JP5790840B2 - Robot control apparatus and robot posture interpolation method - Google Patents

Description

  The present invention relates to a control apparatus and method for interpolating the posture of a robot when a hand of a robot grips the workpiece and moves the workpiece relative to a fixed tool to perform machining.

  2. Description of the Related Art Conventionally, when a workpiece is processed using a robot, it is common to fix the workpiece and work with the tool of the robot's hand. However, for some tasks such as applying seals and deburring, there are cases where tools such as a sealing gun and a router are fixed, and the hand of a vertical articulated robot grips the work. The work form may be referred to as an external TCP (Tool Center Point).

In the case of the external TCP, for example, as shown in FIG. 12, in the case of seal application, the plane formed with the workpiece processing point (application point) is always maintained at an angle of 90 degrees with respect to the paint discharge direction by the sealing gun. It is desirable to do. As shown in FIG. 12A, when the shape of the workpiece is a plane, there is no problem because only the position is changed while maintaining the posture of the hand. However, as shown in FIG. 12B, when the shape of the workpiece includes a curved surface, the robot needs to change the above-mentioned angle to 90 degrees by continuously and finely changing the posture of the hand. . Therefore, if the posture is not maintained as originally taught, the machining accuracy of the workpiece may be lowered.
Further, in the case of deburring shown in FIG. 13, according to the complicated outer shape of the workpiece, the posture of the hand of the robot is taught so that each point of the edge always hits the router at an optimum angle. It is necessary to control, and if the posture is not maintained, burrs cannot be removed and the machining accuracy is lowered.

  Although it is not a technique that is conscious of application to external TCP, for example, in Patent Document 1, when the robot's hand moves continuously via teaching points P1, P2, and P3, the posture (rotation matrix) at each teaching point is described. ) R1, R2, and R3, an equivalent rotation angle vector K12 related to the amount of change from the posture R1 to the posture R2 and an equivalent rotation angle vector K23 from the posture R2 to the posture R3 are calculated. A virtual inward trajectory in the vicinity of the teaching point P2 is calculated, sampling interpolation is performed at a rotation speed designated for the virtual inward trajectory, and a change amount to an interpolation point adjacent to the moving direction at each interpolation point A technique for calculating an equivalent rotation angle vector and generating an operation locus based on the equivalent rotation angle vector is disclosed. Further, in Patent Document 2, in order to suppress a sudden change in the posture of the robot, an interpolation command value related to the posture near the route change point is superimposed as an interpolation command value related to the posture of the robot near the route change point. Techniques using values obtained in this manner are disclosed.

JP 2005-275484 A Japanese Patent Laid-Open No. 11-237910

  However, since the technique disclosed in Patent Document 1 is a technique for generating a trajectory that goes around a specific teaching point, for example, as shown in FIG. A trajectory that does not pass is generated. Since the posture is a recalculated posture for each interpolation point on the trajectory, the initially taught posture may not be maintained. Further, when the control mode for interpolating the posture of the robot as in Patent Document 2 is applied to the external TCP as it is, the posture changes in the case of the external TCP, so that the machining accuracy of the workpiece is lowered. Therefore, these conventional techniques cannot be applied to external TCP.

  Basically, if teaching is performed precisely at minute intervals, it is possible to reproduce the position and posture of the hand as taught, but as a practical matter it is not possible to perform teaching as such. It is efficient and inevitably introduces a process called interpolation. Therefore, it is required to solve the problem on the assumption that teaching is performed at regular intervals and the teaching points are interpolated.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a robot control device and a robot posture interpolation method capable of improving work accuracy when a hand of the robot grips a workpiece to perform processing. Is to provide.

According to the robot control apparatus of the first aspect, the robot is composed of a plurality of X components, Y components, Z components, X-direction rotation components, Y-direction rotation components, and Z-direction rotation components which are already taught hand movement destinations. for been moving target point, the posture coordinate conversion processing unit, X at each moving target point, Y, the rotating components of the Z, performs rotation matrix conversion process to convert the rotational angle coordinates, rotating component To calculate a posture at a moving target point defined by a predetermined two-direction component among the X, Y, and Z direction posture coordinate values coordinated by the X component, the Y component, and the Z component. The coordinate value in any two directions is calculated. The posture vector amplification processing means virtualizes the posture vector determined by the coordinate values in the two directions calculated by the posture coordinate conversion processing means for each movement target point to an arbitrary length with each movement target point as the origin . Posture points are determined, and the coordinate values for the two directions at the virtual posture points are set as amplified first and second amplified direction posture coordinate values, respectively. The amplification posture coordinate value connection processing means performs a trajectory process for connecting the respective coordinate values as smooth trajectories for the amplified first and second direction posture coordinate values existing for each movement target point, and the obtained trajectory. The upper interpolation point is moved along a trajectory that has been corrected so that the two directions satisfy the orthonormal condition, and the posture between the movement target points is complemented.

