WO2014049766A1

WO2014049766A1 - Recognition device for substrate processing machine

Info

Publication number: WO2014049766A1
Application number: PCT/JP2012/074846
Authority: WO
Inventors: 裕司勝見
Original assignee: 富士機械製造株式会社
Priority date: 2012-09-27
Filing date: 2012-09-27
Publication date: 2014-04-03
Also published as: JPWO2014049766A1; JP5767754B2; CN104663016A; CN104663016B

Abstract

A recognition device for substrate processing machine comprises: a reading device for reading an object; a feature parameter preparation device for preparing a feature parameter from the read object; a first memory for storing a reference parameter associated with the feature parameter; and a signal output device for outputting a maintenance request signal on the basis of the hit rate representing the correlation between the feature parameter and the reference parameter. The signal output device outputs the maintenance request signal when the hit rate reaches a maintenance threshold different from an error threshold in a stage where the hit rate does not reach the error threshold.

Description

Recognition device for substrate working machine

The technology disclosed in this specification relates to a recognition apparatus for a substrate working machine. The recognition apparatus disclosed in this specification includes not only those directly provided on the substrate working machine but also those provided on equipment used in connection with the substrate working machine. The board working machine here performs a predetermined work on the circuit board. For example, a solder printing machine for printing cream solder on the circuit board, a mounting machine for mounting electronic components on the circuit board (surface mounting machine) Or a board inspection machine for inspecting a circuit board on which electronic components are mounted.

In order to improve the work accuracy, the substrate work machine recognizes the recognition target in various scenes of the work process, and performs feedback control using the recognition result. For example, a mounting machine which is an example of a board working machine is provided with a recognition device having a mark camera or a parts camera in order to improve the mounting accuracy of electronic components. The mark camera recognizes the mark marked on each of the component parts and the mark marked on the circuit board to be conveyed, and confirms the position. The parts camera recognizes the electronic component sucked by the suction nozzle, and performs identification of the electronic component and confirmation of the suction posture. Examples of these techniques are disclosed in Japanese Patent Application Laid-Open No. 2005-116869 and Japanese Patent Application Laid-Open No. 2010-199630.

For example, if the mark marked on the component is deteriorated due to dirt, damage, etc., the recognition device cannot recognize the mark and a recognition error occurs. In addition, if the continuous operation time of the substrate working machine becomes long, the suction posture of the electronic component by the suction nozzle deteriorates due to various factors, and the electronic component cannot be recognized, and a recognition error occurs. In the conventional substrate working machine, when such a recognition error occurs, the work is forcibly stopped, and maintenance of the cause is performed by the operator. As described above, in the method of performing the maintenance of the cause portion after the recognition error is generated by the recognition device, a situation in which the production plan does not proceed as scheduled frequently occurs. In this specification, it aims at providing the recognition apparatus which can avoid the malfunction by such a recognition error.

One embodiment of a recognition apparatus for a substrate working machine disclosed in this specification stores a reading device that reads an object, a feature parameter creation device that creates a feature parameter from the read object, and a standard parameter related to the feature parameter. And a signal output device that outputs a maintenance request signal based on a hit rate indicating a correlation between the characteristic parameter and the standard parameter. The signal output device outputs a maintenance request signal when the hit rate does not reach the error threshold and reaches a maintenance threshold different from the error threshold.

The signal output device of the recognition device is configured to output a maintenance request signal before a recognition error of a recognition target occurs. For this reason, for example, if the recognition target is maintained in accordance with the non-operation period of the substrate working machine after the maintenance request signal is output, the recognition error of the recognition target can be avoided. Can be prevented from being forcibly stopped.

The perspective view of a mounting machine is shown. The structure of a feeder is shown typically. The top view of a carrier tape is shown typically. The top view of the component adsorption | suction position vicinity of a feeder is shown typically. A sectional view of a parts imaging device is typically shown. The top view of a nozzle stocker is shown typically. A perspective view of a part of substrate transportation device is typically shown. An outline of a mounting head and a mark imaging device is schematically shown. The perspective view of an example of a mounting head is typically shown. The perspective view of another example of a mounting head is typically shown. The top view of the protrusion part of the front-end | tip vicinity of a suction nozzle is shown typically. The system configuration of the mounting machine is shown. An explanatory view of image processing using a seek line is shown. An explanatory view of image processing using a seek line is shown (when a part of a reference mark is missing). An example of how the hit rate changes with time is shown. Another example of how the hit rate changes with time is shown. Another example of how the hit rate changes with time is shown. The flow of the fiducial mark recognition program is shown. Another example of how the hit rate changes with time is shown. The flow of the program which specifies the cause of the hit rate fall among reference mark recognition programs is shown.

The following summarizes the features of the technology disclosed in this specification. The items described below have technical usefulness independently.

