Robot control device

Abstract

To provide a robot control device capable of identifying a weld wire with higher accuracy.SOLUTION: A robot control device 1 that identifies a weld wire by controlling a manipulator 2 equipped with a sensor 3 which performs sensing by irradiating a welding target 5 with light, includes: an acquisition unit 11 for acquiring a three-dimentional shape of the welding target 5 by using a sensing result of the sensor 3; a control unit 12 for controlling the manipulator 2 and the sensor 3 so as to perform first sensing on the entire welding target 5 and second sensing in which light is applied from a normal direction of each surface of the welding target 5 based on the three-dimentional shape of the welding target 5 acquired by the first sensing: and an identification unit 13 for identifying a weld wire by using the three-dimentional shape of the welding target 5 acquired by the second sensing. In this manner, more accurate identification of a weld wire can be achieved by using a result of the second sensing with an increased amount of received light in primary reflection and reduced influence of secondary reflection.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a robot control device that acquires a three-dimensional shape of an object to be welded and identifies a weld line. [Background technology]

[0002] Conventionally, in order to improve the efficiency of robot teaching work, a slit light was projected onto the workpiece, and the bright line image of the slit light was captured by a camera to perform three-dimensional measurement of the work line and generate the robot's movement path (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2002-172575 Summary of the Invention [Problem to be solved by the invention]

[0004] When measuring the shape of a welded workpiece by projecting light onto it, the measurement is performed using the primarily reflected light, but if the welded workpiece is irradiated at an angle, sufficient reflected light cannot be obtained, resulting in a problem of poor measurement accuracy. Furthermore, irradiating the workpiece from an oblique angle can cause secondary reflections, and when measuring the interior angles of structures, etc., receiving the secondary reflected light can worsen measurement accuracy, resulting in the problem of being unable to properly identify the weld seam.

[0005] The present invention has been made to solve the above-mentioned problems, and aims to provide a robot control device that can improve measurement accuracy by obtaining sufficient primary reflected light for a sensor and identify a weld line by reducing the influence of secondary reflection of light from the welding object. [Means for solving the problem]

[0006] In order to achieve the above-mentioned object, one aspect of the present invention provides a robot control device that identifies a weld line on an object to be welded by controlling a manipulator equipped with a sensor that performs sensing by irradiating light onto the object to be welded, and includes: an acquisition unit that acquires the three-dimensional shape of the object to be welded using the sensing results of the sensor; a control unit that controls the manipulator and the sensor to perform first sensing of the entire object to be welded and second sensing in which light is irradiated from the normal direction of each face that defines an intersection line on the object to be welded based on the three-dimensional shape of the object to be welded acquired by the first sensing; and an identification unit that identifies the weld line using the three-dimensional shape of the object to be welded acquired by the second sensing. [Effects of the Invention]

[0007] According to a robot control device of one embodiment of the present invention, after performing a first sensing, a second sensing is performed that can receive sufficient primary reflected light and reduces the influence of secondary reflection, thereby making it possible to obtain the three-dimensional shape of the welding object with high accuracy, and as a result, to identify the welding line with even higher accuracy. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a schematic diagram showing the configuration of a welding robot system according to an embodiment of the present invention. [Figure 2] A flowchart showing the operation of the robot control device according to the embodiment. [Figure 3A] FIG. 10 is a diagram for explaining first sensing in the embodiment. [Figure 3B] FIG. 10 is a diagram for explaining second sensing in the embodiment. [Figure 3C] FIG. 10 is a diagram for explaining second sensing in the embodiment. [Figure 4A] FIG. 10 is a diagram showing another example of a welding target in the embodiment; [Figure 4B] FIG. 10 is a diagram showing another example of a welding target in the embodiment; [Figure 5] FIG. 10 is a diagram showing another example of a welding target in the embodiment; [Figure 6A] FIG. 10 is a diagram showing an example of a highly accurate sensing result in the embodiment. [Figure 6B] FIG. 10 is a diagram showing an example of a low-accuracy sensing result in the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, a robot control device according to the present invention will be described using embodiments. Note that in the following embodiments, components and steps with the same reference numerals are the same or equivalent, and repeated description may be omitted. The robot control device according to this embodiment performs first sensing of the entire object to be welded, and then performs second sensing in which light is irradiated from the normal direction to each surface that defines the intersection line of the object to be welded. This makes it possible to increase the amount of light received from primary reflections in the second sensing and reduce the influence of secondary reflections, thereby making it possible to obtain the three-dimensional shape of the object to be welded with higher accuracy.

