Setting method and setting apparatus for operation path for articulated robot

This application is the national phase under 35 U.S.C. § 371 of PCT International Application No. PCT/JP01/10202 which has an International filing date of Nov. 22, 2001, which designated the United States of America.

1. Technical Field

The present invention relates to a setting method and a setting apparatus for an operation path for an articulated robot. Specifically, the present invention relates to a setting method and a setting apparatus for an operation path for an articulated robot, for setting the path for operating an end effector provided at a forward end of the articulated robot, between predetermined operation points.

2. Background Art

Conventionally, if an articulated robot installed for a production line is directly operated to perform the teaching of the operation posture, an operator skilled in the operation of the articulated robot should perform the operation at the working site of the production line. Accordingly, the operation becomes inefficient. The above operation should also be performed with the production line being stopped. Therefore, the operation rate of the production line is decreased.

Recently, the teaching (off-line teaching) is performed based upon an off-line procedure to efficiently perform the teaching operation or to maintain the operation rate of the production line. In the off-line teaching, a model, which includes an articulated robot, a workpiece as an operation objective, and peripheral structures, is constructed on a computer. Teaching data is prepared by using the model, and then the teaching data is supplied to the articulated robot installed at the working site. Therefore, it is unnecessary to stop the production line during the preparation of the teaching data.

The conventional off-line teaching is not necessarily used widely for the following reason.

Naturally, the articulated robot should not interfere with (for example, contact) various peripheral structures, workpieces or the like. When various peripheral structures exist or when the workpiece is of a complicated shape, it is difficult to set an operation path to avoid such obstacles.

More specifically, the round-robin method, in which the interference is investigated as to all postures of the articulated robot, is not practical, because the amount of calculation is enormous. No solution exists in some cases in the optimizing method such as the so-called mathematical programming. Further, according to the stochastic technique using random numbers, the convergence of solution is not assured and the calculation has no reproducibility.

Several techniques have been suggested to solve the above problems.

For example, a technique is known, which utilizes a flat plane including a start point and an arrival point (see Japanese Patent Publication No. 2875498). In this technique, an off-limit area, in which a cross section of an obstacle is appropriately enlarged, is defined on a prescribed plane. An operation path, which passes through the apex of the off-limit area, is set to avoid the interference. However, in this technique, the operation path is set by verifying the interference with the off-limit area at every time. For this reason, the verifying operation is complex, and the operation path is complicated. Even if the operation path is proper, it is also impossible to verify whether the articulated robot can actually operate on the operation path from a viewpoint of operation ranges of respective axes.

Another technique is also known, for example, in which the position and the shape of an obstacle are inputted and instructed with an exclusively used controller in a production site to set an operation path (see Japanese Laid-Open Patent Publication No. 9-81228). However, in this technique, the operation path cannot be set automatically, because the teaching is performed while operating the actual machine at the production site.

Accordingly, the above off-line teaching relies on the manual operation to set the operation path for avoiding the obstacles at present.

However, the manual operation needs a long period of time to extract a non-interference area in which the robot does not interfere with the workpiece and other equipments. The judgment also differs depending on individual persons. It is inevitable to cause any oversight and/or any omission for the extraction point.

As described above, when the posture of the robot is determined by means of the off-line teaching, the operation required therefor is not necessarily easy. Especially, it is difficult to retrieve a path for retracting a gun unit from a welding point so that it may not interfere with a workpiece, on a monitor screen, when the workpiece is of a complicated three-dimensional shape. It takes a long period of time to perform the teaching.

In consideration of the above problems, it is an object of the present invention to provide a setting method and a setting apparatus for an operation path for an articulated robot, in which steps for determining the path are automatically performed, and teaching data can be prepared in a short period of time without requiring any skill, when off-line teaching is performed for a withdrawing path to make no interference with a workpiece, in a narrow-area operation path for withdrawing an end effector from an operation point on the workpiece, of operation paths for an articulated robot.

Another object of the present invention is to provide a setting method and a setting apparatus for an operation path for an articulated robot, in which a wide-area operation path for making movement between operation points or between withdrawing positions can be set automatically and efficiently without performing any complicated calculation which may be affected by the shape of a workpiece and/or an obstacle.

Still another object of the present invention is to provide a setting method and a setting apparatus for an operation path for an articulated robot, in which a narrow-area operation path and a wide-area operation path can be set automatically and efficiently.

According to the present invention, there is provided a method for setting an operation path for an articulated robot including an end effector, the method comprising an internal space-defining step of defining an internal space which is partially surrounded by an arm or electrodes of the end effector; an extracting step of extracting an objective workpiece portion which exists in the internal space, of a workpiece to be welded; and an interference-investigating step of investigating whether interference occurs between the end effector and the objective workpiece portion when the articulated robot is operated.

Accordingly, the steps for determining the path are automatically performed, and teaching data can be prepared in a short period of time without requiring any skill, when off-line teaching is performed for a withdrawing path to make no interference with a workpiece, in a narrow-area operation path for withdrawing an end effector from an operation point on the workpiece.

In this case, the articulated robot, the end effector, the workpiece, and peripheral structures are virtual ones constructed as a model in accordance with a program processing effected by a computer.

The workpiece may be a model which is approximated with a plurality of blocks.

The internal space may be a model which is approximated with a plurality of blocks.

Further, the interference-investigating step may comprise a reference line-defining step of defining a reference line passing through a substantially central portion of the objective workpiece portion; an investigation end position-defining step of setting an investigation end position for the end effector on the reference line; and a first detailed interference-investigating step of investigating whether interference occurs between the end effector and the objective workpiece portion by operating the end effector from an investigation start position to the investigation end position.

The interference-investigating step may comprise a reference line-defining step of defining a reference line passing through a substantially central portion of the objective workpiece portion; a center of gravity position-defining step of defining a center of gravity position of the objective workpiece portion based upon the reference line; and a second detailed interference-investigating step of investigating whether interference occurs between the end effector and the objective workpiece portion by operating the end effector from an investigation start position to the center of gravity position.

A portion of the objective workpiece portion, which is located closely to an opening as compared with the center of gravity position of the objective workpiece portion, may be extracted as a new objective workpiece portion with which the objective workpiece portion is replaced to perform the center of gravity position-defining step and the second detailed interference-investigating step.

According to another aspect of the present invention, there is provided an apparatus for setting an operation path for an articulated robot provided with an end effector, the apparatus comprising an internal space-defining section for defining an internal space which is partially surrounded by an arm or electrodes of the end effector; a workpiece-extracting section for extracting an objective workpiece portion which exists in the internal space, of a workpiece to be welded; and an interference-investigating section for investigating whether interference occurs between the end effector and the objective workpiece portion when the end effector is operated.

According to still another aspect of the present invention, there is provided a method for setting an operation path for an articulated robot for operating an end effector from a start point to an arrival point, the method comprising an operation-investigating step of setting a path for connecting the start point and the arrival point to investigate whether the end effector can be operated along the path; and a retracting path-setting step of setting a retracting path for operating the end effector by a prescribed distance in a prescribed direction from the start point or the arrival point if the end effector cannot be operated along the path in the operation-investigating step.

Accordingly, a wide-area operation path for making movement between operation points or between withdrawing positions can be set automatically and efficiently without performing any complicated calculation which may be affected by the shape of a workpiece or an obstacle.

The prescribed direction may be a predetermined direction based on a posture of the end effector at the start point or the arrival point.

The prescribed direction may be a direction to connect the start point or the arrival point and an established point in space.

