JPH05165509A - Routing method for deburring robot - Google Patents

Abstract

PURPOSE:To simplify the work by adding a working condition of a shape of a deburrign tool to shape information of parts to route for the robot. CONSTITUTION:Drawing information of parts is read out of a CAD system, and based thereon, each line element of a place of parts 6 in which a deburring work is executed is designated, and position information of a deburring tool 5 is generated. Accordingly, even if data of a tool number and a tool inclination angle, etc., are varied, the route of the corresponding robot 1 can be calculated immediately. Also, the designation is executed by each line element unit, and also, a shape and size of the deburring tool 5, a target chamber angle of the parts 6, etc., are added as working conditions to the drawing information of the parts 6, and obtained position information of the tool 5 is converted to a sequence of points. Accordingly, it will suffice that the working conditions such as the chamber angle, etc., are added at every line element, and it is unnecessary to add minutely the working conditions in each point converted to the sequence of points. In such a way, the work can be simplified.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a route generating method for a deburring robot, and more particularly to a deburring robot for generating a position / orientation to be taught to a deburring robot based on drawing information of parts prepared by a CAD system or the like. The present invention relates to a route generation method for a deburring robot and a deburring robot.

[0002]

2. Description of the Related Art A conventional robot teaching method includes direct teaching in which a human such as a playback robot actually moves the arm of the robot once or several times so as to learn the operation. Further, Japanese Patent Laid-Open No. 63-58505 and Japanese Patent Laid-Open No. 61-175775 disclose methods for generating a robot teaching position based on drawing information created by a CAD system. These are created offline from CAD data. The taught position is compared with the image actually seen by the camera, the positional deviation is detected, and the teaching position is determined by correction. Japanese Patent Laid-Open No. 63-273912 discloses a robot,
This is to create teaching data while simulating the operation of a robot from a workpiece, jig, etc., and finally convert it into a robot language. In addition, in welding robots that perform arc melting while vibrating the welding torch in the groove width direction and detect and correct the positional deviation of the welding torch during vibration, the method of making the welding torch follow the arc welding line is especially fair. -34086.

[0003]

In the robot teaching by the direct teaching of the above-mentioned prior art, the work for adding / correcting the teaching position is considerably complicated, and compared with the offline teaching, when the target parts are changed, It can be said that the influence on the production line in actual operation is great because the robot is taught each time.
In addition, regarding the technology that creates the robot teaching position based on the drawing information created by the CAD system, most of the actual target parts are still simple and uniform in shape, It can be said that it is difficult to accurately and universally create a teaching position of a part having a free curve in which a straight line (a size) and an arc are continuously formed. Thus, there is a need for a method of determining a robot path from drawing information having any free-form curve.

Further, in the case of creating a robot path for the purpose of removing burrs adhering to machined parts,
Depending on the material of the machined parts and the shape of the parts where the burrs are attached, the deburring tool such as a grinder provided in the robot must be replaced arbitrarily. Further, since the grinder becomes worn when a certain place is used for a long time, it is necessary to shift the position of the grinder with respect to the machined part and change the contact position between the two as necessary. In the case of such deburring work, the robot path naturally changes due to the change in the position of the grinder. Furthermore, in order to perform such burr removal,
The machined part is gripped and fixed to the jig for the work, but if the mounting position and the direction when gripping and fixing are changed, the robot path is changed. In such a case, there is a problem that it takes a lot of time to generate a new route again from the component drawing information.

An object of the present invention is, in order to simplify the teaching work for a robot, a method for generating a robot path based on drawing information having a free curve portion created by a CAD system or the like, particularly a machined part. Deburring robot that removes burrs adhering to the deburring robot, easily generates a desired robot path even with changes in the deburring tool shape, tool tilt angle, attachment position and orientation to the jig It is to provide a method of path generation.

Further, a deburring robot for generating a path for a deburring robot has a deburring robot which generates a desired robot path only by a simple operation by converting input data into a command to minimize the occurrence of erroneous operation. It is to provide a route generation method.

[0007]

In order to solve the above-mentioned problems, the drawing information of the part to be deburred is read from the drawing information of the part created by the CAD system, and the deburring work is performed from the read drawing information of the part. The part shape is decomposed into a plurality of straight line parts and circular arc parts by designating the part to be performed, and the shape and size of the deburring tool attached to the robot and the size of the chamfering angle of the part are added to the decomposed part shape information. Is added to the part shape information as a machining condition, and tool position information is created from the part shape information to which the machining condition is added. The straight line part of the tool position information is the start point and the end point, and the arc part is on the arc. It is decided to convert the point sequence into multiple points and determine the position and orientation of the robot from this point sequence information group.

Further, in order to reduce erroneous operation, which is cut with respect to the shape and size of the deburring tool attached to the robot, the size of the chamfering angle of the part, and the path based on the disassembled part shape information. Information (hereinafter referred to as cutting side), a virtual compliance parameter for force control is added to the part shape information by a command that can be specified as a processing condition, and tool position information is created from the part shape information to which the processing condition is added. The straight line part of the tool position information is made into a start point and an end point, the arc part is made into a point sequence into a plurality of points on the arc, and the robot operation information is added to each of the point sequence points as point attribute data. , Coordinate conversion information corresponding to the machining conditions specified by the command, chamfer angle,
Read detailed machining condition information such as virtual compliance parameters and compare it with the tool data in the tool file that stores the shape and size of the tool corresponding to the tool number. It was decided to determine the position and attitude.