[0010] 1 is a schematic diagram showing the configuration of a welding robot system 100 according to this embodiment. The welding robot system 100 according to this embodiment includes a robot control device 1, a manipulator 2 equipped with a welding torch 2a, a sensor 3 that senses an object to be welded (workpiece) 5, and a welding power source 4.

[0011] The robot control device 1 controls the manipulator 2 and the sensor 3, and identifies a weld line on the welding object 5 using the sensing results of the sensor 3 regarding the welding object 5. The robot control device 1 may also control the manipulator 2 and the welding power source 4, for example, so that welding is performed along the identified weld line. Each component of the robot control device 1 will be described later.

[0012] The manipulator 2 has a plurality of arms connected by joints driven by motors. A welding torch 2a is attached to the tip of the manipulator 2. If a welding wire is used for welding, the manipulator 2 may be attached with a wire feeder for feeding the welding wire. The manipulator 2 is also attached with a sensor 3. The attachment position of the sensor 3 on the manipulator 2 is not particularly limited, but it is preferable that the sensor 3 be attached to a position where the welding object 5 can be sensed by the sensor 3 from a desired direction. Therefore, the sensor 3 may be attached, for example, to the tip side of the manipulator 2. The tip side of the manipulator 2 may be, for example, a position on the base end side of the welding torch 2a. The manipulator 2 is not particularly limited, but may be, for example, a vertical articulated robot.

[0013] The sensor 3 performs sensing by irradiating the welding object 5 with light. Typically, the sensing object is composed of multiple welding objects 5. The welding object 5 may be, for example, a metal plate. In this embodiment, a case will be mainly described in which the welding object 5 is a T-joint and a weld line for fillet welding is identified in the T-joint. Needless to say, as will be described later, the welding object 5 is not limited to a T-joint. Furthermore, the light irradiated by the sensor 3 to the welding object 5 is not particularly limited, and may be, for example, visible light, infrared light, ultraviolet light, or the like. Furthermore, the type of the sensor 3 is not limited. The sensor 3 may perform sensing using, for example, a light projection method, a ToF (Time of Flight) method, or a method combining pattern projection and a stereo camera method. While each sensing method is already known, it will be briefly described below.

[0014] The light projection method is a sensing method in which spot light, slit light (linear light), or a predetermined planar pattern (for example, a random dot pattern) is projected onto the object to be measured, and the reflection from the object is observed from a different direction. When projecting spot light or slit light, the three-dimensional shape is usually obtained by scanning it in two or one dimensions.

[0015] The ToF method is a sensing method that calculates the distance to a measurement target based on the delay time between emitted light and reflected light. For example, it is used in LiDAR (Light Detection and Ranging). Methods for measuring the round-trip time of light include, for example, a method that measures the time from when a light pulse is emitted to when reflected light is received, and a method that measures the round-trip time of light based on the phase difference between emitted light and reflected light.

[0016] The method that combines pattern projection and stereo camera techniques is a sensing method in which a planar pattern with a two-dimensional extent is projected onto the object to be measured and the object is then photographed with a stereo camera. Even for objects with featureless surfaces, such as flat plates, the shape can be acquired by projecting a planar pattern. This planar pattern may be, for example, a dot pattern in which each dot has a random shape.

[0017] The sensing result by the sensor 3 is not particularly limited, but may be, for example, three-dimensional point cloud data. In this embodiment, the case where the sensing result is three-dimensional point cloud data will be mainly described.

[0018] Welding power source 4 supplies a high voltage used in welding to welding torch 2a and welding object 5. When a welding wire is used for welding, welding power source 4 may also control the feeding of the welding wire. Note that the configuration of welding power source 4 is already known, and a detailed description thereof will be omitted.

[0019] 1, the robot control device 1 includes an acquisition unit 11, a control unit 12, and an identification unit 13. The robot control device 1 is also connected to a manipulator 2, a sensor 3, and a welding power source 4 so as to be able to exchange information with them.

