JP2018005470A - Autonomous mobile device, autonomous mobile method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To distinguish a movable object with other obstacles to immediately reflect movement of the movable object on a map.SOLUTION: An autonomous mobile device 100 comprises a depth sensor 31 for acquiring depth data on distance to an object, a movable object detection sensor 32 for detecting the movable object, and a control unit 10. The control unit 10 creates a map recording presence probability of an object for each predetermined area on the basis of the depth data that the depth sensor 31 has acquired and, if the movable object detection sensor 32 detects the movable object, changes the presence probability in the predetermined area where the movable object has been detected.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an autonomous mobile device, an autonomous mobile method, and a program.

  Autonomous mobile devices that move autonomously according to their use have become widespread. For example, an autonomous mobile device that moves autonomously for indoor cleaning is known. Such an autonomous mobile device often creates an occupied grid map as a map in real space. The occupied grid map is a map in which values of grid points indicate the presence of an object (obstacle).

  When creating an occupancy grid map, a distance to a surrounding object is detected using a depth (ranging) sensor, a ranging sensor, or the like. Since the distances detected by these sensors usually include an error, the map is often provided with hysteresis in order to reduce the influence of the error. For example, Patent Document 1 discloses a technique for giving hysteresis to a map by using the value of each grid point on the map as the existence probability of the object and expressing the existence of the object based on whether the existence probability is equal to or higher than a threshold value. Yes.

JP 2014-219723 A

  However, if the map is given hysteresis by the technique disclosed in Patent Document 1, even if the object (obstacle) disappears, it is not immediately reflected in the map. Therefore, when a movable object such as a person is recorded on the map as an obstacle, even if the movable object moves from the recorded location, it remains on the map for a while. In this case, since the autonomous mobile device that moves using this map moves so as to avoid obstacles, for example, when following a person, there is a problem that a route that avoids the person is selected.

  The present invention has been made to solve the above problem, and by distinguishing a movable object such as a person from other obstacles, it is possible to immediately reflect the movement of the movable object on a map. It is an object to provide a mobile device, an autonomous mobile method, and a program.

In order to achieve the above object, the autonomous mobile device of the present invention provides:
An autonomous mobile device that creates a map and moves based on the map,
With a control unit,
The controller is
Repeatedly confirm the existence of movable objects and other stationary objects,
The movable object existence probability and the stationary object existence probability are recorded for each predetermined area of the map, and the movable object existence probability and the stationary object existence probability are made different.

  According to the present invention, it is possible to immediately reflect the movement of the movable object on the map by distinguishing the movable object from other obstacles.

It is a figure which shows the function structure of the autonomous mobile apparatus which concerns on Embodiment 1 of this invention. (A) It is a figure which shows the example of a scan in a depth sensor. (B) It is a figure which shows the example of the depth data obtained with a depth sensor. It is a figure which shows the whole structure of the software module of the autonomous mobile apparatus which concerns on Embodiment 1. FIG. It is a flowchart of the map creation module which concerns on Embodiment 1. FIG. It is a figure which shows the specific example of the occupation grid map which the map preparation part which concerns on Embodiment 1 produces. 4 is a flowchart of a path planning module according to the first embodiment. 3 is a flowchart of a movement control module according to the first embodiment. It is a figure which shows the function structure of the autonomous mobile apparatus which concerns on Embodiment 2 of this invention. It is a flowchart of the map creation module which concerns on Embodiment 2. FIG. 10 is a flowchart of a route planning module according to the second embodiment. 10 is a flowchart of a movement control module according to the second embodiment.

  Hereinafter, an autonomous mobile device according to an embodiment of the present invention will be described with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals.

(Embodiment 1)
The autonomous mobile device according to the embodiment of the present invention is a device that autonomously moves according to the application while creating a surrounding map. Examples of this use include security monitoring, indoor cleaning, pets, and toys.

  As shown in FIG. 1, the autonomous mobile device 100 according to the first embodiment of the present invention includes a control unit 10, a storage unit 20, a sensor unit 30, an imaging unit 41, a drive unit 42, and a communication unit 43.

  The control unit 10 is configured by a CPU (Central Processing Unit) or the like, and by executing a program stored in the storage unit 20, each unit described later (position and orientation estimation unit 11, map creation unit 12, movable object region map change) The functions of the unit 13, the route plan unit 14, the movement control unit 15, the depth data acquisition unit 16, and the movable object presence direction acquisition unit 17) are realized. Moreover, the control part 10 is provided with a timer (not shown), and can count elapsed time.

  The storage unit 20 includes a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. Functionally, the image storage unit 21, the map storage unit 22, the depth data storage unit 23, and the movable object presence direction storage unit 24 are provided. including. The ROM stores programs executed by the CPU of the control unit 10 and data necessary for executing the programs in advance. The RAM stores data that is created or changed during program execution.

  The image storage unit 21 stores an image captured by the imaging unit 41. However, in order to save the storage capacity, it is not necessary to store all captured images. The autonomous mobile device 100 uses a plurality of images stored in the image storage unit 21 to create a map and estimate its own position through SLAM (Simultaneous Localization And Mapping) processing. Regarding the image used for estimating the position of the own device, information on the own device (the position and orientation of the own device) when the image is taken is stored together with information on the image.

  The map storage unit 22 stores a map created by the map creation unit 12 based on information from the SLAM process and the sensor unit 30 described later.

  The depth data storage unit 23 stores depth data for each angle. As shown in FIG. 2, the depth data is a depth sensor 31 (described later) and has a predetermined angular interval (for example, 1 °) in the horizontal plane around the depth sensor 31 (autonomous mobile device 100). This is data obtained by measuring the distance to an object existing in the direction. If there is no object in that direction, there will be no depth data for that angle.

  The movable object presence direction storage unit 24 stores a direction (movable object presence direction) in which the presence of a movable object is detected using a movable object detection sensor 32 described later.

  The sensor unit 30 includes a depth sensor 31 and a movable object detection sensor 32. The depth sensor 31 is a sensor capable of scanning while changing an angle and measuring a distance to an object existing in the direction of the angle. The range of angles to be scanned and the resolution of the angles may have various values depending on the depth sensor to be used. However, in this embodiment, the range of angles to be scanned is in the range of −30 ° to 30 °. The resolution is ΔA (for example, ΔA = 0.1 °).

