KR102384102B1 - Autonomous robot and method for driving using the same - Google Patents

Abstract

In the first driving mode, the vehicle travels along a wall created by approximating the surrounding structures while measuring the distances to the surrounding structures by angles through the lidar. Check whether there is a discontinuity point based on the distances to the surrounding structures, and if there is a discontinuity point, change to the second driving mode. The vehicle is driven in the second driving mode until is 0. Here, the rotation angle is an angle at which the robot is directed to an intermediate point between the coordinates of the maximum right distance and the coordinates of the maximum left distance measured based on the driving direction.

Description

Autonomous robot and method for driving using the same

The present invention relates to an autonomous driving robot and a driving method of the autonomous driving robot for easily driving avoiding obstacles installed at narrow intervals in an environment in which curved driving paths are frequently generated, such as a restaurant.

With the recent growth of the untact-related market, cases in which autonomous driving robots provide hospitality services on behalf of employees in various places where customer service is required (eg, hotels or restaurants) are increasing. Hotels and restaurants are places where it is difficult for an autonomous driving robot to drive because the characteristics that make up the space, such as the movement path of the autonomous driving robot or the congestion level of users, are very diverse.

In particular, since tables and chairs are installed at a narrow distance in a restaurant, when the table space is minimized, the autonomous driving robot frequently collides with a table or chair while driving in a curved direction. Accordingly, the driving difficulty of the autonomous driving robot in restaurants increases, and the commercialization of the autonomous driving robot is slow.

In the past, in order to plan a curved driving path for an autonomous driving robot, the path was set in a right-angled triangle method. The right-angled triangle method is a technology that sets the path so that the height side of the right-angled triangle and the approximate wall coincide with the lidar installed in the autonomous driving robot.

However, if an error occurs in the path of the autonomous driving robot when using the right triangle method, an area error occurs between the height side, which is a reference for path setting, and the approximate wall. Accordingly, in an environment where the distance between the obstacles is narrow or a corner driving is required in which the wall surface changes rapidly, the autonomous driving robot mostly collides with an obstacle or a wall surface or stops running.

Accordingly, the present invention provides an autonomous driving robot and a driving method of the autonomous driving robot in which obstacles are installed at narrow intervals by checking the existence of a corner section in advance before entering the corner section.

As a driving method of a robot, which is a feature of the present invention for achieving the technical problem of the present invention,

In the first driving mode, while measuring the distances to the surrounding structures by angle through the lidar, driving along the wall generated by approximating the surrounding structures, checking whether there is a discontinuous point based on the distances to the surrounding structures and changing to the second driving mode if there is the discontinuous point, and while rotating at a rotation angle calculated based on the distance for each angle collected through the lidar, the second driving mode until the rotation angle becomes 0 Comprising the step of driving in 2 driving mode, wherein the rotation angle is, the robot is directed so that the robot is directed at the coordinates of the midpoint between the coordinates of the maximum right distance and the coordinates of the maximum left distance measured based on the driving direction. angle to control.

The driving along the wall may include approximating the coordinates of each point where the laser output from the lidar to the surrounding structure is reflected for each angle with one straight line to generate the wall surface.

The step of driving along the wall comprises calculating an area error between a side from the ground to the height of the lidar and the wall, and maintaining the distance and angle with the wall so that the area error converges to zero, It may include the step of traveling along the wall surface.

The step of determining whether there is a discontinuous point includes checking the right maximum distance among the right distances for each angle measured from the right side based on the driving direction, and if there is a short right distance below a threshold value compared to the right maximum distance determining that there is a corner section on the right side of the driving direction, and checking the left maximum distance among the left distances for each angle measured from the left side based on the driving direction, and the left side shorter than a threshold value compared to the left maximum distance If the distance exists, determining that there is a corner section on the left side of the driving direction.

The step of driving in the second driving mode may include: confirming the right angle at which the right maximum distance and the right maximum distance are measured as coordinates of the right maximum distance; Checking the angle as the coordinates of the left maximum distance, and calculating the rotation angle to pass the coordinates of the intermediate point, which is the bisector of the side generated between the coordinates of the right maximum distance and the left maximum distance coordinates and the rotation angle may be updated based on the distance for each angle collected through the lidar until passing the discontinuous point.

After the step of driving in the second driving mode, if the rotational driving passes the discontinuous point while rotating so that the rotation angle converges to 0, the driving mode is changed to approximate the surrounding structures at the position passing the discontinuous point. may include running along the wall.

As another feature of the present invention for achieving the technical problem of the present invention, the driving method of the robot,

Step of driving while measuring the distance to surrounding structures, check the longest right distance among the right distances for each angle measured on the right side of the driving direction, and if there is a short right distance below the threshold value compared to the longest right distance , determining that there is a corner section on the right side of the driving direction, checking the longest left distance among the left distances for each angle measured in the left side of the driving direction, and the shortest left distance below the threshold compared to the longest left distance If present, determining that there is a corner section on the left side of the driving direction, and passing the entered corner section while changing the driving angle in the direction of the right corner section or the left corner section.

