Robot control method and robot control device

Abstract

A robot control method acquires coordinate information of a designated point included in a movement path of a torch tip when a robot is operated, and calculates a target position of the torch tip on the basis of a database in which coordinates of the target position of the torch tip and an amount of shift of an arm tip with respect to the target position are associated with each other. The amount of offset associated with the vertices of the lattice included in the specified point is determined, and the amount of offset of the tip of the welding torch from the specified point when the operation is performed with the specified point as the target position is derived on the basis of the determined amount of offset and the coordinates of the specified point and the vertices. A target position of the robot with respect to the designated point is corrected based on the derived offset.

Description

Technical Field

[0001] The invention relates to a robot control method and a robot control device. Background Art

[0002] Conventionally, a robot is used to stack weld beads to shape a stacked object. In addition, when welding is performed automatically, an operator pre-teach the welding path and posture. When performing stacked molding, in order to improve the molding accuracy, it is necessary to control the position deviation caused by the robot's posture, structural error, etc.

[0003] For example, Patent Document 1 discloses detecting the angles of the axes of a robot when a plurality of postures are taught to an arbitrary point in space, calculating the setting errors required to match the coordinate axes of the tool tip taught in the plurality of postures, and performing automatic correction.

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: Japanese Patent Application Publication No. 61-133409 Summary of the invention

[0007] Problems to be solved by the invention

[0008] For example, in stacking modeling using welding wire, the target position is sometimes offset due to the performance of the robot used, manufacturing errors, etc., and there are cases where this offset affects the shape quality and internal quality of the modeled object. For example, when the pass becomes longer in a large modeling object, the impact caused by the position offset also becomes larger, which also affects the final quality. In particular, in stacking modeling, there are also cases where the layers are stacked in various postures, and the length equivalent to one pass becomes very long. In addition, there is a characteristic that the number of passes tends to become very large. Therefore, it is not realistic to teach each pass in advance and correct the position offset each time. In addition, due to the individual differences of the robots, the position offset needs to be corrected for each robot. In Patent Document 1, separate teaching is required, and it is impossible to reduce the effort in the above-mentioned case. Therefore, a method is sought to automatically correct the offset of the target position at any point of the stacking plan involved in stacking modeling.

[0009] In view of the above problems, an object of the present invention is to automatically correct a positional deviation of a target position of a welding torch tip of a robot.

[0010] Solutions to Solve Problems

[0011] In order to solve the above-mentioned problems, the present invention has the following structure. That is, a control method is a control method of a robot having a welding torch at the front end of an arm, wherein:

[0012] The control method comprises:

[0013] an acquisition step of acquiring coordinate information of a designated point included in a movement path of the front end of the welding torch when the robot is moved;

[0014] A determination step, based on a database that establishes a correspondence between the coordinates of the target position of the front end of the welding torch and the offset of the front end of the welding torch relative to the target position, determines information on the offset that establishes a correspondence with the vertices of the grid that encloses the specified point;

[0015] a deriving step of deriving, based on the offset determined in the determining step and the coordinates of the designated point and the vertex, an offset of the tip of the welding torch from the designated point when the robot is moved with the designated point as a target position; and

[0016] The correction step corrects the target position of the robot relative to the designated point based on the offset amount derived in the deriving step.

[0017] In addition, as another aspect of the present invention, there is the following structure. That is, a control device is a control device for a robot having a welding torch at the front end of an arm, wherein:

[0018] The control device has:

[0019] A holding unit, which holds a database in which the coordinates of the target position of the front end of the welding torch and the offset of the front end of the welding torch relative to the target position are established in a corresponding relationship;

[0020] an acquisition unit that acquires coordinate information of a designated point included in a movement path of the front end of the welding torch when the robot is moved;

[0021] a determination unit, which determines, according to the database, information on the offset that establishes a corresponding relationship with the vertices of the grid that encloses the specified point;

[0022] an deriving unit that derives, based on the offset determined by the determining unit and the coordinates of the designated point and the vertex, an offset of the tip of the welding torch from the designated point when the robot is moved with the designated point as a target position; and

[0023] A correction unit corrects a target position of the robot relative to the designated point based on the offset derived by the deriving unit.

[0024] Effects of the Invention

[0025] According to the present invention, it is possible to automatically correct the positional deviation of the target position of the welding torch tip of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic configuration diagram showing a configuration example of a stacking molding system according to an embodiment of the present invention.

[0027] Figure 2A This is a schematic diagram showing the concept of a DB structure for deriving a correction amount according to one embodiment of the present invention.

[0028] Figure 2B This is a table diagram showing an example of the DB structure for deriving the correction amount according to one embodiment of the present invention.

[0029] Figure 3 This is a conceptual diagram showing an example of derivation of a correction amount according to one embodiment of the present invention.

[0030] Figure 4 This is a flowchart of a positional deviation correction process according to one embodiment of the present invention.

[0031] Figure 5A Graph showing the derivation result of the correction amount according to one embodiment of the present invention.

[0032] Figure 5B Graph showing the derivation result of the correction amount according to one embodiment of the present invention.

[0033] Figure 5C Graph showing the derivation result of the correction amount according to one embodiment of the present invention.

[0034] Fig. 6A This is a graph for explaining the addition of designated points according to one embodiment of the present invention.

[0035] Figure 6B This is a graph for explaining the addition of designated points according to one embodiment of the present invention.

