Method for preventing robot collisions and apparatus for preventing robot collisions using same
By generating repulsive and attractive fields from three-dimensional coordinates, the method and device guide robots to avoid obstacles, improving safety and efficiency in industrial environments.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- HANYANG ROBOTICS
- Filing Date
- 2025-03-07
- Publication Date
- 2026-07-23
AI Technical Summary
Autonomous robots in industrial settings face challenges in avoiding collisions with obstacles, particularly when unexpected movements occur or when operators cannot continuously monitor their movements, leading to potential accidents and reduced work efficiency.
A method and device that utilize three-dimensional coordinate systems to generate repulsive and attractive fields, synthesizing them into a potential field to guide the robot along an optimal path, minimizing collision risk by continuously updating the field as the robot moves, using sensors to detect obstacles and adjust movement paths in real-time.
The solution effectively prevents collisions, enhancing work speed and reducing accidents by enabling the robot to navigate optimally around static and dynamic obstacles, ensuring safe and efficient operation.
Smart Images
Figure KR2025099623_23072026_PF_FP_ABST
Abstract
Description
Robot collision prevention method and robot collision prevention device using the same
[0001] The present invention relates to a robot collision prevention method and a robot collision prevention device using the same, and more specifically, to a robot collision prevention method and a robot collision prevention device using the same that prevent an autonomous robot used in an industrial site from colliding with an obstacle while moving, thereby improving work speed and reducing the occurrence of accidents.
[0002] Autonomous robots used in industrial settings are configured to move from their current location to a target teaching location, and the robot moves to the teaching location to perform a specific task. However, when the autonomous robot uses the "Jog to" function to move directly from the current location to the taught location, safety accidents may occur if the robot fails to reduce speed or proceeds carelessly, resulting in a collision with an obstacle. In particular, the risk of accidents increases significantly in cases of unexpected movements, such as obstacles unexpectedly entering the path. For instance, if an operator needs to perform additional tasks while the autonomous robot is operating in automatic driving mode, there is a risk of the robot colliding with the operator. Furthermore, it is difficult for an operator to continuously monitor the autonomous robot's movements to prevent unexpected obstacles from entering the path.
[0003] Specifically, in the case of an extraction robot for removing injection-molded parts, the robot may have a sequence of operations such as waiting to grasp the part, removing it after breaking it, and lifting it while it is grasped. However, if an error occurs in this sequence or if the sequence is entered incorrectly by an operator, the extraction robot may interfere with structures surrounding the injection molding machine, such as pillars, during operation, resulting in a collision. Furthermore, in an environment pre-set to prevent collisions by considering obstacles during the extraction robot's operation, unexpected collisions may occur between the extraction robot and the structures if additional structures are installed. To prepare for such situations, it is currently difficult for an operator to constantly monitor the robot's movements.
[0004] Accordingly, the present invention proposes a robot collision avoidance method and device capable of avoiding collisions by considering obstacles present in the surroundings during the movement of an autonomous robot.
[0005] The present invention has been devised to solve the requirements described above, and aims to provide a robot collision prevention method and a robot collision prevention device using the same, which prevent autonomous robots used in industrial sites from colliding with obstacles while moving, thereby improving work speed and reducing the occurrence of accidents.
[0006] A robot collision prevention method according to one aspect of the present invention comprises: a first step (S1) of obtaining three-dimensional coordinate values of an obstacle; a second step (S2) of determining the area of the obstacle by projecting the three-dimensional coordinate values of the obstacle onto a plane; a third step (S3) of generating a repulsive field by the obstacle area at a first position of the robot; a fourth step (S4) of generating an attractive field by the target point to which the robot is to move at the first position of the robot; a fifth step (S5) of generating a potential field by synthesizing the repulsive field and the attractive field at the first position of the robot; a sixth step (S6) of moving the robot to a second position corresponding to the minimum value of the potential field; and a seventh step (S7) of moving the robot to the target point by repeating the first step (S1) to the sixth step (S6) at predetermined time intervals until the second position coincides with the target point, when the robot moves to the second position, with the second position as the new first position.
[0007] In addition, the step of obtaining three-dimensional coordinate values of the obstacle is obtained using an orthogonal coordinate system for the obstacle, and the coordinate values based on the orthogonal coordinate system can be converted into coordinate values based on a spherical coordinate system.
