Image analysis device, control device, mechanical system, image analysis method, and computer program product for image analysis

Abstract

Between the shape of a workpiece after being machined by an industrial machine and the target shape of the workpiece, errors may occur due to various factors. Conventionally, a technique for easily determining the causative factors of such errors has been required. An image analysis device (150) includes: a first image generation unit (152) that generates first image data representing a first distribution of the generation positions in the workpiece of the error between the shape of the workpiece after being machined by an industrial machine (50) and the target shape of the workpiece prepared in advance; a second image generation unit (158) that generates second image data representing a second distribution of the generation positions in the workpiece of the error between the instruction sent to the industrial machine (50) for machining the workpiece and the feedback from the industrial machine (50) corresponding to the instruction; and a correlation acquisition unit (162) that obtains the correlation between the first distribution and the second distribution based on the first image data and the second image data.

Description

Technical Field

[0001] The present invention relates to an image analysis device, a control device, a mechanical system, an image analysis method, and a computer program product for image analysis. Background Art

[0002] There is known a device that generates an image of the movement locus of a tool of an industrial machine and displays positions on the locus where a position deviation has occurred on the image (for example, Patent Document 1).

[0003] Prior Art Documents

[0004] Patent Documents

[0005] Patent Document 1: Japanese Unexamined Patent Application Publication No. 2011-022688 Summary of the Invention

[0006] Problems to be Solved by the Invention

[0007] Between the shape of a workpiece machined by an industrial machine and the target shape of the workpiece, an error may occur due to various factors. Conventionally, a technique for easily determining the cause of such an error has been required.

[0008] Means for Solving the Problems

[0009] In one aspect of the present disclosure, an image analysis device includes: a first image generation unit that generates first image data representing a first distribution of the generation positions of errors between the shape of a workpiece machined by an industrial machine and the target shape of the workpiece prepared in advance in the workpiece; a second image generation unit that generates second image data representing a second distribution of the generation positions of errors between an instruction sent to the industrial machine for machining the workpiece and the feedback from the industrial machine corresponding to the instruction in the workpiece; and a correlation acquisition unit that obtains the correlation between the first distribution and the second distribution based on the first image data and the second image data.

[0010] In other aspects of the present disclosure, in an image analysis method, first image data representing a first distribution of the generation positions of errors between the shape of a workpiece machined by an industrial machine and the target shape of the workpiece in the workpiece is generated, second image data representing a second distribution of the generation positions of errors between an instruction sent to the industrial machine for machining the workpiece and the feedback from the industrial machine corresponding to the instruction in the workpiece is generated, and the correlation between the first distribution and the second distribution is obtained based on the first image data and the second image data.

[0011] In still another aspect of the present disclosure, a computer program causes a computer to function as the following for image analysis: a first image generation unit that generates first image data representing a first distribution of locations in a workpiece where errors occur between the shape of the workpiece after being machined by an industrial machine and a target shape of the workpiece prepared in advance; a second image generation unit that generates second image data representing a second distribution of locations in the workpiece where errors occur between an instruction sent to the industrial machine for machining the workpiece and feedback from the industrial machine corresponding to the instruction; and a correlation acquisition unit that obtains a correlation between the first distribution and the second distribution based on the first image data and the second image data.

[0012] Advantageous Effects of the Invention

[0013] According to the present disclosure, an operator can determine, by considering the correlation between the first distribution and the second distribution, whether there is a high likelihood that the cause of the error between the shape of the machined workpiece and the target shape lies in the error between the instruction and the feedback. Therefore, it is easy for the operator to identify the cause of the error between the shape of the machined workpiece and the target shape. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a block diagram of a mechanical system according to an embodiment.

[0015] Figure 2 is a perspective view of an industrial machine according to an embodiment.

[0016] Figure 3 represents an example of a workpiece Figure 2 machined by the industrial machine shown.

[0017] Figure 4 represents an example of a workpiece model showing the target shape of the workpiece.

[0018] Figure 5 represents an example of a trajectory model formed by the movement trajectory of the industrial machine.

[0019] Figure 6 is Figure 5 an enlarged view of region VI in

[0020] Figure 7 is in Figure 5 a diagram showing the distribution of errors between the instruction and the feedback in the first drive unit on the trajectory model shown.

[0021] Figure 8 is in Figure 5 a diagram showing the distribution of errors between the instruction and the feedback in the second drive unit on the trajectory model shown.

[0022] Figure 9 is inFigure 5 A graph showing the distribution of the error between the command and the feedback in the third drive unit on the trajectory model shown.

[0023] Figure 10 It is Figure 5 A graph showing the distribution of the error between the command and the feedback in the fourth drive unit on the trajectory model shown.

[0024] Figure 11 It is Figure 5 A graph showing the distribution of the error between the command and the feedback in the fifth drive unit on the trajectory model shown.

[0025] Figure 12 It is a graph that arranges and displays the two-dimensional images of the region S1 in Figure 4 and the two-dimensional image of the region S1 in Figure 7 .

[0026] Figure 13 It is Figure 4 a graph that overlays and displays the sequential image data in the image data shown.

[0027] Figure 14 It is Figure 4 a graph that overlays and displays the recognition image data in the image data shown.

[0028] Figure 15 It is a block diagram of a mechanical system in another way.

[0029] Figure 16 It is a block diagram of an image analysis device in another embodiment. Detailed Embodiment

[0030] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In addition, in the various embodiments described below, the same reference numerals are assigned to the same elements, and duplicate descriptions are omitted. First, with reference to Figure 1 , a mechanical system 10 of an embodiment will be described. The mechanical system 10 includes: a control device 12, a display device 14, an input device 16, a measurement device 18, and an industrial machine 50.

[0031] The control device 12 controls the operations of the display device 14, the input device 16, the measurement device 18, and the industrial machine 50. Specifically, the control device 12 is a computer having a processor 20, a memory 22, and an I / O interface 24. The processor 20 has a CPU or a GPU, etc., and performs arithmetic processing for executing various functions described later. The processor 20 is communicably connected to the memory 22 and the I / O interface 24 via a bus 26.

[0032] The memory 22 has a ROM, a RAM, etc., and temporarily or permanently stores various data. The I / O interface 24 communicates with an external device under the instruction from the processor 20, receives data from the external device, and sends data to the external device. In the present embodiment, the display device 14, the input device 16, the measurement device 18, and the industrial machine 50 are connected to the I / O interface 24 in a communicable manner by wireless or wired means.

[0033] The display device 14 has an LCD, an organic EL display, etc. The processor 20 sends image data to the display device 14 via the I / O interface 24 to cause the display device 14 to display an image. The input device 16 has a keyboard, a mouse, a touch sensor, etc., and sends the information input by the operator to the processor 20 via the I / O interface 24. In addition, the display device 14 and the input device 16 may be integrally provided in the control device 12 or may be provided separately from the control device 12.

[0034] The measurement device 18 is a laser scanning type 3D scanner, a 3D measuring machine having a stereo camera, etc., and measures the 3D shape of an object such as a workpiece W described later. The measurement device 18 sends the measurement data of the shape of the measured object to the processor 20 via the I / O interface 24.

