JPH0890465A - Robot pose performance automatic measurement system - Google Patents

Abstract

(57) [Abstract] [Purpose] It is an object of the present invention to provide an automatic measurement system that enables easy and accurate measurement of the pose performance of a robot. [Structure] A target 1 having three surfaces perpendicular to each other and connected to a robot 3, and a multi-sensor-measuring head 4 having three surfaces perpendicular to each other having a non-contact sensor-5.
And between the sensor 5 and the computer system 12 to mediate exchange of signals between them, and between the robot controller 30 and the computer system 12 between them. Interface control unit 6 for transmitting and receiving signals, and software 2 for pose calculation based on at least the cubic localization measurement principle.
And a computer system 12 having zero.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system for automatically measuring the pose performance of an industrial robot based on the ISO9283 standard. Here, the pose of the robot means the position and direction in the space of the arm of the robot, and the vector quantity p = (x,
It can be represented by y, z, a, b, c). Among them, (x, y, z) represents a position, and (a, b, c) represents a direction.

[0002]

2. Description of the Related Art There are various types of robot performance.
Users are interested in structural shapes, workspaces,
Freedom, limited load, position or pose performance, and price. One of the basic conditions for using a robot is that the wrist of the robot has a satisfactory pose repeatability under a specific work environment.

If the user wants the robot to operate in an off-line programming scheme, then he must place a demand on the accuracy of the pose, and he wants the robot to move exactly along the specified trajectory. If so, it should make a request for the path characteristics of the robot. In quality control in robot manufacturing process, performance inspection of robot products, robot error correction, problem diagnosis, robotization method in production process, research on precision equipment method, etc.
Measuring the pose performance of a robot is necessary and important.

However, accurate description and survey of the pose performance of a robot is not easy. It is due to the following causes. 1) Generally speaking, industrial robots have an articulated structure with some sort of seriality. Although the position of each joint can be controlled by the closed loop and high localization accuracy can be obtained, the posture of the wrist cannot be fed back to the control unit of the robot at any time. The minute position errors of each joint are magnified by the length of the arm, and these errors also add or cancel depending on the state in which the arm of the robot is extended. In other words, the robot poses differently at different positions in the work space depending on its structure and control characteristics.
There are even cases where the difference is very large.

2) The pose performance that can be realized by the robot during the actual work is affected by usage conditions such as load load, motion speed and direction, and stability of mounting base. 3) It is necessary to measure a large amount of data in order to perform high reliability and full evaluation. Each measurement must be coordinated with robot motion control.

The method of measuring the position repeatability of a robot using a contact type micrometer is a simple method often used in factories, but the disadvantages of this method are shown below. 1) The position repeatability can be measured, but the direction repeatability cannot be measured. 2) Measurement efficiency is poor. Only one position element can be measured at a time. 3) Dynamic characteristics cannot be measured. 4) Not suitable for automatic measurement.

General-purpose optical measuring instruments such as theodolite,
Even if a laser interferometer-type distance measuring device or a multi-camera system is used, it is not possible to satisfy the measurement requirement of the pose performance of the robot. The main causes are shown below. 1) Normal position data can be measured, but direction data is difficult to measure. 2) Communication and cooperative operation with the robot controller cannot be performed well. Therefore, it is difficult to realize automatic measurement. 3) Use is restricted by environmental factors such as temperature, humidity, dust and site space.

[0008]

SUMMARY OF THE INVENTION It is an object of the present invention to provide an automatic measuring system which overcomes the drawbacks of the conventional measuring instruments described above and makes it possible to measure the pose performance of a robot easily and accurately. This is an issue. More specifically, the present invention is based on the non-contact cubic position measuring principle,
It is possible to obtain position and direction data at the same time, complete measurement control and data processing functions, easily connect to the robot under measurement, and various pose performance specified in ISO9283. To provide an automatic pose performance measurement system for industrial robots, which can automatically measure, have an excellent noise prevention function, can perform stable operation in the work environment, and can create the required measurement report. Is an issue.