[0008] In addition, after the above three-dimensional coordinate values are converted into coordinate values based on the spherical coordinate system, data corresponding to a predetermined reference plane forming the floor in the above orthogonal coordinate system can be removed.
[0009] In addition, when projecting the three-dimensional coordinate values of the obstacle onto a plane, a reference value spaced apart from the robot can be set, and the coordinate values included in the range below the reference value from the robot can be projected onto the plane.
[0010] Additionally, the area of the obstacle can be set as a rectangle that captures the perimeter of the projected plane.
[0011] In addition, the repulsive force field can be calculated symmetrically or asymmetrically with respect to the obstacle area set as the rectangle, by considering the center of the obstacle area and the center of mass of the projected obstacle.
[0012]
[0013] Meanwhile, a robot collision prevention device according to another aspect of the present invention comprises: a sensor unit that acquires three-dimensional coordinate values of an obstacle using an orthogonal coordinate system; an obstacle area determination unit that converts the coordinate values acquired by the sensor unit into coordinate values corresponding to a spherical coordinate system and orthographically projects the coordinate values based on the spherical coordinate system onto a plane corresponding to a horizontal angle and a vertical angle to determine the area of the obstacle; a field generation unit that generates a repulsive field by the obstacle area at a first position of the robot, an attractive field by the target point to which the robot is to move at the first position of the robot, and a potential field by synthesizing the repulsive field and the attractive field at the first position of the robot. The invention is characterized by comprising a control unit that moves the robot from a first position to a second position corresponding to the minimum value of the potential field, and when the robot moves to the second position, the field generation unit generates a new reaction force field, an attractive force field, and a potential field by making the second position a new first position, and repeatedly generates the potential field at predetermined time intervals until the second position coincides with the target point, thereby controlling the robot to track the minimum value of the potential field and reach the target point.
[0014] Here, the obstacle area determination unit can remove data corresponding to a predetermined reference plane forming the floor in the orthogonal coordinate system after the three-dimensional coordinate values are converted into coordinate values based on the spherical coordinate system.
[0015] Here, the obstacle area determination unit can set a reference value spaced at a predetermined distance from the robot and project coordinate values included in a range less than or equal to the reference value from the robot onto a plane.
[0016] Here, the obstacle area can be set as a rectangle that captures the perimeter of the projected plane.
[0017] Here, the field generating unit can generate the repulsive force field symmetrically or asymmetrically for the obstacle area set as the rectangle by considering the center of the obstacle area and the center of mass of the projected obstacle.
[0018] The robot collision prevention method and the robot collision prevention device using the same according to the present invention provide the effect of preventing an autonomous robot used in an industrial site from colliding with an obstacle while moving, thereby improving work speed and reducing the occurrence of accidents.
[0019] By scanning the surrounding environment during the robot's movement, the robot moves along an optimal path that minimizes the risk of collision with obstacles, thereby avoiding obstacles in real time to prevent collisions and allowing the robot to stop if necessary, the effect of significantly reducing the occurrence of accidents is provided.
[0020] FIG. 1 is a flowchart of a robot collision prevention method according to an embodiment of the present invention.
[0021] Fig. 2 is a conceptual diagram of a potential field,
[0022] FIG. 3 is a conceptual drawing illustrating the conversion of coordinate values for an obstacle from a rectangular coordinate system to a spherical coordinate system.
[0023] FIG. 4 is a flowchart specifically illustrating the flowchart of FIG. 1.
[0024] FIG. 5 is a drawing illustrating an example of an exceptional situation.
[0025] FIG. 6 is a drawing showing a situation in which the present invention is applied.
[0026] Fig. 7 is a flowchart illustrating the key parts of Fig. 4 in detail.
[0027] FIG. 8 is a flowchart illustrating another key part of FIG. 4 in detail.
[0028] FIG. 9 is a drawing illustrating an example of a second position transformed by a spherical coordinate system and an example of an exceptional situation.
[0029] FIG. 10 is a drawing showing an example of the changing appearance of the repulsive force field.
[0030] FIG. 11 is a diagram showing an example of the changing appearance of an asymmetric repulsive force field.
[0031] FIG. 12 is a drawing showing the appearance of an obstacle area.
[0032] FIG. 13 is a drawing showing the generation of a symmetric or asymmetric repulsive force field with respect to an obstacle area.
