JP2021103593A - Autonomous mobile device, map information processing method, and program - Google Patents

Abstract

To improve a technology of creating a map of an autonomous mobile device.SOLUTION: An autonomous mobile device 100 includes a barrier detection part 30, a map creation part 12, a barrier erasing part 13, and a route setting part 14. The barrier detection part 30 detects a barrier. The map creation part 12 records, into an environment map, information of the barrier detected by the barrier detection part 30. The barrier erasing part 13 erases, from the environment map, the information of the barrier recorded by the map creation part 12. The route setting part 14 sets a moving route based on the information recorded in the environment map.SELECTED DRAWING: Figure 1

Description

The present invention relates to a technique for creating a map and autonomously moving.

Autonomous mobile devices that move autonomously according to the application have become widespread. For example, autonomous mobile devices that move autonomously for indoor cleaning are known. Such an autonomous mobile device can detect the distance to an object (obstacle) existing in the surroundings by using a range sensor or the like, and create a map showing where the obstacle exists in the surrounding environment. There are many. Since some obstacles move, technological developments are being made to deal with the moving obstacles. For example, Patent Document 1 states that by removing a moving obstacle ghost (data that remains in the past position of the obstacle and appears to still exist there) from the map. An autonomous mobile device capable of efficient movement is described.

Japanese Unexamined Patent Publication No. 2010-102485

The autonomous moving device described in Patent Document 1 can determine whether or not an obstacle detected by a sensor is moving, and if it is moving, can remove the ghost of the obstacle from the map. However, ghosts of obstacles that cannot be detected by the sensor cannot be removed from the map. Therefore, in the conventional autonomous mobile device, there is room for improvement in the technique for creating a map.

An object of the present invention is to improve a technique for creating a map of an autonomous mobile device.

In order to achieve the above object, the autonomous mobile device of the present invention
Obstacle detection unit that detects obstacles and
A cartography unit that records information on obstacles detected by the obstacle detection unit on an environmental map,
An obstacle erasing unit that erases obstacle information recorded by the map-creating unit from the environmental map over time, and an obstacle erasing unit.
A route setting unit that sets a movement route based on the information recorded on the environment map, and
To be equipped.

According to the present invention, it is possible to improve the technique for creating a map of an autonomous mobile device.

It is a figure which shows the functional structure of the autonomous mobile device which concerns on Embodiment 1 of this invention. It is a figure which shows the appearance of the autonomous mobile device which concerns on Embodiment 1. FIG. (A) It is a figure which shows the scan example by a range sensor. (B) It is a figure which shows the example of the distance measurement data obtained by the range sensor. It is a figure which shows the whole structure of the software module of the autonomous mobile device which concerns on Embodiment 1. FIG. It is a figure explaining the problem which the autonomous mobile device which concerns on Embodiment 1 solves. It is a flowchart of the map making module which concerns on Embodiment 1. It is a flowchart of the area sensor map update process which concerns on Embodiment 1. It is a flowchart of the collision sensor map update process which concerns on Embodiment 1. It is a flowchart of the route setting module which concerns on Embodiment 1. It is a flowchart of the movement control module which concerns on Embodiment 1. It is a flowchart of the area sensor map update processing which concerns on modification 1 of this invention. It is a figure which shows the functional structure of the autonomous mobile device which concerns on Embodiment 2 of this invention. It is a figure which shows the appearance of the autonomous mobile device which concerns on Embodiment 2. It is a figure which shows the whole structure of the software module of the autonomous mobile device which concerns on Embodiment 2. It is a flowchart of the map making module which concerns on Embodiment 2. It is a flowchart of the cliff sensor map update process which concerns on Embodiment 2. It is a figure which shows the functional structure of the autonomous mobile device which concerns on Embodiment 3 of this invention. It is a figure which shows the whole structure of the software module of the autonomous mobile device which concerns on Embodiment 3. It is a flowchart of the movement control module which concerns on Embodiment 3. It is a flowchart of the map making module which concerns on Embodiment 3. It is a flowchart of the area sensor map update processing which concerns on Embodiment 3. It is a flowchart of the collision sensor map update process which concerns on Embodiment 3.

Hereinafter, embodiments of the present invention will be described with reference to figures and tables. The same or corresponding parts in the figure are designated by the same reference numerals.

(Embodiment 1)
The autonomous moving device according to the embodiment of the present invention is a device that autonomously moves according to an application while creating a map of the surroundings. This application is, for example, for security monitoring, indoor cleaning, pets, toys, and the like.

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 failure detection unit 30, an imaging unit 41, a drive unit 42, a range sensor 43, and a communication unit 44. , Equipped with.

The control unit 10 is composed of a CPU (Central Processing Unit) or the like, and by executing a program stored in the storage unit 20, each unit (position / orientation estimation unit 11, map creation unit 12, fault erasing unit 13, etc.) described later will be executed. The functions of the route setting unit 14 and the movement control unit 15) are realized. Further, the control unit 10 includes a timer (not shown) and can count the elapsed time.

The storage unit 20 is composed of a ROM (Read Only Memory), a RAM (Random Access Memory), etc., and is functionally composed of an image storage unit 21, a feature point storage unit 22, a map storage unit 23, an update amount storage unit 24, and a collision. The data storage unit 25 and the distance measurement data storage unit 26 are included. The ROM stores a program executed by the CPU of the control unit 10 and data necessary for executing the program in advance. Data that is created or changed during program execution is stored in the RAM.

The image storage unit 21 stores the image taken by the image pickup unit 41. However, in order to save storage capacity, it is not necessary to store all the captured images. The autonomous moving device 100 estimates its own position (self-position and orientation) by SLAM (Simultaneus Localization And Mapping) processing using a plurality of images stored in the image storage unit 21. For the image used for estimating the position of the own machine, the information of the position of the own machine at the time when the image is taken is stored together with the information of the image.

In the feature point storage unit 22, among the feature points included in the image stored in the image storage unit 21, the feature points for which the three-dimensional positions (X, Y, Z) in the real space have been obtained are the three. The dimensional position and the feature amount of the feature point are associated and stored. A feature point is a characteristic part in an image such as a corner part in the image, and can be acquired by using an algorithm such as SIFT (Scale-Invariant Feature Transfer) or SURF (Speeded Up Robot Features). .. The feature amount of the feature point is a feature amount obtained by, for example, SIFT. In the above-mentioned SLAM processing, since the self-position is estimated based on the three-dimensional position of the feature point stored in the feature point storage unit 22, the feature point storage unit 22 stores the map information for SLAM. You can think of it as being.

In the map storage unit 23, an environmental map (collision sensor map described later) created by the map creation unit 12 based on the information from the fault detection unit 30 and a map creation unit 12 based on the information from the range sensor 43. The created environment map (range sensor map described later) is stored. In the environmental map, the floor surface is divided into, for example, a grid of 5 cm × 5 cm, and the probability of existence of obstacles at the positions corresponding to each grid point (region centered on the grid point and having the same shape as the grid). Is an occupied grid map represented as the value of the grid point. In addition, instead of "the position corresponding to each grid point", "the position corresponding to each grid (the region of the grid)" may be used. The value of each grid point on the occupied grid map becomes larger as the probability that an obstacle exists at that location becomes larger, and becomes smaller as the probability that an obstacle does not exist is higher. Therefore, for convenience, "existence" It does not have to be a strict existence probability, just called "probability". The value of the existence probability P (real number in the range of 0 to 1) itself may be used as the numerical value of the existence probability recorded as the value of each lattice point of the occupied grid map, but in the present embodiment, the logarithmic odds L (integer) Is used. The relationship between P and L can be expressed by the following equation (1), where k is a positive constant.
P = 1-1 / (1 + exp (kL)) ... (1)

As can be seen from the equation (1), L = 0 when P = 0.5, L = −∞ when P = 0, and L = ∞ when P = 1, and the probabilities are adjusted by using logarithmic odds. It can be handled numerically. Then, the presence or absence of an obstacle at the position corresponding to the grid point can be expressed by the magnitude when the value L of each grid point is compared with the reference value. In the present embodiment, the minimum value (Lmin) is -127, the maximum value (Lmax) is 127, and the initial value of each grid point of the occupied grid map is L = 0. And. Then, for example, if L is larger than the existence reference value (for example, 10), there is an obstacle at the grid point, and if L is less than the non-existence reference value (for example, -10), there is no obstacle at the grid point. If L is greater than or equal to the absence reference value and less than the existence reference value, the presence or absence of obstacles at the grid points is treated as unknown. However, these values can be changed arbitrarily.

The update amount storage unit 24 stores the update amount (the amount for updating the value of each grid point) for each map stored in the map storage unit 23. As the update amount for the range sensor map, a value Δr (for example, 5) common to all grid points is stored. Further, as the update amount for the collision sensor map, the update amount Δc (for example, 10 as an initial value) of the value of the grid points is stored for each grid point of the collision sensor map. The update amount Δc can be considered as the speed (erasing speed) of erasing the obstacle information from the collision sensor map.

The collision data storage unit 25 is based on the self-position and orientation of the autonomous moving device 100 when the collision sensor 31, which will be described later, detects a collision, and the position of the collision sensor 31 that has detected the collision in the autonomous moving device 100. Data (collision data) indicating the position of the collided object (obstacle) is stored.

The distance measurement data storage unit 26 stores distance measurement data indicating the distance to the object detected by the range sensor 43 together with information on the angle at which the object is detected.

The obstacle detection unit 30 includes a sensor that detects an obstacle when the autonomous mobile device 100 approaches the obstacle. The obstacle is a general term for a place where the autonomous moving device 100 cannot move, and includes a wall, an object, a hole, a cliff, and the like. In the first embodiment, the obstacle detection unit 30 includes a collision sensor 31 that detects an obstacle by detecting the collision with the obstacle. The number of collision sensors 31 is arbitrary, but as shown in FIG. 2, the autonomous moving device 100 includes a collision sensor 31b directly in front of the autonomous moving device 100 and a collision sensor 31a slightly to the right of the front as the collision sensor 31. , A collision sensor 31c slightly to the left of the front is provided. When an object collides with these collision sensors 31, the collided collision sensor 31 notifies the control unit 10 that the collision has been detected. Then, the control unit 10 outputs collision data indicating the position of the collided object based on the self-position and orientation of the autonomous moving device 100 at that time and the position of the collision sensor 31 in the autonomous moving device 100 that has notified the collision detection. Register in the collision data storage unit 25.

