JPS60136805A - Robot control device - Google Patents

Abstract

PURPOSE:To attain speedy operation by reproducing signals smaller than the number of coordinate value signals provided at the time of instruction at a simple part in the course of a robot. CONSTITUTION:At the time of instruction, coordinate value signals obtained when an instructor moves an arm of a robot are successively stored and the data of the coordinate value signals are compressed and stored as driving coordinate signals. At the time of execution, the arm of the robot is moved in accordance with the driving coordinate value signals successively read out from a data storage part 103 to reproduce prescribed work. As to the data compression, the compressibility is increased when the arm of the robot moves on a straight or a moderately curved course, and reduced at the movement on a sharply curved course. Thus, the required storage capacity is reduced while maintaining the accuracy of reproduction to speed up the operation at the execution time.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical field to which the invention pertains] The present invention relates to a robot control device, and particularly to a robot control device for causing a robot (including a robot manipulator) to perform work.

[Prior art]

The most important method when considering the execution form of robot work is the so-called direct teaching-reproduction form.

In this case, the instructor moves the robot's arm during teaching to make the robot perform a desired task. At this time, the sensor included in the robot generates a sensor output according to the position of the robot's arm, and according to this sensor output, a coordinate value signal corresponding to the position of the robot's arm is generated, and this is The information is sequentially stored in the data storage unit.

During execution, the robot's arm strength is increased by 41 degrees in accordance with the coordinate value signals sequentially read out from the data storage section, and the intended work is reproduced.

Conventional robot control devices include an internal world measuring means that sequentially samples the position of the robot's arm during teaching and outputs it as a coordinate value signal, and a coordinate emotion holding device that sequentially holds the coordinate value signal as a driving coordinate value signal. means, drive information calculation means for sequentially reading out the drive coordinate value signal during execution and calculating drive b data for each degree of freedom of the robot, and drive means for fully driving the arm of the robot based on the drive data. It consists of:

Next, a conventional robot control device will be described in detail with reference to the drawings.

FIG. 1 is a block diagram including an example of a conventional robot control device.

In the robot control device shown in FIG. 1, the instructor moves the arms of the robot body 101 during teaching to perform a desired task, and at each predetermined sampling time, each arm of the robot body 101 can be freely moved. The spatial position data of the arm of the robot main body 101 is measured by the internal world measurement unit 102 from the center output 110 outputted from the sensor that detects the amount of displacement for each degree, and the coordinate value signal q of the position data is outputted; It is held in the data storage unit 103.

During execution, the coordinate value signals q held in the data storage unit 103 are sequentially read out, the drive information calculation unit 104 calculates the drive unit @j112 for each degree of freedom of the robot, and the drive unit 1
05 generates a driving force 113 to drive the robot body 101 and reproduce the movement during teaching.

Here, the coordinate value signal q is a three-dimensional position vector signal.

FIG. 2 is a two-dimensional display diagram of the coordinate value signal q held in the data storage unit 103 using such a teaching method. However, since there is no need to limit the x, y, and z components of the position data indicated by each coordinate axis and each coordinate value signal here, their display is omitted.

The display diagram shown in FIG. 2 is an example of a point sequence obtained when an instructor moves the robot's arm at a constant speed on a parabolic path and samples are taken at a constant sampling interval. The plurality of constituent points 201 are points on a route to be sampled, and this route includes a portion with a large bend such as a bend 202 and a portion with a gentle bend such as a gentle bend 203. do. In the case of such a route, the robot control device shown in FIG. 1 must sample the entirety at sufficiently fine sampling times so that the large part of the curve can be sufficiently reproduced during playback.

Therefore, the data storage unit 103 must hold a large amount of coordinate value signals in order to accurately reproduce the teaching content during execution, and the drive information calculation unit 104 and drive unit 105 must hold a large amount of coordinate value signals.
must be capable of repeatedly reading and processing these coordinate value signals at high speed.

On the other hand, in the teaching content, that is, "the spatial position of the robot's arm at each specified sampling time," what is essentially important is the spatial connection of the positions, in other words, the path, and the sampling time. The relationship with the teaching content is secondary, and therefore, during reproduction, there is no problem even if the position is reproduced in relation to a time different from that of the teaching content.