Furthermore, between the points of the tool position information in the form of a point sequence, the coefficient of the three-dimensional curve is set to the inclination at two adjacent points,
It was calculated by using the coordinates of the two points and interpolated by the three-dimensional curve represented by the calculated coefficient.

[0010]

The drawing information of the part is read out from the CAD system, each line element of the place of the part where the deburring work is performed is specified based on this, and the tool position information is created. It is possible to immediately calculate the corresponding robot path even if the data of changes. In addition, each line element is specified, and the shape and size of the deburring tool, the target chamfering angle of the part, etc. are added to the drawing information of the part as processing conditions, and the obtained tool position information is converted into a point sequence. Therefore, it suffices to add a machining condition such as a chamfer angle for each line element, and it is not necessary to add a machining condition for each point arranged in a dot sequence.

The above-mentioned point shape of the part shape is divided into three points.
The interpolation from the next curve (y = ax ³ + bx ² + cx) makes it possible to expand from a point to a line, and
A free curve can be formed by the values of the coefficients (a, b, c) of the next curve. For example, when the cubic curve coefficient a = b = 0, it becomes a linear expression, which represents a straight line, and when a = 0, it represents an arc in a quadratic plane. Therefore, by forming the part shape into a point sequence and setting the coefficient of the cubic curve that interpolates between the points according to the shape of the part, a free curve can be generated without difficulty. Further, the definition of the jig surface coordinate system and the component surface coordinate system can grasp the relative position of the jig surface coordinate system viewed from the jig coordinate system and the relative position of the component surface coordinate system viewed from the component coordinate system. After all, parts side → parts → jig side →
Coordinate conversion that shows the relative relationship to the jig becomes easy, and if the same parts are attached to the A and B surfaces of one jig in the same way, the relative relationship from the jig surface to the jig That is, it suffices to change only the jig plane coordinate system, and it is possible to utilize the same data such as the data of the correction amount of the tool.

[0012]

Embodiments of the present invention will be described below with reference to FIGS. 1 to 28.

First, the target parts in the present invention will be described.
FIG. 2 is an example of a machined part having a free curve portion, and the ridge lines and holes of the part shown by 4 are parts where burrs are generated during machining. The present invention will be described in detail by following the method of creating the robot path data for removing the burr 1 in FIG. FIG. 1 is a schematic external view of a deburring robot and its robot controller, CRT, input device, and the like. The robot path data generated by the present invention is input to the robot controller 2 by the input device 3
Then, the controller 2 operates the robot 1 based on the input robot path data. When the CPU of the robot controller 2 receives various input signals such as a robot start signal, the CPU controls the robot according to the programs stored in the program memory. This system is equipped with a CRT so that the input contents and various output contents can be confirmed. The deburring tool 5 is sucked by air at the tip of the hand of the robot 1, and an arbitrary tool can be selected from several kinds of tools in a tool storage case (not shown). The machined part 6 is attached to a jig 7 which can be rotated 360 ° about the Z axis, and the jig origin and the robot coordinate system origin are accurately positioned relative to each other.

Next, a data creating system mechanism for creating a robot path will be described. FIG. 3 shows the overall configuration. FIG. 3 shows an overall configuration diagram thereof. A part drawing of a machined part is created in the CAD system 8 and stored in the drawing file 9.

The tool path generation program 11 specifies which part of the component 6 the tool (grinder, etc.) 5 deburrs based on the data of the component drawing stored in the drawing file 9. For this designation, the parts diagram is displayed on the CRT,
This is performed by designating each element or line element in the part drawing with a mouse or the like, and the tool path is divided into a plurality of straight line portions and arc portions. At this time, not only the deburring point of the part is specified, but also the moving method of the tool before and after deburring (movement of the robot) is set with reference to the graphic screen. Further, a command that can specify processing conditions such as coordinate conversion information, target chamfering angle, tool number, virtual compliance parameter (viscosity constant C, spring constant k, mass m) is added to the specified information, and is shown in FIG. CL (Cutter L
ocation) Create a file.

In this CL file, commands are described line by line, and the left part delimited by `/` represents the function of the command and is called the main part.