  The movable object detection sensor 32 is a sensor that detects a movable object such as a person, and is an infrared camera (thermography), for example. If an infrared camera is used, not only humans but also pets and animals can be detected. Alternatively, the movable object may be detected by recognizing a movable object such as a person from an image acquired by the imaging unit 41 described later. When recognizing a movable object from an image acquired by the imaging unit 41, the independent movable object detection sensor 32 is unnecessary, and the imaging unit 41 also serves as the movable object detection sensor 32. In this case, it is possible to detect not only people and pets but also moving toys such as cars and trains, robots, other autonomous mobile devices 100, etc. by image recognition. Hereinafter, these persons, pets, moving toys, robots, other autonomous mobile devices 100 and the like that can be detected by the movable object detection sensor 32 are collectively referred to as movable objects. The sensor unit 30 is a sensor other than these, for example, a bumper sensor (not shown) that detects collision with another object, an angular velocity sensor or an acceleration sensor (not shown) that detects the movement of the device itself. ) Etc. may be provided.

  The imaging unit 41 includes a monocular imaging device (camera). The imaging unit 41 acquires an image (frame) at 30 fps, for example. The autonomous mobile device 100 performs autonomous movement while recognizing its own position and surrounding environment in real time by SLAM processing based on images sequentially acquired by the imaging unit 41.

  The drive unit 42 is an independent two-wheel drive type, and is a moving means including wheels and a motor. The autonomous mobile device 100 translates back and forth (translation) by driving the two wheels in the same direction, and rotates (changes direction) on the spot by driving the two wheels in the reverse direction. A swivel movement (translation + rotation (direction change) movement) can be performed by the changed drive. Each wheel is also equipped with a rotary encoder, which measures the rotational speed of the wheel with the rotary encoder and uses the geometric relationship such as the wheel diameter and the distance between the wheels to translate and move. The amount can be calculated.

For example, if the wheel diameter is D and the rotation speed is R (measured by a rotary encoder), the translational movement amount at the ground contact portion of the wheel is π · D · R. Further, if the wheel diameter is D, the distance between the wheels is I, the rotation speed of the right wheel is R _R , and the rotation speed of the left wheel is _RL , the rotation amount of the direction change is 360 (when the right rotation is positive). ° × D × (R _L −R _R ) / (2 × I). By sequentially adding the translational movement amount and the rotation amount, the drive unit 42 functions as a so-called odometry, and measures its own position (position and orientation with reference to the position and orientation at the start of movement). Can do. In this specification, the “orientation” of the own device is also referred to as the “posture” of the own device.

  In addition, you may make it provide a crawler instead of a wheel, and you may make it move by providing a plurality (for example, two) leg | foot and walking with a leg | foot. Also in these cases, the position and orientation (posture) of the own device can be measured based on the movements of the two crawlers and the movements of the feet, as in the case of the wheels.

  The communication unit 43 is a module for communicating with an external device, and is a wireless module including an antenna when wirelessly communicating with the external device. For example, the communication unit 43 is a wireless module for performing short-range wireless communication based on Bluetooth (registered trademark). By using the communication unit 43, the autonomous mobile device 100 can exchange data with the outside. For example, when instructing the destination to the autonomous mobile device 100, the destination information is transmitted via the communication unit 43.

  Next, functions of the control unit 10 will be described. The control unit 10 includes a position / orientation estimation unit 11, a map creation unit 12, a movable object region map change unit 13, a route plan unit 14, a movement control unit 15, a depth data acquisition unit 16, and a movable object presence direction acquisition unit 17. The movement control of the autonomous mobile device 100 is performed. Further, the control unit 10 corresponds to the multi-thread function, and can execute a plurality of threads (different processing flows) in parallel.

  The position / orientation estimation unit 11 estimates the position and orientation of its own device by SLAM processing using a plurality of images taken by the imaging unit 41 and stored in the image storage unit 21. In addition, when estimating the position and orientation, odometry information that can be acquired from the drive unit 42 can also be used.

  The map creation unit 12 creates an occupied grid map in which the position of an obstacle (object) is recorded using information from the depth sensor 31 and the movable object detection sensor 32, and stores it in the map storage unit 22. This occupied grid map is obtained by dividing a plane into grid points and recording “observation count”, “detection count”, and “object presence / absence” at each grid point. The “number of observations” is the number of times that the lattice point is observed by the depth sensor 31, and the “number of detections” is the number of times that the depth sensor 31 detects that an object exists at the position of the lattice point. In other words, it can be considered that the existence probability of the object at the lattice point obtained by “detection function” ÷ “number of observations” is recorded on the map. Further, the “presence / absence of object” takes three values: “with object”, “without object”, and “unobserved”.

  The movable object area map changing unit 13 can immediately delete the presence of the movable object from the map when the movable object existing in the map created by the map creating unit 12 is not moved. The map data corresponding to the area where the movable object exists is changed.

  The route planning unit 14 plans a route from the current position of the own device to the destination based on the map created by the map creating unit 12.

  The movement control unit 15 determines the moving direction and moving speed of the own machine based on the map created by the map creating unit 12 and the route planned by the route planning unit 14 and controls the driving unit 42.

  The depth data acquisition unit 16 acquires depth data (distance to an object) from the depth sensor 31 for each predetermined direction (angle). If there is no object in that direction, there is no depth data and it is not acquired.

  The movable object presence direction acquisition unit 17 acquires the direction in which the detected movable object exists (movable object presence direction) based on the detection information of the movable object detection sensor 32. If no movable object is detected, there is no information on the direction of presence of the movable object.

  The functional configuration of the autonomous mobile device 100 has been described above. Next, the overall configuration of the software module of the autonomous mobile device 100 will be described with reference to FIG. In FIG. 3, the position / orientation estimation module 51 corresponds to the position / orientation estimation unit 11 described in the above functional configuration, the map creation module 52 corresponds to the map creation unit 12, and the route planning module 53 corresponds to the route planning unit 14. The movement control module 54 corresponds to the movement control unit 15.