  Here, “smooth” means mathematically continuous (that each point is differentiable). By controlling in this way, the posture of the hand is taught at each position when the workpiece is moved by moving the robot's hand to the fixed tool. The hand moves so that the movement trajectories of the posture determination coordinates of the respective axes that are set at the time are connected smoothly, that is, the posture at the time of teaching is maintained. Therefore, the angle formed by the surface including the machining point of the workpiece with respect to the fixed tool axis is maintained as taught. As a result, it is not necessary to increase teaching points in order to improve the machining accuracy of the workpiece, so that the teaching work can be simplified.

The figure which is one Example and shows the procedure of posture interpolation in an image The figure which shows the relation between teaching posture and each axis posture point Flow chart of posture interpolation processing (A) is a present Example, (b) is a figure which shows the attitude | position deviation of a prior art. (A) is a positional deviation numerical value, (b) is a posture deviation numerical value, showing the present embodiment and the prior art. Flow chart of hand movement path control Diagram showing how to find a tangent vector The figure which shows the command position and the curve which connected those command positions Graph showing speed pattern Diagram showing the system configuration of the robot Functional block diagram with a focus on control devices Diagram for explaining the prior art (part 1) Fig. 12 equivalent (2) Diagram explaining conventional position and orientation interpolation processing

  Hereinafter, an embodiment will be described with reference to FIGS. As shown in FIG. 10, the robot includes a robot main body 1, a control device (posture determination coordinate setting means, posture interpolation means) 2 for controlling the robot main body 1, and a teaching pendant 3 as a teaching device. In this embodiment, a robot main body 1 which is a control target of the control device 2 is configured as a vertical articulated work robot, for example, and includes a base 4 and a shoulder portion provided on the base 4 so as to be able to turn in the horizontal direction. 5, a lower arm 6 provided on the shoulder portion 5 so as to be turnable in the vertical direction, an upper arm 7 provided on the lower arm 6 so as to be turnable in the vertical direction and capable of being twisted and rotated, and the upper arm 7 And a wrist 8 provided so as to be turnable in the vertical direction.

The wrist 8 is provided with a flange 9 that can be twisted and rotated at the tip end. A hand 21 is attached to the flange 9, and the hand 21 grips a work 22. Then, the horizontal turning operation of the shoulder 5, the vertical turning operation of the lower arm 6, the vertical turning operation of the upper arm 7, the rotating operation of the upper arm 7, the vertical turning operation of the wrist 8, and the flange 9. The joints are rotated by driving the joints via an appropriate transmission mechanism (not shown) by a motor 10 (see FIG. 11), for example, a DC servo motor as a drive source.
In this embodiment, the workpiece 22 is machined by the above-described external TCP. The tool 23 is fixed, and the robot body 1 moves the workpiece 22 to contact the tool 23 to perform machining.

  As shown in FIG. 11, the control device 2 includes a CPU 11 as a control unit, a drive circuit 12 as drive means for driving the motor 10 of each joint, a detection circuit 13 and the like. The CPU 11 is connected to a ROM 14 that stores a system program for the entire robot, a RAM 15 that stores an operation program for the robot body 1, and an interface 16 for connecting the teaching pendant 3. In FIG. 5, the shoulder portion 5, the lower arm 6, the upper arm 7, the wrist 8, and the flange 9 are shown as a single block as a movable portion, and according to this, one motor 10 that is a driving source for these joints is also provided. Only shown.