Determining that there is a corner section on the right side may include determining the location of the structure in which the longest right distance is measured based on the angle at which the longest right distance and the longest right distance are measured. there is.

The step of determining that there is a corner section on the left side may include determining the location of the structure in which the longest left distance is measured, based on the angle at which the longest left distance and the longest left distance are measured. there is.

In the step of passing the corner section, the rotation angle is calculated based on the traveling direction of the robot based on the angle of the structure in which the right distance is measured, the angle of the structure in which the left distance is measured, and the position of the robot. may include steps.

The rotation angle may be an angle at which the robot is rotated so as to bisect the position of the structure in which the right distance is measured and the position of the structure in which the left distance is measured.

The step of passing the corner section may further include calculating the angle to the driving target position based on the position of the structure in which the longest right distance is measured and the position of the structure in which the longest left distance is measured. can

In the step of passing the corner section, if there is no short right distance or short left distance less than the threshold value, obtaining a plurality of coordinates of the structures for each angle based on the distance to the structures, the plurality of coordinates The method may include generating an approximate wall surface by approximating them to a straight line, and generating and driving a driving path so as to run parallel to the approximate wall surface at a location separated from the approximate wall surface by a predetermined distance.

Before the driving step, generating an initial map while driving in a space where the structures are installed, and transmitting the initial map to a server interworking with the robot, and a driving map in which update information is reflected in the initial map from the server and receiving, wherein the update information may include the size of the robot, the size of the structure, the shape of the structure, and boundary line processing information in which a plurality of structures are grouped into one structure.

Receiving the driving map may further include generating a wall by approximating the surrounding structures, and driving along the wall while measuring distances to the surrounding structures by angles through lidar. there is.

As another feature of the present invention for achieving the technical problem of the present invention, a robot that changes a driving route during autonomous driving,

A sensor and a processor for measuring distances to surrounding structures by angle, wherein the processor sets a driving mode to travel along an approximate wall generated based on the surrounding structures, and is discontinuous based on the distances to the surrounding structures. Check the point, and when the discontinuous point is confirmed, the rotation angle to rotate the robot is calculated based on the right reference distance and the right angle, and the left reference distance and the left angle among the distances, and until passing the discontinuous point, the Change the driving mode so that it rotates as much as the rotation angle and drives.

The processor sets a right maximum distance among distances to the surrounding structures as the right reference distance, sets a left maximum distance as the left reference distance, the right reference distance, the right angle, and the left reference distance and the left angle can confirm the position of the surrounding structure.

The processor may calculate a rotation angle to travel to a specific position between the position of the structure in which the right distance is measured and the position of the structure in which the left distance is measured.

The processor may generate the approximate wall surface by obtaining a plurality of coordinates at which the distance of the structure is measured for each angle, and approximating the plurality of coordinates with a straight line.

The processor may set the driving mode to travel along the approximate wall surface to be separated from the approximate wall surface by a predetermined distance and maintain a predetermined angle to the approximate wall surface.

According to the present invention, even in an environment where obstacles such as a restaurant are installed at narrow intervals and frequently generate a curved driving path, the autonomous driving robot facilitates driving in a corner section without colliding with an obstacle through the triangular bisector corner recognition method. can do.

1 is an exemplary diagram of a right-angled triangle method used for setting a curved driving route.
2 is an exemplary view showing a cause of a collision with a structure when driving a corner.
3 is an exemplary diagram of an environment to which an autonomous driving robot is applied according to an embodiment of the present invention.
4 is a structural diagram of an autonomous driving robot according to an embodiment of the present invention.
5 is a flowchart of a driving method according to an embodiment of the present invention.
6 is a flowchart of a method for checking a corner point according to an embodiment of the present invention.
7 is an exemplary view in which a target path is set by recognizing a left turn corner according to an embodiment of the present invention.
8 is an exemplary view for recognizing a left turn corner according to another embodiment of the present invention
9 is an exemplary diagram of an initial map generated by a robot according to an embodiment of the present invention.
10 is an exemplary diagram of update information to be reflected in an initial map according to the first embodiment of the present invention.
11 is an exemplary diagram of update information to be reflected in an initial map according to a second embodiment of the present invention.
12 is an exemplary diagram of update information to be reflected in an initial map according to a third embodiment of the present invention.
13 is an exemplary diagram of a driving map according to an embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, the embodiments of the present invention will be described in detail so that those of ordinary skill in the art to which the present invention pertains can easily implement them. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part "includes" a certain element, it means that other elements may be further included, rather than excluding other elements, unless otherwise stated.

Hereinafter, an autonomous driving robot and a method for generating a curved driving path of the autonomous driving robot according to an embodiment of the present invention will be described in detail with reference to the drawings. Before describing the embodiment of the present invention, curved driving using a right-angled triangle method, which is a wall-following traveling method, will first be described with reference to FIGS. 1 and 2 .

1 is an exemplary diagram of a right-angled triangle method used for setting a driving route.

When the robot 10 plans the travel route, the robot 10 sets the travel route in a right-angled triangle manner. That is, the robot 10 sets the approximate wall surface by approximating the wall surface, which is an irregular structure, as a straight wall surface, and maintains a right-angled triangle with respect to the approximate wall surface to run parallel to the approximate wall surface at a predetermined distance.

