JP2010137298A - Method of preparing working program for double arm robot - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for preparing a working program for a double arm robot, efficiently preparing a working program for a plurality of manipulators according to an assembling procedure intended by a teaching worker in a short time. <P>SOLUTION: A three-dimensional model for a plurality of parts is prepared, manipulators for holding and operating respective parts connected to each other are selected from a plurality of manipulators, the positions of the points through which the parts move in the assembling operations and the attitudes of the parts at these route points are set up, numbers for the route points are allocated in the order of the arrival of the manipulators, the route point data list including route point numbers, part names to be operated, route point positions, and attitudes is prepared for each manipulator, and working macros are combined according to the contents of working which are performed by the manipulators at the route points. Thus, a working program for a plurality of manipulators is prepared. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a method for creating a work program for a robot composed of a plurality of manipulators, and more particularly to a method for automatically generating a work program related to a part assembling work using a plurality of manipulators.

Normally, teaching of a teaching playback type robot is performed using a manual operation means called a programming pendant. Operate each axis of the robot with the axis operation keys on the programming pendant, jog the hand to the work position for the target workpiece, and then press the memory key to store the angle data of each joint axis at that time along with the movement command . A work program composed of a robot language is created by repeating the operation of jogging the robot to a desired position and storing it. In the reproduction operation in which the robot actually performs the work, the angle data of each joint axis taught is sequentially read out, an interpolation command to each axis motor of the robot is generated, and the tip of the robot is moved by position speed control.
On the other hand, when teaching complicated work such as assembly of parts by a robot, such a method is complicated and takes time to teach. In particular, when the assembly work is performed by a multi-arm robot having two or more manipulators as disclosed in Patent Document 1, the work performed by the robot itself can be made efficient, but the work is taught. It is difficult even for a skilled person to do. Patent Document 1 discloses a multi-arm robot that performs assembly work of components including a harness, but does not particularly describe a method of teaching assembly work.
Therefore, as a method for easily teaching such complicated work, there is a method of generating a robot work program using a three-dimensional CAD or an offline simulator.
As a specific method, a moving procedure of a part is defined by three-dimensional CAD, and a robot operation is determined so as to realize the movement (for example, Patent Document 2).
In addition, as a method for creating a work program for a robot composed of a plurality of manipulators, a procedure for moving parts for assembly is defined for each manipulator, and the priority of the manipulator is set for each movement operation unit. There is a robot work program creation method (for example, Patent Document 3) in which the operation procedure of each manipulator is created in combination, and the operation procedure that enables work and the work time is the shortest is selected.
JP-A-7-223179 (FIG. 1) JP 2006-350620 A (FIG. 2) JP 63-295192 A (Table 2)

However, the conventional work program creation method disclosed in Patent Document 2 is applied to work by a single manipulator, and when teaching a robot composed of a plurality of manipulators, each manipulator There was a problem that the operation could not be determined.
Moreover, although the conventional work program creation method disclosed in Patent Document 3 sets a part moving procedure for each of a plurality of manipulators, it is necessary to set a priority for each manipulator and each operation unit. For this reason, there is a problem that the number of combinations of operation procedures becomes very large in the case of work with many types of operations, and the amount of calculation for searching for the optimum operation procedure becomes very large.
In addition, there are many combinations of operation procedures that are actually inoperable due to interference between manipulators, assembly procedures, and the like, resulting in increased useless calculations. Furthermore, it is unclear to the teaching worker what operation procedure is selected until the end (which parts are moved first and which parts are assembled first, etc.), and the teaching worker has no intention. There was a problem that it might become an operation procedure that does not.
Furthermore, when a work item is added, the operation procedure must be searched again from the beginning, and there is a problem that it takes time to create a work program.
In addition, as a problem common to both, when actually performing assembly work with a robot, not only the movement of the part but also the part position is measured by a sensor, and the parts are fitted together while controlling the force. However, in the conventional method, robot commands related to such work need to be added separately after generating the work program, and teaching work where to insert the command in the work program There was also a problem that it was difficult for people to understand.
The present invention has been made in view of such problems. A work program for a plurality of manipulators can be generated based on an assembly procedure intended by a teaching worker, and the work program can be efficiently and quickly generated. An object of the present invention is to provide a work program creation method for a multi-arm robot that can be generated.

