JP2017056546A - Measurement system used for calibrating mechanical parameters of robot - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a measurement system capable of performing measurement in a simpler manner when performing measurement plural times for a plurality of targets for the purpose of improving positioning accuracy of a robot.SOLUTION: A measurement system 10 includes a multi-articulated robot 1 having a light receiving device 4 at a front end of an arm, and a machine tool 8, and utilizes the light receiving device 4 to measure a target 6 fixed to the machine tool 8. The robot 1 is moved such that, in a state where the machine tool 8 is positioned at each stop position and the robot 1 is positioned in the measurement position and posture, the difference between a target 6 on a light receiving surface of the light receiving device 4 and an image acquired by the light receiving device 4 falls within a predetermined error. The system stores positions and postures of the robot 1 after being moved as end points. Based on the plurality of stop positions and the plurality of end points, the system simultaneously obtains the error of the mechanical parameters of the robot 1 and the relative relationship between the robot coordinate system Σb and the machine coordinate system Σm.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a measuring apparatus including a robot and a machine tool.

  It is known to automatically calibrate mechanism parameters in order to improve positioning accuracy over the entire operating range of industrial articulated robots.

  In Patent Document 1, a light receiving device for recognizing a target is used to automatically measure the relationship between the tip position of a robot and the displacement of a controllable drive shaft. A method for determining a formalizable error is disclosed. Patent Document 1 describes that a plurality of targets are installed and measurement is performed a plurality of times for each target in order to improve accuracy over the entire operation range of the robot.

  In order to improve the positioning accuracy evenly within the robot work range, if multiple targets are installed within the robot work range and each target is measured multiple times, It is necessary to add an error to the parameter to be identified and calculated. Therefore, it is necessary to increase the number of measurements as the number of installed targets increases. For example, in the invention described in Patent Document 1, when the position of the target to be measured increases by one, the error parameter of the target position increases by three. If the number of measurement positions and postures is increased by one, the number of formulas that can be used for identification calculation increases by three. Therefore, the measurement position and posture are increased by one each time the target position is increased by one in order to cope with the increase in error parameters. There is a need.

JP 2008-012604 A JP 2005-201824 A

  There is a need for a measurement device that can perform measurement in a simpler manner when performing multiple measurements on a plurality of targets for the purpose of uniformly improving positioning accuracy within the robot's work range.

According to a first invention of the present application, an articulated robot provided with a light receiving device at the tip of an arm, and a machine tool provided within an operation range of the robot, the light receiving device is used. Measuring a target fixed to the machine tool, wherein the target is information on a position of the target imaged on a light receiving surface of the light receiving device, and between the light receiving device and the target. A geometric feature that enables identification of length information related to a distance of the machine tool, and a plurality of stop positions of the machine tool, and when the machine tool is disposed at the plurality of stop positions. A measurement position and orientation determination unit for determining a plurality of measurement positions and orientations of the robot such that the target is included in the field of view of the light receiving device; and the machine tool is stopped in the plurality of stops. And sequentially positioning the robot at the plurality of measurement positions and postures respectively corresponding to the plurality of stop positions, and imaging the target on the light receiving surface of the light receiving device. Based on information on the position of the target and information on the length, a target detection unit that detects information on a distance between the light receiving device and the target, and the machine tool is disposed at the stop position and In a state where the robot is arranged at the measurement position and posture, the difference between the position of the target on the light receiving surface of the light receiving device and a predetermined portion of the image acquired by the light receiving device is within a predetermined error. And the difference between the distance information and a predetermined value is within a predetermined error. The robot moving unit that moves the robot, and the position and posture of the robot after being moved by the robot moving unit, the end point associated with the stop position and the measurement position and posture of the machine tool An error of a mechanism parameter of the robot based on the plurality of stop positions and the plurality of end positions and the plurality of stop positions respectively corresponding to the plurality of stop positions and the plurality of measurement positions and postures; There is provided a measuring device comprising: a calculation unit that simultaneously obtains a relative relationship between the coordinate system of the robot and the coordinate system of the machine tool.
According to a second invention of the present application, in the measurement apparatus according to the first invention, the measurement position / orientation determination unit has a constant distance between the light receiving device and a predetermined point of the target, and The plurality of measurement positions and orientations are determined such that the line-of-sight inclinations of the light receiving devices passing through a fixed point are different.
According to a third invention of the present application, in the measurement apparatus according to the first invention, the measurement position / orientation determination unit has an attitude of a line of sight of the light receiving device passing through a predetermined point of the target, and the light receiving The plurality of measurement positions and postures are determined such that distances between a device and the predetermined point of the target are different.
According to the fourth invention of the present application, in the measurement apparatus according to the first invention, the measurement position and orientation determination unit is configured to automatically generate the plurality of measurement positions and orientations.
According to a fifth invention of the present application, in the measuring apparatus according to any one of the first to fourth inventions, the calculation unit performs a nonlinear problem optimization method including a Newton-Raphson method, a genetic algorithm, and a neural network. By using this, it is configured to obtain an error of the mechanism parameter of the robot.
According to a sixth invention of the present application, in the measuring apparatus according to any one of the first to fifth inventions, the light receiving device is a CCD camera configured to capture a two-dimensional image.
According to a seventh invention of the present application, in the measuring apparatus according to the first invention, the target has a circular mark, and the length information is imaged on a light receiving surface of the light receiving device. The major axis length of the ellipse corresponding to the mark is included.
According to the eighth invention of the present application, in the measuring apparatus according to any one of the first to fifth inventions, the light receiving device is a PSD configured to obtain a center of gravity of the received light amount distribution.
According to the ninth aspect of the present application, in the measuring apparatus according to the eighth aspect, the target is a light emitter.
According to a tenth invention of the present application, in the measuring apparatus according to any one of the first to ninth inventions, the calculation unit includes a position of a gazing point of the light receiving device and a mark formed on the target. Based on an error from the position of the fixed point, an error in the mechanism parameter of the robot is obtained.
According to an eleventh invention of the present application, using a measuring device including an articulated robot having a light receiving device at the tip of an arm, and a machine tool provided in the operation range of the robot, A calibration method for calibrating mechanism parameters, wherein a plurality of stop positions of the machine tool and a target fixed to the machine tool when the machine tool is disposed at the plurality of stop positions are received by the light receiving device. Determining a plurality of measurement positions and postures of the robot to be included in the field of view of the device, sequentially positioning the machine tool at the plurality of stop positions, and corresponding to the plurality of stop positions, respectively; The robot is sequentially positioned at the measurement position and posture of the image, the target is imaged on the light receiving surface of the light receiving device, and the The distance between the light receiving device and the target is detected based on the position information of the target and the length information, the machine tool is placed at the stop position, and the robot is moved to the measurement position and orientation. The difference between the position of the target on the light receiving surface of the light receiving device and a predetermined portion of the image acquired by the light receiving device is within a predetermined error, and the distance information The robot is moved so that a difference from a predetermined value is within a predetermined error, and the position and posture of the robot after moving based on the difference are set as the stop position of the machine tool and the It is stored as an end point associated with the measurement position and posture, and is stored in the plurality of stop positions and the plurality of measurement positions and postures. Based on a plurality of corresponding end points and a plurality of stop positions, an error of a mechanism parameter of the robot and a relative relationship between the coordinate system of the robot and the coordinate system of the machine tool are simultaneously obtained. A calibration method is provided.
According to the twelfth invention of the present application, in the calibration method according to the eleventh invention, when the error of the mechanism parameter of the robot is obtained, the position of the gazing point of the light receiving device and the mark formed on the target Determining an error in the mechanism parameter of the robot based on an error from the position of the predetermined point.