[0033] FIG. 14 is a drawing illustrating an example in which the coordinate values of an obstacle are obtained by a sensor unit mounted on a robot.
[0034] FIG. 15 is a drawing illustrating the setting of an obstacle area for the example of FIG. 14.
[0035] FIG. 16 is a block diagram of a robot collision prevention device according to another embodiment of the present invention.
[0036] Hereinafter, various embodiments of the present invention are described in conjunction with the accompanying drawings. Since various embodiments of the present invention may be subject to various modifications and may have various forms, specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the various embodiments of the present invention to specific forms, and it should be understood that they include all modifications and / or equivalents and substitutions that fall within the spirit and scope of the various embodiments of the present invention. In relation to the description of the drawings, similar reference numerals have been used for similar components.
[0037] Expressions such as "comprising" or "may comprise" that may be used in various embodiments of the present invention indicate the existence of the disclosed function, operation, or component, etc., and do not limit one or more additional functions, operations, or components, etc. Furthermore, in various embodiments of the present invention, terms such as "comprising" or "having" are intended to specify the existence of the features, numbers, steps, operations, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.
[0038] When it is stated that a component is "combined" to another component, it should be understood that the component may be directly connected to the other component, or that a new component may exist between the component and the other component. On the other hand, when it is stated that a component is "directly combined" or "directly in contact" with another component, it should be understood that no new component exists between the component and the other component.
[0039] The terms used in the various embodiments of the present invention are used merely to describe specific embodiments and are not intended to limit the various embodiments of the present invention. The singular expression includes the plural expression unless the context clearly indicates otherwise.
[0040] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the various embodiments of the present invention pertain.
[0041] Terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined in the various embodiments of the present invention.
[0042]
[0043] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a flowchart of a robot collision prevention method according to an embodiment of the present invention, and FIG. 2 is a conceptual diagram of a potential field. FIG. 3 is a conceptual diagram illustrating the conversion of coordinate values for an obstacle from a Cartesian coordinate system to a spherical coordinate system, and FIG. 4 is a flowchart that elaborates on the flowchart of FIG. 1. FIG. 5 is a diagram illustrating an example of an exceptional situation, and FIG. 6 is a diagram showing a situation in which the present invention is applied. FIG. 7 is a flowchart illustrating a key part of FIG. 4 in detail, and FIG. 8 is a flowchart illustrating another key part of FIG. 4 in detail. FIG. 9 is a diagram illustrating an example of a second position converted by the spherical coordinate system and an example of an exceptional situation. FIG. 10 is a diagram showing an example of a change in the repulsion force field, and FIG. 11 is a diagram showing an example of a change in the asymmetric repulsion force field. FIG. 12 is a diagram showing the appearance of an obstacle area, and FIG. 13 is a diagram showing the generation of a symmetric or asymmetric repulsive force field with respect to the obstacle area. FIG. 14 is a diagram illustrating an example in which the coordinate values of an obstacle are obtained by a sensor unit mounted on a robot, and FIG. 15 is a diagram illustrating the setting of an obstacle area for the example of FIG. 14. FIG. 16 is a block diagram of a robot collision prevention device according to another embodiment of the present invention.
[0044]
[0045] The present invention relates to a method and apparatus for preventing collision with obstacles when a robot used in an industrial site moves to a predetermined target point. The robot may be a multi-joint robot, for example, an automated robot for extracting injection molded parts. The extraction robot is used to move to a work area, grasp an injection molded part, and move it to a desired location, and may include a sequence of approaching the injection molded part, grasping the injection molded part, and moving it to a desired location.
[0046]
[0047] A robot collision prevention method according to an embodiment of the present invention comprises, as illustrated in FIG. 1, a first step (S1) of obtaining three-dimensional coordinate values of an obstacle (4); a second step (S2) of determining the area of the obstacle (4) by projecting the three-dimensional coordinate values of the obstacle (4) onto a plane; a third step (S3) of generating a repulsive field by the obstacle area (3) at a first position of the robot (1); a fourth step (S4) of generating an attractive field at a target point to which the robot (1) is to move at the first position of the robot (1); a fifth step (S5) of generating a potential field by synthesizing the repulsive field and the attractive field at the first position of the robot (1); a sixth step (S6) of moving the robot (1) to a second position corresponding to the minimum value of the potential field; and repeating the first step (S1) to the sixth step (S6) at predetermined time intervals until the second position coincides with the target point. It includes a seventh step (S7) of moving the robot (1) to the above-mentioned target point. As illustrated in FIG. 2, the present invention relates to a method and apparatus for generating a potential field in which a repulsive force field and an attractive force field are combined, and for moving to a target point while tracking an optimal movement path using said potential field.
