JPH01127284A - Operation controller for robot - Google Patents

Abstract

PURPOSE: To calculate a coordinate transformation required for controlling the movement of a revolute joint at low cost and at high speed, by providing a function calculation section in a movement control section for outputting sine and cosine values at the same time when moving one joint angle of a robot mechanism as an input signal. CONSTITUTION: A motion control section 7 for transforming coordinates for controlling the revolute joint of a robot mechanism has a function calculation section having function tables of sine and cosine. The function tables give approximate values of sine or cosine by one touch and hence sine or cosine of desired accuracy can be determined at the same time by making simple primary corrections for the values. A high-performance robot operation control device 2 can be obtained at low cost by providing the function calculation section like this.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a motion control device for a robot, and in particular a device for high-speed execution of coordinate transformation calculations necessary for motion control of industrial robots such as articulated robots. The present invention relates to a robot motion control device suitable for controlling the robot motion.

[Prior Art] In order to control the operation of a robot mechanism having at least one rotary joint, such as industrial robots that have become popular in industry, especially multi-jointed robots, it is necessary to perform coordinate transformation calculations in calculation means. I need.

The coordinate transformation calculation formula for an articulated robot is - for example,
An example of a horizontal articulated robot with two joints is as follows.

X = a 1cos θ1+ a 2cos(θ1+0
2) Y = a 1sinθ, +a, 5in(θ□+0
2) Here, X, Y are the positional components of the robot arm tip in the stationary coordinate system, a + 82 is the length of the first and second arms of the articulated robot, θ1. θ2 is the joint angle of the first joint and the second joint.

An example of such an arithmetic expression is, for example,
No. 4 and Japanese Patent Application Laid-open No. 136258/1983.

Therefore, in order to operate the robot based on a stationary coordinate system, the joint angles θ□ and θ2 are given by the calculation means provided in the robot control device, and the position components are calculated.

Calculation to obtain Y, or conversely, give position components X and Y and calculate rotation angle θ1. It is necessary to process the calculation for determining θ2 online at high speed.

[Problems to be solved by the invention] In the above example, since it is a simple example of a two-joint arm, -
Although the amount of calculations may appear to be small, the calculation of trigonometric function values is not that simple.For a robot arm with five or six joints, the calculations are extremely complex and the calculation load is quite large. Therefore, efforts have been made to provide a plurality of calculation means to reduce the calculation load, and also to provide a plurality of processors dedicated to numerical calculation in order to compensate for the lack of numerical calculation capability. However, such a numerical calculation processor is expensive and has a complicated system configuration, so that the practical effect obtained is low.

In addition, in order to avoid this, if an attempt is made to implement the calculation algorithm of a processor dedicated to numerical calculations within the calculation means of the robot control device, the algorithm of such a numerical calculation processor may be implemented, for example, by Nikkei Electronics'
As shown on pages 171 to 185 of 86-7-14, the method of calculating the function value by expanding the function into a series is used, so the calculation load becomes excessive, so it is difficult to implement it. I can't.

The present invention has been made to solve the above-mentioned problems, and the coordinate transformation calculation necessary for controlling the operation of the one-rotation node is performed at high speed using the calculation means of a normal control device, and the calculation of the calculation means is performed at high speed. The purpose of this invention is to provide a low-cost, high-performance robot motion control device that reduces the load, does not require the addition of a special calculation unit, and does not complicate the system.

[Means for Solving the Problems] In order to achieve the above object, a robot motion control device according to the present invention has a configuration including a robot mechanism section having at least one rotary joint, and a rotary joint of the robot mechanism section. and a motion control section that performs coordinate transformation processing to control the motion control section, and when the motion control section is activated with at least one joint angle of the robot mechanism section as an input signal,
A function calculation unit is provided that can simultaneously obtain and output a sine value and a cosine value.

As an additional note, the function calculation section is equipped with a function table for sine values or sine values and cosine values, and calculates the required sine and cosine values by searching the table using the function table and performing linear interpolation of the searched table values. This is what has become.

[Effect] The operation of the above technical means is as follows.

The function table for sine value or sine value and cosine value is 1
The sine or cosine value with the required precision can be obtained by simply giving an approximate value of the sine or cosine value by touching and applying simple first-order correction to it. A significant improvement in calculation speed can be obtained.

Specifically, the function table will be stored in a ROM (read only memory). This R.O.
Since M is a program storage area of the robot control device and a fixed data storage area, and is an area normally accessed by the calculation means, it does not complicate the system configuration and does not normally change the system in any way. do not have.

In addition, while the degree of integration of semiconductor memories such as the ROM is improving, prices are also rapidly decreasing, so even if a fairly large capacity table is provided, there is almost no increase in the cost of the calculation means, and the control table does not become large in size.

