Axis angle determination method for six-axis robot and control apparatus for six-axis robot

This application is based on and claims the benefit of priority from Japanese Patent Application No. 2013-225423, filed Oct. 30, 2013, the disclosure of which is incorporated herein in its entirety by reference.

1. Technical Field

The present invention relates to a method and an apparatus for determining an angle of each axis of a six-axis robot based on a result of teaching of a position and an attitude (position/attitude) of a hand of the six-axis robot through teaching.

2. Related Art

A six-axis robot includes first to sixth axes, and is capable of taking an attitude (posture) in which a rotation axis of the fourth axis and a rotation axis of the sixth axis robot match.

For example, in a perpendicular articulated six-axis robot, the rotation axes of a fourth axis and a sixth axis of the six-axis robot match when the six-axis robot takes an attitude (posture) in which a hand of the six-axis robot faces the vertical direction. In this attitude, the number of combinations of the respective angles of the fourth axis and the sixth axis is infinite. Therefore, the attitude is a singular point (singularity) at which a solution of the angle is not uniquely determined.

When the six-axis robot is taught by continuous pass (CP) teaching, the angle of each axis for each position/attitude is uniquely determined. Therefore, the singular point is not an issue. In addition, the singular point is also not an issue when point-to-point (PTP) teaching is performed, if the user specifies the angle of each axis for each position/attitude. The singular point becomes an issue in PTP teaching when only the position/attitude is specified and the angle of each axis is to be determined by inverse transformation.

Conventionally, the singular point is processed as an error. Therefore, a user is forced to avoid singular points. Alternatively, the user is forced to take measures, such as individually specifying the angles of not only the fourth axis and the sixth axis, but also the angles of the remaining axes. As a result, work load on the user also increases.

It is thus desired to provide an axis angle determination method for a six-axis robot and a control apparatus for a six-axis robot by which each axis angle can be determined for a singular point without placing load on the user.

According to one aspect of the present disclosure, there is provided an axis angle determination method for a six-axis robot. In the method, the angle or position of each axis is determined based on the result of the robot being taught a position/attitude of a hand of the six-axis robot through point-to-point (PTP) teaching. In this instance, the attitude of the robot of which the angle or position of each angle is to be determined next may be a singular point. The singular point is a state in which the rotation axes of a fourth axis and a sixth axis match. When the attitude of the robot is a singular point, the angle of either of the fourth axis and the sixth axis is fixed to a current value and the angle of the other axis is determined to move the robot to the singular point.

In other words, if the angle of either of the fourth axis and the sixth axis is determined to be the current value, the angle of the other axis can be determined so that the six-axis robot can take the position/attitude of the singular point by only the other axis being rotated by a predetermined angle. Therefore, in this method, a singular point is not processed as an error. The angles of the fourth axis and the sixth axis can be determined with certainty.

In the axis angle determination method for a six-axis robot of the present invention, the first axis to third axis of the six-axis robot may be configured by linear-motion axes that are perpendicular to one another. In a six-axis robot configured in this way, the first axis to third axis respectively correspond to three-dimensional coordinates X, Y and Z. The fourth axis to sixth axis are moved in the directions of the coordinate axes. Under a presumption that the six-axis robot is made to perform an operation such as pick-and-place, the six-axis robot often takes an attitude in which the approach vector of the hand indicates the negative direction of the Z axis, and the rotation axes of the fourth axis and the sixth axis match. If this attitude is handled by error processing as in conventional technology, error processing frequently occurs. The work load on the user increases. Therefore, when the six-axis robot configured as described above is the subject, the increase in work load on the user can be prevented. The invention of the present application is very effective.

In the axis angle determination method for a six-axis robot in the present invention, the axis of which the angle is to be fixed to the current value may be the fourth axis. In other words, in the configuration of a common six-axis robot, the rotation range of the fourth axis on the base portion side is often restricted to within±180 degrees, for example. Conversely, unlike the fourth axis, the rotation range of the sixth axis on the hand side is not restricted. For example, the sixth axis is often capable of rotating 360 degrees or more in one direction. This is because the wires and the like that are attached to the robot generally become fewer towards the hand. In other words, for the robot to take the attitude of a singular point, for example, the fourth axis is required to be rotated by half of a full rotation or more when the restriction on the rotation range applies. However, when the sixth axis is rotated, the rotation amount is often minimal. Therefore, the six-axis robot can take the attitude of the singular point in a shorter amount of time.