The right part is a predicate, which is a term for defining the function in detail. Since the coordinate conversion information, the virtual compliance parameter, the designation on the cutting side, and the target chamfer angle are additional information, the main part is set to "INSERT", and the predicate is "JIG" or "JFAC" for coordinate conversion information.
"E", "PFACE", and "PARTS NUM" are represented by "VCSET" for virtual compliance parameters, "CUTSIDE" for cutting side designation, and "CUTANGLE" for target chamfering angles. Here, the coordinate conversion information includes a jig coordinate system, a jig surface coordinate system, a component coordinate system,
Since it is necessary to define the component surface coordinate system, these are respectively the above "JIG", "JFACE", "PFACE",
It corresponds to “PARTS NUM”. Then, in order to define each coordinate system, a number associated with one body 1 is designated for each coordinate system. For example, jig surface coordinate system and jig surface number 1
If they correspond to each other, it may be described as INSERT / JFACE INSERT / 1 in the CL file. "INSERT / 1" on the second line is 1
This shows the content corresponding to "INSERT / JFACE" in the line. Similarly, a jig coordinate system, a component coordinate system, and a component coordinate system are defined. The command for the parts coordinate system is "PARTS NUM" because it is related to the input of the number of parts. In other words, when the number of parts is 3, it is assumed that the parts of the jig to which the three parts are attached are predetermined, and the number of parts and the coordinate system of each part are associated with one body. FIG. 4E shows a virtual compliance parameter for force control 1
It is represented by one group number, for example, group number 1 is designated. In (F) of FIG. 4, if the cutting side is on the right, 1 is specified, and if it is on the left side, 0 is specified in the next line of (F). In FIG. 4G, the target chamfering angle is 45 °.

The straight line part of the free curve for deburring is described as GOLFT (LINE) /72.104127, -40.0, -48.0 shown in FIG. The three numerical values represent coordinate values in the x, y, z directions from a certain origin in the drawing, and 72.104127, respectively in the x, y, z directions from the current position.
-40.0, -48.0 (unit is, for example, mm) It means to move linearly. The arc part is (B)
At (D), GOLFT (ARC) / 22.945260, -58.493145, -48.0 Center = (27.773075, -50.897606, -48.0) radius = 9.0, such as x = 22.9 from the current position to the target position.
Positions up to 4526, y = −58.493145, z = −48.0 (x = 27.773075, y = −50.8976)
(06, z = -48.0) is the center of movement, and it means to move on an arc circumference having a radius of 9.0.

Next, the robot motion determination program 13 creates a robot motion information file 14 based on the CL data 12.

In the robot motion determination program 13, if the location of the part specified by the tool path generation program 10 is a straight line portion, its start point and end point are automatically generated. It is stored in the robot motion information file 14. Further, when the sequence of points in the robot operation determination program 13 is different from the mere movement of the robot, force control needs to be performed in the actual deburring work, and therefore each point is in the process of cutting (CUT). Is stored as attribute data of points, and whether each point is a point on a circular arc is also stored as attribute data. FIG. 5 shows an example of the description.

The numerical value immediately after the CUT or MOVE indicates the number of points. In the case of the CUT, 0 and 1 in the next row mean whether the points are straight points or arc points.

In FIG. 5, the postures (Fx, Fy, Fz), (Gx, Gy, Gz), (H) are first moved to the positions (Px, Py, Pz) before the tool path reaches the deburring position.
x, Hy, Hz) in each posture.

In the robot motion information file 14, in addition to the data input to the robot motion determination program 13,
Coordinate conversion information input in the tool path generation program 10,
The target chamfer angle, tool shape to be used, size, tilt angle, tool number, virtual compliance parameter for force control, and jig mounting position are stored as data.

The robot operation language output program, based on the data in the robot operation information file, compares with the data in the tool information file, and corresponds to the tool used at the position of the part drawing which is the data in the CL file. Position correction is performed, and while the tool speed and the like are automatically determined, the coordinates of the data are further converted into the robot coordinate system and the data is output to the robot command file.

The above processing flow is shown in FIG.

First, in step 1, the CAD system 8 creates a part drawing of a machined part.

Next, in step 2, based on the data of the parts drawing stored in the drawing file 9, the tool path generation program 10 determines which part of the part 6 the tool (grinder, etc.) 5 deburrs. The drawing is displayed on the CRT, and each element or line element in the part drawing is designated by using a mouse or the like to create a tool path.

In step 3, in order to actually teach the robot 1 to perform the deburring work, information other than the tool path created from the drawing information of the parts is required. For example, when removing the burr of the component 6, the reaction force applied to the force sensor attached to the tool 5 of the robot 1 is detected, and the cutting force as the grinder torque, the feed speed of the grinder rotation, and the like are virtually controlled. The machining parameters such as the target chamfering angle, the shape of the tool, the size, and the tool number such that the information of the tool such as the waveform can be added to the tool path as the compliance parameters such as the viscosity constant C, the mass M, and the spring constant K.

In step 4, the CL file 10 is created based on the tool path created in step 2 and the processing condition information added in step 3.

In step 5, if the location of the part specified in step 2 is a straight line portion, its start point and end point are automatically generated. If it is a circular arc portion, a plurality of points on the circular arc are automatically generated, and the tool path is converted into a point sequence. A cubic curve coefficient for interpolating between columns is determined. This interpolation algorithm will be described later.

In step 6, at each point, whether the point is in cutting (CUT) or moving (MOVE) is distinguished and stored as point attribute data.