As shown in FIG. 2, the imaging unit 41 includes a monocular imaging device (camera). The image pickup unit 41 acquires an image (frame) at, for example, 30 fps (frames per second). The autonomous movement device 100 performs autonomous movement while recognizing its own position and the surrounding environment in real time by SLAM processing based on the 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. In FIG. 2, the right wheel is shown as a drive unit 42a, and the left wheel is shown as a drive unit 42b. The autonomous movement device 100 performs parallel movement (translational movement) in the front-rear direction by driving the two wheels in the same direction, rotates on the spot (changes direction) by driving the two wheels in the opposite direction, and speeds each of the two wheels. Turning movement (translation + rotation (direction change) movement) can be performed by the changed drive. In addition, each wheel is equipped with a rotary encoder, which measures the number of rotations of the wheel and uses geometrical relationships such as the diameter of the wheel and the distance between the wheels to increase the amount of translational movement and rotation. You can calculate the amount.

For example, if the diameter of the wheel is D and the rotation speed is R (measured by the rotary encoder), the translational movement amount at the ground contact portion of the wheel is π, D, R. If the diameter of the wheel is D, the distance between the wheels is I, the number of rotations of the right wheel is _RR , and the number of rotations of the left wheel is _RL , the amount of rotation for changing the direction is 360 (assuming that the right rotation is positive). ° × D × ( _{RL −} 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 determines the position of the own machine (self-position and orientation based on the self-position and orientation at the start of movement). It can be used to estimate. Here, the "direction" of the own machine is also referred to as the "attitude" of the own machine.

A crawler may be provided instead of the wheels, or a plurality of (for example, two) legs may be provided and the movement may be performed by walking with the legs. In these cases as well, the self-position and orientation (posture) can be estimated based on the movements of the two crawlers and the movements of the feet, as in the case of the wheels.

The range sensor 43 detects an object (obstacle) existing in the surroundings and acquires the distance (distance measurement data) to the object. As shown in FIG. 2, the range sensor 43 is composed of, for example, a two-dimensional laser scanner provided on the upper part of the autonomous moving device 100, detects an object that is not close to the object from a distance, and acquires the distance to the object. can do. As shown in FIG. 3, the range sensor 43 scans with a laser beam in a predetermined angle range (for example, a range of -30 degrees to +30 degrees) to reach an object existing in the surrounding angle range. The distance (distance measurement data) can be acquired for each predetermined angle (for example, 1 degree). If there is no object in that direction, there will be no ranging data for that angle. For example, when the range sensor 43 scans as shown in FIG. 3 (A), the distance measurement data shown in FIG. 3 (B) can be acquired. When the range sensor 43 periodically (for example, at a frequency of 30 times per second) observes (scans) the surroundings and detects an object, distance measurement data indicating the distance to the object and the object are detected. Information on the angle is transmitted to the control unit 10. Then, the control unit 10 registers these information (distance to the object and the angle at which the object is detected) in the distance measurement data storage unit 26.

The communication unit 44 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 44 is a wireless module for performing short-range wireless communication based on Bluetooth (registered trademark). By using the communication unit 44, the autonomous mobile device 100 can exchange data with the outside. For example, when the user instructs the autonomous mobile device 100 to indicate the destination, the information on the destination may be transmitted via the communication unit 44.

Next, the function of the control unit 10 will be described. The control unit 10 includes a position / orientation estimation unit 11, a map creation unit 12, an obstacle elimination unit 13, a route setting unit 14, and a movement control unit 15, and performs environment map creation, movement control of the autonomous movement device 100, and the like. Further, the control unit 10 supports a multi-thread function, and can execute a plurality of threads (different processing flows) in parallel.

The position / orientation estimation unit 11 estimates the self-position and orientation of the autonomous moving device 100 by SLAM processing using a plurality of images taken by the imaging unit 41 and stored in the image storage unit 21. Further, when estimating the self-position and the posture, the position-posture estimation unit 11 can also use the odometry information that can be acquired from the drive unit 42.

The map creation unit 12 creates an environment map in which the positions of obstacles (objects) are recorded using the information from the collision sensor 31 and the range sensor 43, and stores the environment map in the map storage unit 23. This environmental map is an occupied grid map in which a plane is divided into grid points as described above, and numerical values indicating the existence probabilities of objects (obstacles) detected by the collision sensor 31 and the range sensor 43 are recorded at each grid point. Is. The map creation unit 12 creates a separate occupied grid map for each sensor that detects obstacles. Therefore, in order to distinguish each of these maps, the occupied grid map created by the map creation unit 12 based on the information from the collision sensor 31 is called a collision sensor map, and the map creation unit 12 refers to the information from the range sensor 43. The occupied grid map created based on is called a range sensor map.

For example, the map creation unit 12 records the presence of an obstacle at the grid point corresponding to the collision position on the collision sensor map based on the information from the collision sensor 31. Further, the map creation unit 12 records the existence of obstacles at each grid point detected by the range sensor 43 on the range sensor map based on the information from the range sensor 43.

The obstacle erasing unit 13 deletes the obstacle information recorded on the environmental map (collision sensor map) by the map creating unit 12 from the environmental map (collision sensor map) according to the passage of time. This erasure is performed by the collision sensor map update process described later, but the details will be described later.

The route setting unit 14 sets a movement route to the destination based on the information of the environment map stored in the map storage unit 23.

The movement control unit 15 controls the drive unit 42 so as to move the own machine along the movement route set by the route setting unit 14.

The functional configuration of the autonomous mobile device 100 has been described above. Next, the overall configuration of the software module executed by the control unit 10 of the autonomous mobile device 100 will be described with reference to FIG. In FIG. 4, 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 setting module 53 corresponds to the route setting unit 14. , The movement control module 54 corresponds to the movement control unit 15.

When the power of the autonomous mobile device 100 is turned on, these software modules are started as separate threads, and execution is started in parallel. The position / orientation estimation module 51 performs SLAM processing using the image information acquired from the imaging unit 41, and estimates its own position and orientation using the odometry information acquired from the driving unit 42. The map creation module 52 creates an environment map based on the self-position and attitude estimated by the position / orientation estimation module 51, the collision data obtained from the collision sensor 31, and the distance measurement data obtained from the range sensor 43. ..

The route setting module 53 sets a route based on the self-position and attitude estimated by the position / orientation estimation module 51 and the environment map created by the map creation module 52. The movement control module 54 controls the drive unit 42 based on the self-position and attitude estimated by the position / orientation estimation module 51, the environment map created by the map creation module 52, and the route set by the route setting module 53. Generates movement control information (mainly speed information). Then, the drive unit 42 is driven based on the movement control information generated by the movement control module 54, and moves to the destination.

Here, the problem solved by the autonomous mobile device 100 will be described with reference to FIG. As shown in (1) of FIG. 5, the autonomous mobile device 100 cannot detect the obstacle 70 with the range sensor 43 when the obstacle 70 is small, and the obstacle 70 is not recorded on the environmental map. .. Then, when the autonomous mobile device 100 moves forward and collides with the obstacle 70 as shown in FIG. 5 (2), the obstacle 70 is detected by the collision sensor 31, and the environment map (collision sensor map) by the collision sensor 31 is detected. ), The obstacle 70 is recorded. Then, after that, when the autonomous moving device 100 continues to move while avoiding the obstacle 70, even if the obstacle 70 is removed, the autonomous moving device 100 cannot detect that, and (1) in FIG. As shown in 3), the obstacle 70 will remain on the environmental map (collision sensor map) indefinitely.

Then, the mechanism by which the autonomous mobile device 100 erases the obstacle 70 in (3) of FIG. 5 from the environmental map (collision sensor map) will be described below. First, the processing contents of the map creation module 52 will be described with reference to FIG. In the following, the range sensor map is a two-dimensional array variable MA [i, j], the collision sensor map is a two-dimensional array variable MB [i, j], and the range sensor map and the collision sensor map are integrated into an environmental map. Is represented by the two-dimensional array variable MI [i, j], and the update amount for the collision sensor map is represented by the two-dimensional array variable DC [i, j]. Here, if xsize is the coordinate of the maximum grid point in the X direction of the environment map and ysize is the coordinate of the maximum grid point in the Y direction of the environment map, then 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

First, the map creation unit 12 initializes the range sensor map stored in the map storage unit 23 (step S101). The initialization of the map is performed by setting the values of all the grid points of the occupied grid map to 0. Specifically, MA [i, j] = 0 is executed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

Next, the map creation unit 12 initializes the collision sensor map stored in the map storage unit 23 (step S102). Specifically, MB [i, j] = 0 is executed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

Next, the failure erasing unit 13 initializes the update amount Δc for the collision sensor map stored in the update amount storage unit 24 (step S103). The initial value at this time is arbitrary, but is set to, for example, 10. As a specific process, DC [i, j] = 10 is executed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

Next, the control unit 10 determines whether or not the autonomous mobile device 100 ends its operation (step S104). When the user turns off the power of the autonomous mobile device 100, or when the remaining amount of the battery falls below a predetermined amount (for example, the remaining 3% or the like), the autonomous mobile device 100 ends its operation. When the autonomous mobile device 100 ends the operation (step S104; Yes), the process ends. If the operation is not completed (step S104; No), the map creation unit 12 acquires the self-position and posture estimated by the position / orientation estimation unit 11 at that time (step S105).

Next, the map creation unit 12 performs the area sensor map update process (step S106). The details of the range sensor map update process will be described later. Next, the map creation unit 12 performs the collision sensor map update process (step S107). The details of the collision sensor map update process will be described later. Next, the map creation unit 12 creates an environment map in which the range sensor map and the collision sensor map are integrated (step S108), and returns to step S104.