[0020] The acquisition unit 11 acquires the three-dimensional shape of the welding object 5 using the sensing results of the sensor 3. The three-dimensional shape of the welding object 5 acquired using the sensing results is usually a shape in the coordinate system of the sensor 3. However, the acquisition unit 11 can convert that shape into a shape in the coordinate system of the manipulator 2 by using the angles of each axis of the manipulator 2. In this embodiment, a case where the acquisition unit 11 acquires the three-dimensional shape of the welding object 5 in the coordinate system of the manipulator 2 in this manner will be mainly described. Note that acquisition of the three-dimensional shape may also mean, for example, acquiring information indicating a plane or a curved surface in a three-dimensional space (e.g., an equation indicating a plane or an equation indicating a curved surface) based on three-dimensional point cloud data. For example, if the welding object 5 is a T-joint, the acquisition unit 11 may acquire information indicating the planes of two flat plates that constitute the T-joint. Note that if the shape of the welding object 5 (e.g., a planar shape, a cylindrical shape, etc.) is known in advance, the acquisition unit 11 may acquire the three-dimensional shape using the known shape.

[0021] The control unit 12 controls the manipulator 2 and the sensor 3 so that first sensing is performed on the entire welding object 5. The first sensing on the entire welding object 5 may be, for example, sensing that can acquire the shape of each of multiple surfaces that define the welding line. More specifically, when the welding line is defined by a surface extending horizontally and a surface extending vertically, the first sensing may be performed by irradiating light toward the welding object 5 from a direction that is angled with respect to the vertical direction. The angle between the vertical direction and the light irradiation direction may be, for example, within a range of 30 to 60 degrees, or may be 45 degrees. Furthermore, in the first sensing, the entire welding object 5 may be included in the sensing range. Furthermore, the sensor 3 may be located in a single position when the first sensing is performed.

[0022] Furthermore, the control unit 12 controls the manipulator 2 and the sensor 3 to perform second sensing, based on the three-dimensional shape of the welding object 5 acquired by the acquisition unit 11 through the first sensing, such that light is irradiated from the normal direction of each face defining the intersection line in the welding object 5. That is, each face defining the intersection line in the welding object 5 is identified based on the results of the first sensing, and the manipulator 2 and the sensor 3 are controlled so that sensing is performed from the normal direction of each identified face in the second sensing. Note that the intersection line in the welding object 5 may be, for example, a line corresponding to a weld line. Furthermore, the face defining the intersection line may be any face as long as the position of the intersection line can be identified as a result. For example, the face defining the intersection line may be a face forming the intersection line. The face forming the intersection line may be, for example, a face continuous with the intersection line. Identification of each face may be performed by identifying a predetermined face, such as a horizontal face or a vertical face, or by identifying the intersection line and then identifying each face continuous with the intersection line. The target to be sensed in the second sensing may be one surface or multiple surfaces. The surface may or may not be a plane. In this embodiment, a case where the target surface to be sensed in the second sensing is multiple planes will be mainly described. A case where the target surface is not a plane will be described later with reference to FIG. 5. Note that, for example, the control unit 12 may use the three-dimensional shape acquired in the first sensing to identify each surface defining the intersection line in the welding target 5 and identify the position of the sensor 3 for irradiating each surface with light from the normal direction. Note that if the relative positional relationship between each surface of the welding target 5 and the position of the sensor 3 when irradiating each surface with light from the normal direction is predetermined, the control unit 12 may identify each position of the sensor 3 in the second sensing using the three-dimensional shape of the welding target 5 acquired in the first sensing and the relative positional relationship. Then, the control unit 12 may control the manipulator 2 and the sensor 3 so that sensing is performed from the identified position of the sensor 3 in the second sensing.

[0023] In sensing, when light is irradiated so as to scan a predetermined area or when light is irradiated within a predetermined range, there are multiple directions in which the irradiated light is directed. In this case, performing sensing in the second sensing such that light is irradiated from the normal direction of the surfaces of the objects to be welded may mean, for example, performing sensing in the second sensing such that at least a portion of the irradiated light is irradiated from the normal direction of the surfaces. In this case, for example, the central light of the light irradiated from sensor 3 may be light irradiated from the normal direction of the surfaces of the objects to be welded.

