JP2007144538A - Teaching data creating method for robot - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To quickly and easily create teaching data by effectively utilizing data on the retreat point before and after the work point in existing teaching data. <P>SOLUTION: A group is set including the data on the retreat point having relationship by each data on the work point from the existing teaching data Q, R (step S2). One of the groups is applied by each work point P1 to P4 of newly created teaching data P (step S3 to step S7). The selected group is rotated and converted according to the TCP of the work point (step S8). The ends of the groups are connected to each other to set teaching data P. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a teaching data creation method including the posture of a robot at each work point of a workpiece, and more particularly, to a teaching data creation method for a robot that sets teaching data for another workpiece using teaching data for an existing workpiece.

  Off-line teaching (hereinafter referred to as off-line teaching) is performed in order to increase the efficiency of teaching work of an articulated robot installed in a manufacturing line or to improve the operating rate of the manufacturing line. In off-line teaching, models of articulated robots and workpieces and surrounding structures that are work objects are built on a computer, teaching data is created using these models, and then the teaching data is supplied to the articulated robots on site. By doing so, teaching data can be created without stopping the production line.

  According to the off-line teaching, it is not necessary to stop the production line. However, if it is performed inefficiently, it is necessary to use a teaching computer for a long time, and the burden on the operator who performs the teaching work is large. The articulated robot is provided in the vicinity of a predetermined transfer line. For example, the articulated robot performs a welding operation on a vehicle transferred on the transfer line. However, the type of the transferred vehicle is not limited to one. First, when the corresponding vehicle type is changed, teaching data is newly created.

  In order to solve this problem, Patent Document 1 proposes a method of converting existing teaching data already created for an existing workpiece in accordance with the new workpiece and creating new teaching data. Yes.

  Specifically, when there is existing teaching data, one of a plurality of parameters indicating the posture of the end effector at the conversion destination work point is fixedly set, and the corresponding work point in the existing teaching data is set. The end effector posture is converted so as to match the fixed parameters. According to such a technique, the existing teaching data can be used effectively, and the use time of the computer and the burden on the operator can be reduced.

JP 2004-362018 A

  By the way, in the method proposed in Patent Document 1, existing data is effectively used for a work point at which a robot end effector performs work on a work. Some relay points and relay points to be relayed are set, and data such as these save points are not always effectively used.

  In addition, since existing data is optimally set for the corresponding production line, etc., when using it for new data, at least the work point and the evacuation point in the vicinity of it should not be modified as much as possible. It should be applicable.

  Furthermore, the setting of the evacuation point in the teaching data requires an appropriate setting corresponding to the work point data, so it takes time by trial and error, and the operator has a considerable amount of proficiency. Desired.

  The present invention has been made in consideration of such problems, and makes it possible to effectively use existing teaching data for an existing workpiece and efficiently create teaching data for another workpiece. An object of the present invention is to provide a teaching data creation method for a robot that can effectively and quickly perform the retraction point data corresponding to the work point in the teaching data.

  The teaching data creation method for a robot according to the present invention is a teaching data creation method for a robot in which work is sequentially performed on a plurality of work points of a work by an end effector, and corresponds to the plurality of work points and the work points in the first work. A dividing step of setting a group by including data of retraction points relevant to the work points for each work point data from the first teaching data of the robot in which data relating to the retraction points is recorded; A selection step of selecting any one of the groups for each work point with respect to second teaching data including data relating to one or more work points of two workpieces, and a group selected in the selection step as the second A conversion step of converting based on data relating to the work point of teaching data.

  In this way, a group is set by including the data of the evacuation points relevant to each work point data from the first teaching data, and any one of the groups is selected for each work point of the second teaching data. Thus, the retreat point is used while being associated with the work point. Therefore, it is not necessary to newly set a retreat point corresponding to each work point, and it is possible to perform offline teaching quickly and easily by effectively using the retraction point data in the existing first teaching data.

  In this case, the data relating to the work point and the evacuation point of the first teaching data are recorded in association with the operation order, and in the dividing step, the evacuation points at least before and after the operation order with respect to the work point are grouped. It should be included. As a result, the intrusion operation and the retreat operation with respect to the work point are also performed fairly accurately on the second workpiece.