One embodiment of the recognition device disclosed in this specification includes a reading device, a characteristic parameter creation device, a first memory, and a signal output device. The reading device reads a reading target, and typically may read a mark and / or an electronic component. The mark may include a pattern having a specific shape for image recognition, a barcode, a two-dimensional code, a magnetic code, an IC tag, or the like. These marks may be marked on a substrate working machine or a component of equipment used in connection with the substrate working machine. When the mark has a pattern having a specific shape for image recognition, it is desirable to use a camera, a scanner, a laser scanner, or an imaging device of a fixed image sensor such as a CCD as the reading device. The feature parameter creation device creates feature parameters from the read target. When the read target is a pattern having a specific shape for image recognition, typically, a feature parameter related to an edge of the pattern having the specific shape may be created. The first memory stores standard parameters related to feature parameters. For example, when there are a plurality of types of reading targets, the first memory may store standard parameters related to the characteristic parameters of each reading target. This makes it possible to identify a specific reading object from a plurality of types of reading objects. The signal output device outputs a maintenance request signal based on the hit rate indicating the correlation between the characteristic parameter and the standard parameter. Here, the hit rate may be a higher index as the similarity between the feature parameter and the standard parameter is higher, and conversely, it may be a higher index as the dissimilarity between the feature parameter and the standard parameter is higher. In any case, the hit rate is an index indicating the correlation between the feature parameter and the standard parameter. The hit rate is calculated as an index for determining whether or not to identify a reading target even in a known image processing technique. The signal output device outputs a maintenance request signal when the hit rate does not reach the error threshold and reaches a maintenance threshold different from the error threshold. Here, various information may be included in the maintenance request signal. For example, the maintenance request signal is not limited to requesting the maintenance of the reading target, but when requesting the maintenance of the component related to the reading target together with the maintenance of the reading target, other components different from the reading target This includes the case where only maintenance is required.

The recognition apparatus disclosed in this specification may further include a second memory that stores time-series data of hit rates. Thereby, various useful processes can be performed by using the time-series data of the hit rate. For example, by using time-series data of hit rates, the signal output device can be configured to output a maintenance request signal when the change rate of the hit rate exceeds an abnormal value. An abrupt change in the hit rate change rate is likely to cause an abnormal event. For this reason, the signal output device is useful for early detection of such abnormal events. Further, by using the time-series data of the hit rate, the signal output device outputs a maintenance request signal with reference to the hit rate of another target when the hit rate of one target reaches the maintenance threshold. It can be constituted as follows. In this way, by referring to time-series data of other hit rates, it is possible to analyze the cause of the deteriorated hit rate. The signal output device is useful for identifying the cause of the deterioration of the hit rate.

In the recognition device disclosed in this specification, the reading device may be an imaging device. Further, the imaging apparatus may include a light source that irradiates the target and a measurement device that measures the illuminance irradiated on the target. In this case, the signal output device may output a maintenance request signal including a light source maintenance request signal when the hit rate reaches the maintenance threshold value and the illuminance reaches the illuminance threshold value. There are various causes for the deterioration of the hit rate, but when the reading device is an imaging device, there may be a problem with the illuminance of the light source. In the above embodiment, by comparing the measured illuminance of the light source with the illuminance threshold, it is possible to determine whether or not the cause of the deterioration of the hit rate is the illuminance of the light source.

The recognition apparatus disclosed in this specification may be provided in a mounting machine that mounts electronic components on a circuit board. The mounting machine has an imaging device used as a reading device. The imaging apparatus may include a mark camera that captures an image of a first mark marked on a component constituting the mounting machine and a second mark marked on a circuit board on which the electronic component is mounted. In this case, the maintenance threshold value may include a first maintenance threshold value for the first mark and a second maintenance threshold value for the second mark. The signal output device may output a maintenance request signal including the first request signal when the hit ratio of the first mark reaches the first maintenance threshold. Furthermore, the signal output device may output a maintenance request signal including a second request signal when the hit rate of the second mark reaches the second maintenance threshold. In the above embodiment, the mark camera is configured to be able to image different marks, the first mark marked on the component and the second mark marked on the circuit board. In the above embodiment, the maintenance request signal output when the hit ratio of the first mark reaches the first maintenance threshold and the maintenance request signal output when the hit ratio of the second mark reaches the second maintenance threshold. Therefore, it is possible to distinguish which mark the maintenance request signal is output to. Moreover, in the said embodiment, the 1st mark may be marked on the some component. In this case, the type of the first mark may be different depending on the type of the plurality of component parts. Further, the plurality of component parts are divided into a plurality of groups, and the signal generator belongs to the same group when the hit rate of the first mark marked on one component part reaches the first maintenance threshold value. A maintenance request signal including information on other components may be output. For example, by grouping component parts having similar mark deterioration rates, it is possible to infer the deterioration of other component parts belonging to the same group from the mark deterioration of one component part.

The recognition apparatus disclosed in this specification may be provided in a mounting machine that mounts electronic components on a circuit board. The mounting machine has an imaging device used as a reading device. The imaging device may have a parts camera that images an electronic component. In this case, the parts camera may also take an image of the first mark marked on the component existing in the imaging field when imaging the electronic component. The maintenance threshold value may include a third maintenance threshold value for the electronic component and a first maintenance threshold value for the first mark. The signal output device may output a maintenance request signal including a third request signal when the hit rate of the electronic component reaches the third maintenance threshold. Further, the signal output device may output a maintenance request signal including the first request signal when the hit rate of the first mark reaches the first maintenance threshold value. In the said embodiment, the parts camera is comprised so that imaging of the 1st mark and electronic component which are marked on the component is possible. In the above embodiment, the maintenance request signal output when the hit rate of the first mark reaches the first maintenance threshold and the maintenance request signal output when the hit rate of the electronic component reaches the third maintenance threshold are provided. It is possible to distinguish whether the maintenance request signal is output to which of the first mark or the electronic component. In the above-described embodiment, a correction request signal output device that outputs a correction request signal for a standard parameter of an electronic component based on the hit rate of the electronic component may be further provided. The standard parameters of the electronic component are determined in advance according to the type of the electronic component. However, for example, when an electronic component having a different manufacturing rod is to be recognized using the same standard parameter, the hit rate of the electronic component may differ depending on the manufacturing rod. In such a case, it is possible to suppress deterioration of the recognition accuracy of the electronic component by correcting the standard parameter.

