JP2011054082A - Autonomous moving apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable an autonomous moving apparatus to pass an area, where a certain number or more of moving objects exist, stably at high speed. <P>SOLUTION: The autonomous moving apparatus 1 includes an external sensor part 2, a moving object detecting means 40, an own position prediction means 41, a follow-up object choosing means 42, a passage creating means 43, and a driving part 3. Based on object position coordinate 108 of the moving object and velocity vector 109, the follow-up object choosing means 42 chooses the moving object nearest to the direction of a target of this apparatus 1 as the follow-up target, the passage creating means 43 generates the running path to the back of the follow-up target. The follow-up object choosing means 42 obtains the object position coordinate 108 and the velocity vector 109 of other moving object even during the follow-up of the target object, and if determines that it is better to change the follow-up target, the follow-up target is switched and the passage creating means 43 generates the running path up to the target object after the switching. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a technology of an autonomous mobile device (hereinafter sometimes referred to as “this device”) that travels in a region where a certain number or more of moving or stationary objects exist. Specifically, it can be used for a moving device that can move at a speed equal to or higher than that of a person in an environment where there are moving objects such as people, vehicles, animals, or static or dynamic objects. It relates to technology that can be used for service robots in various environments such as autonomous mobile vehicles, automatic transfer devices, automatic cleaning devices, security robots, factories, hospitals, offices, and roads that move in the same area.

  In recent years, along with developments such as sensing technology and improvement in calculation speed, mechanization and automation technology of human work is rapidly progressing. In this context, the importance of autonomous mobile technology that enables only machines to move within the operating environment without human control is increasing, and development of practical applications is also actively underway. Autonomous mobile devices that have already been put into practical use include automatic guided vehicles (AGVs), automatic cleaning devices, security robots, service robots in hospitals, offices, department stores, and the like. At present, these are applied only when there are few objects that move around (hereinafter sometimes referred to as “moving objects”). In the future, there will be a certain number of moving objects such as in stations and towns. It is expected that an autonomous mobile device that performs this work will be put to practical use. For this reason, there is an increasing demand for a technique for safely and quickly passing through a region where there are many moving objects in the vicinity without contacting the object.

In a region where a plurality of objects exist, a potential method is given as a representative example of the prior art that travels while avoiding each object (see Patent Document 1). In the potential method, it is assumed that there is a high potential field in the existence area of the object that becomes an obstacle (moving obstacle: moving object), and there is a low potential field in the destination where the object does not exist. On the other hand, it is a technology that generates a force in the repulsive direction from the potential field of the object, generates a force in the attractive direction from the potential field of the destination, and controls the autonomous mobile device using the resultant force. Patent Document 1 further discloses that a probability potential field is generated by reflecting the existence probability of a moving obstacle in the potential field, and a moving object can be avoided by using the probability potential field.
However, in the method using the conventional method such as the potential field shown in Patent Document 1 or the like, it is necessary to travel while avoiding all the moving objects in an area where the number and density of the moving objects are large. There are problems such as inability to generate a route and time-consuming passage. Furthermore, in autonomous running by sensing from this device, the situation after avoiding an obstacle is unknown, so the risk of contact with an object that becomes an obstacle increases.

  In addition, among autonomous movement techniques, there is a technique that not only avoids moving obstacles but also performs follow-up movement (see Patent Document 2). The tracking movement technique is a technique for controlling the apparatus to travel at the same speed behind the object using the moving direction and speed of the tracking target. This can be applied to, for example, a service robot that carries the luggage of the person to be followed. Patent Document 2 further discloses that the position and moving direction of the apparatus at the time of following can be changed from the movement of a peripheral object or the following target person to the front, rear, left or right of the following target person. Such a tracking function makes it possible to avoid collisions while moving with a tracking target person, but when used for the purpose of passing through a region with a high human density, the tracking target person himself / herself (autonomous mobile device) There is a problem that it does not always go in the direction (target position).

Further, there is a technique that uses the following function as described above for passing through a region having a high human density (see Non-Patent Document 1). This is because when moving to an area with a high human density, the object is moved to its target position while following the target object by selecting an object having a target direction of the apparatus and a moving direction within a certain threshold. Technology. In this technique, a candidate object nearest to the apparatus is selected as a tracking target from candidate objects having a moving direction within the range of the target direction and the threshold angle range, and the tracking target selects the moving direction within the threshold. A control method that continues to follow as long as it is held is executed. Non-Patent Document 1 discloses that by using this control method, it is possible to reduce the risk of contact with a moving obstacle other than the tracking target being tracked.
However, this control cannot change the tracking target even when there is a better tracking target candidate during the tracking movement. Therefore, in order to reach the target point, it is necessary to perform autonomous traveling in the direction of the target point by avoidance control when the tracking target is away from the target position, and to search for a new tracking candidate, which takes time. Further, since the nearest candidate is selected, there is a possibility that the candidate whose movement direction is closest to the target direction is missed.