The environmental map that integrates the range sensor map and the collision sensor map is a map that enables the autonomous mobile device 100 to recognize information recorded in only one of the range sensor map and the collision sensor map. is there. Specifically, the following processing is performed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.
(1) If MA [i, j] <0 and MB [i, j] <0, then MI [i, j] = min (MA [i, j], MB [i, j])
(2) If MA [i, j] ≧ 0 or MB [i, j] ≧ 0, MI [i, j] = max (MA [i, j], MB [i, j])

Next, the range sensor map update process performed in step S106 of FIG. 6 will be described with reference to FIG. 7. First, the map creation unit 12 determines whether or not the distance measurement data newly registered in the distance measurement data storage unit 26 exists (step S201). If there is no distance measurement data newly registered in the distance measurement data storage unit 26 (step S201; No), the range sensor map update process is terminated.

If the distance measurement data newly registered in the distance measurement data storage unit 26 exists (step S201; Yes), the map creation unit 12 acquires the distance measurement data from the distance measurement data storage unit 26, and the map creation unit 12 acquires the distance measurement data from the distance measurement data storage unit 26. Using the self-position and orientation acquired in step S105, the distance measurement data is converted into coordinates in the range sensor map (step S202). These coordinates indicate the grid points (areas) where obstacles are detected by the range sensor 43 on the range sensor map, and in this step, the map creation unit 12 acquires the position of the obstacle from a distance. Functions as an obstacle position acquisition unit. Here, for example, it is assumed that an obstacle is detected at the coordinates of [m, n] on the range sensor map.

Then, the map creation unit 12 adds the update amount Δr to the value of the grid point where the obstacle is detected (step S203). Specifically, MA [m, n] = MA [m, n] + Δr is executed. However, at this time, if MA [m, n] exceeds the maximum value Lmax, MA [m, n] = Lmax.

Next, the map creation unit 12 subtracts the update amount Δr from the value of the grid point where no obstacle is detected within the observation (scan) range of the range sensor 43 (step S204). Specifically, all the [i, j] before [m, n] are in the direction in which the obstacle is detected, and all in the observation range of the range sensor 43 in the direction in which the obstacle is not detected. MA [i, j] = MA [i, j] −Δr is executed for [i, j] of. However, at this time, if MA [i, j] is less than the minimum value Lmin, MA [i, j] = Lmin.

Then, the map creation unit 12 deletes the distance measurement data acquired in step S202 from the distance measurement data storage unit 26 (step S205), and returns to step S201. By the above-mentioned area sensor map update process, the area sensor map is updated, and the existence probability of the obstacle is represented by the value of each grid point of the area sensor map.

In addition, as the value of each grid point in the range sensor map, instead of the logarithmic odds L, the number of observations (the number of times the grid point is observed (scanned) by the range sensor 43) and the number of detections (at the position of the grid point) A pair of (the number of times that the range sensor 43 detects the existence of an obstacle) may be recorded, and the probability of existence of the obstacle at each grid point may be defined as the ratio of the number of detections to the number of observations.

Next, the collision sensor map update process performed in step S107 of FIG. 6 will be described with reference to FIG. First, the fault erasing unit 13 attenuates the value of each grid point on the collision sensor map (step S301). Step S301 is also called a failure elimination step. Specifically, in this step, the fault erasing unit 13 performs the following processing for all i and j.
(1) If MB [i, j]> 0, MB [i, j] = MB [i, j] -DC [i, j]
(2) If MB [i, j] <0, then MB [i, j] = MB [i, j] + DC [i, j]

Then, the map creation unit 12 sets the value of the grid points of the collision sensor map corresponding to the self-position acquired in step S105 of FIG. 6 to the minimum value Lmin (step S302). Specifically, assuming that the coordinates of the self-position are [p, q], MB [p, q] = Lmin is executed. Then, the fault erasing unit 13 increases the update amount Δc of the value of the grid point of the collision sensor map corresponding to the self-position by the reference non-failure adjustment value (for example, 5) (step S303). Specifically, assuming that the coordinates of the self-position are [p, q], DC [p, q] = DC [p, q] + 5 is executed. If DC [p, q] exceeds Lmax here, DC [p, q] = Lmax.

Next, the map creation unit 12 determines whether or not the collision data newly registered in the collision data storage unit 25 exists (step S304). Step S304 is also referred to as a failure detection step. If there is no newly registered collision data in the collision data storage unit 25 (step S304; No), the collision sensor map update process is terminated.

If there is collision data newly registered in the collision data storage unit 25 (step S304; Yes), the map creation unit 12 indicates the collision data (the position where the collision obstacle exists) from the collision data storage unit 25. (Coordinates) is acquired, and the value of the lattice point of the collision sensor map where the collision obstacle exists is set to the maximum value Lmax (step S305). Specifically, assuming that the coordinates of the colliding obstacle are [v, w], MB [v, w] = Lmax is executed. Step S305 is also called a cartography step. Then, the obstacle erasing unit 13 reduces the update amount Δc of the value of the grid point where the colliding obstacle exists by the reference obstacle adjustment value (for example, 5) (step S306). Specifically, assuming that the coordinates of the colliding obstacle are [v, w], DC [v, w] = DC [v, w] -5 is executed. If DC [p, q] is 0 or less here, DC [v, w] = 1.

Then, the map creation unit 12 deletes the collision data acquired in step S305 from the collision data storage unit 25 (step S307), and returns to step S304. The collision sensor map is updated by the above collision sensor map update process, and the attenuation process in step S301 prevents the ghost of obstacles from remaining. In addition, by increasing or decreasing the update amount Δc of the grid point value itself, the attenuation rate is slowed down at the grid points where there is a high possibility of obstacles, and at the grid points where obstacles are unlikely to exist. The decay rate is increased. This has the effect of reducing the frequency of collisions with non-moving obstacles again and increasing the tracking speed of moving obstacles.

Next, the processing content of the route setting module 53 that sets the route to the destination using the map created by the map creation module 52 will be described with reference to FIG.

First, the control unit 10 determines whether or not the autonomous mobile device 100 ends its operation (step S401). If the operation is terminated (step 401; Yes), the process is terminated. If the operation is not completed (step S401; No), the route setting unit 14 determines whether or not the destination has been set (step S402). The destination may be set by the user of the autonomous mobile device 100 via the communication unit 44, or the destination may be set by the autonomous mobile device 100 as necessary (for example, the remaining battery level is less than a predetermined amount (for example, 10%)). In some cases, the charging station may be set as the destination, etc.) autonomously.

If the destination has not been set (step S402; No), the process returns to step S401. If the destination is set (step S402; Yes), the route setting unit 14 acquires the latest environmental map created by the map creation unit 12 at that time (step S403). This environmental map is an environmental map (a map represented by the two-dimensional array variable MI [i, j]) that integrates the range sensor map and the collision sensor map. Next, the route setting unit 14 acquires the current self-position and attitude estimated by the position / attitude estimation unit 11 (step S404). Then, the route setting unit 14 sets the route from the current position to the destination based on the acquired environment map, the self-position and the posture, and the set destination (step S405). Step S405 is also called a route setting step.

Then, the route setting unit 14 determines whether or not the route can be set (whether or not the route exists) (step S406). If the route does not exist (step S406; No), error processing such as notifying the user to that effect is performed (step S408), and the process returns to step S401. If a route exists (step S406; Yes), the route setting unit 14 notifies the movement control module 54 of the set route (step S407), and returns to step S401.

By the above processing of the route setting module 53, a route to the destination can be obtained. Next, the processing content 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 unit 10 determines whether or not the autonomous mobile device 100 ends its operation (step S501). If the operation is terminated (step S501; Yes), the process is terminated. If the operation is not terminated (step S501; No), the movement control unit 15 determines whether or not the route is set by the route setting module 53 (whether or not the route set by the route setting module 53 is notified) (whether or not the route set by the route setting module 53 is notified). Step S502). If the route has not been notified (step S502; No), it means that the route has not been set yet, so the process returns to step S501.

If the route is notified (step S502; Yes), it means that the route has been set, so that the movement control unit 15 acquires the route set by the route setting module 53 (step S503). Next, the movement control unit 15 acquires the current self-position and attitude estimated by the position / attitude estimation unit 11 (step S504). Then, the movement control unit 15 determines whether or not the own machine has arrived at the destination (step S505). Since the route information includes the 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.

When the destination is reached (step S505; Yes), the process returns to step S501. If the destination has not arrived (step S505; No), the movement control unit 15 acquires the latest environmental map created by the map creation unit 12 at that time (step S506). This environmental map is an environmental map (a map represented by the two-dimensional array variable MI [i, j]) that integrates the range sensor map and the collision sensor map. Then, the movement control unit 15 determines whether or not it is possible to move along the route based on the acquired environment map and route (step S507).

If it is not possible to move along the route (step S507; No), the movement control unit 15 controls the drive unit 42 to stop the movement (step S509), and notifies the route setting module 53 of an error (step S510). ), Return to step S501. If it is possible to move along the route (step S507; Yes), the drive unit 42 is controlled to move along the route (step S508), and the process returns to step S504.

By the above processing of the movement control module 54, the autonomous movement device 100 can move to the destination. Then, as described above, the obstacle erasing unit 13 reduces the existence probability of the obstacle in the collision sensor map with the passage of time. Further, since the obstacle erasing unit 13 can set the speed at which the existence probability is reduced to a different value for each grid point of the collision sensor map, the existence probability can be efficiently reduced according to the property of the object. Can be done. Therefore, even in a place where an obstacle has collided in the past, after a certain period of time has passed, the route setting module 53 can set a route that does not avoid the place, and the autonomous mobile device 100 is essentially unnecessary. Obstacle avoidance operation can be prevented, and it becomes possible to move efficiently.