According to the technology disclosed in this specification, it is possible to construct a management system having a third memory that stores time-series data obtained by the above recognition device. For example, the third memory may store time-series data of hit rates collected from recognition devices provided in at least two of the substrate working machine, the nozzle cleaning machine, the nozzle inspection machine, and the feeder maintenance unit. Good. Using such a management system, the time-series data on the hit rate for a specific recognition target is accumulated from multiple devices, so the time-series data on the hit rate is continuous without interruption and more effective management It can be performed.

The management system may include a prediction device that predicts when the hit rate reaches at least one of the maintenance threshold and the error threshold based on the time-series data of the hit rate stored in the third memory. Good. In this management system, it is possible to grasp the tendency of the hit rate deterioration rate, and it is possible to provide information useful for creating a production plan schedule. For example, the management system may include a fourth memory that stores production plan data describing an operation period and a non-operation period of the substrate working machine. In this case, it may further include a planning device that creates maintenance plan data based on the prediction time predicted by the prediction device and the production plan data stored in the fourth memory. The planning device may be configured to plan the maintenance execution time in the non-operation period before the operation period when the prediction period is included in the operation period.

Hereinafter, a mounting machine used when forming an integrated circuit on a circuit board will be described with reference to the drawings. The integrated circuit on the circuit board is formed by being sequentially conveyed through a solder printer, a mounting machine, and a reflow furnace. The solder printer prints cream solder on the electronic component mounting position on the circuit board. The mounting machine mounts an electronic component at a mounting position on a circuit board on which cream solder is printed. The reflow furnace solders an electronic component on a circuit board by performing heat treatment on the circuit board.

As shown in FIG. 1, the mounting machine 1 includes a plurality of modules 2 arranged adjacent to each other on a table 3. Each of the modules 2 has a common configuration, and includes a plurality of devices in a frame 4 that functions as a housing. FIG. 1 shows an example in which two modules 2 are adjacent to each other, and shows an exterior plate of one of the two modules 2 in a watermarked state. The module 2 includes a component supply device 10, a parts imaging device 20 having a parts camera, a nozzle stocker 30, a substrate transfer device 40, a component transfer device 50, a mounting head 60, and a mark imaging device 70 having a mark camera. Each of the modules 2 includes an input / output device 6 configured to be able to communicate with a control device described later on the surface of the exterior plate. The input / output device 6 is configured to be able to receive input from an operator and to display status information of the module 2 and instructions to various operators described later.

(Configuration of component supply apparatus 10)
As shown in FIG. 1, the component supply device 10 is a device that supplies a plurality of types of electronic components mounted on the circuit board 5 to a component suction position, and is configured by arranging a plurality of cassette-type feeders 11 in parallel. Yes. The feeder main body 12 of the feeder 11 is configured to be detachable from the frame 4. As shown in FIG. 2, the feeder 11 includes a tape reel 13 around which the carrier tape 14 is wound, a tape guide 15 that guides the carrier tape 14 drawn from the tape reel 13 to a component suction position, and the carrier tape 14 as a component. A sprocket 16 that pitch-feeds to the suction position and a motor (not shown) that rotationally drives the sprocket 16 are provided. Further, a communication connector 17 connected to a connector (not shown) provided on the main body of the mounting machine 1 and positioning pins 18 and 19 for determining a mounting position with respect to the main body of the mounting machine 1 are provided on the front end surface of the feeder 11. It has been.

As shown in FIG. 3, the carrier tape 14 is formed with cavities 14a for accommodating the electronic components P at a constant pitch interval. The upper opening of the cavity 14a is covered with a cover tape 14b, and the electronic component P is taped by the carrier tape 14 and the cover tape 14b. The cover tape 14b is peeled off from the carrier tape 14 before the component suction position. A sprocket 16 is engaged with a feed hole 14 c formed in the carrier tape 14. As a result, the carrier tape 14 is fed out by a predetermined pitch and sequentially supplies the electronic components P accommodated in the cavity 14a to the component suction positions. As shown in FIG. 4, an opening 12 a is formed in the upper portion of the feeder main body 12 at a position corresponding to the component suction position. The electronic component P supplied to the component suction position is sucked by the suction nozzle inserted into the opening 12a and taken out from the cavity 14a. On the upper surface of the feeder body 12, a reference mark M1 is marked in the vicinity of the component suction position. The reference mark M1 is imaged by a mark camera described later. The mounting machine 1 uses the image data of the reference mark M1 imaged by the mark camera to identify the reference mark M1 and acquire its position information. The mounting machine 1 uses the positional information of the reference mark M1 to calculate the positional deviation of the component suction position from the ideal position, moves the suction nozzle based on the positional deviation, and sucks the electronic component P.