JP 2003-241836 A JP 2007-316924 A

"Operation planning of multiple mobile robot systems using tracking strategies that can generate group order" The Journal of the Robotics Society of Japan Vol.12, No.4, pp603-608

As described above, although inventions for avoiding collisions with moving obstacles have been made in the past, the target is, after all, the movement of all moving obstacles can be grasped, and avoidance routes can be generated with low human density. It is a simple situation. However, these inventions cannot be applied to movement in areas where there are many moving obstacles such as stations and busy streets in consideration of autonomous movement in the same environment as humans, which is surely required in the future. In this area, the blind spots of the moving obstacles are increased, and the distance between the moving obstacles is narrowed so that the travel route cannot be generated safely. In other words, in order to avoid all existing objects, the avoidance path can only be done within the sensing range, it is not possible to grasp what is in the area that is the blind spot after avoidance, if there is a moving object In such a case, the avoidance movement becomes abrupt and the contact risk is high.
Further, in the solution using the tracking function (see Non-Patent Document 1), the tracking target performs mutual avoidance with other moving objects, so the risk of contact with other than the tracking target is reduced. However, in areas where the human density is high, there is often a gap between the target position to be tracked and its own target position, and it is very common to find a person to be tracked that reaches the target position. Takes time. Furthermore, when no follow-up target is found, it is even more necessary because it is necessary to avoid the moving object.

  Therefore, the problem to be solved by the present invention is to realize safe and high-speed traveling in an area where a large number of moving objects exist.

In order to solve the above problems, the present invention is realized by the following means.
That is, the autonomous mobile device of the present invention acquires information on the outside world, detects the position and speed vector of the moving obstacle based on the information, and based on the position and speed vector, In order to select a candidate object having a near velocity vector as a tracking target and follow the selected tracking target, control is performed by generating a route to the rear of the object. If there is a candidate with a velocity vector closest to the target direction using the position and velocity vector of another moving object while following the tracking target, change the tracking target and run behind the target object By generating a route, control to travel to the destination point is performed.
This shortens the time to reach the target while reducing the risk of contact with the moving object even in a region where a certain number of moving objects exist.
Details will be described later.

  ADVANTAGE OF THE INVENTION According to this invention, the safe and high-speed driving | running | working in the area | region where many moving objects exist is realizable.

It is a conceptual diagram which shows a mode that the autonomous mobile apparatus selects the object which follows in an area | region with many moving objects. It is a conceptual diagram which shows a mode that the target which an autonomous mobile apparatus tracks in the area | region with many moving objects is changed. It is a functional lineblock diagram of the autonomous mobile device of a first embodiment. It is a flowchart which shows the specific process sequence which selects a tracking target. It is a functional block diagram of the autonomous mobile apparatus in 2nd embodiment. It is a conceptual diagram which shows a mode that the path | route to the target point of an autonomous mobile apparatus is produced | generated in the area | region with many moving objects. It is a system configuration | structure figure of the network which an autonomous mobile apparatus connects so that communication is possible. It is a functional block diagram of the autonomous mobile apparatus of 3rd embodiment. It is a conceptual diagram which shows a mode that the intermediate point of an autonomous mobile apparatus is produced | generated in the area | region with many moving objects.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
≪First embodiment≫
First, operations targeted by the autonomous mobile device of the present embodiment will be described with reference to FIGS. 1 and 2.

  FIG. 1 is a conceptual diagram illustrating a state in which an autonomous mobile device follows a target (hereinafter, may be referred to as a “follow target”) in an area where there are many moving objects. In FIG. 1, A to D are objects moving around the apparatus 1, and are moved by velocity vectors Va to Vd in the absolute coordinate system, respectively. At this time, in order to reduce the risk of contact with the moving object and the apparatus 1 passes through the region, the moving object that is directed toward the target point 10 of the apparatus 1 is selected and followed by the object other than the tracking target. Reduce contact opportunities.

  The specific selection method includes a direction vector (hereinafter sometimes referred to as “target direction vector”) from the current point of the device 1 to the target point 10 of the device 1 and the velocity vectors Va to V of the objects A to D. An angle formed by Vd (hereinafter, also referred to as “direction angle difference”) θa to θd is calculated, and the object C having the smallest θc among the direction angle differences θa to θd is selected as the target direction of the apparatus 1 and the target direction. It is assumed that it is the closest and is selected as the tracking target. When calculating the direction angle difference, the velocity vector can be calculated clockwise and counterclockwise with respect to the target direction vector, but the smaller of these two values is calculated as the direction angle difference. To do. For example, referring to FIG. 1, for the object A, the direction angle difference θa is calculated by specifying the counterclockwise velocity vector Va that decreases in value.