  The data relating to the work point and the data relating to the retraction point include data on the entry direction or retraction direction of the end effector, and in the dividing step, the data enters or retreats with respect to the entry or retraction direction of the end effector at the work point. A retreat point where the angular difference in direction is within a predetermined range may be included in the group. As a result, the approach and retract directions of the end effector at the work point and the retreat point are maintained substantially coincident, and the operation of the end effector is suppressed to a small level. Therefore, work with a low possibility of interference during the operation between the work point and the retreat point is performed. Points and evacuation points can be grouped.

  Further, in the dividing step, it is preferable that the group includes a retreat point whose distance from the corresponding work point is within a predetermined range. As a result, distant retreat points are excluded, and a group with a predetermined range is obtained, which is easy to apply to the second teaching data.

  In the first teaching data, it is confirmed that the working point, the retracting point, and the posture of the robot between the working point and the retracting point do not interfere with the workpiece or the like. The group may be moved and rotated while maintaining the relative positions of the work point and the retreat point. As a result, even in the work on the second workpiece by the second teaching data at the conversion destination, the probability of interference is considerably reduced, and the labor for correcting the data is reduced.

  According to the teaching data creation method for a robot according to the present invention, a group is set for each work point data from the first teaching data by including retraction point data relevant to the work point, and the second teaching data. By selecting any one of the groups for each work point, the evacuation point is used while being associated with the work point. Therefore, there is no need to newly set a retraction point for each work point, and the off-line teaching can be performed quickly and easily by effectively using the retraction point data in the existing first teaching data.

  Furthermore, by using the evacuation point while being associated with the work point, the degree of subsequent conversion is small, and the first teaching data is applied while maintaining the reliability and the optimality.

  Hereinafter, a robot teaching data creation method according to the present invention will be described with reference to FIGS. The robot teaching data creation method according to the present embodiment is used in the robot system 10 shown in FIG.

  As shown in FIG. 1, the robot system 10 includes an offline teaching device 11 and a robot 12 to which teaching data created by the offline teaching device 11 is applied.

  The robot 12 is an industrial articulated type, and includes a base portion 14, and a first arm 16, a second arm 18, and a third arm 20 in order with respect to the base portion 14, and the third arm An end effector 22, which is a welding gun, is provided at the tip of 20. The end effector 22 is detachable from the third arm 20. The first arm 16 can be rotated by axes J1 and J2 that can be rotated horizontally and vertically with respect to the base portion 14. The second arm 18 is connected to the first arm 16 so as to be rotatable about the axis J3. The second arm 18 can be twisted and rotated by the axis J4. The third arm 20 is connected to the second arm 18 so as to be rotatable about an axis J5. The third arm 20 can be twisted and rotated by the axis J6. Each of the shaft J4 and the shaft J6 can be twisted and rotated by 360 ° or more.

  The end effector 22 is a C-type welding gun having a pair of electrodes 22a and 22b that open and close on the axis L. When the electrodes 22a and 22b are closed, the work point on the axis L (hereinafter referred to as TCP (Tool Center Point)). ))) And contact the workpiece. A direction that coincides with the axial center of the electrodes 22a and 22b on the main body side from the TCP is a vector Z, and an opening direction of the C-shaped welding gun that is orthogonal to the vector Z and is the end effector 22 is a vector X. The direction of the vector Z can be referred to as the gun pressurizing direction. A direction perpendicular to the vectors X and Z is a vector Y.

  The driving mechanisms of the axes J1, J2, J3, J4, J5 and J6 and the opening / closing mechanisms of the electrodes 22a and 22b are driven by actuators (not shown), respectively, and the TCP coordinates are the rotation angles of the axes J1 to J6 and the various parts of the robot 12. Determined by dimensions.

  By the operation of the robot 12 having such a six-axis configuration, the end effector 22 connected to the tip portion can be moved to an arbitrary position in the vicinity of the vehicle to be transported and can be set to an arbitrary direction. . In other words, the end effector 22 can move with six degrees of freedom. The robot 12 may have an operation unit such as an expansion / contraction operation and a parallel link operation in addition to the rotation operation. The robot 12 operates according to teaching data set in the robot control unit 24.

  As shown in FIG. 2, two robots 12 are arranged adjacent to the vicinity of the vehicle production line 25 and operate simultaneously under the action of the robot controller 24 to weld the vehicle as a workpiece. To do. In this case, for example, one side performs welding of the front row side door frame and the other side performs welding of the rear row side door frame. The arrangement direction C of the two robots 12 is set in parallel with the conveyance direction of the production line 25, but the arrangement direction C is set non-parallel to the conveyance direction depending on the specifications of the robot 12 and the workpiece. In some cases. For example, when welding is performed using two ceiling-suspended robots, they may be arranged at right angles to the conveying direction.