Furthermore, the use of an arithmetic algorithm that simultaneously calculates the sine and cosine values for a given angular component of a rotational joint utilizes the characteristic that sine and cosine calculations are always required in pairs. Compared to calculating each separately, for example, data pre-processing and post-processing can be shared, reducing calculation overhead. In other words, calculation loss can be reduced, which further speeds up the calculation of the − layer. It will be done.

[Embodiments] Hereinafter, each embodiment of the present invention will be described with reference to FIGS. 1 to 4.

First, FIG. 1 is a flowchart diagram showing arithmetic processing procedures in robot motion control according to an embodiment of the present invention.
FIG. 2 is a sine value table diagram showing an example of a function value table used in the arithmetic processing of FIG. 1.

Here, in the case of this embodiment, the data structure is defined as follows.

(i) The sine or cosine value, that is, the sine or cosine value = 1.0, is expressed as 100OOOOOH in hexadecimal.

(n) Angle 0": 000 0000H 90@: 200 0000H 180": 400 0000H 360"': 800 0000H In other words, the angle is Module 800 0000
Defined by H.

Here, the table of FIG. 2 is assumed to be a table in which sine values in the range of angles O to 90'C are given.

However, the definition range of the data is not limited to that shown in FIG.
Range of @~90°, range of -360@~360', etc.
It is sufficient to prepare in a form that is easy to realize in comparison with the system design requirements, that is, the required calculation speed.

Furthermore, instead of only sine values, a cosine value table may be prepared. Regarding this case, another embodiment will be described later.

The configuration of the table in Figure 2 is angle ooo.

000H, 0010000H, 0020000H,...
200 0000H (7) It is assumed that the sine value of the above-mentioned decimal point specifications for each Reni is stored. The table configuration is not limited to this, either, and the table density may be made coarser or more precise depending on the required precision of the results.

When the table is configured in this way, the value indicated by the upper bits of the specified angle, with the lower 16 bits discarded, gives the address of the table.

Therefore, if the lower 16 bits are O, for example, the angle is 0.
In the case of 03 0000H, upper data 0003H gives a table address, and the sine value is directly obtained by searching the table data of this address. bottom 1
In the general case where 6 bits are not 0, for example, the angle 003 0
In the case of 058H, the sine value is obtained by taking out the table address data 25B2DH1 indicated by the upper data 0O03H and the next table address data 3ED45H, and performing the next linear compensation operation.

SIN (0030058H) = (Data 25B2DH of table address 003H)
+ [((subtraction of lower bits ED45H-582DH)
*(Lowest 16 bits of angle data 0058H))) Upper 16 bits of the calculation result] Next, the case of calculating the cosine value will be described.

The following relationship exists between the sine value and the cosine value.

cos(θ)=sin(90'-〇) Since the sine table in Fig. 2 is configured in the range of 0 to 90°, the above formula can be expressed as cos(θ) = sin ( −〇). Therefore, to calculate the cosine value, use the value obtained by inverting the bit pattern in the angle range of 0 to 90' as the basic data for table search, and perform the same calculation as the calculation for solving the sine value described above. I know what's good.

An algorithm for simultaneously obtaining a sine value sin (θ) and a cosine value cos (θ) by giving an angle θ using the table structure and table search method described above will be explained according to the blocks in the flowchart of FIG.

First, the calculation means calculates the joint angle of the rotary joint of the robot. It is started with θ specified.

First, in block 10, the necessary preparations for calculating sine and cosine values are made. Next, in block 2o, the angular range of the specified joint angle θ is checked. That is, joint angle 0 is 0 to 906 or 180 to 27
If it is in the range of 0', the calculation procedures shown in blocks 2101 to 106 are executed, and the joint angle θ is in the range of 90 to 18 degrees.
or in the range of 270 to 360 degrees, the calculation procedures shown in blocks 201 to 206 are executed.

Here, in the case of this example, the joint angle θ is.

Since it is expressed as Module 800 0000H, the above joint angle check is 200 0000H.
If the value of the bit position of the data indicated by is O, the former,
If it is 1, it can be classified as the latter.

Next, after being classified as described above, blocks 101-
106 is shown below.

First, in block 101, joint angles are stored.

In this case, the range of joint angle data is o.

0 0000H ~ IFF FFFFH * or 4゜0
Because it is 0000H to 3FF FFFFH.

If you cut off the upper bits of 200 0000H or more,
An angle of 0~90'' is obtained (block 102).

Next, in block 103, using this joint angle,
The sine value is determined by the table search and one sweep interval described above. Subsequently, the cosine value is determined by inverting the angle data bit pattern (block 104), table lookup, and interpolation (block 1.degree. 5).

Finally, based on the memorized joint angles, in block 106, if the joint angles are specified in the range of 180 to 27 degrees, the signs of the obtained sine and cosine values are reversed and the final Obtain the sine and cosine values of.

