JP7719430B2 - Autonomous driving device and autonomous driving method - Google Patents

Description

The present invention relates to an autonomous driving device and an autonomous driving method, and in particular to an autonomous driving device and an autonomous driving method suitable for a cleaning device used to clean and remove deposits (accumulated powder) within a specified area.

For example, in coke plants, coal dust (fallen dust) accumulates at the top of coke ovens due to leaks from coal loading cars that load coal into the oven, or from coal storage. If this accumulated coal dust is left unattended, the sulfur components in the coal dust can accelerate corrosion of metal structures such as riser pipes, and it can also scatter into the surrounding area, causing adverse effects on the environment. Therefore, to protect the oven body and prevent dust from scattering, the coal dust accumulated at the top is cleaned and removed. However, cleaning and removing coal dust accumulated at the top of the oven is hard work performed in a high-temperature, dusty environment, and is dangerous work that poses the risk of heatstroke and dust inhalation.

As an example of a cleaning device that performs automatic cleaning work, Patent Document 1 discloses a cleaning device that cleans an entire area while performing mapping. This cleaning device classifies the entire cleaning area into three sections: unknown area, uncleaned area, and cleaned area, and has autonomous driving logic that operates the robot to eliminate the unknown area. Once the unknown area is eliminated, the boundary between the uncleaned area and the cleaned area is moved and cleaning is carried out. This type of technology is extremely useful when vacuuming and cleaning small amounts of dust scattered over an entire area in a cleaning area that varies in scope and where unknown areas exist each time.

Patent No. 6410704

When a large amount of fallen dust (usually 1 kg/ ^m2 or more) exists, such as in the top of a coke oven, a method of suction-collecting and storing it in a cleaning device requires a very large device. Therefore, a method of scraping up the fallen dust using a scraper or the like and accumulating it in a designated location is preferable. However, the cleaning device of Patent Document 1 re-creates the mapping each time, and the origin and coordinate system change, so it is not possible to perform operations such as cleaning the same area in the same pattern each time, as in the top of a coke oven, and then accumulating the fallen dust in a designated location. Furthermore, there is a problem in that the process of searching for unknown areas when cleaning the same area in the same pattern each time reduces time efficiency.

In contrast, with conventional SLAM technology, which uses a map of the cleaning area for autonomous navigation, the autonomous vehicle can clean the same area in the same pattern each time and can also scrape up and accumulate the fallen powder in a designated location.However, if a large amount of fallen powder accumulates at the top of the coke oven, exceeding the pushing force of the autonomous vehicle, causing it to become stuck when the scraper is unable to push it all away, or if an object is placed there that completely blocks the autonomous vehicle's movement path, the autonomous vehicle can become stuck and be unable to continue cleaning.
Therefore, an object of the present invention is to provide an autonomous driving device and an autonomous driving method that solve the problems of the conventional technology described above, and that can travel along a fixed route easily and in a short time, while reducing the risk of getting stuck along the route.

To solve the above problems, the present invention provides an autonomous driving device and autonomous driving method that causes a mobile vehicle to autonomously travel via a plurality of way points p within a designated area. When the mobile vehicle is autonomously traveling to way point _p , the device estimates its own position and generates way point information b from pre-stored obstacle position information a1 within the designated area, designated area information a2, and final destination point information a3. A travel route to way point _p along which the mobile vehicle should travel is calculated based on the estimated own position and the generated way point information b. The method then moves the mobile vehicle to way point _p along this travel route. To prevent the mobile vehicle from becoming stuck during autonomous traveling, if it is determined that the moving mobile vehicle is stuck and cannot reach way point _p , the device changes the way point p as the destination of the mobile vehicle. More preferably, the device also performs abnormality processing to change the way point p as the destination of the mobile vehicle when it is determined that a travel route for the mobile vehicle cannot be generated based on obstacle information a acquired by the range measurement device. That is, for example, when applied to an autonomous driving device that uses a cleaning scraper to scrape up and clean deposits within a specified area while autonomously driving, if the cleaning scraper becomes unable to push through a large amount of deposits and becomes stuck, or if the driving route is blocked by an obstacle that is not included in the pre-stored obstacle position information a1, abnormality processing is performed to change the waypoint p to which the mobile cart is to be moved, so that the mobile cart does not get stuck while autonomously driving.

That is, the features of the autonomous driving device and autonomous driving method of the present invention for solving the above problems are as follows.
[1] An autonomous driving device that autonomously drives a mobile carriage (1) equipped with a driving device (2) for driving, while passing through a plurality of via points (p) within a designated area (A),
a calculation and control device (4) that estimates the self-position of the mobile carriage (1), generates waypoint information (b) from obstacle position information a1 within a designated area (A) that is stored in advance, designated area information a2, and final target point information a3, calculates a travel route to a waypoint (p _i ) along which the mobile carriage (1) should move based on the estimated self-position and the generated waypoint information (b), and controls a drive device (2) so that the mobile carriage (1) moves to the waypoint (p _i ) according to the travel route;
The arithmetic and control device (4) determines whether a mobile carriage (1) moving toward a waypoint (p _i ) is stuck and unable to reach the waypoint (p _i ), and if it determines that the mobile carriage (1) is stuck, changes the waypoint (p) to which the mobile carriage (1) should be moved, and controls the drive device (2) to move the mobile carriage (1) to the new waypoint (p).

[2] The autonomous driving device of the above [1] further comprises a range measurement device (3) that measures the surrounding environment of the mobile carriage (1) and acquires obstacle information a,
An autonomous driving device characterized in that an arithmetic and control device (4) estimates the self-position of a mobile cart (1) by comparing obstacle position information a1 within a designated area (A) stored in advance with obstacle information a acquired by a range measurement device (3).
[3] In the autonomous driving device of [2] above, the arithmetic and control device (4) determines whether or not a driving route to a via point (p _i ) to which the mobile carriage (1) should move can be generated based on the obstacle information a acquired by the range measurement device (3), and if it determines that the route cannot be generated, changes the via point (p) to which the mobile carriage (1) should move, and controls the drive device (2) to move the mobile carriage (1) to the changed via point (p).

[4] In the autonomous driving device according to any one of [1] to [3] above, the mobile carriage (1) is provided with a work tool (5) for performing a predetermined task within the designated area (A) and an actuator (6) for driving the work tool (5);
The waypoint information b includes information on the carriage position of the mobile carriage (1), the carriage attitude of the mobile carriage (1), and operation commands for the actuator (6).
[5] In the autonomous mobile device of [4] above, the work tool (5) is a cleaning scraper (5a) that scrapes and cleans deposits within the designated area (A),
The autonomous mobile device is characterized in that the actuator (6) drives the cleaning scraper (5a) so that the cleaning scraper (5a) moves up and down to come into contact with or separate from the road surface.

[6] In the autonomous driving device of any of the above [1] to [5], the arithmetic and control device (4) calculates a target travel time T to the waypoint ( _pi ) from the travel distance ΔL to the waypoint ( _pi ) and the attitude angle change amount Δθ when the mobile carriage (1) moves to the waypoint ( _pi ), and determines whether the mobile carriage (1) is stuck by comparing the actual value t of the elapsed time from the start of movement of the mobile carriage (1) with the target travel time T.
[7] In any one of the autonomous driving devices [2] to [6] above, a movement amount detection device (7) is further provided that detects the movement amount of the mobile carriage (1) from the operation amount of the drive device (2),
An autonomous driving device characterized in that the arithmetic and control device (4) compares obstacle position information a1 within a designated area (A) stored in advance with obstacle information a acquired by the range measurement device (3), and also estimates the self-position of the mobile cart (1) by a particle filter based on a SLAM algorithm using the amount of movement of the mobile cart (1) detected by a movement amount detection device (7).