In order to solve the above problem, the present invention adopts the following method.
The invention according to claim 1 is a work program creation method for a multi-arm robot that operates a multi-arm robot having a plurality of manipulators and generates a work program for performing assembly work of a plurality of parts off-line. For each of a plurality of parts, a first step of selecting and setting a manipulator for gripping and operating the part from the plurality of manipulators;
The second step of setting the position of each point through which each part passes in the assembling operation, the posture of each part at each of the via points, and the manipulator for each via point set in the second step Are assigned numbers in the order of arrival, and the position and orientation of the part at the waypoint are defined as the position and orientation of the waypoint. For each manipulator, the number of the waypoint, the name of the part to be operated, the position and posture of the waypoint A third step of creating a route point data list comprising: a fourth step of selecting work contents to be executed by the manipulator at each route point of the route point data list from a pre-registered work list; Depending on the work contents set for the waypoints, a combination of work macros consisting of a group of instructions for the manipulator A fifth step of creating a plurality of manipulators of the work program, that consists of the features.

  According to the second aspect of the present invention, in the second step, in addition to the plurality of parts, a dummy part is virtually defined, and the position and orientation of the via point of the dummy part are set, thereby setting the part. The operation of only the manipulator is created without gripping.

  The invention according to claim 3 is characterized in that, in the fourth step, a work condition is set as a parameter relating to the work content.

  According to a fourth aspect of the present invention, when the plurality of manipulators are operated in synchronization in the assembly work, a synchronization command and a synchronization operation are performed when setting the work contents in each manipulator in the fourth step. It is characterized by designating a part to be a target and a waypoint number of the part.

  In the fifth aspect of the invention, in the fifth step, each joint target angle of the manipulator is calculated from the position and orientation of the via point using inverse kinematics calculation, and the manipulator passes a singular point. Changes the target posture of the manipulator hand within a predetermined range, obtains an evaluation value by a predetermined evaluation method for the changed target posture, and changes the target posture of the manipulator according to the evaluation value.

  In the invention according to claim 6, in the fifth step, an interpolation operation of the manipulator between the via points is simulated, and when the manipulator passes a singular point in the interpolation simulation, The target posture is changed within a predetermined range, an evaluation value by a predetermined evaluation method is obtained for the changed target posture, and the target posture of the manipulator is changed according to the evaluation value.

  According to a seventh aspect of the present invention, when the manipulator has a redundancy degree of freedom of 7 degrees or more, the manipulator is based on an evaluation value obtained by the evaluation method when a redundant degree of freedom attitude of the manipulator is changed within a predetermined range. Then, the attitude of the redundant degree of freedom of the manipulator is changed.

  In the invention according to claim 8, when the manipulator is mounted on an external shaft that moves the pedestal of the manipulator, the evaluation by the evaluation method when the position of the external shaft is changed within a predetermined range The position of the external shaft is changed based on the value.

