JP2004001226A - Automatic preparation system for welding robot operation program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a welding robot operation program automatic preparation system capable of sufficiently securing teachingless enabled and highly efficient welding quality in the case of welding a work having a complicated three-dimensional shape by utilizing a plurality of welding robots. <P>SOLUTION: A work model is generated from the work shape information of CAD data, a welding model for a basic welding line 120 consisting of the fitting line of a fitting member 110 is generated and a cell model is generated by dividing a work 100 by area division lines 311 to 314 for determining the operation ranges of welding robots. The existence of interference between a work and a welding robot is checked in each basic welding line, and when interference is generated, a welding line shortened to a range having no interference generation is generated, the welding direction, order and route of the welding line are determined in each area and then welding design information is specified to prepare an operation program. <P>COPYRIGHT: (C)2004,JPO

Description

The present invention relates to the application of welding robots in heavy industries such as shipbuilding and bridges, and more particularly to a CAD / CAM system when a plurality of welding robots are applied to a large variety of small-sized and large-sized workpieces. The present invention relates to an automatic generation system for an operation program of a welding robot using the above.

Conventionally, the application of a welding robot in heavy industries such as shipbuilding, bridges, and steel frames has taken a simple application form in which one welding robot is applied to one work. In such an application form, work efficiency has been improved by eliminating teaching using a welding robot at a site using an off-line teaching method using a personal computer.
However, in an application mode in which a plurality of welding robots are applied to one work, teaching is necessary even in an offline mode, so that it takes too much time for teaching on a personal computer. It was difficult to apply to the production. As a method for solving this problem, the technique disclosed in Patent Document 1 of the applicant's prior application has been used. This automatically generates operation programs for a plurality of welding robots collectively using only the workpiece shape information and the welding design information of the CAD system, and enables teaching-less. However, in this system, the operation program of the welding robot is automatically determined by a relatively simple algorithm.
JP-A-6-214625

If the target work is of a large variety and small quantity, and is a work with a complicated shape including many narrow parts due to three-dimensional mounting members, the welding robot operation program including the welding path is automatically executed using a simple algorithm. I can't decide. In particular, the determination of the welding direction and the welding sequence in consideration of the welding quality and work efficiency largely depends on the experience of the robot operator. Therefore, it has been difficult to automatically determine the welding path by the computer.
Further, in an application form in which a plurality of welding robots simultaneously perform welding of one or a plurality of workpieces at the same time, one of the welding robots stops for control in order to avoid interference between the welding robots. As a result, a large amount of dead time is generated. Determining the welding direction and welding order to avoid such interference between welding robots has been a major issue.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and in the case of welding a work having a complicated three-dimensional shape using a plurality of welding robots, teaching-less is possible and more efficiently. It is an object of the present invention to provide an automatic generation system of a welding robot operation program that can sufficiently secure welding quality.

An automatic generation system of a welding robot operation program according to the present invention reads data registered in a CAD / CAM system and welds each of the welding robots when a workpiece having a plurality of mounting members is welded by a plurality of welding robots. In the system that automatically generates the operation program of
Means for generating a work model corresponding to the work from the work shape data of the read data,
Means for generating a welding model having a basic welding line consisting of attachment lines of the constituent members of the work,
Means for generating a cell model by dividing the work model by a region dividing line so as to equalize the workload of the plurality of welding robots,
Means for checking whether or not the workpiece and the welding robot interfere with each other for each of the basic welding lines, and a means for generating a welding line in a range where interference does not occur;
Means for determining a welding direction, a welding order, and a welding path for each of the regions divided by the region dividing line, for all the welding lines after the interference check included in the region,
Means for generating a series of operation programs by allocating data necessary for welding by selecting an operation sequence group called an operation pattern previously determined from the data, for the determined welding line,
It is characterized by having.

Here, the data registered in the CAD / CAM system is mainly CAD data. As shown in Table 1, the CAD data includes work management information, work shape information, welding line information, It is classified into four types of welding design information, and by reading these information, a process of generating a work model and a welding model is performed.

(4) The work management information and the work shape information are data required when a work model corresponding to the work is generated, and data such as geometric data and material are registered for each component. In addition, welding line information and welding design information are data required when generating a welding model. Among them, data on the cross-sectional shape of the mounting member, that is, scantling data, checks for interference with the welding torch and robot arm. It is the basic data when doing.

The work model generation means generates a work shape based on work management information and work shape information. For example, in the case of an outer panel for shipbuilding, a work model is generated for the outer shape of the panel of the base material and the arrangement of various mounting members such as a longge and a transformer mounted on the panel.

(4) The welding model generation means generates a basic welding line that is the basis for attaching the constituent members of the work. In the case of a mounting member, two mounting lines are generated in principle, taking into account the thickness. Generate a welding model consisting of three-dimensional basic welding lines.

The cell model generating means divides the work by a region dividing line that defines the operation range of the plurality of welding robots, and generates a cell model for each divided region. In dividing the area, the work amount of the welding robot is set to be equalized. Further, the basic welding line may be divided by the region dividing line.

The interference checking means checks whether the mounting member and the welding robot interfere with each other for each basic welding line. At the time of this interference check, the basic robot posture is fixed to a plurality of types, and based on the basic robot posture, an interference check is performed on the interference candidate member extracted from the mounting member and its surrounding members, and interference occurs. The torch angles α and β are determined within an allowable range. Here, the basic robot posture is a welding posture in a torch angle range necessary for ensuring welding quality without causing interference with the mounting member. Therefore, the posture of the robot changes depending on whether the depth of the mounting member is small or deep.
The reason why the basic robot posture is fixed to a plurality of types is that when several multi-joint welding robots are suspended on external axes, which are three orthogonal axes, respectively, to weld a complex three-dimensional workpiece, a certain number of types are used. This is because it is easier to determine the crab welding posture (robot posture) and perform the interference check. The use of the predetermined basic robot posture and the scantling data of the mounting member, that is, the scantling change point and the MAX scantling data, makes it possible to easily perform an interference check by automatic processing. .

The members to be subjected to the interference check are the attachment member (own member) and the proximity members located on the back side and the side surface of the attachment member. Among the proximity members, the members overlap the extraction area set for the own member. The member is selected as an interference candidate member. Therefore, interference is checked for the own member and the members extracted as the interference candidate members among the back surface member and the side surface member, respectively, using the scantling data. Then, if interference occurs as a result of the interference check, a weld stop point is generated at the welding line.

に For each area divided by the area dividing line by the cell model generating means, determine the welding direction, welding order, and welding path for all the welding lines included in the area. If the welding line satisfies a certain condition, it is selected as a twin welding line. All others are single welding lines.

The present invention is based on the well-known techniques such as high-speed rotation arc welding, end position detection of an attachment member by an arc sensor, automatic welding line scanning control, and bead joining. A predetermined rule is set so that the above technology can be utilized. As this rule, for example, the end of the welding line is determined so that the end position can be detected by the arc sensor.
Also, when the welding line is divided by the area dividing line or when the welding line is divided by another mounting member connected to the mounting member, the bead joint is also formed on such a divided welding line. For the possible welding lines, the welding direction is set to be the same so that the welding quality of the bead joint is ensured. Further, in principle, a direction toward the counterclockwise direction with respect to the mounting member is defined as a welding direction.

In the welding order determining means, for example, a rule is set so that, for a mounting member having a groove or an inclination (falling), a side on which a groove is processed or a wide angle side of a mounting angle is welded first. . In principle, the welding order is determined in principle such that the surface where the welding start point includes the welding line closest to the robot system coordinate origin is determined on the mounting member, and then the welding start point of the welding line group in the determined surface is determined. The welding line closest to the robot system coordinate origin is ranked first.
These basic rules of the welding direction and the welding order are designed to minimize the occurrence of interference between welding robots.

