JP2017026479A - Position specification device and positioning method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technology that does not restrict a degree of working freedom regarding a device employing means for specifying a position, and allows a simple structure to specify a position relative to a measurement workpiece.SOLUTION: An output member position specification device comprises: a stage member that has a concavity or convexity-shaped three-dimensional shape part provided on a top surface; an output member that is oppositely arranged on the top surface of the stage member; movement means that moves the output member on the stage member; distance measurement means that is provided in the output member, and measures each of a distance to the top surface of the stage member, and a distance to the three-dimensional shape part; and position specification means that specifies a position of the output member relative to the three-dimensional shape part on the basis of a detection result of the distance measurement means.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a means for specifying a relative position from a reference plane with respect to a certain member.

  Conventionally, in applications that teach a positional relationship between an industrial robot output member (Tool Center Position) and a jig, or to calibrate a CMM that measures a member in three dimensions, the position of the member It is done to identify.

  There are various device specifications in each field, but when adopting the position identification method, a method that can be realized under conditions such as the specifications specified for the device, the components provided, and the degree of freedom of operation should be selected. It is.

  In order to realize the position specifying means described above as an example, a technique that is generally used is a technique in which a member is operated with respect to its reference and measurement is performed using a measuring means.

  Regarding this method, there are various methods such as manual and automatic. The most primitive method is a method in which a member is directly operated by a human operation and a position is taught according to a target point on a reference.

  This method has the advantage that it can be recognized by humans so that the reference position is at a level that can be recognized by humans, and it can be flexibly handled even if the standard changes. is there.

  However, there is a possibility that the accuracy may vary greatly depending on the person to be operated, and it takes time and effort, and the work time will be different.

  Next, as a technique for eliminating the accuracy and speed difference caused by passing a person, there is a technique in which the apparatus itself performs teaching or calibration with respect to the reference of the member by automatically measuring the member with respect to the reference.

  This method has the advantage that the higher the accuracy of the measuring means, the higher the accuracy of the position teaching, and the repeated result within the accuracy difference of the apparatus itself. On the other hand, in this method, it is necessary to provide a specific pattern with respect to target coordinates on a reference, and it is necessary that the member or the device itself has a function for recognizing the pattern.

  Naturally, the pattern is desired to be a general and simple pattern that is easy to process and accurate, and is simpler and less expensive than those that require a large processing capacity or expensive in the function. It is desirable to be.

  Patent Documents 1 to 3 introduced below are examples of apparatuses that realize the position specifying means in accordance with each configuration and application.

JP2011-177845A JP 2011-52599 A JP 2012-83192 A

  With respect to Patent Document 1, an industrial robot recognizes a mark on a reference position using a camera installed on an output member for a mark having a plurality of light emitters at a reference position whose distance is relatively fixed. Then, a calibration method is introduced in which the attitude of the output member is adjusted based on the measured values.

  By adopting a light emitter for the mark, the illumination of the camera itself is not necessary, and the coordinates of the mark itself are taught in advance to the extent that the camera viewing angle is entered, thereby enabling an automatic calibration operation.

  An apparatus target in Patent Document 1 is a scalar robot and there is one camera for imaging, and this is an example in which the specifying means is realized for the purpose of performing calibration in a two-dimensional space.

  However, the technique of Patent Document 1 still has a problem that it cannot be applied as a calibration method using a more general-purpose robot. In addition, high accuracy is expected in the alignment by imaging the target point and the output member with a camera, but there is a problem that a half-side high-precision image processing apparatus is expensive and the system itself is complicated.

Patent Document 2 also introduces a method of automatically calibrating a robot output member with reference to a jig operated by an industrial robot. By using a contact-type positioning jig installed on the output member with a jig that has a target value stored in advance, the three planes of the jig are brought into contact with each other, and the position is detected. This is an example in which the specifying means is realized by a method of correcting a coordinate system having a connected point as an origin.

  In the method of Patent Document 2, high-precision calibration can be performed regardless of the skill level of the operator, but a robot capable of contacting three planes in parallel has at least five degrees of freedom in terms of performance. The problem that it is necessary and the device is not versatile remains.