The established point may be a central point of an original axis of the articulated robot.

An end point of the retracting path may be defined as a new start point or a new arrival point to execute the operation-investigating step or the retracting path-setting step again.

The retracting path, in which the prescribed distance is corrected, may be set again if an end point of the retracting path is a point at which the articulated robot cannot arrive or a point at which interference occurs.

According to still another aspect of the present invention, there is provided an apparatus for setting an operation path for an articulated robot for operating an end effector from a start point to an arrival point, the apparatus comprising a path-investigating section for setting a path for connecting the start point and the arrival point to investigate whether the end effector can be operated along the path; and a wide-area operation path-setting section for setting a retracting path for operating the end effector by a prescribed distance in a prescribed direction from the start point or the arrival point if the path-investigating section judges that the end effector cannot be operated along the path.

According to still another aspect of the present invention, there is provided a method for setting an operation path for an articulated robot for operating an end effector between operation points for a workpiece, the method comprising a narrow-area operation path-setting step of setting a narrow-area operation path along which the end effector arranged at the operation point for the workpiece is retracted from the operation point to a point located near an end of the workpiece while maintaining a non-interference state with respect to the workpiece and another obstacle, based upon shapes of the obstacle and the workpiece near the operation point; and a wide-area operation path-setting step of setting a wide-area operation path for effecting operation from a start point to an arrival point by combining predetermined prescribed operations provided that the start point and the arrival point reside in predetermined points of points located near the end.

Accordingly, it is possible to set the narrow-area operation path and the wide-area operation path automatically and efficiently.

The narrow-area operation path-setting step may comprise an internal space-defining step of defining an internal space which is partially surrounded by an arm or electrodes of the end effector; an extracting step of extracting an objective workpiece portion which exists in the internal space, of the workpiece; and an interference-investigating step of investigating whether interference occurs between the end effector and the objective workpiece portion when the articulated robot is operated.

The wide-area operation path-setting step may comprise an operation-investigating step of setting a path for connecting the start point and the arrival point to investigate whether the end effector can be operated along the path; and a retracting path-setting step of setting a retracting path for operating the end effector by a prescribed distance in a prescribed direction from the start point or the arrival point if the end effector cannot be operated along the path in the operation-investigating step.

The prescribed direction may be a predetermined direction based on a posture of the end effector at the start point or the arrival point.

The prescribed direction may be a direction to connect the start point or the arrival point and an established point in space.

According to still another aspect of the present invention, there is provided an apparatus for setting an operation path for an articulated robot for operating an end effector between operation points for a workpiece, the apparatus comprising a narrow-area operation path-setting section for setting a narrow-area operation path along which the end effector arranged at the operation point for the workpiece is retracted from the operation point to a point located near an end of the workpiece while maintaining a non-interference state with respect to the workpiece and another obstacle, based upon shapes of the obstacle and the workpiece near the operation point; and a wide-area operation path-setting section for setting a wide-area operation path for effecting operation from a start point to an arrival point by combining predetermined prescribed operations provided that the start point and the arrival point reside in predetermined points of points located near the end.

The above and other objects, features, and advantages of the present invention will become more apparent from the following description when taken in conjunction with the accompanying drawings in which a preferred embodiment of the present invention is shown by way of illustrative example.

FIG. 1 illustrates an off-line teaching apparatus and a robot apparatus to be used in an embodiment of the present invention;

FIG. 2 shows a block diagram illustrating an arrangement of the off-line teaching apparatus;

FIG. 3 illustrates an arrangement of an articulated robot;

FIG. 4 illustrates an X-type welding gun;

FIG. 5 illustrates welding points for a workpiece;

FIG. 6 shows a flow chart illustrating an entire operation path-setting method for the articulated robot according to the embodiment of the present invention;

FIG. 7 shows a flow chart (No. 1) illustrating a narrow-area operation path-setting method for the articulated robot according to the embodiment of the present invention;

FIG. 8 shows a flow chart (No. 2) illustrating the narrow-area operation path-setting method for the articulated robot according to the embodiment of the present invention;

FIG. 9 shows a flow chart (No. 3) illustrating the narrow-area operation path-setting method for the articulated robot according to the embodiment of the present invention;

FIG. 10 shows a path table;

FIG. 11A illustrates a procedure for setting lines radially from the central point;

FIG. 11B illustrates a procedure for extracting points of intersection in a closed space by drawing lines in a lattice-shaped configuration;

FIG. 12A illustrates a procedure for setting solids about centers of points of intersection;

FIG. 12B illustrates a procedure for extracting overlapped portions of the solids and the workpiece;

FIG. 12C shows an extracted workpiece model;

FIG. 13A illustrates central points of the respective solids;

FIG. 13B shows a procedure for determining a principal component line;

FIG. 14 shows a procedure for determining a withdrawing point and a withdrawing path (V 1 );

FIG. 15 shows a withdrawing path (V 2 );

FIG. 16 illustrates a mask process;

FIG. 17 illustrates an operation path from a start point to an arrival point;

FIG. 18 shows a flow chart (No. 1) illustrating a wide-area operation path-setting method for the articulated robot according to the embodiment of the present invention;

FIG. 19 shows a flow chart (No. 2) illustrating the wide-area operation path-setting method for the articulated robot according to the embodiment of the present invention;

FIG. 20 shows a flow chart (No. 3) illustrating the wide-area operation path-setting method for the articulated robot according to the embodiment of the present invention; and

FIG. 21 illustrates the operation of first and second templates.

Illustrative embodiments of the setting method and the setting apparatus for the operation path for the articulated robot according to the present invention will be explained below with reference to FIGS. 1 to 21 .

Basically, in the setting method and the setting apparatus for the operation path for the articulated robot according to the embodiment of the present invention, the operation path is set while extracting the portion in which the workpiece exists to investigate the interference in the internal space surrounded by the gun unit during the narrow-area operation in which the end effector provided at the forward end of the articulated robot is withdrawn from the operation point on the workpiece. During the wide-area operation for making movement between the withdrawing positions, the operation path is set to move to the arrival point while avoiding the obstacle by operating while combining the template operations for making movement from the start point by the prescribed distance in the prescribed direction.

As shown in FIG. 1 , an off-line teaching apparatus (operation path-setting apparatus) 10 , which is used in the embodiment of the present invention, performs teaching of the operation of an articulated robot 50 . The apparatus 10 is linked to a robot apparatus 12 for performing desired operation for an operation objective based upon prepared teaching data.

The robot apparatus 12 comprises the articulated robot 50 , and a robot control unit 22 for controlling the operation of the articulated robot 50 based upon the teaching data.

As shown in FIG. 2 , a control unit 14 , which constitutes the off-line teaching apparatus 10 , includes CPU (computer) 26 as a control means for controlling the entire off-line teaching apparatus 10 , ROM 28 and RAM 29 as storage sections, a hard disk drive (HDD) 39 for effecting access of data with respect to the hard disk 34 , a drawing control circuit 30 for effecting drawing control on a screen of a monitor 16 , an interface circuit 32 to which a keyboard 18 and a mouse 20 as input apparatuses are connected, a recording medium drive 36 for controlling an external recording medium 36 a (for example, a flexible disk or a compact disk), a data-preparing circuit 38 for preparing teaching data, and a simulation circuit 40 for effecting simulation on the screen of the monitor 16 based upon the teaching data. The simulation circuit 40 is based on three-dimensional CAD, and it has, for example, the function to prepare the model and investigate the mutual interference of the model (interference-investigating section 40 a ).