  These and other objects, features and advantages of the present invention will become more apparent by referring to the detailed description of exemplary embodiments of the present invention shown in the accompanying drawings.

  According to the present invention, measurement is performed on targets arranged at a plurality of positions by moving the machine tool, and the mechanism parameters of the robot are corrected. Thereby, the accuracy can be improved uniformly over the operating range of the robot with a smaller number of measurements than when a plurality of targets are installed. Further, according to the present invention, simultaneously with the error of the mechanism parameter, the error of the rotation matrix representing the relative posture relationship between the robot coordinate system and the machine tool coordinate system can be obtained.

It is the figure which showed schematically the structure of the measuring device which concerns on one Embodiment. It is the figure which showed the block configuration of the robot control apparatus which concerns on one Embodiment. It is the figure which showed the block configuration of the image processing apparatus which concerns on one Embodiment. It is a flowchart which shows the outline of the process of the machine tool performed according to one Embodiment. It is a flowchart which shows the outline of the process of the robot performed according to one Embodiment. It is explanatory drawing relevant to the process in step R7 of the flowchart of FIG. It is a figure which shows the relative positional relationship of a camera and a target. It is a figure which shows the monitor screen corresponding to FIG. 7A. It is a figure which shows the relative positional relationship of a camera and a target. It is a figure which shows the monitor screen corresponding to FIG. 8A. It is a flowchart which shows the outline of the process in step R8 of the flowchart of FIG.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. The components of the illustrated embodiment are appropriately scaled to assist in understanding the present invention. The same reference numerals are used for identical or corresponding components.

  FIG. 1 is a diagram schematically illustrating a configuration of a measurement apparatus 10 according to an embodiment. The measuring device 10 includes an articulated robot 1 and a machine tool 8 provided within the operation range of the robot 1. As shown in FIG. 1, the robot 1 is a robot having any known configuration including an arm 1a and a base 1b. The robot 1 is connected to a robot control device 5 that controls the robot 1. The camera 4 is attached to the tip of the arm 1a, that is, the tool attachment surface 32.

  The robot 1 is set with a robot coordinate system Σb and a mechanical interface coordinate system Σf. The robot coordinate system Σb is a coordinate system fixed to the base 1b of the robot 1, that is, fixed to the work space. The mechanical interface coordinate system Σf is a coordinate system that is fixed to the tool mounting surface 32 and that changes its position and posture according to the operation of the robot 1. The robot controller 5 is configured to constantly acquire the position and orientation of the origin of the mechanical interface coordinate system Σf.

  A known teaching operation panel 18 having operation keys is connected to the robot control device 5. The operator can operate the robot 1 by manually operating the operation keys.

  A machine tool coordinate system Σm is set for the machine tool 8. The machine tool 8 includes known operation keys, and an operator can operate the machine tool 8 by operating the operation keys, and the machine tool 8 of the target 6 fixed to the machine tool 8 can be operated. The amount of change in position in the machine coordinate system Σm can be acquired.

  The camera 4 is a CCD camera, for example, and is a known light receiving device having a function of detecting a two-dimensional image on the light receiving surface (on the CCD array surface) by imaging. The camera 4 is connected to an image processing apparatus 2 having a monitor 3 made of LCD or CRT. In the present embodiment, the camera 4 is used to image the mark 7 of the target 6 fixed to the machine tool 8. In one embodiment, the target 6 may be a light emitter that includes a light source. In this case, a PSD (Position Sensitive Detector) configured to obtain the center of gravity of the received light amount distribution is used as the light receiving device.