[8] An autonomous traveling method for autonomously traveling a mobile carriage (1) equipped with a driving device (2) for traveling while passing through a plurality of via points (p) within a designated area (A),
A step (a) of generating waypoint information (b) from obstacle position information (a1) within a designated area (A), designated area information (a2), and final target point information (a3) stored in advance;
A step (a) of estimating the self-position of the mobile cart (1);
a step (c) of calculating a travel route to a via point (p _i ) along which the mobile carriage (1) will move based on the estimated self-position and the generated via point information b, and controlling a drive device (2) so that the mobile carriage (1) moves to the via point (p _i ) along the travel route;
An autonomous driving method characterized by comprising a step (d) of determining whether a mobile carriage (1) moving toward a waypoint (p _i ) is stuck and unable to reach the waypoint (p _i ), and if it is determined that the mobile carriage (1) is stuck, changing the waypoint (p) to which the mobile carriage (1) should be moved, and controlling a drive device (2) to move the mobile carriage (1) to the new waypoint (p).

[9] In the autonomous driving method of [8] above, in step (i), the surrounding environment of the mobile cart (1) is measured by a range measurement device (3) to acquire obstacle information a, and the self-position of the mobile cart (1) is estimated by comparing obstacle position information a1 within a designated area (A) stored in advance with the obstacle information a acquired by the range measurement device (3).
[10] The autonomous driving method of [9] above, further comprising a step (e) of determining whether or not a driving route to a via point (p _i ) to which the mobile carriage (1) should move can be generated based on the obstacle information a acquired by the range measurement device (3), and if it is determined that the route cannot be generated, changing the via point (p) to which the mobile carriage (1) should move, and controlling the drive device (2) to move the mobile carriage (1) to the changed via point (p).

[11] In the autonomous traveling method according to any one of [8] to [10] above, the mobile carriage (1) comprises a work tool (5) for performing a predetermined task within the designated area (A), and an actuator (6) for driving the work tool (5);
An autonomous traveling method characterized by performing work with a work tool (5) while traveling a mobile carriage (1) within a designated area (A).
[12] In the autonomous traveling method according to [11] above, the work tool (5) is a cleaning scraper (5a) that scrapes and cleans deposits on the ground within the designated area (A),
The actuator (6) drives the cleaning scraper (5a) so that the cleaning scraper (5a) moves up and down to come into contact with or separate from the road surface;
The autonomous traveling method comprises traveling a mobile cart (1) within a designated area (A), driving a cleaning scraper (5a) with an actuator (6), scraping up deposits with the cleaning scraper (5a), and moving the scraped up deposits to a predetermined location and accumulating them.

In this invention, a mobile carriage is made to travel autonomously within a designated area, passing through a plurality of way points p. When the mobile carriage is made to travel autonomously to way point _p , the mobile carriage's own position is estimated and way point information b is generated from pre-stored obstacle position information a1 within the designated area, designated area information a2, and final destination point information a3. A travel route to way point _p along which the mobile carriage should travel is calculated based on the estimated own position and the generated way point information b. The mobile carriage is then moved to way point _p along this travel route. This makes it possible to move the mobile carriage to a desired location along a set route easily and in a short time. For example, if this invention is applied to an autonomous traveling device that cleans a designated area by scraping up deposits within the designated area with a cleaning scraper while autonomously traveling, the same area can be cleaned in the same pattern each time and the deposits can be accumulated in a designated location, allowing the designated area to be cleaned easily, quickly, and efficiently.

Furthermore, if it is determined that the moving mobile cart is stuck and cannot reach way point p _i , the way point p to which the mobile cart is to be moved is changed, and more preferably, if it is determined that a travel path for the mobile cart cannot be generated based on the obstacle information a acquired by the range measurement device, abnormality processing is performed to change the way point p to which the mobile cart is to be moved.This reduces the risk of the mobile cart getting stuck during autonomous driving, and allows the mobile cart to be moved to the desired location more reliably and in a shorter time by autonomous driving.

Therefore, for example, when applied to an autonomous mobile device that cleans a designated area by scraping up deposits with a cleaning scraper while autonomously traveling, if the cleaning scraper becomes unable to push through a large amount of deposits and becomes stuck, or if the travel path is blocked by an obstacle that is not included in the pre-stored obstacle position information a1, abnormality processing is performed to change the way point p of the mobile carrier's destination, preventing the mobile carrier from getting stuck during autonomous traveling, thereby allowing for more efficient cleaning of the designated area. Also, even when the way point p of the mobile carrier's destination is changed when a stuck state occurs or a travel path cannot be generated, the new way point p can be specified in advance, which has the advantage of eliminating the need to implement a complex abnormality processing program.
From the above, the autonomous mobile device and method of the present invention are suitable as a cleaning device for the top of a coke oven where coal dust (fallen dust) accumulates, and can efficiently and appropriately carry out cleaning and removal work of the deposits.

FIG. 1 is a side view schematically illustrating an embodiment of an autonomous driving device of the present invention. 2 is a plan view of the autonomous driving device of the embodiment of FIG. 1 ; FIG. 2 is a block diagram showing a control system for each component of the autonomous driving device of the embodiment of FIG. 1. Block diagram of the arithmetic and control device shown in FIG. 1 is a flow diagram of the control logic of the autonomous driving device of the present invention. FIG. 1 is an explanatory diagram illustrating an example of a method for setting a via point and generating via point information according to the present invention. FIG. 1 is an explanatory diagram illustrating an example of a method for setting a via point and generating via point information according to the present invention. FIG. 1 is an explanatory diagram illustrating an example of a method for setting a via point and generating via point information according to the present invention. FIG. 1 is a diagram showing a schematic diagram of via points set and via point information generated in the present invention in the order of the via points. FIG. 1 is an explanatory diagram showing a method for creating obstacle position information a1 using a 3D shape measuring device in the present invention. FIG. 1 is an explanatory diagram showing a cleaning area in the first embodiment; FIG. 10 is an explanatory diagram showing a movement trajectory of the cleaning device in the first embodiment. FIG. 10 is an explanatory diagram showing a cleaning area in a second embodiment; FIG. 10 is an explanatory diagram showing the movement trajectory of the cleaning device according to the second embodiment of the present invention. FIG. 10 is an explanatory diagram showing a movement trajectory of a cleaning device of a comparative example in Example 2. FIG. 10 is an explanatory diagram showing a cleaning area in the third embodiment. FIG. 10 is an explanatory diagram showing the movement trajectory of the cleaning device according to the third embodiment of the present invention. FIG. 10 is an explanatory diagram showing a movement trajectory of a cleaning device of a comparative example in Example 3.

The present invention provides an autonomous driving device and autonomous driving method that allows a mobile carriage 1 equipped with a driving device 2 for driving to travel autonomously within a designated area A while passing through a plurality of way points p. When the mobile carriage 1 is caused to travel autonomously to way point p _i (the i-th way point p), the device basically estimates the self-position of the mobile carriage 1 and generates way point information b from pre-stored obstacle position information a1 within the designated area A, designated area information a2, and final target point information a3, calculates a driving route to way point p _i along which the mobile carriage will move based on the estimated self-position and the generated way point information b, and moves the mobile carriage 1 to way point p _i along that driving route. In order to prevent the mobile carriage 1 from getting stuck during this autonomous travel, if it is determined that the mobile carriage 1 is stuck and cannot reach way point p _i while moving, the way point p to which the mobile carriage 1 is to be moved is changed, and more preferably, if it is determined that a travel route for the mobile carriage 1 cannot be generated based on the obstacle information a acquired by the range measurement device 3, abnormality processing is performed to change the way point p to which the mobile carriage 1 is to be moved.

For this reason, the autonomous driving device of the present invention comprises a mobile carriage 1 equipped with a drive unit 2 for driving, a calculation and control unit 4 that estimates the mobile carriage 1's own position, generates waypoint information b based on the above information a1 to a3, calculates the driving route, and determines whether the vehicle is stuck, and controls the drive unit 2 based on these results to cause the mobile carriage 1 to drive autonomously. More preferably, the autonomous driving device also comprises a range measurement device 3 that measures the environment around the mobile carriage 1 to obtain obstacle information a, and a movement amount detection device 7 that detects the movement amount of the mobile carriage 1 from the amount of operation of the drive unit 2. The components of the device of the present invention will be described in detail below using the embodiment shown in Figures 1 to 3 as an example.