(Modification example 1)
The amount of update of the value of each grid point of the collision sensor map according to the first embodiment can be increased or decreased only for the position on the movement path of the autonomous moving device 100 and the grid points corresponding to the collision positions. A modified example 1 in which the frequency of adjusting the update amount can be improved by increasing or decreasing the update amount in the range sensor update process will be described.

The modified example 1 differs from the first embodiment only in the range sensor map update process described with reference to FIG. 7. Therefore, the range sensor map update process according to the first modification will be described with reference to FIG. Further, in FIG. 11, the processes from step S201 to step S205 are the same as those in FIG. 7, and thus the description thereof will be omitted.

In the range sensor map update process according to the first modification, after step S204, the obstacle erasing unit 13 updates the value of the grid point of the collision sensor map corresponding to the position of the obstacle detected by the range sensor 43. The value of the quantity Δc is reduced by the reference fault adjustment value (for example, 5) (step S211). Specifically, assuming that the coordinates of the obstacle detected by the range sensor 43 are [m, n], DC [m, n] = DC [m, n] -5 is executed. However, at this time, if DC [m, n] becomes 0 or less, DC [m, n] = 0. In this case, the location of [m, n] is likely to be an obstacle that does not move like a wall, and the presence or absence of the obstacle that can be detected by the range sensor 43 can be frequently confirmed. This is because if an obstacle moves or is removed, it can be immediately reflected in Δc (without colliding with the obstacle).

Next, the obstacle erasing unit 13 refers to the value of the update amount Δc of the grid point value of the collision sensor map corresponding to the position where the obstacle is not detected within the observation (scan) range of the range sensor 43. The adjustment value (for example, 5) is increased (step S212). Specifically, assuming that the coordinates of the obstacle detected by the range sensor 43 are [m, n], the directions in which the obstacle is detected are all [i, j] before [m, n]. ], DC [i, j] = DC [i, j] +5 is executed for all [i, j] within the observation range of the range sensor 43 in the direction in which the obstacle is not detected. .. However, at this time, if DC [i, j] exceeds Lmax, DC [i, j] = Lmax.

Then, the process proceeds to step S205. After that, it is the same as the range sensor map update process according to the first embodiment. In the modification 1 described above, the update amount Δc of the value of the grid points of the collision sensor map can be increased or decreased by using the detection status of the obstacle by the range sensor 43. Therefore, when the collision sensor map is updated, the non-moving obstacles are not deleted from the collision sensor map, and the moving obstacles can be deleted from the collision sensor map even faster. Therefore, the autonomous movement device 100 according to the first modification can reflect the movement of the obstacle on the environment map in a short time, so that the movement can be performed more efficiently.

(Modification 2)
In the process of step S211 of the first modification, the obstacle erasing unit 13 sets the value of the update amount Δc of the value of the grid point of the collision sensor map corresponding to the position of the obstacle detected by the range sensor 43 to the initial value (for example). It may be reset to 10). Specifically, assuming that the coordinates of the obstacle detected by the range sensor 43 are [m, n], DC [m, n] = 10 is executed. When an obstacle suddenly appears in a place where there was no obstacle so far, the value of the update amount Δc corresponding to the place may become very large. In that case, if the collision sensor map is attenuated by the collision sensor map update process, the information on the obstacle may be immediately deleted from the collision sensor map. However, if the value of the update amount Δc is set to the initial value in the process of step S211 as described above, this problem disappears. Therefore, the autonomous moving device 100 according to the second modification can move more efficiently because it can set a route to avoid even a suddenly appearing obstacle.

(Modification example 3)
Further, the processes of the first modification and the second modification can be combined. For example, a reset reference value (for example, 100) for determining whether or not to reset Δc is set. Then, in step S211 if the value of the update amount Δc of the value of the grid point of the collision sensor map corresponding to the position of the obstacle detected by the range sensor 43 exceeds the reset reference value, the obstacle erasing unit 13 It may be reset to the initial value (for example, 10), and if it is equal to or less than the reset reference value, a process of reducing the reference failure adjustment value may be performed. In this way, the autonomous moving device 100 according to the third modification can appropriately reflect the movement of obstacles and sudden appearances on the environment map, so that the autonomous moving device 100 can move more efficiently.

(Embodiment 2)
Examples of the sensor that detects the obstacle by approaching the obstacle include not only a collision sensor but also a cliff sensor. The cliff sensor is a sensor that detects a cliff (hole, cliff, etc.) that is a place where the floor surface does not exist by detecting whether or not the floor surface exists in the surroundings. By providing the autonomous moving device with a cliff sensor, it is possible to prevent the autonomous moving device from falling from a table such as a table. The second embodiment in which the autonomous mobile device includes a cliff sensor will be described.

In the autonomous mobile device 101 according to the second embodiment, as shown in FIG. 12, a cliff data storage unit 27 is added to the storage unit 20 of the autonomous mobile device 100 according to the first embodiment, and a cliff sensor 32 is added to the failure detection unit 30. It is an added configuration. Further, the map storage unit 23 also stores the cliff sensor map created by using the data from the cliff sensor 32, and the update amount storage unit 24 also stores the update amount Δg of the value of each grid point of the cliff sensor map. Will be done. The update amount Δg can be considered as the speed at which the cliff information is erased (erasing speed) from the cliff sensor map. In the following, the cliff sensor map is a two-dimensional array variable MC [i, j], and the environment map that integrates the range sensor map, the collision sensor map, and the cliff sensor map is the two-dimensional array variable MI [i, j]. The update amount for the cliff sensor map shall be represented by the two-dimensional array variable DG [i, j].

The cliff data storage unit 27 obtains the self-position and orientation of the autonomous moving device 101 when the cliff sensor 32 detects the cliff, and the position of the cliff sensor 32 in the autonomous moving device 101 that detects the cliff. Data (cliff data) indicating the position of the cliff (hole, cliff, etc.) to be generated is stored.

The cliff sensor 32 is a sensor that detects a cliff (hole, cliff, etc.) that is a place where the floor surface does not exist by detecting whether or not the floor surface exists in the surroundings. The number of cliff sensors 32 is arbitrary, but as shown in FIG. 13, the autonomous moving device 101 includes one cliff sensor 32 at a position directly in front of the autonomous moving device. When the cliff sensor 32 detects the cliff, the cliff sensor 32 notifies the control unit 10 that the cliff has been detected. Then, the control unit 10 stores cliff data indicating the position of the cliff based on the self-position and orientation of the autonomous moving device 101 at that time and the position of the cliff sensor 32 that has detected the cliff in the autonomous moving device 101. Register in department 27.

As shown in FIG. 14, the overall configuration of the software module executed by the control unit 10 of the autonomous mobile device 101 is such that the cliff sensor 32 is added to the overall configuration of the software module of the autonomous mobile device 100 described with reference to FIG. It has a configured structure. The map creation module 52 of the autonomous mobile device 101 also creates an environment map using the cliff data obtained from the cliff sensor 32. Except for this point, it is the same as the autonomous mobile device 100.

The processing content of the map creation module 52 of the autonomous mobile device 101 will be described with reference to FIG. Since this process is common to the process content of the map creation module 52 of the autonomous mobile device 100 described with reference to FIG. 6 except for a part, the differences will be mainly described.

The processing from step S101 to step S103 is the same as the processing of FIG. Next to step S103, the fault erasing unit 13 initializes the update amount Δg for the cliff sensor map stored in the update amount storage unit 24 (step S111). The initial value at this time is arbitrary, but is set to, for example, 10. As a specific process, DG [i, j] = 10 is executed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

Subsequent processing from step S104 to step S107 is the same as the processing of FIG. Following step S107, the map creation unit 12 performs the cliff sensor map update process (step S112). The details of the cliff sensor map update process will be described later. After that, the map creation unit 12 creates an environment map in which the range sensor map, the collision sensor map, and the cliff sensor map are integrated (step S113), and returns to step S104.

The environment map that integrates the range sensor map, the collision sensor map, and the cliff sensor map is the information recorded in only one of the range sensor map, the collision sensor map, and the cliff sensor map. It is a map to make it recognizable. Specifically, the following processing is performed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.
(1) If MA [i, j] <0 and MB [i, j] <0 and MC [i, j] <0, then MI [i, j] = min (MA [i, j], MB [ i, j], MC [i, j])
(2) If MA [i, j] ≧ 0 or MB [i, j] ≧ 0 or MC [i, j] ≧ 0, MI [i, j] = max (MA [i, j], MB [ i, j], MC [i, j])

Next, the cliff sensor map update process performed in step S112 of FIG. 15 will be described with reference to FIG. First, the fault erasing unit 13 attenuates the value of each grid point on the cliff sensor map (step S601). Specifically, the following processing is performed for all i and j.
(1) If MC [i, j]> 0, MC [i, j] = MC [i, j] -DG [i, j]
(2) If MC [i, j] <0, then MC [i, j] = MC [i, j] + DG [i, j]

Then, the map creation unit 12 sets the value of the grid point of the cliff sensor map corresponding to the self-position acquired in step S105 of FIG. 15 to the minimum value Lmin (step S602). Specifically, assuming that the coordinates of the self-position are [p, q], MC [p, q] = Lmin is executed. Then, the fault erasing unit 13 increases the update amount Δg of the grid point value of the cliff sensor map corresponding to the self-position by the reference non-failure adjustment value (for example, 5) (step S603). Specifically, assuming that the coordinates of the self-position are [p, q], DG [p, q] = DG [p, q] + 5 is executed. If DG [p, q] exceeds Lmax here, DG [p, q] = Lmax.

Next, the map creation unit 12 determines whether or not the cliff data newly registered in the cliff data storage unit 27 exists (step S604). If there is no newly registered cliff data in the cliff data storage unit 27 (step S604; No), the cliff sensor map update process is terminated.

