed only in the upper two divided areas A31 and A32.
Thus, since the obstacle detection is not performed in the divided area A33 where the ground is detected, it is possible to prevent the ground from being erroneously detected as the obstacle, and to reduce unnecessary deceleration processing and stoppage.

  16B, 16C, and 16D, the vehicle further travels in the traveling direction, and after the rear wheel descends from the convex surface G, the vehicle travels on a substantially flat road surface. In this case, the inclination in the vertical direction is no longer detected by the inertial measurement unit 62, so that an area for detecting an obstacle is divided into three rectangular divided areas (parallel to the ground) as shown in FIG. A11, A12, and A13) may be returned to the determination area, and the area weight may be returned to the weight shown in FIG.

Reference Signs List 1 autonomous traveling device, 2 monitoring unit, 3 control unit, 5 management server, 6 network, 10 body, 12R right side, 12L left side, 13 front, 14 rear, 15 bottom, 16 housing space, 18 cover, 21 front wheel, 21a axle, 21b sprocket, 22 rear wheel, 22a axle, 22b sprocket, 23 belt, 31 front wheel, 31a axle, 31b sprocket, 32 rear wheel, 32a axle, 32b sprocket, 33 belt, 40 battery, 41R electric motor, 41L electric motor Motor, 42R motor shaft, 42L motor shaft, 43R gearbox, 43L gearbox, 44R bearing, 44L bearing, 50 control unit, 51 distance detection unit (LIDAR), 51a Light section, 51b light receiving section, 51c scanning control section, 51d laser, 52 running control section, 53 wheels, 54 communication section, 55 camera, 56 image recognition section, 57 obstacle detection section, 58 position information acquisition section, 59 rechargeable battery , 60 speed detection unit, 61 monitoring information acquisition unit, 62 inertia measurement unit, 63 area weight calculation unit, 70 storage unit, 71 input image data, 72 measurement distance information, 73 current position information, 74 route information, 75 transmission monitoring information , 76 Set speed information, 77 Obstacle information, 78 Judgment area information, 79 Area weight value, 91 Communication section, 92 Monitoring control section, 93 Storage section, 93a Reception monitoring information, 100 objects

Claims (13)

The body and
A travel control unit that controls a drive member that drives the vehicle body,
A predetermined light is emitted to a predetermined obstacle determination area including a forward space in the traveling direction, and a light reflected by an object existing in the obstacle determination area is received to detect a distance to the object. A distance detector to perform
An inertial measurement unit that detects a current traveling direction of the vehicle body by measuring a displacement generated by traveling of the vehicle body,
Route information that defines a traveling route of the vehicle body, determination area information that sets the position and size of each divided area obtained by dividing the obstacle determination area into a plurality of pieces, and information that is set based on the traveling state of the vehicle body A storage unit that stores a region weight value indicating the importance of detecting an obstacle in the divided region, and a control unit,
The control unit may be configured to divide the travel direction of the vehicle body based on a current travel direction detected by the inertial measurement unit or a travel direction to be traveled in a travel route determined in the route information. An autonomous traveling apparatus, wherein an area weight value is set so that a relatively large area weight value is set in an area. Furthermore, autonomous driving apparatus of claim 1, in several divided region near the divided area including the traveling direction of the vehicle body, and sets a relatively large region weight value. When it is determined that the object whose distance has been detected exists in the divided region in which the relatively large region weight value is set, a deceleration process or a stop process of the vehicle body is performed in accordance with the region weight value. The autonomous traveling device according to claim 1 or 2, wherein: Wherein a relatively large area divided area weight value is set, if the object is determined not to exist, autonomy according to claim 1 or 2, characterized in that does not perform the deceleration process and stop process of the vehicle body Traveling device. In a divided region other than the divided region in which the relatively large region weight value is set,
The autonomous traveling apparatus according to any one of claims 1 to 4, wherein when it is determined that an object whose distance has been detected is present, deceleration processing and stop processing of the vehicle body are not performed. In the divided region in which the relatively large region weight value is set,
The autonomous traveling apparatus according to any one of claims 2 to 5, wherein a divided area A including the traveling direction of the vehicle body and two divided areas of the divided area A are included.   2. The system according to claim 1, wherein a largest area weight value is set for a divided area including the traveling direction of the vehicle body, and a smaller area weight value is set for other divided areas as the distance from the traveling direction increases. 6. The autonomous traveling device according to any one of claims 1 to 5, When the vehicle body is going to turn left or right in a stationary state,
In the obstacle determination area, a relatively large area weight is set for a plurality of divided areas existing in the direction in which the vehicle body advances due to the turn, and the divided areas exist in the direction opposite to the direction in which the vehicle body advances due to the turn. The autonomous traveling apparatus according to any one of claims 1 to 5, wherein a relatively small area weight value is set in each of the divided areas. When the inertia measuring unit detects that the vehicle body has skid in a specific direction,
Of the obstacle determination areas, a relatively large area weight value is set for a plurality of divided areas existing in a direction in which the vehicle body advances due to the sideslip,
The autonomous traveling apparatus according to any one of claims 1 to 5, wherein a relatively small area weight value is set in a plurality of divided areas existing in a direction opposite to a direction in which the vehicle body advances due to the side slip. . With the light emission position as the center, the light is scanned in a direction parallel to the ground,
When the horizontal fan-shaped area that the light can reach is the obstacle determination area,
10. The autonomous traveling apparatus according to claim 1, wherein each of the divided areas is an area obtained by dividing the obstacle determination area into a plurality of sector shapes having a predetermined central angle. With the light emission position as the center, the light is scanned in a direction perpendicular to the ground,
When the vertical fan-shaped area that the light can reach is an obstacle determination area,
Each of the divided areas is a rectangular area in which the obstacle determination area is divided into a plurality of pieces in the vertical direction, and the area weight value is changed in accordance with the height of each divided area from the ground, so that the divided area is closest to the ground. 10. The autonomous traveling apparatus according to claim 1, wherein an area weight value smaller than that of the other divided areas is set in the area. The distance detection unit includes a light-emitting unit that emits light to the obstacle determination area in the traveling direction, a light-receiving unit that receives light, and the light that is emitted toward a plurality of measurement points in the obstacle determination area. Scanning control unit that scans the light emission direction,
The distance detecting unit uses the time difference between the time at which light is emitted from the light emitting unit and the time at which the light reflected by the object is received by the light receiving unit, and the light emitting unit and The autonomous traveling apparatus according to any one of claims 1 to 11, wherein a distance between the plurality of measurement points is calculated, and a distance to the object is detected from the calculated distance.   The distance detector uses a LIDAR that emits a laser to a two-dimensional space or a three-dimensional space in a predetermined obstacle determination area and measures distances at a plurality of measurement points in the obstacle determination area. The autonomous traveling device according to any one of claims 1 to 12.

Images