The welding path determining means determines the welding paths of all the mounting members according to a predetermined algorithm based on the welding direction and the welding order determined as described above. The welding path determining means includes an interlock area setting means for setting an interlock area for avoiding mutual interference of welding robots, and a welding line selecting means for selecting a welding line included in the set interlock area. Configuration. Further, as a predetermined algorithm, the interlock area setting means may be configured such that, when the operating area of the plurality of welding robots is divided into three in the vertical or horizontal direction, the number of welding lines even partially included in the area on the right side or the upper side is the largest. And the area on the right or upper side of the selected division direction is set as an interlock area. When the welding line selection means divides the operation area of the welding robot into three in the selected division direction, A welding line group (LCR welding line group) extending over three regions is selected, and a welding line group (LC welding line group or CR welding line group) included in one or both of the two regions is selected. The welding path is determined in the order of the LCR welding line group, the LC welding line group, and the CR welding line group.

For the welding line for which the welding direction, sequence, and route have been determined in this manner, an operation sequence called an operation pattern that has been prepared in a database in advance is utilized by utilizing the geometric characteristics of the registered CAD data. By selecting, data necessary for welding is given, and a series of operation programs is generated.
Then, after confirming by the operation simulation, the operation program is transmitted to each welding robot, and the welding robot is actually operated.

As described above, according to the present invention, it is possible to automatically generate operation programs for a plurality of welding robots using only work shape information and welding design information of CAD data registered in a CAD / CAM system. Automatic welding of large-sized workpieces with complicated three-dimensional shapes can be performed with high efficiency and high quality.
In addition, using the cross-sectional shape data (scantling data) of the mounting member and the basic robot posture fixed to a plurality of types, the interference check between the mounting member and neighboring members around the mounting member is performed. There is an advantage that interference check can be easily performed.

FIG. 1 is a flowchart of the CAM system according to the system for generating a welding robot operation program of the present invention, and FIGS. 2 to 6 are sub-flowcharts in each processing of steps S2 to S5 in FIG. FIG. 7 is a plan view showing an example of a work to be welded by the CAM system, and FIG. 8 is a side view. The work illustrates the case of a shipbuilding panel. 7 and 8 show examples of the arrangement of works on a portable surface plate. Usually, two or more works 100 are combined and arranged. Reference numeral 200 denotes a surface plate for setting the work 100, and the work 100 is supported by the tips of a number of bar-shaped projections 201 provided on the surface plate 200. When the support surface of the work 100 is curved or inclined, the support is performed by changing the length of the protrusion 201 as shown in FIG.

FIG. 9 illustrates an outline of a robot mechanism in which a plurality of welding robots are combined in one welding robot stage. An articulated robot 300 having six rotation axes is further provided with three axes of orthogonal axes as external axes. A robot mechanism that is supported by a portal frame 304 in a ceiling-hanging position so as to move up and down, left and right, and back and forth by sliding mechanisms 301, 302, and 303, and is a combination of ten welding robots 300 having nine axes in total. It has become. A well-known welding torch 1 capable of high-speed rotation arc welding is attached to the tip of each welding robot 300. This also uses a known arc sensor to detect the end position of the mounting member, detect a bead joint, and detect a welding line. Automatic copying control is possible. The work 100 supported on the surface plate 200 is carried into the lower part of the robot mechanism by a conveying means such as a conveyor, and the welding robot 300 performs welding on the work 100. In addition, as shown in FIG. 7, region dividing lines 311, 312, 313, and 314 that define the operation region of each welding robot with respect to the workpiece 100 are set vertically and horizontally. Therefore, some of the welding lines of the mounting member are divided by these region dividing lines.

First, the outline of the shipbuilding panel will be described.
The panel base material 101 of the work 100 is composed of three panel pieces 102, 103, and 104 that have been subjected to butt welding in advance, and are joined by butt welding lines 105 and 106, respectively. It is assumed that an opening peripheral member 108 that forms the periphery of the opening 107 is attached to the center panel piece 103 in the previous step. The attachment member indicated by the reference numeral 110 is a member to be welded to the panel base material 101 in accordance with a procedure described below.
Here, since the respective mounting members 110 are arranged symmetrically, the left portion from the center of the workpiece 100 will be described. The right part will be referred to as necessary.
The attachment members on the left side are denoted by A, B, C,... S as shown below. Crosspieces A and B are mounted over three panel pieces 102-104. The vertical members C, D, E, F, G, H and the horizontal members I, J are attached to the left panel piece 102. The cross members K and L are attached across the left panel piece 102 and the center panel piece 103. The vertical member M, the diagonal members N and P, and the reinforcing members Q, R and S of the opening peripheral member 108 are attached to the center panel piece 103.

Next, a generation procedure of a welding robot operation program created for welding the workpiece 100 as described above will be described with reference to FIGS.

(1) Reading CAD data (step S1)
Since the CAD system outputs the CAD data shown in Table 1 above, the work shape information is sequentially read and stored in a predetermined area in the CAM system while referring to the work management information in the CAD output data.

(1-1) Work management information Work management information is used when work shape information and welding line information for a specific work are repeatedly extracted. As shown in Table 1, since the name of the assembling process is described for each mounting member, only the welding line information relating to the mounting member for the base material can be extracted at the corresponding welding robot stage.

(1-2) Work shape information Work shape information is output as data of position information from a specific work coordinate origin. This work shape information is used for checking the interference between the work and the robot in the robot simulator.
FIG. 10 shows a method of defining vertex data (V1 to V8), ridgeline data (E1 to E12), and surface data (F1 to F6) as work shape information.

(1-3) Welding line information The welding line information defines information on welding for each member. This welding line information is information serving as a basis for generating a welding robot program.
When welding line information of the base material is used as the basis of welding line information as in the case of shipbuilding and bridge members, welding line information output from CAD is output only on one side (reference plane) of the member, and is output on the opposite side ( The welding line information on the back side may be generated in a welding model generation process.

(A) Member name This is the name of the mounting member.

(B) Member type The type of the member (for example, deck, trans panel, long, frame, etc.). Basically, this is a design classification, but it is used as a selection parameter of a welding condition database for the purpose of determining welding conditions for each member type and controlling the welding order.

(C) Member end type The shape classification of the member end (for example, flat, scalloped, smip, etc.). Used to determine the welding process at the end of the member.

(D) Target surface data Used to specify which side of the mounting member is the welding line.

(E) Front and back type When the mounting line information of the base material is the basis of the welding line information as in the case of shipbuilding and bridge members, the welding line information output from CAD is output only on one side (reference plane) of the member. Instead, the welding line information on the opposite side (back side) may be generated in a welding model generation process. Here, the reference plane side is defined as front, and the opposite reference plane side is defined as back.

(F) Weld line data A weld line number and a weld point number related to the weld line are defined. In the welding model generation processing, the basic welding line is the welding line in this unit.

(G) Processing step It shows in which step the welding line is processed. If a name other than the processing step name of the corresponding robot is described, it means that welding is performed in another step.

(H) Welding process To change the welding process for each welding line, specify the welding execution process name. If the welding process is not a robot, even if the processing steps are the same, they are excluded from the robot.

(I) Weld point data This is position data of a weld point included in a welding line. Generally, the end point of the member is shown, but when the member is not a straight line, the change point is also output. Also, it is output at the point where the shape (skunt ring) of the mounting member changes.

