JP6129058B2 - Teaching point correction apparatus and teaching point correction method - Google Patents

Description

  The present invention relates to a teaching point correction apparatus and a teaching point correction method for correcting the position of a teaching point in a robot system.

  In the robot system, it is necessary to correct the teaching point set in the robot in accordance with the installation error of the work to be worked. Therefore, a correction value is calculated using a method of calculating a correction value based on the position coordinates when the touch sensor detects contact with the main surface of the workpiece (for example, refer to Patent Document 1), or using a two-dimensional displacement sensor. A calculation method (see, for example, Patent Document 2) has been proposed.

JP 2011-152599 A (paragraphs 0018 to 0043, FIGS. 2 to 3) JP 2011-183535 A (paragraphs 0021 to 0030, FIGS. 1 to 6)

  However, in the method of obtaining the position coordinates with the touch sensor, in order to obtain the three-dimensional position information, the touch sensor needs to be in contact with the three surfaces, and the robot axis direction is determined according to the surface. There was a need to change. When the robot operates, it is accompanied by a position shift (difference between the teaching data and the actual movement amount), but the position shift when changing the direction of the axis is larger than the parallel movement. The accuracy is greatly affected, and highly accurate position correction becomes difficult.

  On the other hand, in the method using a two-dimensional displacement sensor, since at least two pieces of position information can be obtained by one measurement, the number of operations of the robot can be reduced. However, in order to obtain three-dimensional position information, It is necessary to acquire position information in two orthogonal plane operations. That is, in any method, in order to obtain three-dimensional position information, it is necessary to change the axial direction, and highly accurate position correction is difficult.

  The present invention has been made to solve the above-described problems, and it is an object of the present invention to obtain a teaching point correction apparatus and a teaching point correction method capable of highly accurate three-dimensional position correction. is there.

A teaching point correction apparatus according to the present invention is a teaching point correction apparatus that corrects a teaching point of a robot set for a reference workpiece with respect to a work target workpiece installed in place of the reference workpiece. A position sensor that detects the distance to the surface and measures the position of the object, and an imager that matches the direction in which the position sensor detects the direction of imaging, and is connected to the tip of the robot ; a sensor unit measuring operation is that access in one direction, and a measurement control section for controlling the measurement operation by cooperation with the sensor unit and the robot, the position data measured by the position sensor, the image acquired by the imager A teaching point correction unit that calculates a displacement amount of the work target workpiece with respect to the reference workpiece based on the data, and corrects the teaching point based on the calculated displacement amount A reference block serving as a reference for a position is fixed to an installation surface on which one of the reference workpiece and the work target workpiece is installed, and the measurement control unit includes the position sensor. While controlling the measurement operation so as to acquire the position data and the image data for the one workpiece and the reference block while keeping the direction to be detected facing the installation surface, the teaching point correction unit Based on the position data and the image data acquired from the one direction , the displacement amount is calculated as a six-degree-of-freedom component composed of a three-axis parallel translation component and a three-axis rotation component. .

A teaching point correction method according to the present invention is a teaching point correction method for correcting a teaching point of a robot set for a reference workpiece with respect to a work target workpiece installed in place of the reference workpiece. A step of setting the reference workpiece on an installation surface to which a reference block is fixed, setting the teaching point, a position sensor for detecting the distance to the surface of the object and measuring the position of the object; A sensor unit provided with an imager whose orientation is detected by the position sensor is connected to the tip of the robot, and the measurement operation is accessed in one direction . A first measurement step of measuring a position with respect to the reference block; a step of installing the work target workpiece on the installation surface instead of the reference workpiece; and the sensor unit And calculating a displacement amount of the work target workpiece with respect to the reference workpiece from the measurement results of the first measurement step and the second measurement step. And correcting the teaching point based on the calculated displacement amount, and in the first measurement step and the second measurement step, respectively, while the direction detected by the position sensor is directed toward the installation surface, While controlling the operation of the robot and the sensor unit in a coordinated manner so as to obtain the position data and image data of the reference block and the work installed on the installation surface of the reference work and the work target work, wherein in the step of correcting the teaching point, on the basis of the said position data obtained from one direction the image data, and perpendicular to the displacement amount And calculating a component of six degrees of freedom consisting of rotational components of parallel movement component and three axes of three axes, characterized in.