  When the autonomous mobile device 100 is powered on, these software modules are activated as separate threads, and are executed in parallel. The position / orientation estimation module 51 performs SLAM processing using the image information acquired from the imaging unit 41, and estimates the position and orientation of the own device also using odometry information acquired from the driving unit. Based on the position and orientation estimated by the position / orientation estimation module 51, the depth data acquired from the depth sensor 31, and the information on the direction of presence of the movable object acquired from the movable object detection sensor 32, the map creation module 52 Create

  The route planning module 53 plans a route based on the position and orientation estimated by the position / orientation estimation module 51 and the map created by the map creation module 52. The movement control module 54 controls the driving unit 42 based on the position and orientation estimated by the position / orientation estimation module 51, the map created by the map creation module 52, and the route planned by the route planning module 53. Information (mainly speed information) is generated. And the drive part 42 is driven based on the movement control information which the movement control module 54 produced | generated, and the movement to a destination is performed.

  Now, the processing contents of the map creation module 52 for creating a map that can immediately reflect the movement of the movable object will be described with reference to FIG.

  First, the map creation unit 12 initializes a map stored in the map storage unit 22 (step S101). The initialization of the map is performed by setting “observation count” and “detection count” at all grid points of the occupied grid map to 0 and “presence / absence of object” to “unobserved”. Next, the control part 10 determines whether the autonomous mobile apparatus 100 complete | finishes operation | movement (step S102). The autonomous mobile device 100 ends its operation when the user turns off the autonomous mobile device 100 or when the remaining battery level falls below a predetermined amount (for example, 3% remaining). If the autonomous mobile device 100 ends the operation (step S102; Yes), the process ends. If the operation is not finished (step S102; No), the depth data acquisition unit 16 acquires depth data while changing the angle of the depth sensor 31 from −30 ° to + 30 °, and stores it in the depth data storage unit 23 (step). S103).

  Next, the control unit 10 acquires the position and orientation of the own device estimated by the position / orientation estimation unit 11 at that time (step S104). Then, the movable object presence direction acquisition unit 17 acquires the direction (angle range) in which the movable object exists using the movable object detection sensor 32, and stores it in the movable object presence direction storage unit 24 (step S105). Next, in order to create a map based on the depth data at each angle, the control unit 10 sets A = −30 ° as an initial value of the angle (step S106). And the control part 10 determines whether the depth data in the angle A exists (step S107).

  If there is no depth data at the angle A (step S107; No), the map creating unit 12 adds 1 to the “number of observations” of all grid points corresponding to the angle A on the map (step S108). This process will be described with a specific example with reference to the occupation grid map 60 shown in FIG. If the line indicated by L1 in FIG. 5 is a line corresponding to the angle A from the depth sensor 31, the process of step S108 adds 1 to the “number of observations” of all grid points indicated by gray through which L1 in FIG. To do. Then, the process proceeds to step S112. Note that the lattice points on the line corresponding to the angle A can be obtained from the position and orientation values of the autonomous mobile device 100 acquired in step S104 and the polar coordinate values with the depth sensor 31 as the origin.

  If there is depth data at the angle A (step S107; Yes), the map creating unit 12 obtains the depth data at the angle A (position in polar coordinates with the depth sensor 31 as the origin) of the own device acquired in step S104. Using the position and orientation values, the coordinates are converted into the coordinates of the grid points, and the grid points where the object exists are acquired (step S109). Then, the map creation unit 12 adds 1 to the “number of detections” of the grid point where the object exists (step S110). Next, the map creation unit 12 adds 1 to the “number of observations” of all grid points from the depth sensor 31 to the object (step S111).

  This process will be described again with reference to FIG. If the line indicated by L2 in FIG. 5 is a line having depth data at the angle A, the process of step S110 adds 1 to the “detection count” of the grid point indicated by the dark gray on the left of the obstacle in FIG. To do. Then, in the process of step S111, 1 is added to the “number of observations” of all grid points indicated by gray and dark gray through which the line indicated by L2 in FIG. 5 passes. Then, the process proceeds to step S112. Note that all the lattice points from the depth sensor 31 to the object corresponding to the angle A are the values of the position and orientation of the own device acquired in step S104, the polar coordinate values with the depth sensor 31 as the origin, and the step S109. It can be obtained from the acquired grid point where the object exists.

  Then, the control unit 10 adds ΔA (depth sensor resolution, for example, 0.1 °) to the value of A (step S112), and determines whether A is 30 ° or less (step S113). If A is 30 ° or less (step S113; Yes), the process returns to step S107.

  If A exceeds 30 ° (step S113; No), in order to set “presence / absence of object” for all lattice points, the map creating unit 12 sets the position of the lattice point for setting “presence / absence of object”. Is initialized (step S114). Here, the lattice point is represented by (X, Y), and X and Y are set to 0 as initial values. Next, the map creating unit 12 determines whether or not the “number of observations” at the grid point (X, Y) is greater than 0 (step S115).

  If the “number of observations” of the grid point (X, Y) is 0 (step S115; No), the map creation unit 12 sets “presence / absence of object” of the grid point to “unobserved” (step S116). Proceed to step S122. Since the “presence / absence of object” of all grid points has been set to “unobserved” by the map initialization process in step S101, the process may proceed to step S122 without the process of step S116.

  If the “observation count” of the grid point (X, Y) is greater than 0 (step S115; Yes), the map creation unit 12 determines whether “detection count” ÷ “observation count” of the grid point is equal to or greater than a threshold value. Is determined (step S117). Here, the threshold value is for giving hysteresis to the map, and usually a value in the order of 0.5 to 0.8 (for example, 0.6) is set.

  If “the number of detections” ÷ “the number of observations” is less than the threshold value (step S117; No), the map creating unit 12 sets the “presence / absence of object” of the grid point to “no object” (step S118), and step S122. Proceed to

  If “the number of detections” ÷ “the number of observations” is equal to or greater than the threshold (step S117; Yes), the map creating unit 12 sets “the presence / absence of object” at the grid point to “there is object” (step S119). Then, the movable object area map changing unit 13 determines whether the lattice point (X, Y) is an area where a movable object exists (step S120). In this determination, the angle θ when the lattice point (X, Y) is converted into polar coordinates (radial radius r, angle θ) with the position and orientation of the autonomous mobile device 100 as a reference (the origin of polar coordinates) is determined in step S105. The determination can be made based on whether or not the acquired movable object is within an angle range.