  With respect to Patent Document 3, a calibration method for a three-dimensional measuring instrument is introduced. In order to register the coordinate system of the rotary table of the three-dimensional measuring device, a reference is provided on the table surface, and a spherical probe is used as a means for measuring. On the other hand, the reference is a shape that makes three-point contact with the probe, and the measuring instrument realizes the relative position specifying means by specifying the three-dimensional position of the probe tip from the angular position of the three points. is there.

  In the method of Patent Document 3, the reference position is stored in advance by the device so that calibration work can be performed without human intervention. However, it is necessary to unify the curvatures of the three reference spheres related to accuracy. In addition, there is a problem that the positioning and installation process of the reference itself is complicated and sophisticated because it is necessary to perform positioning so as to contact the spherical sensor provided in the probe evenly.

  The present invention relates to an apparatus that employs a position specifying means, and provides a technique for specifying a position with a simple structure with respect to a measurement member without limiting the degree of freedom of operation.

The position specifying device of the present invention includes a stage member having a concave or convex three-dimensional shape portion provided on the upper surface, an output member disposed opposite to the upper surface of the stage member, and moving the output member on the stage member Based on the detection results of the distance measuring means, the distance measuring means provided on the output member, and the distance measuring means for measuring the distance to the upper surface of the stage member and the distance to the three-dimensional shape portion, respectively. Position specifying means for specifying the position of the output member with respect to the three-dimensionally shaped portion.
According to this aspect of the present invention, it is possible to specify a position with a simple structure with respect to the measurement member without limiting the degree of freedom of operation with respect to an apparatus that employs a position specifying means.

  Further, in the position measurement (identification) method of the present invention, in a plane that serves as a reference for identifying a member for which a certain position is to be identified, the plane has at least one pattern, and the pattern includes: Ranging means having an axis perpendicular to the plane, the bottom surface is a right cone or right frustoconical protrusion or depression in contact with the plane, and the member is capable of measuring a distance in one direction; A moving unit capable of moving in parallel to the plane; using the moving unit, the moving unit moves to a position where the pattern does not exist on the plane; and a distance perpendicular to the plane from the member Is obtained by measuring the distance perpendicular to the plane from the member by moving to a position estimated as the pattern on the plane using the moving means. Distance Is D1, and from D0, the distance perpendicular to the plane from the apex or top of the pattern where D1 was measured is added if the pattern is a protrusion, and subtracted if the pattern is a depression When the distance obtained as D2 is D2, D2, and the angle between the side of the pattern where the D1 is measured and the axis, the axis of the pattern and the plane of the measurement position P1 where the D1 is measured A parallel separation distance L1 is obtained.

  In the present invention, in order to further reduce the separation distance from the axis of the pattern, first, after obtaining the separation distance L1 according to claim 1, the member is further moved to the pattern using the moving means. The distance obtained by using the position measurement method according to claim 1 is L2, the measurement position where L2 is obtained is P2, the L1 and L2 are compared, and second, , If L1 is greater than L2, identify P2 as the position PL2 closer to the axis of the pattern; and third, if L1 is less than L2, move the member to P1 using the moving means; When the distance L2 ′ obtained by using the moving means is moved to a position different from the position where L2 is measured and the position measuring method is further compared with L1, and L1 is larger than L2 ′ Is L2 ' The measured measurement position is specified as a position PL2 closer to the axis of the pattern, and when L1 is smaller than L2 ′, the member is moved again to P1 using the moving means, and L2 is replaced with L2 ′. By repeating a process equivalent to the third process, a position PL2 closer to the axis of the pattern than the measurement position at which L1 is measured is specified, and the more specified PL2 is set as a new L1 in the first process. By repeating the above, it is possible to make the separation distance parallel to the axis of the pattern and the plane of the member closer.

  Further, the present invention may be characterized in that alignment on the three-dimensional coordinates with respect to the plane of the member is performed using the separation distance L1 and the distance D0 and the alignment method.