  The offline teaching device 11 is configured by a computer. As shown in FIG. 3, the control unit 26 includes a CPU 28 that controls the entire offline teaching device 11, a ROM 30 and a RAM 32 that are recording units, and a hard disk. Creates a hard disk 36 to / from which data is read / written by a drive (HDD) 34, a recording medium drive 40 for reading / writing teaching data and the like to / from an external recording medium 38 such as a flexible disk and a compact disk, and teaching data for the robot 12. A teaching data creation circuit 42 and a simulation circuit 44 that performs an operation simulation of the robot 12 based on the created teaching data are provided. A display 46 for assisting teaching work by an operator, displaying a simulation image, and the like is connected to the control unit 26 via a drawing control circuit 48 and a keyboard 52 as an input device via an interface 50. And a mouse 54 are connected.

  The hard disk 36 includes a teaching data creation program 56 for creating teaching data of the robot 12, shape data 58 relating to the robot 12, work object and other equipment, and a robot including the operation specifications of each axis of the robot 12. Specification data 60 is recorded.

  As shown in FIG. 4, the work to be welded by the robot 12 corresponds to the vehicle door frames (first work) 100 and 102 for which corresponding teaching data (first teaching data) Q and R have already been created. The teaching data (second teaching data) P to be performed is the door frame (second workpiece) 104 of the vehicle for which the creation has not been performed, and welding is performed on each door frame. The door frames 100, 102, and 104 are door frames corresponding to the left side of the front row of the vehicle, respectively, and have a shape approximated to some extent.

  In FIG. 4, each door frame 100 to 104 is schematically and redundantly shown in a state where each corresponding vehicle is transported on the production line 25, carried into a station welded by the robot 12, and stopped. Yes. That is, the case where it arrange | positions virtually on the same position on the manufacturing line 25 on the basis of the robot 12 is shown.

  The door frame 100 is relatively vertically long, and the door frame 102 is relatively horizontally long. The door frame 104 has a substantially intermediate shape between the door frames 100 and 102. Teaching data Q and R are existing data and data that has been actually applied to the door frames 100 and 102. Therefore, in the operation based on teaching data Q and R, it has been confirmed that there is no interference with obstacles, mutual interference between the two robots 12, and that the cycle time is within the required value. Each R is optimized for the production line 25.

  Teaching data Q corresponding to the door frame 100 includes a plurality of teaching points Qa to Qv (lower case subscripts indicate an operation order (hereinafter the same). Qa to Qv are also representatively indicated as teaching points Qφ). It is data. Among these, the six teaching points Qb, Qf, Qj, Qn, Qq, and Qu are work points for welding the door frame 100, and the rest are retreat points. In the teaching data Q, welding is sequentially performed while moving clockwise with respect to the door frame 100 (for example, in the case of the left door frame). Teaching points Qa to Qv are recorded in association with each other in the order of operation.

  The retreat point is a teaching point that relays between each work point, and is set so that a predetermined obstacle can be avoided and the work point can be smoothly entered or retreated.

  Also, in FIG. 4, each teaching point Qφ typically indicates the position of the TCP of the end effector 22, and actually, as shown in FIG. 5, each axis J1 to J1 of the robot 12 for each point. The angle of J6 and the direction of TCP are defined. That is, as shown in FIG. 6, the orientation of the end effector 22 and TCP is defined three-dimensionally by orthogonal vectors X, Y, and Z with respect to the teaching point (for example, Qb). In FIG. 4, a vector X is representatively shown for teaching points Qb, Qf, Qj, Qn, Qg, and Qu that are work points. As apparent from FIGS. 4 and 6, the vector X is defined in a direction substantially perpendicular to the end face of the door frame 100. Further, the retract points (for example, Qa, Qc) whose operation order is front and back with respect to the work point are points in the vicinity of the corresponding work point (for example, Qb), and the door frame 100 is based on the work point. It is set at a position substantially perpendicular to the end face. Further, at the retract points before and after these work points, the vector X is set to be substantially perpendicular to the end face of the door frame 100. With such a setting, the end effector 22 is a C-type welding gun, so that the vector X is surely and easily moved into and out of the end face of the door frame 100. The retraction points other than the retraction points before and after the work point (for example, Qd) are appropriately relayed so that they can move in a short time without interference in consideration of the shape of the door frame 100 and other obstacles. It is set as a point.