According to the first aspect of the present invention, it is possible to easily create a work program for a plurality of manipulators based on an assembling procedure intended by a teaching worker, and to easily add work items. Further, since the work content can be selected from the work list, the burden on the teaching worker for creating the work program can be reduced.
According to the second aspect of the present invention, since dummy parts that do not actually exist are defined and the position and orientation for each movement are set, the air cut operation and the retreat operation that are not related to the movement of the parts can be performed. A work program can be generated.
According to the third aspect of the present invention, since the work conditions are set according to the work content, it is possible to deal with various assembling work, and the versatility of the multi-arm robot can be increased.
According to the invention described in claim 4, since the waypoint number of the component to be synchronized with the component to be synchronized with the work content is designated, it becomes possible for the plurality of manipulators to operate in synchronization, and the plurality of manipulators to be operated. It is possible to create a work program that operates in cooperation.
According to the fifth to eighth aspects of the present invention, even when the position or posture of the specified part or its moving operation cannot be realized, work that can be performed by automatically changing the posture of the manipulator or the position of the external shaft. A program can be generated.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows the configuration of a multi-arm robot and apparatus for implementing the method of the present invention. Here, a double-arm robot having two left and right manipulators will be described as an example. In the figure, 101 is a left arm manipulator having joints J1 to J6, and 102 is a right arm manipulator having joints J1 to J6. The manipulators 101 and 102 are provided with grippers that open and close at the respective tip portions, and the parts on the work table 105 are gripped by the grippers to perform assembly work. In the work of assembling parts in this way, the manipulator is often subjected to force control and assembled while controlling the force. In such a case, a force sensor (not shown) is mounted between the tip of the manipulator and the gripper.
Reference numeral 103 denotes a control device for the manipulators 101 and 102. The command generation unit 1032 generates commands to both manipulators according to the operation commands described in the work program read by the work program reading unit 103, and servo control of each manipulator is performed. Position commands are sent to units 1033 and 1034.
Servo controllers 1033 and 1034 control the position and speed of each axis motor of each manipulator.
In the example of FIG. 1, the gripper operates by air pressure, and the gripper is opened or closed by supplying or stopping supply of air pressure to the gripper by an I / O command described in the work program.
When executing force control, the command generation unit 1032 sends a force command to the servo control units 1033 and 1034 based on the measurement result of the force sensor, and the servo control units 1033 and 1034 also control force of each axis motor of each manipulator. Execute.

The work program used by the control device 103 is created by the work program creation device 104. The work program creation method of the present invention is also executed on the work program creation device 104.
In the work program creation device 104, the robot models of the manipulators 101 and 102 and the work table 105 are constructed as models of 3D computer graphics based on 3D CAD data and the like. Information such as the positional relationship of the manipulators, the axis configuration of each manipulator, and the operation area are also included in the robot model. Further, the assembly target model 1041 is converted from the three-dimensional CAD data or the like, and is constructed in the work program creation apparatus 104 as a three-dimensional computer graphics model.
The work information input unit 1042 inputs various information such as work contents and work conditions for creating a work program using a GUI. Information by each model and GUI is displayed on a display device 106 connected to the work program creation device 104.
The work program creation device 104 includes a work program generation unit from a part macro defined by the assembly target model 1041, information input by the work information input unit 1042, and a work macro 1043 prepared in advance corresponding to each work content. In 1044, a work program is automatically generated. This work program creation device is realized by, for example, a personal computer and off-line teaching software operating on the personal computer. Alternatively, it may be implemented as a function of a programming pendant added to the robot controller for operation.
When the work program creation device is realized by a personal computer and offline teaching software, the work information input unit 1024 is specifically a keyboard or a mouse. Moreover, when implementing as a function of a programming pendant, it becomes a key or touch panel provided on the programming pendant. Since the input method by GUI using the work information input unit 1024 is the same as that of a general personal computer, the description is omitted.
In FIG. 1, the work program creation device 104 and the display device 106 are separated. However, when the work program creation device is implemented as a function of a programming pendant, the display device 106 is integrated with the work program creation device 104. It may be.

  FIG. 2 shows an example of a component to be assembled and its assembly. As shown in FIG. 2A, the part A (202) and the part B (203) are assembled to the part 201 placed on the work table as shown in FIG. Here, the component 201 has holes 2011 and 2012 for assembling the component 202 and the component 203, respectively, and the component 202 and the component 203 are connected by a cable 204. The operation of assembling two parts connected by such a cable cannot be performed by only one manipulator. It is necessary that the two manipulators carry them in a coordinated manner so that no tension is applied to the cable 204 in a state where the gripper grips the parts 202 and 203 with the respective grippers, and assemble them to the part 201.

FIG. 3 is a flowchart showing a process of creating a work program executed by the work program creating apparatus 104 described with reference to FIG.
First, an assembly target model 1041 is created (step S0). Here, it can be modeled as three-dimensional computer graphics by reading the three-dimensional CAD data to be assembled shown in FIG.
Prior to the assembly work, the parts are supplied by a parts supply device installed around the robot, and this is set as the initial position of the parts.
Next, a manipulator that operates each component to be assembled is designated (step S1).
As a specific designation method, the assembly target model read in step S0 is displayed on the display device 106 using the GUI of the offline teaching software, and each component (202, 203) is selected on the display device 106. 203 is designated to be gripped and operated by the left manipulator, and the part 202 is designated to be gripped and operated by the right manipulator. However, since the component 201 is not directly operated by the manipulator and is only fixed on the work table, the manipulator for gripping and operating the component 201 is not specified.