  In the present invention, a reference plane is prepared for a member whose position is to be specified. The member has means (hereinafter, measuring means) capable of measuring a distance in one direction, and the plane has a plurality of patterns for the member to measure. The member further has means (hereinafter referred to as moving means) that can move in parallel with the plane, and by using this means, it is possible to operate up to all the patterns existing on the plane. . The pattern has an axis perpendicular to the plane, and the bottom surface is a right cone or right frustoconical protrusion or depression in contact with the plane. The plurality of patterns are all assumed to have the same shape. A process for specifying the three-dimensional position of the member using these means and features will be described below. First, the moving means is used to move onto the plane where the pattern does not exist, and after arrival, the distance in the direction perpendicular to the plane is measured using the measuring means. The distance obtained by this processing is set as the first distance. A distance perpendicular to the plane from the apex (or upper base) of the pattern to the plane, a first distance and subtraction when the pattern is a protrusion, and a first when the pattern is a depression The distance obtained by performing addition with the distance is taken as the second distance. Third, the moving means is used to move on the plane estimated to be on the pattern, and after arrival, the distance to the plane perpendicular to the pattern is measured using the measuring means. The distance obtained by this processing is the third distance. From the second distance, the third distance, and the angle obtained from the side and the axis of the pattern from which the third distance was measured, with respect to the plane of the measurement position at which the third distance was measured from the axis of the pattern Since a parallel distance is obtained, a two-dimensional position parallel to the plane can be specified from the center of the pattern. Furthermore, the distance information in the vertical direction with respect to the plane is added from the information on the first distance, and the three-dimensional position of the member with respect to the plane can be specified.

  The present invention relates to an apparatus that employs position specifying means, and can specify the position with a simple structure with respect to the measurement member without limiting the degree of freedom of operation. By using the above-described method, the degree of freedom of movement only needs to be able to move in a direction parallel or perpendicular to the plane. Since it is only necessary to have a reference plane and measuring means only in one direction, it can be realized with a simple structure.

The figure which shows the outline | summary of this discovery. FIG. 3 is a structural diagram related to the robot of the present invention. The figure showing the calibration means of a Z direction in the Example of this invention. The figure showing the calibration process of a Z direction in the Example of this invention. The figure showing the determination method of a Z direction and a type | formula in the calibration of the Z direction of FIG. The figure showing the calibration means of a XY direction in the Example of this invention. The figure showing the calibration process of a XY direction in the Example of this invention. FIG. 8 is a diagram illustrating measurement and determination formulas in the XY calibration in FIG. 7. FIG. 8 is a diagram illustrating processing for moving a measurement point within a threshold range in the XY calibration in FIG. 7. FIG. 8 is a diagram illustrating end processing in calibration in the XY direction in FIG. 7. The figure showing the other robot structure in the Example of this invention.

  Hereinafter, the present invention will be described in detail based on embodiments. Specifically, with respect to the alignment method according to the present invention, an embodiment in which the steps and processes are embodied will be described with reference to the drawings. It should be noted that the following embodiments are merely examples, and the structure specifically adopted in the content does not limit the scope of the claims.

  In this example, the calibration method regarding the position of the industrial robot and its output member is adopted as a specific example and introduced below.

  FIG. 2 shows the configuration of the robot in the case of this embodiment. The robot main body A200 is a serial link robot having six degrees of freedom in this example, and a distance sensor used for calibration of the robot main body is installed on the output member A205. The robot main body has a reference plane (upper surface of the stage member) A101, and four patterns (a concave or convex three-dimensional shape portion) A102 are formed on the reference plane. The coordinate system in this example is a three-dimensional orthogonal coordinate system represented by XYZ, and Z = 0 on the reference plane, that is, the reference of the XY plane. The robot body A200 is also installed on the XY plane.

  The robot body A200 further includes a communication unit A201 for obtaining an operation target point from the external information terminal A204, information taught from the external information terminal, a correction value for the output member A205, a coordinate value for performing calibration, and an operation procedure. A storage unit A203 for holding, and a control unit A202 for controlling the robot from information in the storage unit. In carrying out the present embodiment, first, a measurement position group to be performed later is stored in the storage unit through the communication unit from the external information terminal.

  FIG. 3 shows a calibration method for the reference plane A101 in the Z direction of the output member A205. In calibrating the position of the output member of the robot main body A200 with respect to the reference plane, the calibration starts with distance calibration in the Z direction with respect to the reference plane of the output member. A distance sensor A100 (displacement sensor employing a laser in this example) A100 has a function of measuring a distance in one direction by contact, and operates with the output member.