(Configuration of parts imaging device 20)
As shown in FIG. 1, the part imaging device 20 is fixed to the frame 4 and is disposed between the component supply device 10 and the board transfer device 40. As shown in FIG. 5, the parts imaging apparatus 20 includes a parts camera 21, a support base 22, a connecting member 23, an upper end member 24, a side light source 25, and a cover glass 26. The parts camera 21 is fixed to the frame 4 via a support base 22. As the part camera 21, for example, a CCD camera is used. The support base 22 and the upper end member 24 are fixed via a connecting member 23. The upper end member 24 has a shape in which the upper surface and the bottom surface are opened and the side portions are curved. A number of side light sources 25 are provided on the inner wall surface of the upper end member 24, and the upper opening of the upper end member 24 is covered with a cover glass 26. For example, an LED is used as the side light source 25.

The electronic component P is disposed above the part imaging device 20 after being attracted at the component attracting position of the feeder 11 and before being mounted on the circuit board 5 positioned on the substrate transport device 40. Imaged. The mounting machine 1 uses the image data of the electronic component P imaged by the parts imaging device 20 to identify the electronic component P and confirm the suction posture of the electronic component P. The identification of the electronic component P is performed by comparing image data of the electronic component P with previously acquired part data and determining whether or not the electronic component P is a desired electronic component P. The confirmation of the suction posture of the electronic component P uses the image data of the electronic component P to detect the positional deviation of the electronic component P from the central axis of the suction nozzle and the angular deviation of the electronic component P around the central axis of the suction nozzle. Is done by doing. The suction posture of the electronic component P is corrected by rotating the suction nozzle before the electronic component P is mounted on the circuit board 5.

In addition, the part imaging device 20 also includes a reference mark (for example, a corner dog) marked on a component (for example, the mounting head 60) that exists in the imaging field when the electronic component P is captured as necessary. You may image simultaneously. The mounting machine 1 uses the position information of the reference mark to estimate the position of the center axis of the suction nozzle that sucks the electronic component P, and detects the displacement of the electronic component P from the center axis of the suction nozzle. May be. Further, the parts imaging apparatus 20 may image a specific reference mark before imaging the electronic component P. The mounting machine 1 may correct the position of the central axis of the suction nozzle when the electronic component P is imaged using the position information of the reference mark.

The parts imaging device 20 includes an illuminance measuring device (not shown) that measures the illuminance irradiated to the imaging target. The illuminance measuring apparatus is configured to be able to measure the illuminance that decreases due to, for example, a decrease in the output of the side light source 25 and dirt on the cover glass 26.

(Configuration of nozzle stocker 30)
As shown in FIG. 1, the nozzle stocker 30 is provided in the vicinity of the part imaging device 20 and accommodates a plurality of types of suction nozzles (described later). As shown in FIG. 6, the nozzle stocker 30 includes a plurality of nozzle housing holes 31 for housing the suction nozzles. For example, the nozzle stocker 30 contains a non-defective product suction nozzle of the same type as the suction nozzle in use, and when the suction nozzle in use becomes defective, the good product suction nozzle stored in the nozzle stocker 30 is Replaced with defective suction nozzle. Alternatively, the nozzle stocker 30 accommodates a plurality of types of suction nozzles, and a predetermined suction nozzle is selected according to the type of electronic component mounted on the circuit board 5.

As shown in FIG. 6, a reference mark M2 and a two-dimensional code C1 are marked on the upper surface of the nozzle stocker 30. The two-dimensional code C1 stores information associating the type of suction nozzle accommodated with the nozzle accommodation hole 31. The reference mark M2 and the two-dimensional code C1 are imaged by a mark camera described later. The mounting machine 1 uses the image data of the reference mark M2 imaged by the mark camera to identify the reference mark M2 and acquire its position information. The mounting machine 1 uses the positional information of the reference mark M2 to calculate the positional deviation from the ideal position of the nozzle receiving hole 31 of the nozzle stocker 30, and based on the positional deviation, the mounting nozzle 1 of the suction nozzle from the nozzle receiving hole 31 is calculated. Take out and store in suction nozzle. The mounting machine 1 selects a desired suction nozzle from a plurality of suction nozzles accommodated in the nozzle stocker 30 using information recorded in the two-dimensional code C1 captured by the mark camera.

(Configuration of the substrate transfer device 40)
As shown in FIG. 1, the board transport device 40 is fixed to the frame 4, transports the circuit board 5 along the transport direction (X-axis direction), and positions the circuit board 5 at a predetermined work position. To do. The substrate transfer device 40 is a double conveyor type in which the first transfer device 41 and the second transfer device 41 are arranged in two rows. The first transport device 41 has a pair of

transport guide rails

41a and 41b arranged to face each other in parallel. Similarly, the second transport device 42 has a pair of

transport guide rails

42a and 42b arranged to face each other in parallel.

As shown in FIG. 7, each of the pair of

transport guide rails

41 a and 41 b of the first transport device 41 is formed with holding

plate portions

43 a and 43 b whose upper end portions protrude inward. In addition,

belt conveyors

44a and 44b are provided on the inner wall surfaces of the

transport guide rails

41a and 41b, respectively. The circuit board 5 is transported in the transport direction on the pair of

belt conveyors

44a and 44b and stopped at a predetermined work position. The first transport device 41 is provided with a clamp device (not shown). The clamp device includes a back plate that can be moved up and down by a lifting device, and a back pin that can be replaced at a predetermined position on the back plate. The clamp device raises the circuit board 5 through the back pin by raising the back plate. The circuit board 5 is sandwiched between the back pins and the back surfaces of the holding

plate portions

43a and 44b and positioned at a predetermined work position.