[0035] The industrial machine 50 processes the workpiece W. Hereinafter, with reference to Figure 2 , an industrial machine 50 according to an embodiment will be described. The industrial machine 50 according to the present embodiment is a so-called 5-axis machining center and includes: a base table 52, a translation movement mechanism 54, a support table 56, a swing movement mechanism 58, a swing member 60, a rotation movement mechanism 62, a workpiece table 64, a spindle head 66, a tool 68, and a spindle movement mechanism 70.

[0036] The base table 52 has a bottom plate 72 and a pivot support portion 74. The bottom plate 72 is a substantially rectangular flat plate-shaped member and is disposed above the translation movement mechanism 54. The pivot support portion 74 is integrally formed on the bottom plate 72 so as to protrude upward from the upper surface 72a of the bottom plate 72.

[0037] The translation movement mechanism 54 reciprocates the base table 52 in the x-axis direction and the y-axis direction of the machine coordinate system CM. Specifically, the translation movement mechanism 54 includes: an x-axis ball screw mechanism (not shown) that reciprocates the base table 52 in the x-axis direction; a y-axis ball screw mechanism (not shown) that reciprocates the base table 52 in the y-axis direction; a first drive unit 76 that drives the x-axis ball screw mechanism; and a second drive unit 78 that drives the y-axis ball screw mechanism.

[0038] The first drive unit 76 is, for example, a servo motor, and rotates its rotating shaft according to an instruction from the control device 12. The x-axis ball screw mechanism converts the rotational motion of the output shaft of the first drive unit 76 into a reciprocating motion along the x-axis of the machine coordinate system CM. Similarly, the second drive unit 78 is, for example, a servo motor, and rotates its rotating shaft according to an instruction from the control device 12. The y-axis ball screw mechanism converts the rotational motion of the output shaft of the second drive unit 78 into a reciprocating motion along the y-axis of the machine coordinate system CM.

[0039] The support table 56 is fixed to the base worktable 52. Specifically, the support table 56 has a base portion 80 and a drive unit housing portion 82. The base portion 80 is a substantially quadrangular prism-shaped hollow member, and is fixed to the upper surface 72a of the bottom plate 72 so as to protrude upward from the upper surface 72a. The drive unit housing portion 82 is a substantially semicircular hollow member, and is integrally formed at the upper end portion of the base portion 80.

[0040] The swing movement mechanism 58 has a third drive unit 84 and a speed reducer 86. The third drive unit 84 and the speed reducer 86 are provided inside the base portion 80 and the drive unit housing portion 82. The third drive unit 84 is, for example, a servo motor, and rotates its output shaft according to an instruction from the control device 12. The speed reducer 86 reduces the rotational speed of the output shaft of the third drive unit 84 and transmits it to the swing member 60. In this way, the swing movement mechanism 58 rotates the swing member 60 about the axis A1.

[0041] The swing member 60 is supported by the support table 56 and the pivot support portion 74 so as to be rotatable about the axis A1. Specifically, the swing member 60 has a pair of holding portions 88 and 90 that are opposed to each other in the x-axis direction, and a drive unit housing portion 92 that is fixed to the holding portions 88 and 90. The holding portion 88 is mechanically connected to the swing movement mechanism 58 (specifically, the output shaft of the third drive unit 84). On the other hand, the holding portion 90 is pivotally supported by the pivot support portion 74 via a support shaft (not shown). The drive unit housing portion 92 is a substantially cylindrical hollow member, and is disposed between the holding portions 88 and 90 and integrally formed with the holding portions 88 and 90.

[0042] The rotation movement mechanism 62 has a fourth drive unit 94 and a speed reducer 96. The fourth drive unit 94 and the speed reducer 96 are provided inside the drive unit housing portion 92. The fourth drive unit 94 is, for example, a servo motor, and rotates its output shaft according to an instruction from the control device 12. The speed reducer 96 reduces the rotational speed of the output shaft of the fourth drive unit 94 and transmits it to the workpiece table 64.

[0043] Thus, the rotation moving mechanism 62 is rotated to rotate the workpiece table 64 about the axis A2. The axis A2 is orthogonal to the axis A1 and rotates about the axis A1 together with the swing member 60. The workpiece table 64 is a substantially disk-shaped member and is disposed above the drive unit housing portion 92 so as to be rotatable about the axis A2. The workpiece table 64 is mechanically connected to the rotation moving mechanism 62 (specifically, the output shaft of the fourth drive unit 94), and a workpiece W is provided on the workpiece table 64 via a jig (not shown).

[0044] The spindle head 66 is provided so as to be movable in the z-axis direction of the machine coordinate system CM, and a tool 68 is detachably attached to the front end thereof. The spindle head 66 drives the tool 68 to rotate about the axis A3, and the workpiece W provided on the workpiece table 64 is machined by the rotating tool 68. The axis A3 is orthogonal to the axis A1.

[0045] The spindle moving mechanism 70 includes a ball screw mechanism 98 that reciprocates the spindle head 66 in the z-axis direction and a fifth drive unit 100 that drives the ball screw mechanism 98. The fifth drive unit 100 is, for example, a servo motor, and rotates its rotation shaft according to an instruction from the control device 12. The ball screw mechanism 98 converts the rotation motion of the output shaft of the fifth drive unit 100 into a reciprocating motion along the z-axis of the machine coordinate system CM.

[0046] A machine coordinate system CM is set in the industrial machine 50. The machine coordinate system CM is fixed in a three-dimensional space and is an orthogonal coordinate system that serves as a reference when automatically controlling the operation of the industrial machine 50. In the present embodiment, the machine coordinate system CM is set such that its x-axis is parallel to the rotation axis A1 of the swing member 60 and its z-axis is parallel to the vertical direction.

[0047] The industrial machine 50 relatively moves the tool 68 in five directions with respect to the workpiece W provided on the workpiece table 64 by the translation moving mechanism 54, the swing moving mechanism 58, the rotation moving mechanism 62, and the spindle moving mechanism 70. Therefore, the translation moving mechanism 54, the swing moving mechanism 58, the rotation moving mechanism 62, and the spindle moving mechanism 70 constitute a moving mechanism 102 that relatively moves the tool 68 and the workpiece W.

[0048] As Figure 1 shown, the industrial machine 50 further includes a first sensor 104, a second sensor 106, a third sensor 108, a fourth sensor 110, and a fifth sensor 112. The first sensor 104 is disposed on the first drive unit 76, detects state data of the first drive unit 76, and transmits it to the control device 12 as feedback FB1.

[0049] For example, the first sensor 104 has a rotation detection sensor (such as an encoder or a Hall element) that detects the rotation position R1 (or rotation angle) of the output shaft of the first drive unit 76. In this case, the first sensor 104 detects the rotation position R1 and the speed V1 of the first drive unit 76 as the state data of the first drive unit 76. The speed V1 can be obtained by taking the first derivative of the rotation position R1 with respect to time t (V1 = δR1 / δt). The first sensor 104 sends the position feedback indicating the rotation position R1 and the speed feedback indicating the speed V1 to the control device 12 as feedback FB1.