(J) Weld point type This is the type classification of the weld points. This classification characterizes the welding point. For example, it can be recognized that the scantling change point is a point used for interference check.
The position where the welding point is output from the CAD is as follows.
1) Position to be the start or end point of welding (example: end of member)
Since the welding direction has not been determined, it cannot be determined at this stage which will be the welding start point.
2) The position where the scan ring of the mounting member is changing. 3) The position where the mounting angle of the mounting member is changing.
4) The position at which the mating member changes (eg, the joint part of the base material)
5) Position where the welding direction is changed (other than arc, eg bent member)
6) Both end positions and intermediate positions of the arc 7) Position where the inclination angle of the welding line changes (if the angle changes gradually, a predetermined angle pitch)
8) Position where welding design information changes (eg, change point of leg length)

(K) Scanning type This is a classification of the sectional shape of the member. By defining a member shape by a member cross-sectional shape called a scan ring, a major feature is that interference check with a welding torch and a robot arm can be performed more easily than an interference check using a mere three-dimensional model.

(L) Scanning data This is shape data that can define the cross-sectional shape of a member. FIGS. 11A to 11F show scanling data of some member types. In the drawing, L1 represents the web height, L2 represents the web thickness, L3 represents the flange width, L4 represents the flange thickness, and L5 represents the distance between the web surface and the flange end or between the web end and the flange surface (in the case of a deformed member). .

(M) Mounting Angle As shown in FIG. 12, the mounting angle θF of the mounting member. By rotating the scan ring data by this angle, the shape of the mounting member in the sectional direction can be defined. Wmax and Hmax in the figure represent horizontal and vertical MAX scanning data at the mounting angle θF.

(N) Inclination angle As shown in FIG. 13, when the welding line is inclined, how much the welding line is inclined in the front-rear direction of the welding progress with respect to the horizontal plane (vertical inclination angle θL), and , Information on how much the camera is inclined in the left-right direction (left-right inclination angle θT). It is a parameter for selecting welding conditions in inclined welding.

(O) Name of mating member This is the name of the mating member to which the mounting member is attached. Generally, the base material is the mating member. As an exception, other attachment members may be mating members. For example, a vertical welding line at the intersection of two mounting members. When performing an interference check, a member shape is referred to by this name.

(1-4) Welding design information This is information on what kind of welding is to be performed in design.

(A) Welding method name This is a classification of welding methods such as fillet welding and butt welding, and a classification of welding posture such as horizontal and vertical.

(B) Groove type This is a classification of groove shapes such as a groove groove, an X groove and a V groove.

(C) Groove data This is shape data on the classified groove shapes. When the groove type is fillet welding, the leg length is defined. FIG. 14 shows an example of fillet welding and V-shaped groove data. In the figure, θv is the groove angle, BL is the groove width, Ga is the root gap, and tb is the throat thickness.

(2) Generation of work model (step S2)
As shown in FIG. 2, the work model generation process includes generation of a panel base material (step S21) and generation of an attachment member (step S22).
That is, referring to the work management information in Table 1, first, a panel base material as a base is generated by connecting several base material shape data (step S21). Next, a three-dimensional three-dimensional work shape is generated by repeating the process of reading the workpiece shape information of the mounting members mounted on each base material and further reading the mounting member information to be secondarily attached to each mounting member. (Step S22).
In the work model generation processing, the following processing is performed.

(2-1) Member Shape Stretching Process When generating a work model, as shown in FIG. 15, by specifying each of the magnifications u, v, and w in the X, Y, and Z3 directions with reference to a specific point P0, Change the member shape. Point P ′ is the coordinate position when the coordinates of point P are extended by a predetermined calculation formula. This is a processing on processing in consideration of a change in shrinkage after welding.

(2-2) Mirror processing In the case of a work that is symmetrical with respect to a specific center line such as the shape of a ship, a symmetrical shape is generated by performing data conversion on the symmetrical line. FIG. 16 shows an example of the mirror processing. A left-right symmetric shape is generated by inverting the left figure shape with respect to the symmetry line 10. This can simplify the data generation processing.

(2-3) Reverse Distortion Addition Processing When welding a panel-shaped workpiece, a problem is thermal deformation after welding. Normally, a straightening process is performed by a press or the like after welding, but if reverse distortion is added to the panel before welding, thermal deformation is suppressed. For this purpose, it is a process of deforming the work shape on the CAM. Specifically, a line segment that gives reverse distortion and the amount of reverse distortion are specified, and only the component in the height direction of the work is changed.

FIG. 17 is a flowchart of a processing screen in the inverse distortion adding processing.
First, when a button 401 on the inverse distortion processing screen 400 is pressed, a small group name list 402 is displayed. When one small set 403 is selected from the small set name list 402, the shape 404 of the selected small set 403 is graphically displayed. Then, a reference line 405 for giving reverse distortion is designated by two points Pa and Pb. Next, a point PC is designated on the inverse distortion reference line 405. Next, a window 406 for inputting the amount of inverse distortion is displayed, and the amount of inverse distortion 407 is input. A message 408 indicating that the inverse distortion processing is being performed is displayed, and the inverse distortion processing is performed. When the inverse distortion processing ends, the screen returns to the original screen 400.

18 and 19 show how to give the inverse distortion amount.
The inverse distortion reference line 405 is designated by two points Pa and Pb, and a position Pc at which reverse distortion is applied on the inverse distortion reference line 405 is designated. As shown in FIG. 19A, when the strain amount λc at the point Pc is set, each point of the member and the welding line between Pa and Pc changes the strain amount λ corresponding to the reverse strain reference line 405 in the Z direction. to add. Similarly, a distortion amount λ corresponding to the inverse distortion reference line 405 is added between Pc and Pb. In addition, each point of the member and the welding line other than between Pa-Pc-Pb has zero strain. In the above example, the number of intermediate points Pc between Pa and Pb is one, but a plurality of points can be specified as shown in FIG.

(3) Generation of welding model (step S3)
As shown in FIG. 3, the welding model generation processing includes generation of a basic welding line (step S31), generation of a back surface basic welding line (step S32), deletion of an unnecessary welding line (step S33), and connection of the basic welding line. (Step S34) and generation of a welding line (Step S35). Based on the welding line information output from the CAD data, the basic robot posture, torch angle, and interference avoidance required for robot welding for each welding line. This is a step of generating operation data such as the above, and determining welding conditions.

(3-1) Generation of basic welding line (step S31)
One welding line is generated by joining the welding point data of the welding line information. Since a portion where the attachment member needs to be welded is recognized, this continuous line segment will be referred to as a basic welding line. Basic welding lines include not only straight lines but also arcs and broken lines.

(3-2) Generation of Backside Basic Welding Line (Step S32)
If the welding line information output from the CAD data is defined only on the front surface (reference surface) of the member, the back surface is located at a position separated by the thickness using the thickness data of the scantling data. A basic welding line is generated. When welding line information on the back side is output, the information is read as the welding line information. If the back surface welding line information is not output, the back surface information is copied when the welding line is generated.

(3-3) Deletion of unnecessary welding line (step S33)
If the welding line information output from the CAD data includes that of a plurality of welding robot stages, or if the same stage is manually pre-welded, the welding line information is deleted.
Specifically, in the following cases, the welding line is excluded from the target.
1) Different processing steps 2) Different welding processes

(3-4) Connection of basic welding line (Step S34)
A welding line whose welding start or end position is the same as another basic welding line is connected to form a single welding line. For example, when two members having different plate thicknesses are welded in the previous process to form one attachment member, two welding lines are output from the CAD. In this case, since welding can be performed continuously, one basic welding line is used. The connection conditions and connection disable conditions of the welding line are as follows.
a. Welding line connection conditions 1) The processing steps of the two basic welding lines to be connected must be the same.
2) The welding process of two connected basic welding lines is the same.
3) The welding design information of the two basic welding lines to be connected must be the same.
b. Conditions for disabling connection of welding lines 1) If two or more basic welding lines are connected and the operating range of the welding robot is exceeded, the connection is disabled (turning range is exceeded).