[0036] Figure 7 This is a graph for explaining changes in the offset amount according to one embodiment of the present invention.

[0037] Figure 8 This is a graph for explaining changes in the offset amount according to the posture according to one embodiment of the present invention.

[0038] Fig.9A This is a graph for explaining changes in the offset amount according to the posture according to one embodiment of the present invention.

[0039] Fig. 9BThis is a graph for explaining the conversion of the offset amount according to the posture according to one embodiment of the present invention. DETAILED DESCRIPTION

[0040] Hereinafter, the present specific embodiment will be described with reference to the accompanying drawings, etc. It should be noted that the embodiment described below is used to illustrate an embodiment of the present invention and is not intended to limit the present invention. In addition, the structures described in each embodiment may not all be structures necessary to solve the problems of the present invention. In addition, in each of the accompanying drawings, the same reference numerals are marked on the same constituent elements to indicate the corresponding relationship.

[0041] <First embodiment>

[0042] The first embodiment of the present invention is described below. Here, a stacking molding system capable of performing stacking molding is taken as an example for description. However, as long as the essential part of the present invention can be installed, it is not limited to this.

[0043] [System structure]

[0044] Figure 1 This is a schematic structural diagram of a stacking molding system to which the present invention can be applied.

[0045] The stacking molding system 1 of this embodiment is configured to include a stacking molding device 100 and an information processing device 200 for comprehensively controlling the stacking molding device 100. Figure 1 , a three-dimensional coordinate system represented by an X-axis, a Y-axis, and a Z-axis is shown. The origin position of the three-dimensional coordinate system is not particularly limited, but an arbitrary position is set, and the stacking molding system 1 performs stacking molding operations based on the origin position.

[0046] The stacking molding apparatus 100 includes a welding robot 104 , a filler material supply unit 105 that supplies a filler material (welding wire) M to the welding torch 102 , a robot controller 106 that controls the welding robot 104 , and a power source 107 .

[0047] The welding robot 104 is a multi-joint robot, and the welding torch 102 provided at the front end shaft supports the filler material M so as to be continuously supplied. The welding torch 102 holds the filler material M in a state protruding from the front end. The position and posture of the welding torch 102 can be arbitrarily set three-dimensionally within the range of the degree of freedom of the robot arm constituting the welding robot 104.

[0048] The welding torch 102 has a shielding nozzle (not shown), and shielding gas is supplied from the shielding nozzle. The shielding gas blocks the atmosphere, prevents oxidation and nitridation of the molten metal during welding, and suppresses welding defects. The arc welding method used in this embodiment may be any of a consumable electrode type such as covered arc welding and carbon dioxide gas arc welding, a non-consumable electrode type such as TIG welding and plasma arc welding, and is appropriately selected according to the stacked object W to be formed.

[0049] A shape sensor 101 that can move following the movement of the welding torch 102 is provided near the welding torch 102. The shape sensor 101 detects the shape of the stacked object W formed on the base 103. In the present embodiment, the shape sensor 101 can be used to detect the height, position, width, etc. of the welded bead 108 (also simply referred to as a "weld bead") constituting the stacked object W. The information detected by the shape sensor 101 is sent to the information processing device 200. It should be noted that the structure of the shape sensor 101 is not particularly limited, and it may be a structure that detects the shape by contact (contact sensor) or a structure that detects the shape by using a laser or the like (non-contact sensor). It should be noted that the mechanism for deriving the shape of the formed weld bead is not limited to the shape sensor 101 provided near the welding torch 102. For example, it may be a structure that indirectly derives the shape of the formed weld bead. As an example, a DB (database) indicating the curves of the welding current and the feed speed of the filler material M and the tendency of the height of the weld bead may be defined in advance, and the height of the formed weld bead may be derived based on the welding conditions during modeling.

[0050] In the welding robot 104, when the arc welding method is a consumable electrode type, a conductive tip is arranged inside the shielded nozzle, and the filler material M supplied with molten current is held by the conductive tip. The welding torch 102 generates an arc from the front end of the filler material M under a shielding gas atmosphere while holding the filler material M. The filler material M is fed from the filler material supply unit 105 to the welding torch 102 by a delivery mechanism (not shown) installed on a robot arm or the like. And, when the continuously fed filler material M is melted and solidified while the welding torch 102 is moved, a linear weld bead 108 as a molten solidified body of the filler material M is formed on the base 103. By stacking the weld bead 108, a stacked object W is formed.

[0051] It should be noted that the heat source for melting the filler material M is not limited to the above-mentioned arc. For example, a heat source based on a heating method based on a combination of arc and laser, a heating method using plasma, a heating method using electron beam, a laser, or other methods may be used. In the case of heating using an electron beam or a laser, the amount of heating can be more finely controlled, and the state of the weld bead 108 can be more appropriately maintained, which can contribute to further improvement of the quality of the stacked object W.

[0052] The robot controller 106 drives the welding robot 104 using a prescribed driving program based on the instruction from the information processing device 200, and shapes the stacked object W on the base 103. That is, the welding robot 104 moves the welding torch 102 while melting the filler material M using an arc according to the instruction from the robot controller 106. The power supply 107 is a welding power supply that supplies the power required for welding to the robot controller 106. The power supply 107 can operate in a plurality of control modes, and can switch the power (current, voltage, etc.) when supplying the power to the robot controller 106 according to the control mode. The filler material supply unit 105 controls the supply of the filler material M to the welding torch 102 of the welding robot 104 and the feed speed based on the instruction from the information processing device 200.