[0050] In addition, the first sensor 104 has a current sensor that detects the current EC1 flowing through the first drive unit 76. The first sensor 104 detects the current EC1 as the state data of the first drive unit 76 and sends the current feedback indicating the current EC1 to the control device 12 as feedback FB1.

[0051] Similarly, the second sensor 106 has a rotation detection sensor that detects the rotation position R2 of the output shaft of the second drive unit 78 and a current sensor that detects the current EC2 flowing through the second drive unit 78, and detects the rotation position R2, the speed V2 (= δR2 / δt), and the current EC2 as the state data of the second drive unit 78. And the second sensor 106 sends the position feedback of the rotation position R2, the speed feedback of the speed V2, and the current feedback of the current EC2 to the control device 12 as feedback FB2.

[0052] Similarly, the third sensor 108 has a rotation detection sensor that detects the rotation position R3 of the output shaft of the third drive unit 84 and a current sensor that detects the current EC3 flowing through the third drive unit 84, and detects the rotation position R3, the speed V3 (= δR3 / δt), and the current EC3 as the state data of the third drive unit 84. And the third sensor 108 sends the position feedback of the rotation position R3, the speed feedback of the speed V3, and the current feedback of the current EC3 to the control device 12 as feedback FB3.

[0053] Similarly, the fourth sensor 110 has a rotation detection sensor that detects the rotation position R4 of the output shaft of the fourth drive unit 94 and a current sensor that detects the current EC4 flowing through the fourth drive unit 94, and detects the rotation position R4, the speed V4 (= δR4 / δt), and the current EC4 as the state data of the fourth drive unit 94. And the fourth sensor 110 sends the position feedback of the rotation position R4, the speed feedback of the speed V4, and the current feedback of the current EC4 to the control device 12 as feedback FB4.

[0054] Similarly, the fifth sensor 112 has a rotation detection sensor that detects the rotational position R5 of the output shaft of the fifth drive unit 100 and a current sensor that detects the current EC5 flowing through the fifth drive unit 100, and detects the rotational position R5, the speed V5 (= δR5 / δt), and the current EC5 as state data of the fifth drive unit 100. Further, the fifth sensor 112 sends a position feedback of the rotational position R5, a speed feedback of the speed V5, and a current feedback of the current EC5 as a feedback FB5 to the control device 12.

[0055] When machining a workpiece by the industrial machine 50, the processor 20 sends commands CD1, CD2, CD3, CD4, and CD5 to the first drive unit 76, the second drive unit 78, the third drive unit 84, the fourth drive unit 94, and the fifth drive unit 100 respectively according to the machining program WP. The command CD1 sent to the first drive unit 76 includes, for example, at least one of a position command CDP1, a speed command CDV1, a torque command CDτ1, and a current command CDE1.

[0056] The position command CDP1 is a command that specifies the target rotational position of the output shaft of the first drive unit 76. The speed command CDV1 is a command that specifies the target speed of the first drive unit 76. The torque command CDτ1 is a command that specifies the target torque of the first drive unit 76. Further, the current command CDE1 is a command that specifies the current input to the first drive unit 76.

[0057] Similarly, the command CD2 sent to the second drive unit 78 includes, for example, at least one of a position command CDP2, a speed command CDV2, a torque command CDτ2, and a current command CDE2. Further, the command CD3 sent to the third drive unit 84 includes, for example, at least one of a position command CDP3, a speed command CDV3, a torque command CDτ3, and a current command CDE3.

[0058] Further, the command CD4 sent to the fourth drive unit 94 includes, for example, at least one of a position command CDP4, a speed command CDV4, a torque command CDτ4, and a current command CDE4. Further, the command CD5 sent to the fifth drive unit 100 includes, for example, at least one of a position command CDP5, a speed command CDV5, a torque command CDτ5, and a current command CDE5.

[0059] The industrial machine 50 operates the moving mechanism 102 (specifically, the translational moving mechanism 54, the swing moving mechanism 58, the rotational moving mechanism 62, and the spindle moving mechanism 70) according to the commands CD1, CD2, CD3, CD4, and CD5 from the processor 20, and machines the workpiece W with the tool 68 while relatively moving the tool 68 and the workpiece W. Figure 3An example of a workpiece W processed by an industrial machine 50 is shown. In the present embodiment, the workpiece W has a substantially rectangular flat base portion 120 and a cylindrical portion 124 protruding upward from the upper surface 122 of the base portion 120.

[0060] The machining program WP is a computer program (e.g., a G-code program) including a plurality of command statements that specify a plurality of target positions where the tool 68 should be disposed relative to the workpiece W, minute line segments connecting two adjacent target positions, or the target speed of the tool 68 relative to the workpiece W, etc.

[0061] When generating the machining program WP, an operator uses a drafting device such as CAD to create a workpiece model WM obtained by modeling the target shape of the workpiece W as a product. Figure 4 The workpiece model WM is shown. The workpiece model WM is three-dimensional model data and has a base portion model 120M obtained by modeling the base portion 120 and a cylindrical portion model 124M obtained by modeling the cylindrical portion 124. In a three-dimensional virtual model space where the model is created by the drafting device, a model coordinate system CW is set, and the workpiece model WM is composed of component models (point models, line models, surface models) set in the model coordinate system CW.

[0062] Next, the operator inputs the created workpiece model WM into a program generation device such as CAM, and the program generation device generates the machining program WP based on the workpiece model WM. In this way, the machining program WP is created based on the pre-prepared workpiece model WM and is stored in the memory 22 in advance.

[0063] After the industrial machine 50 has machined the workpiece W, the operator sets the machined workpiece W in the measuring device 18, and the measuring device 18 measures the shape of the machined workpiece W. Here, an error α may occur between the shape of the workpiece W machined by the industrial machine 50 according to the machining program WP and the target shape of the workpiece W prepared in advance (i.e., the workpiece model WM).

[0064] In the present embodiment, the processor 20 generates first image data ID1 representing the distribution Dα (first distribution) of the occurrence positions of the error α in the workpiece W. Specifically, the measuring device 18 receives the input of the workpiece model WM, compares the workpiece model WM with the measurement data of the measured shape of the workpiece W, and measures the error α. In addition, the measuring device 18 may be configured to set a prescribed threshold value for the error α and measure only the error α above the threshold value.

[0065] The processor 20 acquires the workpiece model WM via the I / O interface 24, and acquires the measurement data of the error α from the measuring device 18, and generates Figure 4 the first image data ID1 shown. InFigure 4 In the first image data ID1 shown, a distribution Dα of the generation sites of the error α is shown on the workpiece model WM, and this distribution Dα includes distribution regions E1, E2, and E3 that are displayed in a visually recognizable manner.