  The control unit A202 reads out the target coordinate value for performing the calibration in the Z direction stored in the storage unit A203 in advance with respect to the reference plane with respect to the reference plane, and measures the output member in parallel with the reference plane. Move to part A300. At this time, each measurement unit A300 is on the reference plane not on the pattern. After the movement is completed, the distance sensor is irradiated to perform measurement. This operation is repeatedly measured at four locations, and the aggregated set A302 of the measured values A301 is stored in the storage unit.

  FIG. 4 shows a calibration process for the reference plane A101 in the Z direction of the output member A205. First, here, n is designated by a value of n = 0 to 3 (S400), and the output member is moved to the target coordinates parallel to the reference plane (S401). First, it is determined whether or not the output member has been correctly moved to the target coordinates (S402). If there is a problem, the process proceeds to error processing (S406). When the movement is completed without any problem, the distance perpendicular to the reference plane is measured using the distance sensor A100 at that location (S403), and the measurement result is determined (S404). Regarding this determination, there is a threshold value of the measurement value in the storage unit A203, and the control unit A202 performs comparison from the value. (See FIG. 5, A501, A502) If there is no problem with the measured value, the value is stored in the storage unit (S405), and the process proceeds to the next target coordinate.

  FIG. 5 shows a calculation formula for determining a correction value for the actually measured value. In the measurement in this example, it is assumed that all the measurement values are included in S501 and S503, and this threshold value is determined with respect to the required accuracy of the robot.

  First, in the process of S404, the measurement result is within an allowable range by comparing the upper limit determination threshold value S501 and the lower limit determination threshold value S503 with reference to the reference value S502 with respect to the Z measurement value in the storage unit A203. Check if it is in.

  When all four points are within the allowable range, an average is taken from the measured values of these four points, and the difference is set as a correction value in the Z direction. By performing the above processing, the distance in the Z direction perpendicular to the reference plane of the output member is determined (S504), and the subsequent processing is based on the distance, and the output member is relative to the reference plane. And move in parallel.

  FIG. 6 shows a calibration method for the reference plane A101 in the XY direction of the output member A205.

  Four patterns A102 for calibration in the X and Y directions are provided on the reference plane in this example, and a distance perpendicular to the reference plane is measured using the distance sensor A100 with respect to this pattern. I do.

  The pattern has a right cone shape (A602) that is recessed perpendicular to the XY plane. For example, the radius is 5 mm and the height is 10 mm. It is desirable to change this shape according to the positioning accuracy of the robot body. Furthermore, when the angle of inclination of the conical side is small, for example, when measuring with a distance sensor using a laser, it is difficult to receive reflected light. Since the value may be violated, it is desirable to adopt a shape whose tip is not too acute. The target coordinate in this pattern is the apex of the cone, which is the measurement distance at A601. However, in actuality, a slightly shifted place such as A600 is measured.

  FIG. 7 shows a calibration process for the reference plane A101 in the XY direction of the output member A205.

  First, n is designated by a value of n = 0 to 3 (S700), and the output member is moved in parallel to the reference plane to the target value designated by n (S701). First, it is determined whether or not the output member has been correctly moved to the moving position (S702). If there is a problem, the process proceeds to error processing (S709). When the movement is completed without any problem, the distance perpendicular to the reference plane is measured using the distance sensor A100 at that place (S703), and the control unit A202 determines the measurement result (S704). With respect to this determination, it is determined whether or not the measurement location is included in the pattern A102, and the measurement value measured by the distance sensor is the reference from the output member and the reference plane obtained in step S504. If the result measured in S703 is included within a distance ± (fixed value) perpendicular to the plane, the measured value is determined to be outside the pattern, and the process proceeds to error processing S709. The fixed value is determined by the positioning accuracy of the robot body.

  If there is no problem in the measurement location in S704, in step S705, the control unit A202 measures the distance D1 measured perpendicularly to the reference plane between the output member and the pattern and the vertex of the pattern from the output member. The distance D2 perpendicular to the reference plane and the angle θ at which the axis and side of the pattern intersect each other, and the distance D parallel to the plane between the axis of the pattern being measured and the measurement location Ask for.