In step 7, in the processing up to step 6, the path of the tool is defined as the intersection of the part of the deburring tool 5 and the rotation axis of the tool (hereinafter, virtual center point O ₁ ) as shown in FIG. Represents the robot path data when it coincides with the deburring position. Therefore, depending on the size of the diameter of the tool 5 and the size of the tool tilt angle that is the deviation between the tilt angle of the tool 5 and the target chamfer angle. To correct the robot path. The tool path correction will be described later.

In step 8, each component is fixed on the surface of the jig 7 determined for each component, and the component surface of the component 6 (front view, right side view, etc.).
Cutting the burr. CL file 12 is a part to be cut 6
Since the position data is defined and created for each of the component surfaces, the position data of the certain component surface in the CL file 12 is viewed from the jig coordinate system in which the relative relationship with the robot coordinate system is accurate. Convert coordinates to data. This corresponds to the jig number and the jig coordinate system that corresponds to one body, the jig surface coordinate system that corresponds to one body for each surface of the jig, and one body for each surface of the component Data such as the attached component surface coordinate system is required. This will also be described later.

In this way, the file created by extracting the deburring position data of the parts and the position data of the tools before and after deburring from the CL file 12 and adding 11 the target chamfer angle, coordinate conversion data, etc. It will be called the information file 11. In the robot operation information file 14, CL data for each part surface of the part 6 to be machined is stored in the order of tool movement paths (cutting order).

Since the numbers corresponding to the coordinate system data of the jig, the jig surface and the component surface are added to the CL data for each component surface, the robot is controlled with respect to the CL data, particularly the position data. It is possible to perform coordinate conversion to a reference coordinate system that is the origin. This coordinate-converted CL file will now be called the robot command file 17. The actual robot 1 performs deburring by performing position correction control on the path position data thus created in real time.

In addition to this real-time position correction, the robot controller 2 has to perform many tasks such as sensor control and communication with the component transfer mechanism, and a lot of load is applied to the controller 2 for processing teaching data. Having it is not desirable. Therefore, when the robot command file 17 is created, all control commands to be sent to the robot 1 such as a tool replacement command, a movement command, and a deburring execution command should have a command system that runs on the control controller device 2 actually used. .. It should be noted that the processes from the CL file 12 to the creation of the robot command file 17 are not executed by one program, but the robot operation information file 14 and the robot command file 17 are created by executing different programs respectively. There is. This may be because the program size may become too large to handle a lot of data at one time, or if data correction occurs, the robot movement may be greater than the position data handled from the reference coordinate system (jig coordinate system). This is because it may be easier to correct the position data from the coordinate system set for each component surface like the information file 14, and in the future, additional data will increase as the types and materials of the target components become diversified. Therefore, it is necessary to flexibly deal with the corrections and improvements for the generalization of programs caused by this. If the program is divided in this way, there is an advantage that the program debugging at that time can be performed efficiently. However, even if data is created all at once up to the robot command file 17 without dividing the program in this way, it lacks versatility. It can be implemented in the same manner, except that no advantage is obtained.

Next, regarding the algorithm for determining the coefficient of the three-dimensional curve which stores the space between the point sequences in step 5,
explain.

[0038] To make a route into a point sequence means to determine position data. The position data mentioned here is shown in FIG.
As shown in FIG. 3, not only the positions (Px, Py, Pz) but also the postures (Fx, Fy, Fz), (Gx, Gy, Gz), (H
x, Hy, Hz).

First, when the straight line portion of the free curve is formed into a point sequence, the start point and the end point of the straight line are taken. The start point has the current position, and the end point has the coordinate value immediately after GOLFT (LINE) as the position data (Px, Py, Pz).

Teaching data for the robot includes the position F,
The posture information displayed by G and H is also required and the position is positioned at a predetermined place. Therefore, as for the posture, as shown in FIG. 7, the tangent line at that point, that is, the unit vector connecting the current position P1 and the target position P2 is defined as the posture F (F
x, Fy, Fx). In addition, as shown in FIG. 8, the Z direction of the coordinate system 21 in the component drawing 20 is always the posture H (H
x, Hy, Hz). Posture G (Gx, Gy, G
z) may be obtained from the outer product of H and F.

Next, a case where the circular arc portion is converted into a point sequence will be described. It is conceivable that the number of point sequences generated on the circular arc is determined depending on the central angle of the circular arc, the radius of the circular arc, the speed of the robot, and the machining accuracy, but in the present embodiment, as shown in FIG. A case will be described in which a passing point on an arc is automatically created according to the size of the central angle of the arc. The coordinate value of the equation (C) in the succeeding row of GOLFT (ARC) shown in FIG.
A point sequence is generated on the circular arc in order to move it to the position of the formula (B) through the circular arc having the radius as the value. Like the straight line mode
Even in the arc mode, the posture F is tangent to each point, and the posture H
Is defined as the Z-axis direction, that is, the paper surface direction. The tangent line at each point on the arc is obtained as follows. As shown in FIG. 10, it is necessary to match the tangents of the connection points P1 and P2 of the straight line and the circular arc. First, assuming that the tangent line f0 at the point P0 is Len0 ² = (P ₁ x−P ₀ x) ² + (P ₁ y−P ₀ y) ² , the component of the unit vector of the posture F on the straight line is f ₀ x = _{(P 1 x-P 0 x} ) / Len0 f 0 y = (P 1 y-P 0 y) / Len0 and since the tangent f1 of the point P1 as linear modes _{f 1 x = f 0 x f} 1 y = F ₀ y.