  According to the present invention, three-dimensional position information can be obtained without changing the axial direction of the robot. Therefore, a teaching point correction apparatus and a teaching point correction method capable of highly accurate three-dimensional position correction are obtained. be able to.

1 is an overall configuration diagram of a robot system including a teaching point correction apparatus according to a first embodiment of the present invention. It is a block diagram of the sensor unit which comprises the teaching point correction apparatus concerning Embodiment 1 of this invention. It is a flowchart which shows the part which produces | generates the teaching point and position information with respect to a reference | standard workpiece among the teaching point correction methods concerning Embodiment 1 of this invention. It is a flowchart which shows the part which produces | generates the displacement amount with respect to the position of the reference | standard workpiece of the position of a work target workpiece | work among the teaching point correction methods concerning Embodiment 1 of this invention. It is a figure which shows the structure at the time of creating a teaching point in the teaching point correction apparatus concerning Embodiment 1 of this invention. It is a figure which shows the example at the time of creating a teaching point with respect to a reference | standard workpiece | work with the teaching point correction apparatus concerning Embodiment 1 of this invention. It is a figure which shows the example of a measurement by the position sensor for calculating a certain positional information ((theta) x0) in the teaching point correction apparatus concerning Embodiment 1 of this invention. It is a figure which shows the example which calculates one of positional information ((theta) x0) with the measured value of a position sensor with the teaching point correction apparatus or teaching point correction method concerning Embodiment 1 of this invention. It is a figure which shows the example of a measurement by the position sensor for calculating a certain positional information ((theta) y0) in the teaching point correction apparatus concerning Embodiment 1 of this invention. It is a figure which shows the measurement (imaging) example by the camera for calculating some positional information (x, y, (theta) z) with the teaching point correction apparatus concerning Embodiment 1 of this invention. It is a figure which shows the example which calculates position information (x, y, (theta) z) from the image information imaged with the teaching point correction apparatus or teaching point correction method concerning Embodiment 1 of this invention.

Embodiment 1 FIG.
FIGS. 1-11 is for demonstrating the teaching point correction apparatus and teaching point correction method concerning Embodiment 1 of this invention. 1 to 4 are for explaining the mechanical configuration of the robot system using the teaching point correcting apparatus and the operation in the teaching point correcting method, and FIG. 1 is a robot equipped with the teaching point correcting apparatus. FIG. 2 is a block diagram of the sensor unit constituting the teaching point correction apparatus, and FIG. 3 is a part of the teaching point correction apparatus and the teaching point correction method that generates a teaching point using a reference workpiece. FIG. 4 is a flowchart showing a part for generating a displacement amount of the work to be actually worked with respect to the reference work.

  FIGS. 5 and 6 show examples of teaching point generation for a reference workpiece. FIG. 5 shows a robot system equipped with a teaching point correction device, in which a reference block is arranged on an installation surface, and a robot arm tip FIG. 6 is a diagram showing a state at each stage when the insertion jig is brought close to the reference workpiece in creating a teaching point for the reference workpiece.

  FIGS. 7 to 11 are diagrams for explaining an example of a measurement or calculation method for generating position information indicating a positional relationship of a workpiece with respect to a reference block, and FIG. 7 shows a rotation component around the x axis. FIG. 8 is a diagram illustrating an example of measurement by a position sensor for calculating position information (θx0), and FIG. 8 is an example of calculating position information (θx0) indicating a rotation component around the x axis based on the measurement value of the position sensor. FIG. 9 is a diagram showing an example of measurement by a position sensor for calculating position information (θy0) indicating a rotation component around the y axis, and FIG. 10 is a translation component in the x direction of the position information. FIG. 11 is a diagram showing an example of measurement (imaging) by a camera for calculating a translation component in the y direction and a rotation component (x, y, θz) about the x axis, and FIG. 11 shows position information based on the captured image information. Calculate (x, y, θz) It is a figure which shows the example to take out.

  As shown in FIG. 1, a robot system 1 according to a first embodiment of the present invention includes a robot 2 that is a so-called articulated robot controlled by a control device 6 in order to perform work on a workpiece 4. It is. A sensor unit 3 is attached to the arm tip 2 e of the robot 2. As shown in FIG. 2, in the sensor unit 3, a housing 31 provided with a position sensor 32, a camera 33, and an illumination 34 is held by an auto tool changer 2c at an arm tip 2e. The auto tool changer 2c automatically attaches / detaches the assembly tool (not shown) such as the sensor unit 3 and the hand, and automatically performs the position correction operation and the assembly operation in the robot system 1.