Not only in the case of robots, but in general, in the process of teaching or learning, it is reasonable to take different measures depending on the difficulty of the content, but if we apply this idea to the movement of a robot's arm, Complex parts of the route, such as the bending part 202 in FIG. 2, are relatively gentle, while simple parts, such as the gentle bending part 203 in FIG. This means that the gentle parts of the curve should be moved quickly. From this point of view, in the conventional teaching-reproduction mode, which reproduces the taught movement as it is, the internal world measurement unit
All the points sampled by 02 are stored in the data storage unit 103.
It was not necessarily rational in that it did not have any data compression function or data structure depending on the importance of the point.

In other words, in conventional robot control devices, the number of coordinate value signals to be stored is the same as the number of sampled constituent points, so a data storage unit with a large storage capacity is required, and it is difficult to read out the data at the time of execution. Since the coordinate value signal obtained is the same as the number of sampled constituent points, the operation speed is slow and the working time is long.

[Purpose of the invention]

An object of the present invention is to provide a robot control device that can reduce the storage capacity of a data storage unit and shorten working time.

In other words, an object of the present invention is to reduce unnecessary portions of the robot's arm movement path that are simple depending on the complexity of the robot's arm motion path while keeping the outline of the path. By omitting certain constituent points, reducing the need for data storage for playback, and moving the robot's arm quickly on simple parts of the path and carefully on complex parts, the L movement can be made faster. It is an object of the present invention to provide a robot control device that can simultaneously realize the functions of robot control and smooth operation of the robot.

[Structure of the invention]

The robot control device of the present invention includes: internal world measuring means for sequentially sampling the position of the robot's arm during teaching and outputting it as a coordinate value signal; and temporary coordinate emotion holding means for sequentially holding the coordinate value signal. The designated one of the coordinate value signals is used as an evaluation starting point coordinate value signal, and the designated one of the coordinate value signals held following this evaluation starting point coordinate value signal is used as an evaluation point coordinate value signal. is read out from the temporary coordinate emotion holding means, and a differential operation is performed based on the evaluation start point coordinate value signal and the two coordinate value signals held immediately thereafter, and the four arithmetic square root operations are further performed to obtain the evaluation start point unit tangent vector. A unit tangent vector that generates a signal, performs a differential operation based on the evaluation point coordinate value signal and the two coordinate value signals held immediately after it, and further performs the four arithmetic square root operations to generate an evaluation point unit tangent vector signal. a calculation means, an angle calculation means for calculating the absolute value of the difference between the evaluation start point unit tangent vector signal and the evaluation point unit tangent vector signal and outputting it as an angle signal, and each time the evaluation point coordinate value signal is specified. a distance calculating means for calculating the absolute value of the difference between this evaluation point coordinate value signal and the immediately preceding coordinate value signal, cumulatively calculating the absolute value, and outputting the result as a distance signal; and a product of the angle signal and the distance signal. ? evaluation function calculation means for outputting the evaluation function signal as an evaluation function signal; threshold determination means for comparing the evaluation function signal with a predetermined threshold signal and outputting a determination signal; coordinate emotion holding means for holding the evaluation point coordinate value signal as a driving coordinate value signal when the evaluation point coordinate value signal indicates that the evaluation function signal is larger; The coordinates held immediately after the evaluation point coordinate value signal when the point coordinate value signal is designated as a new evaluation start point coordinate value signal and the determination signal indicates that the evaluation function signal is not larger. processing control means for specifying a value signal as a new evaluation point coordinate value signal; drive information calculation means for calculating drive data for each degree of freedom of the robot by sequentially reading out the drive coordinate value signal during execution; and a driving means for driving the arm of the robot based on data.

That is, when the path along which the robot's arm should move is taught in the form of a data string consisting of the coordinate values of constituent points arranged in the order of the path, the robot control device of the present invention calculates the complexity of the path as a curve. It is detected in the form of the product of the angle between the tangents at two points on the curve and the length of the path between those two points,
The size of the detected product is compared with a pre-given threshold value, and points with a product exceeding the threshold are set as constituent points of a new route, thereby compressing data and controlling the robot's arm. Constructed to increase speed and smooth operation.