[0053] The information processing device 200 may be, for example, an information processing device such as a PC (Personal Computer). Figure 1 Each function shown can be realized by a control unit (not shown) reading out and executing the program of the function of the present embodiment stored in a storage unit (not shown). As a storage unit, a RAM (Random Access Memory) as a volatile storage area, a ROM (Read Only Memory) as a non-volatile storage area, an HDD (Hard Disk Drive), etc. can be included. In addition, a CPU (Central Processing Unit), a GPU (Graphical Processing Unit), or a GPGPU (General-Purpose computing on Graphics Processing Units) can also be used as a control unit.

[0054] The information processing device 200 is configured to include a molding control unit 201, a power supply control unit 202, a feed control unit 203, a DB management unit 204, a shape data acquisition unit 205, a teaching data acquisition unit 206, a correction amount calculation unit 207, and a peripheral device control unit 208. The molding control unit 201 generates a control signal for the robot controller 106 during molding based on the design data (for example, CAD / CAM data, etc.) of the stacked molding object W to be molded. The control signal here includes the movement trajectory of the welding torch 102 by the welding robot 104, the welding conditions when the weld bead 108 is formed, the feed speed of the filler material M by the filler material supply unit 105, etc. The movement trajectory of the welding torch 102 is not limited to the trajectory of the welding torch 102 in the process of forming the weld bead 108 on the base 103, and for example, it also includes the movement trajectory of the welding torch 102 to the starting position of forming the weld bead 108.

[0055] The power supply control unit 202 controls the power supply (control mode) from the power supply 107 to the robot controller 106. Depending on the control mode, the current and voltage values, the current waveform (pulse), etc. when forming a weld bead of the same shape may be different. In addition, the power supply control unit 202 obtains information on the current and voltage supplied to the robot controller 106 from the power supply 107 at an appropriate time.

[0056] The feed control unit 203 controls the feed speed and feed timing of the filler material M based on the filler material supply unit 105. The feed control of the filler material M here includes not only sending out (forward feeding) but also returning (reverse feeding). The DB management unit 204 manages the DB (database) of this embodiment. The DB of this embodiment includes a database that specifies the correction amount for correcting the position deviation described later. The details of the structure of the DB will be described later. The shape data acquisition unit 205 acquires the shape data of the weld bead 108 formed on the base 103 detected by the shape sensor 101.

[0057] The teaching data acquisition unit 206 acquires teaching data including parameters taught by an operator via a teaching pendant (not shown), for example. The molding control unit 201 controls the welding robot 104 by using the design data of the stacked molding W and the teaching data so as to mold the stacked molding W. In the present embodiment, these data are also summarized and recorded as "stacked planning data". It should be noted that the data included in the stacked planning data is not particularly limited. The correction amount calculation unit 207 calculates the correction amount for the position deviation caused by the manufacturing error of the welding robot 104, etc., through the processing described later.

[0058] The peripheral device control unit 208 controls the operation of peripheral devices provided around the welding robot 104 provided in the stacking molding system 1. Figure 1Although not shown in the figure, for example, a positioner for adjusting the position and posture of the base 103 for shaping the stacked object W, a slider that enables the welding robot 104 to slide in a predetermined direction, etc. Therefore, the welding robot 104 can also perform stacking shaping in conjunction with these peripheral devices.

[0059] The coordinate system in the design data of this embodiment corresponds to the coordinate system of the welding robot 104 , and the three axes (X axis, Y axis, and Z axis) of the coordinate system are set so as to define a three-dimensional position with an arbitrary position as the origin.

[0060] The stacked molding system 1 of the above structure melts the filler material M while moving the welding torch 102 according to the moving trajectory of the welding torch 102 specified by the set design data, and supplies the molten filler material M onto the base 103. As a result, a stacked molding W in which a plurality of linear weld beads 108 are arranged and stacked is molded on the upper surface of the base 103.

[0061] [database]

[0062] In the present embodiment, in the shape data for the stacked object W represented by the design data, a point (hereinafter referred to as a "specified point") on the path (hereinafter also referred to as a "pass") during the shaping is specified. When the tip of the welding torch 102 is configured to shape the specified point, a positional deviation may occur due to a manufacturing error of the welding robot 104 or the like. Therefore, in order to derive a correction amount for correcting the positional deviation, in the present embodiment, a database (hereinafter also referred to as a "robot error map") that defines the correction amount for each coordinate and posture in the three-dimensional coordinate system is used. The database is pre-defined based on, for example, an actual measurement value when the tip of the welding torch 102 of the welding robot 104 is moved relative to the coordinates of the object.

[0063] Figure 2A : is a diagram showing the conceptual structure of the robot error map of this embodiment. Figure 2A As shown in FIG. 1 , the robot error map 300 defines a correction amount for each coordinate point 301 on a grid in a three-dimensional space. In this case, the intervals between the coordinate points are not particularly limited. The unit grid, which is the minimum unit of the grid structure, is not limited to Figure 2A The shape of the grid structure may be a cube, or other shapes such as a triclinic grid. In addition, in the grid structure, it is not necessary to define parameters for all grid points, and the parameters of some grid points may not be defined. The processing when the designated point included in the path of the design data is a coordinate other than the coordinate point specified by the robot error map 300 will be described later.