  FIG. 2 is a conceptual diagram showing a state in which the target that the autonomous mobile device follows in an area where there are many moving objects. E and F in FIG. 2 are objects that are moving around the apparatus 1, and are moved by velocity vectors Ve and Vf in the absolute coordinate system, respectively. When this apparatus 1 heads for the target point 11, it first selects the object F as the tracking target and moves while following. Even during the following movement, the apparatus 1 continues to acquire the velocity vector Ve of the surrounding object E and the velocity vector Vf of the object F to be followed, and the direction of the object E toward the target point 11 of the apparatus E and When the direction angle difference θe is smaller than the direction angle difference θf of the object F, the tracking target is changed to the object E. That is, the tracking target is changed to the object E when θe becomes smaller than θf while following the object F.

The configuration of the autonomous mobile device 1 necessary for realizing the operation described with reference to FIGS. 1 and 2 will be described with reference to FIG. FIG. 3 is a functional configuration diagram of the autonomous mobile device of the first embodiment. The autonomous mobile device 1 includes an external sensor unit 2, a drive unit 3, and a control unit 4. The control unit 4 includes a moving object detection unit 40, a self-position estimation unit 41, a tracking target selection unit 42, and a route generation unit 43. Composed.
Such an autonomous mobile device 1 includes, for example, an input unit realized as the external sensor unit 2, an output unit realized as the drive unit 3, a control unit (corresponding to the control unit 4), and a storage unit (not shown). It is a computer. The control unit includes a CPU (Central Processing Unit), an LSI (Large Scale Integration Circuit), and the like, and the storage unit includes a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. The storage unit stores a program for realizing each of the above means, and the control unit 4 reads the program and executes processing. Each component will be described below.

The external sensor unit 2 is a range sensor that performs sensing and acquires shape information of a moving object around the apparatus 1, and uses a stereo camera, a laser scanner, a radar, or the like. For example, a description will be given below with an apparatus configured using a stereo camera. The shape information acquired by the external sensor unit 2 is output as sensor data 100 to the control unit 4.
The drive unit 3 has a function of causing the device 1 to travel along a predetermined route in response to a command from the control unit 4.
The control unit 4 has a function of determining the movement of the apparatus 1 based on the input from the external sensor unit 2 and outputting it to the drive unit 3. The detailed configuration is shown below.

  The moving object detection means 40 detects an object existing in the vicinity from the sensor data 100 acquired from the external sensor unit 2. In addition, a function of deriving the relative distance and direction of the object from the apparatus 1 is provided. For example, when a stereo camera is used, it is possible to calculate and derive the relative coordinates with the present apparatus 1 obtained from the parallax information of the left and right cameras as the object position coordinates 108. This can be achieved by a detection method using an external sensor such as another laser scanner, a monocular camera, or a radar, but the effect obtained in any case is basically the same.

  Further, by tracking the object, the relative velocity of the object is obtained from the difference between the object position coordinate 108 one sampling step before and the current object position coordinate 108. When the control unit 4 calculates the translation speed and the angular velocity of the device 1 from the wheel rotation speed and the steering amount measured by the drive unit 3, the moving object detection means 40 estimates the device 1 from the relative speed. The velocity vector 109 of the peripheral object is derived by removing the velocity component (translation velocity and angular velocity of the device 1).

The self-position estimating means 41 estimates the self-position coordinates 101 in the absolute coordinate system of the device 1 by accumulating the translation speed and the angular velocity of the device 1 obtained from the control unit 4. Further, the error caused by the cumulative calculation is corrected by matching the object position coordinates 108 of the object detected by the moving object detection means 40 with the map information previously owned (stored) in the storage unit.
This means can be realized by the method disclosed in the following reference, for example.

References: "Vehicle" Kimio Kanai Measurement and Control Technology Series Chapter of the Society of Instrument and Control Engineers Chapter 7 [Position Recognition of Moving Objects] pp142-173

The storage unit may store an initial value of the self-position coordinate 101 of the apparatus 1 in order to perform the cumulative calculation.

  The tracking target selecting means 42 has a function of selecting an object having a velocity vector closest to the target direction of the apparatus 1 as a tracking target from among the peripheral objects, as outlined by FIGS. Details of specific functions will be described later. The object position coordinates of the object selected as the tracking target are referred to as tracking target position coordinates 105a, and the velocity vector of the object selected as the tracking target is referred to as tracking target speed vector 105b.