[0009]

According to the present invention, a target having three surfaces perpendicular to each other and connected to a measured portion of a robot via a mechanical interface, and a non-contact type sensor are provided. A multi-sensor measuring head having three surfaces perpendicular to each other and fixed at a measuring position in the working space of the robot, and a signal located between the sensor and the computer system. An interface control unit that mediates the transfer and the transfer of signals between the robot controller and the computer system, and a port based on at least the cubic localization measurement principle. The above problem is solved by a robot pose performance automatic measuring system comprising the computer system having software for calculating the size. On the contrary to the above, the target may be fixed to the measuring position in the working space of the robot and the multi-sensor measuring head may be attached to the robot.

[0010]

[Operation] A multi-sensor having three non-contacting surfaces having a non-contact type sensor-a measuring head is opposed to a target having three surfaces which are perpendicular to each other, and the relative movement of the targets is cubically oriented. Poses are calculated based on the measurement principle,
Position and direction data are measured simultaneously.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described with reference to the accompanying drawings. The industrial robot pose performance automatic measurement system according to the present invention comprises at least the following four organically related parts.

1) Target 1 having three front planes which are perpendicular to each other. Target 1 is an ordinary cube, but a rectangular parallelepiped and others.
A solid having three planes perpendicular to each other is sufficient. Target
The get 1 is connected to the arm (or operating device interface) of the robot 3 to be measured through a mechanical interface 2.

2) Multi-sensor-measuring head 4 with three surfaces perpendicular to each other The measuring head 4 has at least six non-contact sensors-5 arranged on said three surfaces and is mechanically The robot 3 can be fixed at any position in the work space of the robot 3. When the target 1 attached to the robot 3 approaches the measuring head 4, each sensor 5 measures the distance to the surface of the target 1.
Data is sent. The displacement and rotation angle of the target 1 with respect to the measuring head 4 are calculated using these data and the pose calculation algorithm. Using this displacement and the rotation angle, the pose performance of the measured robot 3 is evaluated.

3) Electronic interface control unit 6
(EICU) The EICU 6 provides a power supply and a signal necessary for the sensor 5 through electronic parts, filters a signal output from the sensor 5 with an analog filter-7, and A / D converter 8
After converting to a digital signal through the computer 12
Send to. The EICU 6 also receives control signals from the computer 12 and the robot under test 3 through the digital input / output (DI / O) interface 9, and communicates between the measurement system and the robot through the logic control circuit 10. , And other functions such as auxiliary control, voltage level matching, and electrical insulation are realized.

4) Complete computer system The computer system 12 includes at least a CPU, a memory, a serial and parallel interface 15, and a hardware.
Disk and driver 16, floppy disk driver 17, display 18, keyboard 19, OS system windows and dedicated software package 20.
including. By using software, functions such as measurement management and control, data collection and processing, and automatic generation of measurement result report can be realized. The EICU 6 may be housed in a separate cabinet, or part or all may be housed in a computer cabinet.

In the above-mentioned example, the target 1 is fixed to the robot 3 to be measured, and its movement is monitored by the multi-sensor-measuring head 4 fixed at the measuring position in the working space of the robot. Alternatively, the target 1 may be fixed to a measurement position in the work space of the robot, and the measurement head 4 may be connected to the robot 3. Also in this case, by monitoring the relative movement of the target 1, the same measurement result as in the above example can be obtained.

The feature of the present invention is that the position and direction data can be measured at the same time by adopting the non-contact type cubic localization measurement principle. Since there is no mechanical contact between the measurement head 4 and the target 1 during measurement, it is possible to avoid interference with the measurement result due to the contact force in the contact type measurement. Since the non-contact sensor-5 has no mechanical moving element, frictional damage does not occur. Therefore, the reliability of the system can be improved.

Another important aspect of the present invention is that it is an integrated measurement method. One platform
And a high-accuracy displacement measuring instrument (for example, laser interferometer type distance measuring instrument) are used to measure the target 1, the measuring head 4, the analog filter 7, and the A / D converter 8. Calibrate the chain. Once the calibration is completed, a calibration data table including all measurement chains (number of measurement chains = number of sensors) can be obtained. It is stored in the computer as a data file.

In the subsequent measurement process, the distance between each sensor and the target is obtained by interpolation calculation of the data from the sensor 5 and the calibration data table. Then, the pose characteristic is obtained by the pose calculation method. In the integrated calibration, a pseudo trihedron (measuring head coordinate system) consisting of sensors is created. It also increases the accuracy of the system so as to minimize system errors that occur in the measuring chain.