The driving device 2 for travel provided on the mobile carriage 1 includes a rotating body for travel such as a wheel or caterpillar, a driving source such as a motor for driving the rotating body for rotation, and a steering means for the rotating body for travel.
Furthermore, the self-position of the mobile cart 1 is estimated, for example, by providing a range measurement device 3 in the autonomous driving device, measuring the surrounding environment of the mobile cart 1 using this range measurement device 3 to acquire obstacle information a, and then comparing the obstacle position information a1 within the designated area A that has been stored in advance with the obstacle information a acquired by the range measurement device 3 using the arithmetic and control device 4.
The range measurement device 3 may be any device that acquires obstacle information a around the mobile cart 1 in real time, and may, for example, be one that acquires obstacle information a by measuring the surrounding environment of the mobile cart 1 using a camera or laser range finder.
In addition to using the obstacle information a acquired by the range measurement device 3, the position of the mobile carriage 1 may be estimated based on position information obtained by a satellite positioning system such as GPS.

The obstacle information a is information about the shape and position of obstacles (for example, structures, structural parts of buildings, equipment, and other objects) that exist around the mobile cart 1.
Obstacle position information a1 is position information (e.g., planar position coordinates) of an object that may obstruct the movement of the mobile cart 1 within the designated area A, and is typically position information of structures, structural parts of buildings, equipment, etc., but is not limited to these.
The designated area information a2 is position information (e.g., planar position coordinates of the outer edge (periphery) of the designated area A) of the designated area A through which the autonomous mobile device moves autonomously to perform a specified task.For example, if the autonomous mobile device is a cleaning device, this may include position information (cleaning area information a2) about the cleaning area, which is the designated area A.
The final target point information a3 is position information (e.g., planar position coordinates) of the point where the autonomous mobile device will ultimately arrive to complete a specified task. For example, if the autonomous mobile device is a cleaning device, this may include position information about the point where the collected sediment will ultimately be accumulated (deposit accumulation location information a3).
The obstacle position information a1, the designated area information a2, and the final target point information a3 are stored in advance in a storage means (storage device).

The way point information b is information about the way points (travel route) that the mobile vehicle 1 follows when moving to the target point within the designated area A to perform a specified task, the attitude of the mobile vehicle 1 at each way point, and the operation of the actuator 6 that drives the work tool 5, and includes the vehicle position (x coordinate, y coordinate) of the mobile vehicle 1 at each way point p, the vehicle attitude of the mobile vehicle 1 (carriage attitude angle θ), and an actuator operation command m in the case where the mobile vehicle 1 is equipped with a work tool 5 and an actuator 6 for driving it as described below, and is data that lists j way points _p1 to _pj that the mobile vehicle 1 travels through in order. Note that the vehicle attitude is the direction (azimuth) that the mobile vehicle 1 is facing in the xy coordinate system, i.e., the orientation (azimuth) at the time of arriving at a way point.

As described above, the arithmetic and control device 4 (i) estimates the self-position of the mobile cart 1, (ii) generates way point information b from the above information a1 to a3, (iii) calculates a travel route to way point p _i to which the mobile cart 1 should be moved based on the estimated self-position and the generated way point information b, (iv) determines whether the mobile cart 1 is stuck while moving towards way point p _i and cannot reach way point p _i , and if it determines that the mobile cart 1 is stuck, changes way point p to which the mobile cart 1 should be moved, and (v) further, as necessary, determines whether a travel route to way point p _i to which the mobile cart 1 should be moved can be generated based on the obstacle information a acquired by the range measurement device 3, and if it determines that it cannot be generated, changes way point p to which the mobile cart 1 should be moved, and so on, and based on these, controls the drive device 2 to cause the mobile cart 1 to travel autonomously. The waypoint information b generated in (ii) above is stored in a storage means (memory device) and is read in (iii) above when heading towards each waypoint p, and a travel route is calculated.
Examples of cases in which the mobile cart 1 becomes stuck during autonomous driving include, but are not limited to, cases in which, when applied to an autonomous driving device (cleaning device) that uses a cleaning scraper to scrape up and clean deposits within a designated area A while autonomously driving, the cleaning scraper is unable to push through the large amount of deposits, causing the mobile cart 1 to become stuck.

The mobile carriage 1 can be equipped with a work tool 5 for performing a predetermined task within the designated area A, and an actuator 6 for driving this work tool 5.
There are no particular restrictions on the work tool 5 that the mobile cart 1 is equipped with, but the autonomous driving device of the present invention is suitable as a cleaning device that scrapes up and cleans deposits within a designated area A while traveling autonomously, and in this case the work tool 5 is a cleaning scraper 5a that scrapes up and cleans deposits within the designated area A.
Furthermore, the mobile carriage 1 can be equipped with various work tools 5 for performing predetermined tasks within the designated area A, in addition to the above-mentioned cleaning scraper 5a. Examples include a sprinkler device that sprinkles cleaning liquid (such as water), an arm that lifts objects (such as the lid of a coal inlet) within the designated area A, etc. In this case, the actuator 6 drives the work tool 5, such as the sprinkler device or the arm, to move between an in-use position and a non-use position.
It should be noted that the mobile carriage 1 does not necessarily need to be equipped with the work tool 5 .

1 and 2 are schematic diagrams showing one embodiment of an autonomous mobile device of the present invention, with FIG. 1 being a side view and FIG. 2 being a plan view. The autonomous mobile device of this embodiment is a cleaning device that uses a cleaning scraper 5a to scrape and clean deposits within a designated area A while autonomously traveling. FIG. 3 is a configuration diagram showing the control system for each component of this autonomous mobile device.
The autonomous driving device is equipped with a mobile cart 1, which is equipped with a driving device 2 for driving, including wheels 20, a driving source for the wheels 20 (such as a motor), and a steering mechanism for the wheels 20, and this driving device 2 allows the mobile cart 1 to drive (forward or backward) in any direction.

The mobile carriage 1 is equipped with a range measurement device 3, an arithmetic and control device 4, a movement amount detection device 7, a storage device 8, etc. Furthermore, the mobile carriage 1 is equipped with a cleaning scraper 5a (work tool 5) that scrapes and cleans deposits on the road surface (ground), and an actuator 6 that drives (mainly moves up and down) the cleaning scraper 5a, and the cleaning scraper 5a is held by the actuator 6 so that it can move up and down.
This autonomous driving device (hereinafter sometimes referred to as the "cleaning device") cleans the road surface by using the cleaning scraper 5a to scrape up deposits (accumulated powder) on the road surface as it travels along the road surface, and therefore does not have a deposit suction mechanism or a mechanism for storing the sucked up deposits.

The range measurement device 3 is a device that measures the distance between the mobile cart 1 and surrounding obstacles (for example, structures, structural parts of buildings, equipment, and other objects) or captures images of the surroundings to obtain two-dimensional or three-dimensional shape and position information of the obstacles as obstacle information a. Any type of range measurement device 3 can be used, and for example, one or more types such as a laser range finder, an infrared sensor, an ultrasonic sensor, a radar sensor using radio waves, a depth camera, and a stereo camera can be used, but from the viewpoint of measurement accuracy, a 2D or 3D laser type laser range finder is preferable.
The movement amount detection device 7 is configured by, for example, a wheel encoder, and detects (estimates) the movement amount of the mobile carriage 1 from the movement amount of the wheels 20 that constitute the drive device 2 .

The storage device 8 pre-stores obstacle position information a1 within the designated area A, designated area information a2 (in this embodiment, this is information about the cleaning area, so hereinafter referred to as "cleaning area information a2"), and final target point information a3 (in this embodiment, this is information about the accumulation location of deposits, so hereinafter referred to as "deposit accumulation location information a3").
In this embodiment, the obstacle position information a1 is information on the planar position coordinates of obstacles (structures, etc.) present within the designated area A, the cleaning area information a2 is information on the planar position coordinates of the outer edge (periphery) of the designated area A, which is the cleaning area, and the deposit accumulation location information a3 is information on the planar position coordinates of the deposit accumulation position (location), which is the final target position of movement of the mobile cart 1. The obstacle position information a1, cleaning area information a2, and deposit accumulation location information a3 may be information given in advance on a map (obstacle map) or the like, or may be information measured and stored in advance by the range measurement device 3, or information measured and stored in advance using a device other than the range measurement device 3, etc.