  Next, the position and orientation of a point through which each component passes during the component assembling operation are set (step S2). Here, for the parts 202 and 203, the position / posture for each waypoint when moving from the initial position / posture shown in FIG. 2 (a) to the assembly completion position shown in FIG. 2 (b) is set. The posture of the via point is set by the posture of the orthogonal coordinate system provided for each of the components 202 and 203.

FIG. 4 shows a specific example of the position / posture of each via point of the parts 202 and 203 in the part assembling work. Reference numerals 2021 and 2031 denote orthogonal coordinate systems set for the parts 202 and 203, respectively. The coordinate systems 2021 and 2031 are for representing the position and orientation of the representative point set for each component.
4A, the part 203 is gripped by the gripper of the left manipulator hand and the part 202 is gripped by the gripper of the right manipulator hand as specified in step S1, and lifted as shown in FIG. 4B. At this time, the posture of the component 203 is rotated by 90 ° from the state (a) by the left manipulator in order to make the component 203 suitable for assembly to the component 201. In (c), it is conveyed onto the part 201, and in (d), the part 202 is moved to the approach position for fitting into the hole 2011 of the part 201.
After fitting the part 202 into the hole 2011, the gripper of the right manipulator hand is released to release the grip of the part 202, and the part 203 is moved to the approach position for fitting the part 203 into the hole 2012 in (e). After fitting the component 201 into the hole 2012, the gripper of the component 201 is released by releasing the gripper of the left manipulator hand.
In this way, the position and orientation at each via point of the coordinate system 2021 of the component 202 and the coordinate system 2031 of the component 203 are designated according to the component assembly procedure assumed by the teaching worker.
As a specific designation method, in the GUI environment of the off-line teaching software including the three-dimensional computer graphics of the assembly target model created in step S0, the teaching worker displays the assembly target model displayed on the display device 106 as shown in FIG. As shown in (a) to (e), the position is recorded by moving it to a via point, and the coordinate system of the parts 202 and 203 at the via point of each part is added to determine the axial direction. By repeating this operation for each via point, the position and orientation at each via point of the coordinate system 2021 and the coordinate system 2031 are determined.

Next, a route point data list of each actual manipulator is generated based on the manipulator designated to operate each component and the route points of the component (step S3).
The waypoint data list is automatically generated by the work program generation unit 1044 for the left manipulator list and the right manipulator list, and includes a waypoint number, a part name, and a position and orientation of the waypoint. The waypoint number is a number assigned in the order in which the parts move, and the position and orientation of the waypoint are the position and orientation of the waypoint of each part set in step S2.

Next, the work contents to be executed at the waypoints of the respective parts set as described above are set (step S4). The work content includes “gripping”, “fitting”, “grip cancellation” and the like in the work example of FIG. 2, but in addition to this, “passing”, “stop”, and the like are included as one of the work contents. The work content is selected from those registered in advance.
Also, work conditions are input according to the selected work content.
For example, working conditions for “fitting” include force control parameters, amount of fitting, and the like. A plurality of such work conditions are prepared in advance, a unique number is assigned to each work condition, and the work condition number is set as necessary. The teaching worker sets such work contents and work conditions by the work information input unit 1042 while watching the screen displayed on the display device 106 as appropriate.
FIG. 5 shows an example of a GUI screen for inputting work contents and work conditions in step S4. The via point number, part name, work content, and work condition number are shown from the left of the table. Here, the waypoint number and the part name are shown in a table as shown in FIG. 5 for each manipulator according to the waypoint data list generated in step S3, and the teaching worker sequentially inputs work contents and work condition numbers there. I will do it.

Finally, a work program for the manipulator is automatically generated by combining work macros corresponding to work contents at each waypoint number in the table completed in step S4 (step S5).
A work macro is a group of instructions composed of one or more robot languages for realizing each work content with a manipulator. Robot language includes various operation commands for performing PTP (point-to-point) control, I / O commands for opening and closing the gripper, work commands for fitting using force control, sensor commands for obtaining sensor information, etc. The operation macro is a combination thereof and is registered in advance.
In order to increase the versatility of work macros, the movement commands in the work macros generally specify relative movement amounts such as “movement amount from the current position”, and specify the destination directly. It is configured so that it can be used from various work programs.