  The detection circuit 13 is for detecting the current position (rotation angle) and current speed (rotation speed) of each joint. The detection circuit 13 includes a rotary encoder provided in the motor 10 that drives each joint. 17 is connected. The rotary encoder 17 serves both as a position sensor and a speed sensor, and outputs a pulse signal corresponding to the rotation angle of each motor 10, and the pulse signal is given to the detection circuit 13. The detection circuit 13 detects each motor 10 and thus the current position of each joint based on the pulse signal from each rotary encoder 17, and each motor 10 based on the number of pulse signals input from each rotary encoder 17 per unit time. As a result, the current speed of each joint is detected, and the information on the position and speed is given to the drive circuit 12 and the CPU 11 of each motor 10.

  Each drive circuit 12 compares the position command value and speed command value given from the CPU 11 with the current position and current speed given from the detection circuit 13, and supplies a current corresponding to the deviation to each motor 10. Drive them. As a result, the central portion of the flange 9 that is the tip (hand) of the robot arm moves along a route that sequentially passes through the command position, and performs various operations.

  The movement path of the robot arm tip is given by a teaching operation performed using the teaching pendant 3. In this teaching work, a plurality of positions on the trajectory to be followed by the robot arm tip are sequentially taught as command positions, the posture of the robot arm tip at each command position is taught, and the commanded position and posture taught are Stored in the RAM 15. Then, in actual robot work, the controller 2 sets a curve that smoothly interpolates between a plurality of given command positions and smoothly follows the command positions so that the tip of the robot arm moves on the curve. To control.

  Note that the position of the tip of the robot arm is indicated by the position of the origin of the three-dimensional coordinates (tool coordinates) fixed to the flange 9 on the robot coordinates (base coordinates). Further, the posture of the tip of the robot arm is defined by the orientation indicated by the unit vector on two predetermined axes among the three axes of the three-dimensional coordinates fixed to the flange 9 on the robot coordinates.

  The curve interpolation of the plurality of command positions is performed by a spline interpolation method. An example of this spline interpolation will be described with reference to FIG. 8, R0, R1,... Rn are (n + 1) command positions, R0 is a movement start position, and Rn is a movement end position. Curves G1, G2,... Gn connecting the command positions R0 to Rn are spline curve segments defined for each section with the command positions R0 to Rn as one section. In spline interpolation, a spline curve passing through a command position is defined by a spline segment function corresponding to the spline curve segment.

Each point on each spline curve segment can be determined by the following (1) to (3) as a point on a three-dimensional robot coordinate. Note that t is a parameter.
X (t) = a ₃ t ³ + a ₂ t ² + a ₁ t + a (1)
Y (t) = b ₃ t ³ + b ₂ t ² + b ₁ t + b (2)
Z (t) = c ₃ t ³ + c ₂ t ² + c ₁ t + c (3)
Summarizing the equations (1) to (3), the position of each point on the spline curve segment is defined by a cubic general equation (spline segment function) shown in the following equation (4).

よって、上記（４）式は、次の（９）式のようになる。

Therefore, the above equation (4) becomes the following equation (9).

以上のように指令位置相互間を一区画としたとき、各区画のスプライン曲線セグメントは上記（９）式の一般式によって表されるが、それら各区画のスプライン曲線セグメントを接続するための条件として、接線ベクトルＴ（ｋ−１）、Ｔｋを次の条件（ア）〜（ウ）により定める。

As described above, when the command positions are defined as one section, the spline curve segment of each section is expressed by the general formula of the above formula (9). As a condition for connecting the spline curve segments of each section, The tangent vectors T (k−1) and Tk are determined according to the following conditions (a) to (c).

(A) Connection points must match.
(B) The size of the tangent vector at both ends (start point side command position and end point side command position) of a section is determined to be proportional to the length of the straight line connecting both ends, and the direction of the tangent vector at both ends is The angle between the straight line connecting both ends and the straight line connecting both ends of the preceding one section and the straight line connecting both ends of the following one section is determined in a direction that bisects each.
(C) The direction of the start point side tangent vector determined as described above matches the direction of the end point side tangent vector of the previous section, and the direction of the end point side tangent vector matches the direction of the start point side tangent vector of the subsequent section. about.

  The conditions (a) to (c) will be specifically described focusing on the second section (k = 2; the section between the command positions R1 and R2) shown in FIG. FIG. 7 shows the command positions R0 to R3 in FIG. 8 in an enlarged manner. In FIG. 7, a straight line connecting both ends (command positions R0 and R1) of the first section (k = 1) is between L1 and between both ends (command positions R1 and R2) of the second section (k = 2). Let L2 be a straight line to be connected, and L3 be a straight line connecting both ends (command positions R2 and R3) of the third section (k = 3).

