JPH03104527A - Assembly of car body - Google Patents

Abstract

PURPOSE:To manage the optimum correction quantity for a position teaching data with its grasping quantitatively and to shorten the time until obtaining a specific car body assembly accuracy, by finding the correction quantity of the position teaching data of a work positioning device with the statistical processing of a measuring data. CONSTITUTION:A statistical data is prepared based on the measurement data obtained by the measurement carried out whenever a specific car body assembly is executed, the deviation between the peak value in the histogram of this statistical data and a car body design data 40 is found and a specific correction, namely, the correction reducing the deviation at least is applied on the position teaching data of a work positioning device 11, in the case of this deviation exceeding a statistical allowable range. Due to this correction being to improve the future assembly stage based on the assessment of the product of the past assembly stage and the result of the analysis, the optimum correction quantity for the position teaching data can be managed with its quantitative grasping and the time until obtaining a desired car body assembly accuracy can be shortened.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention provides position teaching data of a workpiece positioning device that positions a plurality of workpieces constituting a car body or main parts of a car body during car body assembly, based on statistics of measurement data. The present invention relates to a method for assembling a vehicle body by making corrections using correction amounts determined through processing.

(Prior Art) As a conventional method for assembling an automobile body or the main parts of a car body, a plurality of gauges are arranged as a workpiece positioning device at a predetermined assembly position on an assembly line based on position teaching data, and these gauges are used to determine the position of the workpiece. A plurality of car body panels are mutually positioned, the body panels in this positioned state are joined together by spot welding, etc. to assemble the main part of the car body, and then another car body panel or main part of the car body is attached to such a main part of the car body. There is a method of joining to form a vehicle body, and in such a method, measurement data after assembly (
The position of the gauge is adjusted by making corrections based on the feedback of the vehicle body measurement data and the measurement data of the positioning device in a relative relationship with the vehicle body measurement data, and then the above-mentioned vehicle body assembly process is repeated (by humans). It is common to improve the accuracy of vehicle body assembly to a predetermined level.

(Problem to be Solved by the Invention) However, simply adding correction based on the feedback measurement data to the position teaching data is not enough to eliminate the deviation of this measurement data from the vehicle body design data. When simply teaching the position to move the workpiece positioning device (gauge), the amount of movement of the gauge and the amount of movement of the car body due to this position teaching are based on factors such as springback of the car body panel, welding distortion, and differences in the rigidity distribution of each part of the car body. Because the relationship is not always linear depending on the influence, it is difficult to predict what the position of the vehicle body will be in the actual assembled state. Therefore, it is necessary to quantitatively understand the appropriate correction amount for the position teaching data. was difficult. For this reason, it is necessary to repeat the aforementioned series of feedback steps many times until the desired vehicle body assembly accuracy is achieved, and a large amount of time is wasted for adjustment.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems by determining the amount of correction of position teaching data of a workpiece positioning device through statistical processing of measurement data.

(Means for Solving the Problems) For this purpose, the vehicle body assembly method of the present invention teaches the position of a plurality of workpieces constituting the vehicle body or the main parts of the vehicle body, with corrections made in accordance with feedback vehicle body measurement data. When assembling the car body or the main parts of the car body by joining each other under the positioning state based on the operation of the workpiece positioning device based on the data, the deviation between the workpiece measurement data measured for the workpiece alone and the car body design data regarding the workpiece. If this deviation is within a predetermined range, the assembly is performed as is, statistical data is created based on the vehicle body measurement data after this assembly, and the deviation between the peak value in the histogram of this statistical data and the vehicle body design data is calculated. is calculated, and when this deviation exceeds a statistical tolerance range, a correction is applied to the position teaching data to at least reduce the deviation; When assembling the vehicle body or the main parts of the vehicle body by joining them together under the positioning state based on the operation of the workpiece positioning device using the position teaching data corrected according to the vehicle body measurement data or the positioning device measurement data, First statistical data is created based on workpiece measurement data measured before and after assembly, second statistical data is created based on vehicle body measurement data measured after assembly, and peak values in a histogram of these statistical data and the vehicle body design are created. The present invention is characterized in that the deviations from the data are determined, and when these deviations exceed a statistical tolerance range, corrections are applied to the position teaching data to at least reduce the deviations.