The hard disk 34 stores, for example, an operation path-setting program 35 having the function to set the operation path for an articulated robot 50 , condition data 37 as the condition for setting the operation path, and unillustrated OS.

The operation path-setting program 35 includes a narrow-area operation path-setting section 35 a for setting, for example, based upon the shape of a workpiece 80 , the narrow-area operation path along which a gun unit (end effector) 68 , which is arranged on a point on the workpiece 80 (see FIG. 5 ), for example, on a welding point T 0 , is retracted to a point located near the end of the workpiece 80 while maintaining the non-interference state with the workpiece 80 and other components, and a wide-area operation path-setting section 35 b for setting the wide-area operation path along which the operation is effected from a start point P 1 to an arrival point P 2 by combining predetermined prescribed operations provided that the start point P 1 (see FIG. 17 ) and the arrival point P 2 reside in arbitrary two points in the space.

The operation path-setting program 35 has a path-investigating section 35 c for investigating wither or not the gun unit 68 can be operated on the path obtained by connecting two arbitrary points.

The operation path-setting program 35 further includes an internal space-defining section 35 d for defining a predetermined internal space in the virtual space, and a workpiece-extracting section 35 e for extracting a portion of the workpiece 80 to be welded existing in a predetermined space.

As shown in FIG. 3 , a second base 56 , a first link 58 , a second link 60 , a third link 62 , a fourth link 64 , and a gun attachment section 66 are connected to a first base 54 as an attachment stand of the articulated robot 50 in this order toward the forward end. The gun unit 68 is connected to the gun attachment section 66 disposed at the forward end.

The second base 56 is rotatable supported with respect to the first base 54 about the center of the axis J 1 as the vertical axis. The proximal end of the first link 58 is supported tiltably with respect to the second base 56 with the axis J 2 as the horizontal axis. The proximal end of the second link 60 is supported swingably with respect to the forward end of the first link 58 with the axis J 3 as the horizontal axis. The third link 62 is connected on the forward end side of the second link 60 with the axis J 4 as the common central axis for rotation. Further, the proximal end of the fourth link 64 is supported swingably with respect to the forward end of the second link 62 with the axis J 5 located in the right-angled direction with respect to the axis J 4 . The gun attachment section 66 is connected on the forward end side of the fourth link 64 with the axis J 6 as the common central axis for rotation.

The gun unit 68 , which is connected to the gun attachment section 66 , is a so-called C-type welding gun, and it has, at both ends of an arch-shaped arm 74 , a pair of electrodes 70 , 72 which are openable/closable along the axis J 6 . In the closed state, the electrodes 70 , 72 make contact with the workpiece 80 at the welding operation point (hereinafter referred to as “TCP (Tool Center Point)”) for the axis J 6 .

The direction, which is directed from TCP and which is coincident with the axial center of the electrode 72 of the main body, is designated as “vector Zr”. The direction, which is perpendicular to the vector Zr and which is directed outside of the gun unit 68 , is designated as “vector Xr”. The direction, which is mutually perpendicular to the vector Xr and the vector Zr, is designated as “vector Yr”.

The driving mechanism for the axes J 1 to J 6 and the opening/closing mechanism for the electrodes 70 , 72 are driven by unillustrated actuators respectively. TCP is determined by the values of respective angles of rotation θ 1 to θ 6 of the axes J 1 to J 6 and the sizes of the respective sections of the articulated robot 50 .

The gun unit 68 is not limited to the C-type welding gun. For example, an X-type welding gun shown in FIG. 4 (welding gun provided with a pair of opening/closing gun arms rotatably supported by a common support shaft) 68 a may be used for the gun unit 68 .

The point of intersection between the axis J 1 and the axis J 2 is defined as the origin (central point of the original axis) O as the reference point for the coordinate calculation and the control in relation to the articulated robot 50 . With the reference of the origin O, the vertically upward direction is represented by the height Z, the direction of the axis J 2 obtained when the angle of rotation θ 1 satisfies θ 1 =0 is represented by the depth Y, and the direction perpendicular to the height Z and the depth Y is represented by the width X. The three-dimensional orthogonal coordinate is expressed with the height Z, the width X, and the depth Y.

Next, explanation will be made with reference to FIGS. 5 and 6 for the procedure for setting the operation path for the articulated robot 50 by using the off-line teaching apparatus 10 and the operation path-setting program 35 constructed as described above.

In the following description, an example will be explained as shown in FIG. 5 in which the gun unit 68 is successively moved between a plurality of welding points (operation points) Tn (n=0, 1, 2, . . . ) for performing the welding for the workpiece 80 which is a thin plate.

The welding point Tn is represented by six values in total including three-dimensional orthogonal coordinate values (X, Y, Z) in the space in which the welding is performed and three parameters of TCP for indicating the posture of the gun unit 68 .

Further, it has been already verified that the gun unit 68 of the articulated robot 50 is capable of arriving at the welding point Tn, and the posture of the gun unit 68 when the welding point Tn is welded, i.e., the values of the vector Xr, the vector Yr, and the vector Zr are determined as well.

According to the embodiment of the present invention, further, the articulated robot 50 , the workpiece 80 , and the peripheral structures are dealt with as virtual models in the off-line teaching apparatus 10 . However, in the following description, these components will be represented-by the same reference numerals as those of the actual apparatus.

The workpiece 80 is dealt with as the model composed of a plurality of blocks in order to obtain a high speed of the processing.

In step S 1 shown in FIG. 6 , an operator for the off-line teaching apparatus 10 starts up the operation path-setting program 35 by a predetermined operation method. OS, which is incorporated in the off-line teaching apparatus 10 , loads the operation path-setting program 35 stored on the hard disk 34 onto RAM 29 to execute the operation path-setting program 35 . The processing of the next step S 2 and the followings are executed by the operation path-setting program 35 .

Subsequently, in step S 2 , a temporary operation path 90 (see FIG. 5 ), which is obtained by connecting the welding points Tn, is set. The operation path 90 may be linear as shown in FIG. 5 , or it may be an arbitrary curve along which the articulated robot 50 is operated with ease. Operation paths 100 , 102 , 104 , 110 , 112 described later on may be set in the same manner as described above.

Subsequently, in step S 3 , it is investigated whether the articulated robot 50 is capable of setting the posture when the gun unit 68 is operated along the temporary operation path 90 . Further, it is investigated whether the gun unit 68 interferes with other structures or components in the operation path 90 .

Specifically, dividing points, which are obtained by dividing the operation path 90 into those having minute lengths, are set. The postures of the articulated robot 50 , i.e., the angles of rotation θ 1 to θ 6 , which are provided when the gun unit 68 is arranged at the respective dividing points, are determined. As for the calculation method for the angles of rotation θ 1 to θ 6 , a well-known matrix calculation method (hereinafter referred to as “inverse operation”) may be applied, for example, for the sizes of the respective sections of the articulated robot 50 and the six values in total defined by the vector Xr, the vector Yr, and the vector Zr for representing the posture of the gun unit 68 and the spatial position coordinates (X, Y, Z) of the dividing points.

When the posture of the gun unit 68 differs between the welding points T 0 and T 1 , the vector Xr, the vector Yr, and the vector Zr may be defined at the respective dividing points in a manner of linear interpolation. In this investigation, it is assumed that the electrodes 70 , 72 are opened so that they may not interfere with the workpiece 80 .

If the posture of the articulated robot 50 holds at each of the dividing points, the operation from the welding point T 0 to the welding point T 1 is actually assured.