[0064] Figure 2B: is a table diagram showing a structural example of the robot error map 300 of the present embodiment. In the present embodiment, as parameters representing the position and posture of the welding robot 104 during shaping, coordinate information (x, y, z) representing the coordinates of the X-axis, Y-axis, and Z-axis of the three-dimensional coordinate system and posture information (α, β, γ) representing the rotation angle of the front end portion (e.g., welding torch 102) of the welding robot 104 around the X-axis, Y-axis, and Z-axis are used. It should be noted that the coordinate information is not limited to the above. For example, parameters (r, θ, z) corresponding to a cylindrical coordinate system may also be used.

[0065] In addition, dx, dy, and dz represent the correction amounts of the X-axis, Y-axis, and Z-axis, respectively. In other words, correction is performed assuming that the positional deviation is the amount of the correction amount shown here. Figure 2B In the formula, the correction amount (dx, dy, dz) is defined corresponding to (x, y, z, α, β, γ). 111 ,y 111 , z 111 , α 111 , β 111 , γ 111.1 ) = (dx 111.1 ,dy 111.1 ,dz 111.1 ). The suffix of each parameter is used to uniquely identify Figure 2A The grid points are shown in Figure 2. Figure 2A In the example of , there are 27 grid points (i.e., coordinate points), which are set to identify them using the positions of the three axes. In addition, even the same coordinate points may have different position offsets depending on the posture, so for convenience, the difference in posture is expressed by using the decimal point after the suffix (.1, .2, ...).

[0066] It should be noted that the correction amount specified by the robot error map 300 is not limited to the offset between the target position to be specified by the control of the welding robot 104 and the tip position of the welding torch 102. For example, it can also be the deviation between the trajectory (for example, a straight line) that geometrically connects the multiple target positions specified by the control and the trajectory that the tip of the welding torch 102 actually passes. The robot error maps related to these correction amounts can be defined separately. The target position of the specified point specified by the stacking plan data is derived by referring to the former, and the area connecting the specified points can be derived by referring to the latter.

[0067] The robot error map 300 of this embodiment can be updated as appropriate, and the updating method is not particularly limited. In addition, the robot error map 300 can also be updated when the structure of the welding robot 104 changes or a certain driving time has passed.

[0068] Next, a method for deriving a correction amount for a designated point whose coordinates are not specified by the robot error map 300 is described. First, when the designated point coincides with a coordinate point specified by the robot error map 300, the correction amount for the designated point can use the value of the robot error map 300. On the other hand, when the designated point coincides with any of the coordinate points specified by the robot error map 300, the correction amount is derived from the coordinate points located around the designated point.

[0069] Figure 3 This is a conceptual diagram for explaining the method of deriving the correction amount of the present embodiment. Here, a designated point P corresponding to the coordinates of the correction amount not specified in the robot error map 300 is taken as an example for explanation. The parameters of the designated point P are set to (x, y, z, α, β, γ). First, the coordinate point closest to the position of the designated point P among the coordinate points specified in the robot error map 300 (hereinafter referred to as the "closest point") is determined. Here, the closest point is set to the coordinate point Q ijk，nearest For explanation.

[0070] Get the coordinate point Q from the robot error map 300 ijk，nearest Here, as the correction amount corresponding to the coordinate point Q ijk，nearest The corresponding correction values ​​for the X-axis, Y-axis, and Z-axis are obtained by ijk ,y ijk , z ijk , α ijk , β ijk , γ ijk ),dy(x ijk ,y ijk , z ijk , α ijk , β ijk , γ ijk )、dz(x ijk ,y ijk , z ijk , α ijk , β ijk , γ ijk ). At this time, between the specified point P and the coordinate point Q ijk，nearest If the distance is closer than the specified threshold, the coordinate point Q ijk，nearest The same value as the prescribed correction amount is used as the correction amount for the designated point P. The prescribed threshold value for the distance between points here may be prescribed in advance.

[0071] It should be noted that at the coordinate point Q obtained from the robot error map 300 ijk，nearest The posture condition θ q and the posture condition θ set for the specified point P p Under different circumstances, with attitude condition θp For example, when θ p =(α,β,γ),θ q =(αu, βv, γw), α=αu, β=βv, γ≠γw, and linear interpolation is performed with respect to γ, the correction amount d can be derived using the following equation (1).

[0072] d ijk (γ) = (1-t)d ijk (γ w1 )+td ijk (γ w2 ) … (1)

[0073] Among them, 0<t<1, γ w1 ≠γ w2 , γ w1 <γ<γ w2

[0074] Similarly, bilinear interpolation (α=αi, β≠βi, γ≠γi) and intra-element interpolation (α≠αi, β≠βi, γ≠γi) can be used for interpolation. These methods are well known and detailed descriptions are omitted here. In addition, the interpolation method is not limited to linear interpolation, and a nonlinear function can also be defined in the manner of Lagrangian interpolation to obtain.