In the accompanying drawings:

FIG. 1 is a flowchart of a process for determining angles of a fourth axis and a sixth axis of a six-axis robot according to a first embodiment;

FIG. 2A is a joint model diagram showing a singular point of the six-axis robot;

FIG. 2B is a joint model diagram showing a non-singular point of the six-axis robot;

FIG. 3 is a perspective view of a configuration of the six-axis robot according to the first embodiment;

FIG. 4 is a functional block diagram of a configuration of a controller and a driving unit of the six-axis robot shown in FIG. 3 ; and

FIG. 5 is a flowchart of a process for determining the angles of a fourth axis and a sixth axis of a six-axis robot according to a second embodiment.

First Embodiment

A first embodiment will hereinafter be described with reference to FIGS. 1 to 4 . FIG. 3 is a perspective view of a configuration of a six-axis robot. A robot 1 has a first axis J 1 , a second axis J 2 , a third axis J 3 , a fourth axis J 4 , a fourth axis J 5 , and a sixth axis J 6 . Of the axes J 1 to J 6 , the first axis J 1 , the second axis J 2 , and the third axis J 3 are configured by linear-motion axes that are perpendicular to each other.

The first axis J 1 is disposed along the X-axis direction in FIG. 3 . The first axis J 1 is provided with a ball screw therein (not shown). The first axis J 1 also has a nut (not shown). A driving motor 20 ( 1 ) (see FIG. 4 ) rotates the ball screw, thereby fastening the nut to the ball screw. A movable body 2 is disposed on the top surface side of the first axis J 1 . When the driving motor 20 ( 1 ) rotates the ball screw, the movable body 2 moves linearly along the first axis J 1 .

The second axis J 2 is disposed along the Y-axis direction in FIG. 3 . One end side of the second axis J 2 is fixed to the top surface of the movable body 2 . The second axis J 2 is also provided with a ball screw and a driving motor 20 ( 2 ) therein, in a manner similar to the first axis J 1 . A movable body 3 is disposed on a surface of the second axis J 2 on the positive-direction side of the X axis in FIG. 3 . When the driving motor 20 ( 2 ) rotates the ball screw, the movable body 3 moves linearly along the second axis J 2 .

The third axis J 3 is disposed along the Z-axis direction in FIG. 3 . The bottom end side of the third axis J 3 is fixed to a side surface of the movable body 3 on the positive-direction side of the X axis in FIG. 3 . The third axis J 3 is also provided with a ball screw and a driving motor 20 ( 3 ) therein, in a manner similar to that described above. A movable body 4 is disposed on a surface of the third axis J 3 on the positive-direction side of the X axis in FIG. 3 . When the driving motor 20 ( 3 ) rotates the ball screw, the movable body 4 moves linearly along the third axis J 3

An arm portion 5 is connected to the movable body 4 . The arm portion 5 is composed of the fourth axis J 4 to the sixth axis J 6 . The fourth axis J 4 is provided on the base portion side of the arm portion 5 . A driving motor 20 ( 4 ) is provided within the movable body 4 . The fourth axis J 4 is rotated by the driving motor 20 ( 4 ) with the Z axis as the rotation axis. The fifth axis J 5 is connected to the tip end side (bottom end side in FIG. 3 ) of the fourth axis J 4 . The sixth axis J 6 is connected to the tip end side of the fifth axis J 5 .

The rotation axis of the fifth axis J 5 is perpendicular to the rotation axis of the fourth axis J 4 . A driving motor 20 ( 5 ) is disposed on one end side of the rotation axis of the fifth axis J 5 . When the fifth axis J 5 is driven by the driving motor 20 ( 5 ), the tip end side of the fifth axis J 5 is rotated.

The rotation axis of the sixth axis J 6 is perpendicular to the rotation axis of the fifth axis J 5 . A driving motor 20 ( 6 ) is provided within the sixth axis J 6 . The sixth axis J 6 is rotatably driven by the driving motor 20 ( 6 ). The robot 1 is fixed in a state in which, for example, both ends of the first axis J 1 are hung from a ceiling in a factory or the like. The position and attitude of a hand H 1 of the robot 1 at the tip of the sixth axis J 6 are changed by each of the first axis J 1 to sixth axis J 6 being moved or rotated.

The robot 1 is connected to a controller 10 (control apparatus including singular point judging means, angle determining means, and rotation axis determining means) and a teaching pendant 11 . As shown in FIG. 4 , the controller 10 includes a central processing unit (CPU) 12 , a read-only memory (ROM) 13 , a random access memory (RAM) 14 , a drive circuit 15 , a position detecting circuit 16 , and the like. The CPU 12 controls the overall controller 10 based on programs stored in the ROM 13 or the RAM 14 and the like. In addition, in a manner that is well known, the CPU 12 outputs various commands to the drive circuit 15 to enable the robot 1 to repeatedly perform a work cycle that is determined in advance.