[0066] Each of the distribution regions E1, E2, and E3 is composed of an aggregate of the generation sites (i.e., points) of the error α measured by the measuring device 18, and is represented as the coordinates of the model coordinate system CW. Each of the distribution regions E1, E2, and E3 corresponds to the following region: a region where the surface of the machined workpiece W protrudes outward or depresses inward with respect to the surface model of the workpiece model WM corresponding to this surface. In addition, the processor 20 may also display each of the distribution regions E1, E2, and E3 in a specific color (red, blue, yellow).

[0067] Thus, in the present embodiment, the processor 20 functions as the first image generation unit 152 that generates the first image data ID1 ( Figure 1 ). In addition, the processor 20 may also display the generated first image data ID1 on the display device 14. In this case, the processor 20 may also change the direction of observing the virtual model space defined by the model coordinate system CW according to the operation of the operator on the input device 16. In this case, the operator can observe the workpiece model WM and the distribution Dα displayed on the display device 14 from various directions.

[0068] The processor 20 generates second image data ID2, and the second image data ID2 represents the commands CD (CD1, CD2, CD3, CD4, and CD5) sent to the industrial machine 50 (specifically, the first drive unit 76, the second drive unit 78, the third drive unit 84, the fourth drive unit 94, the fifth drive unit 100) for machining the workpiece W and the distribution Dβ (second distribution) of the generation sites of the error β of the feedback FB (FB1, FB2, FB3, FB4, and FB5) from the industrial machine 50 (specifically, the first sensor 104, the second sensor 106, the third sensor 108, the fourth sensor 110, and the fifth sensor 112) corresponding to the command CD in the workpiece W.

[0069] Hereinafter, an example of a method for generating the second image data ID2 will be described. First, the processor 20 acquires the time series data of the instruction CD issued during the machining of the workpiece W and the feedback FB acquired during the machining. Here, during the machining of the workpiece W, the processor 20 associates the instruction CD and the feedback FB with the time t (for example, the time from the start of machining or the standard time), and stores them in the memory 22 as the time series data of the instruction CD and the feedback FB. When generating the second image data ID2, the processor 20 reads out and acquires the time series data of the instruction CD and the feedback FB from the memory 22. Therefore, in the present embodiment, the processor 20 functions as the time series data acquisition unit 154 ( Figure 1 ).

[0070] In addition, the processor 20 generates the movement trajectory MP of the industrial machine 50 when machining the workpiece W. As an example, the processor 20 generates the movement trajectory MP1 specified by the machining program WP. The movement trajectory MP1 is an aggregate of minute line segments specified by the machining program WP, and is the movement trajectory in terms of the control of the tool 68 (or TCP) relative to the workpiece W. The processor 20 can generate the movement trajectory MP1 in the three-dimensional space by analyzing the machining program WP.

[0071] As another example, the processor 20 can also generate the movement trajectory MP2 of the industrial machine 50 based on the feedback FB acquired during the machining. The movement trajectory MP2 can be obtained, for example, by performing calculations based on the position feedbacks R1, R2, R3, R4, and R5 detected by the sensors 104, 106, 108, 110, and 112 during the machining. The movement trajectory MP2 is the actual movement trajectory of the tool 68 (or TCP) relative to the workpiece W.

[0072] In this way, the processor 20 generates the movement trajectory MP (MP1 or MP2) of the industrial machine 50. Therefore, in the present embodiment, the processor 20 functions as the trajectory generation unit 156 ( Figure 1 ) for generating the movement trajectory MP. Figure 5 An example of the trajectory model PM that three-dimensionally displays the generated movement trajectory MP is shown. Figure 6 It is Figure 5 an enlarged view of the region VI of. As Figure 6 shown, the trajectory model PM is composed of the movement trajectory MP. The trajectory model PM is a model having an outer shape that substantially coincides with the workpiece model WM.

[0073] In addition, the processor 20 can set the model coordinate system CW' for the trajectory model PM. Here, the movement trajectory MP (MP1, MP2) is obtained as a result of executing the machining program WP created based on the workpiece model WM. Therefore, the processor 20 can set it in a manner consistent with Figure 4The origin position of the model coordinate system CW shown and the positional relationship of each axis direction are the same with respect to the origin position of the workpiece model WM, and Figure 5 set the origin position of the model coordinate system CW’ and the directions of each axis for the trajectory model PM shown. The trajectory model PM is coordinated in the model coordinate system CW’.

[0074] Next, the processor 20 displays the position on the movement trajectory MP where the error β between the command CD and the feedback FB occurs, on the movement trajectory MP, based on the time series data of the command CD, the feedback FB, and the movement trajectory MP. For example, the processor 20 dot-displays (plots) the position on the movement trajectory MP where the error β occurs, on the movement trajectory MP of the trajectory model PM.

[0075] Here, the movement trajectory MP is correlated with the time series data of the command CD and the feedback FB via the time t. Specifically, the movement trajectory MP1 is specified by the machining program WP, and is correlated via the time t with the time series data of the command CD generated according to the machining program WP, and the time series data of the feedback FB corresponding to the command CD. In addition, the movement trajectory MP2 is generated based on the feedback FB, and is correlated via the time t with the time series data of the feedback FB and the time series data of the command CD corresponding to the feedback FB.

[0076] Therefore, the processor 20 can determine the time t when the error β occurs, based on the time series data of the command CD and the feedback FB, and determine the point on the movement trajectory MP at that time t. In this way, the processor 20 generates the second image data ID2 representing the distribution Dβ by displaying the position on the movement trajectory MP where the error β occurs, on the movement trajectory MP of the trajectory model PM. In addition, the processor 20 may be configured to set a prescribed threshold for the error β, and display only the error β above the threshold on the movement trajectory MP.

[0077] Figure 7 An example of the second image data ID2_1 is shown, where the second image data ID2_1 represents the distribution Dβ1 of the generation positions of the error β1 between the command CD1 to the first drive unit 76 and the feedback FB1 corresponding to the command CD1. The error β1 is, for example, the error between the position command CDP1 and the position feedback R1, the error between the speed command CDV1 and the speed feedback V1, or the error between the torque command CDτ1 (or the current command CDE1) and the current feedback EC1.

[0078] In Figure 7In the second image data ID2_1 shown, the distribution Dβ1 of the generation positions of the error β1 is shown on the trajectory model PM, and this distribution Dβ1 includes the distribution regions F1 and F2. Each of the distribution regions F1 and F2 is an aggregate of the generation positions (point displays) of the error β1 displayed (plotted) on the movement trajectory MP.

[0079] Figure 8 An example of the second image data ID2_2 is shown, and the second image data ID2_2 represents the distribution Dβ2 of the generation positions of the error β2 between the command CD2 to the second drive unit 78 and the feedback FB2 corresponding to this command CD2. The error β2 is, for example, the error between the position command CDP2 and the position feedback R2, the error between the speed command CDV2 and the speed feedback V2, or the error between the torque command CDτ2 (or the current command CDE2) and the current feedback EC2. In Figure 8 the second image data ID2_2 shown, the distribution Dβ2 including the distribution regions G1 and G2 is shown on the trajectory model PM.