  The robot controller 5 has a known block configuration as shown in FIG. That is, the robot control device 5 has a memory 12, a teaching operation panel interface 13, and an input / output interface for an external device connected to a bus 17 connected to a main CPU (hereinafter simply referred to as "CPU") 11. The face 16, the servo control unit 15, and the communication interface 14 are connected in parallel. The memory 12 includes a RAM, a ROM, a nonvolatile memory, and the like.

  The teaching operation panel 18 connected to the teaching operation panel interface 13 has a known configuration including a display. By manually operating the teaching operation panel 18, the operator can execute creation, correction and registration of the robot operation program, setting of various parameters, reproduction operation of the taught operation program, jog feed, and the like.

  A system program for controlling basic functions of the robot 1 and the robot controller 5 is stored in the ROM of the memory 12. The robot operation program taught in accordance with the application and the related setting data are stored in the nonvolatile memory of the memory 12. In addition, the nonvolatile memory of the memory 12 also stores programs, parameters, and other data for various processes described later (processing for moving the robot when obtaining the mechanism parameters and communicating with the image processing apparatus for that purpose). Stored.

  The RAM of the memory 12 is used for temporarily storing data related to various processes executed by the CPU 11. The servo control unit 15 includes servo controllers # 1 to #n (n is the total number of axes of the robot, for example, n = 6). The servo control unit 15 outputs a torque command to the servo amplifiers A1 to An based on the movement command and a feedback signal received from a pulse coder (not shown) provided on each axis. The movement command is created by a known method in accordance with arithmetic processing for controlling the robot 1 (trajectory plan creation and interpolation and inverse transformation based thereon).

  Each of the servo amplifiers A1 to An supplies current to the servo motors of the respective axes based on the corresponding torque commands and drives them. The communication interface 14 is connected to the image processing apparatus 2 (see FIG. 1). Via the communication interface 14, commands related to measurement and measurement result data, which will be described later, are transmitted and received between the image processing apparatus 2 and the robot control apparatus 5.

  The image processing apparatus 2 has a known block configuration as shown in FIG. The image processing apparatus 2 has a CPU 20 composed of a microprocessor. The CPU 20 has a ROM 21, an image processor 22, a camera interface 23, a monitor interface 24, an input device (I / O) 25, a frame memory (image memory) 26, a nonvolatile memory 27, via a bus line 30. A RAM 28 and a communication interface 29 are connected to each other.

  The camera interface 23 is connected to a camera 4 (for example, a CCD camera) that is an imaging unit. The camera 4 has an electronic shutter function, and is configured to execute shooting in response to a shooting command received via the camera interface 23. Video data photographed by the camera 4 is stored in the frame memory 26 in the form of a gray scale signal via the camera interface 23.

  The monitor interface 24 is connected to the monitor 3 (see FIG. 1). The monitor 3 displays an image being captured by the camera 4, a past image stored in the frame memory 26, an image processed by the image processor 22, and the like as necessary.

  As shown in FIG. 1, the camera 4 attached to the tool attachment surface 32 of the robot 1 images the mark 7 of the target 6 fixed to the machine tool 8. The video signal of the mark 7 stored in the frame memory 26 is analyzed using the image processor 22 and its two-dimensional position, size, etc. are obtained (details will be described later). The mark 7 makes it possible to specify the position information of the target 6 that forms an image on the light receiving surface of the camera 4 and the length information related to the distance between the camera 4 and the target 6. It is.

  An analysis program and parameters necessary for image processing are stored in the nonvolatile memory 27. The RAM 28 is used for temporarily storing data related to various processes executed by the CPU 20. The communication interface 29 is connected to the robot control device 5 via the communication interface 14 of the robot control device 5.

  Referring again to FIG. 1, the line of sight 40 of the camera 4 is drawn as a straight line extending from the representative point of the camera 4 (for example, the center of the camera lens) toward the target 6. A coordinate system Σv shown in FIG. 1 is a coordinate system representing the line of sight 40. The coordinate system Σv is set so that the origin is located on the line of sight 40 and one coordinate axis (for example, the Z axis) coincides with the line of sight 40.

  Note that the mechanical interface coordinate system Σf is a coordinate system that represents the position and orientation of the tool mounting surface 32 as described above, but is also a coordinate system that represents the position and orientation of the robot 1 in this specification. . That is, unless otherwise specified, the “robot position” means the position of the origin of the mechanical interface coordinate system Σf in the robot coordinate system Σb. When the “robot position” includes the posture of the robot, the “robot position” means the position and posture of the origin of the mechanical interface coordinate system Σf in the robot coordinate system Σb.

  The measuring apparatus 10 according to the present embodiment detects the mark 7 of the target 6 in order to calibrate the mechanism parameter of the robot 1. As shown in FIG. 1, the robot control device 5 includes a measurement position and orientation determination unit 51, a positioning unit 52, a target detection unit 53, a robot moving unit 54, a robot end point storage unit 55, and a calculation unit 56. It is equipped with.

  The measurement position / orientation determination unit 51 includes a plurality of stop positions of the machine tool 8 and the robot 1 in which the target 6 is included in the field of view of the camera 4 when the machine tool 8 is disposed at the plurality of stop positions. A plurality of measurement positions and postures are determined.

  The measurement position and orientation determination unit 51 determines a plurality of measurement positions and orientations such that the distance between the camera 4 and the predetermined point of the target 6 is constant and the inclination of the line of sight of the camera 4 passing through the predetermined point is different. It may be configured as follows.