  If the lattice point (X, Y) is not an area where a movable object exists (step S120; No), the process proceeds to step S122. On the other hand, if the grid point (X, Y) is an area where a movable object exists (step S120; Yes), the movable object area map changing unit 13 determines the “detection count” of the grid point (X, Y) at the point. “Observation count” × threshold value is set (step S121). If `` Observation count '' x threshold value is not an integer, but only an integer value can be set as `` Detection count '', the value obtained by rounding up the decimal part of `` Observation count '' x threshold value is set as the `` Detection count '' at that point. Set.

  Then, the map creating unit 12 advances the grid point to be set for “presence / absence of object” (step S122). In this processing, for example, 1 is added to X, and when X exceeds a specified range, X is returned to 0 and 1 is added to Y. Then, the map creation unit 12 determines whether or not the “presence / absence of object” of all grid points (X, Y) has been set (step S123). As described above, if 1 is added to X or Y in step S122 as described above, it may be determined whether or not the value of Y exceeds the specified range.

  If “presence / absence of object” for all grid points has not been set yet (step S123; No), the process returns to step S115, and if “presence / absence of object” for all grid points has been set (step S123; Yes). ), The process returns to step S102.

  With the above processing, the presence / absence of an actually observed object is set in the “presence / absence of object” of all lattice points observed by the depth sensor 31. In addition, since the “detection count” of the lattice point where the movable object is present, which is observed by the movable object detection sensor 32, is set to exceed the threshold value in the determination in step S117, if the movable object is no longer moving, As a result of the determination in S117, the value immediately falls below the threshold value, and in step S118, the lattice point is set to “no object”. Therefore, it is possible to immediately reflect on the map that the movable object has moved.

  Next, processing contents of the route planning module 53 for planning a route to the destination using the map created by the map creating module 52 will be described with reference to FIG.

  First, the control part 10 determines whether the autonomous mobile apparatus 100 complete | finishes operation | movement (step S201). If the operation is finished (step S201; Yes), the process is finished. If the operation is not terminated (step S201; No), the route planning unit 14 determines whether or not the destination is set (step S202). The destination may be set by the user of the autonomous mobile device 100 via the communication unit 43, or the autonomous mobile device 100 may need to set the destination (for example, the remaining battery level falls below a predetermined amount (for example, 10%)). In some cases, the charging station may be set autonomously (such as setting the charging station as the destination).

  If the destination has been set (step S202; Yes), the process proceeds to step S204 to obtain a map. If the destination is not set (step S202; No), the route planning unit 14 determines whether an error is notified from the movement control module 54 (step S203). If no error has been notified (step S203; No), the process returns to step S201. If an error is notified (step S203; Yes), it will progress to step S204 in order to acquire a map again.

  Then, the route planning unit 14 acquires the latest map created by the map creation module 52 at that time (step S204). Next, the path planning unit 14 acquires the current position and orientation of the own device estimated by the position / orientation estimation module 51 (step S205). Then, the route planning unit 14 plans a route from the current position to the destination based on the acquired map, the position and orientation, and the set destination (step S206).

  Then, the route planning unit 14 determines whether the route has been planned (whether the route exists) (step S207). If there is no path (step S207; No), error processing such as notifying the user of that is performed (step S208), and the process returns to step S201. If a route exists (step S207; Yes), the route plan unit 14 notifies the movement control module 54 of the planned route (step S209), and returns to step S201.

  The route to the destination is obtained by the processing of the route planning module 53 described above. Next, processing contents of the movement control module 54 that performs movement control to the destination using this route will be described with reference to FIG.

  First, the control part 10 determines whether the autonomous mobile apparatus 100 complete | finishes operation | movement (step S301). If the operation is finished (step S301; Yes), the process is finished. If the operation is not finished (step S301; No), the movement control unit 15 determines whether or not a route is set by the route planning module 53 (whether or not a route planned by the route planning module 53 is notified) (step S301; No). Step S302). If the route is not notified (step S302; No), it means that the route has not been set yet, and the process returns to step S301.

  If the route is notified (step S302; Yes), it means that the route has been set, and the movement control unit 15 acquires the route planned by the route planning module 53 (step S303). Next, the movement control unit 15 acquires the current position and orientation of the own device estimated by the position and orientation estimation module 51 (step S304). Then, the movement control unit 15 determines whether or not the own device has arrived at the destination (step S305). Since the route information includes destination information, the movement control unit 15 arrives at the destination by determining whether or not the current position matches the destination included in the route. It can be determined whether or not.

  If the destination has been reached (step S305; Yes), the process returns to step S301. If the destination has not been reached (step S305; No), the movement control unit 15 acquires the latest map created by the map creation module 52 at that time (step S306). And the movement control part 15 determines whether it can move along a path | route based on the acquired map and path | route (step S307).

  If movement along the route is not possible (step S307: No), an error is notified to the route planning module 53 (step S308), and the process returns to step S301. If it can move along the route (step S307; Yes), the drive unit 42 is controlled to move along the route (step S309), and the process returns to step S304.

  Through the processing of the movement control module 54 described above, the autonomous mobile device 100 can move to the destination. Since the map created by the map creation module 52 is a map that immediately reflects the movement of the movable object, even if a movable object such as a person has existed in the past, the movable object currently exists. If it is not a place, the route planning module 53 can plan a route that does not avoid such a place, and the autonomous mobile device 100 can be prevented from performing an unnecessary avoidance operation.

(Modification 1 of Embodiment 1)
In the processing of the map creation module 52 of the first embodiment, in step S121 in FIG. 4, “number of times of detection” ÷ “number of times of observation” matches the threshold value by setting “number of times of observation” × threshold value to “number of times of detection”. However, the “number of detections” may be set so that “the number of detections” ÷ “the number of observations” is slightly larger than the threshold. In this way, a small hysteresis can be generated even for the presence of a movable object. That is, after that, even if the detection of the object by the depth sensor 31 is not performed once or twice at the place where the movable object exists, “the number of detections” ÷ “the number of observations” can be kept above the threshold value. Although the depth sensor 31 is rarely able to detect the presence of the object even though the object is present, if a small hysteresis is caused by such processing, the detection of the depth sensor 31 is not detected. Even if there is, you can create a map without losing sight of the presence of moving objects.