(3-5) Generation of welding line (step S35)
Based on the basic welding line, the welding torch posture and the basic robot posture are determined, and interference between members near the welding line and the welding robot is checked. That is, the nearby members are classified into three members, a self member, a back member, and a side member, interference with each member is checked, and interference avoidance processing is performed. As one of the interference avoiding processes, the basic welding line may be divided into a plurality of welding lines. By this interference avoidance processing, a welding line at which the welding robot can operate is determined. At that time, some new welding points are created and added.
An executable welding line for robot welding that has undergone these processes is simply defined as a welding line.
a. Welding point added by interference avoidance processing 1) Point where torch angles α and β change by interference avoidance processing 2) Point where welding line is divided by interference avoidance processing b. Case where welding point is deleted by interference avoidance processing 1) Interference check is performed at the own member scantling change point, but it is not necessary to change torch angles α and β 2) Even if mounting angle changes The torch angles α and β do not need to be changed c. Other 1) When the welding design information does not change even if the mating member changes

FIG. 4 shows a flowchart of the interference check process in the welding line generation process.
This interference check process is configured by first checking the interference of its own member (steps S351 to S354), then checking the interference of the back member (steps S355 to S359), and checking the interference of the side members (steps S360 to S365). I have.

(3-5-1) Generation of self-member interference check point (step S351)
A scantling change point included in the welding line information of the attachment member is generated as a self-member interference check point.

(3-5-2) Determination of Torch Angle αmax (Step S352)
As shown in FIG. 20, at the interference check point (scannt change point) of its own member, the interference between each scant ring point K0 to K12 and the welding torch outline T1 to T3 at the robot tip is checked. The torch angle αmax at which the line segment T is in contact with any of the points is determined. The torch angle α is a tilt angle of the welding torch 1 with respect to a horizontal plane. α ranges from 22.5 ° to 90 °. In the case of fillet welding, α is generally 45 °.

(3-5-3) Determination of Basic Robot Posture (Step S353)
When the torch angle changes as a result of the interference check, the position of the connection point between the robot arm 305 and the welding torch 1 also changes. Therefore, using the torch angle αmax, as shown in FIG. 21, the torch connection point r1 of the arm outline approximation line segment R1 in the basic robot posture is corrected for the own member.

(3-5-4) Check for interference with own member (step S354)
Using the MAX scanning data of the own member, the interference with the arm outline approximating line segments R1 and R2 is checked, and if the interference occurs, it is checked with the next basic robot posture.
Here, the basic robot posture refers to a posture determined in a plurality of forms as shown in FIG. 22, and is a robot posture that can correspond to a depth depth by an attachment angle of an attachment member or the like. The range of the hatched portion 310 in FIG. 22 is the interference range between the robot arm and the workpiece in the basic robot postures 1), 2), and 3), respectively. In the figure, Ti and To are the inner and outer ridges of the welding torch 1, Ri1 and Ri2 are the inner lines of the robot arm 305, and Ro1 and Ro2 are the outer lines of the robot arm 305.

(3) The basic robot postures 1) to 3) are determined as three types of horizontal offset amounts LH between the tip of the welding torch 1 and the center of the support block 306 of the robot arm 305, and are set as the respective basic postures. At the time of the interference check, first, the presence or absence of interference is checked in the basic robot posture 1), and in the case of interference, the next basic robot posture 2) is checked. Further interference is checked in the basic robot posture 3). If the interference cannot be avoided, the welding line is divided at that point, and the division point is additionally generated as a new welding point. By thus limiting the basic robot posture to some, automatic collision avoidance processing on a computer is enabled.

(3-5-5) Selection of interference candidate member for rear member (step S355)
Next, an interference candidate member on the robot back surface member is selected for the basic robot posture determined as a result of the interference check with the own member. For this purpose, first, as shown in FIG. 23, an area for extracting an interference candidate member is set around the own member, and a member covering this area is set as an interference candidate member.
In FIG. 23, Wc1 × Lmc is a region for extracting a back surface interference candidate member in the basic robot posture 1), and Wc2 × Lmc is a region for extracting a back surface interference candidate member in the basic robot posture 2). Although an area for extracting the back surface interference candidate member in the basic robot posture 3) is not shown, an area of Wc3 × Lmc can be set in the same manner. Lsc and Les are areas for extracting interference candidate members as side members, which will be described later. Lc is the total length of the extraction area, and t is the plate width (including the flange width) of the own member.

(3-5-6) Generation of Interference Check Point of Back Member (Step S356)
By the above method, when the interference candidate member on the back surface is extracted, a perpendicular line is drawn down from the scantling change point of the interference candidate member to the welding line of the own member, and an intersection with the perpendicular line is set as an interference check point. Generate.
For example, when the determined basic robot posture is 1), the members that hang over the area Wc1 are the B member and the D member in FIG. Therefore, an intersection P1 between the scantling change point Kb1 of the member B and the perpendicular drawn down to the welding line 120 of the member and the connection points P3 and P4 between the member D and the member are generated as interference check points. When the basic robot posture 2) is determined, the interference candidate members are the B member, the C member, and the D member, and the interference check points are calculated from the P1 of the B member and the scantling change point Kc1 of the C member. The intersection point P2 with the lowered perpendicular is the point P3, P4 of the D member.

(3-5-7) Determination of Torch Angle αmin (Step S357)
At each of the above interference check points, as shown in FIG. 25, the interference between the torch outlines T4 and T5 and the scantling points K0 to K3 is checked, and the torch at which any point of K and the line T are in contact with each other. Obtain the angle αmin.

(3-5-8) Interference check with back member by basic robot posture (step S358)
As in step S353, the torch connection point r4 of the arm outline R3 in the basic robot posture is corrected using the torch angle αmin as shown in FIG. 21, and the torch connection points are selected as interference candidates with the line segments R3, R4. An interference check with the back member is performed at each interference check point.

(3-5-9) Determination of Torch Angle α (Step S359)
The torch angle α is determined from αmax and αmin at each interference check point (scanning change point) obtained by interference check with the own member and the back member.
That is, if αmin ≦ 45 ° ≦ αmax, the torch angle α = 45 °. This angle is the standard torch position that ensures stable quality in fillet welding.
In addition, when the basic robot postures 1) to 3) cannot be obtained by the above-described interference check processing, or when the torch angles αmin and αmax are inappropriate values for welding, for example, αmax <22.5 ° and αmin> 60 °. In the case of, the basic welding line is divided at the interference check point.

(3-5-10) Selection of Interference Candidate Member for Side Member (Step S360)
In FIG. 23, members extracted as interference candidate members of the side surface members are:
1) A member hanging on the start end side region Lsc or the terminal end side region Lec. 2) A member hanging on two regions Lsc and Wci (i = 1 to 3) or Lec and Wci. However, the member of 2) has already been subjected to the interference check as described above as an interference candidate member of the back surface member, and as a result of the interference check processing, if no interference occurs, there is no interference with the side member. This means that the side member is not selected. However, since the member 2) relates to the start or end of the welding line of the own member, if interference occurs, the side member is further selected and the following interference check is performed.