[0075] On the other hand, between the designated point P and the coordinate point Q as the closest point ijk，nearest In this embodiment, when the distance is greater than a predetermined threshold, a correction amount is newly derived. In this embodiment, based on a plurality of coordinate points Q located around the designated point P, ijk The correction amount corresponding to the specified point P is derived by interpolation. ijk The coordinate point Q can be included as the closest point ijk，nearest , for example you can use Figure 3 The eight coordinate points of the unit grid shown in FIG. 8 are very close to the intervals between the specified points and the vertices on the grid, so it is possible to derive a correction amount that matches the actual offset amount well. The multiple coordinate points focused on here may also be coordinate points at other locations. For example, if the information on the offset amount at the four points on the bottom of the coordinate points on the unit grid is missing, when there is a deviation in the amount of information, you can refer to the coordinate points on a grid larger than the unit grid, or you can use the information from Figure 2A When the center point of the grid is observed, there are six coordinate points in total, namely, the coordinate points in the front, back, left, right, and right above and below. In addition, multiple coordinate points included in the range of a sphere formed by a predetermined radius R with the designated point P as the center may also be used.

[0076] As described above, in this embodiment, in the lattice structure, vertices of the lattice are selected so as to include the designated point. It should be noted that the number and positions of the coordinate points selected may be specified based on the positional relationship between the designated point and the coordinate points near it.

[0077] The posture condition θ at the specified point P p With Q ijk The posture condition θ q Different situations, through Q ijk，nearest The same steps are performed with the posture condition θ p Interpolate in the corresponding way.

[0078] As a calculation method in this case, for example, the least square method can be used. First, the correction amount dx in the x-axis direction is calculated based on the coordinate point Q ijk The information of defines the following equations (2) to (5). Then, the coefficient w is obtained using the least square method using equation (6), and the correction amount in the x-axis direction at the specified point P, that is, the offset amount b, is calculated.

[0079] [Formula 1]

[0080]

[0081] [Formula 2]

[0082]

[0083] [Formula 3]

[0084]

[0085] [Formula 4]

[0086]

[0087] [Formula 5]

[0088]

[0089] It should be noted that the above-mentioned equation is an equation for obtaining the offset amount in the X-axis direction, but the offset amount in the Y-axis direction and the Z-axis direction can also be calculated in the same manner.

[0090] [Processing Flow]

[0091] Figure 4 This is a flowchart of the correction process of the position deviation of the specified point in this embodiment. This process flow is executed by the information processing device 200. For example, the processing unit such as the CPU of the information processing device 200 can be used to implement Figure 1The programs of the various parts shown are read out from a storage unit (not shown) and executed. In addition, before executing this processing flow, a robot error map 300 is specified, and stacking plan data for stacking modeling is prepared. It should be noted that this processing flow can be implemented in advance before the stacking modeling action starts, and can also be executed in parallel during the stacking modeling action.

[0092] In S401, the information processing device 200 acquires information on the movement path for modeling from the stacking plan data. The movement path is composed of a plurality of designated points, and this information is acquired.

[0093] In S402, the information processing device 200 focuses on one designated point (hereinafter referred to as the "focused designated point") from among the designated points obtained in S401, and obtains the parameters related to the layered modeling. The parameters here are the six parameters (x, y, z, α, β, γ) as described above.

[0094] In S403, the information processing device 200 determines the closest point to the eye designated point from the predetermined robot error map 300, and obtains its correction amount information. The correction amount information here is set as the three parameters of dx, dy, and dz, which are the correction amounts of the X axis, Y axis, and Z axis respectively. As described above, there is a case where the coordinates of the eye designated point and the coordinates of the closest point are consistent.

[0095] In S404, the information processing device 200 determines whether the distance between the designated point of interest and the closest point is less than a predetermined threshold value. The predetermined threshold value here can be predetermined in advance. In the case where the distance between the designated point of interest and the closest point is less than the predetermined threshold value (Yes in S404), the processing of the information processing device 200 proceeds to S405. On the other hand, in the case where the distance between the designated point of interest and the closest point is greater than the predetermined threshold value (No in S404), the processing of the information processing device 200 proceeds to S406.

[0096] In S405, the information processing device 200 sets the correction amount information of the closest point acquired in S403 as the correction amount for the designated point of interest. Then, the processing of the information processing device 200 proceeds to S408.

[0097] In S406, the information processing device 200 determines a plurality of nearby coordinate points located near the target designated point from the robot error map 300, and obtains their correction amount information. The plurality of nearby coordinate points may be, for example, Figure 3 8 nearby coordinate points shown. If there is a nearby coordinate point with undefined parameters among the multiple nearby coordinate points, the parameters of other nearby coordinate points may be used instead of the nearby coordinate point. In addition, the remaining nearby coordinate points, that is, 7 nearby coordinate points in the above example, may also be used.

[0098] In S407, the information processing device 200 derives the correction amount for the designated point of interest based on the correction amount information of the plurality of nearby coordinate points obtained in S406. The derivation method here can be derived based on the least squares method as described above. Then, after the correction amount is derived, the processing of the information processing device 200 proceeds to S408.

[0099] In S408 , the information processing device 200 corrects the coordinates of the designated point of interest, that is, the target position of the welding torch 102 , using the correction amount obtained in S405 or S407 .

[0100] In S409, the information processing device 200 determines whether there is an unprocessed designated point among the designated points obtained in S401. If there is an unprocessed designated point (Yes in S409), the processing of the information processing device 200 returns to S402 and the processing is repeated. On the other hand, if there is no unprocessed designated point (No in S409), this processing flow ends.

[0101] It should be noted that the designated points used as processing objects in the above flowchart may be processed for all designated points of the moving path represented by the stacking plan data, or may be processed for only a portion of the designated points included in the moving path. In addition, the correction may be performed only on the value of any one of the X-axis, Y-axis, and Z-axis.