[Principle of the invention]

The robot control device of the present invention sequentially holds coordinate value signals obtained when a teacher moves the robot's arm during teaching, performs data compression on these coordinate value signals, and stores them as driving coordinate value signals. During execution, the robot's arm moves in accordance with drive coordinate value signals sequentially read out from this data storage section, and the intended work is reproduced.

Here, data compression increases the compression ratio when the robot's arm moves along a straight or gently curved path, and decreases the compression ratio when the robot's arm moves along a curved path with a large curve. By doing this, we can reduce the amount of memory required for the data storage unit while maintaining reproducibility accuracy.
Moreover, it allows for faster execution.

[Explanation of Examples]

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

FIG. 3 is a block diagram showing one embodiment of the present invention.

The temporary data storage section 301 temporarily holds the coordinate value signal q from the internal world measuring section 102, and can refer to it as needed in each section.

This temporary data storage section 301 does not have to hold all the coordinate value signals, but the data that needs to be held is generated by the circuits below the differentiating circuits 303 and 304 operating in real time at the same time as the teaching. Since the operation is to sequentially refer to the evaluation start point, evaluation point, one point before it, and two points after it, the coordinate value signal q that must be held at the same time corresponds to the four points mentioned above. It is sufficient that only the evaluation point is given, and as the teaching progresses, the evaluation starting point and evaluation points are changed accordingly.

The coordinate value signal distribution section 302 selectively passes the coordinate value signal q to either the differentiating circuit 303 or 304 based on the identification signal 351 from the control section 308 . Further, when the supply of the last data to be processed by the processing sections below the differentiating circuits 303 and 304 is completed, an end signal 353 is supplied to the control section 368, and in response to this, the control section 308 ends the processing.

FIG. 4 is a principle explanatory diagram for explaining the operating principle of the unit tangent vector calculation section and the distance calculation section shown in FIG. It is a principle explanatory diagram for explaining the operation to be evaluated.

The path 401 is a curved line representing the path of the robot's arm when taught, and intersects with a short line.4. ■ Nariten Rin, I1g+1. Blow +2. ~． This is the point where n0p, .

However, n and p are subscripts that indicate the order of the points. The coordinate value signal qn of the evaluation starting point Qn has x, y, and z components of 1 for xn, yn, and zn, respectively, as shown in the following fl) formula. is expressed as a vertical vector.

The count values n and p are counted by a counter in the coordinate value signal distribution section 302. When the identification signal 351 generated by the control unit 308 selects the coordinate value signal q as the evaluation point coordinate value signal, the Hiro count value pe'i' is incremented, and when it is selected as the evaluation start point coordinate value signal, it increases the count value n of the counter. n+
After setting it to p, the count value p of the counter is set to Jl. The details will be described later.

The product of the angle between the evaluation start point and the tangent to the route at the evaluation point and the length along the curve between the evaluation start point and the evaluation point is employed as the evaluation function. A signal obtained by multiplying the unit vectors touching the path 401 at the evaluation start point Qn and the evaluation point Qn+p by n unit vector calculation units is the nest position tangent vector signal tn + 'n+p. Further, the angle formed by the unit tangent vectors Tn and Tn+p is the angle ΔQ, the distance along the path 401 between the evaluation starting point Q11 and the evaluation point Qn+p is the distance ΔS, and the signals corresponding to the angle ΔO and the distance ΔS are respectively An angle signal Δθ and a distance signal ΔS are assumed. Then,
The above evaluation function signal J is given by the following equation (21).

J∆S∆θ (2) First, the coordinate value signal q that generates the first constituent point unconditionally passes through the coordinate value signal distribution unit 302 and becomes the evaluation starting point coordinate value signal insect, It is supplied to the differentiation circuit 303. The following coordinate value signal distribution unit 3 receives the identification signal 351 from the control unit 308.
The coordinate value signal q that has passed through 02 as the evaluation starting point coordinate value signal qn is guided to the differentiation circuit 303. Differential circuit 30
3 differentiates the evaluation starting point coordinate value signal qn with respect to time and outputs a speed signal Vi shown in the following equation (3).

However, since it is a sample value system, the differentiation circuit 303 performs numerical differentiation,
For example, it operates based on equation (4).