(Modification 2 of Embodiment 1)
In the first embodiment, the “number of detections” of the lattice point corresponding to the place where the movable object exists is changed so as to exceed the threshold value in the determination in step S117. However, at each grid point, the correct existence probability of the object at that grid point is obtained based on “number of detections” ÷ “number of observations”. Therefore, it is considered that there is a case where it is desired to keep the correct existence probability as it is without changing the “number of detections”. In such a case, the map information includes “threshold” in addition to “observation count”, “detection count”, “presence / absence of object”, and “threshold value” of the grid point where the movable object is detected. May be set to a “threshold for movable object” having a larger value (eg, 0.9) than a normal threshold value (eg, 0.6). In this case, when the map is initialized, a normal threshold is set as the “threshold” for each grid point.

  By doing this, even if the “observation count” and “detection count” are not changed, as in the first embodiment, if the movable object is not moved, the determination in step S117 is “detection count” ÷ “ The “number of observations” immediately falls below the “threshold” (threshold for movable objects). Therefore, it is possible to immediately reflect on the map that the movable object has moved. Moreover, at each grid point, the correct existence probability of the object at that grid point can be obtained based on “number of detections” ÷ “number of observations”. For a grid point whose “threshold value” is changed to “threshold value for movable object”, if the “number of detections” ÷ “number of observations” of the grid point thereafter falls below a normal threshold value (for example, 0.6), “threshold value” Is returned from the movable object threshold value to the normal threshold value.

(Embodiment 2)
In the autonomous mobile device 100 according to the first embodiment, the map created by the map creation module 52 is a map that immediately reflects the movement of the movable object, but the movable object and other stationary objects are not distinguished. Is remembered. Therefore, the route planning module 53 plans a route that avoids a movable object that moves and immediately disappears depending on the timing of acquiring a map. In this case, the autonomous mobile device 100 performs a movement for avoiding unnecessary. In order to avoid such a situation, it is conceivable to plan a route based only on a stationary object other than a movable object. When moving along a path based only on a stationary object, there is a possibility of colliding with a movable object. However, in order to prevent a collision, the autonomous mobile device may perform movement control to avoid the movable object during the movement. . A second embodiment based on this concept will be described.

  As shown in FIG. 8, the functional configuration of the autonomous mobile device 101 according to the second embodiment of the present invention is the same as the functional configuration of the autonomous mobile device 100 according to the first embodiment. The storage unit 25 is added. Other configurations are the same as those of the autonomous mobile device 100 according to the first embodiment. However, the “presence / absence of object” recorded at each grid point by the map storage unit 22 takes the values of “with object”, “without object”, “unobserved” and “with movable object”, so there are four types in total. Take the value of

  The movable object-free map creating unit 18 uses the information from the depth sensor 31 and the movable object detection sensor 32 to create an occupied grid map in which the position of the obstacle (object) is recorded while ignoring the presence of the movable object. Then, the map is stored in the movable object-free map storage unit 25. That is, even if the depth sensor 31 detects an object, if the movable object detection sensor 32 detects that the object is a movable object, an occupied grid map is created assuming that the object does not exist. Similarly to the occupied grid map created by the map creating unit 12, this occupied grid map also divides the plane into grid points and records “number of observations”, “number of detections”, and “presence / absence of object” at each grid point. Is.

  The map created by the movable object-free map creating unit 18 is stored in the movable object-free map storage unit 25.

  The overall configuration of the software module of the autonomous mobile device 101 according to Embodiment 2 is the same as the overall configuration of the software module of the autonomous mobile device 100 according to Embodiment 1 shown in FIG.

  Now, the processing content of the map creation module 52 according to the second embodiment will be described with reference to FIG. However, since this processing content overlaps considerably with the processing content of the map creation module 52 according to the first embodiment described with reference to FIG.

  First, steps S401 to S409 are the same as steps S101 to S109 in FIG.

  In step S410, the movable object-free map creation unit 18 determines whether or not the lattice point where the object acquired in step S409 exists is an area where the movable object exists. This determination can be made based on whether or not the angle A is within the range of the angle at which the movable object obtained in step S405 exists.

  If the lattice point where the object exists is an area where the movable object exists (step S410; Yes), the process proceeds to step S413. If the grid point where the object exists is not the area where the movable object exists (step S410; No), the map creating unit 12 adds 1 to the “detection count” of the grid point where the object exists (step S411). Next, the map creating unit 12 adds 1 to the “number of observations” of all grid points from the depth sensor 31 to the object (step S412).

  Steps S413 to S420 are the same as steps S112 to S119 in FIG. Step S421 is the same as step S122 in FIG.

  In step S422, the map creating unit 12 determines whether or not the “presence / absence of object” of all grid points (X, Y) has been set. If “presence / absence of object” for all grid points has not been set yet (step S422; No), the process returns to step S416, and if “presence / absence of object” for all grid points has been set (step S422; Yes). ) The movable object-free map creating unit 18 stores the map created by the map creating unit 12 so far in the movable object-free map storage unit 25 (step S423).

  Next, in order to create a map including a movable object based on the depth data at each angle, the control unit 10 sets A = −30 ° as an initial value of the angle (step S424). And the control part 10 determines whether the depth data in the angle A exists (step S425). If there is no depth data at the angle A (step S425; No), the process proceeds to step S429.

  If there is depth data at the angle A (step S425; Yes), the map creating unit 12 acquires the depth data at the angle A (position in polar coordinates with the depth sensor 31 as the origin) of the own device acquired in step S404. Using the position and orientation values, the coordinates are converted into the coordinates of the grid points, and the grid points where the object exists are acquired (step S426). Then, it is determined whether or not the lattice point where the object exists is an area where the movable object exists (step S427). This determination can be made based on whether or not the angle A is within the range of the angle at which the movable object obtained in step S405 exists.

  If the lattice point where this object exists is not the area where the movable object exists (step S427; No), the process proceeds to step S429. If this lattice point is an area where a movable object exists (step S427; Yes), the “presence / absence of object” of the lattice point is set to “movable object present” (step S428).

  In step S429, the control unit 10 adds ΔA (depth sensor resolution, for example, 0.1 °) to the value of A. And the control part 10 determines whether A after addition is 30 degrees or less (step S430). If A is 30 ° or less (step S430; Yes), the process returns to step S425. If A exceeds 30 ° (step S430; No), the map creation unit 12 stores the occupied grid map stored in the map storage unit 22 as a movable object presence map (step S431), and the process proceeds to step S402. Return.