That is, the robot 10 may transmit a laser for each angle using the lidar 11 , and obtain the distance to the structure and the coordinates of the reflected point based on the reflected wave reflected by the structure. In this case, the structures of the space in which the robot 10 travels have irregular shapes such as a square table, a round table, and a round column on the wall.

Accordingly, the robot 10 should define the structures as approximate straight lines. To this end, the robot 10 generates an approximate wall surface by approximating the group of coordinates of the point where the laser is reflected to a single straight line. Then, the robot 10 travels along the approximate wall surface, and in this case, an error distance with the structure may occur.

When an error distance occurs, an area error ( _Re ) occurs between the height side and the approximate wall surface as shown in FIG. 1 . Here, the area error is calculated as the sum of the rectangle ( _Re1 ) and the right triangle (T _e2 ) as shown in Equation 1 below.

D is the length of the base of the right triangle, and D _e is the distance error of the right triangle based on the approximate wall surface. And θ _e is the error angle between the height side of the right triangle and the approximate wall surface.

When θ _e is sufficiently small, R _e1 and T _e2 may be approximated by Equations 2 and 3, respectively.

Here, when _{De and θ e converge} to 0 through the feedback control of the robot 10 , the robot 10 runs while maintaining a constant distance and angle with the approximate wall surface.

As described above, when using the right-angled triangle method to drive in a place where the table interval is narrow or the wall surface changes rapidly, the robot 10 mostly collides with a wall or an obstacle during corner driving or stops. In this case, the cause of the collision will be described with reference to FIG. 2 .

2 is an exemplary view illustrating a cause of a collision with a structure when driving a corner.

As shown in FIG. 2 , when the robot 10 maintains a right triangle with the right wall (or table) through a right triangle method, it approaches the corner. And, even after passing the G ₀ point where the distance to the front wall becomes D, the front wall is not recognized until the data is updated after the lidar scan, and it goes straight ahead by an additional distance (Δ).

After passing the G ₁ point moved by the additional distance (Δ), the robot 10 starts to recognize the front wall. When the front wall is recognized, the robot 10 starts curved driving and recognizes the approximate wall surface as the front while passing the G ₂ point.

Here, when the approximate wall surface abruptly changes from the right to the front with respect to the robot 10, the area error _Re increases rapidly. Accordingly, the robot 10 rotates when the steering angle is set to the maximum and the minimum turning radius R is maintained.

At this time, in order for the robot 10 to rotate without colliding with the approximate wall surface, a frontal distance greater than the sum of the minimum turning radius R and the additional distance Δ must be secured. Since the additional distance Δ increases the turning radius of the robot 10, if the sum of the minimum turning radius R and the additional distance Δ is less than D, the robot 10 collides with the approximate wall surface.

Therefore, when the robot 10 runs in a dining space where the table spacing is as narrow as 100 cm, the robot 10 unconditionally collides with a table or a wall. This problem can be solved by widening the table installation interval, but there is a disadvantage in that the space utilization efficiency is lowered. In addition, most of the restrictions occur in restaurants, and when the robot turning radius R is reduced, the risk to the robot 10 itself increases due to the drift, slip, and rotation of the robot 10 .

Accordingly, in the embodiment of the present invention, an autonomous driving robot (hereinafter, referred to as a 'robot' for convenience of description) 100 that autonomously travels at high speed in a dining environment with a narrow table interval is a corner without colliding with an obstacle such as a table. We propose a triangle bisector corner recognition method for driving.

The triangle bisector corner recognition method is a method in which a corner is recognized through an algorithm that detects a corner in advance before entering a corner, and the robot 100 runs through a triangular bisector target path setting algorithm. By using the triangle bisector corner recognition method, it can recognize the structures in the dining room, such as tables, chair corners, and walls, which can be corners, and enable corner driving.

Next, the environment of the robot 100 will be described with reference to FIG. 3 . In the embodiment of the present invention, a restaurant is described as an example of a space in which the robot 100 travels, but any space having conditions similar to that of a restaurant can be applied to any space.

3 is an exemplary diagram of an environment to which an autonomous driving robot is applied according to an embodiment of the present invention.

As shown in FIG. 3 , in order to travel in a space in which obstacles are installed at narrow intervals or where a curved travel path is frequently generated, the robot 100 generates an initial map of the space in which it is located. Here, the initial map is a map generated while the robot 100 initially travels the entire dining space, in which the robot 100 can run and an area where it cannot run (eg, a wall, a column, etc.) ) are separated and created.

The robot 100 transmits the generated initial map to the map management server 200 . The map management server 200 reflects the padding value input by the map manager to the initial map, and updates the driving map. The driving map updated by the map management server 200 is transmitted to the robot 100 .

Here, the padding value means a value input by the map manager to adjust the approach distance to the customer or obstacle when the robot 100 travels in space. That is, when the robot 100 provides a service while driving in a restaurant, a collision with the customer may occur when the robot 100 runs within a range in which the chair is moved by the restaurant customer.