(Function) In the method of the present invention, statistical data is created based on measurement data obtained by measurements performed each time a predetermined car body is assembled, and the deviation between the peak value in the histogram of this statistical data and the car body design data is determined. , when this deviation exceeds a statistical tolerance range, a predetermined correction is made to the position teaching data of the workpiece positioning device, that is, a correction that at least reduces the deviation.

At this time, since this correction improves future assembly processes based on the results of evaluation and analysis of products in past assembly processes, it is possible to quantitatively understand and manage the appropriate correction amount for position teaching data. , it is possible to shorten the time required to obtain the desired vehicle body assembly accuracy.

(Example) Hereinafter, embodiments of the present invention will be described in detail based on the drawings.

FIG. 1 is a perspective view showing a vehicle body assembly line used in the first and second embodiments of the vehicle body assembly method of the present invention.
Reference numeral 1 denotes a workpiece positioning device disposed at a vehicle body main part assembly stage A of this assembly line, and reference numeral 12 denotes a vehicle body measurement device disposed at a vehicle body measurement stage B of this assembly line.

This car body assembly line has stage A for assembling the main parts of the car body.
Vehicle body panels such as the main floor, left and right side panels, rear panel, cowl top assembly, shelf panel, etc. are brought in as workpieces, and these vehicle body panels are positioned relative to each other at stage A and temporarily fastened to each other to form the main body part L3. The main body part 13 is then conveyed to a stage (not shown), and at that stage, body panels such as front and rear roof rails and roof panels are further assembled to the main part 13 of the car body, and the car body 14 is assembled. As shown in the figure, the vehicle body 14 is moved to the vehicle body measurement stage B.
The assembly accuracy of the vehicle body 14 is measured on the vehicle body measurement stage B. In addition, in the stage not shown in the pre-process of stage A, the measurement of a single car body panel shall be carried out (details of panel measurement can be found in Tokushan Sho 63).
- see No. 310202, etc.).

In stage A, in order to position each of the vehicle body panels constituting the main body part 13, the workpiece positioning device 1l includes a frame 15, a number of robots 16 attached to the frame 15, and a robot 16 as a hand portion thereof. Each robot 16 on the stand is equipped with one or more workpiece receiving jigs 17, and the robot 16 can
A work clamp device 18 driven by an air cylinder is also provided as a hand section.

Here, as shown in FIG. 2, the workpiece receiving jig 17 uses a servomotor to drive a large number of adjacent strips forward and backward in the same direction one by one, and arranges the leading edges of the strips in a predetermined shape. After that, by fixing the strips integrally with a fixing device, a gauge part 17a that can position and hold the workpiece can be formed (for details, see Japanese Patent Application Laid-Open No. 59-1445).
(See No. 95 or Japanese Patent Application No. 12193/1982). Using such a function, the workpiece positioning device 11 determines the shape of each of the numerous positions to be positioned on each of the vehicle body panels constituting the main body portion l3, and determines the gauge portion of the workpiece receiving jig 17 corresponding to that position. 17a, and at the same time, the workpiece receiving jigs 17 and, by extension, the gauge portions 17a are placed at predetermined positions by the operation of each robot 16, thereby positioning the vehicle body panels with respect to each other. However, the vehicle body panel can be set by operating each work clamp device 18.

A control device l9 for controlling the workpiece positioning device 11
will be established. As shown in FIG. 2, the control device 19 includes a calculation section 20 consisting of a normal CPU, a memory 2l used for storing operation instruction signals and calculation processing, and position data of a gauge section 17a of the work receiving jig. An interface 22 for outputting human power including operation instruction signals, operation confirmation signals, and measurement data to be described later, and operation of servo motors provided for each joint of the robot 16 and servo motors for driving small pieces of the work receiving jig 17. Servo controller 2 to control
3, and a controller (not shown) that controls the operation of the piece setting device of the workpiece receiving jig 17 and the operation of each workpiece clamping device 18, based on the operation instruction signal inputted via the interface 22. Each robot 16
, the work receiving jig 17, and the work clamping device 18, and when the operations are completed, outputs an operation confirmation signal and the measurement data.