Here, an example of a method for generating obstacle position information a1 using a device other than the range measurement device 3 is a method using a 3D shape measurement device 10, such as a 3D scanner, as shown in FIG. 10. Specifically, the 3D shape of the specified area A is scanned using a 3D shape measurement device 10 separate from the mobile platform 1, a 3D CAD model is created from the scan data, and obstacle position information a1 is generated from the cross-sectional shape at a specific height or the results of a simulation using the 3D CAD model. In this way, generating obstacle position information a1 using the 3D shape measurement device 10 reduces measurement errors compared to generating obstacle position information a1 using the range measurement device 3 of the mobile platform 1. Furthermore, compared to using a pre-defined map (obstacle map), information such as deformation of obstacles due to added equipment or aging can be reflected, making it possible to generate accurate obstacle position information a1.

The arithmetic and control device 4 causes the mobile vehicle 1 to travel autonomously along the calculated travel route, visiting waypoints p in order and arriving at the destination point (final waypoint p). At this time, the self-position of the mobile vehicle 1 is estimated by comparing obstacle information a acquired by measuring the surrounding environment using the range measurement device 3 with obstacle position information a1 within the designated area A that is pre-stored (stored) in the storage device 8. This self-position estimation method preferably uses a SLAM algorithm using a Kalman filter, extended Kalman filter, particle filter, Bayes filter, etc., but it is particularly preferable to estimate the self-position of the mobile vehicle using a particle filter based on the SLAM algorithm.
Furthermore, when estimating the self-position using the SLAM algorithm, it is possible to reduce the calculation load and estimate the self-position of the mobile carriage 1 with high accuracy by combining odometry from a movement amount detection device 7 such as a wheel encoder in addition to the obstacle position information a1 and the obstacle information a acquired by measurement using the range measurement device 3. Furthermore, instead of odometry from the movement amount detection device 7, it is also possible to combine information such as GPS positioning and IMU output, which can be expected to produce similar effects.

The arithmetic and control device 4 generates waypoint information b (waypoint information for autonomous movement) from the obstacle position information a1, cleaning area information a2, and deposit accumulation location information a3 within the above-mentioned designated area A. Specifically, in order to efficiently clean the entire cleaning area and accumulate the deposits in a predetermined location, the arithmetic and control device 4 determines the positions of the waypoints that the mobile carriage 1 should follow, the attitude (attitude angle) of the mobile carriage 1 at each waypoint, the operating mode of the actuator 6 that drives the cleaning scraper 5a, etc. based on the above-mentioned information a1 to a3, and generates the waypoint information b.
Here, the waypoint information b includes information such as the carriage position (x coordinate, y coordinate) of the mobile carriage 1 at each waypoint p, the carriage attitude (carriage attitude angle θ) of the mobile carriage 1, and an operation command m for the actuator 6. By giving an actuator operation command m that specifies the carriage position (x coordinate, y coordinate), carriage attitude (carriage attitude angle θ), and actuator operation (raising and lowering operation) that the mobile carriage 1 should take at each waypoint p, the mobile carriage 1 can be moved to a desired position, and the sediment in a desired area can be scraped up and accumulated at the desired position.
The waypoint information b is data in which the x-coordinate of the carriage position, the y-coordinate of the carriage position, the carriage attitude angle θ, and the actuator operation command m are listed in j order of waypoints _p1 to _pj that the mobile carriage 1 circulates, and when the ith waypoint p is reached, autonomous driving is repeated with the waypoint in the (i+1)th waypoint information b as the next waypoint p, and autonomous driving continues until the final jth waypoint p is reached.

The arithmetic and control device 4 calculates a travel route to way point p _i (i-th way point p) along which the mobile carriage 1 will move (generates a travel route), based on the estimated self-position and the generated way point information b, and controls the drive unit 2 so that the mobile carriage 1 moves to way point p _i along the travel route. The calculation of the travel route and the control of the drive unit 2 based on the calculated travel route are repeated for each way point p that the mobile carriage 1 travels through, thereby causing the mobile carriage 1 to autonomously travel to the destination point (final way point p). That is, when the mobile carriage 1 travels autonomously to the final way point p by sequentially visiting predetermined way points p, as the mobile carriage 1 moves toward the i-th way point p, the way point information b is read and set as the next way point, a travel route is generated (calculated), and the mobile carriage 1 performs autonomous travel. When the mobile carriage 1 reaches the i-th way point p, the way point information b for the (i+1)-th way point p is read and set as the next way point, and the mobile carriage 1 performs autonomous travel. This process is repeated until the final way point p is reached. When the mobile vehicle 1 autonomously travels toward the i-th waypoint p, it normally estimates its own position at regular time intervals until it reaches the target point (i-th waypoint p), generates a travel route to the target point based on the estimated own position, and travels along that travel route. This procedure is repeated while traveling toward the target point, and when the distance between the successively estimated own position and the target point becomes equal to or less than a certain value, it is determined that the target point (i-th waypoint p) has been reached.
In addition, the arithmetic and control device 4 controls the actuator 6 based on the actuator operation command m in the i-th way point information b "while the mobile cart 1 is moving toward the i-th way point p" and/or "after arriving at the i-th way point p", and drives the cleaning scraper 5a so that the cleaning scraper 5a moves in the vertical direction and comes into contact with or moves away from the road surface.

Furthermore, to prevent the mobile carriage 1 from getting stuck during autonomous travel, the arithmetic and control device 4 determines whether the mobile carriage 1 is stuck while moving toward way point p _i and cannot reach way point p _i . If it determines that the mobile carriage 1 is stuck, it changes and sets way point p to which the mobile carriage 1 should next move, and controls the drive device 2 to move the mobile carriage 1 to the new way point p. When changing way point p, the new way point p is determined in advance as the "nth way point in the future." If it is determined that the mobile carriage 1 is stuck while moving toward way point p i , the (i+n)th (n is a predetermined natural number) way point information b is changed and set to the next way point p, thereby preventing the mobile carriage 1 from getting stuck. Typically, n is set to "1," but this is not limited to this.
As explained above, examples of cases in which the mobile cart 1 becomes stuck during autonomous travel include, but are not limited to, cases in which the cleaning scraper 5a is unable to push through a large amount of accumulated material, causing the mobile cart 1 to become stuck.

Whether or not the mobile carriage 1 is stuck can be determined, for example, from the travel time to way point p _i (the elapsed time t from the start of movement). In this case, when the mobile carriage 1 moves to way point p _i , the arithmetic and control device 4 calculates the target travel time T (target travel time) to way point p _i from the travel distance ΔL to the way point p _i and the amount of change in attitude angle Δθ (the change in attitude angle θ of the carriage), and determines whether or not the mobile carriage 1 is stuck by comparing the elapsed time t (actual value) from the start of movement of the mobile carriage 1 with the target travel time T. In other words, if the elapsed time t exceeds the target travel time T (t>T), it is determined that the mobile carriage 1 is stuck.
Here, the target travel time T is calculated, for example, when the i-th way point p is reached and the (i+1)-th way point p is set as the next way point, using the distance ΔL between the i-th and (i+1)-th way points p and the amount of change in attitude angle Δθ (change in attitude angle θ of the carriage). The target travel time T is preferably calculated using the following formula (1) using coefficients α, β, and γ determined taking into account the translational speed and rotational speed of the mobile carriage 1, but it may also be derived using other formulas taking into account the translational acceleration/deceleration, rotational acceleration/deceleration, etc. of the mobile carriage 1.
T=αΔL+βΔθ+γ…(1)

Other methods for determining whether a mobile vehicle 1 is stuck include, in the case of an autonomous mobile device that uses a cleaning scraper 5a to scrape up and clean deposits within a designated area A while autonomously traveling as in this embodiment, (i) a load detector such as a load cell is installed between the actuator 6 and the cleaning scraper 5a, and this load detector measures the load on the cleaning scraper 5a (= the pushing force of the cleaning scraper 5a) due to the weight of the deposits, and compares the load detected by this load detector with a threshold value. If the load is equal to or greater than the threshold, it is determined that the mobile vehicle 1 is stuck; (ii) a camera is attached to the mobile vehicle 1 to photograph the deposits that the cleaning scraper 5a is trying to scrape up, and the state of the deposits is observed using image recognition to determine whether the mobile vehicle 1 is stuck; and (iii) a method is used in which the change in the mobile vehicle 1's own position over time is monitored and if it is detected that the mobile vehicle 1 has not moved for a certain period of time, it is determined that the mobile vehicle 1 is stuck.