Accordingly, the map manager sets the range using the padding value to prevent the robot 100 from traveling. Since the method for the map manager to determine the driving range using the padding value is variously performed, the embodiment of the present invention is not limited to any one method.

The robot 100 checks whether there is a discontinuous point, that is, a corner section, using a sensor installed in the robot 100 while driving the dining space using a right-angled triangle method, which is a wall-following driving method, based on the driving map. And when the corner section is sensed, the robot 100 changes the driving mode from the right-angled triangle method to the triangular bisector method for corner driving, and drives the corner in the changed driving mode.

Here, the robot 100 recognizes a discontinuous point, that is, a corner section in various ways. In an embodiment of the present invention, a method of first identifying a position having the longest distance value among the collected distance values, and the collected distance increase ratio It will be described including how to check the position of the section that has increased by more than a threshold value in the information about it. At this time, the robot 100 according to the embodiment of the present invention is described as an example of measuring the distance value to the structure using a LiDAR (Light Detection And Ranging) sensor, but radar, depth camera, and various sensors etc. can be used, but is not necessarily limited thereto.

Here, the structure of the robot 100 will be described with reference to FIG. 4 .

4 is a structural diagram of an autonomous driving robot according to an embodiment of the present invention.

As shown in FIG. 4 , the robot 100 includes a processor 110 , a memory 130 , a sensor 140 , an interface 150 , a storage device 160 , and a driving device that communicate through a bus 120 . 170).

The processor 110 is a device for controlling the operation of the robot 100, and may be a processor of various types that processes instructions included in a program, for example, a central processing unit (CPU), a micro processor unit (MPU) , a micro controller unit (MCU), a graphic processing unit (GPU), or the like. Alternatively, the processor 110 may be a semiconductor device that executes instructions stored in the memory 130 or the storage device 160 . The processor 110 may be configured to execute functions and methods of the robot 100 to be described later.

The processor 110 creates an initial map when the robot 100 first travels in space. A method for the processor 110 to create an initial map is already known, and a detailed description thereof will be omitted in the exemplary embodiment of the present invention.

The processor 110 transmits the generated initial map to the map management server 200 connected through the network. Here, since the shapes of the obstacles mapped to the initial map are different from the shapes of the actual obstacles, when the robot 100 travels to the initial map, there is a high possibility of colliding with the obstacles or structures. In addition, according to a request in the space in which the robot 100 is installed, the robot 100 must run based on the robot service guidelines.

Accordingly, the manager who manages the robot 100 updates the initial map registered in the map management server 200 as a driving map according to the robot service guidelines. In this case, a method of updating the initial map to the driving map will be described in detail later.

The processor 110 uses the driving map updated by the map management server 200 to drive the sensor 140 in a right-angled triangle manner (hereinafter also referred to as 'straight driving mode' or 'first driving mode'). The laser is output for each angle based on the traveling direction of the robot 100 using the , and distance values calculated by reflecting the output laser on the structure are calculated.

Here, it is assumed that the structure is installed on both sides of the front side of the robot 100 and the robot 100 in the traveling direction of the space. A method for the processor 110 to calculate the distance value to the structure for each angle through the sensor 140 is already known, and a detailed description thereof will be omitted in the embodiment of the present invention.

And the processor 110 is the longest left distance (D _L ) in the left direction based on the driving direction of the robot, among the distance values for a plurality of points reflected from the structure and measured, the left angle at which the distance is measured, the right Extract the longest right distance (D _R ) in the direction and the right angle at which the distance was measured.

When there is a point where the ratio of two neighboring distances with respect to the robot 100 is equal to or greater than a preset threshold, the processor 110 identifies the point as a discontinuous point. Here, the discontinuous point means that a corner is formed at the corresponding point.

When the discontinuous point is identified, the processor 110 changes the driving mode from the straight driving mode to the triangular bisector corner recognition method (hereinafter also referred to as a 'corner driving mode' or a 'second driving mode') that is a corner driving mode. In addition, the processor 110 calculates a rotation angle to be controlled by the robot 100 based on the longest right distance and right angle, and the longest left distance and left angle.

The processor 110 controls the robot 100 to rotate at the confirmed rotation angle, and travels at the rotated angle. Here, the rotation angle is the longest right distance and the right angle, the longest left distance and the left angle, and the point corresponding to the longest right distance and the right angle in the triangle generated based on the position of the robot 100, and the most This is the angle that controls the movement to the target path passing through the coordinates of the triangle bisector between the long left distance and the point corresponding to the left angle. The processor 110 controls the driving device 170 so that the calculated rotation angle converges to zero.

On the other hand, the processor 110 may calculate the distance to the surrounding structures of the robot 100 for each angle based on the reflected signal received after the laser output by the sensor 140 for each angle is reflected by the structure. In this case, if the structure is continuously maintained like a wall, the distance from the robot 100 to the structure is lengthened or shortened in a predetermined pattern according to the angle.