[0080] Figure 9 An example of the second image data ID2_3 is shown, and the second image data ID2_3 represents the distribution Dβ3 of the generation positions of the error β3 between the command CD3 to the third drive unit 84 and the feedback FB3 corresponding to this command CD3. The error β3 is, for example, the error between the position command CDP3 and the position feedback R3, the error between the speed command CDV3 and the speed feedback V3, or the error between the torque command CDτ3 (or the current command CDE3) and the current feedback EC3. In Figure 9 the second image data ID2_3 shown, the distribution Dβ3 including the distribution regions H1 and H2 is shown on the trajectory model PM.

[0081] Figure 10 An example of the second image data ID2_4 is shown, and the second image data ID2_4 represents the distribution Dβ4 of the generation positions of the error β4 between the command CD4 to the fourth drive unit 94 and the feedback FB4 corresponding to this command CD4. The error β4 is, for example, the error between the position command CDP4 and the position feedback R4, the error between the speed command CDV4 and the speed feedback V4, or the error between the torque command CDτ4 (or the current command CDE4) and the current feedback EC4. In Figure 10 the second image data ID2_4 shown, the distribution Dβ4 including the distribution region I3 is shown on the trajectory model PM.

[0082] Figure 11An example of the second image data ID2_5 is shown, where the second image data ID2_5 represents the distribution Dβ5 of the generation sites of the error β5 between the command CD5 to the fifth drive unit 100 and the feedback FB5 corresponding to the command CD5. The error β5 is, for example, the error between the position command CDP5 and the position feedback R5, the error between the speed command CDV5 and the speed feedback V5, or the error between the torque command CDτ5 (or the current command CDE5) and the current feedback EC5. In Figure 11 In the second image data ID2_5 shown, the distribution Dβ5 including the distribution region J3 is shown on the trajectory model PM.

[0083] As Figures 7 - 11 shown, the second image data ID2_1, ID2_2, ID2_3, ID2_4, ID2_5 show the distributions Dβ1, Dβ2, Dβ3, Dβ4, Dβ5 of the errors β1, β2, β3, β4, β5 for the drive units 76, 78, 84, 94, 100. In addition, the processor 20 may also generate the image data ID2_1, ID2_2, ID2_3, ID2_4, and ID2_5 individually as the second image data ID2. In this case, the second image data ID2 includes a total of 5 image data ID2_1, ID2_2, ID2_3, ID2_4, and ID2_5.

[0084] Instead, the processor 20 may also generate the second image data ID2 as image data in a form in which the image data ID2_1, ID2_2, ID2_3, ID2_4, and ID2_5 are integrated into 1 image data. In this case, the second image data ID2 is image data including the trajectory model PM and the distribution regions F1, F2, G1, G2, H1, H2, I3, and J3 distributed on the trajectory model PM.

[0085] As described above, in the present embodiment, the processor 20 functions as the second image generation unit 158 that generates the second image data ID2 ( Figure 1 ). In addition, the processor 20 may also display the generated second image data ID2 on the display device 14. In this case, the processor 20 may also change the viewing direction of the virtual model space defined by the model coordinate system CW' according to the operation of the operator on the input device 16. Thereby, the operator can observe the trajectory model PM and the distribution Dβ displayed on the display device 14 from various directions. In addition, the processor 20 may also arrange and display the first image data ID1 and the second image data ID2 on the display device 14.

[0086] The processor 20 calculates the correlation CR between the distribution Dα of the first image data ID1 and the distribution Dβ of the second image data ID2 based on the first image data ID1 and the second image data ID2. In the present embodiment, first, an operator operates the input device 16 to specify a specific region S for the distribution Dα of the first image data ID1.

[0087] For example, as Figure 4 shown, while visually confirming the first image data ID1 displayed on the display device 14, the operator operates the input device 16 and, in the first image data ID1, specifies a region S1 so as to include the distribution regions E1 and E2 in the distribution Dα. The processor 20 receives the input information input to the input device 16 via the I / O interface 24.

[0088] In this way, in the present embodiment, the processor 20 functions as an input receiving unit 160 ( Figure 1 ) that receives the input information specifying the region S1 in the first image data ID1. The processor 20 extracts the distribution regions E1 and E2 included in the specified region S1 according to the input information.

[0089] Next, as Figures 7 - 11 shown, the processor 20 sets the region S1 at the position (specifically, the same position) corresponding to the region S1 specified in the first image data ID1 in the second image data ID2_1, ID2_2, ID2_3, ID2_4, and ID2_5.

[0090] As described above, the positional relationship of the workpiece model WM in the model coordinate system CW is consistent with the positional relationship of the trajectory model PM in the model coordinate system CW'. Therefore, the processor 20 can set the region S1 at the same position as the first image data ID1 in the second image data ID2_1, ID2_2, ID2_3, ID2_4, and ID2_5 based on the input information of the region S1.

[0091] And, the processor 20 extracts the distribution regions F1 and F2 included in the set region S1 in the distribution Dβ1 in the Figure 7 shown second image data ID2_1. In addition, the processor 20 extracts the distribution regions G1 and G2 included in the region S1 in the distribution Dβ2 in the Figure 8 shown second image data ID2_2. In addition, the processor 20 extracts the distribution regions H1 and H2 included in the region S1 in the distribution Dβ3 in the Figure 9 shown second image data ID2_3. On the other hand, in Figure 10 and Figure 11In the second image data ID2_4 and ID2_5 shown, the distributions Dβ4 and Dβ5 are not included in the set region S1.

[0092] Furthermore, the processor 20 calculates Figure 4 the correlation CR1_1 between the distribution Dα (distribution regions E1 and E2) in the region S1 shown and Figure 7 the distribution Dβ1 (distribution regions F1 and F2) in the region S1 shown. Hereinafter, the method for calculating the correlation CR1_1 will be described. As an example, the processor 20 converts Figure 4 the distribution regions E1 and E2 in the region S1 shown into 2D image data. In addition, as Figure 7 shown, the processor 20 converts the distribution regions F1 and F2 in the region S1 when observing the trajectory model PM from the same direction as the direction in which the workpiece model WM is observed in Figure 4 into 2D image data.

[0093] Figure 12 An example of imaging such 2D image data is shown. And the processor 20 calculates, through an operation, the similarity between the images (or shapes) of the distribution regions E1 and E2 obtained by converting them into 2D image data as shown in Figure 12 and the images of the distribution regions F1 and F2, as a parameter representing the correlation CR1_1. This similarity is a parameter indicating the degree of similarity between two images (shapes). For example, it can be calculated based on the correlation matching of the brightness of the two images or the distance between intermediate images obtained by performing an orthogonal transformation (Fourier transform, discrete cosine transform) on each image.

[0094] The larger (or smaller) the value of the similarity, the more similar the two images (shapes) are (i.e., the higher the correlation). Therefore, when calculating the correlation CR1_1 as the similarity, if the similarity between the distribution regions E1 and E2 and the distribution regions F1 and F2 is large (or small), it quantitatively indicates that their correlation is high.