  The measurement position / posture determination unit 51 has a plurality of measurement position / postures in which the posture of the line of sight of the camera 4 passing through a predetermined point of the target 6 is constant and the distance between the camera 4 and the predetermined point of the target 6 is different. May be configured to determine.

  The measurement position and orientation determination unit 51 may be configured to automatically generate a plurality of measurement positions and orientations.

  The positioning unit 52 sequentially positions the machine tool 8 at a plurality of stop positions, and sequentially positions the robot 1 at a plurality of measurement positions and postures respectively corresponding to the plurality of stop positions.

  The target detection unit 53 forms an image of the target 6 on the light receiving surface of the camera 4 and detects information on the distance between the camera 4 and the target 6 based on the position information and length information of the target 6. To do.

  The robot moving unit 54 positions the target 6 on the light receiving surface of the camera 4 and the image acquired by the camera 4 in a state where the machine tool 8 is disposed at the stop position and the robot 1 is disposed at the measurement position and posture. The robot 1 is moved so that the difference from the predetermined location is within a predetermined error, and the difference between the distance information and the predetermined value is within a predetermined error.

  The robot end point storage unit 55 stores the position and posture of the robot 1 after being moved by the robot moving unit 54 as an end point associated with the stop position, measurement position, and posture of the machine tool 8.

  Based on the plurality of end points and the plurality of stop positions respectively corresponding to the plurality of stop positions and the plurality of measurement positions and postures, the calculation unit 56 determines the error of the mechanism parameter of the robot 1, the robot coordinate system Σb, and the machine tool coordinates. The relative relationship with the system Σm is simultaneously obtained.

  The calculation unit 56 may be configured to obtain an error of the mechanism parameter of the robot 1 by using a nonlinear problem optimization method including a Newton-Raphson method, a genetic algorithm, and a neural network.

  A mechanism parameter calibration method executed in the measurement apparatus according to the present embodiment will be described with reference to the flowcharts of FIGS. 4 and 5. FIG. 4 is a flowchart showing an outline of processing of the machine tool 8, and FIG. 5 is a flowchart showing an outline of processing of the robot 1.

  In step M1, the target 6 is fixed to the machine tool 8. In step R1, the camera 4 is attached to the tool mounting surface 32 of the robot 1. The target 6 and the camera 4 do not need to be installed at accurate positions, but are fixed so that the positions do not change during measurement.

  In step M2, a relationship between the moving direction and moving distance of the target 6 on the image and the moving direction and moving distance of the machine tool 8 is obtained. In step R2, a relationship between the moving direction and moving distance of the target 6 on the image and the moving direction and moving distance of the robot 1 is obtained. These relationships are used to determine how the robot 1 or machine tool 8 can be moved to move the target 6 on the image in the intended direction by the intended distance. For example, each axis of the robot 1 or the machine tool 8 is detected by translating the robot 1 or the machine tool 8 a plurality of times and detecting the feature values of the figure of the target 6 imaged on the light receiving surface of the camera 4 each time. The relationship between the movement direction and the movement distance and the change in the characteristic value of the target 6 is obtained.

In the present embodiment, the measurement is performed at n _i (i = 1, 2,..., M) measurement positions and postures of the robot 1 with respect to m stop positions of the machine tool 8. The number of the stop position of the machine tool 8 is represented by an index “i”. The stop position of the i-th machine tool in the machine tool coordinate system Σm represented by ^M p _i.

In Step M3, “1” is input to the index i. In step M4, the operator moves the machine tool 8 to the initial stop position ^M p ₁ . The machine tool 8 records ^M p ₁ and transmits a movement completion signal to the robot controller 5.

When the robot control device 5 receives the movement completion signal (step R3), the process proceeds to step R4 to determine whether or not the index i is “1”. When the index i is “1”, the process proceeds to step R5 so that the mark 7 of the target 6 falls within the field of view of the camera 4 when the machine tool 8 is located at the stop position ^M p ₁ (for example, The operator moves the robot 1 so that the mark 7 appears on the monitor 3 of the image processing apparatus 2.

In step R6, when the machine tool is arranged at the stop position ^M p _1, and ^B p _t (the initial position of the target 6) a position of a predetermined point of the mark 7 in the robot coordinate system .SIGMA.b, mechanical interface coordinate system Σf The position ^F p _S of the gazing point 31 of the camera 4 is automatically measured.

In order to execute the measurement in step R6, for example, a known technique described in Patent Document 2 can be used. According to this known technique, measurement is performed from a plurality of directions (with a plurality of robot postures) with the gazing point 31 of the camera 4 aligned with a predetermined point of the mark 7 in the robot coordinate system Σb, and the position ^F of the gazing point 31 is determined. and calculates the position ^B p _t of a predetermined point of p _S and the mark 7. If the gaze point 31 is replaced with a tool tip point (TCP (Tool Center Point)) attached to the hand of the robot, this method can mechanically touch up a predetermined point in the robot coordinate system from multiple directions. , Similar to the conventional method of calculating the position of a predetermined point of TCP and mark 7.

  At this stage, it is not necessary to calculate the positions of the camera 4 and the target 6 with high accuracy. In other words, it is sufficient if the target 6 does not deviate from the field of view of the camera 4 when the posture of the robot 1 is changed around the gazing point 31 with the gazing point 31 coincident with a predetermined point of the target 6. .