However, when there is a corner section, the robot 100 can confirm that the distance reflected from the location of the structure where the corner section starts and the distance reflected from the front wall are abruptly lengthened or shortened rather than in a predetermined pattern. In this case, the processor 110 determines that there is a corner section in the abruptly changed section.

The memory 130 may include various types of volatile or non-volatile storage media. For example, the memory 130 loads a corresponding program so that instructions described to execute the operations of the present invention are processed by the processor 110 , a read only memory (ROM) 131 and a random access memory (RAM) (132).

In an embodiment of the present invention, the memory 130 may be located inside or outside the processor 110 , and the memory 130 may be connected to the processor 110 through various known means.

The sensor 140 corresponds to various equipment (eg, lidar, depth camera, inertial measurement unit (IMU), etc.) from which the robot 100 while driving according to the operation of the present invention collects information. do. In the embodiment of the present invention, the use of the lidar as the sensor 140 is described as an example, but the present invention is not necessarily limited thereto.

The interface 150 interworks with the map management server 200 , and transfers the initial map generated by the robot 100 to the map management server 200 . Also, the interface 150 may receive the driving map transmitted from the map management server 200 and transmit it to the processor 110 .

The storage device 160 may store an initial map of a space in which the robot 100 travels. Also, the storage device 160 stores the driving map updated by the map management server 200 .

The driving device 170 causes the robot 100 to travel in the space under the control of the processor 110 .

A method of recognizing a corner and driving while the robot 100 described above travels in space using the driving map will be described with reference to FIG. 5 .

5 is a flowchart of a driving method according to an embodiment of the present invention.

As shown in FIG. 5 , the robot 100 creates an initial map to travel in an arbitrary space ( S100 ). The initial map is a map generated by the robot 100 driving the entire space, and the initial map is divided into an area in which the robot 100 can drive and an area in which the robot 100 cannot drive (eg, a wall, a pillar, etc.) is created by A method for the robot 100 to create an initial map is already known, and a detailed description thereof will be omitted in the embodiment of the present invention.

The robot 100 transmits the created initial map to the map management server 200 (S101). The map management server 200 receiving the initial map inputs update information (eg, robot body size, obstacle or structure size, obstacle shape, padding value, boundary line processing information, etc.) to be reflected on the initial map from the map manager. receive And the map management server 200 updates the driving map reflecting the padding value in the initial map, and transmits it to the robot 100 (S102, S103).

The robot 100 starts autonomous driving in a right-angled triangle method, which is a straight-line driving mode, in an arbitrary space based on the driving map received in step S103 (S104). Here, the arbitrary space will be described as an example of a space such as a restaurant in which structures are installed around the robot 100 . And the structure may be a dining room table or a dining room wall.

The robot 100 outputs lasers for each angle using a sensor 140 such as a lidar sensor while autonomously driving. And when the laser output from the robot 100 is reflected back to the structures in the space, the robot 100 calculates distance values to the structures for each angle based on the driving direction based on the reflected and returned time (S105).

The robot 100 checks whether there is a discontinuous point, that is, a corner section, based on the calculated distance values (S106). In the embodiment of the present invention, two methods are exemplified as a method for checking the presence or absence of a corner section.

The first method checks the two distances having the largest distance values from the right and left sides with respect to the robot 100, and if there is a short left or right distance below the threshold value compared to the checked distance, a corner section is located in the corresponding section. A way to judge that there is. And in the second method, the distance value is first collected at an angle of 180 degrees (or 360 degrees) left and right based on the sensor 140 of the robot 100, and the ratio between the two distance values is changed to be greater than or equal to a preset threshold This is a method of judging that there is a corner section in the section being used.

In the exemplary embodiment of the present invention, a method of determining the presence or absence of a corner section will be described using this exemplary embodiment, but the present invention is not limited thereto.

The robot 100 checks whether a corner section, which is a discontinuous point, exists based on the distance values (S107), and if there is no corner section, repeats the procedure after step S104. However, if there is a corner section, the robot 100 changes the driving mode from the straight driving mode to the corner driving mode (S108), and drives until it exits the corner section in the corner driving mode (S109).

Among the above procedures, a method for the robot 100 to check the discontinuous point in step S106 will be described with reference to FIG. 6 .

6 is a flowchart of a method for checking a corner point according to an embodiment of the present invention.

As shown in FIG. 6 , the robot 100 calculates the distance values and angles in all directions based on the driving direction ( S200 ), and then calculates the ratio of the adjacently measured distances ( S201 ). That is, the robot 100 calculates the ratio of the distances by dividing the two distance values.

The robot 100 checks whether there is a value having a threshold value or more among the values calculated in step S201. Then, the robot 100 recognizes a point at which an adjacent distance value having a value greater than or equal to the threshold value is measured as a corner that is a discontinuous point (S202).

The robot 100 measures the longest right distance (or, referred to as 'right maximum distance') among distance values in all directions, the right coordinate at which the distance is measured, and the longest left distance (or referred to as 'left maximum distance') and a rotation angle is calculated based on the left coordinate at which the distance is measured (S203). Then, the driving continues while controlling the rotation angle to converge to 0 (S204).