An example of a gripping work macro corresponding to the part gripping of the via point number 1 of the left manipulator in FIG. 5 and an example of the working condition thereof, and an example of a working macro corresponding to the part fitting of the via point number of the left manipulator and the working condition thereof As shown in FIG.
FIG. 6A shows a work macro “Grip” corresponding to the work content “gripping”. The work macro “Grip” has a (first line) grip position measurement command using the sensor and a (second line) movement command based on the measurement result by the sensor,
(Third line) It is composed of three instructions for closing the gripper. The fourth line is an instruction for returning to the work program that called the work macro.
At the waypoint number 1, the part is in the initial position as shown in FIG. 4A. However, when the manipulator actually approaches the initial position of the part and grips the part with the gripper, the position of the part is reconfirmed. There is a need to. Therefore, the gripper is provided with a sensor for measuring the position of the component, and the position of the component is first measured in the first line of the work macro “Grip”.
The position of the part to be gripped with reference to the measurement sensor is assigned to P001 by the first line, and the manipulator operates by the value assigned to P001 by the second line to finely adjust the position of the gripper.
Also, the work conditions shown in FIG. 6B are based on the I / O number of the control device corresponding to the opening / closing of the gripper, the object number for measurement by the sensor, and the grip position correction value for determining the grip position based on the measurement result. Become.

Similarly, the waypoint number 5 of the left manipulator in FIG. 5 describes the operation of inserting the part A (203) into the hole (2012) of the part C (201) from the initial position as shown in FIG. Yes. FIG. 6C shows a work macro “Fitting” corresponding to the work content “Fitting”. The work macro “Fitting”
(1st to 3rd lines) A group of instructions relating to the TUKITE instruction for lowering the manipulator until the part A (203) hits the fitting part of the part C (201) (4th to 6th lines) Group of SAGURI instructions that finely move the manipulator to find the hole when it hits the edge of the hole (2012) of part C (7th to 10th lines) Related to the INSERT instruction that inserts part A to the bottom of the hole of part C Consists of instruction groups.
If this work is successful, 0 is returned in the 11th line, and if it is unsuccessful, -1 is returned to the work program that called this work macro in the 13th line.
In addition, as shown in FIG. 6 (d), the operation conditions for the above-described butting, searching, and insertion are input.
Note that “* _INS” on the 7th line and “* _ERROR” on the 12th line are labels indicating the destinations of the third, sixth, and 10th JUMP instructions. “GETSTATUS” is an instruction that takes in a variable such as B001 as a return value whether an immediately preceding instruction such as a hit instruction (TUKIATE instruction) or a search instruction (SAGURI instruction) has succeeded or failed. Instructions such as a TUKIATE instruction, a SAGURI instruction, and an INSERT instruction are prepared in advance as manipulator work instructions.
Work macros and work conditions corresponding to the respective work contents such as “gripping”, “passing”, “fitting”, “holding cancellation”, and “stop” are registered in advance, and in step S 4 By specifying the work conditions, information necessary for generating a work program for the assembly work can be obtained.

As described above, work macros and work conditions are prepared according to the work contents that can be selected, and the work macros are combined in accordance with the order of the table created in step S4, whereby a manipulator work program is automatically generated.
As described above, in the present invention, the manipulator that works for each part is set, the position and orientation of the waypoint of the part is set according to the determined assembling procedure, and the work content corresponding to that is set and set. By taking the procedure of creating a work program for multiple manipulators from a combination of work macros that realize the work content, a work program for multiple manipulators can be generated based on the assembly procedure intended by the teaching worker, In addition, a work program can be efficiently generated in a short time.