  A detailed calculation method in S705 is shown in FIG. Next, it is determined whether or not the separation distance D is within the separation distance threshold (S706). The separation distance threshold, like the separation distance D, represents a distance parallel to the reference plane centered on the axis of the pattern being measured, and the value changes depending on the accuracy of the robot body A200. If it is outside the threshold value, the process shown in FIG. 9 is repeated until it falls within the threshold value (S707). If the measured value falls within the threshold value, it is stored in the storage unit (S708), and the process proceeds to the next target coordinate.

  FIG. 8 shows a procedure for obtaining a separation distance parallel to the plane between the axis of the pattern being measured and the measurement location in the step of S705.

  At the measurement point A800 of the actual measurement value A600 and the conical vertex A801 of the pattern A102, the angle between the vertex of the cone and the reference line extending perpendicularly to the reference plane A101 is θ, and from the measurement point A800 to the vertex A801 of the cone. The distance in the XY plane is A600 (r) and A601 (R).

  r is obtained in step S703, and R is a right cone stored in the storage unit A203 in advance with respect to a distance perpendicular to the reference plane from the reference plane and the output member obtained in step S504. It is obtained by adding a distance perpendicular to the base from the apex of the shape A602.

  From R and r determined above, and the angle θ, the distance A 802 (D) parallel to the reference plane of the axis from the vertex of A 801 and the measurement point A 800 is [D = tan θ * (R · r )].

  As described above, the output member that is currently measuring A800 specifies that the control unit A202 exists at a distance D from the target coordinate A801 parallel to the reference plane by a distance D. That is, the control unit A202 serves as a position specifying unit that specifies the position of the output member with respect to the pattern A102 that is a three-dimensionally shaped part based on the detection result of the distance sensor.

  FIG. 9 shows detailed steps in S706. This step is carried out when the location to be actually measured is included in the mark but the threshold value determined in advance in the storage unit A203 has not been reached.

  The threshold value here is a distance parallel to the plane from A801, and is expressed as A902. First, fine operation A900 is performed in a certain direction from the measurement point (S900). This fine motion is a value sufficiently small with respect to the accuracy of the robot main body A200, and the minimum motion resolution is preferable. As the order of the operation direction, in this example, the operation is performed in the order of -Y, + X, + Y, starting with -X (A901). Measurement is performed after the operation (S901), and if the measured value becomes large, it is approaching the center point which is the target point, and therefore the fine operation is further advanced in the same direction. When the value becomes smaller, the operation proceeds to the next direction (S902). This operation is performed until the output member A205 is finally included in A902 by performing the operation as described above and continuing the operation of determining whether it is within the range again in FIG. 7 or S705.

  FIG. 10 shows a step of determining the correction value in the XY direction after the step until the output member A205 is within the determination reference in the XY direction is completed.

  The deviation amount between the target coordinate and the measurement coordinate in XY at each measurement point calculated in the steps of FIGS. 7 to 8 is averaged to calculate a correction amount (S1000). The correction value is stored in the storage unit A203, and the control unit A202 again performs an operation taking the correction value into account (S1001).

  If it is within the predetermined threshold values at all measurement points, the calibration is terminated as it is, but if there is an out of range, the process from measurement to correction value determination is repeated again (S1002). In this example, the number of repetitions is 4 (S1003), but this number is not limited. Through this step, the final correction value in the XY plane is determined.

  From the finally determined correction values for X, Y, and Z, the calibration with the accuracy error of the cylindrical upper A1000 with a fixed radius Z and having a radius as a threshold in the XY plane is completed.

  FIG. 12 shows that the present invention is adopted in configurations other than this example. In this example, calibration was performed in the XY coordinates with respect to the reference plane A101 using a conical depression as a target coordinate using a distance sensor. As another example, for example, the positional relationship between the robot main body A200 and the reference plane is As shown in A1200 and A1201, calibration can be performed on the reference plane without any problem on any plane.