Next, an example in which the robot 100 recognizes a corner and drives in a corner driving method will be described with reference to FIG. 7 . In an embodiment of the present invention, an embodiment in which a target path is set by recognizing a left-turn right-angle corner will be described.

7 is an exemplary view in which a target path is set by recognizing a left turn corner according to an embodiment of the present invention.

As shown in FIG. 7 , the robot 100 scans a structure in space using a lidar, a sensor installed in the front, and distances from the left and right based on the front direction, which is the running direction of the robot 100, among the scan data. Checks the longest left reference distance (D _L ) and the longest right reference distance (D _R ).

If a corner exists in the space in which the robot 100 is traveling, a discontinuous point indicating a large difference between the left reference distance D _L or the right reference distance D _R is compared with the distance of the immediately adjacent point. .

In FIG. 7 , a discontinuous point occurs at a corner located on the left of the left reference distance D _L , and the distance of the discontinuous point to the discontinuous point is denoted by D _D . In the present invention, if the ratio (D _L /D _D ) of the left reference distance (D _L ) and the distance to the discontinuous point is greater than 2, it is defined as discontinuous and recognized as a corner.

Here, when the ratio (D _L /D _D ) of the left reference distance (D _L ) and the distance to the discontinuous point (D _D ) increases, the robot 100 recognizes a corner section at a point closer to the discontinuous point. Similarly, when the ratio is small, a corner section can be recognized at a point further away from the discontinuity point.

Through this, the robot 100 recognizes the corner section by checking the presence or absence of discontinuous points before entering the corner section, and switches the driving mode from the straight-line driving mode, which is the existing right-angled triangle method, to the corner driving mode, which is the triangular bisector method. . The robot 100 travels in a corner section in a corner driving mode, and can travel stably without collision even in a place where the driving width is narrow due to structures or obstacles.

Here, the robot 100 calculates the left reference distance (D _L ) and the left angle (θ _L ), the right reference distance (DR ) and the right angle (θ _{R )} using the following Equations 4 and 5 to coordinate data can be obtained with

Here, θ _L represents an angle to the left reference distance D _L with respect to the robot 100 , and θ _R represents an angle to the right reference distance D _R from the robot 100 .

)를 확인한다.When the robot 100 checks the coordinates of the first point and the coordinates of the second point, the robot 100 may identify a triangle having the coordinates of the first point, the coordinates of the second point, and the coordinates of the robot 100 as vertices, respectively. And based on the coordinates of the robot 100, the coordinates of the bisector of the triangle (

) is checked.

The robot 100 sets the target path with the coordinates of the triangular bisector point. At this time, the error of the rotation angle θ _e between the existing driving coordinates and the set target path is obtained by Equation 6 below.

Here, α denotes a weight according to a difference in characteristics (eg, deceleration speed, maximum speed, acceleration, etc.) of the robot 100 . It is a weight for correcting the rotation angle calculated through Equation 6 for each difference in the characteristics of the robot 100 .

)를 지나는 목표 경로를 따라 이동하게 된다. 이를, 로봇(100)이 코너에 진입하기 전에 코너를 인식할 수 있고, 로봇(100)의 주행 방향의 정면 벽을 기준으로 목표 경로를 설정할 수 있으므로, 장애물 또는 구조물과의 충돌 없이 코너 구간을 통과할 수 있다.After calculating the rotation angle (θ _e ) error in this way, when the robot 100 travels while rotating so that the rotation angle (θ _e ) converges to 0, the coordinates (

) along the target path passing through In this way, the robot 100 can recognize the corner before entering the corner, and set the target path based on the front wall in the driving direction of the robot 100, so that it passes through the corner section without colliding with obstacles or structures. can do.

Next, an example of recognizing a corner section that is a discontinuous section according to another embodiment of the present invention will be described with reference to FIG. 8 .

8 is an exemplary view for recognizing a left turn corner according to another embodiment of the present invention.

As shown in FIG. 8 , the robot 100 may check the distance to the structure for each preset angle among the left and right 180 degrees through the sensor 140 . When the distance and angle to the structure are checked, the robot 100 may also check the coordinates corresponding to the distance of the structure.

At this time, at an angle (90 degrees) corresponding to the robot traveling direction, the sensor 140 is set to 0 degrees to the left of the robot 100 and 180 degrees to the right of the robot 100 as an example. In addition, although measuring the distance to the structure at an angle of 5 degrees is described as an example, the angle and direction are not necessarily limited as such.

At this time, the robot 100 can see that the distance to the structure measured at an angle of 35 degrees and the distance to the structure measured at an angle of 40 degrees increased rapidly compared to the rate of increase or decrease between the distances measured at other adjacent angles. . Accordingly, the robot 100 may determine the 35 degree section as a discontinuous point, and determine that the corner section starts from the section.

Next, an initial map initially generated by the robot 100, a process in which updated information is reflected in the initial map, and a driving map generated through the initial map will be described with reference to FIGS. 9 to 13 .

9 is an exemplary diagram of an initial map generated by a robot according to an embodiment of the present invention.

An initial map created by the robot 100 through initial driving in a restaurant is shown in FIG. 9 . 9 is an initial map generated by a robot 100 equipped with a lidar while driving a restaurant.