If the cliff data newly registered in the cliff data storage unit 27 exists (step S604; Yes), the map creation unit 12 indicates the cliff data (the position where the detected cliff exists) from the cliff data storage unit 27. Coordinates) are acquired, and the value of the grid point where the detected cliff is present on the cliff sensor map is set to the maximum value Lmax (step S605). Specifically, assuming that the coordinates of the colliding obstacle are [v, w], MC [v, w] = Lmax is executed. Then, the fault erasing unit 13 reduces the update amount Δg of the value of the grid point where the detected cliff exists by the reference fault adjustment value (for example, 5) (step S606). Specifically, assuming that the coordinates of the colliding obstacle are [v, w], DG [v, w] = DG [v, w] -5 is executed. If DG [p, q] is 0 or less here, DG [v, w] = 1.

Then, the map creation unit 12 deletes the cliff data acquired in step S605 from the cliff data storage unit 27 (step S607), and returns to step S604. The cliff sensor map is updated by the above cliff sensor map update process, and the attenuation process of step S601 prevents the ghost of the cliff from remaining. In addition, by increasing or decreasing the update amount Δg of the grid point value itself, the decay rate is slowed down at the grid points where the cliff is likely to be present, and the decay rate is slowed down at the grid points where the cliff is unlikely to be present. Is fast. This has the effect of reducing the frequency of reapproaching unchanged cliffs (such as the edges of tables and tables) and increasing the follow-up speed for changing cliffs (such as holes that open and close the lid, such as manholes). Is getting. Therefore, the autonomous mobile device 101 according to the second embodiment can appropriately reflect the appearance and disappearance of the cliff on the environmental map, so that the autonomous mobile device 101 can move efficiently.

(Modification example 4)
In the autonomous mobile device 101 according to the second embodiment, the fault detection unit 30 includes the collision sensor 31 and the cliff sensor 32, but the fault detection unit 30 is not limited to these, and detects some obstacles by approaching them. Any sensor X can be included as long as it can be used. Then, in that case, the map creation unit 12 also records the sensor X map created by using the data from the sensor X in the map storage unit 23, and creates an environmental map in which all the environmental maps for each sensor provided are integrated. .. Further, the update amount storage unit 24 also stores the update amount Δx of the value of each grid point of the sensor X map. As the sensor X, for example, a floor surface sensor that detects a rough road, a moisture sensor that detects a wet portion of the floor or a puddle, or the like can be considered. By increasing the types of sensors included in the fault detection unit 30, the autonomous mobile device can create an environment map flexibly corresponding to various environments and can move efficiently.

(Embodiment 3)
In each of the above-described embodiments, the obstacle detection unit 30 includes sensors (proximity obstacle detection unit) such as a collision sensor 31 and a cliff sensor 32 that detect an obstacle by approaching the obstacle. Then, the obstacle detected by the obstacle detection unit 30 is assumed to exist reliably, and the value of the grid point at that position on the collision sensor map is set to the maximum value Lmax. However, the obstacle detection unit 30 may include a sensor (distant obstacle detection unit, for example, a range sensor 43) that can detect an obstacle from a distance without being close to each other. In this case, the obstacle detected by the obstacle detection unit 30 is detected not only by the obstacle detected by the range sensor 43 but also by a sensor that detects the obstacle by approaching the obstacle such as the collision sensor 31. (For obstacles), the probability of existence of the obstacle at that position may be updated. Specifically, instead of setting the maximum value Lmax for the value of the grid point at that position (the position of the colliding obstacle) on the collision sensor map, the process of adding the update amount Δr to the value of the grid point at that position. May be done. In this way, the third embodiment in which the collision sensor map and the range sensor map are updated and attenuated in the same manner will be described.

As shown in FIG. 17, the autonomous mobile device 102 according to the third embodiment has a configuration in which the failure detection unit 30 of the autonomous mobile device 100 according to the first embodiment includes not only a collision sensor 31 but also a range sensor 43. .. The map storage unit 23 stores a range sensor map (MA [i, j]), a collision sensor map (MB [i, j]), and an environment map (MI [i, j]) that integrates both sensor maps. .. The update amount storage unit 24 stores the update amount for the collision sensor map and the update amount for the range sensor map. The update amount for the collision sensor map includes a value Δco common to all the grid points and a value DCT [i, j] of the update amount (attenuation amount) with respect to the grid points [i, j]. The update amount for the range sensor map includes a value Δro common to all grid points and a value DRT [i, j] of the update amount (attenuation amount) with respect to the grid points [i, j]. Here, if xsize is the coordinate of the maximum grid point in the X direction of the environment map and ysize is the coordinate of the maximum grid point in the Y direction of the environment map, then 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

The update amount Δco is a value to be subtracted from the value of the grid point of the self-position of the collision sensor map (there is no obstacle because the own machine exists), and when the collision sensor 31 detects an obstacle. It is a value to be added to the value of the grid point at the position of the obstacle. The update amount DCT [i, j] is the amount of attenuation of the value of each grid point used when the collision sensor map is attenuated. All of these values are positive numbers and can be set to, for example, 10. Further, if the value of DCT [i, j] becomes too large, the speed of erasing the obstacle information from the collision sensor map becomes too fast. Therefore, the maximum value DCmax is set and DCT [i, j] is set. ] Is greater than or equal to DCmax, it is desirable to set DCT [i, j] = DCmax. The appropriate value of DCmax differs depending on the time required for erasing obstacles and the decay cycle of the map (the time required for erasing obstacles divided by the decay cycle of the map, divided by Lmax. ), For example, when it is desired to set the time for erasing the obstacle to 60 seconds or more at the shortest, and the period of the attenuation processing is 5 seconds, DCmax = Lmax / 12.

The update amount Δro is a value to be subtracted from the value of the grid point at the position where the obstacle is not detected by the range sensor 43 on the range sensor map, and when the range sensor 43 detects an obstacle. It is a value to be added to the value of the grid point at the position of the obstacle. The update amount DRT [i, j] is the amount of attenuation of the value of each grid point used when the range sensor map is attenuated. All of these values are also positive numbers, and for example, 5 can be set. Further, if the value of DRT [i, j] becomes too large, the speed of erasing the obstacle information from the range sensor map becomes too fast. Therefore, the maximum value DRmax is set and the DRT [i, j] is set. When the value of [j] is greater than or equal to DRmax, it is desirable to set DRT [i, j] = DRmax. The appropriate value of DRmax differs depending on the time required to erase obstacles and the decay cycle of the map (the time required to erase obstacles divided by the decay cycle of the map, divided by Lmax. ), For example, when the time for erasing an obstacle is desired to be 60 seconds or more at the shortest, and the decay processing cycle is 5 seconds, DRmax = Lmax / 12.

As shown in FIG. 18, the overall configuration of the software module executed by the control unit 10 of the autonomous mobile device 102 is deleted from the overall configuration of the software module of the autonomous mobile device 100 described with reference to FIG. It has a configured structure. Actually, as can be seen from the flowchart of the movement control module 54 described later, the combination of the route setting module 53 and the movement control module 54 of the autonomous movement device 100 is the movement control module 54 of the autonomous movement device 102. .. In each of the above embodiments, it has been described that each software module is executed in a separate thread. However, in order to show that some or all of them may be executed in the same thread, in the third embodiment, the route setting module 53 and the movement control module 54 are in the same thread (movement of the autonomous movement device 102). An example executed by the control module 54) will be described.

The processing content of the movement control module 54 of the autonomous movement device 102 will be described with reference to FIG.

First, the control unit 10 determines whether or not the autonomous mobile device 102 ends its operation (step S701). If the operation is terminated (step S701; Yes), the process is terminated. If the operation is not completed (step S701; No), the route setting unit 14 determines whether or not the destination has been set (step S702). The destination may be set by the user of the autonomous mobile device 102 via the communication unit 44, and the autonomous mobile device 102 may have a battery level below a predetermined amount (for example, 10%) as necessary (for example, the remaining battery level is less than a predetermined amount (for example, 10%). In some cases, the charging station may be set as the destination, etc.) autonomously.

If the destination has not been set (step S702; No), the process returns to step S701. If the destination is set (step S702; Yes), the route setting unit 14 acquires the current self-position and attitude estimated by the position / attitude estimation unit 11 (step S703). Next, the route setting unit 14 acquires the latest environment map at that time, which is created by the process of the map creation module 52 described later (step S704). This environmental map is an environmental map (a map represented by the two-dimensional array variable MI [i, j]) that integrates the range sensor map and the collision sensor map. Then, the route setting unit 14 sets the route from the current position to the destination based on the acquired environment map, the self-position and the posture, and the set destination (step S705). Step S705 is also called a route setting step.

Then, the route setting unit 14 determines whether or not the route can be set (whether or not the route exists) (step S706). If the route does not exist (step S706; No), the movement control unit 15 controls the drive unit 42 to stop the movement (step S707), and notifies the user that the route does not exist, for example, via the communication unit 44. (Step S708), and the process returns to step S701.

If the route exists (can be set) (step S706; Yes), the movement control unit 15 determines whether or not the own aircraft has arrived at the destination (step S709). Since the route information set in step S705 also includes destination information, the movement control unit 15 determines whether or not the self-position matches the destination included in the route information. Therefore, it can be determined whether or not the user has arrived at the destination.

When the destination has arrived (step S709; Yes), the user is notified that the destination has arrived (step S710), and the process returns to step S701. If it has not arrived at the destination (step S709; No), the movement control unit 15 determines whether or not it can move along the route (step S711). If it is not movable (step S711; No), the process proceeds to step S707 to perform error processing. If it is movable (step S711; Yes), the drive unit 42 is controlled to move along the path (step S712).

Then, the movement control unit 15 acquires the current self-position and posture estimated by the position / attitude estimation unit 11 (step S713). Next, the movement control unit 15 acquires the latest environmental map created at that time by the process of the map creation module 52 described later (step S714). Then, the movement control unit 15 determines whether or not the acquired environment map has been updated from the previous environment map (step S715). If it has been updated (step S715; Yes), it returns to step S705 and resets the route. If it has not been updated (step S715; No), the process returns to step S709.

By the above processing of the movement control module 54, the autonomous movement device 102 can move to the destination.