(3-5-11) Generation of Interference Check Point of Side Member (Step S361)
As shown in FIG. 26, when A is a self-member and B is a side member extracted as an interference candidate member, the side member B has already been subjected to interference check as a back member, and here, interference occurs at point P1. To do. That is, when the posture of the welding torch 1 is moved from the point P2 where the end point B0 of the member B is projected onto the own member A while maintaining the posture of the welding torch 1 at α = 45 ° and β = 0 °, the outer line of the torch becomes the member B. The contact point Bc is determined, and the intersection P1 with the perpendicular drawn down to the own member A is determined from the point Bc, thereby generating the point P1 as an interference check point.

(3-5-12) Adjustment of Torch Angle γ (Step S362)
At the interference check point P1, the torch angle γ is adjusted so that the welding torch 1 is parallel to the inclination of the side member B as shown in FIG. The torch angle γ is adjusted by rotating the joint shaft 307 of the first robot arm 305a and the second robot arm 305b as shown in FIG.

(3-5-13) Adjustment of Torch Angle β (Step S363)
As shown in FIG. 26, the torch angle β is adjusted so that the welding torch 1 is parallel to the surface of the side member B at the interference check point P1. Here, the torch angle β is the angle of inclination of the welding torch 1 in the horizontal plane with respect to the welding line, and generally β = 0 °, but β is changed within the range of −45 ° to + 45 ° at locations where there is a risk of interference. .

(3-5-14) Check for interference with own member and side member (steps S364, S365)
At each point of P0 and P1, an interference check with the own member A and the side member B in the torch postures of α, β, and γ is performed.

1) Check for interference with own member If welding torch 1 is moved from point P1 to point P0 with torch angle α determined as described above, welding torch 1 may interfere with own member A. For example, as shown in FIG. 29 which is a view taken in the direction of the arrow a in FIG. 26, when the flange width or the web height of the own member A changes in this section, interference occurs.
Accordingly, when interference occurs, α is corrected to 45 ° or less or 45 ° or more within the range of αmin and αmax obtained as described above. If the corrected value of α is smaller than the allowable lower limit value or larger than the allowable upper limit value, the point P1 is generated as a welding end point (end point of welding).

2) Check for Interference with Side Member Even with the torch angles β and γ adjusted as described above, the welding torch 1 may interfere with the side member B as shown in FIG. FIG. 30 is a view on arrow b in FIG. 26. If such interference occurs, β is changed within the above-mentioned angle range. If β does not fall within the allowable angle range, point P1 is generated as a welding end point (a welding stop point).

3) Result of interference check Table 2 shows the presence and absence of interference at points P0 and P1.

複数 個 Since a plurality of works are arranged on the surface plate 200, the processing of the above steps S1 to S3 is repeated n times (see FIG. 1).

(4) Generation of cell model (step S4)
As shown in FIG. 5, the cell model generation processing includes selecting a target work (step S41), arranging the work (step S42), dividing an area (step S43), generating a welding line (step S44), and a twin welding line. (Step S45), determination of the welding direction (step S46), determination of the welding order (step S47), and determination of the welding path (step S48).

(4-1) Selection of target work (step S41)
A work to be welded is selected from the work management information in Table 1 described above.

(4-2) Arrangement of work (Step S42)
A work to be welded is arranged on a cell model composed of the following robot arrangement data.
a. Robot location data 1) Origin position of each robot in the system coordinate system: ROBOT1 (X1, T1, Z1)
2) External axis operation area in the system coordinate system:
(X1, Y1, Z1, X2, Y2, Z2, X3, Y3, Z3, X4, Y4, Z4)
3) Robot's area dividing line in the system coordinate system:
(X1, Y1, Z1, X2, Y2, Z2, X3, Y3, Z3, X4, Y4, Z4)
4) Width of overlap area

b. Automatic work placement processing Checks whether the work outline approximation rectangle fits within the rectangle that indicates the robot operation area.
As the workpiece approximate outer shape, in addition to a quadrangle, a right triangle can be considered. In this case, the same processing is performed.
When a plurality of worksets are possible in the square of the surface plate, a setting with a small number of divisions of the basic welding line by the region dividing line is selected.

FIG. 32 is a flowchart of the automatic work placement processing.
First, the selected n works are read from the work list (step S421). Next, for example, in the case of a work shape as shown in FIG. 33A, the coordinates of the vertices B1, B2, B3 and B4 of the base material are obtained from the vertex data of the base material as shown in FIG. 33B (step S422). ), Outline approximate line segments B1B2, B2B3, B3B4, B4B1 are obtained (step S423), and an approximate rectangle including B1 to B4 is obtained based on the corner sides (step S424). However, rectangles whose outlines extend outside the rectangle are excluded. In the case of FIG. 33A, the line segment B3B4 has a convex outer shape, which makes it difficult to arrange the work. Therefore, a square based on this line segment is excluded.
Then, the one having the smallest area as shown in FIG. 4C is selected as the optimum approximate outer shape (step S425). The length of the long side of the work approximating rectangle is set to BL, and the length of the long side is compared with the platen area size to determine the setting of the work (step S426).

In step S427, the squares included in the surface plate are classified into the following three cases. FIG. 34 shows the outer shape of the surface plate 200 and the element squares of the surface plate included therein. RW and RL are the horizontal and vertical widths of the element rectangle.
As shown in FIG. 34, for the element rectangles A to E, the long side length BL of the work approximation rectangle and the platen widths RW, RL are sequentially compared to determine the work setting (steps S428 to S430).
1) Case 1: BL ≦ RW
In this case, the work is set horizontally in the A element as shown in FIG.
2) Case 2: RW <BL ≦ RL
In this case, the work is set vertically in the A element as shown in FIG.
3) Case 3: RL <BL
In this case, since it does not fit within the A element, it is compared with the next element rectangle B as described above.

Next, in the work setting determined as described above, it is checked whether there is any welding line outside the robot operation area (step S431). If the welding line protrudes from the robot operation area, the work approximate rectangle is parallel-shifted so as to enter the operation area.

The recalculation of the platen area in step S432 is performed as follows.
As shown in FIG. 36, there are eight types of squares A to H that can be considered in the L-shaped surface plate left after one work is set. A to H are sorted in ascending order of RL, and the order is determined.
The above steps S428 to S433 are repeated n times for the number of selected works.

(4-3) Area division (step S43)
In order to arrange a plurality of welding robots capable of performing the same work in a finite work area and complete the work in the shortest time, it is desirable to level the work load on each welding robot. However, when the individual robots are arranged in a suspended manner as shown in FIG. 9, a restriction condition occurs in the work area of the robot. It is divided into two areas: a work area (unique area) assigned to each limited robot, and an area (lap area) that can be assigned to adjacent robots.
FIG. 37 shows a flowchart of the process of dividing the work area of the robot.

First, the average load of each ideal robot in the work area is calculated by the following equation (step S434).
Ave. = (Total work volume) / (number of robots)
Next, as shown in Table 3, the work amount of the unique region and the work amount of the lap region are obtained (steps S435 and S436).

Next, the amount of work in the lap area is divided as shown in FIG. 38, and each robot is assigned (step S437).
In FIG. 38, a = Ave. −FL1
b = FL1-a
a2 = Ave. -FL2 '(however, FL2' = FL2 + b)
b2 = LL2-a2
It is.

The work load of the robot after performing the processing of step S437 is Ave. If there is more than two, there are two adjacent robots. However, the robot No. which is the start of the process 1 and 1 if the last robot is at its peak. If there are three or more peak robots, the work amount dividing process in step S439 is not performed.
Therefore, in step S438, it is determined how many peak robots are present after the processing in step S437, and only when the number is two, the division processing of the workload is performed as described below (step S439). The example of FIG. 39 shows a case where there are two peak robots.