[0024] Furthermore, in each sensing performed in the second sensing, for example, sensing may be performed so that the intersection line of the plurality of welding objects 5 is included in the sensing range, or it does not have to be so.

[0025] When the welding object 5 is a T-joint, for example, there are welding objects 5 that extend vertically and welding objects 5 that extend horizontally, so in the second sensing, two sensing operations are performed for one weld line: sensing from the normal direction of the welding object 5 that extends vertically, and sensing from the normal direction of the welding object 5 that extends horizontally. Also, when the welding object 5 is not flat, such as a cylindrical shape, sensing may be performed from multiple normal directions for the surface of one welding object (for example, the outer peripheral surface of a cylindrical shape).

[0026] Furthermore, the control unit 12 may control the first sensing so that sensing is performed multiple times without changing the position of the sensor 3. Furthermore, the control unit 12 may control the second sensing so that sensing is performed multiple times for one surface without changing the position of the sensor 3. When multiple sensing is performed, for example, multiple measurement results can be obtained for each point of the three-dimensional point cloud data, and the three-dimensional shape may be acquired using a representative point of the multiple measurement results for each point. The representative point of the multiple points may be, for example, the center of gravity of the multiple points.

[0027] Furthermore, after the weld line is identified by the identification unit 13, the control unit 12 may control the manipulator 2 so that welding is performed by the welding torch 2a along the identified weld line. During this welding, the control unit 12 may also control the welding power source 4.

[0028] The identification unit 13 identifies the weld line using the three-dimensional shape of the welding object 5 acquired by the second sensing. The three-dimensional shape used to identify the weld line may be, for example, an integration of multiple three-dimensional shapes acquired by the acquisition unit 11 using multiple sensing results in the second sensing. This integration may be performed by, for example, the acquisition unit 11. The acquisition unit 11 may acquire a single three-dimensional shape by, for example, combining multiple three-dimensional shapes represented in the coordinate system of the manipulator 2. The weld line may be, for example, the intersection line between the surfaces of the two welding objects 5, or may be a line obtained by offsetting the intersection line in a predetermined direction. The weld line may be, for example, a straight line or a curved line. Furthermore, for example, the entire identified intersection line may be the weld line, or a range of the intersection line excluding a predetermined length (for example, 1 centimeter or 2 centimeters) from both ends may be the weld line. The same applies when the weld line is offset from the intersection line. In this way, the identification unit 13 may identify the weld line by, for example, identifying the intersection line in the three-dimensional shape of the welding object 5 according to the second sensing result. Furthermore, if the weld line is offset from the intersection line, the identification unit 13 may identify, for example, a weld line that has a predetermined relationship with the intersection line. Furthermore, the identification of the weld line may be, for example, identifying multiple points along the weld line. The multiple points may be, for example, teaching points. When the identification of the weld line is identification of teaching data including teaching points, the identification unit 13 may, for example, determine one end of the weld line as the welding start point and the other end as the welding end point. Furthermore, the teaching data may also include, for example, the advance / retract angle and the target angle of the welding torch 2a. The advance / retract angle and the target angle may be, for example, predetermined angles.

[0029] Next, the operation of the robot control device 1 will be described with reference to the flowchart of FIG. (Step S101) The control unit 12 controls the manipulator 2 so that the sensor 3 is at a first sensing position. By this control, the sensor 3 is moved to a position where the first sensing is performed. The position of the sensor 3 may be determined in advance, for example.

[0030] (Step S102) The control unit 12 causes the sensor 3 to perform first sensing. That is, light is emitted from the sensor 3, and sensing is performed using the light. The result of the first sensing is passed from the sensor 3 to the acquisition unit 11.

[0031] (Step S103) Acquisition unit 11 acquires the three-dimensional shape of object 5 to be welded using the result of the first sensing.

[0032] (Step S104) The control unit 12 controls the manipulator 2 so that the light emitted from the sensor 3 is in the normal direction of the plane included in the three-dimensional shape acquired in step S103. By this control, the sensor 3 is moved to a position where the second sensing is performed.

[0033] (Step S105) The control unit 12 causes the sensor 3 to perform sensing. That is, light is emitted from the sensor 3, and sensing is performed using the light. The result of the sensing is passed from the sensor 3 to the acquisition unit 11.

[0034] (Step S106) The acquisition unit 11 acquires the three-dimensional shape of the object to be welded 5 using the result of the sensing in step S105.