The opposing distance D1 between the pair of

transport guide rails

41a and 41b of the first transport device 41 is configured to be changeable according to the width of the circuit board 5 to be transported. In this example, the conveyance guide rail 41a on the left side in the figure is a fixed guide rail, and the conveyance guide rail 41b on the right side in the figure is a movable guide rail. The movable guide rail 41b is connected to a driving device (not shown), and is configured such that the facing distance D1 between the movable guide rail 41b and the fixed guide rail 41a can be changed. A reference mark M3 is marked on the upper surface of the movable guide rail 41b. The reference mark M3 is imaged by a mark camera described later. The mounting machine 1 uses the image data of the reference mark M3 imaged by the mark camera to identify the reference mark M3 and acquire its position information. The mounting machine 1 detects the position of the movable guide rail 41b with respect to the fixed guide rail 41a using the position information of the reference mark M3. The configuration of the first transport device 41 described above is the same in the second transport device 42.

As shown in FIG. 7, a reference mark M <b> 4 is marked on each of the diagonal positions on the surface of the circuit board 5. The reference mark M4 is imaged by a mark camera described later. The mounting machine 1 uses the image data of the reference mark M4 captured by the mark camera to identify the reference mark M4 and acquire its position information. The mounting machine 1 uses the positional information of the reference mark M4 to calculate the positional deviation from the ideal position of the mounting position of the electronic component P on the surface of the circuit board 5, and moves the suction nozzle based on the positional deviation. The electronic component P is mounted.

(Configuration of component transfer device 50, mounting head 60, and mark imaging device 70)
As shown in FIG. 1, the component transfer device 50 is fixed to the frame 4 and is an XY robot type. The component transfer device 50 includes a Y-axis slide mechanism and an X-axis slide mechanism. The Y-axis slide mechanism is configured to be able to move the Y-axis slider 52 in the Y-axis direction using a Y-axis servo motor 51.

FIG. 8 shows an outline of the X-axis slide mechanism, the mounting head 60, and the mark imaging device 70. The X-axis slide mechanism is provided on the Y-axis slider 52, and is configured to be able to move the X-axis slider 53 in the X-axis direction using an X-axis servo motor (not shown). The mounting head 60 and the mark image pickup device 70 are fixed to the X-axis slider 53, and the component supply device 10, the part image pickup device 20, the nozzle stocker 30, and the substrate using the Y-axis slide mechanism and the X-axis slide mechanism. It can be moved on the XY plane above the transfer device 40. The mark imaging device 70 includes a mark camera 71 and a light source (not shown). As the mark camera 71, for example, a CCD camera is used. For example, an LED is used as the light source. Further, the mark imaging device 70 includes an illuminance measurement device (not shown) that measures the illuminance irradiated to the imaging target. The illuminance measuring apparatus is configured to be able to measure the illuminance that decreases due to, for example, a decrease in the output of the light source and dirt on the cover of the light source.

The mounting head 60 is configured to be detachable from the X-axis slider 53. In the mounting machine 1, a necessary mounting head 60 is selected from a plurality of types of mounting heads 60 and used. 9A and 9B illustrate two types of mounting

heads

60a and 60b.

The mounting head 60 a shown in FIG. 9A is a revolver type and has a plurality of nozzle holders 61. Each nozzle holder 61 holds the same type or different types of suction nozzles 62. The plurality of nozzle holders 61 are arranged at regular intervals around the central axis, and are configured to be capable of index rotation. Further, each of the nozzle holders 61 is configured to be capable of rotating, and is configured to be movable up and down at one index position. As a result, the suction nozzle 62 held by the nozzle holder 61 is also rotated and can be moved up and down at one index position. The revolver-type mounting head 60a is used, for example, to hold the suction nozzle 62 corresponding to a small electronic component. The suction nozzle 62 is provided with a protruding portion 62a protruding in the radial direction, and a two-dimensional code C2 is marked on the upper surface of the protruding portion 62a (see FIG. 10). In the two-dimensional code C2, the type of the suction nozzle 62 is recorded. The two-dimensional code C2 is imaged by the mark camera when a desired suction nozzle 62 is selected from the plurality of suction nozzles 62 accommodated in the nozzle stocker 30. The mounting machine 1 performs solid identification of the suction nozzle 62 using the information of the two-dimensional code C2.

The mounting head 60 b shown in FIG. 9B is a single type and has one nozzle holder 63. The nozzle holder 63 is configured to be capable of rotating, whereby the suction nozzle 64 held by the nozzle holder 63 is also rotated. The suction nozzle 64 is also provided with a protruding portion 64a protruding in the radial direction, and a two-dimensional code (not shown) is marked on the upper surface of the protruding portion 64a. The type of the suction nozzle 64 is recorded in the two-dimensional code. This two-dimensional code is also imaged by the mark camera when a desired suction nozzle 64 is selected from the plurality of suction nozzles 64 accommodated in the nozzle stocker 30. The mounting machine 1 performs solid identification of the suction nozzle 64 using the information of the two-dimensional code.

(Configuration of image processing unit)
Hereinafter, the image processing unit will be described by taking as an example the case of recognizing the reference marks M1-M4 described above. Here, the above-described reference marks M1-M4 are different depending on the type of the marked component. Note that the same technique as described below can also be applied to confirm the two-dimensional code C1-C2 and the suction posture of the electronic component P.