The differentiating circuit 303 can refer to the coordinate value signal q held in the temporary data storage unit 301 through the coordinate value signal distribution unit 302 as the coordinate value signal q held immediately after the evaluation start point coordinate value signal qn.

Inputting the velocity signal vn outputted from the differentiating circuit 303, the unit vector calculating section 305 calculates the evaluation starting point unit tangent vector signal tn at the evaluation starting point Ql based on equation (5) and outputs the calculated value.

The squarer 321 calculates each X, y, z of the speed signal vn.
The square of the component 341 (xn2, ′y,, 2,
"7.2)", and the adder 323 outputs this square multiplier 3
The sum signal 343 (xn"child piece +", 2
) sound output and this sum signal 343 is supplied to the square square device 3.
25 is its square root signal 345 (ffi〒:t,” +;,
”)'e small output, the divider 327 takes the quotient of the velocity signal vn and the square root signal 345 to obtain a unit tangent vector signal t.
Let it be n.

Here, for the evaluation point Qn+p of the evaluation starting point Q, yυp fish, the identification from the control unit 308 is performed in exactly the same way as the evaluation starting point unit tangent vector signal 1n is output for the evaluation starting point Qn. The coordinate value signal q supplied to the coordinate value signal distribution unit 302 based on the signal 351 is converted into the evaluation point coordinate value signal q.
n + p is supplied to the differentiation circuit 304 and processed in the unit vector calculation unit 306 in the same manner as the unit vector calculation unit 305, and the evaluation point unit tangent vector signal at the evaluation point Qn+, FIG. 5, is the angle shown in FIG. 3. This is an operation explanatory diagram for explaining the operation of the calculation unit, which calculates the angle ΔO formed by the unit tangent vector Tn (!: Tn+), which is composed of the two unit tangent vector signals output from the circuit described above. It is an operation explanatory diagram for explaining the operation of the circuit.The angle calculation unit 391 inputs the evaluation start point unit tangent vector signal t and the evaluation point unit machine vector signal tn+p, and performs the operation shown in equation (8). In other words, the angle Δθ formed by both unit tangent vectors Tnr and Tn+P is output as an angle signal Δθ.
As is clear from FIG. 5, 0 satisfies the following equation (6).

Here, in a range where the angle ΔO is sufficiently small, the following equation (7) holds true.

l Tn+p -Tr+ l'::△■(△@: o
)------(7) Therefore, the following equation (8) holds true for the evaluation start point unit tangent vector signal tn+p, the evaluation point unit tangent vector signal tゎ, and the angle □□□ signal △θ. do.

Therefore, the evaluation start point unit tangent vector signal tn input to the difference unit 307 and the evaluation point unit tangent vector signal "n+
A difference signal tn+p-tn is output from p and is used as an absolute signal.

Next, the operation of the distance calculation section 310 will be explained with reference to FIG. 4. Evaluation point Qn+ from evaluation starting point Qn.

The length ΔS along the route 4,01 leading to is
Constituent points located between the evaluation starting point Qn and the evaluation point Q n +p are sequentially approximated by the length of the connecting broken line,
Referring to the coordinate value signals q corresponding to the evaluation point Qn+ and the component point Qn+p-1 immediately before it in the order of p, the distance calculation unit 310 calculates the distance signal between the two based on equation (9), and sequentially The sum is accumulated and the distance signal ΔS is obtained.

The supplied subtractor 329 outputs a difference vector signal 347 ((i+, +t-s Qn+1) of each X, y, and z component,
The absolute f1σ calculator 330 is supplied with the difference vector signal 347. Bar'F: (D sum of squares ei-1
(rXn+1)2+(yn++-1Yn++a+(
zn+ 1-1-21+1)2) is output, and the accumulator 3
31 outputs the cumulative sum from i==1 to p as a distance signal ΔS.

Bird = 1 + CZn + t-1zn-H) 2 ... Cost 1]) Since the distance calculation unit 310 calculates by cumulative calculation, it is sufficient to refer to only one evaluation point and one component point immediately before it at a time. It is.

The angle signal output from the absolute value calculation unit 312 and the distance signal ΔS output from the distance calculation unit 310 are (2)
As shown in the equation, the evaluation function calculation unit 311 calculates the product of the two and outputs it as an evaluation function signal J.