[0035] (Step S107) The control unit 12 determines whether to continue the second sensing, i.e., whether to repeat the sensing. For example, if the three-dimensional shape acquired using the first sensing result includes two or more surfaces and there is a surface among the two or more surfaces on which sensing has not been performed, the control unit 12 may determine to repeat the sensing. If the sensing is to be repeated, the control unit 12 returns to step S104, and if the sensing is not to be repeated, the control unit 12 proceeds to step S108. Note that if the control unit 12 returns to step S104, sensing may be performed on surfaces on which sensing has not yet been performed. If the sensing of step S105 is performed multiple times, the second sensing is performed by the multiple sensing operations.

[0036] (Step S108) The identification unit 13 identifies a weld line using the three-dimensional shape of the welding object 5 acquired by the acquisition unit 11 using the second sensing result.

[0037] (Step S109) Control unit 12 controls manipulator 2 and welding power source 4 so that welding is performed along the weld line identified in step S108. Then, when this welding is completed, the series of processes for identifying the weld line and welding using the identification results is completed. In the case of the T-joint shown in FIG. 1, after identifying the weld line on one side (the right side in the figure) and completing fillet welding along that weld line, the weld line on the other side (the left side in the figure) may be identified and fillet welding may be performed along that weld line. The order of the processes in the flowchart of FIG. 2 is an example, and the order of the steps may be changed as long as the same results are obtained.

[0038] Next, the operation of the robot control device 1 according to this embodiment will be described using a specific example. In this specific example, as shown in Figure 1, the welding object 5 is a T-joint, and we will explain the case where a weld line is identified and fillet welding is performed at the joint portion of the plate-shaped workpiece on one side (the right side in the figure).

[0039] First, the control unit 12 controls the manipulator 2 so that the light emitted from the sensor 3 is angled at approximately 45 degrees with respect to the vertical direction (step S101). This control may, for example, be control to move the sensor 3 to a predetermined position and direction. Next, the control unit 12 causes the sensor 3 to perform a first sensing (step S102). In this case, the sensor 3 may irradiate the welding object 5 with light, as indicated by the solid arrow in FIG. 3A. Note that if the sensor 3 receives only the light reflected by the primary reflection indicated by the dashed arrow 21, high-precision sensing is achieved. However, in reality, the sensor 3 also receives the light reflected by the secondary reflection indicated by the dashed arrow 22. Furthermore, if the light is irradiated from an oblique angle, the amount of light received by the primary reflection also decreases. This decrease in the amount of light received by the primary reflection and the influence of the secondary reflection reduce the accuracy of the 3D point cloud data acquired by the sensor 3.

[0040] Acquisition unit 11 acquires the three-dimensional shape using the three-dimensional point cloud data acquired by the first sensing, and passes it to control unit 12 (step S103). Upon receiving the three-dimensional shape, control unit 12 identifies each plane of object-to-be-welded 5, i.e., horizontal planes and vertical planes.

[0041] Next, in the second sensing, the control unit 12 first causes the sensor 3 to perform sensing of a horizontal plane. Specifically, as shown in FIG. 3B, the control unit 12 moves the sensor 3 so that light is emitted vertically downward from the sensor 3 located above the horizontal welding object 5 (step S104). Then, the control unit 12 causes the sensor 3 to perform sensing. That is, light is emitted from the sensor 3 located in the position shown in FIG. 3B as indicated by the solid arrow, and the reflected light of the primary reflection indicated by the dashed arrow is received by the sensor 3 (step S105). In this case, since light is irradiated from the normal direction to the plane of the welding object 5, the amount of received light of the primary reflection increases, reducing the possibility of secondary reflection, and enabling the acquisition of more accurate three-dimensional point cloud data. The acquisition unit 11 acquires the three-dimensional shape of the plane of the welding object 5 extending horizontally using the three-dimensional point cloud data acquired by the sensor 3 (step S106).