The vehicle body main part assembly stage A is also provided with a plurality of spot welding robots (not shown), and these robots weld and join the vehicle body panels to each other while they are positioned by the work receiving jig 17. Temporarily fasten those panels.

In stage B, the vehicle body measurement device l2 has a plurality of (first
Six laser beam distance measuring devices 24 (in the figure) and these distance measuring devices 24 are connected to each other by a servo motor 2 within a plane orthogonal to the direction of laser beam emission, as shown by arrow D in the figure.
5 and guided by a guide 26, and a control device 28 that controls the operations of the distance measuring device 24 and the moving device 27. Here, the distance measuring device 24 emits a laser beam, and receives the reflected light of the laser beam from the vehicle body 14 with a one-dimensional optical sensor, and from the light receiving position on the optical sensor, this device and the vehicle body 14 are detected. The moving device 27 moves the distance measuring device 24 within a certain range so that the reflected position of the laser beam on the vehicle body l4 includes a predetermined reference point 14a to be described later. let

A control device 28 is provided to control the vehicle body measurement device,
As shown in FIG. 2, the control device 28 includes an arithmetic section 29 consisting of a normal CPU, a memory 30 used for storing and arithmetic processing of operation instruction signals, measurement data, etc., and inputting operation instruction signals, operation confirmation signals, and the like. A measurement data output interface 31, a servo controller 32 that controls the operation of the servo motor 25 of the moving device 27, and a sensor controller 33 that controls the operation of the one-dimensional optical sensor of the distance measuring device 24 and inputs measurement data. Each distance measuring device W24 is moved to a predetermined position with respect to the vehicle body 14 by each moving device 27 based on an operation instruction signal manually inputted via the Inx Ace 3l, and each distance measuring device 24 moves the device The distance from to the vehicle body 14 is measured within a certain range, and when the measurement is completed, an operation confirmation signal and measurement data are output.

The car body assembly line shown in FIG. 1 is further provided with a main controller 34 having a structure called an expert system in order to operate the workpiece positioning device 11 and the car body measuring device 12 in conjunction with each other. , this main controller 34 is a normal CP, as shown in FIG.
A calculation unit 35 consisting of U, a memory 36 used for storing measurement data sent from the control device 28 and calculation processing,
A human output device 37 for direct manual input and operation confirmation by the worker, a man-machine interface 38 that connects the input/output device 37 to the calculation section 35, and a human-machine interface 38 for outputting operation control signals, operation confirmation signals, and inputting measurement data. and an interface 41 for inputting CAD data 40 recorded on a cassette tape as vehicle body design data designed by a computer-aided design (CAD) system and corresponding to the type of vehicle to be assembled. a CAD data storage unit 42 that stores CAD data, and a vehicle body assembly knowledge base 43 that contains theoretical and advanced empirical knowledge data regarding vehicle body assembly;
and an inference engine 44 that systematically performs inference.

Further, as shown in FIG. 1, a workstation 50 is provided in conjunction with this assembly line. The workstation 50 is configured similarly to the main controller 34 described above,
As shown in FIG. 2, it includes a calculation section 51, a memory 52, an interface 53, a measurement data storage section 54, a statistical data storage section 55, and a knowledge data storage section 56. The workstation 50 receives measurement data before and after assembly (data on each body panel constituting the main body part l3 and position data on the workpiece receiving jig itself) from the control device 19 manually inputted via the interface 53, and the control device 28. Vehicle body measurement data from (position data after assembly),
And panel measurement data from the control device 61 of the panel measurement device 60 (its configuration is similar to the vehicle body measurement device 12 in FIG. 2) for performing the panel measurement described above is stored in the measurement data storage section 54. Then, statistical data is created based on these measurement data and stored in the statistical data storage section 55, and statistical processing of this statistical data and interface 53 are performed.
Based on the comparison with the vehicle body design data (CAD data) from the main control device 34 input via the knowledge data storage section 56 to be memorized.