  As for the position sensor 32, a contact-type displacement sensor is used in this embodiment. The camera 33 and the illumination 34 can be appropriately selected depending on the usage environment. The control device 6 is, for example, a computer including a storage device, and based on data measured by the position sensor 32 and the camera 33, three-dimensional position information (three orthogonal translational components + rotational components). Total 6 degrees of freedom) is calculated and stored. As a result, it is possible to generate teaching points for the workpiece 4 and to perform correction (calibration) described later. That is, the control device 6 is a controller that controls the robot 2 and functions as a teaching point correction device that generates and corrects teaching points in cooperation with the sensor unit 3.

  The workpiece 4 is a target on which the robot 2 works, such as a component mounting pallet and an assembly unit, and is exchanged when the type is switched. The exchange operation itself may be either manual exchange or automatic exchange. The reference blocks 5A and 5B are installed and fixed on the installation surface 7 of the robot system 1, and are the four references in the robot system 1. Therefore, the reference blocks 5A and 5B are not normally removed. The workpiece 4 and the reference blocks 5A and 5B are rigid bodies (for example, metal), and are provided with surfaces (for example, 5Af and 5Bf) that can be used as a reference so that the position sensor 32 can measure them. Further, it is assumed that a characteristic edge portion or marking is applied so that image processing can be performed from an image captured by the camera 33.

Next, the operation of the teaching point correction apparatus or the teaching point correction method according to the present embodiment will be described along the flow of the flowcharts of FIGS.
First, with respect to a plurality of types of workpieces to be worked, teaching points and position information are generated for a reference workpiece (also referred to as a master workpiece) (FIG. 3). First, one of the types of workpieces 4 is selected as the reference workpiece 4M and installed on the installation surface 7 (step S10). Here, as the workpiece 4, as shown in FIG. 5, a hexahedron having a square hole 4h opened on one surface is used, and a rectangular parallelepiped insertion jig 8 corresponding to the hole 4h is provided at the arm tip 2e of the robot 2. An example in which is installed will be described.

  Next, as shown in FIGS. 6A to 6C, the rectangular parallelepiped portion 8p at the tip of the insertion jig 8 is inserted into the rectangular hole 4h portion of the reference workpiece 4M to generate a teaching point ( Step S30). This insertion operation is not generally performed by automatic control, and is performed while finely adjusting the orientation and position of the robot 2. Then, the information (IO) of the translational component and the rotational component immediately before the rectangular parallelepiped portion 8p shown in FIG. 6B is inserted into the hole 4h, and the insertion of the rectangular parallelepiped portion 8p into the hole 4h are completed. Teaching the information (IP) of the translation component and the rotation component as a teaching point.

  Next, position information indicating the positional relationship between the reference work 4M for which teaching points have been taught and the reference blocks 5A and 5B installed and fixed on the installation surface 7 is generated (steps S30 to S50). First, as shown in FIG. 7, the position sensor 32 is brought into contact with two points (PMa and PMb and PAa and PAb) set on the reference work 4M and the reference surfaces 4f and 5Af of the reference block 5A. The position is measured (part of step S30). As the contact position, the value of the x coordinate is kept constant, the value of the y coordinate is set according to the point, and the value of the z coordinate when the position sensor 32 contacts the reference plane is acquired. The position where the position sensor 32 is brought into contact according to the point is set according to the type of the workpiece 4 as a teaching point of the robot 2 and stored in the control device 6.

From the contact position data measured using the position sensor 32 as described above, position information (θx0) indicating a rotation component centered on the x axis is calculated as follows (part of step S50). At this time, for example, if the y-coordinate (the teaching point of the robot 2) and the z-coordinate (value of the position sensor 32) at the time of contact are expressed as follows for convenience, the position sensors on the reference planes 4f and 5Af The relationship in the zy plane of the part where 32 touches is as shown in FIG.
PMa = (y1, t1)
PMb = (y2, t2)
PAa = (y3, t3)
PAb = (y4, t4)

  In the reference blocks 5A and 5B, the straight line connecting the point PMa and the point PMb and the straight line connecting the point PAa and the point PAb have a certain inclination depending on the machining accuracy and the mounting accuracy. The inclinations of the two straight lines led by the two points in the reference workpiece 4M and the reference block 5A are inclinations of the reference plane 4f with respect to the reference plane 5Af. The inclination is the rotation component θx0 around the x axis in the position information, and is derived from the trigonometric addition theorem as shown in the following equation (1).