Next, the processing contents of the map creation module 52 of the autonomous mobile device 102 will be described with reference to FIG. This process is a process in which the process of step S103 is replaced with steps S121 and S122 among the processing contents of the map creation module 52 of the autonomous mobile device 100 described with reference to FIG. The contents of each process of updating the range sensor map in S106 and updating the collision sensor map in step S107 are also different from the processes of the autonomous mobile device 100. Therefore, the processing contents of steps S121 and S122 will be described, and then the contents of each processing of the range sensor map update and the collision sensor map update of the autonomous mobile device 102 will be described.

First, in step S121, the fault erasing unit 13 initializes the DRT [i, j], which is the update amount (attenuation amount) for the range sensor map stored in the update amount storage unit 24. The initial value at the time of this initialization is arbitrary, but can be set to 5, for example. As a specific process, DRT [i, j] = 5 is executed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

Next, in step S122, the fault erasing unit 13 initializes the DCT [i, j] which is the update amount (attenuation amount) for the collision sensor map stored in the update amount storage unit 24. The initial value at the time of this initialization is arbitrary, but can be, for example, 10. As a specific process, DCT [i, j] = 10 is executed for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize.

Next, the range sensor map update process performed in step S106 of the process of the map creation module 52 (FIG. 20) will be described with reference to FIG.

Assuming that the coordinates of the current position (self-position) of the own machine are represented by [p, q], first, the obstacle erasing unit 13 updates the grid points (attenuation amount) of the range sensor map corresponding to the self-position. DRT [p, q] is increased by the reference non-disorder adjustment value (for example, 3) for the range sensor map (step S221). That is, assuming that the reference non-disability adjustment value for the range sensor map is 3, DRT [p, q] = DRT [p, q] +3 is executed. However, when the value of DRT [p, q] is larger than the maximum value DRmax of DRT, DRT [p, q] = DRmax.

Next, the map creation unit 12 determines whether or not the distance measurement data newly registered in the distance measurement data storage unit 26 exists (step S222). If the distance measurement data newly registered in the distance measurement data storage unit 26 exists (step S222; Yes), the map creation unit 12 acquires the distance measurement data from the distance measurement data storage unit 26, and the map creation unit 12 acquires the distance measurement data from the distance measurement data storage unit 26. Using the self-position and orientation acquired in step S105, the distance measurement data is converted into coordinates in the range sensor map (step S223). These coordinates indicate the grid points (areas) where obstacles are detected by the range sensor 43 on the range sensor map, and in this step, the map creation unit 12 acquires the position of the obstacle from a distance. Functions as an obstacle position acquisition unit. Here, for example, it is assumed that an obstacle is detected at the coordinates of [m, n] on the range sensor map.

Then, the map creation unit 12 adds the update amount Δro to the value of the grid point where the obstacle is detected in the range sensor map (step S224). Specifically, MA [m, n] = MA [m, n] + Δro is executed. However, at this time, if MA [m, n] exceeds the maximum value Lmax, MA [m, n] = Lmax.

Next, the map creation unit 12 subtracts the update amount Δro from the value of the grid point where no obstacle is detected within the observation (scan) range of the range sensor 43 in the range sensor map (step S225). Specifically, all the [i, j] before [m, n] are in the direction in which the obstacle is detected, and all in the observation range of the range sensor 43 in the direction in which the obstacle is not detected. MA [i, j] = MA [i, j] -Δro is executed for [i, j] of. However, at this time, if MA [i, j] is less than the minimum value Lmin, MA [i, j] = Lmin.

Then, the obstacle erasing unit 13 sets the value of the update amount (attenuation amount) DRT [m, n] of the range sensor map corresponding to the lattice point where the obstacle is detected to the reference failure adjustment value for the range sensor map ( For example, reduce by 3) and set the update amount (attenuation amount) DRT [i, j] of the range sensor map corresponding to the lattice point where no obstacle was detected as the reference non-obstacle adjustment value for the range sensor map. Increase by (for example, 3) (step S226). If DRT [m, n] is 0 or less here, DRT [m, n] = 1. If DRT [i, j] is greater than or equal to DRmax, then DRT [i, j] = DRmax.

Next, the obstacle erasing unit 13 reduces the value of the collision sensor map update amount (attenuation amount) DCT [m, n] corresponding to the grid point where the obstacle is detected by the reference obstacle adjustment value (for example, 5). , The value of the collision sensor map update amount (attenuation amount) DCT [i, j] corresponding to the grid points where no obstacle is detected is increased by the reference non-obstacle adjustment value (for example, 5) (step S227). If DCT [m, n] is 0 or less here, DCT [m, n] = 1. If the DCT [i, j] is greater than or equal to DCmax, then DCT [i, j] = DRmax.

Then, the map creation unit 12 deletes the distance measurement data acquired in step S223 from the distance measurement data storage unit 26 (step S228), and returns to step S222.

On the other hand, in step S222, if there is no distance measurement data newly registered in the distance measurement data storage unit 26 (step S222; No), the failure erasing unit 13 performs attenuation processing on the previous range sensor map. It is determined whether or not a certain time (a cycle of attenuation processing, for example, 5 seconds) has elapsed since then (step S229). If a certain time has not elapsed (step S229; No), the range sensor map update process is terminated. If a certain time has passed (step S229; Yes), the fault erasing unit 13 attenuates the range sensor map (step S230), and then ends the range sensor map update process.

Step S230 is also called a fault elimination step. In this step, specifically, the obstacle erasing unit 13 applies the following to all the grid points of the range sensor map, that is, all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize. Perform processing. (However, of the following processes, the process of (1) is indispensable, and the process of (2) does not have to be performed.)
(1) If MA [i, j]> 0, then MA [i, j] = MA [i, j] -DRT [i, j]
(2) If MA [i, j] <0, then MA [i, j] = MA [i, j] + DRT [i, j]

By the above-mentioned area sensor map update process, the area sensor map is updated, and the existence probability of the obstacle is represented by the value of each grid point of the area sensor map.

Next, the collision sensor map update process performed in step S107 of the process of the map creation module 52 (FIG. 20) will be described with reference to FIG. 22.

Assuming that the coordinates of the current position (self-position) of the own machine are represented by [p, q], first, the obstacle erasing unit 13 uses the update amount (attenuation amount) of the grid points of the collision sensor map corresponding to the self-position. A certain DCT [p, q] is increased by a reference non-failure adjustment value (for example, 5) (step S321). That is, assuming that the reference non-failure adjustment value is 5, DCT [p, q] = DCT [p, q] + 5 is executed. However, when the value of DCT [p, q] is larger than the maximum value DCmax of DRT, DCT [p, q] = DCmax.

Next, the fault erasing unit 13 subtracts the update amount Δco from the value of the grid points of the collision sensor map corresponding to its own position (step S322). Specifically, MB [p, q] = MB [p, q] −Δco is executed. However, at this time, if MB [p, q] is less than the minimum value Lmin, MB [p, q] = Lmin.

Then, the map creation unit 12 determines whether or not the collision data newly registered in the collision data storage unit 25 exists (step S323). If there is collision data newly registered in the collision data storage unit 25 (step S323; Yes), the map creation unit 12 acquires the collision data from the collision data storage unit 25 and acquires it in step S105 of FIG. The collision data is converted into the coordinates in the collision sensor map by using the self-position and the orientation (step S324). These coordinates indicate the grid points (regions) where obstacles are detected by the collision sensor 31 on the collision sensor map. Here, for example, it is assumed that an obstacle is detected at the coordinates of [m, n] on the collision sensor map.

Then, the map creation unit 12 adds the update amount Δco to the value of the grid point where the obstacle is detected in the collision sensor map (step S325). Specifically, MB [m, n] = MB [m, n] + Δco is executed. However, at this time, if MB [m, n] exceeds the maximum value Lmax, MB [m, n] = Lmax.

Then, the obstacle erasing unit 13 reduces the value of the collision sensor map update amount (attenuation amount) DCT [m, n] corresponding to the grid point where the obstacle is detected by the reference obstacle adjustment value (for example, 5) (for example, 5). Step S326). If DCT [m, n] is 0 or less here, DCT [m, n] = 1.

Then, the map creation unit 12 deletes the collision data acquired in step S324 from the collision data storage unit 25 (step S327), and returns to step S323.

On the other hand, if there is no collision data newly registered in the collision data storage unit 25 in step S323 (step S323; No), the failure erasing unit 13 is constant after the previous collision sensor map is attenuated. It is determined whether or not the time (the cycle of the attenuation processing, for example, 5 seconds) has elapsed (step S328). If the fixed time has not elapsed (step S328; No), the collision sensor map update process is terminated. If a certain time has passed (step S328; Yes), the fault erasing unit 13 attenuates the collision sensor map (step S329), and then ends the range sensor map update process.

Step S329 is also called a fault elimination step. In this step, specifically, the fault erasing unit 13 performs the following processing for all the grid points of the collision sensor map, that is, all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize. I do. (However, of the following processes, the process of (1) is indispensable, and the process of (2) does not have to be performed.)
(1) If MB [i, j]> 0, MB [i, j] = MB [i, j] -DCT [i, j]
(2) If MB [i, j] <0, then MB [i, j] = MB [i, j] + DCT [i, j]

By the above collision sensor map update process, the collision sensor map is updated, and the value of each grid point of the collision sensor map represents the existence probability of an obstacle.

By each of the above processes, in the autonomous mobile device 102, the obstacle erasing unit 13 not only lowers the obstacle existence probability on the collision sensor map with time, but also reduces the obstacle existence probability on the range sensor map with time. Decrease according to. Then, the obstacle erasing unit 13 makes it possible to set the speed at which the existence probability is lowered to a different value for each grid point of the collision sensor map and the range sensor map, so that the existence probability can be set according to the property of the object. It can be reduced efficiently.

For example, since it is obvious that there is no obstacle at the self-position, the speed at which the probability of existence of the obstacle is lowered is increased by increasing the update amount (attenuation amount) of the value of the grid point at the self-position. .. In addition, by reducing the update amount (attenuation amount) of the value of the grid point corresponding to the place where the obstacle is detected, the speed of reducing the existence probability of the obstacle is reduced immediately after the obstacle is detected. There is. In addition, by increasing the update amount (attenuation amount) of the grid point value corresponding to the place where the obstacle is not detected, the speed at which the existence probability of the obstacle remaining on the map even though it is not detected is reduced. Is getting bigger.