In step R7, n _i measurement positions and postures ^B P ′ _{i, j} (j = 1, 2,...) With respect to the position of the target 6 when the machine tool 8 is disposed at the stop position ^M p _i. , N _i ), and the robot controller 5 records the measurement position and orientation ^B P ′ _{i, j} (j = 1, 2,..., N _i ).

In this embodiment, the measurement position and orientation ^B P ′ _{i, j} (j = 1, 2,..., N _i ) are automatically calculated as described later. First, when the index i is “1”, the n positions and postures of the robot 1 relative to the position of the target 6 when the machine tool is placed at the stop position ^M p ₁ are calculated in step R6. and the initial position ^B p _t of the target 6 in the coordinate system .SIGMA.b, a position ^F p _S of the focus point 31 of the camera 4 in the mechanical interface coordinate system .SIGMA.f, based on, automatically calculated and recorded.

  As a first example, the n positions and postures are a plurality of positions where the distance between the camera 4 and the predetermined point of the mark 7 is constant and the inclination of the line of sight 40 of the camera 4 passing through the predetermined point is different. And may include posture. Alternatively, as a second example, the n positions and postures are plural in which the posture of the line of sight 40 of the camera 4 passing through the predetermined point of the mark 7 is constant and the distance between the camera 4 and the predetermined point is different. May include the position and posture.

  FIG. 6 shows three positions and postures in which the distance between the camera 4 and the predetermined point of the mark 7 is constant and the inclination of the line of sight 40 of the camera 4 passing through the predetermined point is different according to the first example. Yes.

  In order to automatically calculate n positions and postures, the position and posture of the robot 1 obtained by aligning the gazing point 31 of the camera 4 with the target 6 in step R6 is set as a basic position and posture. Then, a predetermined value may be added to the attitude angle of the camera 4 while keeping the position of the gazing point 31 of the camera 4 constant. Thereby, a plurality of positions and postures included in the operation range of the robot 1 are automatically generated.

  Alternatively, the range of the posture angle of the camera 4 corresponding to the movement range of the robot 1 is calculated with the position of the gazing point 31 of the camera 4 being fixed, and any of a plurality of angles ( For example, a method of generating a plurality of positions and postures with respect to an angle obtained by equally dividing the posture angle range may be employed.

In this way, after automatically calculating and recording the n positions and postures of the robot with respect to the position of the target 6 when the machine tool 8 is placed at the stop position ^M p ₁ , the index i is “1”. When n ₁ positions and postures are selected from the n positions and postures, n ₁ of the robot 1 with respect to the position of the target 6 when the machine tool 8 is located at the stop position ^M p ₁ is selected. Measurement position and orientation ^B P ′ _{1, j} (j = 1, 2,..., N ₁ ) and recorded.

In step R8, n _i measurement positions and postures ^B P ′ _{i, j} (j = 1, 2, _j ) of the robot 1 with respect to the position of the target 6 when the machine tool 8 is disposed at the stop position ^M p _i . .., N _i ) as starting points, the “automatic touch-up process” is executed in order. The position and orientation ^B P _{i, j} (j = 1, 2,..., N _i ) of the robot 1 at the end of the “automatic touch-up process” are recorded as end points, and the robot controller 5 performs “automatic touch-up process”. "End signal is transmitted to the machine tool 8." In the “automatic touch-up process”, the robot 1 is moved so that the value or parameter indicating the geometric feature of the shape of the mark 7 formed on the light receiving surface of the camera 4 satisfies a predetermined condition.

  Details of the “automatic touch-up process” will be described with reference to FIGS. 7A, 7B, 8A, 8B, and 9. FIG. For example, the mark 7 has a circular shape in which a cross line indicating the center position is drawn, and the geometric feature value of the target 6 imaged on the light receiving surface of the camera 4 is the center position of the ellipse. And the long axis length (see FIG. 7B). Usually, since the light receiving surface of the camera 4 is inclined with respect to the plane of the mark 7 of the target 6, the circular mark 7 appears as an ellipse on the image. The mark 7 may be a shape other than a circle, or a character or a symbol.

  When the positions of the camera 4 and the target 6 are shifted (see FIG. 7A), as shown in FIG. 7B, an elliptical mark image 7a formed on the light receiving surface is a predetermined point (for example, It is deviated from the center M) of the light receiving surface, and the major axis length of the mark image 7a is shorter than the diameter of the mark 7.

  Therefore, as shown in FIGS. 8A and 8B, by moving the robot 1, the center of the mark image 7a coincides with the center M of the light receiving surface, and the major axis length of the mark image 7a is the mark 7 The relative positional relationship between the camera 4 and the target 6 is changed so as to match the diameter.

  More specifically, the difference between the center of the mark image 7a and the center M of the light receiving surface on the light receiving surface, and the difference between the major axis length of the mark image 7a and the diameter of the mark 7 are within a predetermined error. The robot 1 is automatically moved. Since the feature value of the mark image 7a can be detected by image processing on the light receiving surface, the above processing can be executed according to the detection result.

  With reference to FIG. 9, an “automatic touch-up process” for moving the mark image 7a captured by the camera 4 to the center M on the image and adapting the size of the image will be described.

  In step S <b> 1, the mark 7 is imaged by the camera 4. Thereby, for example, a mark image 7a shown in FIG. 7B is obtained.

  In step S2, the image processing device 2 detects the position and size of the figure of the mark 7 on the image (for example, the center position of the mark image 7a on the screen and the long axis length of the mark image 7a).