The drive circuit 15 is configured by, for example, an inverter circuit that converts the commands outputted from the CPU 12 to electrical signals. The electrical signals are then used to provide the robot 1 with the commands. The robot 1 includes the plurality of motors 20 ( 1 ) to 20 ( 6 ) that respectively drive the first axis J 1 to sixth axis J 6 . The robot 1 also includes encoders 21 ( 1 ) to 21 ( 6 ) that to correspond to the plurality of motors 20 ( 1 ) to 20 ( 6 ).

The position detecting circuit 16 receives input of signals from the encoders 21 ( 1 ) to 21 ( 6 ) for detecting the rotation positions of the motors 20 ( 1 ) to 20 ( 6 ). The position detecting circuit 16 thereby detects the rotation positions of the motors 20 ( 1 ) to 20 ( 6 ).

The CPU 12 performs feedback control of the motors 20 ( 1 ) to 20 ( 6 ) based on the rotation positions of the motors 20 ( 1 ) to 20 ( 6 ) detected by the position detecting circuit 16 . In the robot 1 , the output from the motors 20 ( 1 ) to 20 ( 6 ) are transmitted to the first axis J 1 to sixth axis J 6 via reducers 22 ( 1 ) to 22 ( 6 ). The motors 20 ( 1 ) to 20 ( 6 ) and the reducers 22 ( 1 ) to 22 ( 6 ) configure a driving unit 23 .

As in the attitude shown in FIG. 3 , a state in which the rotation axis of the sixth axis J 6 matches the Z axis in the same manner as the fourth axis J 4 is an attitude that is a singular point. As shown in the joint model of the six-axis robot 1 in FIG. 2A , this attitude is a state in which the approach vector of the hand H 1 indicates the negative direction of the Z axis. In addition, FIG. 2B shows an attitude that is a non-singular point in which the rotation axis of the sixth axis J 6 does not match the rotation axis of the fourth axis J 4 . Under a presumption that the six-axis robot 1 having the structure shown in FIG. 3 is taught an operation such as pick-and-place, it can be easily recalled that instances in which the position/attitude becomes the singular point occur frequently.

Next, the working according to the present embodiment will be described with reference to FIG. 1 . FIG. 1 is flowchart of a process for determining the angles of the fourth axis J 4 and the sixth axis J 6 for each position/attitude (of a plurality of teaching points) taught to the robot 1 through PTP teaching using the teaching pendant 11 . The angles of the other axes and the positions of the movable bodies 2 to 4 are determined in a manner similar to a well-known manner. Therefore, descriptions thereof are omitted. Data indicating each position/attitude of the plurality of teaching points taught to the robot 1 are stored in the ROM 13 or RAM 14 of the controller 10 .

First, the controller 10 determines whether or not a position/attitude of the robot 1 that has been taught by PTP teaching is a singular point (step S 1 ). As shown in FIG. 2A , when an approach vector of the attitude of the robot 1 that has been taught is A=(X, Y, Z), the attitude is a singular point when the approach vector A=(0, 0, −1). Therefore, the approach vector A that is the singular point is A=(0, 0, −1) as shown in FIG. 2A . Therefore, the controller 10 determines that the position/attitude of the robot 1 that has been taught is a singular point if (YES), and proceeds to step S 2 . On the other hand, when determined that the position/attitude of the robot 1 is not a singular point, the controller 10 performs ordinary angle calculation in the same manner as that in the past. The controller 10 thereby determines the angles of the fourth axis J 4 and the sixth axis J 6 (step S 4 ). Here, step corresponds to singular point judging means (equivalent to a singular point judging unit).

At step S 2 , the controller 10 determines the sum of the angles of the fourth axis J 4 and the sixth axis J 6 based on a normal vector of the attitude of the robot 1 that has been taught. This process will be described hereafter. An attitude matrix of the fourth axis J 4 to sixth axis 76 of the six-axis robot 1 is expressed as follows.

In the attitude matrix, the first column is components of a normal vector (see FIG. 2 A), and the second column is components of an orientation vector (see FIG. 2A ), the third column is components of an approach vector (see FIG. 2A ), C i and S i are respectively the cosine and sine of the angle of the to i(=4 to 6) axis. At the singular point, the angle of the fifth axis J 5 is zero degrees. Therefore, C 5 =1 and S 5 =0. The attitude matrix becomes as follows.