[0095] As another example of the method for calculating the correlation CR1_1, the processor 20, as Figure 12 shown, overlaps the distribution regions E1 and E2 obtained by converting them into 2D image data with the distribution regions F1 and F2 in such a way that the outer frames of the image data match. And the processor 20 obtains the area or the number of pixels of the region where the distribution regions E1 and E2 overlap with the distribution regions F1 and F2 as the correlation CR1_1. The larger this area or the number of pixels, the higher the correlation between the distribution regions E1 and E2 and the distribution regions F1 and F2 is quantitatively indicated.

[0096] As still another example of the method for calculating the correlation CR1_1, the processor 20 converts Figure 7The distribution regions F1 and F2 in the shown model coordinate system CW’ are transformed into Figure 4 the shown model coordinate system CW, in which they coincide with the distribution regions E1 and E2. Here, each display point constituting the distribution Dβ1 in the second image data ID2_1 is represented by the coordinates of the model coordinate system CW’. In addition, as described above, the position of the workpiece model WM in the model coordinate system CW is consistent with the position of the trajectory model PM in the model coordinate system CW’.

[0097] Therefore, by plotting the coordinates of the distribution regions F1 and F2 in the model coordinate system CW’ of Figure 7 in the model coordinate system CW of Figure 4 , the distribution regions E1 and E2 can be made to coincide with the distribution regions F1 and F2 in the model coordinate system CW. And the processor 20 obtains, in the model coordinate system CW, the area of the region where the distribution regions E1 and E2 coincide (or are close within a specified distance) with the transformed distribution regions F1 and F2 as the correlation CR1_1. The larger this area, the higher the correlation between the distribution Dα and the distribution Dβ1 quantitatively. Using the method as above, the correlation CR1_1 can be obtained.

[0098] Using the same method, the processor 20 can respectively obtain Figure 4 the correlation CR1_2 between the distribution Dα (E1 and E2) in the region S1 of Figure 8 and the distribution Dβ2 (G1 and G2) in the region S1 of Figure 4 , the correlation CR1_3 between the distribution Dα in the region S1 of Figure 9 and the distribution Dβ3 (H1 and G2) in the region S1 of Figure 4 , the correlation CR1_4 between the distribution Dα in the region S1 of Figure 10 and the distribution Dβ2 (nonexistent) in the region S1 of Figure 4 , and the correlation CR1_5 between the distribution Dα in the region S1 of Figure 11 and the distribution Dβ2 (nonexistent) in the region S1 of

[0099] As described above, the processor 20 obtains the correlation CR1 between the distribution Dα and the distribution Dβ based on the first image data ID1 and the second image data ID2. Therefore, the processor 20 functions as the correlation acquisition unit 162 ( Figure 1 ) that obtains the correlation CR1. In addition, in the present embodiment, the processor 20 respectively obtains the correlations CR1_1, CR1_2, CR1_3, CR1_4, and CR1_5 between the distribution Dα and the distributions Dβ1, Dβ2, Dβ3, Dβ4, and Dβ5 of each of the drive units 76, 78, 84, 94, and 100.

[0100] Next, the processor 20 generates sequential image data OD1, and the sequential image data OD1 displays the first driving unit 76, the second driving unit 78, the third driving unit 84, the fourth driving unit 94, and the fifth driving unit 100 in descending order of the obtained correlations CR1_1, CR1_2, CR1_3, CR1_4, and CR1_5. Figure 13 An example of an image representing the sequential image data OD1.

[0101] As Figure 13 shown, in the present embodiment, the processor 20 overlays and displays the sequential image data OD1 and the first image data ID1. In the present embodiment, regarding the height of the correlation CR1, CR1_1 > CR1_2 > CR1_3 > CR1_4 = CR1_5 = 0. Therefore, in the set region S1, the distribution regions F1 and F2 of the error β1 related to the first driving unit 76 have the highest correlation CR1_1 with the distribution Dα (distribution regions E1 and E2), the distribution regions G1 and G2 of the error β2 related to the second driving unit 78 have the second highest correlation CR1_2, and the distribution regions H1 and H2 of the error β3 related to the third driving unit 84 have the third highest correlation CR1_3.

[0102] On the other hand, regarding the distributions Dβ of the errors β4 and β5 related to the fourth driving unit 94 and the fifth driving unit 100, the correlation CR1 with the distribution Dα is zero (i.e., there is no correlation). Therefore, as Figure 13 shown, in the sequential image data OD1, the plurality of driving units 76, 78, 84, 94, and 100 are displayed in the order of displaying the first driving unit 76 in the first place, the second driving unit 78 in the second place, the third driving unit 84 in the third place, and the fourth driving unit 94 and the fifth driving unit 100 in the fourth place.

[0103] In addition, in the present embodiment, the sequential image data OD1 displays an identification information column K and a correlation information column L together with the order of the driving units 76, 78, 84, 94, and 100. The identification information column K represents information for identifying the driving unit (for example, a string, an identification number, a symbol, etc.), and the correlation information column L represents information on the correlation CR1 (for example, a numerical value).

[0104] In this way, in the present embodiment, the processor 20 functions as the third image generation unit 164 ( Figure 3 ) that generates the sequential image data OD1. Through this sequential image data OD1, the operator can visually recognize the order of the correlation CR1 between the distribution Dα (E1, E2) of the error α in the specified region S1 and the distributions Dβ of the errors β of the respective driving units 76, 78, 84, 94, and 100.

[0105] In addition, the processor 20 can also function as the third image generation unit 164 to generate the recognition image data DD1 of the drive unit 76 with the highest recognition correlation CR1 instead of (or on the basis of) the sequential image data OD1. Figure 14 An example of such recognition image data DD1 is shown. In Figure 14 the example shown, the recognition image data DD1 is composed of an arrow pointing to the area S1 and a mark (specifically, the number inside the circle mark) of the first drive unit 76 with the highest recognition correlation CR1, and is shown in the first image data ID1. Through the recognition image data DD1, the operator can visually recognize that the distribution Dα (E1, E2) of the error within the specified area S1 has the highest correlation with the distribution Dβ1 (F1, F2) of the error regarding the first drive unit 76.

[0106] Similarly, as Figure 4 shown, while visually confirming the first image data ID1 displayed on the display device 14, the operator operates the input device 16 and specifies the area S2 in the first image data ID1 in such a way as to include the distribution area E3 in the distribution Dα. The processor 20 functions as the input receiving unit 160, receives the input information of the specified area S2, and as Figures 7 - 11 shown, sets the area S2 at the same position as the area S2 specified in the first image data ID1 in the second image data ID2_1, ID2_2, ID2_3, ID2_4, and ID2_5.