In step S3, the position and size of the figure of the target 6 obtained in step S2 are respectively equal to a predetermined point (for example, the center point M of the light receiving surface) and a predetermined length (for example, the diameter of the mark 7) on the image. Determine whether you are doing it. Specifically, the distance between the center point M and the center of the mark image 7a and the difference between the major axis length of the mark image 7a and the diameter of the mark 7 are all determined in advance that is acceptable on the image. If the error is equal to or smaller than the threshold distance δ _image, it is determined that the position and size of the target 6 “match” the predetermined point and the predetermined length, and the process is terminated.

When at least one of the distance difference and the difference between the major axis length and the diameter is larger than the threshold distance δ _image , the position and size of the target 6 are “mismatched” with respect to the predetermined point and the predetermined length. And the process proceeds to step S4. The distance on the image can be measured, for example, by replacing it with the number of square “pixels” corresponding to the distance.

  In step S4, the mark image 7a is moved to the center M on the image, and a robot translation command is created so that the long axis length of the mark image 7a matches a predetermined value. Here, the robot translation command means a movement command for moving the robot 1 without changing the posture of the robot 1, that is, the posture of the mechanical interface coordinate system Σf, in the robot coordinate system Σb fixed in space. .

  In step S5, the robot 1 is moved in accordance with the robot translation command described above. When the movement of the robot 1 is completed, the process returns to step S1. Then, the processes in steps S1 to S5 are repeated until “match” is determined in step S3.

The robot position (end point) at the time when the above-mentioned “automatic touch-up process” is completed is the distance between the center position of the mark 7 on the image and the predetermined point, the long axis length of the mark image 7a and the predetermined size. The difference between them is less than the threshold distance δ _image .

Referring to FIG. 4 again, when the machine tool 8 receives the “automatic touch-up process” end signal in step M5, the process proceeds to step M6. In step M6, it is determined whether or not the index i is equal to “m”. If the index i is smaller than “m”, the index i is incremented by “1” in step M7, and the process returns to step M4. In step M 4, the operator moves the machine tool 8 to the next stop position ^M p _i , and the machine tool 8 records the stop position ^M p _i and transmits a movement completion signal to the robot controller 5.

Referring to FIG. 5, when the robot controller 5 receives the movement completion signal in step R3, the process proceeds to step R4. In Step R4, it is determined whether or not the index i is equal to “1”. When the index i is larger than “1”, the process proceeds to step R7, where n _i measurement positions and postures ^B P ′ _{i, j} (j with respect to the target position when the machine tool 8 is disposed at the stop position ^M p _i = 1, 2,..., N _i ). In the present embodiment, the measurement position and orientation are automatically calculated using the n positions and orientations recorded in step R7 when the index i is “1”.

Next, a method for calculating the measurement position and orientation will be described. First, the relationship between the movement of the target 6 and the movement of the machine tool 8 on the image obtained in step M2, and the movement between the movement of the target 6 and the robot 1 on the image obtained in step R2. the relationship, based on the obtained rotation matrix ^B R _M representing the relative orientation relationship between the robot coordinate system Σb a machine tool coordinate system .SIGMA.m.

At this stage, the rotation matrix ^B R _M may not be calculated with high accuracy. For example, the target in the case, the rotation matrix ^B R _M calculated robot coordinate system Σb from the target 6 is a machine tool 8 from the state of being included in the field of view of the camera 4 by translational movement to move the target 6 6 It is sufficient that the robot 1 is translated by the movement direction and the movement distance, and the target 6 does not deviate from the field of view of the camera 4.

Then, when the index i is “1”, n _i positions and orientations are selected from the n positions and orientations recorded in step R 7, and only ^B R _M ( ^M p _i ^−M p ₁ ) is selected. the position and orientation obtained by parallel movement, n _i pieces of measurement position and orientation ^B P _'i of the robot 1 relative to the position of the target 6 when the machine tool 8 is arranged at the stop position ^M p _{i, j (j} = 1, 2,..., N _i ). The robot controller 5 records the measurement position and orientation ^B P ′ _{i, j} (j = 1, 2,..., N _i ).

In step _{^{R8, B P 'i, j}} (j = 1,2, ···, n i) to execute the "automatic touch-up process" in turn as a starting point. The robot position / posture ^B P _{i, j} (j = 1, 2,..., N _i ) at the end of the “automatic touch-up process” is recorded as the end point, and the robot controller 5 ends the “automatic touch-up process”. A signal is transmitted to the machine tool 8.

  When the machine tool 8 receives the “automatic touch-up process” end signal (step M5), the process proceeds to step M6 to determine whether or not the index i is equal to “m”. When the index i is smaller than “m”, the process returns to step M7, and the processes of steps M7, M4, R3, R4, R7, R8, M5, and M6 are repeatedly executed.

  If it is determined in step M6 that the index i is equal to “m”, this means that the movement of the machine tool 8 to all the stop positions has been completed. In this case, the process proceeds to step M8, and the machine tool 8 transmits a measurement end signal to the robot control device 5.

  When the index i is “m”, if step R3 is executed before the process of step M4 is executed, or if step R9 is executed before the process of step M8 is executed, step R10 is executed. , R9, R3 are repeatedly executed.