Ave. Is obtained from the following equation.
Pv = Σ (FLn-ave.)
Next, the average value Pave is obtained from the following equation.
Pave = Pv / 3
Attention is paid to the peak robot having the largest amount of work in the unique region (the robot of No. 3 corresponds to this in the example of FIG. 39). Here, the difference of the movable lap area is obtained from Pave, and the work amount is sequentially moved as shown in FIG. Further, in step S440, the number of peak robots after the work amount division processing is determined, and the above processing is repeated only when the number is two.
As described above, the workload of the robot is leveled.

(4-4) Generation of welding line (step S44)
When the welding line is divided by a region dividing line representing the operation region of the welding robot, the division point is generated as a welding point. Therefore, one basic welding line is constituted by a plurality of welding lines.

FIG. 41 shows an example of the workpiece 100 divided by the region dividing lines 311 to 314 of the welding robot. Here, one work 100 is divided into eight regions. In the drawing, reference numeral 120 denotes a welding line of each mounting member.

(4-5) Selection of twin welding line (Step S45)
After the weld lines are generated as described above, twin weld lines are selected only if the following criteria are met.
1) The length of the welding line on both sides of the mounting member must be the same.
2) The welding design conditions must be the same.
3) The two welding robots do not interfere with each other.
For welding lines that do not satisfy the above three conditions, a single welding line is selected.

(4-6) Determination of welding direction (step S46)
The welding direction is determined for each welding line in the generated divided region. The rules for determining the welding direction are, for example, as follows.

Rule R1: The welding direction is set so that the end that can be detected by the arc sensor is the welding end point (end).
Since the allowable range of the end detection sensor (arc sensor) has a restriction that the basic torch angle β is within ± 45 °, the side where the basic torch angle β at the end of the welding line falls within this range is defined as the welding end point. . For example, when another mounting member 113 is connected to one side of the mounting member 110 as shown in FIG. 42 (a), the torch angle β is set so as to avoid interference between the welding torch 1 and the work. The member 113 is greatly inclined with respect to the welding line (β> 45 °). On the other hand, the opposite end has no interference, so that the torch angle β is within the allowable range of the end detection sensor. Therefore, with respect to the mounting member 110, the end on the mounting member 113 side is the welding start point WS, and the end on the side without interference is the welding end point WE, and the welding direction is determined in the direction shown by the arrow. .
On the other hand, as shown in FIG. 42 (b), when the attachment member 110 is sufficiently separated from the attachment member 113, either may be set as the welding start point, but in such a case, a rule R4 described later is used. Is used to set the welding direction in a counterclockwise direction with respect to the mounting member 110. In this case, turning welding may be performed at the end of the mounting member 110. When turning welding is performed, the welding start point and the welding end point match.

Further, as another example, when there are a plurality of welding lines on one side of the mounting member, for example, as shown in FIG. 43, when the mounting member 110 is connected to another mounting member 114, 115, Three welding lines 121, 122, and 123 (for the sake of convenience in the following description, the symbols that represent the welding line and the welding direction at the same time) are present on one side of the mounting member 110. In such a case, the welding directions of the welding line 121 and the welding line 123 are divided into a clockwise direction and a counterclockwise direction according to the above rule R1. On the other hand, since the welding line 122 is out of the allowable range of the terminal detection sensor on both sides, the welding direction is not determined. Also in such a case, the welding direction is set in the counterclockwise direction with respect to the mounting member 110 using the rule R4 described later.

Rule R2: When one or both of the welding lines form a bead joint, the welding direction is from the side closer to the robot system coordinate origin to the side farther from the origin.
This is, for example, when one welding line is divided due to interference or the like even in the same welding robot operation area as shown in FIG. 44, and the welding lines 121 to 123 can be connected by the bead joining process. In such a case, it is for stabilizing the welding quality by defining the welding direction as one so that the overlapping manner of welding at the bead joints 141 and 142 is uniform.
In the example of FIG. 44, the scalloping portion 118 is provided on each of the other attaching members 116 and 117 connected to the attaching member 110, so that the beads can be joined through the scalloped portion 118.

Rule R3: When one or both of the welding lines are divided by an area dividing line that defines the operating area of the welding robot, the welding direction of the welding line is from the side closer to the robot system coordinate origin to the side farther away. And
This is because, as shown in FIG. 45, one attachment member 110 extends over a plurality of operation areas of the welding robot, and the welding lines 121 to 123 and welding lines 124 to 126 on both sides are divided by the area dividing lines 311, 312. In such a case, by defining the welding direction in one direction, the welding quality of the bead joint at the region division points 151 and 152 is stabilized.

Rule R4: If the welding direction is not uniquely determined by the rules of R1 and R2, the welding direction of the welding line is assumed to be a counterclockwise direction.
The above rules are applied in the order of R3 → R2 → R1 → R4.

(4-7) Determination of welding order (step S47)
The principle rule Ra for determining the welding order is as follows.
First, a surface including a welding line whose welding start point is closest to the robot system coordinate origin in the mounting member is determined. That is, it is determined which of the front and back sides of the mounting member is to be welded first. In order to determine the surface including the welding line to be welded first, it is determined whether or not the normal vector of the surface is directed to the robot system coordinate origin. For example, in FIG. 43, when the robot system coordinates OXY are taken as shown, the normal vector Va perpendicular to the surface 110a of the mounting member 110 is directed toward the robot system coordinate origin O (X of the normal vector Va). Since the sum of the component Vax and the Y component Vay becomes a minus sign, it is directed toward the origin O). On the other hand, the normal vector Vb perpendicular to the back surface 110b of the mounting member 110 is directed in the direction opposite to the robot system coordinate origin O (the sum of the X component Vbx and the Y component Vby of the normal vector Vb is a + sign. Therefore, the direction is opposite to the origin O). Therefore, the welding line included in the surface 110a is firstly welded to the mounting member 110.
Next, welding is started with the welding line having the welding start point closest to the robot system coordinate origin O as the first place among the welding line group included in the surface. In the example of FIG. 43, among the welding line groups 121 to 123 included in the surface 110a, the welding line whose welding start point WS is closest to the robot system coordinate origin O is the welding line 121, so that the welding line 121 is first. Welded. Therefore, the welding order is determined as the order of the welding line 121 on the front surface 110a → the welding line 122 → the welding line 123 → the welding line 124 on the back surface 110b.

Also, the following rules apply.
Rule Rb: Weld the side with the groove first.
This rule is to minimize the occurrence of defects such as blowholes caused by the primer when the groove is coated with the primer. For example, in the case of the attachment member as shown in FIG. 46, the welding line on the groove 112 side (left side) is welded first.

(4-8) Determination of welding route (step S48)
Based on the welding direction and welding order for each welding line determined above, the welding path is determined so that the interference between the welding robots is minimized and the air cut time is minimized.
In the case of a multi-robot, minimizing interference between welding robots is an important factor for improving the arc time rate. As shown in FIG. 47, the specific processing includes an interlock area setting means 481 for avoiding interference, a welding line selecting means 482 for selecting a welding line included in the set interlock area, and The welding path determination means 483 determines a welding path for each welding line.
As a method of avoiding interference between the welding robots, the welding robots perform welding while taking an interlock (synchronization) for each of the above-mentioned regions, thereby minimizing the probability that two welding robots will simultaneously enter adjacent regions. I keep it low.