[0107] And the processor 20 functions as the correlation acquisition unit 162, and uses the above method to respectively obtain Figure 4 the correlation CR2_1 between the distribution Dα (E3) within the area S2 in Figure 7 and the distribution Dβ1 (nonexistent) within the area S2 in Figure 4 the correlation CR2_2 between the distribution Dα within the area S2 in Figure 8 and the distribution Dβ2 (nonexistent) within the area S2 in Figure 4 the correlation CR2_3 between the distribution Dα within the area S2 in Figure 9 and the distribution Dβ3 (nonexistent) within the area S2 in Figure 4 the correlation CR2_4 between the distribution Dα within the area S2 in Figure 10 and the distribution Dβ4 (I3) within the area S2 shown in Figure 4 and the correlation CR2_5 between the distribution Dα within the area S2 in Figure 11 and the distribution Dβ5 (J3) within the area S2 shown in

[0108] Next, the processor 20 functions as the third image generation unit 164 to generate sequential image data OD2( Figure 13 ), and the sequential image data OD2 displays the first driving unit 76, the second driving unit 78, the third driving unit 84, the fourth driving unit 94, and the fifth driving unit 100 in descending order of the obtained correlation coefficients CR2_1, CR2_2, CR2_3, CR2_4, and CR2_5.

[0109] In the present embodiment, regarding the height of the correlation coefficient CR2, CR2_4 > CR2_5 > CR2_1 = CR2_2 = CR2_3 = 0. Therefore, in the set region S2, the distribution region I3 of the error β4 related to the fourth driving unit 94 has the highest correlation coefficient CR2_4 with the distribution Dα (distribution region E3) within the region S2, and the distribution region J3 of the error β5 related to the fifth driving unit 100 has the second highest correlation coefficient CR2_5.

[0110] On the other hand, regarding the distributions Dβ of the errors β1, β2, and β3 related to the first driving unit 76, the second driving unit 78, and the third driving unit 84, the correlation coefficient CR2 with the distribution Dα is zero (i.e., there is no correlation). Therefore, as Figure 13 shown, in the sequential image data OD2, the multiple driving units 76, 78, 84, 94, and 100 are displayed in the order of displaying the fourth driving unit 94 in the first place, the fifth driving unit 100 in the second place, and the first driving unit 76, the second driving unit 78, and the third driving unit 84 in the third place.

[0111] In addition, the processor 20 can also function as the third image generation unit 164 to generate recognition image data DD2 for recognizing the driving unit 94 with the highest correlation coefficient CR2, as Figure 14 shown. The recognition image data DD2 is composed of an arrow pointing to the region S2 and a mark for recognizing the fourth driving unit 94 with the highest correlation coefficient CR2, and is shown in the first image data ID1. In addition, the processor 20 can either overlap and display the sequential image data OD1 and OD2, or the recognition image data DD1 and DD2 with the second image data ID2, or can also be displayed as image data different from the first image data ID1 and the second image data ID2.

[0112] As described above, in the present embodiment, the processor 20 functions as the first image generation unit 152, the second image generation unit 158, the third image generation unit 164, the correlation acquisition unit 162, the input reception unit 160, the time series data acquisition unit 154, and the trajectory generation unit 156. These first image generation unit 152, second image generation unit 158, third image generation unit 164, correlation acquisition unit 162, input reception unit 160, time series data acquisition unit 154, and trajectory generation unit 156 constitute the image analysis device 150( Figure 1 ).

[0113] In addition, the image analysis device 150 can be configured as a computer program (i.e., software). This computer program causes a computer (processor 20) to function as the first image generation unit 152, the second image generation unit 158, the third image generation unit 164, the correlation acquisition unit 162, the input reception unit 160, the time series data acquisition unit 154, and the trajectory generation unit 156 for image analysis.

[0114] According to the present embodiment, the operator can determine whether the cause of the error α is likely to be the error β by considering the correlation CR between the distribution Dα of the error α and the distribution Dβ of the error β. Therefore, it is easy for the operator to determine the cause of the error α. As a result, it is possible to improve the processing accuracy of the industrial machine 50 and to streamline the startup process of the mechanical system 10.

[0115] In addition, in the present embodiment, the processor 20 generates second image data ID2_1, ID2_2, ID2_3, ID2_4, ID2_5 representing the distributions Dβ1, Dβ2, Dβ3, Dβ4, Dβ5 of each of the drive units 76, 78, 84, 94, 100, and respectively obtains the correlations CR1, CR2 between the distribution Dα of the error α and the distributions Dβ1, Dβ2, Dβ3, Dβ4, Dβ5 of each of the drive units 76, 78, 84, 94, 100.

[0116] With this configuration, it is easy for the operator to estimate which of the plurality of drive units 76, 78, 84, 94, and 100 is likely to be the cause of the error α. For example, in the case of the above-described embodiment, the operator can presume that for Figure 4 the distribution regions E1 and E2 in the distribution Dα of the error α, the error β1 related to the first drive unit 76 is likely to be the cause. In addition, the operator can presume that for Figure 4 the distribution region E3 in the distribution Dα of the error α, the error β4 related to the fourth drive unit 94 is likely to be the cause.

[0117] In addition, in the present embodiment, the processor 20 receives the input information of the designated regions S1 (S2), and obtains the correlations CR1 (CR2) between the distribution regions E1 and E2 (E3) included in the region S1 (S2) of the distribution Dα and the distribution regions F1 and F2, G1 and G2, H1 and H2 (I3, J3) included in the region S1 of the distribution Dβ. With this configuration, the operator can easily determine the cause of the error α generated in the desired region by considering the correlation CR for the desired region in the distribution Dα.

[0118] In addition, it is also possible to Figure 1 set at least one of the functions of the first image generation unit 152, the second image generation unit 158, the third image generation unit 164, the correlation acquisition unit 162, the input reception unit 160, the time series data acquisition unit 154, and the trajectory generation unit 156 shown in an external device of the control device 12. Figure 15 This represents such a mode.

[0119] Figure 15 The mechanical system 170 shown has: a control device 12, a display device 14, an input device 16, a measurement device 18, an industrial machine 50, and a design support device 172. The design support device 172 is, for example, a device integrating a drawing device such as CAD and a program generation device such as CAM, and is connected to the I / O interface 24 of the control device 12.

[0120] In the present embodiment, the measurement device 18 functions as the first image generation unit 152, measures the shape of the processed workpiece W, and generates the first image data ID1. In addition, the design support device 172 functions as the second image generation unit 158 and generates the second image data ID2. Therefore, in the present embodiment, the processor 20, the measurement device 18, and the design support device 172 of the control device 12 constitute the image analysis device 150.

[0121] In addition, in the above-described embodiment, the case where the processor 20 functions as the input reception unit 160 that receives the input information designating the regions S1 and S2 has been described. However, the input reception unit 160 can also be omitted. In this case, the processor 20 can also obtain the correlation CR between the entire distribution Dα of the first image data ID1 and the entire distribution Dβ (Dβ1, Dβ2, Dβ3, Dβ4, Dβ5) of the second image data ID2.

[0122] In addition, in the above-described embodiment, the case where the processor 20 generates the second image data ID2 by displaying the position on the movement locus MP where the error β has occurred on the movement locus MP has been described. However, it is not limited thereto. For example, the processor 20 may also generate the second image data ID2 that shows the distribution Dβ of the error β on the workpiece model WM based on the instruction CD, the feedback FB, and the workpiece model WM. In this case, the locus generation unit 156 can be omitted.