Using the various devices shown in FIG. 1, the method of each embodiment of the present invention assembles a car body while correcting the position teaching data of the workpiece positioning device based on the control program shown in FIG. 3 or 4. I do.

That is, in the first embodiment, step 1 in FIG.
Panel measurement is performed in step 01, and the deviation d between the panel measurement data and the vehicle body design data (CAD data) is determined for each panel. In the next step 102, based on this deviation d, if d<-β or d>β (β is the boundary value of whether the vehicle body panel is a manufactured defective product), in step 103 it is determined as NG (defective product). This car body panel is not used for car body assembly,
If it is an assembleable panel with -β<d<-α or α<d<β (α is a smaller value than β), the control proceeds to step 104 or later, and if it is an optimal panel with -α<d<α, the control is continued. Proceed to step 105 and subsequent steps. Note that this determination is based on the one with the lowest manufacturing precision among the vehicle body panels used for assembly.

In step 104, a correction amount of the position teaching data of the workpiece positioning device 1l is calculated (this correction amount is determined based on the deviation d, for example). This correction amount is fed back to the workpiece positioning device l1 in step 106, and in step 107, the workpiece positioning device 11 operates based on the position teaching data including this correction amount to position a plurality of car body panels as workpieces. The car body 14 or main body part 13 is assembled by joining these car body panels to each other under the following conditions, and after the assembly is completed, step 10 is performed.
Perform online measurement in step 8. This online measurement is the first
The vehicle body measurement stage B shown in the figure shall be used to measure the vehicle body 1.
4, for example, center position data and pinch dimension data of reference holes (mounting holes, positioning holes).

In the next step 109, a correction amount is learned based on the data measured in step 108. This learning involves, for example, correlating measurement data input with a first-order lag to understand trends in the assembly accuracy of car bodies as complete products as a function, and this function has become statistically significant. The learning result will be reflected in the position teaching data in step 104 as knowledge data.

On the other hand, the control is performed in step 1 from the determination in step 102.
In the case of the optimal panel proceeding to step 05, assembly is performed in the same manner as in step 107 with the stencil 105, and the stencil 1
At step 10, online measurement is performed in the same manner as step 108, and from the obtained measurement data, statistical data is created at step 111. This set of statistical data is generally in a normal distribution state, as shown in the curve l in the fifth garden, which represents the histogram of the deviation from the CAD data of the measurement data of N samples. The peak value P of this curve l and the corresponding CAD data P. In step 113, it is determined whether the deviation ΔP (ΔP=p-po) is within the allowable range (1a<ΔP<a).

Here, if the deviation ΔP is within the allowable range, the control is terminated as it is because it is not enough to require correction of the position teaching data, and if the deviation ΔP exceeds the allowable range (ΔP≦−a, ΔP
≧a), the position teaching data is corrected, and in step 114, the correction amount of the position teaching data is calculated and fed back to the workpiece positioning device 11. Here, this correction amount is, for example, as indicated by the arrow X in FIG.
However, this may not be done all at once, but may be corrected in stages over several times. As a result, the appropriate correction amount for the position teaching data can be quantitatively grasped and managed, and the time required to obtain the desired vehicle body assembly accuracy can be shortened.

Next, the second embodiment will be explained using the control program shown in FIG. 4 (however, the description of the stamps having the same content as the first embodiment shown in FIG. 3 will be omitted).

First, in step 121 of FIG.
Perform panel measurement in the same manner as in step 1, and use step 122 to record panel measurement data and CAD data (panel master data).
Wet positioning device 11 based on the deviation between the two based on the comparison with
(The flow of data in this example is summarized in FIG. 6).

In the next step 123, the jig is measured, that is, the position data of the work receiving jig 17 itself of the work positioning device (hereinafter referred to as jig position measurement data) is measured, and in step 124
A jig error is calculated from this jig position measurement data and CAD data (jig master data), and in step 125 it is determined whether this jig error is within an allowable range. If the jig error exceeds the allowable range, the step 126 corrects the position teaching data of the workpiece positioning device 11 based on the jig error and the correction amount obtained in step 122, and if the jig error is within the allowable range. If step 126 is skipped, the workpiece is positioned and assembled using the position teaching data in the next step 127.