  The route generation unit 43 generates a route for following the subject to be followed. Specifically, the tracking target position coordinates 105a and the tracking target speed vector 105b of the tracking target object selected by the tracking target selection means 42 are acquired, and a low potential field for attractive force is applied to the tracking target object by, for example, the potential method. Is generated as a follow-up destination. Furthermore, a high potential field for repulsive force is also generated on the moving object at the same time to prevent contact with the object to be followed, which is the follow-up destination, and follow at a position spaced apart by a certain distance. At this time, generation of repulsive force due to a high potential field has a strong influence because the contact with the velocity vector direction (traveling direction) has a high risk of contact, and conversely, a weak force is applied on the opposite side to the traveling direction. 1 also follows at the rear part deviating from the traveling direction of the object to be tracked.

  Further, the path generation means 43 avoids itself from a moving object approaching the device 1 by making the repulsive force from an object other than the object to be tracked weaker during follow-up than when not following it. The avoidance behavior of the device 1 is weakened. As a result, the performance of following the object to be tracked can be increased by reducing the avoidance performance of the apparatus 1 itself from the moving object. An object other than the tracking target object is unlikely to approach the tracking target object, and as a result, the object is unlikely to approach the apparatus 1 that is being tracked. be able to.

  Finally, the path generation unit 43 calculates the drive command value 106 from the sum of the attractive force from the tracking target and the repulsive force from other moving objects, and outputs the drive command value 106 to the drive unit 3. As a result, it is possible to travel at the same speed in the same direction as the tracking target, and the effects obtained using other tracking methods are basically the same. The drive command value 106 includes, for example, a drive amount value required to travel at substantially the same speed in substantially the same direction as the tracking target and a control signal for starting driving of the drive unit 3. In addition, the generation of the follow-up path by the path generation unit 43 includes a process of calculating at least the drive command value 106. However, in addition to this process, include processing to analyze the terrain information of the route itself (road condition (eg, whether it is paved, whether there is a puddle), slope, hardness, etc.). May be. The topographic information can be handled as a part of the sensor data 100 acquired from the external sensor 2, for example. Depending on the terrain information, depending on the case, the autonomous mobile device 1 can be bypassed and controlled to continue following.

FIG. 4 is a flowchart showing a specific processing procedure for selecting a tracking target. This process is executed by the tracking target selection means 42.
As a rough flow of this flowchart, steps 420 to 423 are used to calculate the angle (direction angle difference) between the target direction of the apparatus 1 described in FIG. Ranking. In step 424, either the processing in step 425 or the processing in steps 426 to 428 is selected depending on whether or not the apparatus 1 is currently following. Details of each operation will be described below.

  In step 420, the target direction is calculated from the self-position coordinates 101 of the apparatus 1 acquired from the self-position estimation means 41 and the absolute target position coordinates (position coordinates in the absolute coordinate system of the target point) preset in the storage unit. .

  In step 421, the target direction of the apparatus 1 calculated in step 420 is compared with the velocity vector of each moving object acquired from the moving object detection means 40, and the direction angle difference from the target direction of each velocity vector is calculated.

  In step 422, the direction angle difference from step 421 and the velocity vector of the surrounding moving object are obtained from the moving object detection means 40 and checked, and both the direction angle difference and the velocity vector are within a certain range. For example, the moving object is set as a tracking target candidate. Note that the direction angle difference threshold value and the velocity vector threshold value for determining whether or not they are within the predetermined range are stored in the storage unit.

  In step 423, using the direction angle difference calculated in step 421, the peripheral moving objects that are candidates for tracking are ranked in order of increasing direction angle difference.

In step 424, the process branches depending on whether or not there is a follow target, that is, whether or not the present apparatus 1 is currently performing the follow operation. When the tracking target does not exist (No in step 424), the moving object having the first (top) rank added in step 423 in step 425 is selected as the tracking target, and its position coordinates (following target position coordinates 105a), speed A vector (following target speed vector 105b) is output. After the output, the entire process of selecting the follow target is finished.
On the other hand, if there is a follow target in step 424 (Yes in step 424), the direction angle difference of the moving object that is the follow target in step 426 is compared with the direction angle difference of the top moving object in step 423. To do.