[0042] Next, the control unit 12 determines that sensing should be repeated because there is another plane to be sensed (step S107) and causes the sensor 3 to perform sensing on the vertical plane. Specifically, as shown in FIG. 3C, the control unit 12 moves the sensor 3 so that light is emitted toward the welding object 5 from the sensor 3 located on the right side of the vertical welding object 5 in the drawing (step S104). Then, the control unit 12 causes the sensor 3 to perform sensing. That is, light is emitted from the sensor 3 located in the position shown in FIG. 3C as indicated by the solid arrow, and the reflected light of the primary reflection as indicated by the dashed arrow is received by the sensor 3 (step S105). In this case, too, the amount of light received from the primary reflection increases, reducing the possibility of secondary reflection, and enabling more accurate 3D point cloud data to be acquired. The acquisition unit 11 acquires the 3D shape of the plane of the welding object 5 extending in the vertical direction using the 3D point cloud data acquired by the sensor 3 (step S106). In this specific example, the case where sensing of the horizontal plane is performed followed by sensing of the vertical plane is described, but the sensing order may be reversed.

[0043] When the second sensing is thus completed (step S107), the acquisition unit 11 integrates the two three-dimensional shapes acquired according to the second sensing results to acquire a highly accurate three-dimensional shape of the welding object 5. Then, the identification unit 13 identifies the intersection line of the two planes in the three-dimensional shape to identify the weld line 6 shown in FIG. 3C (step S108). Information indicating the identified weld line 6 is passed to the control unit 12.

[0044] Upon receiving the information indicating the weld line 6, the control unit 12 controls the manipulator 2 and the welding power source 4 to perform welding along the identified weld line (step S109). In this manner, the weld line is identified and welding is performed along the weld line.

[0045] As described above, according to the robot control device 1 of this embodiment, after performing the first sensing, the amount of received light from the primary reflection is increased and the influence of the secondary reflection is reduced. This allows the three-dimensional shape of the welding object 5 to be acquired with high accuracy, and as a result, the weld line can be identified with high accuracy. Therefore, more accurate welding can be achieved. Furthermore, the control unit 12 also controls the welding along the identified weld line, so that automatic welding can be performed.

[0046] Next, a modified example of the robot control device 1 according to this embodiment will be described. For example, if the welding object 5 is a lap joint as shown in FIG. 4A, the control unit 12 may control the manipulator 2 and the sensor 3 to perform second sensing in which light is irradiated in the direction indicated by the solid arrow in the figure. This second sensing may identify, for example, the weld line 6. In the case of a lap joint, if the edge of the upper surface of the upper plate and the upper surface of the lower plate can be identified, the intersection line between the upper plate and the lower plate can be identified. Therefore, in this case, the surfaces defining the intersection line may be, for example, the upper surface of the upper plate and the upper surface of the lower plate. In addition, as in the case of a T-joint, second sensing may be performed in which light is irradiated leftward in FIG. 4A. In other words, second sensing may be performed in which light is irradiated from the normal direction to each surface constituting the intersection line between the upper plate and the lower plate.

[0047] Furthermore, for example, when the welding object 5 is a butt joint shown in FIG. 4B , the control unit 12 may control the manipulator 2 and the sensor 3 to perform second sensing in which light is irradiated in the direction indicated by the solid arrow in the figure. This second sensing may identify, for example, the weld line 6. Note that while FIG. 4B shows a V-shaped groove, the same applies to butt joints with grooves of I-, X-, U-, H-, L-, K-, J-, etc. For example, when the groove shape of the butt joint is V- or X-shaped and the groove is large, second sensing may be performed in which light is irradiated from the normal direction of each face constituting the V- or X-shaped groove (for example, each face at a 45-degree angle with respect to the horizontal direction in FIG. 4B ).

[0048] Furthermore, for example, when welding a cylindrical welding object 5 and a flat welding object 5 shown in FIG. 5 , the second sensing may include sensing in which light is irradiated in the normal direction of the flat welding object 5, i.e., in the direction of the downward arrow in the figure, and multiple sensing operations (e.g., three or four sensing operations) in which light is irradiated in the normal direction of the outer circumferential surface of the cylindrical welding object 5, i.e., in the direction of the horizontal or diagonal arrow in the figure. The multiple sensing operations are preferably performed so as to cover the entire outer circumferential surface of the cylindrical welding object 5. The multiple sensing operations may be performed, for example, so as to cover the entire intersection line between the two welding objects 5, i.e., so as to obtain the shape of the outer circumferential surface of the cylindrical welding object 5 connected to the intersection line. For example, the weld line 6 may be identified by such second sensing. Sensing for obtaining the planar shape of the flat welding object 5 may also be performed multiple times. The multiple sensing operations may be performed so as to cover the entire intersection line of the two objects to be welded 5, that is, so as to obtain the shapes of the flat objects to be welded 5 connected to the intersection line.