  Also, the tangent vectors at both ends R0 and R1 of the first section are T1s (start point side tangent vector) and T1e (end point side tangent vector), respectively, and the tangent vectors of both ends R1 and R2 of the second section are T2s (start point side). The tangent vectors), T2e (end point side tangent vectors), and the tangent vectors of both ends R2, R3 of the third section are T3s (start point side tangent vectors) and T3e (end point side tangent vectors), respectively.

  Then, the end point side tangent vector vector T1e of the first section, the start point side tangent vector T2s and the end point side tangent vector T2e of the second section, the start point side tangent vector T2s and the end point side tangent vector T2e of the third section are (10) to (13). Here, since the first section has no previous section, for the start point side tangent vector T1s, for example, the size is set to a value obtained by multiplying the length of the straight line L1 by the proportional coefficient r, and the direction is the same as the straight line L1. Determine.

上記（１０）式〜（１３）式において、ｒはＴ１ｓの大きさを定める際に使用したと同じ値の比例係数であり、本実施例では１と定めている。

In the above equations (10) to (13), r is a proportional coefficient having the same value as that used when determining the magnitude of T1s, and is set to 1 in this embodiment.

  Here, since P0 to P3 with arrows are position vectors of the command positions R0 to R3, in the expressions (10) and (11), the first expression in parentheses is a unit in a direction that coincides with the straight line L1. The vector and the second expression are used to obtain unit vectors in the direction matching the straight line L2. Similarly, in the expressions (12) and (13), the first expression in parentheses is the direction corresponding to the straight line L2. The second unit vector and the second equation are used to obtain unit vectors in the direction matching the straight line L3. Accordingly, in the equations (10) to (13), the difference between the angles formed by the straight lines L1 and L2 and the difference between the angles formed by the straight lines L2 and L3 are obtained by calculating the expressions in parentheses. .Multidot.5, the direction that bisects the angle formed by the straight lines L1 and L2, and the direction that bisects the angle formed by the straight lines L2 and L3 are obtained.

  In this way, the start point side tangent vector and the end point side tangent vector are obtained for each section. Then, after obtaining tangent vectors at both ends for each section, four spline parameters of a functional expression representing the spline curve segment of each section are obtained as described above. Then, it is possible to obtain a function in which the start point side tangent vector and the end point side tangent vector of each section match the direction to the end point side tangent vector of the previous section and the start point side tangent vector of the subsequent section, respectively. A smoothly continuous spline curve (spline function) can be obtained at the connection point (command position).

  Next, the operation when performing robot work in the above configuration will be described. Prior to performing the robot work, the robot main body 1 is actually moved using the teaching pendant 3, and the movement start position, the passing point and the movement end position of the tip of the robot arm are taught as command positions. The command position and orientation are stored in the RAM 15 (storage means).

  As shown in FIG. 6, when a start operation is performed in order to perform the robot work, the robot control device 2 first reads the command position from the RAM 15 (step S11), and both ends of the movement path, that is, the movement start position and A tangent vector at the movement end position is set (step S12). In this case, the tangent vector at the movement start position faces the same direction as the straight line connecting the movement start position and the next command position, and has a magnitude proportional to the length of the straight line up to the next command position. Also, the tangent vector at the movement end point is in the same direction as the straight line connecting the immediately preceding command position and the movement end point position, and is large in proportion to the length of the straight line connecting the immediately preceding command position and the movement end point position. Say it. Note that the direction and size of the tangent vector at the movement start position and the movement end position are not limited to this, and can be determined in consideration of previous and subsequent operations. good.

  Next, the control device 2 divides the movement start position to the movement end position into a plurality of sections at the command position, and the start-point side tangent vector and the end-point side tangent vector for each section according to the expressions (10) to (13) described above. Are calculated (step S13). And the control apparatus 2 calculates | requires the vector coefficient of the function formula (spline segment function) of the spline curve segment of each area by Formula (5)-(8) mentioned above (step S14).