If the evaluation function signal J does not exceed the threshold value signal THRE, the comparator 314 outputs a determination signal 354 indicating "Z" to the control section 308, and upon receiving this, the control section 308 generates an identification signal 351 and Sends a command to the value signal distribution unit 302 and calculates the count value pt- of the count in the coordinate value signal distribution unit 302.
The coordinate value signal q is read out and guided to the differentiation circuit 304, and the above-described processing is performed again. The evaluation function signal J is the threshold signal T)IR
If it is greater than or equal to E, the comparator 314 outputs a determination signal 354 indicating "large", and upon receiving this, the control unit 308 generates an identification signal 351 and sends a command to the coordinate value signal distribution unit 302. The count value of the counter is set to nen+p, and the count value p of the counter is set to 1, and the process is restarted from a new evaluation starting point. At the same time, the control unit 308 outputs a selection signal 352.
is set to 1ON1, the data switch 309 is closed, and the evaluation point coordinate value signal Qn+p is stored in the data storage section 103 as a driving point coordinate value signal.

A coordinate value signal corresponding to the point sequence shown in FIG. 2 is input to the temporary data storage unit 301 in FIG. An example of a glass displayed two-dimensionally is shown in Part 6.
As shown in the figure.

Constituent points 601 correspond to drive coordinate value signals for the new route remaining after data compression. Second
A bent part 602 corresponding to a large curved part of the hail curve of the original data shown in the figure is sampled finely.
A gently bent portion 603 corresponding to a gently curved portion is sampled coarsely, and the shape of the curve is sufficiently maintained even if the number of data is reduced.

In this way, the hail curve of the original data shown in Fig. 2 can be sufficiently reproduced even from the constituent points 601 according to the drive coordinate value signal which is compressed data, and since the number of constituent points is reduced, the data It is also clear that the amount of memory can be reduced.

In the above embodiment, an example was shown in which a part of the data storage unit was used as the coordinate emotion storage unit and a data storage unit was used as the coordinate emotion storage unit, but it is also possible to use parts of the same unit. .

Further, in the above embodiment, an example was described in which the data before processing is arranged at a constant sampling interval, but in other cases, that is, when the data string before processing includes the sampling time (x, , yn, zn , 1'.〉; Tn is given in the form of the nth sampling time, and even if the interval of Tn is not constant,
Equation (3) may be changed to Equation 11 below and is within the scope of the present invention.

Note that each part of the configuration example can be realized by normal means, and therefore will not be described in detail here. The scope of the present invention also includes devices in which the number or arrangement of each part and circuit is different due to considerations such as device cost or processing speed. Also, the formulas being explained,
The data structures, algorithm configuration examples, etc. are all provided for the purpose of clearly explaining the concept of the present invention, and are not intended to limit the present invention.

〔Effect of the invention〕

The robot control device of the present invention allows the robot to operate quickly in a simple part of the robot's path because the number of driving coordinate value signals read during movement reproduction can be reduced compared to the number of coordinate value signals obtained during teaching. In addition, the number of driving coordinate value signals for complex parts can be reduced compared to simple parts, so the robot's arm can be moved carefully! Instead of holding all the coordinate value signals corresponding to the sampled constituent points, the drive coordinate value signal can be held as data compressed and a much smaller amount of data, which reduces the amount of data stored. It has the advantage that it can be manufactured at low cost.

[Brief explanation of the drawing]