In step 427, when the direction angle difference of the current tracking target object is θ, the direction angle difference of the top moving object is θn, and the threshold value for evaluating the change in the direction angle difference is ε (where ε is the user) Can be set arbitrarily)
| Θ |-| θn |> ε
If the condition is satisfied (Yes in step 427), that is, if the direction angle difference θ has improved beyond the set threshold value ε, the tracking target is changed and step 425 is executed. If not satisfied (No in step 427), that is, if the direction angle difference θ does not improve beyond the set threshold value ε, in step 428, the current position coordinates of the tracking target (tracking target position coordinates 105a), the velocity vector (Following target speed vector 105b) is output. In other words, the current follow-up is continued. After outputting, the entire process is terminated.

  As described above, according to the first embodiment, the autonomous mobile device 1 makes contact with a non-following target by following a moving object in a direction substantially equivalent to that of the device 1 through a region where a certain number of moving objects exist. Can be reduced. Furthermore, by providing a function to change the tracking target to another moving object even during tracking, even if the tracking target is deviated from the target direction of the apparatus 1, the time required for passing the region by changing the transfer is reduced, and contact with another object is achieved. It is possible to reduce the number of sections in which avoidance traveling with high risk is performed. That is, by using the autonomous mobile device of the present invention, it is possible to realize safe and high-speed traveling in an area where a large number of moving objects exist. As a result, for example, an outdoor autonomous mobile robot that can move efficiently in a place where people come and go can be configured.

<< Second Embodiment >>
Next, a second embodiment will be described. The second embodiment not only selects a moving object with a small directional angle difference in the method for selecting the tracking target, but also the candidate with the smallest directional angle difference is too far away from the candidate with the smallest directional angle difference. In order to cope with cases where it is inconvenient to follow, such as when the moving speed is too slow or too fast, the relative distance from the device 1 and the magnitude of the velocity vector are each multiplied by a coefficient that can be arbitrarily set as a weight term and evaluated. This is an embodiment in which a parameter is expressed as a score.
In addition, the same components as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted, and different portions are described by adding new numbers.

  FIG. 5 is a functional configuration diagram of the autonomous mobile device in the second embodiment. The tracking target selection unit 42b in FIG. 5 includes the function provided in the tracking target selection unit 42 of the first embodiment. The target position 101 is acquired from 41 and the target to be followed is selected. The detailed method will be described below. The follow target selection unit 42b includes a target direction calculation unit 421, a score calculation unit 422, and a score comparison unit 423. Moreover, although this apparatus 1 is provided with the target point input means 45, this target point input means 45 may be provided in other embodiment.

The target direction calculation means 421 in FIG. 5 acquires the self position coordinates 101 from the self position estimation means 41 and the target position coordinates 102 from the target point input means 45. Here, the target point input means 45 is an interface for inputting a target point (see reference numeral 10 in FIG. 1) that changes depending on the task of the apparatus 1. Further, the target position coordinate 102 is a position coordinate in the absolute coordinate system of the target point 10.
The target direction calculation means 421 derives and outputs a target direction vector 103 by subtracting the self-position coordinates 101 from the acquired target position coordinates 102.

The score calculation unit 422 acquires the object position coordinates 108 and the velocity vector 109 of the peripheral object from the moving object detection unit 40, and acquires the target direction vector 103 from the target direction calculation unit 421. A relative distance between the apparatus 1 and the object is obtained from the object position coordinates 108, and a direction angle difference is derived from the target direction vector 103 and the velocity vector 109.
By adding the weight term obtained by multiplying the relative distance value W by the coefficient α and the weight term obtained by multiplying the value V of the velocity vector 109 by the coefficient β to the direction angle difference value Θ, The following evaluation function E is created.
E = Θ + αW + βV
The coefficients α and β can be arbitrarily set by the user and are stored in the storage unit.
A score 104 for evaluating the priority order of the tracking target is calculated for each moving object from the evaluation function E and output. It should be noted that the mathematical formula of the evaluation function E can be appropriately changed in design to suit the purpose of the present invention. Further, for example, the higher the score 104, the higher the priority may be set.

  The score comparison unit 423 calculates and compares priorities when following each object using the score 104 of each object acquired from the score calculation unit 422, and selects a tracking target. Then, the tracking target position coordinates 105a and the tracking target speed vector 105b of the selected tracking target object are output. Subsequent processing is the same as in the first embodiment.

  By this process, for example, the object having the highest priority, including the object that is currently following, is selected as the tracking target, and the drive command value corresponding to the tracking target position coordinate 105a and the tracking target speed vector 105b of the selected object. 106 is output to the drive unit 3 to drive the apparatus 1. If the selected tracking target has a lower priority than the moving object currently being tracked, the tracking target may be controlled to continue without changing the tracking target.