Next, at the point P1, the unit vector U1 directed to the center point of the arc is expressed as follows: u ₁ x = (Pcx−P ₁ x) / Len1 u ₁ y = (Pcy−P ₁ y) / Len1 where (Pcx , Pcy) represents the coordinates of the center point of the arc, and Len1 ² = (Pcx−P ₁ x) ² + (Pcy−P ₁ y) ² .

Then, consider the vector U1 ₊ and U1 ₋ obtained by rotating the unit vector U1 by 90 ° or −90 ° around the Z axis as shown in FIG. Where f
The U1 ₊ (U1 ₋ ) vector having the smaller value among the angle formed by 1 and U1 _{+ and} the angle formed by f1 and U1 ₋ is defined as the tangent line at the point P1. In this case, the tangent at point P1 is the vector U1 ₊ . This determines the direction of the tangent line. In order to make the tangent line defined by the straight line mode and the tangent line defined by the arc mode coincide at the connection point P1 of the straight line and the arc, the angle between the two vectors is strictly zero. Is.
By determining the tangential direction at the point P1 in this way, it becomes possible to determine the direction of the arc. The number of passing points on the circular arc is determined according to the central angle of the circular arc, the size of the radius of the circular arc, and the robot's operating speed finishing accuracy.

The central angle θ and the number of passing points are set as follows, for example.

[0045] In this case, the passing point is created by equally dividing the central angle of the arc. If the number of passing points is n, the central angle θ is divided into θ / n °. The position S1 (x,
y) is the following x = Pcx ± rsin (θ / n) (x−Pcx) ² + (y−Pcy) ² = r ² (where r is the radius of the arc) …………………… Although it can be calculated from the equation (1), there are four sets (points indicated by Δ) of the above-mentioned values of (x, y) as shown in FIG. This is because the above formula (1) is the center Pc of the circle (Pcx, P
Since cy) is used as the origin, a set of (x +, y +),
(X +, y-) pair, (x-, y +) pair, (x-, y
The coordinates of the pair of −) come out. Therefore, in order to find the division points along the direction of the arc, as shown in FIG. 12, the tangent line f1 of the point P1 previously found for determining the direction of the arc and the unit vector connecting these four (x, y) points P1 (X, y), which makes the smallest angle α among the angles α1 to α4 formed by, may be used as the position coordinates of the passing point. At this time, in order to remember which quadrant (x, y) is selected from the set of (x, y) coordinates, the correspondence as shown in FIG. 13 is made. The beginning and end of the arc are assumed to be Pn and Ps at the north and south pole positions, as shown in FIG. If the (x, y) that makes the smallest angle α is in the 0th quadrant of the circle, the direction counter is set to 0, and if it is in the 1st quadrant, the counter is set to 1 to the 3rd quadrant. Then, for the subsequent division points S2, S3 ...
Substituting (θ / n) * m (m = 2, 3 ...) Into, four (x, y) are calculated in the same manner. Where (θ / n) *
Depending on how many times m is, which one is selected from the four (x, y) is determined. For example, (θ / n) * m is 90 °
In the following case, the direction counter is the same as the counter obtained at the first passing point. However, if the angle exceeds 90 ° and is 180 ° or less, the direction counter is not the same as the counter obtained at the first passing point.

[0046]

[Table 1]

As shown in (1), when the first direction counter is 0 or 2, that is, the point is passed first, the direction of the arc is clockwise, so the counter increases, but When the first direction counter is 1 or 3, the counter is decremented because it is counterclockwise. If the incremented / decremented counter value is 0, the pair is (x +, y +), and if the counter value is 1, the pair is (x +, y-).
For the division points S2, S3, ... Which of the four (x, y) coordinate sets is selected depends on the central angle of the arc up to the division point. Below is the position of the first division point and (θ / n)
Table 1 shows a correspondence table determined according to the value of * m.

The direction counter is incremented when the arc direction is clockwise and decremented when it is counterclockwise. Therefore, the current position (arc starting point) is (X +, Y +), (X
+, Y-), (X-, Y-), (X +, Y-), determine the direction counter at that point, and determine the size of the center angle of the arc to the next division point. Let me decide. The counting method of the direction counter is performed in the robot motion determination program, the current counter value is stored in the memory, and the value is sequentially increased or decreased by ± 1 according to the arc center angle of the next division point. ing.