[0048]
[0049] The first step (S1) is a step of obtaining three-dimensional coordinate values of an obstacle (4). The first step (S1) is a step of sensing an object within a predetermined area from the robot (1) by a sensor unit (10) installed on the robot (1), such as LiDAR, and obtaining coordinate values for said object. According to the present embodiment, as shown in FIGS. 3 and 4, the three-dimensional coordinate values of said obstacle (4) are obtained using an orthogonal coordinate system. In addition, the coordinate values obtained using said orthogonal coordinate system are converted into coordinate values according to a spherical coordinate system. Accordingly, the coordinate values (X, Y, Z) according to the orthogonal coordinate system are converted into coordinate values (ø, ψ, distance) according to the spherical coordinate system. The above ø represents an angle in the horizontal direction from a reference point (corresponding to the longitude angle, hereinafter referred to as the 'horizontal angle'), and ψ represents an angle measured from the z-axis directly above the robot (1), (corresponding to the latitude angle, hereinafter referred to as the 'vertical angle').
[0050] According to the present embodiment, as shown in (1) of FIG. 7, the coordinate values for the obstacle (4) are converted into coordinate values in the spherical coordinate system, and then, as shown in (2) of FIG. 7, data corresponding to a predetermined reference plane (2) forming the floor in the orthogonal coordinate system is removed. Additionally, according to the present embodiment, noise is removed from the data converted to the spherical coordinate system. The noise includes data regarding non-objects, such as points formed by light or reflected light, and removing such data improves the identification accuracy of the obstacle (4).
[0051]
[0052] The second step (S2) above is a step of determining the area of an obstacle (4) by projecting the three-dimensional coordinate values of the obstacle (4) onto a plane. The obstacle area (3) is defined as the area where the obstacle (4) is identified to exist, and according to the present embodiment, it can be derived by projecting onto an area below a reference value. Specifically, as shown in (3) of FIG. 7, the second step (S2) sets a reference value that is spaced a predetermined distance from the robot (1) when projecting the three-dimensional coordinate values of the obstacle (4) onto a plane, and projects the coordinate values included in the range below the reference value from the robot (1) onto a plane. The reference value can be set by the user as a distance that can avoid a collision when the obstacle (4) is observed, taking into account the working radius of the robot (1) and the movement speed of the robot (1). By projecting the coordinate values of the obstacle (4) that exists below the above reference value onto a plane, the actual possibility of the robot (1) colliding with the obstacle (4) is considered, and at the same time, the efficiency of data processing is substantially improved.
[0053] As illustrated in (4) of FIG. 7, the plane is a ø-ψ plane and can be projected into two-dimensional coordinates for a horizontal angle given as ø and a vertical angle given as ψ. Subsequently, as illustrated in (4) of FIG. 7, the area of the obstacle (4) can be defined by setting a rectangle that captures the perimeter of the projected plane. According to (4) of FIG. 7, the obstacle area (3) is labeled as a bounding box. Once the obstacle area (3) is derived, a repulsive field caused by the obstacle (4) can be generated using the bounding box as shown in (6) of FIG. 7. The process of generating the repulsive field and the attractive field will be explained in detail below.
[0054]
[0055] The third step (S3) above is a step of generating a repulsive field by the obstacle area (3) at the first position of the robot (1). The fourth step (S4) above is a step of generating an attractive field at the target point where the robot (1) is to move at the first position of the robot (1). And, the fifth step (S5) is a step of generating a potential field by synthesizing the repulsive field and the attractive field generated at the first position of the robot (1). The repulsive field is a field formed by the obstacle (4), and has a higher value the closer it is to the obstacle (4) and the greater the influence, and the attractive field acts as a guide to induce the robot (1) to the target point.
[0056]
[0057] In the above third step (S3), the repulsive force field is derived by the following mathematical formula 1.