  As described above, according to the second embodiment, it is impossible to select a candidate that is too far to follow even in an object having various velocities and relative distances, or a candidate that is too slow or too fast for the device 1. The risk of passing through can be reduced.

<< Third embodiment >>
Next, a third embodiment will be described. In order for the apparatus 1 that performs tracking and transfer of moving objects to exert its effect, it is necessary that a moving object heading in a direction substantially equivalent to the apparatus 1 exists in the vicinity. However, even if it follows such a moving object once, it does not make sense to follow if the moving object to be transferred after following is moving in the direction away from the target point.
In the third embodiment, the absolute position coordinates of the moving object (position coordinates of the moving object in the absolute coordinate system) and the velocity vector are added as members to each record in the external moving body information database. However, based on this information, a technique that makes it possible to set a travel route in an area where moving objects traveling in a direction substantially equivalent to that of the present apparatus 1 is gathered will be described. The area includes an area that cannot be sensed by the external sensor unit 2 of the apparatus 1.

FIG. 6 is a conceptual diagram illustrating a state in which a route to a target point of the autonomous mobile device is generated in an area where there are many moving objects. When the autonomous mobile device 1 travels toward the final target point 12, it needs to pass through a region where the density of moving objects is high. However, the external sensor unit 2 mounted on the apparatus 1 can only detect nearby objects, that is, the objects 201 to 205 in FIG. Therefore, when the tracking object is selected by the method of the first and second embodiments, the object 203 having the velocity vector that minimizes the direction angle difference from the target direction vector at the final target point 13 is selected as the tracking target. End up. However, the velocity vector of the object 203 is directed to a region where there are many objects having different target directions from the apparatus 1 such as the regions B and C in the figure, and there is no candidate to change when the object 203 is removed from the tracking target.
The area A is an area where there are many moving objects toward the final target point 12, the area B is an area where there are many moving objects coming toward the apparatus 1, and the area C is a black square. It is an area where there are many persons (moving objects) who are gathered at the street stall shown and are stationary. The reference vector indicates a direction that penetrates the region B from the position of the apparatus 1.

  Therefore, in the present embodiment, by obtaining (indirectly) the velocity vector of a moving object in a region where the device 1 cannot be sensed (located far beyond the sensing range) by the external mobile body information database, A target point X1 is set on the way to a region A where there are many moving objects going in substantially the same direction, and the apparatus 1 is controlled to pass through the region A and head toward the final target point 12. Specific execution means will be described below.

  FIG. 7 is a system configuration diagram of a network to which autonomous mobile devices are connected so as to be able to communicate with each other. In addition to the autonomous mobile device (this device) 1, this network (eg, wireless LAN (Local Area Network), wired network) is arranged as infrastructure equipment such as a server 9 having a mobile body information database 6 and a monitoring camera. One or more fixed monitoring devices (monitoring devices) 7 and one or more other autonomous mobile devices 8 that are autonomous mobile devices other than the autonomous mobile device (this device) 1 are communicably connected.

By providing the communication apparatus 5 in the apparatus 1, communication with the server 9 including the moving body information database 6 is possible, and as information, for example, absolute position coordinates of a moving object, a velocity vector, and the like are transmitted and received. The server 9 including the moving body information database 6 transmits and receives the absolute position coordinates and velocity vectors of the moving object from the fixed monitoring device 7 and the other autonomous mobile device 8 including the communication device, and stores the data. The server 9 is a computer having hardware resources such as an input unit, an output unit, a storage unit, and a control unit, and is disposed in a building in which a monitor for monitoring an area where a plurality of autonomous mobile devices come and go is provided, for example. .
In addition, the same components as those in the first and second embodiments are denoted by the same reference numerals and description thereof is omitted, and different parts are described by adding new numbers. The control unit 4 b includes functions common to the control unit 4.

  FIG. 8 is a functional configuration diagram of the autonomous mobile device of the third embodiment. This functional configuration diagram shows the data flow inside the control unit 4b. The apparatus 1 includes target position generation means 44 in addition to the means already described.

  The target position generation unit 44 acquires the absolute position coordinates 107a and the velocity vector 107b of the moving object (for example, another autonomous mobile device 8) registered in the moving body information database 6, and the self position coordinates 101 from the self position estimation unit 41. And the target position coordinates 102b are output as a target point on the way before reaching the final target point.