In the next step 12B, in-jig measurements, that is, in the vehicle main part assembly stage A, various data before and after welding and assembly of the workpiece (vehicle body panel) under the restraint of the workpiece receiving jig 17, such as panel shape and panel The clamping force is measured, and first statistical data is created from the obtained measurement data in step 129. Similarly, in step 130, online measurement is performed on the vehicle body 14 after assembly is completed, second statistical data is created in step 131 from the obtained measurement data, and statistical processing of the first and second statistical data is performed in step 132. Do each.

This statistical processing is performed by creating a histogram similar to that shown in FIG.
The amount of error for the D data is calculated, and based on this amount of error, the presence or absence of a statistical error is determined in step 134. If a statistical error is found, the amount of statistical correction for the position teaching data is calculated in step 135. Then, this correction amount is fed back to the correction of the position teaching data in step 126 described above (if no statistical error is found, the control is ended as is). As a result, in this second embodiment, in addition to the effects of the first embodiment, by adding the position data of the workpiece positioning device itself and the data of each body panel in the positioning state as measurement data,
It is possible to further improve the accuracy of vehicle body assembly.

(Effects of the Invention) Thus, in the vehicle body assembly method of the present invention, since the amount of correction of the position teaching data of the workpiece positioning device is determined by statistical processing of the measured data, it is possible to quantitatively grasp and manage the appropriate amount of correction for the position teaching data. This makes it possible to shorten the time it takes to achieve the desired vehicle body assembly accuracy.

[Brief explanation of drawings]

Fig. 1 is a perspective view showing a car body assembly line used in the first and second embodiments of the car body assembly method of the present invention, and Fig. 2 is a diagram showing the configuration of equipment arranged on the car body assembly line shown in the first garden. 3 and 4 are flowcharts showing the control program for the vehicle body assembly line in the first embodiment and the second embodiment, respectively, and FIG. 5 is a histogram of the deviation of the measurement data from the CAD data in the first embodiment. The diagram shown in FIG. 6 is a diagram for explaining the flow of each data in the second embodiment. l1...Work positioning device l2...Vehicle body measuring device 13...Main part of the vehicle body l4...Vehicle body 14
a...Reference point 16...Robot 17.
... Workpiece receiving jig 17a ... Gauge part 19
．． 28...Control device 34...Main control device 40...CAD data (vehicle body design data) 50...
Workstation 60...Pafuru measurement equipment 1.61
...Control device Figure 3

Claims (1)

[Claims] 1. A plurality of works constituting the vehicle body or the main parts of the vehicle body,
When assembling a car body or main parts of a car body by joining them together under a positioning state based on the operation of a workpiece positioning device using position teaching data corrected according to feedback of car body measurement data, The deviation between the measured workpiece measurement data and the car body design data regarding the workpiece is determined, and if this deviation is within a predetermined range, the assembly is performed as is, statistical data is created based on the car body measurement data after this assembly, and statistical data is created based on the car body measurement data after this assembly. A vehicle body characterized in that a deviation between a peak value in a histogram of statistical data and the vehicle body design data is determined, and when this deviation exceeds a statistical tolerance range, a correction is applied to the position teaching data to at least reduce the deviation. Assembly method. 2. Multiple workpieces that make up the car body or the main parts of the car body,
When assembling the car body or the main parts of the car body by joining them together under the positioning state based on the operation of the workpiece positioning device using the position teaching data that has been corrected according to the feedback of the car body measurement data or the positioning device measurement data. , create first statistical data based on the workpiece measurement data measured before and after the assembly, create second statistical data based on the vehicle body measurement data measured after the assembly, and calculate the peak value in the histogram of these statistical data and the A method for assembling a vehicle body, characterized in that deviations from vehicle body design data are determined, and when these deviations exceed a statistical tolerance range, corrections are applied to the position teaching data to at least reduce the deviations.