[0123] In addition, in the above-described embodiment, the case where the processor 20 acquires the time-series data of the instruction CD and the feedback FB has been described. However, it is not limited thereto. As long as the processor 20 can acquire the error β, any data can be used. In this case, the time-series data acquisition unit 154 can be omitted. In addition, the third image generation unit 164 may also be omitted from the above-described embodiment. For example, the processor 20 is configured to transmit the obtained correlation CR to an external device (such as a server) via a network (LAN, Internet).

[0124] Figure 16 An image analysis device 180 showing another embodiment. The image analysis device 180 is composed of a computer having a processor (CPU, GPU, etc.) and a memory (ROM, RAM, etc.), or a computer program (software) that operates the computer, and includes: a first image generation unit 152, a second image generation unit 158, and a correlation acquisition unit 162. In the present embodiment, the functions of the above-described input reception unit 160, time-series data acquisition unit 154, locus generation unit 156, and third image generation unit 164 may also be provided to an external device of the image analysis device 180.

[0125] In addition, the above-described instructions CD1, CD2, CD3, CD4, or CD5 may also have a position instruction or a speed instruction of a driven body (for example, the base table 52, the swing member 60, the workpiece table 64, or the spindle head 66) driven by the driving units 76, 78, 84, 94, or 100. In this case, the feedback FB has a position feedback or a speed feedback of the driven body, and the industrial machine 50 has a sensor for detecting the position of the driven body.

[0126] In addition, the processor 20 may also Figure 4 in the shown first image data ID1, display the distribution Dα in different colors according to the magnitude of the error α. Similarly, the processor 20 may also Figures 7 - 11 in the shown second image data ID2, display the distribution Dβ in different colors according to the magnitude of the error β.

[0127] In addition, the processor 20 (or the measurement device 18) may also generate a measured workpiece model obtained by modeling the shape of the workpiece W measured by the measurement device 18. Further, the processor 20 (or the measurement device 18) may generate first image data ID1 representing the distribution α of the error α on the measured workpiece model. Additionally, the processor 20 may generate first image data ID1 representing the distribution α of the error α on the above-described trajectory model PM.

[0128] In addition, in the above-described embodiment, the case where the industrial machine 50 has a total of five drive units 76, 78, 84, 94, and 100 has been described, but it may have any number of drive units. Further, the industrial machine 50 may have, for example, a vertically (or horizontally) articulated robot as the moving mechanism. In this case, the industrial machine 50 may have a tool such as a laser processing head attached to the fingertips of the robot instead of the above-described tool 68, and while moving the tool by the robot, perform laser processing on the workpiece with a laser beam emitted from the tool.

[0129] As described above, the present disclosure has been explained by way of embodiments, but the above-described embodiments do not limit the invention of the claimed technical solution.

[0130] Reference Signs

[0131] 10, 170 Mechanical System

[0132] 12 Control Device

[0133] 50 Industrial Machine

[0134] 68 Tool

[0135] 102 Moving Mechanism

[0136] 150, 180 Image Analysis Device

[0137] 152 First Image Generation Unit

[0138] 154 Time-Series Data Acquisition Unit

[0139] 156 Trajectory Generation Unit

[0140] 158 Second Image Generation Unit

[0141] 160 Input Reception Unit

[0142] 162 Correlation Acquisition Unit

[0143] 164 Third Image Generation Unit.

Claims

1. An image analysis device, characterized in that: The image analysis device has: A first image generation unit that generates first image data representing a first distribution of a generation position in the workpiece of an error between the shape of the workpiece machined by an industrial machine and a target shape of the workpiece prepared in advance; A second image generation unit that generates second image data representing a second distribution of a generation position in the workpiece of an error between an instruction sent to the industrial machine for machining the workpiece and a feedback from the industrial machine corresponding to the instruction; and A correlation acquisition unit that obtains a correlation between the first distribution and the second distribution based on the first image data and the second image data.

2. The image analysis device according to claim 1, characterized in that: The industrial machine has: A tool that machines the workpiece; and A moving mechanism that is a moving mechanism for relatively moving the workpiece and the tool, and has a plurality of driving parts that respectively drive the tool relative to the workpiece in a plurality of directions, The second image generation unit generates the second image data representing the second distribution of the error between the instruction sent to each driving part and the feedback for each driving part, The correlation acquisition unit respectively obtains the correlation between the first distribution and the second distribution of each driving part.

3. The image analysis device according to claim 2, characterized in that: The image analysis device further has: a third image generation unit that generates identification image data for identifying the driving part with the highest correlation, or sequential image data for displaying the plurality of driving parts in descending order of the correlation.

4. The image analysis device according to any one of claims 1 to 3, characterized in that: The image analysis device further has: an input reception unit that receives input information for designating a region in the first image data, The correlation acquisition unit obtains the correlation between a distribution region included in the region in the first distribution and a distribution region included in a region set at a position corresponding to the region in the second image data in the second distribution.

5. The image analysis device according to any one of claims 1 to 3, characterized in that: The instruction includes a position instruction for specifying the position of the industrial machine, The feedback includes a position feedback representing the position of the industrial machine detected by a sensor.

6. The image analysis device according to any one of claims 1 to 3, characterized in that: The image analysis device further has: a time series data acquisition unit that acquires time series data of the instruction and the feedback; and A trajectory generation unit that generates a movement trajectory of the industrial machine when machining the workpiece, The second image generation unit displays, on the movement trajectory, positions on the movement trajectory where the error between the instruction and the feedback has occurred, based on the time series data and the movement trajectory, thereby generating the second image data.

7. A control device for the industrial machine, characterized in that: The control device includes the image analysis device according to any one of claims 1 to 6.

8. A mechanical system, characterized in that: The mechanical system includes: An industrial machine that processes workpieces; A measuring device that measures the shape of the workpiece processed by the industrial machine; and The image analysis device according to any one of claims 1 to 6, The first image generation unit generates the first image data based on the measurement data of the shape of the workpiece measured by the measurement device.

9. An image analysis method, characterized in that: Generate first image data, which represents the first distribution of the generation positions in the workpiece of the error between the shape of the workpiece processed by the industrial machine and the target shape of the workpiece prepared in advance; Generate second image data, which represents the second distribution of the generation positions in the workpiece of the error between the instruction sent to the industrial machine for processing the workpiece and the feedback from the industrial machine corresponding to the instruction; and Based on the first image data and the second image data, obtain the correlation between the first distribution and the second distribution.

10. A computer program product, characterized in that: For image analysis, make the computer function as the following parts: A first image generation unit that generates first image data, which represents the first distribution of the generation positions in the workpiece of the error between the shape of the workpiece processed by the industrial machine and the target shape of the workpiece prepared in advance; A second image generation unit that generates second image data, which represents the second distribution of the generation positions in the workpiece of the error between the instruction sent to the industrial machine for processing the workpiece and the feedback from the industrial machine corresponding to the instruction; and A correlation acquisition unit that obtains the correlation between the first distribution and the second distribution based on the first image data and the second image data.

Images