Subsequently, in step S 4 , it is judged whether the solution of the inverse operation is normally determined at each of the dividing points. That is, it is judged whether TCP is capable of arriving at the dividing point. If the solution is not determined, if the value of the angle is without the rotatable range of the axis J 1 to J 6 even if the solution is determined, or if the articulated robot 50 interferes in the determined posture (for example, interferes with the obstacle 82 , other workpieces, and pillars in the factory), then the routine proceeds to step S 5 . If the solution is normally determined, the solution is within the rotatable range, and no interference occurs, then the routine proceeds to step S 7 .

The investigation for the interference is automatically performed by the function of the simulation circuit 40 . When the simulation circuit 40 is used, it is possible to reliably perform the three-dimensional investigation which is not clear from the screen of the monitor 16 as the two-dimensional expression.

In step S 5 , the narrow-area operation path, which is used to withdraw the gun unit 68 from the welding points T 0 and T 1 , is set by the function of the narrow-area operation path-setting section 35 a . A detailed method therefor will be described later on.

Subsequently, in step S 6 , the two withdrawing positions Ue (see FIG. 14 ), which are obtained by the narrow-area operation path, are set as the start point P 1 and the arrival point P 2 respectively to set the wide-area operation path for moving the gun unit 68 from the start point P 1 to the arrival point P 2 . The setting is performed by the function of the wide-area operation path-setting section 35 b . A detailed method therefor will be described later on.

After setting the narrow-area operation path and the wide-area operation path, the routine proceeds to step S 7 .

In step S 7 , it is confirmed whether the investigation is performed for all of the operation paths 90 set in step S 1 to complete the process. If there is any operation path 90 which is not investigated, the routine returns to step S 3 to continue the investigation.

As described above, in the embodiment of the present invention, the welding points Tn are firstly connected to one another by the operation path 90 . If the operation path 90 is not applied as it is, the narrow-area operation path for avoiding, for example, any projection of the workpiece 80 and the obstacle 82 is set. Further, the wide-area operation path is set in order to make movement between the withdrawing positions Ue obtained by setting the narrow-area operation path.

When the narrow-area operation path is set, the portion, in which the workpiece exists, is extracted to investigate the interference in the internal space which is partially surrounded by the gun unit 68 . Therefore, it is possible to automatically set the path for avoiding any interference with the workpiece.

When the wide-area operation path is set, the template operation is applied, in which the gun unit 68 is moved by a prescribed distance in a prescribed direction. Therefore, it is possible to automatically set the wide-area operation path without performing, any complicated calculation which may be affected by the shapes of the workpiece 80 and the obstacle 82 .

Further, the setting of the narrow-area operation path for withdrawing the gun unit 68 of the articulated robot 50 from the welding point Tn on the workpiece 80 and the setting of the wide-area operation path for making movement from the start point P 1 to the arrival point P 2 are performed by the different processes adapted to the respective processes. Therefore, it is possible to efficiently set the operation path between the welding points Tn.

Next, explanation will be made with reference to FIGS. 7 to 16 for the method for setting the narrow-area operation path in step S 5 shown in FIG. 6 .

When the narrow-area operation path is set, three methods are principally used in order to determine the path for withdrawing the gun unit 68 from the welding portion of the workpiece 80 .

Firstly, a method is used to directly make movement from the welding portion to the withdrawing point. Secondly, a method is used to make movement from the welding point to the center of gravity on the cross section of the workpiece 80 . Thirdly, a method is used to extract only a portion of the workpiece 80 disposed closely to the opening of the gun unit 68 so that the withdrawing path is determined by preferentially using the extracted portion.

In step S 101 shown in FIG. 7 , the gun unit 68 of the articulated robot 50 is set at the position at which the welding point T 0 of the workpiece 80 is welded.

The welding point T 0 gives the adjustment start position (Ts), and hence it is recorded on the temporary path table 12 for the operation data to perform the initialization (see Order 1 shown in FIG. 10 ).

As shown in FIG. 10 , the path table 120 comprises the column 120 a of “Direction of gun unit”, the column 120 b of “Position of TCP”, and the column 120 c of “Angle of each axis”. The column 120 c of “Angle of each axis” includes the angles of rotation θ 1 to θ 6 .

Subsequently, in step S 102 shown in FIG. 7 , TCP of the gun unit 68 located at the welding point To is set as the investigation start position Ts.

Subsequently, in step S 103 , the central point C (see FIG. 11A ) is defined at the substantial center of the gun unit 68 where the arm 74 and the electrodes 70 , 72 are overviewed. Radial straight lines 1090 are set from the central point C at predetermined angle widths to determine points of intersection 1092 on the inner circumferential side of the arm 74 and the electrodes 70 , 72 .

For the simplified explanation, the points of intersection 1092 are determined on the plane. However, actually, the points of intersection are determined in the three-dimensional shape by utilizing the data in the depth direction as well. Accordingly, the workpiece model (objective workpiece portion) 1096 described later on and the solids (or blocks) 1094 described below are dealt with as three-dimensional shapes not as planar shapes.

Subsequently, in step S 104 , as shown in FIG. 11B , the plurality of points of intersection 1092 are connected with a line segment to set an annular line 1092 b for forming a closed interval 1092 a . Lattice-shaped lines, which have predetermined spacing distances, are set in the closed interval 1092 a to extract points of intersection 1092 c existing in the closed interval 1092 a , of points of intersection of the lattice-shaped lines.

Subsequently, in step S 105 , as shown in FIG. 12A , square solids 1094 are embedded about the centers of the extracted points of intersection 1092 c so that no gap is formed to set the internal space of the gun.

The processes of steps S 103 to S 105 are executed by the function of the internal space-defining section 35 d.

Subsequently, in step S 106 , as shown in FIG. 12B , the workpiece 80 is arranged so that the workpiece 80 is matched for relative positions with respect to the gun unit 68 and the gun internal space. A portion, in which the workpiece 80 and the solids 1094 are overlapped with each other, is extracted as a workpiece model 1096 (see FIG. 12C ). Then, a portion 80 a of the workpiece 80 , which is not overlapped with the solids 1094 , is excluded, because the portion is irrelevant to the investigation of the interference. The respective solids 1094 , which constitute the workpiece model 1096 , are distinguished as workpiece solids 1098 . Even if the gun unit 68 is moved, the initial positions are fixed for the workpiece model 1096 and the respective workpiece solids 1098 .

The process in step S 106 is executed by the function of the workpiece-extracting section 35 e.

As described above, the process is easily performed, because the workpiece 80 is dealt with as the model with the plurality of blocks. Further, no useless processing is performed, because any unnecessary portion of the workpiece 80 (for example, non-overlapped portion 80 a ) is automatically excluded.

Subsequently, in step S 107 , the principal component line (or the reference line) M 1 of the workpiece model 1096 is calculated by the technique of principal component analysis.

The method for calculating the principal component line M 1 will be explained in detail. As shown in FIG. 13A , central point coordinates (Xs, Ys, Zs) of the respective workpiece solids 1098 are defined.

Subsequently, as shown in FIG. 13B , the square sum of the distance s between each of the central point coordinates 1098 a and the principal component line M 1 is made minimum. The principal component line M 1 is defined to satisfy the following expression.