[0102] [Control example]

[0103] Figure 5A to Figure 5C This is a diagram showing an example of comparing the correction amount (estimated value) obtained by the method of this embodiment with the actual measured value obtained by measuring the positional deviation of the tip of the welding torch 102 at a specified point, i.e., a target position. The fact that the correction amount can be estimated with good accuracy means that the positional deviation that occurs can be properly corrected. Figure 5A to Figure 5C In the figure, the vertical axis represents the offset [mm], and the horizontal axis represents each specified point. The interval between the specified points is set to 100 mm. Figure 5A to Figure 5C In FIG. 1 , the left side of the dotted line shows the value when any one of the parameters α, β, and γ as posture information is changed. In addition, the right side of the dotted line shows the value when multiple parameters α, β, and γ as posture information are changed.

[0104] Figure 5A It is a diagram showing the results of estimated values ​​and measured values ​​in the X-axis direction. Figure 5B It is a diagram showing the results of estimated values ​​and measured values ​​in the Y-axis direction. Figure 5C : is a graph showing the estimated value and the measured value in the Z-axis direction. Figure 5A to Figure 5CAs shown, the offset amount (ie, correction amount) can be estimated at a value close to the actual measured value in any axial direction.

[0105] As described above, this embodiment can automatically correct the positional deviation of the target position of the welding torch tip of the robot. At this time, the positional deviation can be corrected with high accuracy even for coordinates not included in the predetermined robot error map, which can reduce the operator's work.

[0106] [Variant 1]

[0107] A modified example of this embodiment will be described. Consider a case where the welding torch 102 is moved along a movement path specified by stacking plan data. In this case, the offset amount may vary for each position of the movement path. Fig. 6A 600 is a diagram for explaining the change in the offset amount in the moving path 600, wherein the vertical axis represents the X-axis direction and the horizontal axis represents the Y-axis direction. Fig. 6A The cross-section line in represents the offset on the XY plane. For example, there is a movement path 600 which includes a designated point 601 and a designated point 602.

[0108] The moving path 600 spans four areas with different offsets. In addition, the offsets are different at the designated point 601 and the designated point 602. Here, it is assumed that even if the offsets of the designated point 601 and the designated point 602 are corrected, the accuracy of the middle part deviates from the path specified by the stacking plan data because the offsets of the middle part are different.

[0109] So, if Figure 6B Add a new specified point to the moving path as shown. Figure 6B In the example, two designated points 611 and 612 are added. By adding designated points and treating them as correction objects, the entire movement trajectory can be corrected with better accuracy. The judgment on whether to add designated points can be, for example, Fig. 6A When multiple areas with different offsets are separated, the number of areas crossed by the moving path is compared with the number of designated points included in the moving path. If the number of areas crossed by the moving path is large, additional designated points are added. It should be noted that when dividing the area according to the offset, it is not limited to Fig. 6A The examples shown can also be divided into more detailed categories.

[0110] in addition, Figure 7 is used to Fig. 6ADifferent representations illustrate the distribution of positional offsets. Graphs showing the offset in the Y-axis direction when the welding torch 102 is moved in a movement path along the X-axis direction from the designated point 701 to the designated point 702 in a certain posture of the welding robot 104. Even when the offset is determined based on the coordinates of the designated point 701 and the designated point 702 and corrected, a positional offset (trajectory error) may occur in the middle of the coordinates of the designated point 701 and the designated point 702. Therefore, it is not limited to correcting the positional offset of a designated point based on the offset of a designated point, and the correction amount for the entire path can also be determined based on the offset estimated for multiple designated points included in the movement path.

[0111] [Variant 2]

[0112] Figure 8 β is a diagram showing changes in the offset amount that occurs when β, one of the parameters of the posture of the welding robot 104, is switched for the same target position. Figure 8 In the equation, the vertical axis represents the offset and the horizontal axis represents the value of β. Figure 8 In the example of , when the value of β is 0, the offset can be set to approximately 0. That is, even if the target position is the same, there is a case where the position offset can be suppressed by changing the posture.

[0113] Fig.9A as well as Fig. 9B 1 is a diagram showing changes in the offset amount for each posture of the welding robot 104 . Fig.9A An example is shown in which three designated points 902, 903, and 904 are included on the moving path 901. In addition, an overview is shown in which the welding robot 104 performs layered molding while switching between three postures 906a, 906b, and 906c based on the starting point 905. Here, the welding robot 104 sets the designated point 902 as the target position in the posture 906a, sets the designated point 903 as the target position in the posture 906b, and sets the designated point 904 as the target position in the posture 906c.

[0114] exist Fig. 9B In the case of posture 906a, the offset is Fig. 9B The offset in the case of posture 906b corresponds to Fig. 9B The offset in the case of posture 906c corresponds to the graph 911. Fig. 9B That is, the offset amount generated by the posture (here, parameter β) is different. Fig. 9BIn the case of the example, if the shaping can be performed while the value of β is switched while the posture 906b is maintained, the offset can be suppressed and the shaping with higher accuracy can be performed. Assuming such a situation, the welding robot 104 can be provided with a slider or the like that can be moved while the posture is maintained. Therefore, the information processing device 200 can improve the shaping accuracy by moving the starting point 905 using a slider or the like while the welding robot 104 is fixed in the posture 906b.