It should be noted here that in this method, the starting positions of the arcs are assumed to be the positions of the north pole and the south pole of the circle. As shown in the example of FIG. 14, the start of the arc (point P1 position) is an arbitrary place. Therefore, in reality, as shown in FIG. 14, it is calculated in the XY coordinate system (x, y).
The coordinate values must be converted so as to have the values in the X'-Y 'coordinate system. For that, X'- for the XY coordinate system
How many times the Y'coordinate system rotates about the Z axis (θ °) is calculated by the following formula, and the value of (x, y) in the XY coordinate system is rotated according to the calculated value. Good. This includes
For example, the direction cosine vector l, m, n of each coordinate axis of the X-Y-Z coordinate system is used, and the direction cosine vector of each coordinate axis of the X'-Y'-Z 'coordinate system is l', m ', n. ′, By rotating (l, m, n) by θ °, (l
′, M ′, n ′), 2θsΔθ = (l·l ′ + m · m ′ + n · n′−1) can be used to obtain Δθ.

As for the attitude of the passing point on the arc, the unit vector connecting the points P1 and S1 'is defined as the attitude (Fx, Fy, Fz) as shown in FIG. 15, and the attitude (Hx, Since Hy, Hz) is always (0, 0, 1), the posture (Gx, Gy, Gz) can be obtained.

The method of creating a point sequence on a circular arc from GOLFT (ARC) has been described in the case where the tool moves from the linear mode to the circular arc mode shown in FIG.
A case where the arc mode does not start from a straight line will be described. In the case of a ridge line shape in which arcs are continuous as shown in FIG. 16 (A) or as in the ridge line of a part as shown in FIG. 16 (B), there are holes inside and outside so that burrs must be removed. It corresponds to when. Even in such a case, the same idea is basically created, but the difference from the former case is that the tangent line with the straight line mode cannot be compared at the arc start point P1, and therefore the direction of the arc needs to be separately provided. .. For example, unless otherwise specified, the direction of the arc is counterclockwise, and in the case of FIG. 16B, the point P1 is started and moved to the point P2. In this case, the U _- vector obtained by rotating the unit vector connecting the center of the circle and the point P1 by -90 degrees becomes a tangent line at the point P1, and otherwise the same algorithm as in the arc passing point creation (point sequence) from the straight line mode. Good. Further, in the example shown in FIG. 17, even when the straight line and the circular arc are continuous, the tangents of the straight line and the circular arc do not match at the tangent point P1. In such a case, the straight line and the circular arc are not considered to be continuous, but are divided into three parts, that is, the straight line, the circular arc, and the straight line.

Now, when the position and orientation of the point-shaped component shape is determined in this manner, interpolation is then performed between the points 3
Set the secondary curve coefficient.

In the case of a straight line, the cubic curve is generally y = ax ³
+ Bx ² + cx + d, but as shown in FIG. 19, in the Tx-Ty coordinate system, the point (0, 0),
In a cubic curve connecting two points (1,0), y = ax ³ + b
It may be x ² + cx. At this time, the straight line connecting the two points is y = 0 in the Tx-Ty coordinate system.
= B = c = 0.

Further, a method of determining a coefficient for interpolating a point on a circular arc will be described.

As shown in FIG. 19, a cubic curve y = ax ³
+ Bx ² + cx is the origin (0, 0 in the coordinate system Tx-Ty.
0) and (1,0), and the origin (0,
The gradient of the curve in (0) is α, and the tangent vector of the curve in (1,0) is β. Cubic curve y = ax ³ + bx ² + c
Substituting x = 1 and y = 0 into x gives a + b + c = 0. Moreover, if x = 0 is substituted into α = 3ax ² + 2bx + c, then c = α and x = 0 is substituted into β = 3ax ² + 2bx + cx to obtain β = 3a + 2b + c.

From this, a = α + β b = −β−2α c = α.

Next, as shown in FIG. 20, consider a case where points P1 and P2 are interpolated by a cubic curve. The tangent vectors at the points P1 and P2 are f1 and f2 in the XY coordinate system. These two connection vectors are now X1
2-convert to a tangent vector in the Y12 coordinate system. That is, if the angle between the point P1 and the point P2 and the X axis of the X12-Y12 coordinate system and the X axis of the XY coordinate system is θ, then f1 '= Rf1 f2' for f1 and f2. = Rf2 However,

[0058]

[Equation 1]

Is a tangent vector viewed from the X12-Y12 coordinate system.

Therefore, f1 '= (f1'x, f1'y),
If f2 '= (f2'x, f2'y), then α = f
1′y / f1′x β = f2′y / f2′x and a = α + β b = −β−2α c = α, the cubic curve coefficient can be determined.

The rotation matrix around the Z axis is

[0062]

[Equation 2]