  Through the above processing, two maps are created: a map without moving object that ignores the presence of the moving object, and a map with moving object in which all obstacles (objects) including the moving object are recorded. Next, processing contents of the route planning module 53 according to the second embodiment will be described with reference to FIG. However, since this processing content has a considerable overlap with the processing content of the route planning module 53 according to the first embodiment described with reference to FIG.

  First, step S501 is the same process as step S201 of FIG. Next, the route planning unit 14 determines whether or not a destination has been set (step S502). If the destination is set (step S502; Yes), the route plan unit 14 acquires the movable object-free map created by the map creation module 52 (step S505), and proceeds to step S506.

  If the destination is not set (step S502; No), the route planning unit 14 determines whether an error is notified from the movement control module 54 (step S503). If no error has been notified (step S503; No), the process returns to step S501. If an error is notified (step S503; Yes), the route planning unit 14 acquires the latest moving object presence map created by the map creation module 52 at that time (step S504), and proceeds to step S506.

  Steps S506 to S510 are the same processes as steps S205 to S209 in FIG.

  Through the processing of the route planning module 53 described above, it is possible to plan an efficient route ignoring the presence of a movable object using a map without a movable object at first. If the movement control module 54 is not able to move along the route due to the presence of the movable object, an error is notified from the movement control module 54. Therefore, a route avoiding the movable object can be planned based on the movable object presence map at that time. Next, processing contents of the movement control module 54 that performs movement control to the destination using this route will be described with reference to FIG.

  However, the processing content of the movement control module 54 according to the second embodiment is the same as the processing content of the movement control module 54 according to the first embodiment described with reference to FIG. 7 except for step S606. Will be explained. Since only one occupied grid map is used in the movement control module 54 according to the first embodiment, the single occupied grid map is acquired in step S306. In the movement control module 54 according to the second embodiment, there are two types of occupied grid maps: a map without a movable object and a map with a movable object. In step S606, a map with a movable object is acquired. Except for step S606, the process of the movement control module 54 according to the second embodiment shown in FIG. 11 is the same as the process of the movement control module 54 according to the first embodiment shown in FIG.

  With the above processing, in the case of movement control, it is possible to control the drive unit 42 in consideration of the presence of a movable object, and to move so as not to hit the movable object. In addition, since the route is initially planned using a map without moving objects, an efficient route can be planned ignoring the presence of moving objects.

(Modification 1)
In the embodiment described above, when there is no depth data at the angle A, 1 is added to the number of observations of all points corresponding to the angle A (gray points through which a line indicated by L1 in FIG. 5 passes). However, depending on the depth sensor 31, there is a measurable distance range, and when it is impossible to distinguish whether the object (obstacle) is closer or farther than that range, there is no depth data, and the object (obstacle) The location of can't be recorded on the map. In such a case, step S108 of the map creation module 52 of the first embodiment in FIG. 4 and step S408 of the map creation module 52 of the second embodiment in FIG. 9 may be omitted. Since the processing of step S108 and step S408 takes a considerable amount of processing time when the map is large, the processing of the map creating module 52 can be speeded up by omitting this processing.

(Modification 2)
In any of the embodiments described above, the existence probability p of an object (obstacle) is calculated by “number of detections” ÷ “number of observations”, and whether or not an object is present is determined based on whether p is equal to or greater than a threshold value. doing. In this case, division needs to be performed to calculate the existence probability p, but division requires more calculation time than addition and subtraction. Therefore, by using the concept of log odds, it is also possible to make the processing unnecessary division. Logarithmic odds l is obtained by log (p / (1-p)), and the range of the value of existence probability p is 0 to 1, but the range of values of logarithmic odds l corresponding to this is −∞ to ∞. It becomes. For this reason, the existence probability p and its threshold value need to be treated as decimal values, but the logarithmic odds l and its threshold value corresponding to this can use integer values, and from this point of view, the processing can be considerably speeded up. Become. In addition, by using log odds, not only the speed effect but also the effect of preventing the occurrence of a digit loss when the existence probability is updated near 0 or 1 is obtained.

  Processing in the case of using the logarithmic odds concept will be described focusing on differences in processing of the map creation module 52 (FIG. 4) of the first embodiment. In addition, since it contributes to speeding up, the process of step S108 described in the description of the modification 1 is omitted at the same time. When the logarithmic odds concept is used, “logarithmic odds” are recorded instead of “observation count” and “detection count” at each grid point of the occupied grid map 60 stored in the map storage unit 22. In step S101, “log odds” is set to an initial value 0, and “presence / absence of object” is set to “unobserved”.

  When an object is detected, in step S110, 2 is added to "logarithmic odds" instead of adding 1 to "number of detections". In step S111, instead of adding 1 to the “number of observations” of the grid point from the depth sensor 31 to the object, 2 is subtracted from the “log odds” of the grid point from the depth sensor 31 to the object. That is, 2 is subtracted from the “logarithmic odds” of the gray grid point among the grid points through which the line indicated by L2 in FIG. 5 passes, and 2 is added to the “logarithmic odds” of the dark gray grid point. Note that the value 2 added or subtracted here is an example, and other values may be used. The value to be added may be different from the value to be subtracted.

  For the determination of the presence / absence of an object, two thresholds are prepared: an object presence determination threshold (for example, 1) and an object absence determination threshold (for example, -1). In the determination in step S115, if “log odds” is equal to or greater than the object absence determination threshold and less than the object presence determination threshold, it is determined as an “unobserved region” and the process proceeds to step S116. In step S117, if “logarithmic odds” is less than the object absence determination threshold, it is determined as “object absence region”, and the process proceeds to step S118. If “logarithmic odds” is equal to or larger than the object presence determination threshold, it is determined as “object presence region”, and the process proceeds to step S119. In step S121, an object presence determination threshold is set in “logarithmic odds”. The other processes are the same as those of the map creation module 52 of the first embodiment shown in FIG. In this way, by introducing the concept of log odds, it becomes possible to determine the presence or absence of an object only by adding and subtracting integers, and the processing of the map creation module 52 can be speeded up.

  In the second modification, the method of introducing the logarithmic odds concept based on the first embodiment has been described, but in the second embodiment, the logarithmic odds concept can be obtained by processing the object presence / absence determination in the same manner as described above. Can be introduced.