In order to realize the automation of the pose performance measurement, it is necessary to guarantee the synchronization between the measuring system and the robot controller 30. The present invention provides a simple and effective synchronization method and implementation means. That is, two signals of MS (system measurement state) and RS (robot motion state) are provided. MS = 1 indicates that the measurement system is collecting data and the robot operation is prohibited. MS = 0 is
When data collection is completed, the robot is ready to operate. Further, RS = 1 represents a state in which the robot is operating and measurement is impossible. RS = 0 indicates that the robot arm has reached the specified position (or entered the measurement range) and is ready for measurement. The measuring system and the robot controller 30 control their own control flow by these two signals. This realizes the synchronization between the measuring system and the robot controller 30.

In the present invention, the MS / RS cable 32 connects the digital I / O interface 31 of the robot controller 30 and the EICU 6 of the measurement system. The EICU 6 is electrically shielded and has a voltage level matching function so that the measurement system can be easily connected to various robots having different digital I / O electric characteristics.

In order for the measurement system to remain reliable in the work environment, the present invention uses analog and digital two-layer filter techniques. In other words, the output of the sensor-5 is first passed through the analog filter-7, converted into a digital signal, and then processed by the digital filter-21 to improve the noise resistance of the measurement system. I am trying.

A salient feature of the present invention is that the metrology system has a high degree of automation. In addition to the initial setting of measurement parameters and adjustment of the initial position of the measuring head, measurement, processing and reporting of results are all automated. The present invention is used to measure the following pose performances.

1) Pose stabilization time and pose overshoot amount (PST and POS) or (STO) 2) Pose accuracy and pose repeatability (AP and RP) 3) Multidirectional poise Return accuracy (VAP) 4) Poise characteristic drift (Drift) or (DP)
C)

The automation of the measurement process has become possible due to the design of the system according to the measurement requirements and the use of computer control technology. At the heart of the computer control system is a rationally designed software package 20. In the present invention, the measurement system software must have at least the following functions. 1) Management and control of measurement 2) Collection and processing of data 3) Creation of measurement result report

In order to obtain a complete evaluation of the pose of the robot, the relevant measurement item must be measured at a large number of measurement points. The ISO9283 standard defines five possible measurement points. The present invention uses one measurement process status table to manage the measurement process and data file, thereby avoiding erroneous mistakes such as forgetting measurement, duplication, and misalignment in operation. You can

When one measurement item ends at a certain measurement point (measurement and data storage are all completed), the end date is recorded in the measurement process state table so that it can be displayed on the screen (see FIG. 3). ). When a duplicate measurement occurs, the computer automatically detects it and informs the user. When a misalignment setting occurs (for example, VAP measurement is performed on the measurement point P3), the computer informs the user that this is an incorrect measurement setting. In the present invention, since the measurement data of all the combinations of measurement shown in FIG. 3 are stored in one data file, the data retrieval and processing are simple.

The present invention provides a user-friendly graphics interface. The user can easily set measurement conditions such as measurement items, measurement points, and various measurement parameters according to the presented contents. on the other hand,
The computer control system operates according to the measurement parameters set by the user and the signal RS of the robot operating state.
Control the measurement process. At the same time, in the measurement process, the computer system monitors in real time the output of the precision power adapter providing power to the sensor. If the change in the output exceeds the normal range,
The computer stops the immediate measurement and gives an alarm.

The flow of data collection and processing in the present invention is shown in FIG. The output signal of the sensor is first converted into a digital signal by the A / D converter 8 after the high frequency is cut by using the analog filter-7. Further, by processing the converted digital signal using the digital filter-21, generation of a large error due to external noise is avoided. Then, using the calibration data table of the measurement chain, the non-linear error of the sensor and other system errors in the measurement chain are corrected, and the distance data from each sensor to the target surface is calculated. To do. By the distance data and the pose calculation algorithm, the pose and change of the target 1 with respect to the measuring head 4 can be known.

Another feature of the present invention is that it provides a variety of measurement result representation formats to facilitate further analysis and development of the measurement results. The present invention provides the following three different measurement result formats (display on the display 18 or output by the printer 33).