  Like the teaching data Q, the teaching data R corresponding to the door frame 102 is data including a plurality of teaching points Ra to Ro. Of these, the four teaching points Rb, Rg, Rj, and Rn are working points for welding the door frame 102, and the rest are retreat points. As in the teaching data Q, the teaching data R is set with the angles of the axes J1 to J6 and the direction of the TCP for each of the teaching points Ra to Ro, and moves in the clockwise direction with respect to the door frame 102. Weld. The relative position between the work point and the retreat point and the direction of the TCP are also set in the same manner as the teaching data Q.

  Regarding the teaching data P for the door frame 104, only the position and the direction of TCP are provisionally set for the four work points P1 to P4, and the posture of the robot 12 at the work points P1 to P4 is not set. Also, the position of the retraction point, the orientation of the TCP at the retraction point, and the posture of the robot at the retraction point are not set. The positions of the work points P1 to P4 and the TCP are determined in advance according to design conditions.

  Next, a method for setting the teaching data P corresponding to the door frame 104 based on the existing teaching data Q and R will be described.

  In step S <b> 1 of FIG. 7, information about the shape of the door frame 104 and the four temporarily set work points P <b> 1 to P <b> 4 is input to the offline teaching apparatus 11. This input operation is performed by an operator directly inputting or reading from a predetermined recording medium. The subsequent processing is automatically performed in the off-line teaching device 11. In addition, each workpiece, the robot 12, the end effector 22, and the like in the following processing are handled as model data in a virtual space in the offline teaching apparatus 11.

  In step S2 (dividing step), a group is set by including relevant retraction point data for each work point data based on a predetermined rule from the teaching points Qφ and / or Rφ. By this grouping, for example, as shown in FIG. 4, six groups Gq1 to Gq6 are set. The group Gq1 consists of teaching points Qa to Qc, the group Gq2 consists of teaching points Qd to Qg, the group Gq3 consists of teaching points Qh to Qk, the group Gq4 consists of teaching points Ql to Qo, and the group Gq5 Consists of teaching points Qp to Qs, and group Gq6 consists of teaching points Qt to Qv.

  The teaching points Rφ are also grouped in the same manner. For example, as shown in FIG. 4, four groups Gr1 to Gr4 are set. The group Gr1 is composed of teaching points Ra to Rc, the group Gr2 is composed of teaching points Rd to Rh, the group Gr3 is composed of teaching points Ri to Rl, and the group Gr4 is composed of teaching points Rm to Ro. Each group set in this way can be used as a so-called template, and is used as shown below. The grouping procedure will be described later.

  In steps S3 to S7, a group to be applied to each target work point is selected. At this time, since the teaching data R and Q are already optimized, when the selected group is converted with respect to the target work point in the conversion process described later, the posture difference of the robot 12 before and after conversion is as small as possible. It is recommended to select such a group.

  In step S3, one target work point is selected from P1 to P4 which are work points for the door frame 104, the direct distance between the target work point and the work point of each group is checked, and the direct distance is within a predetermined range γ ( (See FIG. 4). For example, when the teaching point P1 is selected as the target work point, groups Gq1, Gq2, Gr1 and Gr2 are extracted as shown in FIG.

  As described above, since a group having a distance close to the target work point is selected, the robot 12 can be applied in a state where the posture of the robot 12 is substantially the same with respect to the target work point. Group selection by comparison of direct distances is a simple calculation. In addition, a work point where the posture of the robot 12 changes greatly during conversion can be eliminated, and subsequent calculations can be performed efficiently.

  In step S4, the distance in the arrangement direction C (see FIG. 2) between the target work point and the work point of the group selected in step S3 is checked, and a group whose distance is within the predetermined range γc is extracted. As a result, the groups Gq1, Gq2, and Gr1 are extracted from the target work point P1, and the group Gr2 is excluded.

  In this manner, by eliminating the group in which the distance in the arrangement direction C of the two robots 12 exceeds the predetermined range γc, the possibility of mutual interference between the two robots 12 is reduced. In other words, considering mutual interference between articulated robots, one robot tends to have a narrower allowable range of motion in the direction of the other robot. By selecting a work point with a short distance, each robot is less likely to change its posture in the arrangement direction C in the subsequent conversion process, thereby reducing the mutual interference of the robots.