The processing inside the target position generation means 44 will be described in detail. The target position generation unit 44 includes a moving object classification unit 440 and a target candidate most frequent region selection unit 441.
The moving object classification means 440 uses the absolute position coordinates 107a and the velocity vector 107b acquired from the moving object information database 6, and the object position coordinates 108 and the velocity vector 109 of the moving object acquired from the moving object detection means 40 of the present apparatus 1. The direction angle difference of each moving object is derived. A moving object whose derived direction angle difference is equal to or smaller than a certain value is selected, and the absolute position coordinates of the selected moving object are output as candidate moving object coordinates 300. By this selection, it is possible to classify in which region there are moving objects that can and cannot be sensed.
In the present embodiment, the object position coordinate 108 is a relative coordinate with the apparatus 1, but the object position coordinate 108 can be handled as an absolute coordinate in a certain absolute coordinate system by using the self-position coordinate 101, for example. Therefore, in the present embodiment, the object position coordinates 108 are handled as absolute coordinates as necessary.

  The target candidate most frequent area selection unit 441 acquires candidate moving object coordinates 300 of a candidate moving object having a small directional angle difference (below a certain value) from the moving object classification unit 440, and the number of candidates within a certain range on the traveling road surface. Is calculated while changing the center position of the range, and the region in which the number of candidates is the maximum (the range for counting the included moving objects) is moving in a direction substantially equivalent to that of the apparatus 1 Judgment is made and the position coordinates of the center of the area are output as the target position coordinates 102b. A conceptual diagram of this method is shown in FIG.

FIG. 9 is a conceptual diagram showing a state in which a midpoint target point of the autonomous mobile device is generated in an area where there are many moving objects. A method of deriving the target position coordinates 102b in the target candidate most frequent region selecting unit 441 will be described.
Using the absolute position coordinates of the candidate moving object group 503, while counting the number of candidate objects 500 included in the range H having a certain size, the range H is moved relative to the traveling direction from the apparatus 1. Scanning is performed in the X-axis direction of the coordinate system at a distance L on the coordinate axis Y representing the position. The graph shown in the lower part of FIG. 9 is a graph (hereinafter sometimes referred to as “histogram”) in which the calculated number of candidate objects 500 is the vertical axis and the position coordinate 502 on the coordinate axis X is the horizontal axis. . A point 501 having the maximum number of candidate objects 500 is derived from the histogram data. Then, the center coordinates of the range H at that time are output as the target position coordinates 102b.

  Such a target position determination method is updated every time the apparatus 1 reaches the vicinity of the target position coordinates 102b set by the target position generating means 44, and the apparatus 1 enters within a certain distance of the final target point 12. Continue until. Then, it is possible to reach the final target point through an area where there are many objects moving in the same direction as the device 1.

  In addition, since the process by the other means of the autonomous mobile apparatus 1 is the same as the content demonstrated in 1st and 2nd embodiment, description of the process is abbreviate | omitted.

  As described above, according to the third embodiment, by using the information on the moving object registered in the moving body information database 6 in the autonomous mobile device 1, it is possible to grasp the motion of the moving object that exists in the region that cannot be sensed. It is possible to travel in an area where there are many possible candidates. As a result, it becomes easy to find the tracking target, and it is possible to reduce the risk of collision with other moving objects along the traffic flow and to reach the final target point in a short time.

  The above embodiment is a preferred embodiment of the present invention, but the form of implementation is not limited to that described above. Therefore, various modifications can be made to the implementation form without departing from the scope of the present invention.

  For example, in the first embodiment, processing is performed such that the smallest direction angle difference is picked up as a tracking target candidate (step 423 in FIG. 4). However, when the potential generation cannot be set appropriately and a follow-up path cannot be generated, the one having the smallest directional angle difference (or the third, fourth,...) Is taken as a follow target candidate. Subsequent processing may be executed.

  Further, in the second embodiment, processing is performed so that the object with the highest priority, including the moving object that is currently following, is the tracking target. However, when the potential generation cannot be set appropriately and a follow-up path cannot be generated, the lower priority order may be picked up as a follow-up target, and the subsequent processing may be executed.

  In the second embodiment, the moving object that is too slow or too fast is controlled not to be a tracking target. However, by changing the speed of the device 1, the moving object is controlled to follow as a tracking target. Also good. However, for a moving object that is too slow, if a tracking target that can reach the target point sooner is found, it is preferable to switch to the tracking target.

Further, in the third embodiment, the center of the range H in which the number of candidate objects 500 heading to the target point is maximized is set as the target position coordinate 102b, and the apparatus 1 is processed so as to go there (see FIG. 9). . However, if the potential generation cannot be set properly and the following path cannot be generated, the center of the range where the number of candidate objects 500 is the second (or third, fourth,...) Is the target. The position coordinates 102b may be used. Further, for example, there may be a predetermined number stored in the storage unit, for example, one or two or more ranges may be specified, and the center of the range may be set as the target position coordinates 102b.
When the range is specified, the target position is set for a range in which the number of candidate objects 500 is less than or equal to the predetermined rank stored in the storage unit or less than the predetermined number. Control may be performed so that the coordinate 102b is not set, and the target position coordinate 102b is narrowed down so as not to be subject to the follow-up operation.