  When the vector coefficient of the spline segment function for each section is obtained in this way, a spline curve connecting all the command positions from the movement start point position to the movement end point position with a smooth curve is obtained. 2 is to integrate the formula obtained by differentiating the formula (9) by t (for example, 0 ≦ t ≦ 1) from 0 to 1, using the formula (9) which is the spline segment function. Thus, the total curve distance from the movement start position to the movement end position is calculated (step S15).

  Since the obtained total curve distance is the movement distance from the movement start position to the movement end position of the tip of the robot arm, next, the control device 2 generates the speed pattern shown in FIG. 9A from the total curve distance. (Step S16). This speed pattern is determined so that the speed of the tip of the robot arm is trapezoidal, and the speed is increased until the tip of the robot arm reaches a predetermined maximum speed at a predetermined constant acceleration from the movement start position. When it moves and reaches a predetermined maximum speed, it moves at a constant speed at that maximum speed, and generates a speed pattern that decelerates at a predetermined deceleration to reach the movement end position.

  When the velocity pattern is generated, the control device 2 obtains the movement distance of the tip of the robot arm within a certain sampling time from the velocity pattern, and the movement distance within the current sampling time is determined as the movement end position within the previous sampling time. In addition to determining the movement end position on the spline curve at the current sampling time, the posture of the robot arm tip is calculated (step S17).

  Then, the control device 2 calculates the angular velocity and position of each arm so that the tip of the robot arm moves from the movement start position to the movement end position within the sampling time (step S18), and sends a speed command value to the motor 10 of each arm. The position command value is output (step S19). Then, the control device 2 moves the tip of the robot arm to the movement end position (step S20: YES) by repeating the above steps S17 to S19.

  Here, in this embodiment, when the robot body 1 is operated by the external TCP, the interpolation performed for the command position described above is similarly applied to the hand posture to perform the interpolation. Hereinafter, this posture interpolation processing will be described with reference to FIGS. FIG. 3 is a flowchart showing the posture interpolation process. As described above, when the teaching point is set for the position and orientation using the teaching pendant 3 (step S1), the teaching point (the tool coordinates as viewed from the base coordinates, The coordinate conversion is performed so that the position and orientation (fixed tool coordinates viewed from the tool coordinates) to the tool 23 viewed from the tool (hand) are obtained (see FIG. 10) (step S2).

  In addition, although the coordinate transformation determinant shown in FIG. 10 is an external TCP, the hand is referred to as a tool (t) and the fixed tool 23 is referred to as a work (w) in order to match the conventional TCP. The determinant for converting from the base coordinates to the tool coordinates is expressed by the product of the conversion matrix from the base coordinates to the fixed tool coordinates and the inverse matrix of the conversion matrix from the tool coordinates to the fixed tool coordinates. (The second equation on the right side is equivalent to a transformation matrix from fixed tool coordinates to tool coordinates).

  Next, each teaching posture at each teaching position is specified as a coordinate value. Since the taught posture is given as a rotation angle value (X direction rotation component, Y direction rotation component, Z direction rotation component), a spatial coordinate value (X direction posture coordinate value, Y direction posture coordinate) using a rotation matrix. Value, Z-direction posture coordinate value) (rotation matrix conversion processing, posture coordinate conversion processing means). A rotation matrix corresponding to the posture of the hand at each position is represented by the converted coordinates. That is, this rotation matrix becomes a three-dimensional coordinate value representing the normal vector (X-axis), the orientation vector (Y-axis), and the approach vector (Z-axis) that specify the hand posture on the base coordinates. Further, a combination of each direction coordinate value of each teaching position and each direction posture coordinate value at the teaching position corresponds to the movement target position.

  Then, a virtual point obtained by extending an arbitrary length in each of the X, Y, and Z axes with the teaching position as the origin is a posture point (amplified X direction posture coordinate value, amplified Y direction posture coordinate value, amplified Z direction). Posture coordinate value) (step S3, posture vector amplification processing means). That is, as shown in FIG. 2, if the teaching posture D is considered as a direction vector that stands with the teaching position (SC) as the origin, the three-dimensional coordinate values (SN, SO, SA) of the direction vector are determined. The teaching posture is specified. The “arbitrary length” may be any length that can specify the teaching posture, and is, for example, n (> 2) times the unit vector of the coordinate system.