The sum θ 4+6 of the angles of the fourth axis J 4 and the sixth axis J 6 is determined as follows based on components of the normal vector.

Here, tan i =S i /C i .

When the sum θ 4+6 of the angles is obtained at step S 2 , the controller 10 holds the current angle θ 4 for the fourth axis J 4 . The controller 10 determines the angle θ 6 of the sixth axis J 6 (step S 3 ). In other words, the angle θ 6 is determined by

θ 6 =θ 4+6 θ 4

Here, steps S 2 and S 3 correspond to an angle determining means (equivalent to an angle determining unit).

Here, at step S 3 , the fourth axis J 4 is held at the current angle θ 4 and the angle θ 6 of the sixth axis J 6 is determined for the following reason.

In other words, in the robot 1 , the rotation range of the fourth axis J 4 is restricted to within±180 degrees. However, the rotation range of the sixth axis J 6 is not restricted in this way. The sixth axis J 6 is capable of rotating 360 degrees or more in one direction. This is because the wires and the like that are attached to the robot 1 generally become fewer towards the hand H 1 .

In other words, to take the attitude of the singular point, for example, the fourth axis J 4 is required to be rotated by half of a full rotation or more when the restriction on the rotation range applies. However, when the sixth axis J 6 is rotated, the amount of rotation can often be minimal. Therefore, the six-axis robot can take the attitude of the singular point in a shorter amount of time.

According to the present embodiment such as that described above, the angle or position of each of the first axis J 1 to sixth axis J 6 of the six-axis robot 1 is determined based on the teaching result of the position/attitude of the hand H 1 of the of the robot 1 by PTP teaching. In this instance, the attitude of the robot 1 of which the angle or position of each of the first axis J 1 to sixth axis J 6 is to be determined next may be a singular point. The singular point is a state in which the rotation axes of the fourth axis J 4 and the sixth axis J 6 match. When the attitude of the robot 1 is the singular point, the controller 10 fixes either of the angles of the fourth axis J 4 and the sixth axis J 6 to the current value and determines the angle of the other axis to move the robot 1 to the singular point.

In other words, if the angle of either the fourth axis J 4 or the sixth axis J 6 is determined to be the current value, the angle of the other axis can be determined so that the robot 1 can take the position attitude of the singular point by only the other axis being rotated by a predetermined angle. Therefore, as a result of this method, the singular point is not processed as an error. The angles of the fourth axis J 4 and the sixth axis J 6 can be determined with certainty.

The robot 1 of which the first axis J 1 to third axis J 3 are configured by linear-motion axes that are perpendicular to each other is the control subject. Under a presumption that the robot 1 , configured in this way, is made to perform an operation such as pick-and-place, the robot 1 often takes the attitude of the singular point. In this attitude, the approach vector of the hand indicates the negative direction on the Z axis. If this attitude is handled by error processing as in conventional technology, error processing frequently occurs. The work load on the user increases. Therefore, when the robot 1 according to the present embodiment is subjected to control, the increase in work load on the user can be prevented. The present embodiment is extremely effective.

In addition, the axis to be fixed to the current value is the fourth axis J 4 . Therefore, the six-axis robot 1 can take the attitude of the singular point in a shorter amount of time. Furthermore, when the fourth axis J 4 is rotated, the amount of rotation may increase as in the instance described above, compared to when the sixth axis J 6 is rotated.

Therefore, for example, when the user (operator) is a novice, it can be presumed that the user considers the fourth axis J 4 to have rotated significantly more than should be expected based on the teaching result. At this time, the user questions whether or not an error of some sort has occurred in the teaching operation. The user may redo the teaching operation. Redoing an operation that is actually not required is very inefficient. Therefore, a situation in which a novice to the teaching operation is made to have unnecessary doubt can be prevented when the angle of the fourth axis J 4 is fixed to the current value and the sixth axis J 6 is rotated as according to the present embodiment.

In addition, according to the present embodiment, the Z-axis component of the approach vector A is referenced to determine the singular point at step S 1 . Determination of the singular point can also be performed based on the temporarily inversely transformed result of the position/attitude that has been taught. Here, examples of processing time in both methods will be compared in detail. As a condition for comparison, the average processing time of 1600 operations is used.

When the amount of time required to solve the inverse transformation is about 14.7 μs, the amount of time required for the controller 10 to make the determination at step S 1 is about 0.37 μs. The amount of time required for the controller 10 to determine whether or not the attitude is a singular point based on the result of the temporarily solved inverse transformation, and perform the process again when the attitude is determined to be a singular point is about 29.4 μs.