[0049] Furthermore, in the first and second sensing, for example, the accuracy of the first sensing may be lower than the accuracy of the second sensing. This accuracy may be the accuracy in a situation where there is no influence of secondary reflection. Usually, the first sensing only needs to determine the approximate shape of the welding object 5, so a low accuracy of the first sensing is not a problem. Therefore, for example, the first sensing may be performed only once, and the second sensing may be performed multiple times for each surface of the sensing object. This reduces the time required for the first sensing, shortening the overall time required to identify the weld line and improving the accuracy of the second sensing, thereby improving the accuracy of identifying the weld line.

[0050] Furthermore, for example, when sensing is performed at multiple positions of the sensor 3 in the second sensing, i.e., when sensing is performed on multiple surfaces in the second sensing, the control unit 12 may determine, among the multiple positions of the sensor 3 in the second sensing, the position that is closest to the position of the sensor 3 in the first sensing as the position of the sensor 3 in the initial second sensing. This reduces the time required to move the sensor 3 from the position of the sensor 3 in the first sensing to the position of the sensor 3 in the initial second sensing, thereby reducing the overall time required to identify the weld line. In this case, for example, the control unit 12 may determine the position of the sensor 3 in the second sensing using the results of the first sensing, and then control the sensor 3 to move to the identified position of the sensor 3. Note that this distance may be, for example, a physical distance or a temporal distance. The physical distance may be, for example, a linear distance or a travel distance. For example, if there is an obstacle between the origin and destination, the travel distance may be longer than the straight-line distance. The temporal distance may be, for example, travel time.

[0051] Furthermore, for example, the control unit 12 may control the second sensing not to be performed if the weld line can be identified based on the three-dimensional shape of the welding target acquired by the first sensing. As described above, in the first sensing as shown in FIG. 3A, the amount of light received from the primary reflection is typically reduced, and the influence of the secondary reflection is significant, resulting in reduced sensing accuracy. On the other hand, if the welding target 5 is an object with low surface reflectivity, such as a black iron oxide coated surface, there may be little secondary reflection, and the weld line may be identified using the results of the first sensing. In this way, if the first sensing can be performed with high accuracy, the second sensing may not be performed, thereby enabling the weld line to be identified in a shorter time. Note that, if the weld line cannot be identified based on the three-dimensional shape of the welding target acquired by the first sensing, the control unit 12 may control the manipulator 2, etc., to perform the second sensing.

[0052] The determination of whether a weld line can be identified based on the three-dimensional shape of the welding object acquired by the first sensing, i.e., the determination of whether high-precision sensing was achieved in the first sensing, may be made, for example, by performing multiple sensing operations in the first sensing and using the variance of multiple measurement values ​​calculated for each point in the three-dimensional point cloud data. For example, if the variance is small, it may be determined that high-precision sensing was achieved, and if the variance is not small, it may be determined that high-precision sensing was not achieved. A small variance may mean, for example, that a representative value of the variances corresponding to each point in the three-dimensional point cloud data or the sum of all the variances corresponding to each point is smaller than a predetermined threshold, or that the sum of a predetermined number of values ​​with the highest variances corresponding to each point in the three-dimensional point cloud data is smaller than a predetermined threshold. The representative value may be, for example, the mean, median, maximum, minimum, etc.

[0053] The determination of whether high-precision sensing was achieved in the first sensing may be made, for example, using the difference between the three-dimensional shape acquired using the first sensing result and each point of the three-dimensional point cloud data acquired by the first sensing. The difference between the three-dimensional shape and the point may be, for example, the distance between the three-dimensional shape (e.g., a plane) and the point. For example, if the difference is small, it may be determined that high-precision sensing was achieved, and if the difference is not small, it may be determined that high-precision sensing was not achieved. A small difference may mean, for example, that a representative value of the differences corresponding to each point of the three-dimensional point cloud data or a total value of all the differences corresponding to each point is smaller than a predetermined threshold, or that a total value of a predetermined number of the largest differences corresponding to each point of the three-dimensional point cloud data is smaller than a predetermined threshold. The representative value is as described above. For example, if the point cloud data indicated by the black circles in FIG. 6A is acquired by the first sensing, it may be determined that high-precision sensing was achieved, and if the point cloud data indicated by the black circles in FIG. 6B is acquired, it may be determined that high-precision sensing was not achieved. 6A and 6B are diagrams each schematically showing a state in which three-dimensional point cloud data is viewed from the direction of the intersection line between two flat plate-shaped welding objects 5.