More preferably, to prevent the mobile carriage 1 from getting stuck when the travel path is blocked by an obstacle not included in the obstacle position information a1 pre-stored in the storage device 8 (e.g., an object placed on the travel path for a certain task), the arithmetic and control device 4 determines whether a travel path to a via point p _i to which the mobile carriage 1 should move can be generated based on the obstacle information a acquired by measurement using the range measurement device 3. If it is determined that a travel path cannot be generated, the arithmetic and control device 4 changes and sets the via point p to which the mobile carriage 1 should move, and controls the drive device 2 to move the mobile carriage 1 to the new via point p. When changing the via point p, the new via point p is predetermined as the "nth via point ahead." When it is determined that a travel path cannot be generated based on the obstacle information a while moving toward or while moving toward the i-th via point p, the arithmetic and control device 4 changes and sets the i+nth (n is a predetermined natural number) via point information b to the next via point p, thereby preventing the mobile carriage 1 from getting stuck. Typically, n is "1," but this is not limited to this.

4 is a block diagram of the arithmetic and control device 4. The arithmetic and control device 4 includes a self-position estimation unit 40, a waypoint information generation unit 41, a route generation unit 42, a control unit 43, a stuck determination unit 44, a route generation determination unit 45, and the like.
When moving the mobile carriage 1 to way point p _i (i-th way point p), the self-position estimation unit 40 takes in the obstacle position information a1 stored in the storage device 8 and the obstacle information a acquired by measuring the surrounding environment using the range measurement device 3, and compares (collates) both pieces of information to estimate the self-position of the mobile carriage 1. Here, when performing self-position estimation using the SLAM algorithm, the amount of movement of the mobile carriage 1 detected by the movement amount detection device 7 (wheel encoder, etc.) is taken in and combined with the odometry by the movement amount detection device 7, thereby performing self-position estimation.
The waypoint information generating unit 41 takes in the obstacle position information a1, the designated area information a2, and the final target point information a3 stored in the storage device 8, and generates waypoint information b based on this information a1 to a3.

The path generation unit 42 receives information on the self-position estimated by the self-position estimation unit 40 and the way point information b generated by the way point information generation unit 41, and generates (calculates) a travel path to way point p _i along which the mobile carriage 1 will move, based on these, and outputs this to the control unit 43. Based on this command, the control unit 43 controls the drive device 2 so that the mobile carriage 1 moves to way point p _i along the travel path. Here, based on the self-position information received from the self-position estimation unit 40, the path generation unit 42 determines that way point p _i has been reached when the distance between the self-position and way point p _i is equal to or less than a certain value.
The path generating unit 42 also outputs an operation command for the actuator 6 based on the way point information b relating to the way point p _i to the control unit 43, and the control unit 43 controls the actuator 6 based on the command.

The stuck determination unit 44 receives the elapsed time t from the control unit 43 since the start of movement, compares this elapsed time t with the target travel time T, and if t > T, determines that the mobile carriage 1 moving toward way point p _i is stuck and unable to reach way point p _i , and outputs this stuck occurrence information to the path generation unit 42. Based on this, the path generation unit 42 changes way point p to which the mobile carriage 1 should be moved (for example, changing it to the (i+1)th way point p), generates (calculates) a travel route to the new way point p, and outputs this to the control unit 43. Based on this command, the control unit 43 controls the drive unit 2 so that the mobile carriage 1 moves to the new way point p along the travel route. The path generation unit 42 also outputs an operation command for the actuator 6 based on way point information b for the new way point p to the control unit 43, and the control unit 43 controls the actuator 6 based on this command.

The path generation determination unit 45 takes in obstacle information a acquired by measuring the surrounding environment using the range measurement device 3, and determines whether a travel path to a way point p _i to which the mobile carriage 1 should move can be generated based on this obstacle information a. If it determines that the travel path cannot be generated, it outputs path generation impossible information to the path generation unit 42. Based on this, the path generation unit 42 changes the way point p to which the mobile carriage 1 should move (for example, changes it to the (i+1)th way point p), generates (calculates) a travel path to the new way point p, and outputs this to the control unit 43. Based on this command, the control unit 43 controls the drive unit 2 so that the mobile carriage 1 moves to the new way point p along the new travel path. The path generation unit 42 also outputs an operation command for the actuator 6 based on way point information b for the new way point p to the control unit 43, and the control unit 43 controls the actuator 6 based on this command.

As shown in Figures 1 and 2, the cleaning scraper 5a, which is the work tool 5, is a member that scrapes up deposits (accumulated powder) on the road surface as the mobile carriage 1 travels with its lower end in contact with the road surface.In this embodiment, it is composed of a plate-shaped (wall-shaped) dozer member that is arranged along the width direction of the front side of the mobile carriage 1 (the front side in the direction of travel during cleaning), and is supported on the mobile carriage 1 via a support arm 9 so that it can move up and down.
The actuator 6 drives the cleaning scraper 5a via the support arm 9 so that the cleaning scraper 5a moves up and down to come into contact with or separate from the road surface. It is also preferable that the cleaning scraper 5a is constantly pressed against the ground by the actuator 6 so that the lower end of the cleaning scraper 5a is in close contact with the ground (cleaning surface) when the scraper is in contact with the road surface. When the target to be cleaned is coal powder, it is not necessary to control the pressing force, but it is also possible to control the pressing force depending on the particle size and density of the deposits.

A sliding guide mechanism (such as a so-called slide guide) is typically used as a mechanism for supporting the cleaning scraper 5a so that it can move up and down. This guide mechanism is composed of, for example, a guide section provided on the movable carriage 1 along the vertical direction, and a slide member provided on the support arm 9 that slidably engages with the guide section. The actuator 6 moves the slide member of the support arm 9 up and down along the guide section, thereby driving the cleaning scraper 5a so that the cleaning scraper 5a moves up and down and comes into contact with or separates from the road surface. The actuator 6 is composed of, for example, a hydraulic cylinder, a ball screw, or the like.
The cleaning scraper 5a is made up of a plate-shaped main body 50 (dozer main body) that runs along the width direction of the mobile carriage 1, and side plates 51 (side dozer parts) that are connected to both ends of the plate-shaped main body 50 in order to enclose (block) the scraped-up deposits and prevent them from escaping to the outside in the width direction of the mobile carriage. In other words, the cleaning scraper 5a is configured in a U-shape in plan view by the plate-shaped main body 50 and the side plates 51 that are connected to both ends of the plate-shaped main body 50.

The main body of the cleaning scraper 5a is usually made of a metal plate-like member, but at least a portion of the lower end (for example, the lower end of the plate-like main body 50) is preferably made of a rubber flat plate or a resin or metal brush to make it easier to scrape up deposits on the road surface. This is particularly effective when scraping up deposits with small particle size and bulk density, such as coal powder (coal powder usually has a particle size of 1 mm or less and a bulk density of 1.0 g/cm3 or ^less ), and allows for high collection efficiency of the deposits. However, depending on the properties of the cleaning surface (unevenness, etc.) and the particle size and density of the deposits, a resin or metal flat plate may also be used.