In step R9, when the robot control device 5 receives the measurement end signal, the mechanism parameters of the robot 1 and the robot coordinate system Σb are determined based on the position and posture of the robot 1 at each stop position and each end point of the machine tool 8. Calculate and update the rotation matrix with the machine tool coordinate system Σm. Specifically, in step R11, the position of the gazing point 31 and the position of the target 6 in the robot coordinate system Σb coincide with each other at each stop position and end point of the machine tool 8 in the robot coordinate system Σb. , The error of _p mechanism parameters P _{K, 0} = [P _{k, 0,1} , P _{k, 0,2} ,..., P _{k, 0, p} ] and the mechanical interface of the robot The error of the position ^F p _S = [X _S , Y _S , Z _S ] of the gazing point 31 in the coordinate system Σf and the mark 7 in the robot coordinate system Σb when the machine tool 8 is located at the stop position ^M p ₁ the position of a predetermined point ^{_{B p t = [X t,}} Y t, Z t] and the error of identifying calculate the error of the rotation matrix ^B R _M between the robot coordinate system Σb a machine tool coordinate system .SIGMA.m. The rotation matrix ^B R _M can be expressed using three variables R = [φ, θ, ψ] according to Euler angles or the like.

Kinematic parameters P _{K, 0,} position ^F p _S of the gaze point 31, and the position ^B p _t of a predetermined point of the mark 7, the variable parameter P _ID = error in _{R [ΔP k, ΔP S,} ΔP t, ΔR] And The error parameter P _ID is identified and calculated as follows.

Position ^B p _t of the target 6 is the position of the target 6 in the robot coordinate system Σb when machine 8 is arranged at the stop position ^M p _1. Therefore, the position of the target 6 in the robot coordinate system Σb when the machine tool 8 is located at the stop position ^M p _i is based on the movement distance ( ^M p _i ^−M p ₁ ) of the machine tool 8 using the formula: ^B p _t + ^B R _M - is determined by calculating the _{^{(M p i M p 1)}} .

And "automatic touch-up processing" position and posture (end point) ^B P i of the robot _{end, j (j = 1,2, ···} , n i) a vector representing the position of the ^B p _i, and _j, posture ^B R i a rotation matrix representing _the, when _j, the position of the gaze point 31 in the robot coordinate system Σb uses the position ^F p _S of the gaze point 31 in the mechanical interface coordinate system .SIGMA.f, wherein: ^B p _{i, j} + ^B R _i, is determined by calculating the _j ^F p _S.

In an ideal state, that is, when all components of the error parameter P _ID are zero and there is no error in positioning of the machine tool 8, each stop position ^M _pi (i = 1, 2,. ·, m) for the position ^B p _i of the gaze point 31 in the robot coordinate system ^{_{Σb, j + B R i,}} j F p S (j = 1,2, ···, n i) is the robot coordinate system .SIGMA.b will correspond to the ^{_{- (M p 1 M p i}} ) position ^B p _t + ^B R _M of the target 6 in.

In contrast to an ideal state in which no error occurs, in practice, between the position of the gazing point 31 in the robot coordinate system Σb and the position of the target 6 according to the error parameter P _ID including the mechanism parameter or rotation matrix error. Has a position error e _{i, j} = ( ^B p _{i, j} + ^B R _{i, j} ^F p _S ) − ( ^B p _t + ^B R _M ( ^M p _i ^−M p ₁ )). As can be seen from the geometrical relationship, when the value of each drive axis of the robot 1 is fixed at a constant value, ^B p _{i, j} + ^B R _{i, j} ^F p _S is a function of P _ID. . Therefore, e _{i, j} is also a function of P _ID .

Using Newton Raphson method, the stop position of the machine tool 8, the position ^B p i of the gaze point 31 in the robot coordinate system Σb corresponding to the position and posture of the robot 1 in each _{^{endpoint, j + B R i, j}} F p _S and the target position in the robot coordinate system _{^{Σb B p t + B R M}} (M p i - M p 1) error between _{E = [e 1,1, ···,} e i, j, ···, P _ID such that e _{m, nm} ] takes the minimum value is obtained. Then, the robot 1 is updated by replacing the mechanism parameter of the robot 1 with the calculated value.

  According to the calculation system according to the present embodiment, the position of the target 6 fixed to the machine tool 8 is obtained from the moving distance of the machine tool 8 using the high positioning accuracy of the machine tool 8. Therefore, it is not necessary to increase the measurement position and posture each time the position of the target 6 is increased. Specifically, in the present embodiment, although the error parameter of the rotation matrix increases by three, the parameter to be identified and calculated does not increase even if the number m of target positions to be measured is increased. Therefore, when measurement is performed for two target positions (m = 2), the number of measurement positions and orientations necessary for identification calculation is the same as when a plurality of targets are installed. However, when measurement is performed for three or more target positions (m> 2), the number of measurement positions and postures required for identification calculation can be reduced by m-2 compared to the case where a plurality of targets are installed. .

  Thus, according to the present embodiment, the positioning accuracy of the robot 1 can be improved uniformly in the operation range of the robot 1 in a simple manner. Further, simultaneously with the mechanism parameter error, the error of the rotation matrix between the robot coordinate system Σb and the machine tool coordinate system Σm can be obtained.

  Although various embodiments of the present invention have been described above, those skilled in the art will recognize that the functions and effects intended by the present invention can be realized by other embodiments. In particular, the constituent elements of the above-described embodiments can be deleted or replaced without departing from the scope of the present invention, or known means can be further added. In addition, it is obvious to those skilled in the art that the present invention can be implemented by arbitrarily combining features of a plurality of embodiments disclosed explicitly or implicitly in this specification.