  In step S5, the target work point and the TCP of the work point in the group selected in step S4 are compared, and those having a small angle difference are further extracted. For example, when compared with the vector X as a representative, as is apparent from FIG. 4, since the vector X at P1 is in the leftward substantially horizontal direction, the work point similarly has the group X having the leftward substantially horizontal direction X. Gq1 and Gq2 are selected, and the group Gr1 in which the direction of the work point vector X is diagonally upper left is excluded.

  Thus, by selecting a group having a work point with a small angle difference in the direction of the TCP with respect to the target work point, the posture difference of the end effector 22 before and after conversion can be reduced.

  In step S6, each posture of the robot 12 when the end effector 22 is arranged so as to have a TCP orientation defined for each of the target work points is obtained. The posture is obtained by a known inverse calculation method using a determinant, for example, based on the position of the target work point and the orientation of the TCP.

  In step S7, the posture of the robot 12 at the target work point is compared with the posture of the robot 12 at the work point in the group selected in step S5, and the posture of the robot 12 at the work point of the selected group is determined as the target work. When the posture is changed to the posture of the robot 12 at a point, a group of work points with the smallest posture difference of a specific joint of the robot 12 is selected. Specifically, among the six joints of the robot 12, a group having the smallest posture difference between the axes J4, J5, and J6, which are the three rotary joints on the distal end side to which the end effector 22 is connected, is selected. This posture difference is expressed by, for example, the sum of absolute values of angle differences or the square sum of angle differences. That is, in the robot 12, the base-side axes J1, J2, and J3 predominately define the position of the end effector 22, and the front-end axes J4, J5, and J6 are the postures of the end effector 22 (that is, roll, pitch). , Yaw) predominately. Accordingly, appropriate data that substantially maintains the posture of the end effector 22 is selected by suppressing the posture difference between the axes J4 to J6. In the case where a telescopic shaft is provided, the telescopic shaft does not affect the posture of the end effector 22. Therefore, it is preferable to compare the posture difference with respect to the three axes on the tip side of the rotating shaft excluding the telescopic shaft. By such processing in step S7, for example, group Gq1 is selected for P1, which is the target work point.

  A step of selecting a group including a work point at a short distance from the target work point (steps S3 and S4), and a group including a work point whose posture change with respect to the work posture of the end effector 22 with respect to the target work point is within a predetermined range. It is not necessary to perform the process (step S5) and the process (step S7) that includes a work point including a work point having a small attitude difference between the front end side three axes with respect to the target work point in the above order. You can do it in order. The group selected in each selection step is not limited to a method of selecting a group within a predetermined range, and a predetermined number (including 1) may be selected in order from a favorable condition.

  Next, in step S8, rotation conversion processing is performed so that the group finally selected in each of the selection processes corresponds to the target work point. For example, as shown in FIG. 8, even if the selected group Gq1 is moved and the teaching point Qb, which is a work point, is made to coincide with the target work point P1, the direction of TCP may be different. For example, when compared with the vector X as a representative, the position is shifted by the angle θ, the entire data of the group Gq1 is rotated about the vector Z by the angle θ. Thereby, the TCP of the teaching point Qb coincides with the TCP of the target work point, and the other teaching points Qa and Qc move while maintaining the relative position. In the teaching data Q and R, since it has been confirmed that the retract point does not interfere with the workpiece, the conversion process may be performed while maintaining the relative positions of the work point and the retract point in the group. . As a result, even in the work on the second workpiece by the second teaching data at the conversion destination, the probability of interference with the workpiece becomes considerably low, and the labor for correcting the data is reduced. Actually, three-dimensional rotation conversion is performed.

  In step S9, a group is selected for all four work points P1 to P4 of the door frame 104, and it is confirmed whether or not the rotation conversion has been completed. If unsupported work points remain, one of them is set as a new target work point, and the process returns to step S3 to continue the group selection process.

  In this way, groups are associated with each of the work points P1 to P4 by performing the processing of steps S3 to S8. For example, as shown in FIG. Gq1, Gr1, Gq4, and Gq6 are selected and rotationally converted.