  The obtained position information θx0 is a value serving as a reference for the same type of workpiece 4 as the reference workpiece 4M, and is stored in the control device 6.

  On the other hand, with reference to the reference block 5B, as shown in FIG. 9, the position is set at two points (PMc, PMd, PBc, PBd) set on the reference work 4M and the reference surfaces 4f, 5Bf of the reference block 5B. The sensor 32 is brought into contact, and the position at the time of contact is measured (part of step S30). As the contact position, the value of the y coordinate is kept constant, the value of the x coordinate is set according to the point, and the value of the z coordinate when the position sensor 32 contacts the reference plane is acquired. The position where the position sensor 32 is brought into contact according to the point is set according to the type of the workpiece 4 as a teaching point of the robot 2 and stored in the control device 6.

Then, position information (θy0) indicating a rotation component around the y-axis is calculated (part of step S50). Also here, the y-coordinate (the teaching point of the robot 2) and the z-coordinate (value of the position sensor 32) at the time of contact are expressed as follows for convenience.
PMc = (x5, t5)
PMd = (x6, t6)
PBc = (x7, t7)
PBd = (x8, t8)

  The inclinations of the two straight lines led by the two points in the reference workpiece 4M and the reference block 5B are inclinations of the reference plane 4f with respect to the reference plane 5Bf. The inclination is a rotation component θy0 around the y-axis in the position information, and is derived from the trigonometric addition theorem as shown in the following equation (2).

  The obtained position information θy0 is a value serving as a reference for the same type of workpiece 4 as the reference workpiece 4M, and is stored in the control device 6. Further, t1, t2, t3, and t4 and t5, t6, t7, and t8 obtained in the above operation (step S30) are stored in the control device 6 because they become the reference in the z-axis direction.

  Furthermore, as shown in FIG. 10, the camera 33 incorporated in the sensor unit 3 moves the portion where one of the reference blocks 5A and 5B (reference block 5A in the figure) and the reference workpiece 4M are adjacent to each other upward ( An image is taken from a position separated in the z direction (step S40). The number of pixels and the imaging area of the camera 33 can be appropriately selected depending on the correction accuracy. The camera 33 is connected to an image processing system (not shown), and has a general image processing system function such as edge detection and edge distance measurement in a captured image.

  For example, it is assumed that an image (contour shape on the xy plane) as shown in FIG. 11 is obtained by imaging. Then, the image processing system includes the contour EAs that intersects the x axis of the reference block 5A and the contour EAo that faces the reference workpiece 4M, and the contour EMs that intersects the x axis of the reference workpiece 4M and the contour EMO that faces the reference block 5A. To extract. Then, the angle formed by the extracted contours EAs and EMs is the position information θz0 (rotational component around the z axis), the EAo intersection with the EAs, and the intersection of the EMo EMs with the translational component in the y-axis direction. The parallel movement component in the y0 and x-axis direction is calculated as the position information xo (part of step S50).

  Since the obtained x0, y0, and θz0 are values that serve as a reference for the work target workpiece 4W of the same type as the reference workpiece 4M, they are stored in the control device 6. Here, the positional relationship between the reference workpiece 4M and the reference block 5A is obtained, but the positional relationship between the reference workpiece 4M and the reference block 5B may be obtained and the value may be stored in the control device 6 as a reference value. .

  Thereby, the generation of position information serving as a reference position for teaching the insertion operation is completed for a reference work of a certain type of work. Then, the same operation (steps S10 to S50) is repeated until the position information of the necessary type of workpiece with respect to the reference workpiece is obtained (in the case of “N” in step S60). When the position information of the necessary type of workpiece with respect to the reference workpiece is obtained (in the case of “Y” in step S60), the generation of the position information for the reference workpiece is completed.

  Next, regarding the correction of the teaching point for the workpiece 4 that is actually the work target (FIG. 4), the same type of workpiece as the reference workpiece 4M in which a square hole 4h is opened on one side of the hexahedron (not shown, but the reference workpiece 4M is not shown). In order to distinguish them from each other, an example of selecting the work target workpiece 4W) will be described.