[0058] [Mathematical Formula 1]
[0059]
[0060]
[0061] Here, f k is the repulsive force field caused by the k-th obstacle (4). A k is the height value of the repulsion field caused by the k-th object, and is set so that the repulsion field can sufficiently cover the area around the obstacle (4). A k is determined by the following mathematical formula 2, which is a function of the relative distance between the robot (1) and the object. A k The value increases as the distance between the robot (1) and the object decreases, and is determined by considering the relative speed between the robot (1) and the obstacle (4). When calculating the relative speed between the robot (1) and the obstacle (4), the speed of the end of the robot (1) arm (End of Arm Tool; EOAT) can be used.
[0062]
[0063] [Mathematical Formula 2]
[0064]
[0065] Here, is the relative speed of the k-th obstacle (4), s is the speed sensitivity coefficient, dmax is the distance threshold, d k is the distance to the k-th obstacle (4).
[0066] And, (ø i ,ψ j ) indicates the position of the grid point in the ø-ψ plane. i represents the range of the horizontal angle, and j represents the range of the vertical angle. For example, referring to (1) in FIG. 8, -75 ≤ i ≤ +75 and -15 ≤ j ≤ +15. Of course, the ranges of i and j can be set differently depending on the working environment.
[0067] And, a k , b k is a constant determined by the calculated obstacle area (3), and is defined in proportion to the width and length of the obstacle (4) area. As shown in FIG. 12, α k is the vertical width of the obstacle area (3), and β k is the width of the obstacle area (3) above. At this time, a k =ωα k , b k = ωβ k It can be derived by applying a predetermined proportionality constant ω.
[0068] and, and Each represents the degree to which the center of mass of the obstacle (4) deviates from the center point of the obstacle area (3), and can be defined by the following mathematical formula 3.
[0069]
[0070] [Mathematical Formula 3]
[0071] ,
[0072] However, -1≤ ≤+1, -1≤ ≤+1
[0073]
[0074] Here, is the center of the obstacle area (3) of the k-th recognized obstacle (4), and is the center of mass of the k-th recognized obstacle (4), ø(A) is the ø component of A, and likewise ψ(A) is the ψ component of A. According to the present embodiment, with respect to the obstacle area (3) set as a rectangle, the center of the obstacle area (3) and the center of mass of the projected obstacle (4) are considered to form a symmetric or asymmetric shape. As shown in FIG. 7 (6) and FIG. 13, when the center of the obstacle area (3) and the center of mass of the projected obstacle (4) coincide, the repulsion force field is formed symmetrically, whereas when the center of the obstacle area (3) and the center of mass of the projected obstacle (4) do not coincide, the repulsion force field is formed asymmetrically.
[0075] Specifically, the change in the repulsive force field generated in the third step (S3) is as shown in FIG. 10. For example, FIG. 10(a) is a k =1, b k This is the case where =1, and Fig. 10(b) is a k =3, b k This is the case where =1, and Fig. 10(c) is a k =1, b k This is the case where =3. a k As increases, it becomes flatter in the ø-axis direction, and b k As it increases, it becomes gentler in the ψ-axis direction. That is, the repulsive force field is affected by the width and height of the obstacle area (3).
[0076] In addition, as shown in the upper view of FIG. 11(a), As this approaches +1, the repulsive force field changes such that the positive part of the ø-axis rises and the negative part of the ø-axis lowers, as shown in the lower view of FIG. 11(a). As this approaches -1, the repulsive force field changes such that the negative part of the ø-axis rises and the positive part of the ø-axis lowers. Likewise, as shown in the upper view of FIG. 11(b), As this approaches +1, the repulsive force field changes such that the positive part of the ψ-axis rises and the negative part of the ψ-axis lowers, as shown in the lower view of FIG. 11(b), As this approaches -1, the repulsive force field changes such that the negative part of the ψ-axis rises and the positive part of the ψ-axis falls.
[0077]
[0078] In the above 4th step (S4), the attractive field (fatt(ø) i ,ψ j )) is derived by the following mathematical formula 4. As shown in the formula below, the above attractive field is (ø i ,ψ j It can be defined as the magnitude of the attractive force at the lattice points of ). Below, ø goal and ψ goal Each is defined as the horizontal and vertical angles of the target point.