As shown in FIG. 11, each of the modules 2 (see FIG. 1) of the mounting machine 1 includes an image processing unit 100. Note that the image processing unit 100 may be provided in a host computer of a management system that collectively manages a plurality of modules 2, or other substrate working machine, nozzle cleaning machine, nozzle inspection machine, feeder maintenance unit. May be provided in a host computer of a management system including

The image processing unit 100 includes a control device 110 and a memory 120 that stores image processing data. The image processing unit 100, the part imaging device 20, and the mark imaging device 70 are collectively referred to as a recognition device. The control device 110 includes a feature parameter creation device 111, a hit rate calculation device 112, a determination device 113, and a signal output device 114. The feature parameter creation device 111 creates feature parameters from the reference marks M1-M4 imaged by the mark imaging device 70. The feature parameters created here vary depending on the pattern of the reference marks M1-M4, and can be created using a known image recognition technique. For example, the feature parameter includes a pattern edge extracted from the image data of the reference marks M1-M4.

Here, the first memory 121 of the memory 120 stores standard parameters for each of the reference marks M1-M4, and the reference marks M1-M4 are associated with the standard parameters. The hit rate calculation device 112 calculates a hit rate indicating the correlation between the characteristic parameters of the captured reference marks M1-M4 and the corresponding standard parameters. The hit rate indicates the similarity between the feature parameter and the standard parameter, and indicates a higher value as the similarity is higher. Note that the hit rate calculated by the hit rate calculation device 112 is stored in the second memory 122. The second memory 122 stores the hit rate every time the hit rate of the reference mark M1-M4 is calculated by the hit rate calculation device 112, so that the time series of the hit rate of each of the reference marks M1-M4 is stored. Data is stored.

Here, an example of image processing using seek lines is illustrated. Image processing using seek lines is to perform edge extraction of an imaging target using luminance data included in image data. Image processing using seek lines is superior in high-speed processing because edge extraction is performed using only luminance data on the seek lines. Note that image processing using seek lines is disclosed in Japanese Patent Application Laid-Open No. 8-180191, Japanese Patent Application Laid-Open No. 2004-279304, and Japanese Patent Application Laid-Open No. 2006-114821 that have already been filed and published by the present applicant. Technology can be applied.

As shown in FIG. 12, when the reference marks M1-M4 have a circular pattern, a plurality of seek lines SL are preset in the vicinity of the position where the reference marks M1-M4 should exist, A reference point section in which the luminance changes abruptly is calculated from the luminance data of the plurality of reference points, and edge extraction of the reference marks M1-M4 is performed. In this example, eight seek lines SL are illustrated, but the present invention is not limited to this example. The plurality of seek lines SL are prepared in advance as templates in accordance with the types of the reference marks M1-M4, and can be referred to as standard parameters.

The hit rate is affected by deterioration due to dirt, breakage, etc. of the reference marks M1-M4. For example, as shown in FIG. 13, when a part of the reference marks M1-M4 is missing, the edge cannot be extracted on the seek line SL corresponding to the missing part (shown by hatching). In the example of FIG. 13, since edge extraction cannot be performed in one seek line SL, the hit rate is calculated as 7/8 = 87.5%. As described above, when the reference mark M1-M4 is further deteriorated due to dirt, breakage, etc., the similarity between the edge of the reference mark M1-M4 (also referred to as a characteristic parameter) and the seek line SL (also referred to as a standard parameter) decreases. Hit rate decreases.

As shown in FIG. 14, the deterioration of the reference marks M1-M4 usually progresses with time, and the hit rate also decreases. As shown in FIG. 15 and FIG. 16, the decrease in the hit rate is not limited to a linear one over time, but may decrease so as to draw a curve. The elapsed time here is the total time when the mounting machine 1 is operating, the total time including when the mounting machine 1 is operating or not operating, or other reference mark M1- This is the total time that is causally related to the degradation of M4.

As shown in FIG. 14 to FIG. 16, when the hit rate is lower than the error threshold (identification limit value), the determination device 113 determines that the reference marks M1-M4 cannot be identified. When the hit rate becomes lower than the error threshold due to the deterioration of the reference marks M1-M4, the control device 110 forcibly stops the operation of the mounting machine 1.

The determination device 113 determines whether or not the reference marks M1 to M4 need to be maintained based on the calculated hit rate. As shown in FIGS. 14 to 16, the determination device 113 determines whether or not the hit rate has fallen below the maintenance threshold. The maintenance threshold value is set to a value larger than the error threshold value of the hit rate at which the reference marks M1 to M4 cannot be identified. That is, the determination device 113 determines that the maintenance of the reference mark M1-M4 is necessary when the deterioration of the reference mark M1-M4 is progressing, even if the reference mark M1-M4 can be identified. . When the determination device 113 determines that the maintenance of the reference marks M1-M4 is necessary, the control device 110 outputs a maintenance request signal from the signal output device 114, and the input / output device 6 (see FIG. 1) uses the reference mark M1- Notify that M4 maintenance is required. The contents displayed on the input / output device 6 can specify which reference mark M1-M4 needs to be corrected. As will be described later, when the cause of the decrease in the hit rate of the reference mark M1-M4 is not the deterioration of the reference mark M1-M4 but the deterioration of other components, the control device 110 performs maintenance of the components. The input / output device 6 notifies that it is necessary.