  In step S10, the operation between the work points and the retreat points in the groups Gq1, Gr1, Gq4, and Gq6 corresponding to the work points P1 to P4 is simulated to check whether there is any interference between the work and the robot 12.

  If the occurrence of interference is confirmed, the corresponding group is moderately rotated around the target work point, and the simulation is performed again. In this rotation conversion, the entire group is rotated in the same manner as in step S9, and the relative position between the work point and the retreat point is maintained. In addition, the vector Z which is the direction of the TCP gun pressurization at the work point is fixed and the vector X and the vector Y are rotated. Thereby, the holding | maintenance attitude | position of the workpiece | work by electrode 22a, 22b is maintained appropriately.

  In step S11, as shown in FIG. 9, the ends of each group are connected and ordered to temporarily set teaching data P, and an operation simulation based on the teaching data P is performed, and there is no interference between the workpiece and the robot 12. Make sure. If interference occurs, make appropriate corrections. For example, when interference occurs due to the presence of the obstacle 200 on the path between the teaching point Rc and the teaching point Ql, a relay point Px for avoiding the obstacle 200 may be set.

  In this manner, the teaching data P for the door frame 104, which is another workpiece, can be created by effectively using the existing teaching data Q and R for the door frames 100 and 102. In such a teaching data creation method, the relationship between the work points and retreat points in the existing teaching Q and R partial groups is applied to the new teaching data P without breaking, and the shape of the door frame 104 is Since the shapes of the door frames 100 and 102 are approximated to some extent, there is a high possibility that a group satisfying the conditions is selected, and the optimality of the teaching Q and R with respect to the production line 25 is maintained.

  Next, the dividing step in step S2 will be described in detail. For example, the following six methods can be used to group work points and retreat points in the division process.

  In the first method, as shown in FIG. 10, among teaching points Q1 to Q9 (corresponding to the above teaching point Qφ), an intermediate point is centered on teaching points Q1 and Q7 which are working points. Spheres 110 and 112 having the same diameter in contact with each other are set, teaching points Q1 and Q2 included in the sphere 110 are set as a group G11, and teaching points Q6, Q7 and Q8 included in the sphere 112 are set as a group G12. Set to. In this case, the other teaching points Q3, Q4 and Q5 are excluded from the group.

  In the first method, by setting the spheres 110 and 112, evacuation points within an appropriate range centered on the work point are grouped and can be easily applied to teaching data P to be described later. Note that the balls 110 and 112 are for convenience and do not necessarily need to be set. The balls 110 and 112 are limited to a length that is ½ of the distance between the teaching point Q1 and the teaching point Q7, and the direct distance to the teaching points Q1 and Q7. May be selected within this length range.

  The second method is a method of selecting and grouping those whose distances are within a predetermined range α around the teaching points Q1 and Q7 as work points. In this case, if the range α is substantially the same as the radius of the spheres 110 and 112, the same grouping as in the first method is performed, and the group G21 corresponding to the teaching point Q1 includes the teaching points Q1 and Q2. The group G22 corresponding to the teaching point Q7 includes teaching points Q6 to Q8. If the range α is set to be wide and there is a place where the group G21 and the group G22 overlap, and there is a teaching point at the overlapping part, the same as the adjacent teaching point that exists in the non-overlapping range Set to a group. Moreover, you may make it belong to both groups in common.

  In the second method, the retreat points within the predetermined range α are grouped, so that each group is normalized by the range α, and is easy to apply to the teaching data P.

  The third method is a method of adding the teaching points Q3, Q4, and Q5 excluded in the first method or the second method to any group, and the distance on the route indicated by the broken line is any group. It is set so that it belongs to the same group as the one near the teaching point Qφ at the inner end. That is, since the teaching point Q3 is close to the teaching point Q2 at the end of the group G11 (or G21), it is set as the group G31 together with the teaching points Q1 and Q2. Since the teaching points Q4 and Q5 are close to the teaching point Q6 at the end of the group G12 (or G22), they are set as the group G32 together with the teaching points Q6 to Q8. As for the teaching point Q9, which group it belongs to is determined by the distance relationship with the group based on the left work point (not shown).

  In the third method, all teaching points Qφ belong to any group, and existing data is used without waste. In particular, when the door frame 104 corresponding to the teaching data P has a shape similar to that of the existing door frame 100 or 102, all the teaching points Qφ are used as a group, so that the subsequent data The necessity of correction and the degree of data correction are reduced, which is preferable.