In the actual assembly work, the manipulator at the standby position approaches the part, moves away from the work, or moves one part with the other manipulator without directly operating the part. There are parts that need to move only the manipulator.
The operation of such a manipulator defines dummy parts 302 (part D) and 303 (part E) that do not actually exist as shown in FIG. 7 in the process (step S2) of setting the position and orientation of the via point of each part. Furthermore, a manipulator for operating the manipulator can be generated by designating a manipulator for operating the dummy parts and setting the position and orientation 3021 and 3031 of the via points of each dummy part.
Also in step S4, the work content to be executed at the via point of the dummy part set here is shown as No. in FIG. 1 or No. Set as in 8. By doing so, it is possible to include the approach operation to the part of the manipulator without moving the part and the retracting operation of the manipulator after completion of the assembly work in the work program automatically generated in step S5. Become.

Furthermore, when performing work with a plurality of manipulators, it is necessary to prevent the manipulators from interfering with each other. Also, in assembling the parts as shown in FIG. 2, it is necessary that the two manipulators operate in a coordinated manner while holding the parts 202 and 203, respectively, so that no tension is applied to the cable. When a plurality of manipulators must operate synchronously in this way, when setting the work contents of the waypoints of each part in step S4, the parts to be synchronized and the waypoint numbers of the target parts to be synchronized are set. specify. An example of a GUI to which a field for inputting such a synchronization condition is added is shown in FIG.
The left manipulator No. 2 or No. As shown in FIG. 5, the parts to be synchronized (part B operated by the right manipulator in this example) and the via point number (via point number of the part B operated by the right manipulator in this example) are input.
FIG. 9 shows an example of a work program generated based on the work contents of the left and right manipulators input for synchronization in this way. FIG. 9A shows an example of a work program for the left manipulator, and FIG. 9B shows an example of a work program for the right manipulator. A work program as shown in FIG. 9 is created corresponding to the work contents and work conditions specified in FIG. Here, P000 to P004 in the work program are the robot position and orientation data created in step S5 from the position and orientation of the parts specified in step S2. COND is a tag for designating a work condition number. At this time, the synchronization of the left and right manipulators is realized by inserting a SYNC command. The SYNC command is a command to synchronize the left manipulator (R0) and the right manipulator (R1). When this command is executed, it waits until the SYNC command of the counterpart manipulator is executed. Therefore, when both SYNC instructions are executed, the execution of the lower work program is resumed at the same time.
By doing so, it becomes possible for a plurality of manipulators to operate in synchronism, so that it is possible to create a work program that operates in cooperation with the plurality of manipulators.

In the automatic generation of the work program (step S5), the position and orientation of each via point of the part set in step S2 are converted into the position and orientation of the manipulator tip, and this is used as input to calculate each of the manipulators by inverse kinematic transformation calculation. Calculate the joint angle target value.
As a result, the target angle of each joint axis at each via point of each manipulator is obtained, and a work program can be generated together with the movement command. The calculation of the inverse kinematic transformation is a known method, and the work program relating to movement is completed by performing this calculation for all waypoints.
However, the position of the via point is specified in the GUI environment and is not determined by actually operating the manipulator. Therefore, there may be a case where the waypoint is a singular point. If singularities are included, the solution of inverse kinematic transformation cannot be obtained. In some cases, the solution obtained by the calculation of inverse kinematics conversion is out of the movable range of the joint of the manipulator, and the position and orientation of the via point of the part set in step S2 is a position and orientation that cannot be realized by the manipulator.
On the other hand, for via points that simply pass through or via points that do not have any problems if the work posture is changed by changing the gripping method of the part, if only the target posture of the manipulator hand at that via point is changed, the target position will be In many cases, it can be realized. For example, in FIGS. 4B and 4C, the parts 202 and 203 do not necessarily have the postures shown in FIG.
Therefore, when a via point is generated in which the solution cannot be obtained by inverse kinematic conversion in step S5 or the solution is out of the joint movable range, the target posture of the manipulator hand is changed within a predetermined range, The feasible position and orientation of the manipulator at the via point of the part can be obtained by changing the target posture by a predetermined evaluation method.
For example, the evaluation method for determining the target posture may be as follows.
(1) A direction in which the posture is most easily changed is determined from the radius of the manipulable ellipsoid (see “Tsuneo Yoshikawa“ Robot Control Basics ”).
The manipulability ellipsoid represents the maneuverability of the hand. The direction in which the principal axis radius is long is the direction in which the hand speed can be easily obtained, and the direction in which the short direction is small is the direction in which only a small hand speed can be obtained. Therefore, it is possible to know the direction that is easily changed by the manipulable ellipsoid principal axis radius (singular value of the Jacobian matrix).
(2) The posture is swung in a certain range in the above direction, and a range in which the inverse kinematic transformation can be calculated is determined.
(3) Among the above ranges, the solution in which the inverse kinematics conversion is farthest from the limit value of each joint movable range is set as the target posture.
Thus, a work program that can be realized by automatically changing the target posture can be automatically generated.