Σ|s| 2 =min

Specifically, the respective central point coordinates 1098 a are used to calculate the eigen value and the eigen vector of the variance and covariance matrixes, and Xs, Ys, Zs are used to determine the position of the center of gravity G 1 as an average value of the respective coordinates of X, Y, Z. The eigen vector, which passes through the center of gravity position G 1 , is the principal component line M 1 .

In the following steps S 108 to S 112 , as shown in FIG. 14 , it is investigated whether any interference is caused when the operation is performed linearly from the investigation start point Ts to the withdrawing position Ue.

Specifically, in step S 108 , the withdrawing position Ue is determined. As shown in FIG. 14 , the withdrawing position Ue resides in the position on the principal component line M 1 . The vector Xr, which is based on TCP of the gun unit 68 , is moved while making coincidence with the principal component line M 1 . The place, at which the gun unit 68 and the electrodes 70 , 72 do not interfere, is set as the withdrawing position Ue.

Subsequently, in step S 109 , the posture of the articulated robot 50 , i.e., the angles of rotation θ 1 to θ 6 are determined based upon the position and the posture of the gun unit 68 prescribed by the withdrawing position Ue. In this calculation method, the determination may be made by the inverse operation from the six values in total prescribed by the position coordinates (X, Y, Z) in the space of the withdrawing position Ue and the vector Xr, the vector Yr, and the vector Zr for representing the posture of the gun unit 68 .

Subsequently, in step S 110 for the branching judgment, it is judged whether the solution is normally determined in the inverse operation in step S 109 . That is, it is judged whether TCP is capable of arriving at the withdrawing position Ue. If the solution is not determined, if the value of the angle is without the rotatable range of the axis J 1 to J 6 even if the solution is determined, or if the articulated robot 50 interferes with other structures in the determined posture, then the routine proceeds to step S 111 . If the solution is normally determined, the routine proceeds to step S 112 .

In the investigation for the interference, especially when the X-type welding gun 68 a is adopted for the gun unit, the investigation is made for both of the open state and the closed state of the gun unit.

If the solution is not determined normally, the rotation operation is performed in step S 111 to make rotation by α° about the center of the vector Yr. The rotation operation means the fact that the gun unit 68 is rotated about the center of the withdrawing position Ue within a range to cause no interference with the workpiece model 1096 as indicated by two-dot chain lines shown in FIG. 14 . After the vector Xr, the vector Yr, and the vector Zr are determined in this state, the routine returns to step S 109 . The investigation may be performed assuming that the angle α° has angle values in both of plus and minus directions.

If the loop, which is formed by steps S 109 to S 111 , is continuously executed predetermined number of times, the withdrawing position Ue is set again at an appropriate position which is farther on the principal component line M 1 and at which the posture of the articulated robot 50 holds. Next, the routine proceeds to the next step S 112 .

The process for making the rotation by α° is not limited to the process based on the center of the vector Yr. The process may reside in rotation about the axis, for example, the vector Xr or the vector Zr. Such a process may be adopted in the following process for rotation in the same manner as described above.

Next, the routine proceeds to the process shown in FIG. 8 . In step S 112 , as indicated by the path V 1 shown in FIG. 14 , the gun unit 68 is operated linearly from the investigation start position Ts to the withdrawing position Ue to investigate whether any interference occurs between the arm 74 and the electrodes 70 , 72 and the workpiece model 1096 .

In step S 113 for the branching judgment, if it is judged that any interference occurs according to the investigation in step S 112 , the routine proceeds to step S 114 . If it is judged that no interference occurs, the routine proceeds to step S 131 as the termination process, because the withdrawing operation can be performed by one time of the operation.

As described above, if the shape of the workpiece 80 is simple, it is possible to shorten the process time, because the withdrawing path can be determined by one time of the operation.

In the example shown in FIG. 14 , the electrode 70 clearly interferes with the projection 1096 a of the workpiece model 1096 during the movement along the path V 1 . In this case, the routine proceeds to step S 114 .

In the following steps S 114 to S 118 , it is investigated whether any interference occurs when the operation is performed linearly from the investigation start position Ts to the center of gravity position G 1 of the workpiece model 1096 .

Specifically, in step S 114 , as shown in FIG. 15 , the path V 2 , which connects the investigation start position Ts and the center of gravity position G 1 , is defined. The posture of the gun unit 68 is assumed, in which the vector Xr coincides with the path V 2 based upon the center of gravity position G 1 .

In step S 115 , the posture of the articulated robot 50 is determined with the assumed posture by the inverse operation described above.

Subsequently, in step S 116 for the branching judgment, it is investigated whether the solution in the inverse operation is normally determined in the same manner as in step S 110 . Then, in addition to the inverse operation process, it is also preferable to investigate whether the gun unit 68 interferes with the workpiece model 1096 .

If the solution is not determined normally, the rotation operation is performed to rotate by α° about the center of the vector Yr (step S 117 ) in the same manner as in step S 111 . After the vector Xr, the vector Yr, and the vector Zr are determined in this state, the routine returns to step S 115 .

If the solution is determined, the interference is investigated by linearly operating the gun unit 68 along the path V 2 from the investigation start position Ts to the center of gravity position G 1 in step S 118 in the same manner as in step S 112 .

If the loop, which is formed by steps S 115 to S 117 , is continuously executed predetermined number of times, it is judged that the gun unit 68 cannot be arranged at the center of gravity position G 1 . After this processing is finished, the routine proceeds to step S 124 as the mask process.

If it is judged that any interference occurs by the investigation performed in step S 118 described above and step S 130 described later on, the routine proceeds to step S 124 via step S 119 for the branching judgment. If it is judged that no interference occurs, the routine proceeds to the next step S 120 , assuming that the operation is successfully performed up to the center of gravity position.

In step S 120 , the posture of the articulated robot 50 at that point of time is additionally recorded on the path table 120 .

Subsequently, in step S 121 , the operation is made linearly from the position of the gun unit 68 at that point of time to the withdrawing position Ue in the same manner as in step S 112 to investigate whether interference occurs. In the example shown in FIG. 15 , the investigation is made along the principal component line M 1 .

In step S 122 for the branching judgment, if it is judged that any interference occurs by the investigation in step S 121 , the routine proceeds to step S 123 . If it is judged that no interference occurs, the routine proceeds to step S 131 as the termination process, because the withdrawing operation can be performed by this operation.

If there is any interference, the position of the gun unit 68 at that point of time is used as a new investigation start position in step S 123 to perform the updating process to make exchange for the previous investigation start position Ts. That is, in the example shown in FIG. 15 , it is judged that the portion outside the gun internal space needs not to be considered any more, because the gun unit 68 is successively withdrawn up to the center of gravity position G 1 . Therefore, the investigation start position Ts is also updated in order to set the workpiece model 1096 again at that point of time.

The workpiece solids 1096 are extracted and updated in the same manner as in step S 106 described above. A new principal component line M 1 and a new center of gravity position G 1 are determined in the same manner as in step S 107 described above to update them respectively, and then the routine returns to step S 114 . After the routine returns to step S 114 , the processing is continued for the new workpiece solids 1096 , the principal component line M 1 , and the center of gravity position G 1 determined in step S 123 .

As described above, the portion, which is not included in the gun internal space, is successively excluded from the processing objective. Therefore, it is possible to determine the path for withdrawing the gun unit 68 for the workpiece 80 having any complicated shape as well.

However, if the loop, which is formed by steps S 114 to S 123 , is executed not less than predetermined number of times, it is judged that it is extremely difficult to withdraw the gun unit 68 for the workpiece 80 . Therefore, the processing is finished to make the plan again.