  The fourth method is a method of selecting a predetermined number of teaching points at short distances from teaching points Q1 and Q7, which are work points. For example, the teaching points Q2 and Q3 which are two points closest to the teaching point Q1 are selected to set the group G41, and the teaching points Q6, Q5, Q8 and Q9 which are two points closest to the teaching point Q7 are selected. Select to set the group G42.

  In the fourth method, at least one teaching point in the order of operation with respect to the work point is secured, and the entry operation and the retraction operation with respect to the work point are performed fairly accurately also on the door frame 104 which is a new workpiece. .

  The fifth method is a method of setting a group in which the angle difference between the approaching and retracting directions is within a predetermined range with respect to the approaching or retracting direction of the end effector 22 at the work point. That is, as shown in FIG. 11, the direction of TCP at the teaching point Q1, which is a work point, is represented as vectors X0, Y0, Z0, and the direction of TCP at the teaching point Qφ, which is another retreat point, is represented as vectors X1, Y1, Z1. At this time, the vector X0 indicating the approaching or retracting direction is compared with the vector X1. Specifically, the vector X1 of the teaching point Qφ is moved in parallel to the teaching point Q1, and a vector X1y obtained by projecting the vector X1 on the X0-Y0 plane and a vector X1z projected on the X0-Z0 plane are obtained. Further, an angle θ1 formed by the vector X1y and the vector X0 and an angle θ2 formed by the vector X1z and the vector X0 are obtained, and it is confirmed whether or not each of the angles θ1 and θ2 is within a predetermined angle range.

  In this way, the teaching points Qφ whose angles θ1 and θ2 are within the predetermined angle range are set in the same group as the corresponding work points in the order from the work point. When one or more of the angle θ1 and the angle θ2 are outside the predetermined angle range, they are excluded from the group.

  In the fifth method, since the angle difference in the approach or retraction direction is set as a group, the work is similar in shape, so the possibility of interference during entry and retraction is low and applicable. Is easy.

  As shown in FIG. 12, in the sixth method, the approaching and retracting direction (that is, the direction of the vector X0) is perpendicular to the positional relationship between the teaching point Q1 as the work point and the teaching point Qφ as the retracting point. This is a method of obtaining a difference Yq in a specific direction (that is, vector Y0) and grouping those having the difference Yq within a predetermined range β.

  In this way, the teaching points Qφ whose difference Yq is within the predetermined range β are set in the same group as the corresponding work points in order from the work point. When the difference Yq is out of the predetermined angle range β, it is excluded from the group.

  In the sixth method, since the restriction is made by the difference Yq in the direction of the vector Y, it is possible to extract a retreat point set in substantially the same direction as the vector X0 indicating the approach retreat direction at the work point. Thereby, even if it is a different workpiece | work, the possibility of the interference at the time of approach to the work point and retraction is low, and application is easy. In the sixth method, the difference in the direction of the vector X0 is not taken into consideration, but since this direction is the retreat direction, it is unlikely that interference will occur during entry and retreat even if the distance is long.

  Note that the fifth method and the sixth method may be requirements for further narrowing down the groups selected by the first to fourth methods. The fifth method and the sixth method may be applied in common.

  The teaching points Qφ are grouped by any one of the first to sixth methods or a combination thereof. For example, the groups Gq1 to Gq6 and the groups Gr1 to Gr4 are set.

  As described above, in the teaching data creation method of the robot according to the present embodiment, a group is set by including the data of the retraction point relevant to the work point for each work point data from the teaching data Q and R. Then, by selecting any one of the groups for each of the work points P1 to P4 of the teaching data P to be newly created, the evacuation point is used while being associated with the work point. Accordingly, it is possible to reduce the need to newly set the evacuation point for each work point, and to effectively use the data of the evacuation point in the existing teaching data P and Q and perform offline teaching quickly and easily. Further, by using the evacuation point while being associated with the work point, the degree of subsequent conversion can be reduced, and the reliability and optimization of the teaching data Q and R can be maintained and applied to the teaching data P. In particular, as the door frame 104, which is a new workpiece, is closer to the door frames 102, 104, which are existing workpieces, the higher the reliability is, the higher the optimality is.

  The number of existing teaching data to be used is not limited to two teaching data Q and R, but may be three or more or one according to the shape of the corresponding door frame 104 or the like.