The cleaning scraper 5a may be supported on the movable carriage 1 via a support arm so that it can rotate freely up and down. In this case, the actuator 6 rotates (drives) the cleaning scraper 5a up and down relative to the movable carriage 1, and may be, for example, a rotary actuator provided on the support part of the support arm, or a hydraulic cylinder connected between the movable carriage 1 and the support arm.
The cleaning device equipped with the above-described cleaning scraper 5a and actuator 6 can reliably and efficiently scrape up deposits on the road surface by traveling with the lower end of the cleaning scraper 5a in close contact with the road surface using the actuator 6, and can move the scraped up deposits to a predetermined location and collect them. In addition, because the cleaning scraper 5a can move up and down, when cleaning is not being performed, the mobile cart 1 can be made to travel with the lower end of the cleaning scraper 5a away from the road surface.

The autonomous driving method of the autonomous driving device is performed according to the flow shown in FIG.
When autonomous travel (cleaning) begins (starts), waypoint information b is generated by the arithmetic and control device 4 from pre-stored obstacle position information a1, cleaning area information a2, and deposit accumulation location information a3, and this waypoint information b is stored in the storage device 8. The mobile vehicle 1 starts from its initial position, and in step S1, the waypoint number i to which it will head is set to 1 (waypoint p _i = p ₁ ). Then, in step S2, waypoint information b for waypoint number 1 (waypoint p ₁ ) is read, and in step S3, the target travel time T to reach waypoint number 1 is calculated. Furthermore, in step S4, the mobile vehicle 1's own position is estimated, and in step S5, a travel route is generated.

If a driving route can be generated (step S5 is successful), autonomous driving begins toward waypoint number 1 (waypoint _p1 ) in step S6. On the other hand, if a driving route cannot be generated (step S5 is unsuccessful), in step S7 the current waypoint number i is changed to i+n (waypoint number 1 is changed to 1+n), and in step S8 it is determined whether the changed waypoint number i is greater than the number of waypoints j. Here, an example of a case where a driving route cannot be generated is when an unexpected obstacle (such as an intruder) that is not included in the obstacle position information a1 makes it impossible to generate a route. If the changed waypoint number i is equal to or less than the number of waypoints j (step S8 is No), the process returns to step S2, and the waypoint information b for the changed waypoint number i is read. On the other hand, if the waypoint number i is greater than the number of waypoints j (step S8 is Yes), an error is determined in step S9, an error message is displayed, and the flow ends.

During autonomous travel towards waypoint number 1 (waypoint _p1 ) in step S6, the elapsed time t from the start of travel is measured, and in step S10 it is determined whether the elapsed time t has exceeded the target travel time T. If the elapsed time t has not exceeded the target travel time T (No in step S10), it is determined in step S11 whether the mobile carriage 1 has reached the target point (waypoint number 1). On the other hand, if the elapsed time t has exceeded the target travel time T (Yes in step S10), the process proceeds to step S7, and the same flow as above is followed thereafter.
Here, in step S11, whether the target point (way point number 1) has been reached is determined based on the estimation of the vehicle's own position. That is, until the target point is reached, the vehicle estimates its own position at regular time intervals, generates a travel route to the target point based on the estimated own position, and travels along that travel route. This procedure is repeated while traveling toward the target point, and it is determined that the vehicle has reached the target point when the distance between the successively estimated own position and the target point (way point number 1) becomes equal to or less than a certain value.

If it is determined in step S11 that the mobile carriage 1 has reached the target point (way point number 1) (step S11: Yes), the next way point is set (i = i + 1) in step S12, and it is determined in step S13 whether the set way point number i is equal to or greater than the number of way points j (i ≥ j). If way point number i is less than the number of way points j (step S13: No), the process returns to step S2 and repeats the flow from step S2 onwards. On the other hand, if it is determined in step S13 that way point number i is equal to or greater than the number of way points j (step S13: Yes), the process ends. Also, if it is determined in step S11 that the target point (way point number 1) has not been reached (step S11: No), the process returns to step S2 and repeats the flow from step S2 onwards.

Next, an example of a method for setting a cleaning route and generating the waypoint information b will be described with reference to FIGS.
The waypoint information b is generated from obstacle position information a1, cleaning area information a2, and deposit accumulation location information a3 within the designated area A. In the designated area A (cleaning area) shown typically in Figures 6 and 7, obstacle position information a1 is the position information of obstacles within the entire map area, cleaning area information a2 is the coordinates of the cleaning area (a ≤ x ≤ b, c ≤ y ≤ d), and accumulation location information a3 is the coordinates ( _xc , _yc ) of the final deposit accumulation location.

To generate the via point information b, first, as shown in Figure 6, a cleaning route is set up along the x-axis of the cleaning area (a ≤ x ≤ b, c ≤ y ≤ d), scraping up the piles from the left end (x = a) to the accumulation position (x = _xc ) in the order of "start point (a, ds) → end point ( _xc , ds)" and temporarily depositing and accumulating them at the accumulation position (x = _xc ). N such cleaning routes are set up for each scraper width of 2s, up to the "start point (a, d-(2n-1)s) → end point ( _xc , d-(2n-1)s)" route within the cleaning area, and finally, a cleaning route from "start point (a, c+s) → end point ( _xc , c+s)" is set to clean the bottom edge. In other words, n+1 cleaning routes are set up, and (n+1) x 2 = 2n+2 pieces of via point information b (information on the via points that are the start and end points of the cleaning route) are generated.
Furthermore, as shown in Figure 7, a cleaning route is set up along the x-axis direction of the cleaning area (a≦x≦b, c≦y≦d) to collect the deposits from the right end (x= _b ) to the accumulation position (x=xc) in the order of "start point (b, ds) → end point ( _xc , ds)" and temporarily store and accumulate them at the accumulation position (x= _xc ).N+1 such cleaning routes are set up in the same way as above, and (n+1)×2=2n+2 pieces of via point information b (information on the via points that are the start and end points of the cleaning route) are generated accordingly.

Furthermore, as shown in Figure 8, a cleaning route is set up to collect the deposits (deposits temporarily placed and accumulated as described above) from the top end (y=d) to the accumulation position (y= _yc ) along the y-axis direction of the cleaning area (a≦x _≦b , c≦y _≦ d) from the start point ( _xc , d) to the end point ( _xc , yc) and accumulate them at the final accumulation position (xc, _yc ), and a cleaning route is set up to collect the deposits (deposits temporarily placed and accumulated as described above) from the bottom end (y=c) to the accumulation position (y= _yc ) along the y-axis direction of the cleaning area (a≦x _≦b , c≦y _≦d ) from the start point ( _xc , c) to the end point (xc, yc) and accumulate them at the final accumulation position ( _xc , _yc ), and four pieces of via point information b (information on the via points that are the start and end points of the cleaning route) are generated accordingly. Finally, by setting the initial position (0, 0) as the waypoint (final point), j=(2n+2)×2+4+1=4n+9 pieces of waypoint information b are generated as shown in FIG.

Next, the autonomous driving method of the present invention will be described.
This autonomous traveling method is a method for autonomously traveling the mobile carriage 1 using the device configuration described above. Therefore, this is an autonomous traveling method in which a mobile carriage 1 equipped with a driving device 2 for traveling autonomously travels while passing through a plurality of way points p within a designated area A, and includes the steps of: (a) generating way point information b from pre-stored obstacle position information a1 within the designated area A, designated area information a2, and final target point information a3; (b) estimating the self-position of the mobile carriage 1; (c) calculating a traveling route to way point p _i (i-th way point p) along which the mobile carriage 1 should move based on the estimated self-position and the generated way point information b, and controlling the driving device 2 so that the mobile carriage 1 moves to way point p _i along that traveling route; and (d) determining whether the mobile carriage 1 is stuck while moving toward way point p _i and cannot reach way point p _i , and if it is determined that the mobile carriage 1 is stuck, changing the way point p to which the mobile carriage 1 should be moved, and controlling the driving device 2 to move the mobile carriage 1 to the new way point p.