A portion expressed in black is a non-drivable space of the robot 100 , and most of it corresponds to an area where obstacles such as tables and chairs are placed. The table interval of the space mapped with the initial map is about 110 cm, so the driving space is narrower than the space where the robot 100 is generally put in a lot.

Accordingly, the map management server 200 receiving the initial map updates the initial map as a driving map based on the update information input by the map manager.

According to the universal driving guidelines for serving robots that are currently providing autonomous driving services in Korean restaurants, it is recommended that tables installed in restaurants be at least 2m apart. This is a guideline for preventing accidents and stably operating the robot when a crowded situation occurs in the restaurant. However, since the table spacing is a part directly related to a restaurant's sales, preference for autonomous driving services of robots may decrease when guidelines are observed.

Accordingly, the map manager uses the update information to update the initial map so that the robot 100 can drive in compliance with the guidelines when there are many restrictions in space such as a restaurant and there are many obstacles around the robot 100 . can do. This will be described with reference to FIGS. 10 to 12 .

10 is an exemplary diagram of update information to be reflected in an initial map according to the first embodiment of the present invention. 11 is an exemplary diagram of update information to be reflected in an initial map according to a second embodiment of the present invention, and FIG. 12 is an exemplary diagram of update information to be reflected in an initial map according to a third embodiment of the present invention.

The updated information includes the size of the robot 100 body, the size of obstacles including walls, correction of the shape of the obstacles, and boundary line processing information for the approach risk area. In the embodiment of the present invention, in order for the robot 100 to travel in a restaurant, an example of updating the initial map to a driving map so that the table interval is 1 m or less will be described.

In the initial map created by the robot 100, the robot manager creates a driving boundary line except for the serving route. At this time, the range around the chair set on the table is included in the standard by examining the driving boundary.

That is, as shown in Fig. 10, the map manager performs boundary processing around the table periphery as indicated by the first display means (①). When boundary processing is performed like the first display means, the robot 100 cannot run between the table and the table.

At this time, even if the table legs are located only in the middle rather than at each corner of the table top, or in different positions, it will be described as an example that boundary processing is always performed based on the boundary of the table top. Since it is difficult to detect table legs due to the characteristics of lidar, boundary processing is performed based on the boundary of the upper plate to prevent the risk of collision with the upper plate.

In addition, as shown in FIG. 11 , when the robot 100 passes a space with a chair to travel in a narrow space, a boundary line is set centered on the width of the chair as indicated by the second display means (②). do.

In addition to this, as shown in FIG. 12 , when a round table is used in a restaurant, the chairs can be arranged in any direction due to the characteristics of the round table. Therefore, the map manager makes a square based on the diameter of the round table and borders it with a larger square around the width of the chair. Based on this, the robot 100 can stably provide a serving service.

A driving map generated by reflecting the above-described updated information on an initial map will be described with reference to FIG. 13 .

13 is an exemplary diagram of a driving map according to an embodiment of the present invention.

As shown in FIG. 13, since the boundary treatment (③) has been made so that traveling between tables is impossible even in a narrow space with a table interval of about 1 m, the robot 100 can stably provide serving. And when there is a person in the entry path, the robot 100 can also stably set the avoidance path.

At this time, the boundary line is also processed for elements that interfere with the lidar sensor data such as glass, mirrors, and metal in the restaurant to support stable driving.

And, in an embodiment of the present invention, by setting the size of the robot 100 to be larger than the actual size, the map manager can adjust the approach distance to the customer in the restaurant within the range in which the chair can move.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto. is within the scope of the right.

Claims (20)