Next, explanation will be made for steps S 124 to S 130 as the processing to be performed if it is judged in step S 119 that any interference occurs due the operation along the path Vn (n=1, 2, 3, . . . ). In this case, only a portion of the workpiece model 1096 , which is located near the opening of the gun unit 68 , is extracted (or subjected to the mask process) to preferentially use the extracted portion so that the withdrawing path is determined.

In step S 124 shown in FIG. 9 , as shown in FIG. 16 , a portion of the workpiece model 1096 , which is located on the side of the opening of the gun unit 68 , is designated as a new objective workpiece portion 1096 c based upon the center of gravity position G 1 , and the portion is distinguished from a portion 1096 c which is located on the side opposite to the opening. In the distinguishing process, the process is conceived so that the gun unit 68 is withdrawn for only the portion disposed closely to the opening. The processing is reserved for the portion 1096 c located on the side opposite to the opening to extract the new objective workpiece portion 1096 b of the opening. The workpiece model 1096 is replaced with the new objective workpiece portion 1096 b to be dealt with up to steps S 125 to S 130 as the downstream processes.

Subsequently, in step S 125 , the principal component line M 2 and the center of gravity position G 2 are determined in relation to the new objective workpiece portion 1096 b in the same manner as in the process in step S 107 described above.

In step S 126 , the path V 3 for connecting the investigation start position Ts and the center of gravity position G 2 is defined in the same manner as in step S 114 described above to assume the posture of the gun unit 68 in which the vector Xr is allowed to coincide with the path V 3 based upon the center of gravity position G 2 .

Subsequently, in step S 127 , the posture of the articulated robot 50 is determined with the assumed posture by the inverse operation in the same manner as in step S 115 described above.

Subsequently, in step S 128 for the branching judgment, it is investigated whether the solution in the inverse operation is determined normally in the same manner as in step S 116 described above.

If the solution is not determined normally, the rotation operation is performed to make rotation by α° about the center of the vector Yr (step S 129 ) in the same manner as in step S 117 described above. The routine returns to step S 127 .

If the solution is determined, in step S 130 , the gun unit 68 is linearly operated along the path V 3 from the investigation start position Ts to the center of gravity position G 2 to investigate the interference in the same manner as in step S 118 described above. The routine returns to step S 119 to judge the interference investigation.

As described above, even if no appropriate path is found when the path is retrieved for the object of the entire workpiece model 1096 , then only the new objective workpiece portion 1096 b , which is located closely to the opening of the gun unit 68 , can be preferentially used to determine the withdrawing path by applying the mask process to the workpiece model 1096 . Further, in the downstream processing, the workpiece model 1096 is successively converted into one having the simple shape by combining the updating process for the workpiece model 1096 in step S 123 described above, making it easy to determine the withdrawing path.

If the loop, which is formed by steps S 127 to S 129 , is continuously executed predetermined number of times, it is judged that the gun unit 68 cannot be arranged at the center of gravity position G 2 . The routine returns to step S 124 in order to perform the further mask process. However, if the mask process is executed not less than predetermined number of times, it is judged that the mask process is not effective for the shape of the workpiece 80 . The routine returns to step S 120 which is the withdrawing process applied with no mask process to calculate the withdrawing path again.

In step S 131 as the termination process, for example, the coordinate of the withdrawing position Ue as the investigation end position and the vector data are added as the operation data to the path table 120 (see FIG. 10 ). Among them, Un as the operation data is inserted between the respective welding points Tn in the path table 120 . Next, the routine returns to the process shown in FIG. 6 .

As described above, even if no appropriate path is found when the path is retrieved for the object of the entire workpiece model 1096 , then only the portion, which is located closely to the opening of the gun unit 68 , can be preferentially used to determine the withdrawing path by applying the mask process to the workpiece model 1096 . Further, in the downstream processing, the workpiece model 1096 is successively converted into one having the simple shape by combining the updating process for the workpiece model 1096 in step S 123 described above, making it easy to determine the withdrawing path.

In the above explanation, the technique for determining the path to withdraw the gun unit 68 from the welding point of the workpiece 80 has been described. As for the path for advancing the gun unit 68 into the welding point, the advancing path may be obtained by inverting the order in the path table 120 .

The principal component line has been used as the reference line for the workpiece model 1096 . Another reference line such as a straight line based on the least square method or a curve having an arbitrary order may be used if the shape of the workpiece model 1096 is represented by the line or the curve.

Next, explanation will be made with reference to FIGS. 17 to 21 for the method for setting the wide-area operation path in step S 6 shown in FIG. 6 .

In the following description, as shown in FIG. 17 , explanation will be made for an example in which the gun unit 68 is operated from the start point P 1 at which the workpiece 80 as the thin plate is disposed to the arrival point P 2 . It is assumed that the obstacle 82 exists between the start point P 1 and the arrival point P 2 . The withdrawing positions Ue, which are determined in the setting of the narrow-area operation path described above, are dealt with as the start point P 1 and the arrival point P 2 .

In step S 201 shown in FIG. 18 , the wide-area operation path-setting section 35 b of the operation path-setting program 35 is executed by a predetermined operation method by an operator for the off-line teaching apparatus 10 . This process may be continuously performed after the setting of the narrow-area operation path.

In step S 202 , the wide-area operation path-setting section 35 b reads, from the hard disk 34 , the condition data 37 as the condition for setting the operation path, and the data is stored in RAM 29 . Further, the start point P 1 and the arrival point P 2 for setting the operation path as well as the shape of the workpiece 80 and the position of the obstacle 82 or the like are recognized from the condition data 37 .

Subsequently, in step S 203 , the operation path (path) 100 to connect the start point P 1 and the arrival point P 2 is set to investigate the acceptance or rejection of establishment of the posture and the occurrence of any interference when the gun unit 68 is operated along the operation path 100 .

Specifically, dividing points, which are obtained by dividing the operation path 100 into those having minute lengths, are set by the function of the path-investigating section 35 c . The postures of the articulated robot 50 , i.e., the angles of rotation θ 1 to θ 6 , which are obtained when the gun unit 68 is arranged at the respective dividing points, are determined by means of the inverse operation.

When the posture of the gun unit 68 differs between the start point P 1 and the arrival point P 2 , the vector Xr, the vector Yr, and the vector Zr for indicating the posture of the gun unit 68 may be defined at the respective dividing points in a manner of linear interpolation. In this investigation, it is assumed that the electrodes 70 , 72 are opened so that they may not interfere with the workpiece 80 .

If the posture of the articulated robot 50 holds at each of the dividing points, the operation from the start point P 1 to the arrival point P 2 is actually assured.

Steps S 206 , S 212 , S 215 , S 218 , S 224 , and S 227 described later on are also executed by the function of the path-investigating section 35 c.

In step S 204 , it is judged whether the solution of the inverse operation is normally determined at each of the dividing points. Specifically, it is judged whether TCP is capable of arriving at the dividing point. If the solution is not determined, if the value of the angle is without the rotatable range of the axis J 1 to J 6 even if the solution is determined, or if the articulated robot 50 interferes with the obstacle 82 or the like in the determined posture, then the routine proceeds to step S 205 . If the solution is normally determined, the termination process is performed for the setting of the wide-area operation path in step S 229 .

The function of the interference of the simulation circuit 40 may be used for the occurrence of interference.

In step S 205 shown in FIG. 19 , in order to avoid the obstacle 82 or establish the posture, the template operation is applied from the start point P 1 for the gun unit 68 to set a first junction point Q 1 . In this case, the template represents the prescribed operation to be executed by the articulated robot 50 .