In this instance, the determination at step S 1 becomes meaningful in terms of processing time when the percentage of singular points among all teaching points is 2.5% or more (however, it is presumed that the processing time for a singular point according to the present embodiment is also 14.7 μs). Therefore, under the presumption of an actual case, it can be considered that the determination performed according to the present embodiment is substantially more advantageous.

Second Embodiment

FIG. 5 shows a second embodiment. Sections that are the same as those according to the first embodiment are given the same reference numbers. Descriptions thereof will be omitted. Differences will be described below.

As shown in FIG. 5 , according to the second embodiment, when determined “YES” at step S 1 , the controller 10 determines the sum θ 4+6 of the angles at subsequent Step S 11 , in a manner similar to that at step S 2 . In addition, the controller 10 determines a movement amount δθ from the current values of the angles. The movement amount δθ herein is equivalent to the “difference” between the angles determined at step S 3 according to the first embodiment.

At subsequent step S 12 , the controller 10 determines time T4 and time T6 based on the respective rotation speeds of the fourth axis J 4 and the sixth axis J 6 . The time T4 is when the fourth axis J 4 is rotated by the movement amount δθ. The time T6 is when the sixth axis J 6 is rotated by the movement amount δθ. In other words, the capacities of the respective motors 20 ( 4 ) and 20 ( 6 ) and the reduction ratios of the respective reducers 22 ( 4 ) and 22 ( 6 ) of the fourth axis J 4 and the sixth axis J 6 differ. Therefore, the values of the time T4 and the time T6 also differ.

At step S 13 , the controller 10 compares the lengths of the time T4 and the time T6. When determined that T4>T6 (YES), the controller 10 fixes the angle θ 4 of the fourth axis J 4 to the current value. The controller 10 rotates the sixth axis J 6 by δθ (step S 14 ). On the other hand, when determined that T4<T6 (NO), the controller 10 fixes the angle θ 6 of the sixth axis J 6 to the current value. The controller 10 rotates the fourth axis J 4 by δθ (step S 15 ). It goes without saying that either of the fourth axis J 4 and the sixth axis J 6 can be rotated when T4=T6. Steps S 12 and S 13 correspond to a rotation axis determining means.

As described above, according to the second embodiment, the controller 10 determines the movement amount δθ from the current values of the angles when the robot 1 is moved from a current position/attitude to an attitude of a next singular point. The controller 10 determines the time T4 when the fourth axis J 4 is rotated by the movement amount δθ. The controller 10 also determines the time T6 when the sixth axis J 6 is rotated by the movement amount δθ.

The controller 10 compares the lengths of the time T4 and the time T6. The controller 10 fixes the angle of either of the fourth axis J 4 and the sixth axis J 6 that has the longer time to the current value. The controller 10 changes the angle of either of the fourth axis J 4 and the sixth axis J 6 that has the shorter time. Therefore, the amount of time required for movement to the singular point can be shortened. Work efficiency can be improved.

The present invention is not limited to the embodiments described above or shown in the drawings. Modifications and expansions such as those below are also possible.

In the first embodiment, if the rotation range of the fourth axis J 4 is not restricted, the angle of the sixth axis J 6 may be fixed to the current value. The angle of the fourth axis J 4 may be changed. For example, a driving condition that the output torque of the fourth axis J 4 increases as the rotation speed of the sixth axis J 6 increases can be presumed. When focus is placed on torque depending on the load, it may be more advantageous to change the angle of the fourth axis J 4 . In addition, in this instance, play in the cables and apparatuses connected by wires outside of the robot 1 can be disposed near the first axis J 1 to third axis J 3 . The first axis J 1 to third axis J 3 are linear-motion axes. As a result, load placed on the first axis J 1 to third axis J 3 can be reduced.

The configuration of the six-axis robot is not limited to that shown in FIG. 3 . The present invention may also be applied to common perpendicular articulated six-axis robots capable of taking an attitude in which the rotation axes of the fourth axis and the sixth axis match. In this instance, regarding the singular point, a method enabling determination may be used depending on the configuration of the robot.

In the first embodiment, the sum θ 4+6 of the angles of the fourth axis J 4 and the sixth axis J 6 is determined based on components of the normal vector. Alternatively, the sum θ 4+6 of the angles of the fourth axis J 4 and the sixth axis J 6 may be determined as follows based on components of the orientation vector.