Furthermore, this autonomous driving method can take on more specific embodiments such as those shown in (i) to (iv) below.
(i) In step (a), the range measurement device 3 measures the surrounding environment of the mobile cart 1 to obtain obstacle information a, and the obstacle position information a1 within the designated area A stored in advance is compared with the obstacle information a obtained by the range measurement device 3 to estimate the self-position of the mobile cart 1.
(ii) Further, the method includes a step (e) of determining whether or not a travel route to a via point p _i to which the mobile carriage 1 should move can be generated based on the obstacle information a acquired by the range measurement device 3, and if it is determined that the route cannot be generated, changing the via point p to which the mobile carriage 1 should move, and controlling the driving device 2 to move the mobile carriage 1 to the new via point p.

(iii) The mobile cart 1 is equipped with a work tool 5 for performing specified work within the designated area A and an actuator 6 for driving the work tool 5, and work is performed using the work tool 5 while the mobile cart 1 is traveling within the designated area A.
(iv) In the case of (iii) above, the work tool 5 is a cleaning scraper 5a that scrapes up and cleans deposits on the ground within the designated area A, and the actuator 6 drives the cleaning scraper 5a so that the cleaning scraper 5a moves up and down to come into contact with or move away from the road surface. The mobile cart 1 is made to travel within the designated area A, and the cleaning scraper 5a is driven by the actuator 6, while the cleaning scraper 5a scrapes up deposits, and the scraped up deposits are moved to a predetermined location and accumulated.

The details of the autonomous driving method of the present invention and the device used to implement it are the same as those described above for the autonomous driving device, and a detailed description thereof will be omitted.
The autonomous mobile device and method of the present invention are suitable for use as a cleaning device for the top of a coke oven where coal dust (fallen dust) accumulates, and can efficiently and appropriately carry out cleaning and removal work of the deposits.

In order to evaluate the autonomous traveling capability and cleaning ability of the autonomous traveling device (cleaning device) of the present invention equipped with the cleaning scraper 5a and control system as shown in Figures 1 to 4, the following cleaning test was conducted inside a building. In this test, coal powder to be cleaned was scattered in the cleaning area, and while the cleaning device was traveling autonomously, the coal powder was scraped up by the cleaning scraper 5a and collected in a chute (accumulation position) which is a coal powder recovery location, from which it was dropped.
The cleaning scraper 5a provided in the cleaning device is 0.4 m wide and has a flat brush at the bottom.
11 to 18 show the interior of the building where the test was conducted using a coordinate system, and show the cleaning area (the area where coal powder is scattered), the initial position of the cleaning device (the starting point and end point of the movement of the cleaning device), the chute (the final location where the coal powder is accumulated and collected), and part of the surrounding walls.

[Example 1]
In a rectangular cleaning area (0≦x≦5, 1≦y≦4) inside a building as shown in Figure 11, coal powder to be cleaned was spread at a rate of approximately 1 kg/ ^m² (spreading amount: approximately 15 kg) to a nearly uniform thickness. The initial position of the cleaning device was set so that the center position of the cleaning scraper 5a was the origin (0,0). A cleaning path was set as shown schematically in Figures 6 to 8, and the coal powder was collected and dropped into a chute installed at coordinates (2,2). The waypoint information b generated during this process is shown in Table 1. The waypoint information b shown in Table 1 includes, from the left, the x-coordinate and y-coordinate of the carriage position, the carriage attitude angle θ (indicating the direction the carriage is facing when it arrives at the waypoint), and the actuator operation command m (indicating how the cleaning scraper 5a is to be operated on the way to the waypoint). The cleaning device reads the waypoints in order from the first row and performs autonomous navigation. The actuator operation command m is "1" to lower the cleaning scraper 5a onto the road surface and press it against it, and "0" to lift the cleaning scraper 5a up and away from the road surface.

Unless the cleaning device "becomes stuck during movement" or "cannot generate a travel route," it moves to way point _p37 (end point) via way points _p1 to _p36 in order based on way point information b in Table 1, and performs operations based on way point information b (turning of mobile carriage 1 based on carriage attitude angle θ, operation of actuator 6 based on actuator operation command m). On the other hand, if it is determined that "it will become stuck" or "cannot generate a travel route" when moving to way point _p1 (i-th way point p), the way point to move to is changed to way point p1 ₊₁ (i+1-th way point p), and autonomous travel is performed by reading the way point information.

The movement trajectory of the cleaning device is shown in Figure 12. The movement trajectory in Figure 12 shows a portion of the waypoint p. By passing through waypoints _p1 to _p36 , the cleaning device scraped up coal powder within the cleaning area with the cleaning scraper 5a, collected it in the chute located at coordinates (2, 2), and dropped it there before returning to the initial position (the end point of the cleaning device's movement). The amount of coal powder collected by the chute (collected amount) was 12.4 kg, and the collection rate, calculated as (collected amount) / (spread amount) × 100, was 82.7%. This confirmed that the coal powder was properly removed and that the cleaning device had excellent autonomous navigation and cleaning capabilities.

[Example 2]
In an area inside a building as shown in Fig. 13, an obstacle that does not exist in the original obstacle map (an unexpected obstacle that was not included in the obstacle position information a1) was placed at the position of coordinates (5, 3) to physically block passage, and then, as in Example 1, coal powder to be cleaned was spread in the cleaning area to a substantially uniform thickness at about 1 kg/ ^m2 (spreading amount: about 15 kg), and a test similar to that in Example 1 was performed. Here, the arithmetic and control device 4 is configured to determine whether or not a travel route to the ith way point p along which the cleaning device should move can be generated based on the obstacle information a acquired by the range measurement device 3, and if it is determined that the route cannot be generated, change the way point along which the cleaning device should move to the (i+1)th way point p.

The movement trajectory of the cleaning device is shown in Figure 14. The movement trajectory in Figure 14 shows a portion of the way point p. In this example, when the cleaning device was moving from the 20th way point _p20 at coordinates (2,3.4) to the 21st way point _p21 at coordinates (5,3), it was determined that a path to reach way point _p21 could not be generated. Therefore, the way point to which the cleaning device should move was changed to way point _p22 . The cleaning device then moved from the 20th way point _p20 to the 22nd way point _p22 at coordinates (2,3), and then passed through the 23rd and subsequent way points p to return to the initial position (the end point of the cleaning device's movement). The amount of coal powder collected by the chute (collected amount) was 11.5 kg, and the collection rate, calculated by dividing the collected amount by the applied amount x 100, was 76.7%. As described above, although the amount of collected coal was reduced by one pass, the coal dust was properly cleaned to the extent possible, and even when the movement path was blocked by an obstacle, the cleaning was completed without getting stuck. This confirmed that the coal dust was properly cleaned and that the cleaning device has excellent autonomous mobility and cleaning capabilities.
For comparison, the movement trajectory of a cleaning device traveling autonomously using conventional SLAM technology is shown in Figure 15. In this case, the 21st waypoint was completely blocked by an obstacle, making it impossible to continue traveling, and the cleaning device became stuck.

[Example 3]
In an area inside a building as shown in FIG. 16 , coal powder to be cleaned was spread in the cleaning area at a substantially uniform thickness of approximately 1 kg/ ^m² (spreading amount: approximately 15 kg), as in Example 1. An additional 10 kg of coal powder was spread on area G, and a test similar to that in Example 1 was conducted. Here, when the cleaning device moves to the i-th waypoint p, the arithmetic and control device 4 calculates a target travel time T to the waypoint p, compares the actual elapsed time t from the start of movement of the cleaning device with the target travel time T, and determines whether the cleaning device is stuck. If it determines that the cleaning device is stuck, the arithmetic and control device 4 changes the waypoint to which the cleaning device should be moved to the i+1-th waypoint p. The target travel time T was calculated using the above formula (1), and the coefficients α, β, and γ were set to 0.4, 0.4, and 10, respectively.