  First, the work target workpiece 4W is installed on the installation surface 7 (step S110). At this time, the attachment position of the work target workpiece 4W does not coincide with the position where the reference workpiece 4M teaching the teaching point is installed. Therefore, the same measurement as the measurement performed on the reference workpiece 4M (steps S30 to S40) is also performed on the work target workpiece 4W (steps S120 to S130), and the position information is based on the measurement values. Displacement amounts (Δx, Δy, Δz, Δθx, Δθy, Δθz) are calculated (step S140).

Among the displacement amounts, the displacement amounts Δθx and Δθx of the rotation component around the x-axis and the rotation component around the y-axis are first inclined (rotation component) with respect to the reference blocks 5A and 5B of the work target workpiece 4W, as in step S50. θx1, θy1) are calculated. And it calculates by calculating the difference with the positional information (theta) x0 and (theta) y0 of the reference | standard workpiece | work 4M as follows (part of step S140).
Δθx = θx1-θx0
Δθy = θy1-θy0

  If the values of the position sensor 32 obtained for calculating θx1 and θy1 are t9, t10, t11, t12, and t13, t14, t15, t16, respectively, the following equation (3) or equation (4) is used. , Δz which is the displacement of the translation component in the z direction is obtained (part of step S140).

  Strictly speaking, the above expression (3) or (4) is an expression that does not hold when there is a positional shift on the xy plane of the work target workpiece 4W. However, in reality, the influence of the positional deviation on the xy plane of the work target workpiece 4W on the position in the z direction is negligible, and the correction described later is accurately performed by the displacement amount Δz obtained by the above equation. It can be carried out.

The remaining displacement amounts (Δx, Δy, Δθz) are first translated in the x-axis direction with respect to the reference block 5A of the work target workpiece 4W based on the image data obtained in step S130, as in step S50. A component x1, a translation component y1 in the y-axis direction, and a rotation component θz1 around the z-axis are calculated. And it calculates by calculating the difference with the positional information x0, y0, (theta) z of the reference | standard workpiece | work 4M as follows (part of step S140).
Δx = x1−x0
Δy = y1-y0
Δθz = θz1-θz0

  The calculated three-dimensional displacement amounts (Δx, Δy, Δz, Δθx, Δθy, Δθz) are fed back to teaching point information (IO, IP) as correction values in the tool coordinate system with the arm tip 2e as the origin. And corrected (calibrated).

  As described above, the displacement amount necessary for three-dimensional correction (the orthogonal three-axis translational components: Δx, Δy, Δz, orthogonal 3 is only required to access the workpiece 4 and the reference blocks 5A and 5B from one direction. Three rotation components of the shaft: Δθx, Δθy, Δθz) can be obtained. For this reason, it is not necessary to change the axial direction of the robot 2 when generating the position information necessary for correction, and the influence of the inherent positional deviation amount possessed by the robot 2 itself is minimized. Is possible. Furthermore, shortening of calibration time is also expected.

  In addition, since the measurement operation for obtaining the position information necessary for correction is only required to be accessed from one direction, the robot 2 to which the sensor unit 3 is attached accesses the peripheral part of the workpiece 4 from any direction. There is no need to configure it. Therefore, there is a practical advantage that the entire robot system 1 is compact, space saving and equipment cost can be reduced. Moreover, since the assembly equipment can always access the workpiece 4 from one direction, the robot system 1 can be introduced without being restricted by the equipment configuration.

  Here, the correction value is derived by one measurement operation. However, when high position accuracy is required, the accuracy of the correction value may be increased by repeating the measurement operation and the correction operation. good. In the first embodiment, an example in which a contact displacement sensor is used as the position sensor 32 has been described. However, the present invention is not limited to this. For example, a non-contact displacement sensor such as a laser displacement sensor or an ultrasonic sensor may be used. In this case, since all the sensors incorporated in the sensor unit 3 are non-contact type, even if the workpiece 4 is a flexible object, the position correction amount is derived, and the robot teaching point can be calibrated.