Next, the flow of the recognition program for the reference marks M1-M4 will be described with reference to FIG. Hereinafter, the reference mark M3 (see FIG. 7) written on the upper surface of the movable guide rail 41b of the substrate transport apparatus 40 will be described as an example.

In step S1, the mark camera 71 images the reference mark M3 written on the upper surface of the movable guide rail 41b. The image data of the captured reference mark M3 is transmitted to the control device 110, and preprocessing such as noise removal is performed as necessary.

In step S2, the feature parameter creation device 111 extracts the edge of the read image data of the reference mark M3 and creates a feature parameter.

In step S3, the feature parameter and its standard parameter are matched and the hit rate is calculated. Here, in the image processing using the above-described seek line, step S2 and step S3 are performed simultaneously. That is, the control device 110 reads a plurality of seek lines (standard parameters) corresponding to the reference mark M3 stored in the first memory 121, extracts the edge (feature parameter) of the reference mark M3, and also has a hit rate. calculate.

In step S4, the determination device 113 determines whether or not the calculated hit rate exceeds an error threshold (see FIGS. 4 to 16). If the hit rate exceeds the error threshold, it is determined that the mark has been successfully identified, and the process proceeds to step S5. The hit rate at this time is displayed on the input / output device 6 and stored in the second memory 122. If the hit rate does not exceed the error threshold, it is determined that the deterioration of the reference mark M3 has progressed to the point where it cannot be identified or that there is another cause. The control device 110 forcibly stops the mounting machine 1.

In step S5, the determination device 113 determines whether or not the hit rate of the identified reference mark M3 is below the maintenance threshold (see FIGS. 14 to 16). If the hit rate falls below the maintenance threshold value, it is determined that the deterioration of the reference mark M3 has progressed to a range where maintenance is necessary or that there is another cause, and the process proceeds to step S6. If the hit rate does not fall below the maintenance threshold, the process proceeds to step S7.

In step S6, the determination device 113 reads the time-series data of the hit rate stored in the second memory 122, calculates the change rate from the time-series data of the hit rate, and the change rate exceeds the abnormal value. It is determined whether or not. Here, the rate of change is calculated as a value in which the hit rate has decreased in a predetermined time. For example, as shown in FIG. 18, when the current time is t2 and the hit rate is h2, the hit rate h1 of the past time t1 is read, and the hit rate decreases in a predetermined time (between t1 and t2). It is determined whether or not the amount (difference between h1 and h2) exceeds an abnormal value. When the rate of change in the hit rate exceeds the abnormal value, it is estimated that a phenomenon different from normal has occurred. For this reason, the determination apparatus 113 proceeds to step S7 and analyzes the cause. If the change rate of the hit rate does not exceed the abnormal value, the recognition program for the reference mark M3 is terminated, and the operation of the mounting machine 1 is continued.

In step S7, the determination device 113 analyzes the cause of the decrease in the hit rate and identifies the cause of the decrease in the hit rate. When the cause of the decrease in the hit rate is specified, the signal output device 114 outputs a maintenance request signal, and the input / output device 6 is notified so that it can be determined that the cause maintenance is necessary.

Here, with reference to FIG. 19, an example of a program for analyzing the cause of the decrease in the hit rate in step S7 will be described. In this example, it is examined whether there is a cause in the illuminance of the light source.

In step S11, the illuminance of the light source is measured. In step S12, it is determined whether the illuminance of the light source is below the illuminance threshold. If the illuminance of the light source falls below the illuminance threshold, it is determined that the cause of the decrease in the hit rate is a decrease in the illuminance of the light source. The control device 110 outputs a light source maintenance request signal from the signal output device 114, and notifies the input / output device 6 (see FIG. 1) that maintenance is necessary regarding the illuminance of the light source. If the illuminance of the light source does not fall below the illuminance threshold, it is determined that there is a cause in the reference mark M3, and the process proceeds to step S13.

In step S13, the determination device 113 determines whether there is another component that belongs to the same group as the movable guide rail 41b that is determined to require maintenance of the reference mark M3. Here, the plurality of components are divided into a plurality of groups according to the deterioration rate of the reference marks M1-M4. For example, those with the same usage frequency may belong to the same group. In the mounting machine 1, the substrate transfer device 40 is a double conveyor type in which the first transfer device 41 and the second transfer device 41 are arranged in two rows. For this reason, the 1st conveyance apparatus 41 and the 2nd conveyance apparatus 42 belong to the same group, and when the maintenance of the reference mark M3 of one conveyance apparatus is required, the maintenance of the reference mark of the other conveyance apparatus is also necessary. It is. Thus, in the example of the reference mark M3 of the movable guide rail 41b of the first transport device 41, in step S14, information on the movable guide rail of the second transport device 42 belonging to the same group is extracted, and both of them are extracted. The input / output device 6 displays that maintenance is necessary. If there is no other component that belongs to the same group as the component that is determined to require correction of the fiducial mark, only the information of the component that is determined to require correction of the fiducial mark is input / output. Displayed on the device 6.

According to the above-described mark recognition program, before the reference mark M1-M4 reaches the error threshold and the mounting machine 1 is forcibly stopped, the maintenance instruction for the deteriorated reference mark M1-M4 is given by the input / output device 6. Informed. Therefore, it is possible to maintain the reference marks M1-M4 on which the maintenance instruction is displayed during the non-operating period of the mounting machine 1. In this way, by maintaining the reference marks M1-M4 before the identification of the reference marks M1-M4 becomes impossible, a situation where the mounting machine 1 is forcibly stopped can be avoided, and the production plan is met. Smooth production becomes possible.