In addition, when the manipulator has a configuration with seven or more axes having redundant degrees of freedom as shown in FIG. 10, it is possible to avoid a singular point or a joint movable range over by changing the posture of the redundant degrees of freedom. . For example, in a 7-axis manipulator, without changing the position and posture of the manipulator tip, the position of the manipulator elbow (the intersection of the 3rd axis (J3) and 4th axis (J4) axes counted from the base of the robot) It is possible to change the elbow position).
The origin of the manipulator (the point on the axis of the first axis (J1) counted from the base of the robot is located at the same height as the axis of the second axis (J2)) and the wrist point (the axis of the J6 axis and J7 The rotation angle of the elbow position around a straight line connecting the axis of the axis) is called the elbow rotation angle, and is often used as a redundant degree of freedom in a seven-degree-of-freedom manipulator.
Using such redundant degrees of freedom avoids singular points and joint movement range over.
In step S5, when a via point where the solution is not obtained by inverse kinematic conversion or the solution is out of the joint movable range is generated, the elbow rotation angle of the manipulator is changed within a predetermined range, By changing the target elbow rotation angle according to the evaluation method, the feasible position and orientation at the waypoint of the manipulator can be obtained.
For example, the evaluation method for determining the target elbow rotation angle may be as follows.
(1) The elbow rotation angle is changed within a certain range, and the range in which the inverse kinematic transformation can be calculated is determined.
(2) Among the above ranges, a solution having the farthest distance from the limit value of the joint movable range (with respect to each joint angle limit value) as the solution of the inverse kinematic conversion is set as the target elbow rotation angle.
(3) or the above-mentioned range having the highest manipulability is defined as the elbow rotation angle.
By automatically changing the redundancy degree of freedom posture in this way, a feasible work program can be automatically generated.

Furthermore, when the manipulator is mounted on an external traveling platform or turntable whose position can be controlled, it is possible to avoid singular points or over the joint movable range by changing the position of the traveling table or turntable. Is possible.
In step S5, if there is a transit point where the solution cannot be obtained by inverse kinematics conversion or the solution is out of the joint movable range, the position of the traveling table or the turntable on which the manipulator is mounted is set to a predetermined value. The position and orientation of the manipulator that can be realized at the waypoint of the part can be obtained by changing the range and changing the position of the traveling table or the turntable by a predetermined evaluation method.
For example, the evaluation method for determining the position of the traveling table or the rotating table may be as follows.
(1) The direction of change of the traveling platform and the turntable is determined from the radius of the manipulability ellipsoid of the manipulator.
(2) Change the position of the carriage or turntable in a certain range in the above direction, and subtract the change in the position and orientation of the manipulator base from the target position and orientation of the manipulator hand to perform inverse kinematic conversion and calculate it The correct range.
(3) Among the above ranges, a solution in which the inverse kinematics conversion is farthest from the limit value of the joint movable range is set as a target position of the traveling platform or the turntable.
A work program that can be realized can be automatically generated by automatically changing the position of the traveling table or the turntable in this way.

The above shows an example in which the position and orientation of the via point of the part set in step S2 cannot be realized by the manipulator. However, the position and orientation cannot be realized when moving by interpolating between the via points. In some cases. This can be confirmed by performing an interpolated movement simulation between via points using the off-line teaching software for the generated work program.
In this case, at the point where the target position and posture cannot be realized during the movement of the manipulator, the posture is determined by using the same method as the method of changing the posture at the waypoint, the posture of the redundant degree of freedom, and the position of the traveling platform and the turntable. It is also possible to change the position of the redundant degree of freedom, the position of the traveling platform or the turntable, and add the point as a new waypoint to the work program.