It is represented by

Next, the tool diameter correction (step 7) will be described. As shown in FIG. 18, the tool path in the robot operation information file 14 shown in FIG. 3 was such that the virtual center point 0 of the deburring tool 5 and the deburring position of the component 6 were the same. In order to actually perform the deburring work, this data must be corrected according to the tool shape and the target chamfering angle. Now, as shown in FIG. 21, a tool having different blade diameters, blade lengths, and blade angles d, L, and θ is shown. Also,
It is assumed that the intermediate point T of the cutting edge comes into contact with the component. Figure 2
Assuming that the area between P1 and P2 of 2 is a deburring point, first, + l / 2 in the Z direction and −d / 2 in the X direction are shifted. Further, as shown in FIG. 23, when the target chamfering angle is, for example, 45 °, it is necessary to tilt the Z axis toward the component side by δ = 45 ° −θ / 2. This is when the blade shape is conical, but when it becomes a cylinder as shown in FIG. 21 (B), the shift in the x direction becomes -d / 2. If the same part of the tool 5 is used for a long time, the tool 5 will start to wear, but this will lead to deterioration of the deburring finishing accuracy, and the tool 5 will not wear significantly, so the point of contact with the part is shifted from the T point. Need (offset move). To actually generate the robot path, the tool number,
From the robot motion information, the robot motion language output program receives the target chamfer angle, the tool offset amount, the left or right side of the tool with respect to the tool movement direction, etc.
By referring to the tool information file 18, the position corresponding to the tool condition cutting condition used immediately is corrected and a robot path is generated.

Position correction by this tool will be described in more detail.

Since the robot path position in the CL file and the robot motion information file created based on the CAD drawing information is the data of the edge (edge, ridgeline) itself of the part, the diameter d of the tool used and the bite angle are used. It is necessary to shift the route position data according to θ. As a precondition for this method, as shown in FIG. 24, the attitude h of the tool is set in the same direction as the Z axis of the component, and the tool advancing method f is set in the X axis. First, with respect to the orthogonal direction f of the tool, when the cutting side of the component is on the right side as shown in FIG. 24, the predetermined angle is −δ when the orientation of the tool is on the left around the f axis from the following equation.

G ′ = cos δg + (1−cos δ) (f · g) f h ′ = cos δh + (1−cos δ) (f · h) f Next, the above-mentioned machining at the tool position where the tool actually contacts the part The diameter is d '(when the blade shape is conical as described above, -d /
2), the tool position is p = p + sig * d '* g.

However, p is a tool, g is a posture g, g of h, sig is a constant of 1 if the cutting side is right and 0 if it is left.

If this offset amount is added to this, then p = p + sig * d '* g + Of. However, Of is an offset amount.

Next, the coordinate conversion (step 8) will be described.

The method of creating the motion position for the robot 1 from the data of the CL file 12 using the CAD drawing has been described. As described above, the CL data defines coordinates for each surface of the component, and the position data is a value in each coordinate system. It is necessary for the robot 1 to coordinate-convert the position data in each coordinate system into a value viewed from the same reference coordinate system. When the parts 6 are processed, the parts 6 are always attached to a fixed jig 7 for each part to perform the work. Therefore, the same reference coordinate system is assumed to be the coordinate system of each jig (jig coordinate system). The position data created by the robot operation determination file 14 shown in FIG. 3 is coordinate-converted into position data in the jig coordinate system and stored in the robot command file 17. Coordinate conversion from the component surface coordinate system to the jig coordinate system is performed in the order of component surface coordinate system ⇒ component coordinate system ⇒ jig surface coordinate system ⇒ jig coordinate system. FIG. 25 shows an example of a state where parts are fixed to a jig.

The component is attached to the "face" having the jig, and the burr on the "face" having the component is shaved. Therefore, a jig surface coordinate system is required for the jig coordinate system, and a component coordinate system viewed from each jig surface is necessary in order to grasp where each jig surface has a component. The tool path generation program adds a jig number, a jig surface number, and a component surface number corresponding to each coordinate system to CL data created for each component surface, and performs coordinate conversion based on this. The jig surface number and the component surface number may be set for each jig and each component as shown in FIG. 26, for example. Further, four identical parts are attached to a jig surface in FIG. Even if the mounting method is rotated by 180 ° as in this example, if all four identical parts are cut in the same place and in the same cutting order, how many parts are placed on the jig surface in the CL file 10. Therefore, it is sufficient to create CL data for one component by defining the four component coordinate systems (four in this case).

[0073]

According to the present invention, offline robot teaching can be performed without stopping the operating production line, and the work can be greatly simplified. In particular, according to the present invention, a robot path can be easily generated even for a part having a free curve portion. In addition, when removing burrs generated in machined parts, a path with accurate position correction is generated for any shape of the deburring tool, and even if the method of attaching to the jig is different, the same drawing path data can be used. Since it is possible to generate a robot motion language, it is possible to generate a high-speed route corresponding to changes in the deburring work environment.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a deburring robot.

FIG. 2 is a diagram showing an example of a machined part.

FIG. 3 is an overall configuration diagram of a robot route data creation system.

FIG. 4 is a diagram showing an example of a CL file.

FIG. 5 is a diagram showing an example of a robot motion information file.

FIG. 6 is a flowchart for creating a robot command.

FIG. 7 is an explanatory diagram of a posture in a straight line mode.

FIG. 8 is an explanatory diagram of a tool direction and a component coordinate system.

FIG. 9 is a diagram showing passing points on an arc.

FIG. 10 is a diagram showing an example in which a straight line and a circular arc are continuous.

FIG. 11 is an explanatory diagram of a tool direction and a component coordinate system.

FIG. 12 is an explanatory diagram of creating a first division point.

FIG. 13 is an explanatory diagram of setting of a direction counter.