1) General Report On the initial screen of the format of the general report, information about the robot to be measured, the measurement environment, the measuring instrument, the data file and the like and the measurement process status table are shown ( (Fig. 4). In other formats, the stability time and overshoot amount, pose accuracy and repeatability, multidirectional pose return accuracy, and measurement results and conditions of drift of pose characteristics are included. Shown. This general report is automatically created by the metrology system. In addition, the software of the present invention also provides an editing function in the general report menu. The user can use it to comment in the general report in a free format.

2) Graphic Report The graphic processing function of the software of the present invention enables the user to very easily output a graphic display of an arbitrary pose component in the measurement result. FIG. 5 shows an example of one-dimensional graphic output of the pose stabilization time and the overshoot amount. 6 and 7 are examples of one-dimensional and two-dimensional graphic output of AP and RP, respectively. FIG. 8 shows an example of a VAP two-dimensional graphic output. FIG. 9 is an output example of the one-dimensional error distribution diagram of the pose repeatability. FIG. 10 is an example of one-dimensional graphic output of drift of pose characteristics. These figures are very useful for error analysis and problem diagnosis of robots.

3) Distance data table between each sensor and target The accuracy of the processing result can be verified using these data. Also, these data can be used for new data processing which will be developed in the future.

[0034]

Since the present invention is as described above, position and direction data can be obtained at the same time by the pose calculation algorithm based on the non-contact type cube position measuring principle, and it has been completed. Equipped with measurement control and data processing functions, it can be easily connected to the robot to be measured.
It is capable of automatically measuring the various pose performances specified in 1., has an excellent noise prevention function, can perform stable operation in the work environment, and can generate the necessary measurement report. -It is possible to provide an automatic performance measurement system.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of an embodiment of the present invention.

FIG. 2 is a diagram showing a data collection and processing flow in an embodiment of the present invention.

FIG. 3 is a diagram showing an example of a measurement process state table in the example of the present invention.

FIG. 4 is a diagram showing an example of an initial screen of a general report according to the embodiment of the present invention.

FIG. 5 is a diagram showing a one-dimensional graphic output example of the pose stabilization time and the overshoot amount in the embodiment of the present invention.

FIG. 6 is a diagram showing an example of one-dimensional graphic output of pose accuracy and pose repeat accuracy (AP and PR) in the embodiment of the present invention.

FIG. 7 is a diagram showing an example of two-dimensional graphic output of AP and PR in the embodiment of the present invention.

FIG. 8 is a diagram showing an example of a two-dimensional figure output with multi-directional pose return accuracy (VAP) in the embodiment of the present invention.

FIG. 9 is a diagram showing a one-dimensional error distribution diagram output example of the pose repeatability in the embodiment of the present invention.

FIG. 10 is a diagram showing an example of a one-dimensional graphic output of drift of the pose characteristic according to the embodiment of the present invention.

[Explanation of symbols]

 1 Target 2 Mechanical interface 3 Robot to be measured 4 Measuring head 5 Non-contact type sensor 6 Electronic interface control unit (EICU) 12 Computer system 20 Software package 30 Robot controller La

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yan Ning Singapore 0511 Science Park Drive 73 Sintech Building SJ (Singapore) Research and Development Pity Y Limited

Claims (3)

[Claims] 1. A target having three surfaces perpendicular to each other and connected to a measured portion of a robot via a mechanical interface, and three surfaces perpendicular to each other having a non-contact sensor. Have,
The multi-sensor measuring head is fixed to a measuring position in the work space of the robot, and is positioned between the sensor and the computer system to mediate exchange of signals between the both and a robot controller. An interface control unit located between the computer system and the computer system for mediating the exchange of signals between them, and the computer having software for pose calculation based on at least the cubic localization measurement principle. -Automatic measurement system for robot pose performance consisting of a data system. 2. A target having three surfaces which are perpendicular to each other and fixed at a measurement position in a working space of a robot, and three surfaces which are perpendicular to each other having a non-contact type sensor,
A multi-sensor measuring head, which is connected to the measured part of the robot via a mechanical interface, is located between the sensor and the computer system to mediate exchange of signals between them. At the same time, an interface control unit located between the robot controller and the computer system for mediating the exchange of signals between them, and at least for pose calculation based on the cubic localization measurement principle. A system for automatically measuring the pose performance of a robot, which comprises the computer system having the above software. 3. The automatic pose performance measuring system for a robot according to claim 1, wherein a table showing the measuring process is used to assist the management of the measuring process and the data file.