[0115] [Variant 3]

[0116] In the above structure, the correction amount is calculated using the least square method as shown in equations (2) to (6). Alternatively, the correction amount for the specified point P can be derived using a learning model obtained by machine learning. In this case, the learning model can be generated for use by repeatedly performing a learning process in which each parameter of the A matrix in the parameters of equations (2) to (6) is input and each parameter of the b matrix (i.e., the correction amount) is output. It should be noted that the input data (parameters) used in learning can be part or all of the coordinate information and posture information of the welding robot 104. Alternatively, the learning model can be generated for use by repeatedly performing a learning process in which the coordinate information, posture information, and various parameters of the grid points determined from the robot error map 300 are input and the offset is output. The machine learning here uses, for example, a method of deep learning (deep learning) based on a neural network, and supervised learning is used as an example for explanation. It should be noted that a more specific method (algorithm) of deep learning is not particularly limited, and for example, a well-known method such as a convolutional neural network (CNN) can be used. It should be noted that data with interpolated posture information may be used as learning data, thereby increasing the amount of learning data and improving the accuracy of prediction of the offset amount at a specified point.

[0117] It should be noted that the learning process does not necessarily have to be performed by the information processing device 200. For example, the information processing device 200 may provide learning data to a learning server (not shown) provided outside the information processing device 200 and perform learning processing on the server side. In addition, the server may provide the information processing device 200 with a learned model as needed. Such a learning server may be located on a network such as the Internet (not shown), and the learning server and the information processing device 200 may be connected to each other so as to be able to communicate. That is, the information processing device 200 may act as a machine learning device, or an external device may act as a machine learning device.

[0118] <Other embodiments>

[0119] In addition, the present invention can also be implemented by the following processing: a program or application for implementing the functions of one or more of the above-mentioned embodiments is supplied to a system or device using a network or storage medium, and one or more processors in a computer of the system or device reads and executes the program.

[0120] Alternatively, the present invention may be realized by a circuit that realizes one or more functions (for example, an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array)).

[0121] As described above, the following matters are disclosed in this specification.

[0122] (1) A control method is a method for controlling a robot having a welding torch at a front end portion of an arm, wherein:

[0123] The control method comprises:

[0124] an acquisition step of acquiring coordinate information of a designated point included in a movement path of the front end of the welding torch when the robot is moved;

[0125] A determination step, based on a database that establishes a correspondence between the coordinates of the target position of the front end of the welding torch and the offset of the front end of the welding torch relative to the target position, determines information on the offset that establishes a correspondence with the vertices of the grid that encloses the specified point;

[0126] a deriving step of deriving, based on the offset determined in the determining step and the coordinates of the designated point and the vertex, an offset of the tip of the welding torch from the designated point when the robot is moved with the designated point as a target position; and

[0127] The correction step corrects the target position of the robot relative to the designated point based on the offset amount derived in the deriving step.

[0128] This makes it possible to automatically correct the positional deviation of the target position of the welding torch tip of the robot. In this case, the positional deviation can be corrected with high accuracy even for coordinates not included in the predetermined robot error map, which can reduce the operator's work.

[0129] (2) The control method according to (1), wherein:

[0130] In the determining step, information on the offsets corresponding to the plurality of vertices is determined.

[0131] In the derivation step, the offset is derived by interpolation of the offsets associated with the plurality of vertices.

[0132] This makes it possible to accurately correct positional deviations even for coordinates that are not registered in the database.

[0133] (3) The control method according to (1), wherein:

[0134] The database also establishes a corresponding relationship with the parameters of the tilt posture of the welding torch.

[0135] This allows the target position to be corrected in consideration of a plurality of postures that the robot can take, and in particular, allows the tendency of the target position to deviate due to the tilted posture of the welding torch, such as a posture with a large deviation or a posture with a small deviation, to be directly reflected.

[0136] (4) The control method according to (3), wherein:

[0137] In the acquisition step, posture information relative to the specified point is also acquired.

[0138] In the derivation step, the offset is derived by interpolating a difference between the posture information with respect to the designated point and a parameter of the posture indicated by the information on the offset associated with the vertex determined in the determination step.

[0139] Thus, even the positional offset in a posture not registered in the database can be estimated based on the positional offset based on the parameters of the registered posture.

[0140] (5) The control method according to (1), wherein:

[0141] In the acquisition step, posture information relative to the specified point is also acquired.

[0142] In the derivation process, the output obtained by inputting the coordinate information and posture information of the designated point into the learned model is derived as the offset from the designated point. The learned model is generated by taking the coordinate information of the vertex and the posture information of the robot at the vertex as input and learning the offset relative to the vertex as output.

[0143] As a result, the position offset can be estimated using the learned model obtained by machine learning without using a complex mathematical model. In addition, the learned model has learned both coordinate information and posture information in advance, so there is no need to perform posture interpolation and coordinate interpolation step by step, and the offset can be derived in a form that takes into account the influence of posture and coordinates.

[0144] (6) The control method according to (1), wherein:

[0145] The control method further includes an adding step of adding a new designated point to the movement path acquired in the acquiring step.

[0146] This can suppress positional deviation on the moving path, thereby improving the quality of the molded object.

[0147] (7) The control method according to (1), wherein:

[0148] In the correction step, a target position of the robot relative to the movement path when the robot is moved along the movement path is corrected based on the offsets relative to a plurality of designated points included in the movement path.