It is assumed that the first template is applied to the start point P 1 and the arrival point P 2 .

As shown in FIG. 21 , the first template resides in the operation in which the first junction point Q 1 obtained by operating in the prescribed direction by the prescribed distance is set based upon TCP of the gun unit 68 , and the gun unit 68 is moved along the operation path (retracting path) 102 (see FIG. 17 ) for connecting the start point P 1 and the first junction point Q 1 . The first junction point Q 1 is obtained by moving the position of the start point P 1 . It is assumed that the direction of the gun unit 68 possessed by the start point P 1 , i.e., the direction of TCP is unchanged.

In general, in order to properly perform the welding operation, the vector Zr is set to be perpendicular to the workpiece 80 . Therefore, it is preferable that the prescribed direction is the withdrawing direction for the gun unit 68 , i.e., the direction opposite to the vector Xr. A distance, with which the gun unit 68 can be sufficiently disengaged from the workpiece 80 , may be previously prescribed for the prescribed distance depending on the size of the gun unit 68 . In the gun unit of a general size, it is preferable that the prescribed distance is 100 mm.

The first template provides an effective retracting method for the thin plate which is a general workpiece. It is possible to set the operation path in accordance with the predetermined convenient retracting method without being affected by the shape of the workpiece.

Subsequently, in step S 206 , the acceptance or rejection of the posture establishment of the articulated robot 50 at the first junction point Q 1 and the occurrence of interference with the peripheral obstacle are investigated in the same manner as in step S 203 .

Subsequently, in step S 207 , if it is judged that the posture of the articulated robot 50 holds at the first junction point Q 1 and there is no interference as a result of the investigation in step S 206 , the routine proceeds to step S 212 . Otherwise, the routine proceeds to step S 208 .

In step S 208 , in order to obtain the appropriate posture at the first junction point Q 1 , the posture is set, in which the gun unit 68 is rotated by a predetermined angle about the center of the vector Xr, Yr or Zr. The rotating process is performed together with step S 209 as the next judgment process to make successive rotation for all of the vectors Yr, Zr, and Xr.

Subsequently, in step S 209 , it is confirmed whether the added up angle of the rotation by the predetermined angle one by one arrives at 360°. If the added up angle is less than 360°, the routine proceeds to step S 206 to judge the posture of the articulated robot 50 .

If no proper posture is obtained at the first junction point Q 1 even if the rotation is performed by 360° for each of the vector Xr, the vector Yr, and the vector Zr, then the first junction point Q 1 is set again in step S 210 at a position returned by a predetermined distance in the direction toward the start point P 1 . That is, if the first junction point Q 1 is set at the distance of 100 mm from the start point P 1 , the point is returned by 10 mm in the direction toward the start point P 1 to set the point again at the position of 90 mm.

Subsequently, in step S 211 , the added up value of the distance of the return of the first junction point Q 1 is confirmed. If the point is returned to the start point P 1 as the original point, then the process is stopped, and the plan is made again. If the point is not returned to the start point P 1 , i.e., if the range of 10 to 90 mm is given, then the routine proceeds to step S 206 to judge the posture of the articulated robot 50 .

In step S 212 (if it is judged that the posture of the articulated robot 50 holds and no interference is caused in the judgment in step S 207 described above), the investigation is performed by the same process as in step S 203 for the acceptance or rejection of the posture establishment and the occurrence of the interference when the gun unit 68 is operated along the operation path 102 .

Subsequently, in step S 213 , the judgment is made in the same manner as in step S 204 . If it is judged that the posture of the articulated robot 50 holds at the dividing point on the operation path 102 and the operation can be performed along the operation path 102 , then the routine proceed to the next step S 214 . If it is judged that the operation cannot be performed, the routine is returned to step S 210 to further change the position of the first junction point Q 1 .

In step S 214 , it is confirmed that two of the first junction point Q 1 and the first junction point Q 2 are set for the start point P 1 and the arrival point P 2 . The routine proceeds to the next step S 215 . If the first junction point Q 2 corresponding to the arrival point P 2 is not set, the routine is returned to step S 205 shown in FIG. 19 .

Subsequently, in step S 215 , the operation path 104 for connecting the two first junction points Q 1 and Q 2 is set to investigate the acceptance or rejection of the posture establishment and the occurrence of the interference when the gun unit 68 is operated along the operation path 104 .

Specifically, the processing is performed while prescribing that the first junction point Q 1 is the new start point and the first junction point Q 2 is the new arrival point. The investigation is made for the operation path 104 in the same manner as in the investigation for the path between the start point P 1 and the arrival point P 2 in step S 203 described above.

Subsequently, in step S 216 , the judgment is made in the same manner as in step S 204 . If it is judged that the posture of the articulated robot 50 holds at the dividing point on the operation path 104 and the operation can be performed along the operation path 104 , then the termination process is performed for the setting of the wide-area operation path in step S 229 shown in FIG. 18 . If it is judged that the operation cannot be performed, the routine proceeds to the next step S 217 .

In step S 217 shown in FIG. 20 , in order to avoid the obstacle 82 , the template operation is applied from the first junction point Q 1 for the gun unit 68 to set a second junction point R 1 .

It is assumed that the second template is applied to the first junction point Q 1 (and Q 2 ).

As shown in FIG. 21 , the second template is used such that the line 108 for connecting the first junction point Q 1 and the predetermined established point 106 is set, and the second junction point R 1 is defined as the point obtained by moving by a prescribed distance from the first junction point Q 1 on the line 108 .

The second junction point R 1 is obtained by moving only the spatial position for the first junction point Q 1 . It is assumed that the direction of the gun unit 68 possessed by the first junction point Q 1 , i.e., the direction of TCP is unchanged.

The second template is provided for the gun unit 68 having been disengaged from the workpiece 80 in order to operate in the direction in which the interfering obstacle 82 does not exist. The movement is made in the direction toward the origin O with the free space in which the possibility of existence of the obstacle 82 is low. That is, in general, the obstacle 82 tends to be absent near the origin O such that the operation of the articulated robot 50 is not inhibited. When the operation is made in this direction, the possibility of avoiding the obstacle 82 is preferably increased. Further, as for the articulated robot of a general size, the prescribed distance is preferably 100 mm.

Those other than the origin O may be used as the established point 106 . If there is any place at which the obstacle 82 does not exist or if there is any place at which the operation is easily performed, such a place may be used for the established point 106 . For example, when the operation range of the articulated robot 50 is expressed in the space, it is conceived that the degree of freedom of the operation is largest at the central position. Therefore, such a position may be used for the established point 106 .

Subsequently, in step S 218 , the acceptance or rejection of the posture establishment of the articulated robot 50 at the second junction point R 1 and the occurrence of any interference with the peripheral obstacle are investigated in the same manner as in step S 203 .

Subsequently, in step S 219 , if it is judged that the posture of the articulated robot 50 holds at the second junction point R 1 and there is no interference as a result of the investigation in step S 218 , the routine proceeds to step S 224 . Other than the above, the routine proceeds to step S 220 .

In step S 220 , in order to obtain the appropriate posture at the second junction point R 1 , the posture is set, in which the gun unit 68 is rotated by a predetermined angle about the center of the vector Xr, Yr, or Zr in the same manner as in step S 208 .

Subsequently, in step S 221 , it is confirmed whether the added up angle of the rotation by the predetermined angle one by one arrives at 360°. If the added up angle is less than 360°, the routine proceeds to step S 218 to judge the posture of the articulated robot 50 .