Further, the control device 110 may calculate a prediction time when the hit rate falls below the maintenance threshold and / or the error threshold based on the time-series data of the hit rate, and display the prediction time on the input / output device 6. Such a forecast time is useful for making a reasonable production plan. Furthermore, the control device 110 compares such a predicted time with production plan data describing the operating period and the non-operating period of the mounting machine 1, and when the predicted time is included in the operating period, The maintenance time may be planned during the non-operation period and displayed on the input / output device 6. It is useful for planning a more rational production plan.

15 and 16, when the hit rate decreases so as to draw a curve, the determination device 113 indirectly determines whether or not the hit rate has fallen below the maintenance threshold from the change rate of the hit rate. You may judge.

As mentioned above, although the specific example of this technique was demonstrated in detail, these are only illustrations and do not limit a claim. The technology described in the claims includes various modifications and changes of the specific examples illustrated above. The technical elements described in this specification or the drawings exhibit technical usefulness alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing. In addition, the technology illustrated in the present specification or the drawings achieves a plurality of objects at the same time, and has technical utility by achieving one of the objects.

10: Parts supply device 20: Parts imaging device 30: Nozzle stocker 40: Board transfer device 50: Component transfer device 70: Mark imaging device 100: Image processing unit 110: Control device 111: Feature parameter creation device 112: Hit rate calculation Device 113: Determination device 114: Signal output device 120: Memory 121: First memory 122: Second memory

Claims

A recognition device for a substrate working machine,
A reading device for reading an object;
A feature parameter creation device for creating a feature parameter from the read target;
A first memory storing standard parameters relating to the characteristic parameters;
A signal output device that outputs a maintenance request signal based on a hit rate indicating a correlation between the characteristic parameter and the standard parameter,
The signal output device is a recognition device that outputs the maintenance request signal when the hit rate has not reached the error threshold and reaches a maintenance threshold different from the error threshold.
The recognition device according to claim 1, further comprising a second memory for storing time-series data of the hit rate.
The recognition device according to claim 2, wherein the signal output device outputs the maintenance request signal even when a change rate of the hit rate exceeds an abnormal value.
4. The signal output device according to claim 2, wherein when the hit rate of one target reaches the maintenance threshold, the signal output device outputs the maintenance request signal with reference to the hit rate of another target. Recognition device.
The recognition device according to any one of claims 1 to 4, wherein the reading device is an imaging device.
The imaging apparatus includes a light source that irradiates the target, and a measuring device that measures the illuminance irradiated to the target.
The recognition according to claim 5, wherein the signal output device outputs the maintenance request signal including a light source maintenance request signal when the hit rate reaches the maintenance threshold and the illuminance reaches an illuminance threshold. apparatus.
The board working machine is a mounting machine for mounting electronic components on a circuit board,
The imaging apparatus includes a mark camera that images a first mark marked on a component part constituting the mounting machine and a second mark marked on a circuit board on which an electronic component is mounted,
The maintenance threshold includes a first maintenance threshold for the first mark and a second maintenance threshold for the second mark,
The signal output device is:
When the hit rate of the first mark reaches the first maintenance threshold, the maintenance request signal including a first request signal is output;
The recognition apparatus according to claim 5 or 6, wherein when the hit rate of the second mark reaches the second maintenance threshold, the maintenance request signal including a second request signal is output.
The first mark is marked on a plurality of components;
The plurality of components are divided into a plurality of groups;
The signal output device includes information on other component parts belonging to the same group when the hit ratio of the first mark marked on one component part reaches the first maintenance threshold. The recognition device according to claim 7 which outputs a request signal.
The board working machine is a mounting machine for mounting electronic components on a circuit board,
The recognition apparatus according to claim 5, wherein the imaging apparatus includes a parts camera that images the electronic component.
The parts camera is also imaging a first mark marked on the component existing in the imaging field when imaging the electronic component,
The maintenance threshold includes a third maintenance threshold for the electronic component and a first maintenance threshold for the first mark,
The signal output device is:
When the hit rate of the electronic component reaches the third maintenance threshold, the maintenance request signal including a third request signal is output,
The recognition apparatus according to claim 9, wherein the maintenance request signal including a first request signal is output when the hit rate of the first mark reaches the first maintenance threshold.
The recognition apparatus according to claim 9 or 10, further comprising a correction request signal output device that outputs a correction request signal of the standard parameter of the electronic component based on the hit rate of the electronic component.
At least one recognition device according to claim 1;
And a third memory for storing time-series data of hit rates obtained by the at least one recognition device.
The said 3rd memory memorize | stores the time series data of the hit rate collected from the said recognition apparatus provided in at least 2 of a substrate working machine, a nozzle cleaning machine, a nozzle tester, and a feeder maintenance unit. 12. The management system according to 12.
A prediction device that predicts when the hit rate reaches at least one of the maintenance threshold and the error threshold based on the time-series data of the hit rate stored in the third memory. The management system according to 12 or 13.
A fourth memory storing production plan data describing the operation period and non-operation period of the substrate work machine;
A planning device for creating maintenance plan data based on the prediction time predicted by the prediction device and the production plan data stored in the fourth memory;
The management system according to claim 14, wherein when the predicted time is included in the operation period, the planning device plans a maintenance execution time in a non-operation period before the operation period.