[0149] Thus, the position deviation of the robot with respect to the movement path including the plurality of designated points can be corrected based on the position deviation with respect to the plurality of designated points.

[0150] (8) A control device is a control device for a robot having a welding torch at a front end portion of an arm, wherein:

[0151] The control device has:

[0152] A holding unit, which holds a database in which the coordinates of the target position of the front end of the welding torch and the offset of the front end of the welding torch relative to the target position are established in a corresponding relationship;

[0153] an acquisition unit that acquires coordinate information of a designated point included in a movement path of the front end of the welding torch when the robot is moved;

[0154] a determination unit, which determines, according to the database, information on the offset that establishes a corresponding relationship with the vertices of the grid that encloses the specified point;

[0155] an deriving unit that derives, based on the offset determined by the determining unit and the coordinates of the designated point and the vertex, an offset of the tip of the welding torch from the designated point when the robot is moved with the designated point as a target position; and

[0156] A correction unit corrects a target position of the robot relative to the designated point based on the offset derived by the deriving unit.

[0157] This makes it possible to automatically correct the positional deviation of the target position of the welding torch tip of the robot. In this case, the positional deviation can be corrected with high accuracy even for coordinates not included in the predetermined robot error map, which can reduce the operator's work.

[0158] Various embodiments have been described above, but the present invention is certainly not limited to these examples. As long as one is skilled in the art, it is obvious that various variations or modifications can be thought of within the scope of the patent technical solution, and these should certainly be understood to belong to the technical scope of the present invention. In addition, the various constituent elements in the above-mentioned embodiments can also be arbitrarily combined within the scope of the main purpose of the invention.

[0159] It should be noted that the present application is based on the Japanese patent application (Japanese Patent Application No. 2022-114754) filed on July 19, 2022, and the contents thereof are incorporated herein by reference.

[0160] Description of Reference Numerals

[0161] 1. Cascading modeling system

[0162] 100 Layered Modeling Device

[0163] 101 Shape Sensor

[0164] 102 Welding torch

[0165] 103 Base

[0166] 104 Welding Robot

[0167] 106 Robot Controller

[0168] 107 Power Supply

[0169] 108 Deposited weld

[0170] 200 Information processing device

[0171] 201 Modeling Control Department

[0172] 202 Power supply control unit

[0173] 203 Feed control unit

[0174] 204 DB (Database) Management Department

[0175] 205 Shape data acquisition unit

[0176] 206 Teaching data acquisition unit

[0177] 207 Correction amount calculation unit

[0178] 208 Peripheral device control unit

[0179] W Layered Modeling

[0180] M Filling material.

Claims

1. A control method for a robot having a welding torch at a front end of an arm, wherein: The control method comprises: an acquisition step of acquiring coordinate information of a designated point included in a movement path of the front end of the welding torch when the robot is moved; A determination step, based on a database that establishes a correspondence between the coordinates of the target position of the front end of the welding torch and the offset of the front end of the welding torch relative to the target position, determines information on the offset that establishes a correspondence with the vertices of the grid that encloses the specified point; a deriving step of deriving, based on the offset determined in the determining step and the coordinates of the designated point and the vertex, an offset of the tip of the welding torch from the designated point when the robot is moved with the designated point as a target position; and The correction step corrects the target position of the robot relative to the designated point based on the offset derived in the deriving step.

2. The control method according to claim 1, wherein: In the determining step, information on the offsets corresponding to the plurality of vertices is determined. In the derivation step, the offset is derived by interpolation of the offsets associated with the plurality of vertices.

3. The control method according to claim 1, wherein: The database also establishes a corresponding relationship with the parameters of the tilt posture of the welding torch.

4. The control method according to claim 3, wherein: In the acquisition step, posture information relative to the specified point is also acquired. In the derivation step, the offset is derived by interpolating a difference between the posture information with respect to the designated point and a parameter of the posture indicated by the information on the offset associated with the vertex determined in the determination step.

5. The control method according to claim 1, wherein: In the acquisition step, posture information relative to the specified point is also acquired. In the derivation process, the output obtained by inputting the coordinate information and posture information of the designated point into the learned model is derived as the offset from the designated point. The learned model is generated by taking the coordinate information of the vertex and the posture information of the robot at the vertex as input and learning the offset relative to the vertex as output.

6. The control method according to claim 1, wherein: The control method further includes an adding step of adding a new designated point to the movement path acquired in the acquiring step.

7. The control method according to claim 1, wherein: In the correction step, a target position of the robot relative to the movement path when the robot is moved along the movement path is corrected based on the offsets relative to a plurality of designated points included in the movement path.

8. A control device, which is a control device for a robot having a welding torch at the front end of an arm, wherein: The control device has: A holding unit, which holds a database in which the coordinates of the target position of the front end of the welding torch and the offset of the front end of the welding torch relative to the target position are established in a corresponding relationship; an acquisition unit that acquires coordinate information of a designated point included in a movement path of the front end of the welding torch when the robot is moved; a determination unit, which determines, according to the database, information on the offset that establishes a corresponding relationship with the vertices of the grid that encloses the specified point; a deriving unit that derives, based on the offset determined by the determining unit and the coordinates of the designated point and the vertex, an offset of the tip of the welding torch from the designated point when the robot is moved with the designated point as a target position; as well as A correction unit corrects a target position of the robot relative to the designated point based on the offset derived by the deriving unit.

Images