  Here, the coordinate value indicating the posture represented by using the rotation matrix is a value close to zero. The calculation for performing the interpolation is performed by the CPU built in the control device 2. However, since the CPU has a calculation limit digit, when a value close to zero is used from the beginning, each part exceeding the calculation limit digit is used. There are values that affect the posture at the teaching point. However, since the subtle difference is swallowed as a calculation error in the calculation of the CPU, the interpolation value cannot be calculated so that the posture changes smoothly. Therefore, in order to avoid such a situation, the length of the vector is multiplied by n, and a value for calculating so as to smoothly change the posture within the calculation limit digit of the CPU is taken in.

  With the processing in step S3, three posture points are determined as three-dimensional coordinate values for one teaching posture, and the teaching posture can be interpolated by treating these points in the same way as the “position”. That is, trajectory processing is performed and interpolated so that the posture points of X, Y, and Z are each connected as a smooth trajectory by, for example, a spline curve like the teaching position described above (step S4, amplified posture coordinate value connection processing). means).

  Here, FIG. 1 conceptually shows a process of interpolating the teaching position and the teaching posture. In order to avoid complication of illustration, only the Y and Z axes are shown for the posture points. 1 (a) and 1 (b) show the case where each teaching position is subjected to curve interpolation, and FIG. 1 (c) shows the case where the Y-direction posture point and the Z-direction posture point are determined for the teaching posture at each teaching position. Yes, each arrow corresponds to an orientation vector and an approach vector. FIG. 1D shows the case where the Y-direction posture point is subjected to curve interpolation, and FIG. 1E shows the case where the Z-direction posture point is subjected to curve interpolation. Further, curve interpolation is similarly performed for an X direction (normal vector) posture point (not shown).

  In step S4, the posture points are independently interpolated in each of the X, Y, and Z directions. Since the interpolation points in these directions correspond to the respective axes of the three-dimensional coordinates, the relationship between them is an orthonormal condition. It is necessary to satisfy. Therefore, correction is performed so as to satisfy the condition (step S5). Thus, the teaching posture interpolation process is completed.

  Then, when the coordinates of the interpolation point obtained in step S5 are transformed so as to be the position / orientation of the tool viewed from the base (step S6), the hand trajectory (X, Y, Z each obtained by adding the interpolation point) The trajectory of the direction / posture point is converted into angles of the respective axes J1 to J6 (step S7). This completes the posture interpolation process.

  Here, FIG. 4 shows an approach vector teaching posture (SA) at a teaching point, and (b) applies a conventional interpolation technique such as Patent Documents 1 and 2, and the hand passes the teaching point. The approach vector in this case is shown three-dimensionally. In (b), there is a deviation with respect to the teaching posture, but in (a) when the posture interpolation technique of the present embodiment is applied, it can be seen that there is almost no deviation in posture.

  FIG. 5 shows the amount of deviation between the conventional interpolation technique and the interpolation technique of this embodiment for the teaching points of position and orientation. In addition, the workpiece | work 22 and the tool 23 shown in the figure are model things which do not respond | correspond to FIG. Regarding the positional deviation of (a), there is no significant difference between the conventional and the present embodiment, which is an error of less than 0.1 mm. On the other hand, with regard to the posture deviation of (b), the error angle may exceed 5 [deg] in the conventional technique, but in this embodiment, the error angle is 0.1 [deg] or less, which is greatly improved. Has been.

  As described above, according to the present embodiment, the control device 2 already teaches when teaching the position and posture of the hand with respect to the X, Y, and Z axes with the tip of the hand as the coordinate origin. X, Y at each movement target point with respect to a movement target point composed of a plurality of X components, Y components, Z components, X direction rotation components, Y direction rotation components, and Z direction rotation components, which are the movement destinations of the hand. , Z direction rotation components are subjected to rotation matrix conversion processing, and X, Y, and Z direction posture coordinate values coordinated by the X component, Y component, and Z component are calculated for each direction rotation component. Then, for each of the X, Y, and Z direction posture coordinate values that exist for each calculated movement target point, the unit vector is directed to each of the X, Y, and Z direction posture coordinate values with the movement target point as the origin. , The positions obtained by multiplying the vector by n are set as the amplified X, amplified Y, and amplified Z direction posture coordinate values, respectively. Then, with respect to the amplification X, amplification Y, and amplification Z direction posture coordinate values existing for each movement target point, only the amplification X, amplification Y, and amplification Z direction posture coordinate values are targeted, and the respective coordinate values are connected as a smooth trajectory. When the trajectory process is performed, the process of moving along the obtained trajectory with complementing the posture between the respective movement target points is performed.