The procedure for executing blocks 201 to 206 is almost the same, but since the joint angle ranges are different, the bit pattern of the data is opposite to that described above.

Therefore, the order of operations is reversed: first the cosine value, then the sine value. The same applies to the sign correction of the result.

As can be seen from the above explanation, compared to the conventional case where the calculation of the sine value and the calculation of the cosine value are performed separately, the block 10.20 and the blocks 101, 102, 1
06 or blocks 201, 202, and 206 can be shared or can be calculated only once, so the table search and interpolation methods described above can be combined with a shortened calculation time, resulting in a significant reduction in calculation time. can get. Also.

Since the sine and cosine values can be obtained as a pair at the same time, the coordinate transformation calculation itself can also be simplified by using the pair of sine and cosine values.

By the way, when the content explained in this example is implemented in a robot control device that uses a microprocessor with a clock frequency of 8MH1 as a calculation means, the table capacity is 2.
By simply adding bytes to , a computation time of 42 μsec was obtained. This is Nikkei Electronics, '86-7-
14. The calculation time of the 16.67MHz calculation-dedicated processor shown in Table 2 on page 185 is 23.0μsec*2 → 4
Although the clock rate is half that of the 6μ513C, it is faster, and it has been proven that there is no need for an expensive arithmetic processor.

Next, FIG. 3 is a flowchart showing arithmetic processing procedures in robot motion control according to another embodiment of the present invention.

In this embodiment, all calculations are performed using the floating point calculation method. Accordingly, the unit of joint angle is radian. The structure of the table is the same as the example in FIG. 2, but the numbers stored in the table are floating point numbers. Furthermore, unlike the previous embodiment, the range of the table ranges from 0 to 2π radians, with the unit of joint angle being radians.

The table is constructed in the same way as in FIG. 2, but the numbers stored in the table are floating point numbers. In addition, the range of the table configuration is different from the last time, in which a sine value table and a cosine value table of joint angles from 0 to 2π radians are prepared. However, in this case, even if the sine value table and the cosine value table are provided independently, the phase difference between the sine value and the cosine value is shifted by π/2 radians.

An overlap table that shares most of the table may be provided. It goes without saying that this idea of table configuration can be used in the previous example as well, and it is easy to understand that the larger the angular range of the table to be prepared, the smaller the amount of calculations in the case of the previous example. Good morning.

In this embodiment, in addition to the above table, a table of α/β for a constant β and various α is prepared in order to further reduce the amount of calculation. The specific contents will be explained with reference to the flowchart in FIG.

In the embodiment shown in FIG. 3, as in the embodiment shown in FIG.
The calculation means is activated with one of the joint angles θ being specified.

First, in block 501, the value of θ is limited to match the prepared table range 0 to 2π. get.

Next, at block 502, α:In on II:θ*β is calculated. Here, β is the reciprocal of the table resolution (table roughness) and is a constant.

Int [•] means taking the integer part of the real value in [•]. In the calculation of θ and β, processing such as rounding may be added to the calculation result in order to ensure the accuracy of the subsequent calculation.

The obtained value of α is an integer value and provides a table reference address.

Next, in block 503, α/β is obtained by table lookup using α, and the next subtraction is performed.

γ=θ. −α/β Here, γ is the joint angle portion below the table resolution.

Next, in block 504, a sine value sin α and a cosine value cos α are obtained by table lookup using α. However, in proper notation, α of sin α and cos α must represent an angle, which is 1, but in the above example, the table address α is used and is set to 1.

The same applies below.

Next, in blocks 505 and 506, the following operations are performed, respectively, to obtain sinθ and eO8θ.

= sin a÷γcos a =CoSα−γsinαゞAs described above, in the embodiment shown in FIG.
It can be seen that the sine and cosine values can be obtained simultaneously with only one multiplication and two operations. In the case of this embodiment as well, by adopting the method of calculating sine and cosine at the same time, blocks 501 to 504 in FIG. It is clear that the time saving effect is also significant.

In addition, block 501 in FIG. 3 sets the above joint angle range in consideration of the fact that the motion range of the robot is usually -π to +π, or even in the case of excessive motion, the range of -2π to +2π. It is also clear that this calculation step can be omitted if a covering table is prepared.

The arithmetic processing of the embodiments described above can be further speeded up by providing a plurality of arithmetic means and introducing a method of parallelizing the arithmetic operations.

Furthermore, even in a single-function numerical processing processor, if a function is provided to simultaneously calculate and output sine and cosine values for one joint angle input, the processing processor itself can become more sophisticated and faster. , when using this for robots,
Of course, the effect will be greater.

Next, a general embodiment of a robot motion control device employing such arithmetic processing means will be described with reference to FIG. 4.