  Measurement can also be performed using a physical measuring instrument such as A1202. However, in that case, the shape of the protrusion should be selected as A1203. With regard to the shape of the protrusion, it is not a perfect conical shape, and if it is devised, for example, by making the measurement threshold range a flat surface, it becomes easier to measure, and the determination process can be omitted. However, since this method is expected to be accompanied by a decrease in speed during measurement due to physical contact, measurement without contact is preferable.

  The present invention is not limited to the above-described embodiments and examples, but is shown by the description of the scope of claims, and further includes all modifications within the meaning and scope equivalent to the description of the scope of claims.

A100 Distance sensor A101 Reference plane of the robot body A102 Calibration mark A103 Calibration mark shape A200 Robot body A201 Robot body communication unit A202 Robot body control unit A203 Robot body memory unit A204 External information processing terminal A205 Robot output member A300 Height measurement location from reference plane A301 Distance sensor value A302 Measured value A600 Actual measurement value A601 Ideal measurement value A602 Target coordinate A800 Measurement position A801 Target coordinate A802 Distance from measurement position to target coordinate A803 Distance calculation formula A900 Fine Motion direction A901 Fine motion direction search order A902 Tolerance from target coordinates A1000 Accuracy range after calibration A1100 Robot body configuration example 1
A1101 Robot body configuration example 2
A1102 Measurement method example A1103 Measurement target

Claims (4)

A stage member provided with a concave or convex three-dimensional shape portion on the upper surface;
An output member disposed opposite to the upper surface of the stage member;
Moving means for moving the output member on the stage member;
A distance measuring means provided on the output member, each for measuring a distance from the upper surface of the stage member and a distance from the three-dimensional shape part;
An output member position specifying device comprising: position specifying means for specifying a position of the output member with respect to the three-dimensional shape portion based on a detection result of the distance measuring means. In the plane that serves as a reference for specifying the position and the member that wants to specify a position,
The plane has at least one pattern,
The pattern has an axis perpendicular to the plane;
The bottom surface is a right cone or right frustoconical protrusion or depression in contact with the plane,
The member is a distance measuring means capable of measuring a distance in one direction;
Moving means capable of moving parallel to the plane;
Using the moving means, move to a position where the pattern does not exist on the plane, and a value obtained by measuring a distance perpendicular to the plane from the member is D0,
D1 is a distance obtained by moving to the position estimated as the pattern on the plane using the moving means, and measuring a distance perpendicular to the plane from the member,
From D0, the distance perpendicular to the plane from the apex or upper base of the pattern where D1 is measured is added when the pattern is a protrusion, and subtracted when the pattern is a depression. When the distance is D2, the distance parallel to the plane between the axis of the pattern and the measurement position P1 at which D1 is measured is determined from the angle between D1, D2, and the side of the pattern at which D1 is measured and the axis. A position measuring method characterized by obtaining a distance L1.   In order to further reduce the separation distance from the axis of the pattern, first, after obtaining the separation distance L1 according to claim 1, the member is further made smaller than P1 in the pattern by using the moving means. The distance measured by using the position measuring method according to claim 1 is L2, the measurement position where L2 is obtained is P2, L1 and L2 are compared, and secondly, L1 is greater than L2. If it is larger, P2 is specified as a position PL2 closer to the axis of the pattern. Third, if L1 is smaller than L2, the member is moved to P1 by using the moving means, and the position where L2 is measured The distance L2 ′ obtained by using the moving means is moved to a position different from the position by using the moving means, and the distance L2 ′ obtained by using the position measuring method is further compared with L1. When L1 is larger than L2 ′, L2 ′ is Obtained measurement position When the position PL2 closer to the axis of the pattern is specified and L1 is smaller than L2 ′, the member is moved again to P1 by using the moving means, and L2 is replaced with L2 ′. By repeating the same process, the position PL2 closer to the pattern axis than the measurement position where L1 is measured is specified, and the more specified PL2 is set as a new L1, and the process is repeated from the first process. An alignment method, wherein a separation distance parallel to an axis of the pattern and the plane of the member is made closer. Using the separation distance L1 and the distance D0 according to claim 2 and the alignment method according to claim 2, alignment of the member on a three-dimensional coordinate with respect to the plane is performed. Alignment method.

Images