If no proper posture is obtained at the second junction point R 1 even if the rotation is performed by 360° for each of the vector Xr, the vector Yr, and the vector Zr, then the second junction point R 1 is set again in step S 222 at a position obtained by movement by a predetermined distance in the direction toward the established point 106 . That is, if the second junction point R 1 is set at the distance of 100 mm from the first junction point Q 1 , the point is further moved by 100 mm in the direction toward the established point 106 to set the point again at the position of 200 mm.

Subsequently, in step S 223 , the added up value of the distance of the movement of the second junction point R 1 is confirmed. If the point arrives at the established point 106 , then the process is stopped, and the plan is made again. If the point does not arrive at the established point 106 , the routine proceeds to step S 218 to judge the posture of the articulated robot 50 .

In step S 224 (if it is judged that the posture of the articulated robot 50 holds and no interference is caused in the judgment in step S 219 described above), the operation path (retracting path) 110 for connecting the first junction point Q 1 and the second junction point R 1 is set. The investigation is performed by the same process as in step S 203 for the acceptance or rejection of the posture establishment and the occurrence of the interference when the gun unit 68 is operated along the operation path 110 .

Subsequently, in step S 225 , the judgment is made in the same manner as in step S 204 . If it is judged that the posture of the articulated robot 50 holds at the dividing point on the operation path 110 and the operation can be performed along the operation path 110 , then the routine proceed to the next step S 226 . If it is judged that the operation cannot be performed, the routine is returned to step S 222 to further change the position of the first junction point Q 1 .

In step S 226 , it is confirmed that two of the second junction points R 1 and R 2 are set for the first junction points Q 1 and Q 2 . The routine proceeds to the next step S 227 . If the second junction point R 2 corresponding to the first junction point Q 2 is not set, the routine is returned to step S 217 .

Subsequently, in step S 227 , the operation path 112 for connecting the two second junction points R 1 and R 2 is set to perform the investigation for the operation on the operation path 112 in the same manner as in step S 203 .

Subsequently, in step S 228 , the judgment is made in the same manner as in step S 204 . If it is judged that the posture of the articulated robot 50 holds at the dividing point on the operation path 112 and the operation can be performed along the operation path 112 , then the termination process is performed for the setting of the wide-area operation path. If it is judged that the operation cannot be performed due to the interference with the obstacle or the like, then the routine is returned to step S 222 , and the two second junction points R 1 , R 2 are further moved to repeat the process until the operation path holds.

After completing the setting of the operation path from the start point P 1 to the arrival point P 2 , the termination process is performed for the setting of the wide-area operation path in step S 229 shown in FIG. 18 . The termination process includes, for example, the recording of the set wide-area operation path on the path table 120 (see FIG. 10 ). The start point P 1 , the first junction point Q 1 , the second Junction point R 1 , the second junction point R 2 , the first junction point Q 2 , and the arrival point P 2 , which are included in the set operation path, are recorded in an order of operation on the path table 120 . Specifically, the values of the angles of rotation θ 1 to θ 6 about the respective axes of the articulated robot 50 and the values of the vectors Xr, the vector Yr, and the vector Zr indicating TCP and the position coordinates (X, Y, Z) at the respective points are recorded.

The operation path recorded on the path table 120 is converted by the data-preparing circuit 38 into the program data for operating the actual articulated robot 50 , and the data is transmitted to the robot control unit 22 .

The path table 120 is recorded in RAM 29 and the hard disk 34 . However, if necessary, the path table 120 may be printed or displayed on the screen of the monitor 16 .

In the foregoing description, the operation path 104 is the path for connecting the first junction points Q 1 and Q 1 . Alternatively, the first template may be applied to only the side of the start point P 1 to determine the first junction point Q 1 , and the application may be made as it is for the arrival point P 2 to set the path for connecting the first junction point Q 1 and the arrival point P 2 .

As for the operation path 112 , for example, the path for connecting the second junction point R 1 and the first junction point Q 2 may be set in the same manner as described above.

The operation paths 102 , 110 as the retracting path for making the retraction from the start point P 1 may be also used when the operation is made to another point other than the arrival point P 2 .

The prescribed distance, which is firstly applied for the first template, is 100 mm. Alternatively, starting from 10 mm, the distance may be elongated to 20 mm and 30 mm.

The order of application of the first and second templates may be inverted depending on the situation concerning, for example, the workpiece 80 and the obstacle 82 .

The set path table 120 indicates the wide-area operation path from the start point P 1 to the arrival point P 2 or the narrow-area operation path for representing the withdrawing operation from the welding point Tn. However, the operation paths are reversible, and they may be used upon the operation from the arrival point P 2 to the start point P 1 . Further, the path may be utilized up to an intermediate position without using the entire operation path.

Further, the embodiment of the present invention is applicable, for example, to an assembling robot and an applying robot other than the welding robot. The articulated robot 50 may have a seven-axis structure or a structure having, for example, a link mechanism or an expansion/contraction mechanism.

As described above, according to the embodiment of the present invention, the operation path 100 for connecting the start point P 1 and the arrival point P 1 is firstly set to investigate whether the gun unit 68 can be operated along the operation path 100 . Therefore, if the gun unit 68 can be operated along the operation path 100 , the operation path can be set extremely conveniently without providing any junction point or the like for the operation. Even if the operation on the operation path 100 cannot be performed, the first template is applied to operate by the prescribed distance in the direction opposite to the vector Xr as the prescribed direction from the start point P 1 or the arrival point P 2 . Therefore, the first junction points Q 1 and Q 2 can be set automatically and efficiently without performing any complicated calculation and without being affected by the shape of the workpiece 80 .

The first template is used to operate by the prescribed distance with which the gun unit 68 can be sufficiently retracted from the workpiece 80 depending on the size of the gun unit 68 in the prescribed direction set in the direction in which it is conceived to retract the gun unit 68 most easily with respect to the workpiece 80 . Therefore, although the method is convenient, the possibility of the successful and safe retraction from the workpiece 80 is high. Further, for example, in step 206 , the safety is verified. Therefore, there is no fear of interference or the like when the articulated robot 50 is actually operated.

According to the embodiment of the present invention, if the first junction points Q 1 , Q 2 or the second junction points R 1 , R 2 , which are set on the retracting path, are the points at which the articulated robot 50 cannot arrive or at which any interference occurs, the prescribed distances of the first and second templates are corrected to set the positions of the first junction points Q 1 , Q 2 or the second junction points R 1 , R 2 again. Therefore, it is possible to set the preferable retracting position.

As for the second template, the prescribed direction is the direction toward the origin O for the coordinate calculation for the articulated robot 50 . Therefore, the possibility of interfering with the obstacle 82 is low.

Further, according to the embodiment of the present invention, the first template and the second template are applied in combination. The gun unit 68 is firstly retracted from the workpiece 80 with the first template, and then the gun unit 68 is retracted from another obstacle 82 or the like with the second template to thereby verify the safety. Therefore, it is possible to set the retracting path and the wide-area operation path automatically and efficiently without performing any complicated calculation. Thus, it is of course possible to improve the operation efficiency. Further, it is also possible to improve the quality of the off-line teaching data without relying on the skill of the operator.

It is a matter of course that the setting method and the setting apparatus for the operation path for the articulated robot according to the present invention are not limited to the illustrative embodiments described above, which may be embodied in other various forms without deviating from the gist or essential characteristics of the present invention.