1) Interlock area setting means The setting of the interlock area for avoiding interference depends on the number of welding lines included in the operation area of the welding robot. In the interlock area setting means 481, of the two division methods as shown in FIG. 48, the welding line included in the R area (the right or upper interlock area in the operation area of the welding robot) is the most. Choose a splitting method that will increase. In the method shown in (a), focusing on the interference between the left and right welding robots, the operating area of each welding robot is divided into three in the vertical direction, and the hatched portions on each right are interlocked areas 321, 322, and 323. It is set as. On the other hand, the method shown in (b) focuses on the interference between the upper and lower welding robots, divides the operating area of each welding robot into three parts in the horizontal direction, and places an interlock area in the upper hatched part in the lower area. 324 is set. The division ratio of the operation area of the welding robot is usually equal, but may be different depending on the work.
In the example of FIG. 48, since the welding line length included in the method (a) is longer than the welding line length included in the method (b), the method (a) having a high probability of occurrence of an interference problem is employed. You.

2) Welding line selection means As shown in FIG. 49, the welding line selection means 482 provides three regions in the operation region of the welding robot, and classifies the welding lines for each region. Specifically, a welding line group (LCR welding line group) extending over three regions of L, C, and R; a welding line group (LC welding line group) included in one or both of the two regions of L and C; It is classified into a welding line group (CR welding line group) included in one or both of the two regions C and R.

3) Welding path determination means The welding path is determined in the order of the LCR welding line group, the LC welding line group, and the CR welding line group. At the time of actual welding, the interlock is taken for each area by the welding robot group control system. The welding path for each region is selected so that the air cut is minimized. Specifically, start with the mounting member whose welding start point is closest to the robot system coordinate origin, and select another mounting member that has a welding start point closest to the welding end point of the last welding line of that mounting member. repeat.

FIG. 50 shows the relationship between the welding path and the interlock. To explain this drawing, first, in the welding of the LCR welding line group in step S481, it is checked whether or not the LCR welding line group exists in the operation area of the welding robot. If the LCR welding line group exists and the LCR welding line group also exists in the left welding robot adjacent to the left side of the welding robot, the welding is waited until the left welding robot completes welding (welding is not started). If the left welding robot is waiting for an interlock, no interference occurs between the left and right welding robots, so that welding of the welding robot is started. Next, when the LCR welding line group does not exist in the welding robot and the LC welding line group exists, if the welding of the LCR welding line group of the left welding robot is completed, the LC welding line group of the welding robot is completed. Is started (step S482). Finally, if neither the LCR welding line group nor the LC welding line group exists in the welding robot and the CR welding line group exists, the welding robot is completed when the welding of the LC welding line group of the right welding robot is completed. Of the group of CR welding lines is started (step S483).
Then, in the welding of each welding line group, welding is started from a mounting member having a welding start point closest to the robot system coordinate origin, and from the welding end point of the last welding line of the mounting member to the welding end point most. Move to another mounting member with a closer welding start point. This minimizes air cut.

(5) Generation of operation program (step S5)
As shown in FIG. 6, the operation program generation processing includes selecting an operation pattern (step S51), checking for interference based on the operation pattern (step S52), generating a welding line for avoiding interference (step S53), and developing the operation pattern. (Step S54), selection of welding conditions (Step S55), and generation of an operation program (Step S56).

(5-1) Selection of operation pattern (step S51)
Based on the member shape characteristics at each welding point of the welding line generated for each mounting member, an operation pattern that seems to be optimal is sequentially selected and assigned. Here, the member shape characteristic is the upper part that specifies the welding operation at each welding point, such as the presence or absence of a side member at the end of the welding member.
Basic operation patterns are: 1) approach, 2) start end detection (touch sensing), 3) start end / middle / end end welding, 4) end detection (arc sensor), 5) retract, 6) wire cut, 7) Nozzle cleaning, 8) Air cut.
Then, a management table of these basic operation patterns is provided, and an optimum execution pattern is selected and allocated.

(5-2) Interference check by operation pattern (step S52)
It is checked whether or not the welding robot does not interfere with its own member and adjacent members at a specific position in the operation sequence allocated as described above. At the time of this interference check, a check is made with the basic robot posture having the torch angles α and β determined as described above.

(5-3) Generation of welding line for avoiding interference (step S53)
In the interference check in step S52, the welding line that has caused interference with its own member and the adjacent member is fixedly processed to shorten the welding line to a position where no interference occurs in the operation pattern, thereby avoiding the interference. To avoid interference, the assigned basic operation pattern (robot posture) is not changed in principle. This is because if the basic operation pattern is changed, it becomes difficult to secure the welding quality.

(5-4) Development of Operation Pattern (Step S54)
The assigned operation pattern is developed according to the welding path determined in step S48. As the processing, the position of each operation sequence point is calculated and obtained from the member position data.

(5-5) Selection of welding conditions (step S55)
Welding conditions are assigned to each welding line on the welding path. For example, a welding condition data code in the case of horizontal fillet welding is described as shown in FIG. The codes of the respective elements are set in sub-tables as shown in FIGS. 3B to 3G, and the optimum code corresponding to the property of the welding line is set at the position of the welding condition code in the welding condition data bank. The welding conditions are stored.

(5-6) Generation of operation program (step S56)
As described above, the welding line data for the workpiece 100 is created, and this is edited as an operation program of the welding robot.

(6) Operation simulation (Step S6)
Simulate the operation of the welding robot with the above operation program, and finally check whether there is any problem. After that, this operation program is transmitted to each welding robot, and provided for actual welding.

FIG. 52 shows a specific example of the welding path. The origin O of the robot system coordinates is set at the lower left corner of the surface plate 200, and the L, C, and R areas are set according to the method shown in FIG. The arrangement relationship of each welding robot with respect to the workpiece 100 is NO. No. 1 to No. Shown at 8. 53 to 56 show NO. 1 to NO. The weld line in the area of No. 4 is shown in a separate view. The welding line of each mounting member is shown by adding a numeral to the lowercase alphabet. In the figure, reference numerals 151 to 168 denote area division points on the left side of the workpiece 100.

First, the welding robot NO. In region 2, since the LCR welding line group includes the divided welding line of the J member, the welding path is in the order of welding lines j1, j2, j3, and j4 in a counterclockwise direction. Next, since the LC welding line group includes the divided welding lines of the C member, the welding paths are in the order of c1, c2, and c3. Finally, since the CR welding line group includes the division welding line of the B member and the welding lines of the G and H members, the welding path is defined by the welding lines b1 and b2 counterclockwise from the B member near the robot system coordinate origin O. , B3, and b4, and then move to the G member based on the rule of the minimum air cut, and then the welding lines g1, g2 counterclockwise, and then the welding lines h1, h2 of the H member.

Next, the welding robot NO. In region 1, since the LCR welding line group includes the divided welding line of the I member, the welding path is in the order of i1, i2, i3, and i4 in the counterclockwise direction, and then the C path included in the LC welding line group is included. The division welding lines c4, c5, and c6 of the members are in this order. Lastly, among the A, E, and F members included in the CR welding line group, the welding lines e1, e2, and then the A member are moved from the E member near the robot system coordinate origin O to the divided welding lines a1, a2, a3. , A4, and finally, the welding lines f1 and f2 of the F member.

In the area of the welding robot N0.4, since the LCR welding line group includes the divided welding line of the B member, the welding paths are in the order of b5, b6, b7, and b8, and then J, Since the divided welding lines of the D member and the welding lines of the L member are included, first, the welding members j5 and j6 are moved from the J member near the robot system coordinate origin O to the D member, and the welding lines d1, d2, d3, d4, and then the welding lines l1 and l2 of the L member. Finally, since the M and P members are included in the CR welding line group, the division welding lines m1, m2, m3, and m4 of the M member and the welding line p1 of the P member are in this order. The back side of the P member is not welded. Further, twin welding lines r1, r2, s1, and s2 are selected for the R and S members, respectively. The twin welding lines r1 and r2 of the R member are the welding robot NO. 4 and NO. 3 perform welding in synchronization with each other, and the twin welding lines s1 and s2 of the S member are set to the welding robot NO. 4 and NO. 6 performs welding in synchronization.