[0079] [Mathematical Formula 4]
[0080]
[0081]
[0082] The above fifth step (S5) is a step of generating a potential field by synthesizing the reaction field and the attraction field at the first position of the robot (1), as illustrated in FIG. 8. Subsequently, the position of the minimum value is calculated from the potential field. The position corresponding to the minimum value is determined as the second position. As illustrated in FIG. 8 and FIG. 9, the minimum value is an example calculated as (ø,ψ)=(-8,-5), and the robot (1) is shown moving from the initial first position (e.g., origin (0,0)) to a point with a radius of 0.1m, a horizontal angle of -8°, and a vertical angle of -5°.
[0083]
[0084] The above 6th step (S6) and the above 7th step (S7) relate to the process of the robot (1) moving from a first position to a final target position by the continuous generation of a potential field. The above 6th step (S6) is a step of moving the robot (1) to the above 2nd position corresponding to the minimum value of the potential field. The above 7th step (S7) is a step of moving the robot (1) to the above target point by repeating the above 1st step (S1) to the above 6th step (S6) at predetermined time intervals until the above 2nd position coincides with the above target point. At this time, as shown in FIG. 4, the coordinate value of the second position to which the robot (1) is to be moved may be a value converted back to an orthogonal coordinate system, and the control unit (40) may move the robot (1) to the above 2nd position based on the orthogonal coordinate system.
[0085] In step 6 (S6), the robot (1) moves from the first position to the second position. Subsequently, when the robot (1) reaches the second position, the repulsion field, attraction field, and potential field are regenerated at the second position. The second position is treated as the first position again, and the position corresponding to the minimum value in the newly generated potential field is set as the new second position. By repeating this process, when the second position finally reaches the target point, the robot (1) is able to move along the optimal path to the target point while avoiding collision with the obstacle (4). The predetermined time interval can be set, for example, to 200 milliseconds (ms). Of course, the predetermined time interval can be set differently depending on the actual operating environment of the robot (1). By repeating the potential field for a very short period of time in this way and tracking the position corresponding to the minimum value of the potential field to move the robot (1), the effect of the robot (1) moving along the optimal path while substantially continuously preventing collision with the obstacle (4) is provided.
[0086] In addition, as illustrated in FIG. 6, the robot collision prevention method according to the embodiment of the present invention can be effectively applied not only when the obstacle (4) is stationary as in FIG. 6(a), but also when the obstacle (4) is moving as in FIG. 6(b). FIG. 6 illustrates a case where a robot (1) moves from the origin (O) to point A to perform a task, and then moves to point B to perform a task. Even if a static obstacle (4) (Fig. 6(a)) or a dynamic obstacle (4) (Fig. 6(b)) appears during the process of moving from point A to point B, the present invention generates an optimal path by reflecting a real-time potential field, so it can be effectively applied in both cases of FIG. 6(a) and FIG. 6(b).
[0087] Meanwhile, according to the present embodiment, in the process of the robot (1) moving to a target point, the robot (1) may exceptionally maintain a stationary state if it cannot reach a second position due to the obstacle (4) when following the movement mechanism. For example, if left-right bending motion of the robot (1) is impossible as in FIG. 5(a) and up-down bending motion of the robot (1) is possible as in FIG. 5(b), and if the robot (1) is controlled to proceed to the target point by the continuous generation of the potential field when the obstacle (4) is placed in a line between the robot (1) and the target point and access is possible by up-down bending motion, the robot (1) may be controlled to maintain a stationary state.
[0088]
[0089] Meanwhile, according to another aspect of the present invention, a robot collision prevention device is proposed. The robot collision prevention device according to the present embodiment includes a sensor unit (10), an obstacle area determination unit (20), a field generation unit (30), and a control unit (40). The robot collision prevention device according to the present embodiment can be implemented substantially by applying the robot collision prevention method described above.
[0090] The sensor unit (10) obtains three-dimensional coordinate values of an obstacle (4) using an orthogonal coordinate system. According to the present embodiment, the sensor unit (10) may be a LiDAR and is installed at the end of the robot (1) arm to sense the obstacle (4) from the viewpoint of the robot (1) arm. The position of the obstacle (4) is obtained by the sensor unit (10) using an orthogonal coordinate system, thereby substantially performing the first step (S1) described above.