  Further, in the third embodiment, by acquiring information (position and speed of a moving object, etc.) from another autonomous mobile device via a server, it is possible to detect information on a moving object in a region that cannot be detected by itself. did. However, it is not necessary to use a server to acquire such information. For example, each moving object has positioning means such as GPS (Global Positioning System) and communication means such as ad hoc communication so that each autonomous mobile device can share its position and speed, You may make it acquire speed. Further, a peer-to-peer communication method may be introduced.

  Further, a plurality of midway target points (target position coordinates 102b) set in the third embodiment may be provided so as to reach the final target point. In this case, it is possible to perform control so as to provide a convenient midway target point for reaching (provisionally) a certain midway target point.

  Further, in the present embodiment, the tracking method is performed using the potential method, and control is performed so as to generate a potential that generates repulsive force on a moving object other than the tracking target. In this case, control may be performed so that the potential is generated in all moving objects other than the tracking target, or control may be performed so that the potential is generated only in part.

  It goes without saying that an invention in which some or all of the invention-specific matters described in the embodiments are appropriately combined can be configured.

  In addition, software and hardware configuration changes, flowcharts, database designs, and the like can be changed as appropriate.

DESCRIPTION OF SYMBOLS 1 Autonomous mobile device 2 Outside world sensor part 3 Drive part 4, 4b Control part 6 Mobile body information database 7 Fixed monitoring apparatus (monitoring apparatus)
8 Other autonomous mobile devices 40 Moving object detection means 41 Self-position estimation means 42, 42b Tracking target selection means 43 Route generation means 44 Target position generation means 45 Target point input means 421 Target direction calculation means 422 Score calculation means 423 Score comparison means 440 Moving object classification means 441 Target candidate most frequent area selection means

Claims (6)

In an autonomous mobile device that moves in a region where a predetermined number or more of moving objects exist in the vicinity,
Control for calculating the velocity of the detected moving object in the absolute coordinate system by detecting and tracking the moving object;
Control for calculating an angle formed by the direction indicated by the calculated speed and the direction from the position of the autonomous mobile device to the target point of the autonomous mobile device for each detected moving object;
Control for selecting a moving object having the smallest calculated angle;
A potential field that is low enough to generate a predetermined attractive force with respect to the selected moving object, and a potential that is high enough to generate a predetermined repulsive force with respect to some or all of the moving objects other than the selected moving object Control that generates a field and follows the selected moving object as a tracking target;
When there is a moving object having the angle smaller than the angle indicated by the selected moving object during the tracking, the moving object is selected and the control to follow as the tracking target is executed. An autonomous mobile device characterized by Control that calculates the distance between the autonomous mobile device and the detected moving object, and determines the priority of tracking the moving object based on the calculated distance, the calculated angle, and the magnitude of the speed When,
2. The autonomous mobile device according to claim 1, wherein the moving object having the highest priority is selected and the control to follow as the follow target is executed. When the moving object with the highest priority cannot be set as the tracking target, whether a lower moving object is selected in the priority and the selected moving object can be set as the tracking target. The autonomous mobile device according to claim 2, wherein control for determining the tracking target is performed. The autonomous mobile device is communicably connected to a server having a database that registers the position and speed of a moving object in an area where the moving object cannot be detected in the absolute coordinate system for each moving object,
Control to obtain the position and speed of a moving object present in the undetectable area from the server;
Control that identifies a moving object that moves in the same direction as the direction from the position of the autonomous mobile device to the target point of the autonomous mobile device, among the detected moving object and the moving object that exists in the undetectable region When,
Control for setting a range for counting the moving objects to be included, and specifying a range in which the number of the specified moving objects exceeds a predetermined number;
The center of the specified range is set as an intermediate target point of the autonomous mobile device, a moving object that moves in the same direction as the direction from the position of the autonomous mobile device to the intermediate target point is selected and followed as the tracking target. The autonomous mobile device according to claim 1, wherein control is performed. The position and speed of the moving object acquired from the server are:
Acquired from one or both of one or more monitoring devices that are communicably connected to the server and arranged at a predetermined position, and one or more autonomous mobile devices that move in a predetermined area, and are registered in the database The autonomous mobile device according to claim 4, wherein: The control for setting the height of the potential field generated in the moving object to be avoided when the tracking is performed to be lower than that when the tracking is not performed is performed. 1. The autonomous mobile device according to 1.

Images