DESCRIPTION OF SYMBOLS 1 Robot 2 Image processing apparatus 3 Monitor 4 Camera 5 Robot control apparatus 6 Target 7 Mark 8 Machine tool 10 Measuring apparatus 11 Main CPU
12 Memory 13 Interface for teaching operation panel 14 Communication interface (robot controller side)
15 Servo Control Unit 16 Input / Output Interface for External Device 17 Bus 18 Teaching Operation Panel 20 CPU
21 ROM
22 Image Processor 23 Camera Interface 24 Monitor Interface 25 Input Device 26 Frame Memory 27 Nonvolatile Memory 28 RAM
29 Communication interface (image processing device side)
30 Bus line 31 Gaze point 32 Tool mounting surface 40 Line of sight 51 Measurement position / orientation determination unit 52 Positioning unit 53 Target detection unit 54 Robot movement unit 55 Robot end point storage unit 56 Calculation unit

Claims (12)

An articulated robot having a light receiving device at the tip of an arm, and a machine tool provided within an operation range of the robot, and a target fixed to the machine tool using the light receiving device A measuring device for measuring,
The target is a geometric feature that enables identification of information on a position of the target imaged on a light receiving surface of the light receiving device and information on a length related to a distance between the light receiving device and the target. Have
A plurality of stop positions of the machine tool, and a plurality of measurement positions and postures of the robot such that the target is included in the field of view of the light receiving device when the machine tool is disposed at the plurality of stop positions. And a measurement position and orientation determination unit for determining
A positioning unit that sequentially positions the machine tool at the plurality of stop positions and sequentially positions the robot at the plurality of measurement positions and postures respectively corresponding to the plurality of stop positions;
The target forms an image on the light receiving surface of the light receiving device, and detects information on the distance between the light receiving device and the target based on the position information and the length information of the target. A detection unit;
In a state where the machine tool is disposed at the stop position and the robot is disposed at the measurement position and posture, the position of the target on the light receiving surface of the light receiving device and a predetermined portion of an image acquired by the light receiving device A robot moving unit that moves the robot so that a difference between the distance information and a predetermined value is within a predetermined error.
A robot end point storage unit for storing the position and posture of the robot after being moved by the robot moving unit as an end point associated with the stop position of the machine tool and the measurement position and posture;
Based on the plurality of end points and the plurality of stop positions respectively corresponding to the plurality of stop positions and the plurality of measurement positions and postures, the mechanism parameter error of the robot, the robot coordinate system, and the machine tool And a calculation unit for simultaneously obtaining a relative relationship with the coordinate system.   The measurement position and orientation determination unit includes a plurality of measurement position and orientations such that a distance between the light receiving device and a predetermined point of the target is constant, and a line-of-sight inclination of the light receiving device passing through the predetermined point is different. The measuring device of claim 1, wherein the measuring device is configured to determine   The measurement position / orientation determination unit is configured such that the attitude of the line of sight of the light receiving device passing through the predetermined point of the target is constant and the distance between the light receiving device and the predetermined point of the target is different. The measurement device according to claim 1, wherein the measurement device is configured to determine a measurement position and orientation.   The measurement apparatus according to claim 1, wherein the measurement position and orientation determination unit is configured to automatically generate the plurality of measurement positions and orientations.   5. The calculation unit according to claim 1, wherein the calculation unit is configured to obtain an error of a mechanism parameter of the robot by using a nonlinear problem optimization method including a Newton-Raphson method, a genetic algorithm, and a neural network. The measuring apparatus of any one of Claims.   The measurement device according to claim 1, wherein the light receiving device is a CCD camera configured to capture a two-dimensional image.   The measurement according to claim 1, wherein the target has a circular mark, and the length information includes a major axis length of an ellipse corresponding to the mark imaged on a light receiving surface of the light receiving device. apparatus.   The measuring device according to claim 1, wherein the light receiving device is a PSD configured to obtain a center of gravity of a received light amount distribution.   The measuring device according to claim 8, wherein the target is a light emitter.   The calculation unit is configured to obtain an error of a mechanism parameter of the robot based on an error between a position of a gazing point of the light receiving device and a position of a predetermined point of a mark formed on the target. Item 10. The measuring device according to any one of Items 1 to 9. A calibration method for calibrating mechanism parameters of a robot using a measuring device including an articulated robot having a light receiving device at the tip of an arm and a machine tool provided within an operation range of the robot. And
A plurality of stop positions of the machine tool, and a target fixed to the machine tool when the machine tool is disposed at the plurality of stop positions, and included in a field of view of the light receiving device. Determine multiple measurement positions and postures,
Sequentially positioning the machine tool at the plurality of stop positions, and sequentially positioning the robot at the plurality of measurement positions and postures respectively corresponding to the plurality of stop positions;
The target is imaged on the light receiving surface of the light receiving device, and information on the distance between the light receiving device and the target is detected based on information on the position of the target and information on the length,
In a state where the machine tool is disposed at the stop position and the robot is disposed at the measurement position and posture, the position of the target on the light receiving surface of the light receiving device and a predetermined portion of an image acquired by the light receiving device And the robot is moved so that the difference between the distance information is within a predetermined error, and the difference between the distance information and the predetermined value is within a predetermined error,
Storing the position and posture of the robot after moving based on the difference as an end point associated with the stop position and the measurement position and posture of the machine tool;
Based on the plurality of end points and the plurality of stop positions respectively corresponding to the plurality of stop positions and the plurality of measurement positions and postures, the mechanism parameter error of the robot, the robot coordinate system, and the machine tool A calibration method including simultaneously obtaining a relative relationship with a coordinate system of   When determining the error of the mechanism parameter of the robot, the error of the mechanism parameter of the robot is determined based on the error between the position of the gazing point of the light receiving device and the position of the predetermined point of the mark formed on the target. The calibration method according to claim 11, further comprising:

Images