FIG. 1 is a block diagram including a conventional example, FIG. 2 is a display diagram two-dimensionally displaying the coordinate value signal held in the data storage unit shown in FIG. 1, and FIG. 3 is an embodiment of the present invention. FIG. 4 is a diagram illustrating the dynamics of the unit tangent vector calculation unit and distance calculation unit shown in FIG. FIG. 2 is a two-dimensional display diagram. 101...Robot body, 102...
Internal world measurement unit, 1-03...Data storage unit, 10
4... Drive information calculation unit, 105... Drive unit, 110... Sensor output, 112...
...Driving information, 113...Driving force, 120.
...Robot control device, 201...Construction point, 202...Bending part, 203...Gentle bending part, 301...Temporary data storage Department, 302
...Coordinate value signal distribution section, 303, 304...
...Differential circuit, 305, 306...Unit vector calculation section, 307...Differentiator, 308...
...Control unit, 309...Data switch, 3
10...Distance calculation unit, 311...Evaluation function calculation unit, 312...Absolute value calculation unit, 313
...Threshold input section, 314...Comparator,
321, 322...Squarer, 323°324.
... Adder, 325.326 ... Square rooter, 327.328 ... Divider, 329 ...
...Differentiator, 330...Absolute value calculator, 331
... Accumulator, 390 ... Unit tangent vector calculation section, 391 ... Corner layer calculation section, 392 ...
...Processing control section, 393... Interval value judgment section,
341.342... Square signal, 343,344
...sum signal, 345,346...square root signal, 347...difference vector signal, 348...
・! ...Sum of square root signal, 351...Identification signal, 352...Selection signal, 353...End signal, 354...Judgment signal, Qn...・・・・・・
・Evaluation starting point, q... Coordinate value signal, qn...
−... Evaluation starting point coordinate value signal % Qn+p...
・Evaluation point, qn+p...Evaluation point coordinate value signal, v
n...Evaluation starting point speed signal% vn+p...
...Evaluation point speed signal, T1Tn+p ...
Unit tangent vector, tn...Evaluation start point unit tangent vector signal s''n+p...Evaluation point unit tangent vector signal, ΔS...Distance, △S...・Distance signal, ΔO...Angle, Δθ...Angle signal, J...Evaluation function signal %TI-IRE...
...Threshold signal, 4o1/old...route, 6o1...
... Constituent point, 602 ... Bend part, 603 ...
...gentle bending part. 111 noーーー＼ 〜１１１トー) Cause Z′ , − Bead ′°°・,. ε ゛・・ ■ `` ■ 1 喀---1 Kari ■ Figure 1 ■ Tsu 1 Furrow 1 4 Figure 1 Figure 5 Kotobuki 6

Claims (1)

[Claims] internal world measuring means that sequentially samples the position of the robot's arm during teaching and outputs it as a coordinate value signal; a temporary coordinate value string holding means that sequentially holds the coordinate value signal; and an instruction of the coordinate value signal. One of the coordinate value signals held following this evaluation start point coordinate value signal is designated as an evaluation point coordinate value signal, and one designated 7'C is used as an evaluation point coordinate value signal to form the temporary coordinate value string. While reading from the holding means, a differential operation is performed based on the evaluation start point coordinate value signal and the two coordinate value signals held immediately thereafter, and further arithmetic square root operation is performed to generate an evaluation start point unit tangent vector signal. unit tangent vector calculating means for performing a differential operation based on the evaluation point coordinate value signal and the two coordinate value signals held immediately after the evaluation point coordinate value signal and further performing four arithmetic square root operations to generate an evaluation point unit tangent vector signal; ,
angle calculating means for calculating the absolute value of the difference between the evaluation start point unit tangent vector signal and the evaluation point unit tangent vector signal and outputting the result as an angle signal; distance calculation means for taking the absolute value of the difference between the coordinate value signal and the immediately preceding coordinate value signal, cumulatively calculating the absolute value, and outputting the result as a distance signal; and evaluating by multiplying the angle signal and the distance signal. evaluation function calculation means for outputting as a function signal; threshold determination means for comparing the evaluation function signal with a predetermined threshold signal and outputting a determination signal; and the determination signal is larger than the evaluation function signal. coordinate arrangement holding means for holding the evaluation point coordinate value signal as a drive coordinate value signal when
When the judgment signal indicates that the evaluation function signal is larger, the evaluation point coordinate value signal is designated as a new evaluation start point coordinate value signal, and the judgment signal indicates that the evaluation function signal is larger. processing control means for specifying the coordinate value signal held immediately after the evaluation fish coordinate value signal as a new evaluation point coordinate value signal when the evaluation fish coordinate value signal indicates that the evaluation point coordinate value signal is not present;
The method further includes: drive information calculation means for sequentially reading out the drive coordinate value signals during execution to calculate drive data for each degree of freedom of the robot; and drive means for driving the arm of the robot based on the drive data. Characteristic robot control device.