As a driving method of a robot,
In the first driving mode, while measuring the distances to the surrounding structures by angles through the lidar, driving along the wall surface generated by approximating the surrounding structures;
Checking whether there is a discontinuous point based on the distances for each angle to the surrounding structures, and
If there is the discontinuous point, the first driving mode is changed to the second driving mode, and while rotating at a rotation angle calculated based on the distances for each angle collected through the lidar, until the rotation angle becomes 0 driving in the second driving mode
including,
The rotation angle is an angle obtained by correcting an angle for controlling the robot based on a weight according to the speed of the robot. The calculated angle at which the robot is directed to the coordinates of the midpoint between the coordinates of the maximum distance. According to claim 1,
The step of running along the wall is,
A driving method for generating the wall surface by approximating the coordinates of each point reflected by the angle of the laser output from the lidar to the surrounding structure as a single straight line. 3. The method of claim 2,
The step of running along the wall is,
Calculating an area error between a side from the ground to the height of the lidar and the wall surface, and
Driving along the wall while maintaining the distance and angle with the wall so that the area error converges to 0
A driving method comprising: According to claim 1,
The step of checking whether there is a discontinuous point is,
Check the right maximum distance among the right distances for each angle measured from the right side based on the traveling direction of the robot, and if there is a short right distance below the threshold value compared to the right maximum distance, the right side of the robot traveling direction determining that there is a corner section, and
Check the left maximum distance from among the left distances for each angle measured from the left side based on the driving direction of the robot, and if there is a short left distance below the threshold value compared to the left maximum distance, the left side of the robot traveling direction is Step to determine that there is a corner section
A driving method comprising: 5. The method of claim 4,
The step of driving in the second driving mode includes:
checking the right angle at which the right maximum distance and the right maximum distance are measured as the coordinates of the right maximum distance;
Confirming the left angle at which the left maximum distance and the left maximum distance are measured as the coordinates of the left maximum distance; and
calculating the rotation angle to pass through the coordinates of the intermediate point, which is a bisector point on a side generated between the coordinates of the right maximum distance and the coordinates of the left maximum distance
including,
The driving method, wherein the rotation angle is updated based on the distance for each angle collected through the lidar until passing the discontinuous point. According to claim 1,
After the step of driving in the second driving mode,
When passing the discontinuous point while rotating so that the rotation angle converges to 0, changing the driving mode and driving along a new wall surface generated by approximating the surrounding structures at the position passing the discontinuous point
Further comprising, a driving method. As a driving method of a robot,
Driving while measuring the distance to surrounding structures,
Checking the longest right distance among the right distances for each angle measured in the right side of the driving direction, and determining that there is a corner section on the right side of the driving direction if there is a short right distance below the threshold value compared to the longest right distance step,
Check the longest left distance among the left distances for each angle measured in the left side of the driving direction, and if there is a short left distance equal to or less than a threshold value compared to the longest left distance, it is determined that there is a corner section on the left side of the driving direction step, and
passing the entered corner section while changing the rotation angle in the direction of the right corner section or the left corner section
including,
The rotation angle is an angle obtained by correcting an angle for controlling the robot based on a weight according to the speed of the robot. The calculated angle at which the robot is directed to the coordinates of the midpoint between the coordinates of the maximum distance. 8. The method of claim 7,
The step of determining that there is a corner section on the right side,
Based on the angle at which the longest right distance and the longest right distance are measured, confirming the location of the structure in which the longest right distance is measured
A driving method comprising: 9. The method of claim 8,
The step of determining that there is a corner section on the left side,
Based on the angle at which the longest left distance and the longest left distance are measured, confirming the location of the structure in which the longest left distance is measured
A driving method comprising: 10. The method of claim 9,
The step of passing the corner section is,
Calculating a rotation angle based on the traveling direction of the robot based on the angle of the structure in which the right distance is measured, the angle of the structure in which the left distance is measured, and the position of the robot
A driving method comprising: delete 11. The method of claim 10,
The step of passing the corner section is,
Calculating the angle to the driving target position based on the position of the structure in which the longest right distance is measured and the position of the structure in which the longest left distance is measured
Further comprising, a driving method. 13. The method of claim 12,
The step of passing the corner section is,
If there is no short right distance or short left distance below the threshold, obtaining a plurality of coordinates of the structures for each angle based on the distance to the structures;
generating an approximate wall surface by approximating the plurality of coordinates with a straight line; and
generating and driving a driving route so as to run parallel to the approximate wall surface at a location separated from the approximate wall surface by a predetermined distance;
A driving method comprising: 8. The method of claim 7,
Before the driving step,
generating an initial map while driving the space in which the structures are installed; and
transmitting the initial map to a server interworking with the robot, and receiving a driving map in which update information is reflected in the initial map from the server
including,
The update information includes the size of the robot, the size of the structure, the shape of the structure, and boundary line processing information in which a plurality of structures are grouped into one structure and processed. 15. The method of claim 14,
Receiving the driving map comprises:
generating a wall surface by approximating the surrounding structure; and
Driving along the wall while measuring the distances to the surrounding structures by angle through the lidar
Further comprising, a driving method. A robot that changes a driving route during autonomous driving, comprising:
A sensor that measures the distances to the surrounding structures by angle, and
processor
including,
The processor is
The driving mode is set to travel along the approximate wall surface generated based on the surrounding structure, and a discontinuous point is identified based on the distances to the surrounding structure. When the discontinuous point is confirmed, the right reference distance and right angle among the distances , and calculate the rotation angle to rotate the robot based on the left reference distance and the left angle, rotate by the rotation angle until passing the discontinuous point, and change the driving mode to travel until the rotation angle becomes 0 , robot. 17. The method of claim 16,
The processor is
Among the distances to the surrounding structures, the right maximum distance is set as the right reference distance, and the left maximum distance is set as the left reference distance,
The robot, which confirms the position of the surrounding structure with the right reference distance and the right angle, and the left reference distance and the left angle. 18. The method of claim 17,
The processor is
A robot that calculates a rotation angle to travel to a specific position between the position of the structure in which the right maximum distance is measured and the position of the structure in which the left maximum distance is measured. 19. The method of claim 18,
The processor is
A robot for generating the approximate wall surface by obtaining a plurality of coordinates at which the distance of the structure is measured for each angle, and approximating the plurality of coordinates with a straight line. 20. The method of claim 19,
The processor is
and setting a driving mode to travel along the approximate wall surface to be separated from the approximate wall surface by a certain distance and to maintain a certain angle to the approximate wall surface.

Images