The movement trajectory of the cleaning device is shown in Figure 17. The movement trajectory in Figure 17 shows a portion of the waypoint p. In this example, the cleaning device became stuck en route from the 21st waypoint _p21 at coordinates (5,3) to the 22nd waypoint _p22 at coordinates (2,3). Since the elapsed time t from the waypoint _p21 exceeded the target travel time T, it was determined to be stuck, and the waypoint to which the cleaning device should be moved was changed to waypoint _p23 . As a result, the cleaning device passed waypoint _p22 and headed for the 23rd waypoint _p23 at coordinates (5,2.6), and then passed through the 24th and subsequent waypoints p to return to the initial position (the end point of the cleaning device's movement). The amount of coal powder collected by the chute (collected amount) was 11.6 kg, and the collection rate, calculated by dividing the collected amount by the applied amount x 100, was 46.4%. As described above, although the recovery rate decreased because the additional coal could not be recovered, the coal powder was properly cleaned to the extent possible, and even when the vehicle became stuck, it was able to complete the cleaning without getting stuck. This confirmed that the coal powder was properly cleaned and that the cleaning device has excellent autonomous mobility and cleaning capabilities.
For comparison, the movement trajectory of a cleaning device traveling autonomously using general SLAM technology is shown in Figure 18. In this case, the cleaning device was unable to push through the coal powder midway from the 21st waypoint to the 22nd waypoint, coordinates (2, 3), and was unable to continue traveling, becoming stuck.

REFERENCE SIGNS LIST 1 Mobile cart 2 Drive device 3 Range measurement device 4 Arithmetic and control device 5 Work tool 5a Cleaning scraper 6 Actuator 7 Movement amount detection device 8 Storage device 9 Support arm 10 3D shape measurement device 20 Wheel 40 Self-position estimation unit 41 Route point information generation unit 42 Route generation unit 43 Control unit 44 Stack determination unit 45 Route generation determination unit 50 Plate-shaped main body unit 51 Side plate unit A Designated area G Area p, p _i via point

Claims (11)

An autonomous driving device that autonomously drives a mobile carriage (1) equipped with a driving device (2) for driving within a designated area (A) while passing through a plurality of via points (p),
a calculation and control device (4) that estimates the self-position of the mobile carriage (1), generates waypoint information (b) from obstacle position information a1 within a designated area (A) that is stored in advance, designated area information a2, and final target point information a3, calculates a travel route to a waypoint (p _i ) along which the mobile carriage (1) should move based on the estimated self-position and the generated waypoint information (b), and controls a drive device (2) so that the mobile carriage (1) moves to the waypoint (p _i ) according to the travel route;
The arithmetic and control device (4) determines whether the mobile carriage (1) moving toward the way point (p _i ) is stuck and cannot reach the way point (p _i ), and if it determines that the mobile carriage (1) is stuck, changes the way point (p) to which the mobile carriage (1) should be moved to a way point (p) that has been determined in advance as a destination, and controls the drive device (2) to move the mobile carriage (1) to the way point (p) that has been changed to ;
The mobile carriage (1) is provided with a work tool (5) for performing a predetermined task within a designated area (A) and an actuator (6) for driving the work tool (5);
The waypoint information b includes information on the carriage position of the mobile carriage (1), the carriage attitude of the mobile carriage (1), and operation commands for the actuator (6) . The vehicle further includes a range measurement device (3) that measures the surrounding environment of the mobile vehicle (1) and acquires obstacle information a,
The autonomous driving device according to claim 1, characterized in that the arithmetic and control device (4) estimates the self-position of the mobile carriage (1) by comparing obstacle position information a1 within the designated area (A) stored in advance with obstacle information a acquired by the range measurement device (3). The autonomous driving device according to claim 2, characterized in that the arithmetic and control device (4) determines whether or not a driving route to a via point (p _i ) to which the mobile carriage (1) should move can be generated based on the obstacle information a acquired by the range measurement device (3), and if it determines that the route cannot be generated, changes the via point (p) to which the mobile carriage (1) should move to a via point (p) that has been determined in advance as a destination, and controls the drive device (2) to move the mobile carriage (1) to the destination via point (p). The work tool (5) is a cleaning scraper (5a) that scrapes and cleans deposits in the designated area (A),
An autonomous mobile device as described in any one of claims 1 to 3, characterized in that the actuator (6) drives the cleaning scraper (5a) so that the cleaning scraper (5a) moves in an up and down direction to come into contact with or move away from the road surface. The autonomous driving device according to any one of claims 1 to 3, characterized in that when the mobile carriage (1) moves to the waypoint (p _i ), the arithmetic and control device (4) calculates a target travel time T to the waypoint (p _i ) from the travel distance ΔL to the waypoint (p _i ) and the attitude angle change amount Δθ, and determines whether the mobile carriage (1) is stuck by comparing an actual value t of the elapsed time from the start of movement of the mobile carriage (1) with the target travel time T. The system further includes a movement amount detection device (7) that detects the movement amount of the movable carriage (1) from the operation amount of the drive device (2),
The autonomous driving device according to claim 2 or 3, characterized in that the arithmetic and control device (4) compares the obstacle position information a1 within the designated area (A) stored in advance with the obstacle information a acquired by the range measurement device (3), and also estimates the self-position of the mobile cart (1) by a particle filter based on a SLAM algorithm using the amount of movement of the mobile cart (1) detected by the movement amount detection device (7). An autonomous traveling method for autonomously traveling a mobile carriage (1) equipped with a driving device (2) for traveling within a designated area (A) while passing through a plurality of via points (p), comprising:
A step (a) of generating waypoint information (b) from obstacle position information (a1) within a designated area (A), designated area information (a2), and final target point information (a3) stored in advance;
A step (a) of estimating the self-position of the mobile cart (1);
a step (c) of calculating a travel route to a via point (p _i ) along which the mobile carriage (1) will move based on the estimated self-position and the generated via point information b, and controlling a drive device (2) so that the mobile carriage (1) moves to the via point (p _i ) along the travel route;
a step (d) of determining whether a mobile carriage (1) moving toward a waypoint (p _i ) is stuck and unable to reach the waypoint (p _i ), and if it is determined that the mobile carriage (1) is stuck, changing the waypoint (p) to which the mobile carriage (1) should be moved to a waypoint (p) that has been determined in advance as a destination, and controlling a drive device (2) to move the mobile carriage (1) to the waypoint (p) that has been changed to ;
The mobile carriage (1) is provided with a work tool (5) for performing a predetermined task within a designated area (A) and an actuator (6) for driving the work tool (5);
The method for autonomous driving is characterized in that the waypoint information b includes information on the carriage position of the mobile carriage (1), the carriage attitude of the mobile carriage (1), and operation commands for the actuator (6) . 8. The autonomous driving method according to claim 7, wherein in step (i), the surrounding environment of the mobile carriage (1) is measured by a range measurement device (3) to acquire obstacle information a, and the self-position of the mobile carriage (1) is estimated by comparing obstacle position information a1 within the designated area (A) stored in advance with the obstacle information a acquired by the range measurement device ( 3 ). The autonomous driving method according to claim 8, further comprising a step (e) of determining whether or not a driving route to a via point (p _i ) to which the mobile carriage (1) should move can be generated based on the obstacle information a acquired by the range measurement device (3), and if it is determined that the route cannot be generated, changing the via point (p) to which the mobile carriage (1) should move to a via point (p) determined in advance as a destination, and controlling the drive device (2) to move the mobile carriage ( 1 ) to the destination via point (p). 10. The autonomous traveling method according to claim 7 , wherein work is performed with a work tool (5) while the mobile carriage (1) is traveling within a designated area (A). The work tool (5) is a cleaning scraper (5a) that scrapes and cleans deposits on the ground within the designated area (A),
The actuator (6) drives the cleaning scraper (5a) so that the cleaning scraper (5a) moves up and down to come into contact with or separate from the road surface;
The autonomous driving method according to claim 10, characterized in that the mobile cart (1) is caused to travel within the designated area ( A ), the cleaning scraper (5 a) is driven by the actuator (6), the cleaning scraper (5 a) scrapes up the deposits, and the scraped up deposits are moved to a predetermined location and accumulated therein.