By such processing, not only the obstacles that collided in the past but also the obstacles in the place that were detected by the range sensor 43 in the past but are out of the observation (scan) range of the range sensor 43 after that. After a certain amount of time, the information on the obstacle is erased. Therefore, the autonomous mobile device 102 can move by setting a route that does not avoid the place where the obstacle existed in the past. Therefore, the autonomous moving device 102 can prevent the obstacle avoidance operation that is originally unnecessary, and can move efficiently.

Further, in the autonomous mobile device 102, when the place where the obstacle existed in the past is outside the observation range of the range sensor 43, the information of the obstacle is deleted from the map with the passage of time. .. Therefore, since it is determined that there is no obstacle after a certain period of time, the place where the obstacle was present is likely to be included in the movement route to the destination, and as a result, the place is determined by the range sensor 43. The possibility of observing increases.

The autonomous mobile device 102 includes a collision sensor 31 and a range sensor 43 as the fault detection unit 30, but only one of them may be used. In this case, the collision sensor map or the range sensor map can be used as it is as the environment map.

Further, the obstacle erasing unit 13 determines the collision sensor map decay speed (adjusted by the reference obstacle adjustment value and the reference non-obstacle adjustment value), which is the speed at which the obstacle existence probability in the collision sensor map is decreased with the passage of time, and the measurement range. Area sensor map decay speed, which is the speed at which the probability of obstacle presence in the sensor map decreases over time (adjusted by the reference obstacle adjustment value for the area sensor map and the reference non-obstacle adjustment value for the area sensor map) Can be different. Since the collision sensor 31 can detect obstacles only in a very narrow range as compared with the range sensor 43, the adjustment amount of the collision sensor map attenuation speed is generally larger than the adjustment amount of the range sensor map attenuation speed. It is considered better to do it. This is because the follow-up speed to the obstacle in the collision sensor map can be increased.

Therefore, in the above-described third embodiment, the initial value of the DCT is made larger than the initial value of the DRT, and the reference non-disability adjustment value for the collision sensor map is made larger than the reference non-disability adjustment value for the range sensor map. The reference fault adjustment value for the collision sensor map is made larger than the reference fault adjustment value for the range sensor.

In each of the above embodiments, the position / orientation estimation unit 11 estimates the self-position and orientation of the autonomous moving device 100 by SLAM processing, but the present invention is not limited to this. For example, the position / orientation estimation unit 11 may estimate its own position and attitude only from the odometry information that can be acquired from the drive unit 42 without performing SLAM processing. Further, the position / attitude estimation unit 11 may acquire radio waves from, for example, a GPS (Global Positioning System) satellite and estimate its own position and attitude by GPS.

Further, in each of the above embodiments, the distance measurement data storage unit 26 stores distance measurement data indicating the distance to the object detected by the range sensor 43 together with information on the angle at which the object is detected. , Not limited to this. For example, the ranging data storage unit 26 acquires the self-position and orientation of the autonomous moving device when the surroundings are observed (scanned) by the range sensor 43, and the range sensor 43, as in the collision data storage unit 25. Data indicating the position of an object (obstacle) detected by the range sensor 43, which is obtained based on the distance measurement data and the angle obtained, may be stored. In this way, the process of converting the distance measurement data into the coordinates of the map (step S202, step S223) in the range sensor map update process (FIGS. 7, 11, and 21) becomes unnecessary.

Further, in each of the above embodiments, when creating an environment map in which each sensor map is integrated, the maximum value of the value of each grid point of each sensor map is adopted, but the present invention is not limited to this. For example, the product of the existence probabilities of obstacles in each sensor map may be adopted. For example, in step S108 of FIG. 6 or FIG. 20, the map creation unit 12 calculates the range sensor map and the collision sensor map for all i and j of 0 ≦ i ≦ xsize and 0 ≦ j ≦ ysize by the following calculation. You may create an environment map that integrates.
MI [i, j] = 1- (1 / (1 + exp (k × MA [i, j])) × (1 / (1 + exp (k × MB [i, j])))

Further, in the first embodiment, the range sensor 43 may not be provided, and the obstacle may be detected only by the collision sensor 31. In this case, it is not necessary to create and update the range sensor map, and the autonomous mobile device 100 can use only the collision sensor map as the environment map.

Further, in each of the above-described embodiments, instead of providing the range sensor 43, a depth camera or sonar is provided, and obstacles observed by these are recorded on an environment map similar to the range sensor map. good.

Further, in each of the above-described embodiments, the range sensor 43 is not provided, and the obstacle is detected by the visual SLAM using the image taken by the imaging unit 41, and the obstacle detected by the visual SLAM is detected by the range sensor. It may be recorded on an environmental map similar to the map.

Further, in each of the above-described embodiments, the user is notified of an error or the like via, for example, the communication unit 44, but the autonomous mobile device includes an output unit such as a display or a speaker, and the output unit is provided. May notify the user of an error or the like. Further, in each of the above-described embodiments, when receiving an instruction of a destination or the like from a user, for example, it is performed via a communication unit 44, but the autonomous mobile device includes an input unit such as a touch panel or a microphone. , The input unit may accept instructions from the user.

Further, in the range sensor map update process (FIGS. 7 and 11) of the first or second embodiment described above, a process of attenuating the range sensor map immediately before step S201 (collision sensor map update process (FIG. 8). ), The process corresponding to step S301) may be performed. Specifically, the fault erasing unit 13 performs the following processing for all i and j. However, the following Δs (attenuation amount for the range sensor map) is set in advance to a value smaller than the update amount Δr or the update amount Δc, for example, 1 mag.
(1) If MA [i, j]> 0, MA [i, j] = MA [i, j] −Δs
(2) If MA [i, j] <0, then MA [i, j] = MA [i, j] + Δs

Similar to the attenuation processing of the collision sensor map, the attenuation amount Δs of the value of the grid point is stored for each grid point of the survey area sensor map (for example, in DS [i, j]), and each grid point is used. The amount of attenuation Δs of the value of each grid point may be increased or decreased depending on whether or not the obstacle is detected by the range sensor 43. By the above processing, it is possible to prevent the ghost of the obstacle outside the observation (scanning) range of the range sensor 43 from remaining, as in the third embodiment.

The functions of the autonomous mobile devices 100, 101, and 102 can also be performed by a computer such as a normal PC (Personal Computer). Specifically, in the above embodiment, it has been described that the program of the autonomous movement control processing performed by the autonomous movement devices 100, 101, 102 is stored in the ROM of the storage unit 20 in advance. However, the program is stored and distributed in a computer-readable recording medium such as a flexible disk, a CD-ROM (Compact Disk Read Only Memory), a DVD (Digital Versailles Disc), and an MO (Magnet-Optical Disc), and the program is distributed. You may configure a computer that can realize each of the above-mentioned functions by reading and installing the above-mentioned in a computer.

Although the preferred embodiment of the present invention has been described above, the present invention is not limited to the specific embodiment, and the present invention includes the invention described in the claims and the equivalent range thereof. Is done. The inventions described in the claims of the original application of the present application are described below.

(Appendix 1)
Obstacle detection unit that detects obstacles and
A cartography unit that records information on obstacles detected by the obstacle detection unit on an environmental map,
An obstacle erasing unit that erases obstacle information recorded by the map-creating unit from the environmental map over time, and an obstacle erasing unit.
A route setting unit that sets a movement route based on the information recorded on the environment map, and
An autonomous mobile device equipped with.

(Appendix 2)
The obstacle detection unit includes a proximity obstacle detection unit that detects an obstacle by approaching the obstacle.
The autonomous mobile device according to Appendix 1.

(Appendix 3)
The cartography unit records the information of the obstacle on the environment map for each position of the obstacle.
The autonomous mobile device according to Appendix 1 or 2.

(Appendix 4)
The obstacle erasing unit adjusts the erasing speed, which is the speed at which the obstacle erasing unit erases the obstacle information from the environment map, according to the number of times the obstacle detecting unit detects the obstacle at the position. ,
The autonomous mobile device according to Appendix 3.

(Appendix 5)
When the obstacle detecting unit detects the obstacle, the obstacle erasing unit slows the erasing speed at the position where the obstacle is detected by a reference obstacle adjustment value.
The autonomous mobile device according to Appendix 4.

(Appendix 6)
The obstacle detection unit includes a proximity obstacle detection unit that detects the obstacle by approaching the obstacle, and a distant obstacle detection unit that detects the obstacle from a distance.
The reference obstacle adjustment value when the proximity obstacle detection unit detects the obstacle is larger than the reference obstacle adjustment value when the distance obstacle detection unit detects the obstacle.
The autonomous mobile device according to Appendix 5.

(Appendix 7)
If the obstacle detecting unit does not detect the obstacle, the obstacle erasing unit increases the erasing speed at the position where the obstacle is not detected by a reference non-obstacle adjustment value.
The autonomous mobile device according to any one of Appendix 4 to 6.

(Appendix 8)
The obstacle detection unit includes a proximity obstacle detection unit that detects the obstacle by approaching the obstacle, and a distant obstacle detection unit that detects the obstacle from a distance.
The reference non-obstacle adjustment value when the proximity obstacle detection unit does not detect the obstacle is larger than the reference non-obstacle adjustment value when the distant obstacle detection unit does not detect the obstacle.
The autonomous mobile device according to Appendix 7.

(Appendix 9)
The obstacle detection unit includes a collision sensor that detects an obstacle by detecting the collision with the obstacle.
The autonomous mobile device according to any one of Appendix 1 to 8.

(Appendix 10)
The obstacle detection unit includes a cliff sensor that detects an obstacle by detecting a cliff that is a place to fall.
The autonomous mobile device according to any one of Appendix 1 to 9.