  As a result, when the workpiece 22 is moved with respect to the fixed tool 23 by moving the hand of the robot main body 1, the position of the hand is maintained while the taught posture is maintained. Moving. Accordingly, the angle formed by the surface including the machining point of the workpiece 22 with respect to the axis of the fixed tool 23 is maintained as taught, so that the teaching point is set to improve the machining accuracy of the workpiece 22. Since there is no need to increase the number, teaching work can be simplified.

The present invention is not limited to the embodiments described above or shown in the drawings, but can be modified or expanded as follows.
The posture interpolation is not limited to using a spline curve, but may be performed so that the movement trajectories of the posture determination coordinates of each axis are connected smoothly (so as to be mathematically continuous).

  In the drawings, 1 is a robot body, 2 is a control device (posture coordinate conversion processing means, posture vector amplification processing means, amplified posture coordinate value connection processing means), 3 is a teaching pendant, and 9 is a flange (hand).

Claims (2)

When the tip of a vertical articulated robot grips the workpiece and moves the workpiece relative to a fixed tool to perform processing,
Regarding the movement target point composed of a plurality of X components, Y components, Z components, X direction rotation components, Y direction rotation components, and Z direction rotation components, which are the movement destinations of the hand already taught,
X-rotating components in each moving target point, Y-rotation component, the Z-rotating component, the rotation angle performed rotation matrix conversion process to convert the coordinate values, X component for each rotating component, Y component, Z In order to calculate a posture at a moving target point defined by a predetermined two-way component among the X-direction posture coordinate value, the Y-direction posture coordinate value, and the Z-direction posture coordinate value coordinated by the component, Posture coordinate conversion processing means for calculating coordinate values in one direction;
Each moving target point to the origin, the orientation coordinate conversion processing means by have been pre-SL two posture vector determined by the direction of the coordinate value to determine the orientation point is virtually stretched to any length calculated for each moving target point Posture vector amplification processing means for setting the coordinate values for the two directions at the virtual posture point as amplified first direction posture coordinate values and amplified second direction posture coordinate values, respectively;
For the amplified first direction posture coordinate value and the amplified second direction posture coordinate value existing for each moving target point, only the amplified first direction posture coordinate value is used as a target, and a trajectory process for connecting them as a smooth trajectory; Amplifying posture coordinate value connection processing means for performing a trajectory process for connecting only two-direction posture coordinate values as a smooth trajectory,
The interpolation point on the trajectory obtained by this amplified posture coordinate value connection processing means moves by interpolating the posture between each moving target point along the trajectory corrected so that the two directions satisfy the normal orthogonal condition. A robot control apparatus characterized by performing processing to perform When the tip of a vertical articulated robot grips the workpiece and moves the workpiece relative to a fixed tool to perform processing,
Regarding the movement target point composed of a plurality of X components, Y components, Z components, X direction rotation components, Y direction rotation components, and Z direction rotation components, which are the movement destinations of the hand already taught,
X-rotating components in each moving target point, Y-rotation component, the Z-rotating component, the rotation angle performed rotation matrix conversion process to convert the coordinate values, X component for each rotating component, Y component, Z In order to calculate a posture at a moving target point defined by a predetermined two-way component among the X-direction posture coordinate value, the Y-direction posture coordinate value, and the Z-direction posture coordinate value coordinated by the component, Calculate the coordinate values in one direction,
With each moving target point as the origin, the posture vector determined by the coordinate values in the two directions calculated for each moving target point is extended to an arbitrary length to determine a virtual posture point, and the virtual posture point at the virtual posture point is determined. The coordinate values for the two directions are the amplified first direction posture coordinate value and the amplified second direction posture coordinate value, respectively.
For the amplified first direction posture coordinate value and the amplified second direction posture coordinate value existing for each moving target point, only the amplified first direction posture coordinate value is used as a target, and a trajectory process for connecting them as a smooth trajectory; Perform trajectory processing that connects only two-direction posture coordinate values as a smooth trajectory,
The interpolated points on the connected trajectory are subjected to a process of moving along the trajectory corrected so that the two directions satisfy the orthonormal condition, complementing the posture between the moving target points. The robot's posture interpolation method.

Images