FIG. 14 is a diagram showing positions of first division points in an X′-Y ′ coordinate system.

FIG. 15 is an explanatory diagram of a posture of an arc passing point.

FIG. 16 is a diagram showing an example in which a straight line mode is not continuous to an arc mode.

FIG. 17 is a diagram showing an example in which the tangent vectors from the straight line mode to the circular arc mode do not match.

FIG. 18 is a view showing a virtual center point of the deburring tool.

FIG. 19 is a diagram showing a cubic curve.

FIG. 20 is a diagram for obtaining a cubic curve interpolation coefficient.

FIG. 21 is a view showing an example of a deburring tool.

FIG. 22 is a diagram showing a shift amount of the deburring tool.

FIG. 23 is a view showing a tilt angle of a deburring tool.

FIG. 24 is an explanatory diagram of a shift amount of the deburring tool.

FIG. 25 is an explanatory diagram of a coordinate system.

FIG. 26 is a diagram showing an example of surface numbering.

FIG. 27 is a diagram showing an example of mounting a plurality of devices.

[Explanation of symbols]

1 ... Robot, 2 ... Robot control controller device, 3
... I / O device, 4 ... Burrs, 5 ... Deburring tool, 6 ... Machined parts, 7 ... Jig, 9 ... CAD output drawing file, 1
0 ... Tool path generation program, 11 ... Machining conditions, 12 ...
CL file, 13 ... Robot motion determination program, 1
4 ... Robot operation information file, 15 ... Robot operation language output program, 16 ... Coordinate conversion information file, 17
… Robot command file, 18… Tool information file

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Masao Miki 1070 Ige, Katsuta City, Ibaraki Prefecture Hitachi Ltd. Mito Plant

Claims (8)

[Claims] 1. A shape information of a part is read from drawing information of the part created by a CAD system, and processing conditions including at least the shape of a deburring tool attached to a robot are added to the read shape information of the part, A method for generating a path for a deburring robot, which is characterized by generating a path for a robot. 2. A shape of a part by designating a part to be deburred from the drawing information of the part to be deburred from the drawing information of the part created by the CAD system and reading the drawing information of the part to be read out. Is decomposed into a plurality of straight line parts and arc parts, and the decomposed part shape information is added to the part shape information as a processing condition including at least the shape of the deburring tool attached to the robot,
The straight line portion of the part shape information to which the processing conditions are added is made into a start point and an end point, and the arc portion is made into a point sequence into a plurality of points on the arc, and tool position information is created from this point sequence information to make the robot position. And a posture are determined, a path generation method for a deburring robot. 3. The number of points dividing the arc is set according to the size of the center angle of the arc portion, and four points on the circumference calculated from the divided center angles are passed first. The first arc division point position is determined by calculating the included angle between each tangent vector at the four points on the circumference and the tangent vector at the arc start point, and the first division point position is determined from the first division point position. 3. The deburring robot path generation method according to claim 2, wherein the positions of the second and subsequent arc division points are determined according to the size of the arc center angle. 4. The robot operation information is added as point attribute data at each point of the point-sequenced part shape information, and a final robot operation code is collectively created. The robot path generation method according to claim 2 or 3. 5. A coordinate system of the component, a coordinate system of a jig for gripping the component, a component surface coordinate system of each surface of the component, and a jig surface coordinate system of the jig surface of the jig. A coordinate conversion from each component surface coordinate system to a robot reference coordinate system is performed by defining each and adding a number corresponding to each coordinate system to the component shape information. A method for generating a path for the deburring robot described. 6. A jig is mounted by mounting a plurality of parts of the same shape, adding the number of parts of the part to the shape information, and converting the shape data of the one part into coordinates. The method for generating a path for a deburring robot according to claim 2 or 3, wherein the method is applied to the path generation for all other parts. 7. A part shape is made into a plurality of straight line parts by reading out drawing information of the part from drawing information of the part created by a CAD system and designating a part to be deburred from the drawing information of the part read out. The command is added to the part shape information by disassembling it into a circular arc part and incorporating a command capable of specifying the processing condition into the decomposed part shape information, and the straight line part of the part shape information to which the processing condition is added is the start point. At the end point, the arc part is made into a sequence of points on a plurality of points on the arc to create tool position information, robot motion information is added to the tool position information as attribute data of the point, and specified by the command. Coordinate conversion information corresponding to machining conditions, chamfer angle, detailed compliance condition information of virtual compliance parameters are read out, and a small number corresponding to the tool number is read. In addition, the deburring robot is characterized in that the position and orientation of the robot are determined by comparing the tool shape and size with the tool data in a tool file that stores the coordinates and performing coordinate conversion from the coordinate conversion information to the robot coordinate system. Route generation method. 8. Between the points of the tool sequence information in the sequence of points,
The coefficient of a three-dimensional curve is calculated from the inclination of a tangent line to the part shape at two adjacent points and the coordinates of the two points, and the interpolation is performed by the three-dimensional curve represented by the calculated coefficient. 8. A method for generating a route for a deburring robot according to any one of 2 to 7.