  As described above, according to the teaching point correction apparatus according to the first embodiment of the present invention, the teaching point (IO, IP) of the robot 2 (or robot system 1) set for the reference workpiece 4M is used as the reference workpiece. This is a teaching point correction device that corrects a work target workpiece 4W installed instead of 4M, and detects the distance to the surface of the target object and measures the position of the target object, and the direction of imaging. An imager (camera 33) that matches the direction detected by the position sensor 32 is provided, the sensor unit 3 connected to the tip of the robot 2 (arm tip 2 e), and the sensor unit 3 in cooperation with the robot 2. Based on the measurement control unit (formed in the control device 6) for controlling the measurement operation, the position data measured by the position sensor 32, and the image data acquired by the image pickup device (camera 33). A teaching point correction unit (formed in the control device 6) that calculates the amount of displacement of the work target workpiece 4W relative to the reference workpiece 4M and corrects the teaching point (IO, IP) based on the calculated amount of displacement; Reference blocks 5A and 5B serving as position references are fixed to the installation surface 7 on which one of the workpiece 4M and the work target workpiece 4W is installed, and the position sensor 32 detects the measurement control unit. While controlling the measurement operation so as to acquire position data and image data for one work 4 and the reference blocks 5A and 5B, with the direction to be set facing the installation surface 7, the teaching point correction unit Based on the image data, the displacement amount is calculated as a six-degree-of-freedom component (Δx, Δy, Δz, Δθx, Δθy, Δθz) composed of a three-axis parallel translation component and a three-axis rotation component. So to minimize the positional deviation of the robot 2, it is possible to highly accurate three-dimensional position correction.

  In particular, the teaching point correction unit calculates a biaxial rotation component (Δθx, Δθy) perpendicular to the detection direction and a parallel movement component (Δz) in the detection direction out of the components with six degrees of freedom using only the position data. Thus, only the remaining three-degree-of-freedom components (Δx, Δy, Δθz) have to be calculated from the image data. Therefore, in the image analysis, it is only necessary to extract the edge portions of the adjacent workpiece 4 and the reference block 5A as shown in FIG. 11, and the displacement amount can be easily calculated without complicated image analysis. .

  If a non-contact displacement sensor is used as the position sensor 32, three-dimensional position correction can be accurately performed even when a workpiece whose surface is easily deformed is used.

  Further, according to the teaching point correction method according to the first embodiment, the teaching point (IO, IP) of the robot 2 (or robot system 1) set for the reference workpiece 4M is set in place of the reference workpiece 4M. This is a teaching point correction method for correcting the work target workpiece 4W, wherein the reference workpiece 4M is set on the setting surface 7 on which the reference blocks 5A and 5B serving as positions are fixed, and the teaching point (IO, IP) is set. A step (steps S10 to S20), a position sensor 32 that detects the distance to the surface of the object and measures the position of the object, and an image that matches the direction in which the position sensor 32 detects the image capturing direction. The sensor unit 3 provided with a device (camera 33) is connected to the tip of the robot 2 (arm tip 2e), and the sensor unit 3 is used to connect the reference blocks 5A and 5B of the reference workpiece 4M. A first measurement step (steps S30 to S40) for measuring the position to be performed, a step (step S110) of installing the work target workpiece 4W on the installation surface 7 instead of the reference workpiece 4M, and a work target using the sensor unit 3 The second measurement process (steps S120 to S130) for measuring the position of the work 4W with respect to the reference blocks 5A and 5B, and the displacement amount of the work target work 4W with respect to the reference work 4M from the measurement results of the first measurement process and the second measurement process. And a step of correcting the teaching point (IO, IP) based on the calculated displacement amount (steps S50, S140 to S150), and the position sensor 32 detects each in the first measurement step and the second measurement step. The workpiece 4 installed on the installation surface 7 of the reference workpiece 4M and the work target workpiece 4W with the orientation to be directed toward the installation surface 7 In the process of controlling the operation of the robot 2 and the sensor unit 3 in cooperation so as to acquire the position data and image data of the quasi-blocks 5A and 5B, and correcting the teaching point (IO, IP), the position data and the image data Based on the above, the displacement amount is calculated as a component having six degrees of freedom (Δx, Δy, Δz, Δθx, Δθy, Δθz) composed of a three-axis parallel translation component and a three-axis rotation component. The positional deviation of the robot 2 is minimized, and highly accurate three-dimensional position correction is possible.