[0091] FIG. 14 illustrates an example in which an obstacle (4) is recognized through a sensor unit (10) installed at the end of the robot (1) arm. As shown in FIG. 14, a radia is installed at the end of the robot (1) arm, and the red dot indicates the position of the radia. The viewpoint facing the radia is set to 0°, the opposite direction is 180°, the angle of rotation clockwise is set to the (-) direction, and the angle of rotation counterclockwise is set to the (+) direction. The case in which the rotation range in the clockwise and counterclockwise directions is -135° to +135° is illustrated.
[0092] The obstacle area determination unit (20) is provided to determine the area of an obstacle (4) by converting the coordinate values obtained by the sensor unit (10) into coordinate values corresponding to a spherical coordinate system and projecting the coordinate values in the spherical coordinate system onto a plane for horizontal and vertical angles. According to the present embodiment, the obstacle area determination unit (20) removes data corresponding to a predetermined reference plane (2) forming the floor in the orthogonal coordinate system after the three-dimensional coordinate values are converted into coordinate values in the spherical coordinate system. In addition, noise is removed from the data converted to the spherical coordinate system. The noise may include data regarding non-objects, such as points formed by light or reflected light. That is, the obstacle area determination unit (20) can perform a data preprocessing process that converts the coordinates of the data obtained by the sensor unit (10), removes data corresponding to the reference plane (2), and removes noise.
[0093] The obstacle area determination unit (20) substantially performs the process of the second step (S2) described above. The obstacle area determination unit (20) sets a reference value spaced apart from the robot (1) at a predetermined distance and projects coordinate values included in a range below the reference value from the robot (1) onto a plane, and at this time, the obstacle area (3) is set as a rectangle that captures the perimeter of the projected plane. For example, FIG. 15 shows the process of determining the obstacle area (3) by labeling the projected data in the state of FIG. 14. Considering the characteristics of a rotational radia, for example, if an object is measured at a horizontal angle ø of 3˚ and a vertical angle ψ of -16.5˚ to -13.5˚, a binary image is generated on the ø-ψ plane by placing 1 in the cell (ø,ψ)=(3,-15) on the graph of FIG. 15 (processing as if data exists). Using computer vision techniques, labeling can be performed for each obstacle (4) (e.g., by 8-directional labeling), and the center of the obstacle area (3) and the center of mass of the obstacle (4) can be calculated using information about each obstacle (4). In FIG. 15, the box surrounding the white area with a rectangle is the obstacle area (3) (in FIG. 15, obstacle 3 is observed, and the rectangular box for the obstacle in the center is the obstacle area (3)), the circle is the center of the obstacle area (3), and the point inside the obstacle area (3) is the center of mass of the obstacle (4).
[0094] The field generation unit (30) is provided to generate a potential field by combining a repulsive field caused by the obstacle area (3) at the first position of the robot (1), an attractive field at the target point where the robot (1) is to move at the first position of the robot (1), and the repulsive field and the attractive field at the first position of the robot (1). Since the field generation unit (30) substantially performs the third step (S3) to the fifth step (S5), a repetitive description is omitted. Furthermore, since the field generation unit (30) performs the process from the third step (S3) to the fifth step (S5), it can retain the operation or effect of the third step (S3) to the fifth step (S5). For example, the field generating unit (30) can form a repulsive force field using the mathematical formulas 1 to 3, and for the obstacle area (3) set as a rectangle, the repulsive force field is generated symmetrically or asymmetrically by considering the center of the obstacle area (3) and the center of mass of the projected obstacle (4). In addition, the field generating unit (30) forms an attractive field using the mathematical formula 4, and generates a potential field by combining the repulsive force field and the attractive field.
[0095] The control unit (40) controls the robot (1) to move along an optimal path from a first position to a target point. First, the robot (1) is moved from the first position to a second position corresponding to the minimum value of the potential field. When the robot (1) moves to the second position, the field generation unit (30) creates a new reaction force field, a force field, and a potential field by making the second position the new first position. Based on the second position to which the robot (1) has moved, a potential field is created again to continuously move the robot (1) to a position with a predetermined minimum value. Consequently, the control unit (40) controls the robot (1) to reach the target point by tracking the minimum value of the potential field while repeatedly generating the potential field at predetermined time intervals until the second position coincides with the target point. The predetermined time interval can be applied in the same way as the robot collision prevention device described above, and can be set in various ways depending on the working environment of the robot (1).