  The “existence probability” in the present invention is referred to as “existence probability” for convenience as a reference value for determining the presence / absence of an object at each lattice point, and must be an accurate existence probability of an object. There is no. Therefore, in all of the above-described embodiments and all of the modifications, a combination of “number of detections” and “number of observations” recorded at each grid point and “log odds” are also referred to as existence probabilities.

  In the above-described embodiment, the range of the angle scanned by the depth sensor 31 is −30 ° to 30 °, but is not limited thereto. When using the depth sensor 31 that can scan 360 ° in all directions, the scan range may be set to −180 ° to 180 °.

  In each of the above-described embodiments, the image captured by the image capturing unit 41 is stored in the image storage unit 21, and the map is created or automatically created by SLAM processing using a plurality of images stored in the image storage unit 21. The position of the aircraft is estimated. However, even if SLAM processing is not performed, position estimation and map creation can be performed using odometry information from the drive unit 42 and depth data for each angle from the depth sensor 31. Therefore, when the SLAM process is not performed, the autonomous mobile devices 100 and 101 do not need to include the imaging unit 41 and the image storage unit 21.

  Note that the functions of the autonomous mobile devices 100 and 101 of the present invention can also be implemented by a computer such as a normal PC (Personal Computer). Specifically, in the above embodiment, the autonomous movement control processing program performed by the autonomous mobile devices 100 and 101 has been described as being stored in advance in the ROM of the storage unit 20. However, the program is stored and distributed on a computer-readable recording medium such as a flexible disk, CD-ROM (Compact Disc Read Only Memory), DVD (Digital Versatile Disc), and MO (Magneto-Optical Disc). A computer capable of realizing each of the functions described above may be configured by reading and installing the program on a computer.

  As mentioned above, although preferable embodiment of this invention was described, this invention is not limited to the specific embodiment which concerns, This invention includes the invention described in the claim, and its equivalent range It is. Hereinafter, the invention described in the scope of claims of the present application will be appended.

(Appendix 1)
An autonomous mobile device that creates a map and moves based on the map,
With a control unit,
The controller is
Repeatedly confirm the existence of movable objects and other stationary objects,
Autonomy characterized by recording the existence probability of the movable object and the existence probability of the stationary object for each predetermined area of the map, and making the existence probability of the movable object different from the existence probability of the stationary object Mobile equipment.

(Appendix 2)
The movable object is a person, a living thing, or a movable object,
The autonomous mobile device according to attachment 1.

(Appendix 3)
An autonomous mobile device that creates a map and moves based on the map,
A depth sensor that acquires depth data that is the distance to the object;
A movable object detection sensor for detecting a movable object;
A control unit,
The controller is
Based on the depth data acquired by the depth sensor, for each predetermined area, create a map that records the existence probability of the object,
When the movable object detection sensor detects the movable object, the existence probability in the predetermined region where the movable object is detected is changed.
Autonomous mobile device.

(Appendix 4)
The controller is
Let the depth sensor check the presence or absence of an object for each predetermined direction, if there is an object, to obtain depth data that is the distance to the object,
Causing the movable object detection sensor to detect a movable object, and based on information detected by the movable object detection sensor, obtaining a direction in which the detected movable object exists;
Creating a map for recording the existence probability based on the depth data of the object in the predetermined area determined by the depth data acquired by the depth sensor;
Changing the existence probability in the predetermined area corresponding to the movable object existing in the acquired direction;
The autonomous mobile device according to attachment 3.

(Appendix 5)
The map created by the control unit represents whether or not the object is present in the predetermined region depending on whether the existence probability is equal to or higher than a predetermined threshold.
The control unit changes the existence probability in the predetermined region corresponding to the movable object existing in the acquired direction to the predetermined threshold.
The autonomous mobile device according to attachment 4.

(Appendix 6)
The map created by the control unit represents whether or not the object is present in the predetermined region depending on whether the existence probability is equal to or higher than a predetermined threshold.
The control unit changes the existence probability in the predetermined area corresponding to the movable object existing in the acquired direction to a value larger than the predetermined threshold by a predetermined value;
The autonomous mobile device according to attachment 4.

(Appendix 7)
The map created by the control unit represents whether or not the object is present in the predetermined region depending on whether the existence probability is equal to or higher than a predetermined threshold.
The control unit changes the threshold value in the predetermined region corresponding to the movable object existing in the acquired direction to a predetermined movable object threshold value.
The autonomous mobile device according to attachment 4.

(Appendix 8)
The controller is
In a direction other than the acquired direction, a movable object-free map that represents the presence or absence of an object other than the movable object in the region is created based on the existence probability in the predetermined region determined by the depth data acquired by the depth sensor. And
The presence probability of the predetermined area corresponding to the movable object existing in the acquired direction is changed by the movable object-free map.
The autonomous mobile device according to attachment 4.

(Appendix 9)
The controller is
Plan a route to the destination based on the map and the moving object-free map,
Control the movement of the aircraft based on the route,
Immediately after the destination is set, the route to the destination is planned based on the movable object-free map.
The autonomous mobile device according to appendix 8.

(Appendix 10)
The control unit, when it is not possible to control the movement of its own device in the route planned based on the map without moving objects, plans a route based on the map,
The autonomous mobile device according to attachment 9.

(Appendix 11)
The controller is
Recording the existence probability on a map as a ratio of the number of times the depth data is acquired to the number of times the object is confirmed
The autonomous mobile device according to any one of appendices 4 to 10.

(Appendix 12)
The controller is
Recording the probability of existence on a map with logarithmic odds;
The autonomous mobile device according to any one of appendices 3 to 10.

(Appendix 13)
An autonomous movement method for creating a map and moving based on the created map,
When the movable object is detected, the map of the area where the movable object is detected is changed.
Autonomous movement method.