  In particular, in the process of correcting the teaching point, out of the six degrees of freedom components, the biaxial rotation component (Δθx, Δθy) perpendicular to the direction to be detected and the parallel movement component (Δz) in the direction to be detected are obtained from only the position data. Since the calculation is made, only the remaining three-degree-of-freedom components (Δx, Δy, Δθz) need be calculated from the image data. Therefore, in the image analysis, it is only necessary to extract the edge portions of the adjacent workpiece 4 and the reference block 5A as shown in FIG. 11, and the displacement amount can be easily calculated without complicated image analysis. .

1: robot system, 2: robot, 3: sensor unit (teaching point correction device),
4: Work, 4f: Main surface of work, 4M: Standard work, 4W: Work target work,
5: Reference block set, 5A: Reference block A, 5Af: Main surface of reference block A, 5B: Reference block B, 5Bf: Main surface of reference block B, 6 Control device (teaching point correction device), 7: Installation surface 8: Insertion jig 31: Sensor unit housing 32: Position sensor 33: Camera (imaging device) 34: Illumination 35: Auto tool changer
EAo, EAs: measurement object (edge) by imaging of reference block A, EMo, EMs: measurement object (edge) by imaging, PAa to PAb: measurement object (point) by position sensor on main surface of reference block A, PBc to PBd: measurement target (point) by the position sensor on the main surface of the reference block B, PMa to PMd: measurement target (point) by the position sensor on the main surface of the workpiece.

Claims (5)

A teaching point correction device for correcting a teaching point of a robot set for a reference workpiece with respect to a work target workpiece installed instead of the reference workpiece,
A position sensor that detects the distance to the surface of the object and measures the position of the object, and an imager that matches the direction in which the image is captured to the direction that the position sensor detects are provided at the tip of the robot. are connected, a sensor unit measuring operation is that access in one direction,
A measurement control unit for controlling the measurement operation by the sensor unit in cooperation with the robot;
Based on the position data measured by the position sensor and the image data acquired by the imaging device, a displacement amount of the work target workpiece with respect to the reference workpiece is calculated, and the teaching point is corrected based on the calculated displacement amount. A teaching point correction unit,
On the installation surface on which one of the reference workpiece and the work target workpiece is installed, a reference block serving as a reference for the position is fixed,
The measurement control unit controls the measurement operation so as to acquire the position data and the image data with respect to the one workpiece and the reference block while keeping a direction detected by the position sensor toward the installation surface. And
The teaching point correction unit is a six-degree-of-freedom component composed of a three-axis translation component and a three-axis rotation component orthogonal to the displacement amount based on the position data and the image data acquired from the one direction. A teaching point correction apparatus characterized by calculating.   The teaching point correction unit calculates, from the six-degree-of-freedom components, a biaxial rotation component perpendicular to the detection direction and a translation component of the detection direction only from the position data. The teaching point correction apparatus according to claim 1.   The teaching point correction apparatus according to claim 1, wherein the position sensor is a non-contact displacement sensor. A teaching point correction method for correcting a teaching point of a robot set for a reference workpiece with respect to a work target workpiece installed instead of the reference workpiece,
Installing the reference work on an installation surface on which a reference block serving as a reference of position is fixed, and setting the teaching point;
A sensor unit provided with a position sensor that detects the distance to the surface of the object and measures the position of the object, and an imager that matches the direction in which the position is detected by the position sensor. A first measurement step of measuring a position of the reference workpiece with respect to the reference block using the sensor unit connected to a tip and accessed in one direction by a measurement operation ;
In place of the reference workpiece, a step of installing the work target workpiece on the installation surface;
A second measurement step of measuring a position of the work target workpiece with respect to the reference block using the sensor unit;
Calculating a displacement amount of the work target workpiece with respect to the reference workpiece from the measurement results of the first measurement step and the second measurement step, and correcting the teaching point based on the calculated displacement amount,
In the first measurement step and the second measurement step, the workpiece installed on the installation surface among the reference workpiece and the work target workpiece, with the direction detected by the position sensor facing the installation surface, In conjunction with controlling the operation of the robot and the sensor unit so as to obtain the position data and image data of the reference block,
In the step of correcting the teaching point, on the basis of the position data and the image data acquired from the one direction, the six degrees of freedom composed of a three-axis translational component and a three-axis rotation component orthogonal to the displacement amount. A teaching point correction method comprising calculating by a component.   In the step of correcting the teaching point, out of the six-degree-of-freedom components, a biaxial rotation component perpendicular to the detection direction and a translation component of the detection direction are calculated based only on the position data. The teaching point correction method according to claim 4, wherein:

Images