Configuration diagram of a multi-arm robot apparatus to which the method of the present invention is applied The figure which shows an example of the assembly | attachment operation | work which applies the method of this invention. The flowchart which shows the process sequence of the method of this invention. The figure which shows the position and orientation setting method for every movement of the components of the method of this invention The figure which shows the example of the input screen of the work content and work condition of the method of this invention The figure which shows the example of the work macro of the method of this invention, and work conditions The figure which shows the position and orientation setting method in the dummy components of the method of this invention The figure which added the synchronous condition to the work content and work condition input screen of the method of the present invention The figure which shows the example of the work program produced | generated by the method of this invention Diagram showing a manipulator with redundant degrees of freedom

Explanation of symbols

101 Left arm robot (6 degrees of freedom)
102 Right arm robot (6 degrees of freedom)
103 Control Device 104 Program Creation Device 105 Worktable 106 Display Device 201 Part C
2011 A hole for mounting a part B 2012 A hole for mounting a part A 202 A part B
2021 A coordinate system 203 representing the position and orientation of the part B. Part A
2031 Coordinate system 204 representing position and orientation of component A Cable 302 Dummy component D
303 Dummy part C
3021 A coordinate system representing the position and orientation of the dummy component D 3031 A coordinate system representing the position and orientation of the dummy component C

Claims (8)

A work program creation method for a multi-arm robot that operates a multi-arm robot having a plurality of manipulators and generates a work program for assembling a plurality of parts offline,
For each of the plurality of parts, a first step of selecting and setting a manipulator for gripping and operating the part from the plurality of manipulators;
A second step of setting the position of each point through which each component passes in the assembly operation, and the posture of each component at each of the via points;
For each waypoint set in the second step, a number is assigned in the order in which the manipulator arrives, and the position and orientation of the part at the waypoint are set as the position and orientation of the waypoint, and for each manipulator A third step of creating a via point data list including the via point number, the name of the part to be operated, and the via point position and orientation;
A fourth step of selecting a work content to be executed by the manipulator at each waypoint in the waypoint data list from a pre-registered work list;
A fifth step of creating work programs for the plurality of manipulators by combining work macros comprising instructions for the manipulators according to the work content set for each of the via points;
A work program creation method for a multi-arm robot characterized by comprising:   In the second step, in addition to the plurality of parts, a dummy part is virtually defined, and the position and orientation of the via point of the dummy part are set to operate only the manipulator without gripping the part. The work program creation method for a multi-arm robot according to claim 1, wherein:   2. The work program creation method for a multi-arm robot according to claim 1, wherein in the fourth step, a work condition is set as a parameter relating to the work content. When operating the plurality of manipulators in synchronization in the assembly work,
2. The multi-arm according to claim 1, wherein when the work content is set in each of the manipulators in the fourth step, a synchronization command, a part to be synchronized and a via point number of the part are designated. Robot work program creation method. In the fifth step, each joint target angle of the manipulator is calculated from the position and posture of the via point using inverse kinematics calculation, and when the manipulator passes a singular point,
2. The target posture of the manipulator hand is changed within a predetermined range, an evaluation value is obtained by a predetermined evaluation method for the changed target posture, and the target posture of the manipulator is changed according to the evaluation value. The work program creation method of the double-arm robot described. In the fifth step, the interpolation operation of the manipulator between the waypoints is simulated, and when the manipulator passes a singular point in the interpolation simulation,
2. The target posture of the manipulator hand is changed within a predetermined range, an evaluation value is obtained by a predetermined evaluation method for the changed target posture, and the target posture of the manipulator is changed according to the evaluation value. The work program creation method of the double-arm robot described. If the manipulator has more than 7 degrees of freedom,
7. The redundant degree of freedom posture of the manipulator is changed based on an evaluation value obtained by the evaluation method when the redundant degree of freedom posture of the manipulator is changed within a predetermined range. How to create a multi-arm robot work program. When the manipulator is mounted on an external shaft that moves the pedestal of the manipulator,
The multi-arm robot operation according to claim 5 or 6, wherein the position of the external axis is changed based on an evaluation value obtained by the evaluation method when the position of the external axis is changed within a predetermined range. How to create a program.

Images