[0096]
[0097] Thus, the robot collision prevention method and device according to the embodiment of the present invention provide the effect of improving work speed while reducing accident occurrence by enabling an autonomous robot (1) used in an industrial site to move along a path with a low probability of collision with an obstacle (4) in real time.
[0098]
[0099] Although the present invention has been described in detail with respect to preferred embodiments, the present invention is not limited to the above embodiments, and many variations may be provided within the scope of the present invention.
Claims
1. Step 1 (S1) of obtaining 3D coordinate values of an obstacle; A second step (S2) of determining the area of an obstacle by projecting the 3D coordinate values of the obstacle onto a plane; A third step (S3) of generating a repulsive field by the obstacle area at the first position of the robot; A fourth step (S4) of generating an attractive field by the target point where the robot will move from the first position of the robot; A fifth step (S5) of generating a potential field by synthesizing the repulsive force field and the attractive force field at the first position of the robot; Step 6 (S6) of moving the robot to a second position corresponding to the minimum value of the potential field; and A robot collision prevention method characterized by including: a seventh step (S7) in which, when the robot moves to the second position, the second position is treated as a new first position and the first to sixth steps (S1) are repeated at predetermined time intervals until the second position matches the target point, thereby moving the robot to the target point.
2. In Paragraph 1, A robot collision prevention method characterized in that the step of obtaining three-dimensional coordinate values of the obstacle is obtained using an orthogonal coordinate system with respect to the obstacle, and the coordinate values obtained by the orthogonal coordinate system are converted into coordinate values obtained by a spherical coordinate system.
3. In Paragraph 2, A robot collision prevention method characterized by removing data corresponding to a predetermined reference plane forming a floor in the Cartesian coordinate system after the above 3D coordinate values are converted into coordinate values based on the above spherical coordinate system.
4. In Paragraph 1, A robot collision prevention method characterized by setting a reference value spaced at a predetermined distance from the robot when projecting three-dimensional coordinate values of the obstacle onto a plane, and projecting coordinate values included in a range less than or equal to the reference value from the robot onto a plane.
5. In Paragraph 1, A robot collision prevention method characterized in that the area of the obstacle is set as a rectangle that captures the perimeter of the projected plane.
6. In Paragraph 5, A robot collision prevention method characterized by the fact that the repulsive force field is calculated symmetrically or asymmetrically with respect to the obstacle area set as a rectangle, by considering the center of the obstacle area and the center of mass of the projected obstacle.
7. A sensor unit that acquires the 3D coordinate values of an obstacle using an orthogonal coordinate system; An obstacle area determination unit that converts the coordinate values obtained by the sensor unit into coordinate values corresponding to a spherical coordinate system, and orthographically projects the coordinate values based on the spherical coordinate system onto a plane for horizontal and vertical angles to determine the area of an obstacle; A field generating unit that generates a potential field by synthesizing the repulsive field and the attractive field at the first position of the robot, the repulsive field by the obstacle area at the first position of the robot, the attractive field by the target point to which the robot is to move at the first position of the robot, and the repulsive field and the attractive field at the first position of the robot; and A robot collision prevention device characterized by comprising: a control unit that moves the robot from a first position to a second position corresponding to the minimum value of the potential field, and when the robot moves to the second position, the field generation unit creates a new reaction force field, an attractive force field, and a potential field by making the second position a new first position and repeatedly generates the potential field at predetermined time intervals until the second position coincides with the target point, thereby controlling the robot to track the minimum value of the potential field and reach the target point.
8. In Paragraph 7, A robot collision prevention device characterized by the above obstacle area determination unit removing data corresponding to a predetermined reference plane forming a floor in the above orthogonal coordinate system after the above 3D coordinate values are converted into coordinate values based on the above spherical coordinate system.
9. In Paragraph 1, A robot collision prevention device characterized by the obstacle area determination unit setting a reference value spaced apart from the robot at a predetermined distance and projecting coordinate values included in a range less than or equal to the reference value onto a plane from the robot.
10. In Paragraph 1, A robot collision avoidance device characterized in that the obstacle area is set as a rectangle that captures the perimeter of the projected plane.
11. In Paragraph 9, A robot collision prevention device characterized by the field generation unit generating the repulsive force field symmetrically or asymmetrically for the obstacle area set as the rectangle, by considering the center of the obstacle area and the center of mass of the projected obstacle.