FIG. 4 is a block diagram of an operation control device for an articulated spout pot according to an embodiment of the present invention.

In FIG. 4, 1 is a horizontal multi-jointed robot (hereinafter simply referred to as robot) having at least one rotational joint.
It is. 2 is a control device for driving and controlling the robot 1, and is connected to the robot 1 through a power cable and a signal cable. Control device! 2 is the console 3 and the calculation means 4.1F! It consists of a dynamic circuit 5 and a storage means 6.

The console 3 includes input means for operating the robot and teaching the robot to operate, display means for displaying the results of these inputs, displaying the operating state of the robot 1, and the like.

In addition, although not shown, communication means with other computers, communication means with peripheral devices, etc. are also connected.

Based on the input support from the console 3, the calculation means 4 refers to the operation program and operation data stored in the storage means, generates a robot operation control signal, and sends it to the drive circuit 5.

The drive circuit 5 generates a drive signal for the robot 1 based on the command from the calculation means 4, and drives actuators provided at each joint of the robot 1. Further, a feedback signal from the actuator, such as an encoder pulse signal, is sent to the calculation means 4.

In this way, the calculation means 4 executes all calculation processing such as controlling the operation of the robot 1, communicating with the outside, and exchanging signals with the console 3, but the data and programs that form the basis thereof are stored in the storage means 6. Stored.

The information stored in the storage means 6 is 7.7'...
, 7'. These include, for example, a motion control section 7 for controlling the motion of the robot 1, a console control section for decoding and executing instructions from the console, a storage section for robot motion programs, and a storage section for storing operating point positions. etc.

In this device, the part involved in the execution of the arithmetic processing described above is the operation control section 7, and a sine and cosine processing section 8 is located at a lower level thereof, and a function table section 9 is provided attached thereto. There is.

Therefore, the calculation means 4 executes the processing of the motion control section 7 in accordance with the requirements for rotary joint control, but at this time, the motion control section 7 performs coordinate transformation processing for controlling the rotary joint, so as to Te sine.

This uses the cosine processing section 8.

According to this embodiment, the calculation means in the robot control device
A function operation equipped with a sine value or a function table for sine and cosine values that can be used to obtain sine and cosine values by simply performing a simple primary correction on the table search results under the motion control unit when controlling the robot's motion. Since a sine/cosine processing unit is provided for the rotational joint, when a joint angle is given, it is possible to simultaneously obtain the sine and cosine values for the given angle component, and the coordinate transformation process for controlling the rotational joint is It is possible to speed up the calculation of sine and cosine values required for .

Furthermore, by utilizing the fact that sine and cosine values can be obtained simultaneously, other calculation algorithms can be simplified, making it possible to significantly reduce the calculation load on the calculation means.

Furthermore, the cost increases due to the addition of a function calculation section.
Another advantage is that a high-performance robot motion control device can be obtained at low cost and without complicating the system.

In each of the above-mentioned embodiments, the calculation means applied to the motion control device of a horizontal multi-joint robot has been described, but the present invention is not limited to this. Needless to say, the present invention can be generally applied to an operation control unit that performs coordinate transformation processing for controlling the .

[Effects of the Invention] As described above, according to the present invention, the coordinate transformation calculation necessary for controlling the motion of a rotary joint can be executed at high speed using the calculation means of a normal control device, and the calculation load on the calculation means can be reduced. reduce
It is possible to provide a low-cost, high-performance robot motion control device without adding a special calculation unit or complicating the system.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing arithmetic processing procedures in robot motion control according to an embodiment of the present invention, and FIG. 2 is a sine value table showing an example of a function value table used in the arithmetic processing of FIG. FIG. 3 is a flowchart showing the arithmetic processing procedure in the motion control unit of a robot according to another embodiment of the present invention, and FIG. 4 is a motion control device for an articulated robot according to an embodiment of the present invention. FIG. DESCRIPTION OF SYMBOLS 1... Robot, 2... Control device, 4... Arithmetic means, 5... Drive circuit, 6... Storage means, 7... Operation control unit, 8... Sine, cosine processing Part, 9...Function table part.

Claims (1)

[Claims] 1. A robot mechanism unit having at least one rotary joint, and an operation control unit that performs coordinate transformation processing for controlling the rotation joint of the robot mechanism unit, and the operation control unit includes: A robot motion control device is provided with a function calculation section capable of simultaneously determining and outputting a sine value and a cosine value when activated using at least one joint angle of the robot mechanism section as an input signal. . 2. In the device described in claim 1, the function calculation unit includes a function table of sine values or sine values and cosine values, and performs table search using the function table;
A robot motion control device characterized in that required sine and cosine values are obtained by linear interpolation of retrieved table values. 3. The robot motion control device as set forth in claim 1, wherein the function calculation unit is constituted by a dedicated numerical calculation processor.