Welding robot NO. In region 3, since the LCR welding line group includes the division welding line of the A member, the welding path is in the order of a1, a2, a3, and a4, and then the LC welding line group is divided into the I and D members. Since the welding line and the welding line of the K member are included, first, the divided welding lines i5 and i6 of the I member, and then to the D member, and the divided welding lines d5, d6, d7 and d8, and the welding lines k1 and k2 of the K member. It becomes order of. Finally, since the CR welding line group includes the M member split welding line and the N member welding line, the M member split welding lines m5, m6, m7, and m8 are in order, and then the N member welding line n1 is in the order. It becomes. The back side of the N member is not welded similarly to the P member. Further, twin welding lines q1, q2, r1, and r2 are selected for the Q and R members in the same manner as described above.
Hereinafter, the welding path of the welding line is similarly determined for the right side portion of the workpiece 100 as shown in FIG.

ち After the welding of the panel base material and each mounting member is completed as described above, a vertical welding line is welded to join the mounting members. The vertical welding lines are sequentially welded in ascending welding order from the welding line close to the robot system coordinate origin in the operation area of each welding robot.

It is a flowchart which shows the production | generation procedure of the welding robot operation | movement program of this invention. 3 is a flowchart in the work model generation processing of FIG. 1. 2 is a flowchart in a welding model generation process of FIG. 1. 4 is a flowchart of an interference check process in the welding line generation process of FIG. 3 is a flowchart in a cell model generation process of FIG. 1. 3 is a flowchart in the operation program generation processing of FIG. It is a top view showing an example of a work. FIG. 8 is a side view of FIG. 7. It is a schematic diagram of a robot mechanism composed of a plurality of welding robots. FIG. 4 is an explanatory diagram of a method of defining each data of work shape information. It is explanatory drawing of the method of defining the scanling data of various attachment members. FIG. 6 is an explanatory diagram of a method for defining MAX scanning data. It is explanatory drawing which shows the inclination angle of the up-down direction of the welding line, and the inclination angle of the left-right direction. It is explanatory drawing of the method of defining groove shape data. It is explanatory drawing of the extension process of a member shape. It is explanatory drawing of the mirror process of a member shape. It is a flowchart of a reverse distortion addition processing screen. FIG. 4 is an explanatory diagram of a method for giving an inverse distortion amount. FIG. 4 is an explanatory diagram of a method for giving an inverse distortion amount. FIG. 4 is an explanatory diagram of a method for determining a torch angle αmax by checking interference with a self-member. FIG. 3 is a diagram illustrating an outline of a robot arm. It is explanatory drawing of basic robot attitude | position 1)-3). It is explanatory drawing of the area | region for extraction of an interference candidate member. It is explanatory drawing of an interference check point. FIG. 9 is an explanatory diagram of a method for determining a torch angle αmin by checking interference with a back member. It is explanatory drawing of the interference check method with respect to a side member. It is a figure which shows tilting a welding torch in parallel with a side surface member in the interference check with a side surface member. FIG. 3 is an explanatory diagram of α, β, and γ axes of a welding torch and a robot arm. FIG. 26 is a view as viewed from the direction of arrow a in FIG. 25. FIG. 26 is a view taken in the direction of the arrow b in FIG. 25. It is explanatory drawing which shows the method of repeatedly checking between the end parts P0-P1 of an own member in the interference check with a side member. It is a flowchart of an automatic work arrangement process. FIG. 9 is an explanatory diagram of a method of generating a work approximate rectangle in the automatic work placement processing. It is explanatory drawing of the element square of a surface plate in automatic work arrangement processing. It is explanatory drawing of the arrangement | positioning method of the element square of a surface plate, and a work approximation square. It is explanatory drawing which shows the arrangement | positioning method to the remaining part of the surface plate after arrange | positioning one workpiece | work. It is a flowchart of the division | segmentation process of the work area of a welding robot. It is explanatory drawing of the division | segmentation method of the work amount of a welding robot. FIG. 4 is an explanatory diagram of a method for leveling the work amount of a welding robot. FIG. 4 is an explanatory diagram of a method for leveling the work amount of a welding robot. It is a top view showing the welding line of each attachment member on a work. It is explanatory drawing which shows the starting end and the terminal end of a welding direction. It is explanatory drawing which shows the welding direction and order of an attachment member. It is explanatory drawing which shows the welding direction and order when connection of a welding line is possible. It is explanatory drawing which shows the welding direction and order when a welding line is divided. It is explanatory drawing of the mounting member with a groove. It is a block diagram of the determination means of a welding path. FIG. 3 is an explanatory diagram showing a method of dividing an operation area of a welding robot. FIG. 48 is a layout diagram of L, C, and R regions according to the division method of FIG. 47. It is a flowchart which determines a welding path | route by the welding line in L, C, and R area | region and the interlock of a welding robot. It is explanatory drawing of the welding condition data name of a horizontal fillet welding line, and each optimal welding condition table. It is explanatory drawing of the welding route of a workpiece | work. In FIG. It is a division diagram of two areas. In FIG. It is a division view of one area. In FIG. It is a division diagram of four areas. In FIG. It is a division diagram of three areas.

Explanation of reference numerals

DESCRIPTION OF SYMBOLS 1 Welding torch 100 Work 101 Panel base material 110 Mounting member 120 Basic welding line 200 Surface plate 300 Welding robot 305 Robot arm 311 to 314 Area division line

Claims (5)

In a case where a workpiece having a plurality of mounting members is welded by a plurality of welding robots, a system which reads data registered in a CAD / CAM system and automatically generates an operation program for each of the welding robots,
Means for generating a work model corresponding to the work from the work shape data of the read data,
Means for generating a welding model having a basic welding line consisting of attachment lines of the constituent members of the work,
Means for generating a cell model by dividing the work model by a region dividing line so as to equalize the workload of the plurality of welding robots,
Means for checking whether or not the workpiece and the welding robot interfere with each other for each of the basic welding lines, and a means for generating a welding line in a range where interference does not occur;
Means for determining a welding direction, a welding order, and a welding path for each of the welding lines after the interference check included in the region for each region divided by the region dividing line,
Means for generating a series of operation programs by allocating data necessary for welding by selecting an operation sequence group called an operation pattern previously determined from the data for the determined welding line,
An automatic generation system for a welding robot operation program, comprising: The interference checking means fixes the basic robot posture to a plurality of types, and performs an interference check on the interference candidate members extracted from the mounting member and its peripheral members according to the basic robot posture to determine a torch angle. The system for automatically generating a welding robot operation program according to claim 1, further comprising means. The interference checking means sets an extraction area of the interference candidate member on the back side and the side surface of the mounting member, and performs interference by using cross-sectional shape data of the interference candidate member applied to the mounting member and the extraction area. 3. The automatic generation system of a welding robot operation program according to claim 2, wherein the system is configured to check. 2. The system according to claim 1, further comprising means for selecting a twin welding line when the generated welding line satisfies a predetermined condition. The automatic generation of a welding robot operation program according to claim 1, wherein, when the welding direction with respect to the mounting member is determined by the welding direction determining means, the end of the welding line is set so that the terminal position can be detected by an arc sensor. system.

Images