(Appendix 14)
A depth data acquisition step of confirming the presence or absence of an object for each predetermined direction and acquiring depth data that is a distance to the object when the object exists;
A movable object presence direction acquisition step of detecting a movable object and acquiring a direction in which the detected movable object exists;
A map creation step of creating a map indicating the presence or absence of the object in the predetermined region based on the existence probability based on the depth data of the object in the predetermined region determined by the depth data acquired in the depth data acquisition step;
A movable object region map changing step of changing the map of the predetermined region corresponding to the movable object existing in the direction acquired in the movable object presence direction acquiring step;
An autonomous movement method comprising:

(Appendix 15)
On the computer,
A depth data acquisition step of confirming the presence or absence of an object for each predetermined direction and acquiring depth data that is a distance to the object if the object exists;
A movable object presence direction acquisition step of detecting a movable object and acquiring a direction in which the detected movable object exists;
A map creating step for creating a map representing the presence or absence of the object in the predetermined region based on the existence probability based on the depth data of the object in the predetermined region determined by the depth data acquired in the depth data acquiring step;
A movable object region map change step for changing the map of the predetermined region corresponding to the movable object existing in the direction acquired in the movable object presence direction acquisition step;
A program for running

DESCRIPTION OF SYMBOLS 100, 101 ... Autonomous mobile device, 10 ... Control part, 11 ... Position and orientation estimation part, 12 ... Map preparation part, 13 ... Movable object area map change part, 14 ... Path planning part, 15 ... Movement control part, 16 ... Depth Data acquisition unit, 17 ... movable object presence direction acquisition unit, 18 ... movable object-free map creation unit, 20 ... storage unit, 21 ... image storage unit, 22 ... map storage unit, 23 ... depth data storage unit, 24 ... movable object Presence direction storage unit, 25 ... Map storage unit without moving object, 30 ... Sensor unit, 31 ... Depth sensor, 32 ... Movable object detection sensor, 41 ... Imaging unit, 42 ... Drive unit, 43 ... Communication unit, 51 ... Position and orientation Estimating module 52 ... Map creation module 53 ... Route planning module 54 ... Movement control module 60 ... Occupied grid map

Claims (15)

An autonomous mobile device that creates a map and moves based on the map,
With a control unit,
The controller is
Repeatedly confirm the existence of movable objects and other stationary objects,
Autonomy characterized by recording the existence probability of the movable object and the existence probability of the stationary object for each predetermined area of the map, and making the existence probability of the movable object different from the existence probability of the stationary object Mobile equipment. The movable object is a person, a living thing, or a movable object,
The autonomous mobile device according to claim 1. An autonomous mobile device that creates a map and moves based on the map,
A depth sensor that acquires depth data that is the distance to the object;
A movable object detection sensor for detecting a movable object;
A control unit,
The controller is
Based on the depth data acquired by the depth sensor, for each predetermined area, create a map that records the existence probability of the object,
When the movable object detection sensor detects the movable object, the existence probability in the predetermined region where the movable object is detected is changed.
Autonomous mobile device. The controller is
Let the depth sensor check the presence or absence of an object for each predetermined direction, if there is an object, to obtain depth data that is the distance to the object,
Causing the movable object detection sensor to detect a movable object, and based on information detected by the movable object detection sensor, obtaining a direction in which the detected movable object exists;
Creating a map for recording the existence probability based on the depth data of the object in the predetermined area determined by the depth data acquired by the depth sensor;
Changing the existence probability in the predetermined area corresponding to the movable object existing in the acquired direction;
The autonomous mobile device according to claim 3. The map created by the control unit represents whether or not the object is present in the predetermined region depending on whether the existence probability is equal to or higher than a predetermined threshold.
The control unit changes the existence probability in the predetermined region corresponding to the movable object existing in the acquired direction to the predetermined threshold.
The autonomous mobile device according to claim 4. The map created by the control unit represents whether or not the object is present in the predetermined region depending on whether the existence probability is equal to or higher than a predetermined threshold.
The control unit changes the existence probability in the predetermined area corresponding to the movable object existing in the acquired direction to a value larger than the predetermined threshold by a predetermined value;
The autonomous mobile device according to claim 4. The map created by the control unit represents whether or not the object is present in the predetermined region depending on whether the existence probability is equal to or higher than a predetermined threshold.
The control unit changes the threshold value in the predetermined region corresponding to the movable object existing in the acquired direction to a predetermined movable object threshold value.
The autonomous mobile device according to claim 4. The controller is
In a direction other than the acquired direction, a movable object-free map that represents the presence or absence of an object other than the movable object in the region is created based on the existence probability in the predetermined region determined by the depth data acquired by the depth sensor. And
The presence probability of the predetermined area corresponding to the movable object existing in the acquired direction is changed by the movable object-free map.
The autonomous mobile device according to claim 4. The controller is
Plan a route to the destination based on the map and the moving object-free map,
Control the movement of the aircraft based on the route,
Immediately after the destination is set, the route to the destination is planned based on the movable object-free map.
The autonomous mobile device according to claim 8. The control unit, when it is not possible to control the movement of its own device in the route planned based on the map without moving objects, plans a route based on the map,
The autonomous mobile device according to claim 9. The controller is
Recording the existence probability on a map as a ratio of the number of times the depth data is acquired to the number of times the object is confirmed
The autonomous mobile device according to any one of claims 4 to 10. The controller is
Recording the probability of existence on a map with logarithmic odds;
The autonomous mobile device according to any one of claims 3 to 10. An autonomous movement method for creating a map and moving based on the created map,
When the movable object is detected, the map of the area where the movable object is detected is changed.
Autonomous movement method. A depth data acquisition step of confirming the presence or absence of an object for each predetermined direction and acquiring depth data that is a distance to the object when the object exists;
A movable object presence direction acquisition step of detecting a movable object and acquiring a direction in which the detected movable object exists;
A map creation step of creating a map indicating the presence or absence of the object in the predetermined region based on the existence probability based on the depth data of the object in the predetermined region determined by the depth data acquired in the depth data acquisition step;
A movable object region map changing step of changing the map of the predetermined region corresponding to the movable object existing in the direction acquired in the movable object presence direction acquiring step;
An autonomous movement method comprising: On the computer,
A depth data acquisition step of confirming the presence or absence of an object for each predetermined direction and acquiring depth data that is a distance to the object if the object exists;
A movable object presence direction acquisition step of detecting a movable object and acquiring a direction in which the detected movable object exists;
A map creating step for creating a map representing the presence or absence of the object in the predetermined region based on the existence probability based on the depth data of the object in the predetermined region determined by the depth data acquired in the depth data acquiring step;
A movable object region map change step for changing the map of the predetermined region corresponding to the movable object existing in the direction acquired in the movable object presence direction acquisition step;
A program for running

Images