  Also, in the dividing step (step S2), the end effector 22 enters the group at the working point and the retracting point by including in the group a retracting point whose posture difference is within a predetermined range with respect to the posture of the end effector 22 at the working point. The retreat direction is kept substantially the same, and an appropriately associated group is obtained.

  Further, by including the evacuation points whose distance from the corresponding work point is within the predetermined range γ (or γc) in the group, the distant evacuation points are excluded, and a group having a group within the predetermined range is obtained.

  Furthermore, since it has the conversion process (step S8) which converts the selected group based on the data regarding the work points P1 to P4, the teaching data Q and R are made to correspond more appropriately to the teaching data P. Can do.

  The teaching data P and Q confirm that the robot 12 at the retraction point does not interfere with the door frames 100 and 102 and that the cycle time is within the required value. The conversion may be performed while maintaining the relative positions of the point and the retreat point. As a result, even in the work on the door frame 104 by the teaching data P at the conversion destination, the probability of interference is considerably reduced, and the labor for data correction is reduced.

  Most of the above processing is automatically performed under the action of the off-line teaching device 11, so that it is not necessary for the operator to work for a long time by trial and error, and the operation method can be mastered in a short time. .

  The teaching data creation method for the articulated robot according to the present invention is not limited to the above-described embodiment, and it is needless to say that various configurations and processes can be adopted without departing from the gist of the present invention.

It is a block block diagram of an offline teaching apparatus. 1 is a perspective view of a vehicle production line and a robot system. It is a block block diagram of an offline teaching apparatus. It is a schematic diagram which shows the teaching data corresponding to the door frame of each vehicle. It is a table which shows the content of teaching data. It is a perspective view which shows the direction of an end effector and TCP. It is a flowchart which shows the procedure of the teaching data of the robot which concerns on this Embodiment. It is a schematic diagram which shows the rotation conversion process of a group. It is a schematic diagram which shows the new teaching data set with respect to the workpiece | work. It is a schematic diagram which shows a mode that a work point and a save point are grouped. It is a schematic diagram which shows a mode that the retraction point whose angle difference of an approach or retraction direction is in a predetermined range with respect to the approach or retraction direction of the end effector in a work point is included in a group. It is a schematic diagram which shows a mode that the retraction point which is in the vector Y0 direction difference predetermined range on the positional relationship of a work point and a save point is included in a group.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Robot system 12 ... Robot 22 ... End effector 24 ... Robot control part 100-104 ... Door frame C ... Arrangement direction G1-G4, G11, G12, G21, G22, G31, G32, G41, G42, Gq1-Gq6, Gr1 to Gr4 ... groups P, Q, R ... teaching data P1 to P4 ... work point Px ... relay points Q1-Q9, Qa-Qv, Ra-Ro ... teaching point

Claims (5)

In the teaching data creation method for a robot that performs work sequentially on multiple work points of a workpiece by an end effector,
From the existing first teaching data of the robot in which data related to a plurality of work points in the first work and the retreat points corresponding to the work points are recorded, the retraction points relevant to the work points are determined for each work point data. A splitting process to set up groups with data,
A selection step of selecting any one of the groups for each work point with respect to second teaching data including data relating to one or more work points of the second workpiece;
A conversion step of converting the group selected in the selection step based on data on a work point of the second teaching data;
A teaching method for creating teaching data for a robot. The robot teaching data creation method according to claim 1,
Data relating to the working point and the retraction point of the first teaching data are recorded in association with each other in the order of operation,
In the dividing step, a teaching data creation method for a robot, wherein the group includes at least retraction points whose operation order is before and after the work point. In the robot teaching data creation method according to claim 1 or 2,
The data relating to the work point and the data relating to the retraction point include data on the intrusion direction or retraction direction of the end effector,
In the dividing step, the teaching data creation of the robot is characterized in that the group includes a retreat point whose angle difference in the approach or retraction direction is within a predetermined range with respect to the approach or retraction direction of the end effector at the work point. Method. In the robot teaching data creation method according to any one of claims 1 to 3,
In the dividing step, a robot teaching data creation method is characterized in that a retreat point whose distance from a corresponding work point is within a predetermined range is included in a group. In the robot teaching data creation method according to any one of claims 1 to 4,
In the conversion step, the teaching data creation method for a robot, wherein the group is moved and rotated while maintaining the relative positions of the work point and the retreat point in the group.

Images