[0054] In this case, when the second sensing is not performed, the identifying unit 13 may identify the weld line using the three-dimensional shape of the welding object acquired by the first sensing. In this case, since the first sensing is performed with high accuracy, the identifying unit 13 can identify the weld line with high accuracy by using the three-dimensional shape of the welding object acquired by the first sensing. Furthermore, the time required to identify the weld line can be reduced compared to when the second sensing is performed.

[0055] Furthermore, in the present embodiment, when sensing of two or more surfaces is performed in the second sensing, the case where the three-dimensional shape is acquired for each surface and then the acquired results are integrated has been mainly described, but this is not necessarily the case. For example, the acquisition unit 11 may integrate the three-dimensional point cloud data acquired for each surface and then acquire the three-dimensional shape.

[0056] In the present embodiment, the case where control unit 12 controls welding torch 2a to be performed along the identified weld line has been described, but this is not necessarily the case. Control unit 12 does not have to perform control for welding. For example, welding using the identified weld line may be performed by another manipulator. In this case, welding torch 2a may not be attached to manipulator 2. When welding torch 2a is not attached to manipulator 2, sensor 3 may be attached to the tip of manipulator 2.

[0057] Furthermore, in the above embodiments, each process or function may be realized by centralized processing by a single device or a single system, or may be realized by distributed processing by multiple devices or multiple systems.

[0058] In the above embodiments, each component may be configured with dedicated hardware, or components that can be realized by software may be realized by executing a program. For example, each component may be realized by a program execution unit such as a CPU reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory. During execution, the program execution unit may execute the program while accessing a storage unit or recording medium. The program may also be executed by being downloaded from a server or the like, or by being read from a predetermined recording medium. The program may also be executed by a single computer or multiple computers. That is, centralized processing or distributed processing may be performed.

[0059] Furthermore, the present invention is not limited to the above-described embodiment, and various modifications are possible, and it goes without saying that these modifications are also included within the scope of the present invention. [Explanation of symbols]

[0060] 1 robot control device, 2 manipulator, 2a welding torch, 3 sensor, 5 welding object, 11 acquisition unit, 12 control unit, 13 identification unit

Claims

1. A robot control device that identifies a weld line on a welding object by controlling a manipulator equipped with a sensor that performs sensing by irradiating the welding object with light, an acquisition unit that acquires a three-dimensional shape of the welding object using a sensing result of the sensor; a control unit that controls the manipulator and the sensor so as to perform a first sensing of the entire object to be welded and a second sensing in which light is irradiated from a normal direction of each face that defines an intersection line in the object to be welded based on the three-dimensional shape of the object to be welded acquired by the first sensing; an identification unit that identifies a weld line using the three-dimensional shape of the welding object acquired by the second sensing, A robot control device, wherein the first and second sensing are performed using the same sensor.

2. the control unit controls so that the second sensing is not performed when a weld line can be identified based on the three-dimensional shape of the welding object acquired by the first sensing; The robot control device according to claim 1 , wherein the identifying unit identifies the weld line using the three-dimensional shape of the welding object acquired by the first sensing when the second sensing is not performed.

3. The robot control device according to claim 1 , wherein the accuracy of the first sensing is lower than the accuracy of the second sensing.

4. 4. A robot control device according to claim 1, wherein, when sensing is performed at a plurality of positions of the sensor in the second sensing, the control unit determines, among the plurality of positions of the sensor in the second sensing, the position that is the shortest distance from the position of the sensor in the first sensing as the position of the sensor in the initial sensing of the second sensing.

5. The manipulator also has a welding torch attached to it; The robot control device according to claim 1 , wherein the control unit controls the manipulator so that welding is performed by the welding torch along the weld line identified by the identification unit.

Images