(Appendix 11)
Further provided with an obstacle position acquisition unit for acquiring the position of the obstacle from a distance,
The map creation unit records the information on the position of the obstacle acquired by the obstacle position acquisition unit on the environment map.
The autonomous mobile device according to any one of Appendix 1 to 10.

(Appendix 12)
The obstacle erasing unit erases the obstacle information detected by the obstacle detecting unit from the environment map, and does not delete the obstacle information existing at the position acquired by the obstacle position acquisition unit from the environment map.
The autonomous mobile device according to Appendix 11.

(Appendix 13)
The obstacle erasing unit also erases the information of the obstacle existing at the position acquired by the obstacle position acquisition unit from the environment map, and the speed of erasing the information of the obstacle detected by the obstacle detection unit from the environment map is high. , It is faster than the speed of deleting the information of the obstacle existing at the position acquired by the obstacle position acquisition unit from the environment map.
The autonomous mobile device according to Appendix 11.

(Appendix 14)
The obstacle position acquisition unit acquires the position of the obstacle by using the information from the range sensor that acquires the distance measurement data which is the distance to the obstacle.
The autonomous mobile device according to any one of Appendix 11 to 13.

(Appendix 15)
The obstacle position acquisition unit confirms the existence of an obstacle in each predetermined direction, and when the existence of the obstacle is confirmed, acquires distance measurement data which is the distance to the obstacle.
The map creation unit records in the environment map the ratio of the number of times the distance measurement data is acquired to the number of times the obstacle position acquisition unit confirms the existence of the obstacle as the existence probability.
The autonomous mobile device according to any one of Appendix 11 to 14.

(Appendix 16)
The cartographic unit calculates the existence probability of the obstacle at the position on the environmental map corresponding to the position of the obstacle, and records the existence probability in the environmental map.
The autonomous mobile device according to any one of Appendix 1 to 15.

(Appendix 17)
The cartographic unit records the existence probability on the environmental map with logarithmic odds.
The autonomous mobile device according to Appendix 15 or 16.

(Appendix 18)
Obstacle detection steps to detect obstacles and
A map creation step that records information on obstacles detected in the obstacle detection step on an environmental map, and
The obstacle elimination step of deleting from the environmental map by attenuating the obstacle information recorded in the map creation step, and
A route setting step for setting a movement route based on the information recorded on the environment map, and
Autonomous movement methods including.

(Appendix 19)
On the computer
Obstacle detection step to detect obstacles,
A map creation step that records information on obstacles detected in the obstacle detection step on an environmental map.
The obstacle erasing step of deleting the obstacle information recorded in the map creation step from the environmental map according to the passage of time, and the obstacle erasing step.
A route setting step for setting a movement route based on the information recorded on the environment map,
A program to execute.

(Appendix 20)
An obstacle detection unit that detects an obstacle by approaching the obstacle,
A cartography unit that records information on obstacles detected by the obstacle detection unit on an environmental map,
An obstacle erasing unit that erases obstacle information recorded by the map-creating unit from the environmental map, and an obstacle erasing unit.
A route setting unit that sets a movement route based on the information recorded on the environment map, and
An autonomous mobile device equipped with.

(Appendix 21)
An obstacle detection step that detects an obstacle by approaching it, and
A map creation step that records information on obstacles detected in the obstacle detection step on an environmental map, and
The obstacle elimination step of deleting the obstacle information recorded in the map creation step from the environment map, and the obstacle elimination step.
A route setting step for setting a movement route based on the information recorded on the environment map, and
Autonomous movement methods including.

(Appendix 22)
On the computer
Obstacle detection step, which detects an obstacle by being close to the obstacle,
A map creation step that records information on obstacles detected in the obstacle detection step on an environmental map.
The obstacle erasing step for erasing the obstacle information recorded in the map creation step from the environmental map, and the obstacle erasing step.
A route setting step for setting a movement route based on the information recorded on the environment map,
A program to execute.

10 ... Control unit, 11 ... Position / orientation estimation unit, 12 ... Map creation unit, 13 ... Obstacle erasing unit, 14 ... Route setting unit, 15 ... Movement control unit, 20 ... Storage unit, 21 ... Image storage unit, 22 ... Features Point storage unit, 23 ... Map storage unit, 24 ... Update amount storage unit, 25 ... Collision data storage unit, 26 ... Distance measurement data storage unit, 27 ... Cliff data storage unit, 30 ... Failure detection unit, 31, 31a, 31b , 31c ... Collision sensor, 32 ... Cliff sensor, 41 ... Imaging unit, 42, 42a, 42b ... Drive unit, 43 ... Range sensor, 44 ... Communication unit, 51 ... Position and orientation estimation module, 52 ... Map creation module, 53 ... Route setting module, 54 ... Movement control module, 70 ... Obstacles, 100, 101, 102 ... Autonomous mobile device

In order to achieve the above object, the autonomous mobile device of the present invention
A route setting means for setting a movement route so as to avoid the obstacle based on the map information in which the position information corresponding to the obstacle is registered, and
An erasing means for erasing the location information from the map information according to the passage of time after registration, and
Until the position information is deleted, when the position information is registered again by updating the map information by setting the movement route by the route setting means, the position information by the erasing means is deleted. Control means to control the time to increase
To be equipped.

Claims (19)

Obstacle detection unit that detects obstacles and
A cartography unit that records information on obstacles detected by the obstacle detection unit on an environmental map,
An obstacle erasing unit that erases obstacle information recorded by the map-creating unit from the environmental map over time, and an obstacle erasing unit.
A route setting unit that sets a movement route based on the information recorded on the environment map, and
An autonomous mobile device equipped with. The obstacle detection unit includes a proximity obstacle detection unit that detects an obstacle by approaching the obstacle.
The autonomous mobile device according to claim 1. The cartography unit records the information of the obstacle on the environment map for each position of the obstacle.
The autonomous mobile device according to claim 1 or 2. The obstacle erasing unit adjusts the erasing speed, which is the speed at which the obstacle erasing unit erases the obstacle information from the environment map, according to the number of times the obstacle detecting unit detects the obstacle at the position. ,
The autonomous mobile device according to claim 3. When the obstacle detecting unit detects the obstacle, the obstacle erasing unit slows the erasing speed at the position where the obstacle is detected by a reference obstacle adjustment value.
The autonomous mobile device according to claim 4. The obstacle detection unit includes a proximity obstacle detection unit that detects the obstacle by approaching the obstacle, and a distant obstacle detection unit that detects the obstacle from a distance.
The reference obstacle adjustment value when the proximity obstacle detection unit detects the obstacle is larger than the reference obstacle adjustment value when the distance obstacle detection unit detects the obstacle.
The autonomous mobile device according to claim 5. If the obstacle detecting unit does not detect the obstacle, the obstacle erasing unit increases the erasing speed at the position where the obstacle is not detected by a reference non-obstacle adjustment value.
The autonomous mobile device according to any one of claims 4 to 6. The obstacle detection unit includes a proximity obstacle detection unit that detects the obstacle by approaching the obstacle, and a distant obstacle detection unit that detects the obstacle from a distance.
The reference non-obstacle adjustment value when the proximity obstacle detection unit does not detect the obstacle is larger than the reference non-obstacle adjustment value when the distant obstacle detection unit does not detect the obstacle.
The autonomous mobile device according to claim 7. The obstacle detection unit includes a collision sensor that detects an obstacle by detecting the collision with the obstacle.
The autonomous mobile device according to any one of claims 1 to 8. The obstacle detection unit includes a cliff sensor that detects an obstacle by detecting a cliff that is a place to fall.
The autonomous mobile device according to any one of claims 1 to 9. Further provided with an obstacle position acquisition unit for acquiring the position of the obstacle from a distance,
The map creation unit records the information on the position of the obstacle acquired by the obstacle position acquisition unit on the environment map.
The autonomous mobile device according to any one of claims 1 to 10. The obstacle erasing unit erases the obstacle information detected by the obstacle detecting unit from the environment map, and does not delete the obstacle information existing at the position acquired by the obstacle position acquisition unit from the environment map.
The autonomous mobile device according to claim 11. The obstacle erasing unit also erases the information of the obstacle existing at the position acquired by the obstacle position acquisition unit from the environment map, and the speed of erasing the information of the obstacle detected by the obstacle detection unit from the environment map is high. , It is faster than the speed of deleting the information of the obstacle existing at the position acquired by the obstacle position acquisition unit from the environment map.
The autonomous mobile device according to claim 11. The obstacle position acquisition unit acquires the position of the obstacle by using the information from the range sensor that acquires the distance measurement data which is the distance to the obstacle.
The autonomous mobile device according to any one of claims 11 to 13. The obstacle position acquisition unit confirms the existence of an obstacle in each predetermined direction, and when the existence of the obstacle is confirmed, acquires distance measurement data which is the distance to the obstacle.
The map creation unit records in the environment map the ratio of the number of times the distance measurement data is acquired to the number of times the obstacle position acquisition unit confirms the existence of the obstacle as the existence probability.
The autonomous mobile device according to any one of claims 11 to 14. The cartographic unit calculates the existence probability of the obstacle at the position on the environmental map corresponding to the position of the obstacle, and records the existence probability in the environmental map.
The autonomous mobile device according to any one of claims 1 to 15. The cartographic unit records the existence probability on the environmental map with logarithmic odds.
The autonomous mobile device according to claim 15 or 16. Obstacle detection steps to detect obstacles and
A map creation step that records information on obstacles detected in the obstacle detection step on an environmental map, and
The obstacle elimination step of deleting from the environmental map by attenuating the obstacle information recorded in the map creation step, and
A route setting step for setting a movement route based on the information recorded on the environment map, and
Autonomous movement methods including. On the computer
Obstacle detection step to detect obstacles,
A map creation step that records information on obstacles detected in the obstacle detection step on an environmental map.
The obstacle erasing step of deleting the obstacle information recorded in the map creation step from the environmental map according to the passage of time, and the obstacle erasing step.
A route setting step for setting a movement route based on the information recorded on the environment map,
A program to execute.

Images