JPS5941012A - Teaching method of robot - Google Patents

Abstract

PURPOSE:To attain highly accurate positioning by calculating the end point coordinate values of a position to be moved from input data to move a robot to an objective attitude. CONSTITUTION:When the driving of the robot is controlled in accordance with a teaching operation from a teaching box 3, data are inputted to a microprocessor 4 in a controlling device 1 through a serial I/F 8 at the depression of an operation key in accordance with the operation from the box 3 to calculate the objective coordinate values corresponding to the key input. The device 1 and a mechanical part 2 constitute a position controlling system. An output value from a pulse encoder PE is fed back through a counter 7, the difference between the feedback value and the converted value of the objective coordinate value previously found by the device 4 which corresponds to the encoder PE is D/A converted 6 and outputted to drive an actuator.

Description

[Detailed description of the invention] The present invention relates to a method for teaching a robot.

In the conventional robot teaching method, 1%
As disclosed in No. 14, two ITV cameras are used to project the target beggar on the TV screen, and by specifying the movement target point of the robot hand on each screen using a light ben, the six-dimensional target point can be calculated. Since the coordinates were calculated and the robot was controlled to approach a predetermined position, the conditions for objects that could be recognized by the camera were limited, and the addition of one image processing device increased costs.

The purpose of the present invention is to move a robot to an arbitrary position by providing data such as hand posture information (hand coordinate system position 9 hand torsion angle) using a human-powered device. It is an object of the present invention to provide a robot teaching method for calculating the previous end point coordinate value and teaching the robot to move to a target posture.

That is, the present invention is a teaching method for a robot controlled using a point-to-point method, in which a command is given to move the robot to a desired position and stop it, and posture information of the stopped robot is captured in a memory. The robot's posture information is displayed on a display device in an orthogonal coordinate system, and based on the robot's posture information displayed in this orthogonal coordinate system, commands are given in the orthogonal coordinate system to correct the robot to a predetermined posture. This robot teaching method is characterized by storing posture information of the robot in a memory and teaching the robot.

In particular, according to the present invention, in assembly work where a part is inserted into a shaft attached at a predetermined angle, the hand of the robot is finely adjusted so that it is parallel to the shaft direction before insertion, In addition, by aligning the center and moving in hand coordinate mode, smooth insertion is possible. In this case, since the instructor already knows the mounting angle of the shaft and the position of the center of the shaft in the orthogonal coordinate system, all he or she has to do is input the data in the orthogonal coordinate system as the hand posture information.

Quick and highly accurate positioning becomes possible.

The present invention will be described below based on an embodiment shown in the drawings. FIG. 1 shows a hardware configuration for driving and controlling the robot according to instructions from the teaching box 6. As shown in FIG. When the operation key is pressed by the teaching box 3, the data is sent to the control device 10 via the serial INF B.
The microprocessor (J)tic is loaded and performs calculation processing of target coordinate values according to key human power. The control device 1 and the robot mechanism section 2 form a position control system, and the output value from the pulse encoder pE is fed back through the counter 7, and the microprocessor (A) 4 performs an encoder-equivalent conversion of the previously obtained target coordinate value. The difference between the two values is converted into an analog signal using D/16 and output, and the actuator is driven.

FIG. 2 is a plan view showing the teaching box 3 in detail. Move the normal robot χ manually,
A teaching work basic function operation section for storing position data in a storage section, a robot posture information display section 9 and a posture information input device section 10 of the present invention are integrated.

FIG. 3 is a flowchart showing the flow of processing according to the present invention in relation to the flow of teaching work. is shown in more detail.

FIG. 5 shows the relationship of coordinate systems in order to explain the robot posture information described in the present invention.

In other words, posture information includes the position (Y value, Y value, Z value) of the robot's hand position tP in the XYZ coordinate system and the torsion angle (α
The robot's orthogonal coordinate values are expressed in units of mrn and degrees, respectively (11).

α value = angle between the hand coordinate axis XH and the XY plane.

β value = angle that the hand coordinate axis Y makes with the xY plane.

γ value = absolute rotation angle of hand axis (+6) means.

Next, the operating state will be specifically explained using the flowchart in FIG. 3 and the hardware configuration in the first part.

When the microprocessor (4) 4's teaching gummy gram (stored in ROM 2'5) is started, after initializing each data,
Key human commands from the teaching box are interpreted. This is because the microprocessor CB) 12 periodically receives information from the keyboard display I/p 15 about which keys have been pressed on the teaching basic operation section 19 and posture information input device section 10 of the teaching box 3 using a signal from the clock generator 14. , waiting for a data transfer request signal from the control device 1 side, the path line 18, AC
″r, 416, level changer 15 to serial I/
FB, microprocessor (A) via pass line 5
The code data sent to 4 is sent to the microprocessor (A
) 4 is to analyze what kind of data requires processing, pass the data to the general processing routine, and then fly.

Among these, there is a teaching process 20α for importing the robot's position coordinate values into the storage unit 22, and when moving the robot manually, each axis is moved independently in joint mode, or the hand is moved in orthogonal mode along the orthogonal coordinate system. In the mode change process for making the selection ') Oe, 2pf, and hand opening operation process 20!1 to 207, each target value calculation process is performed according to the operation mode determined by the mode conversion, but all of them are performed using orthogonal coordinates. The value (,4) is obtained and stored in the data. Figure 4 shows the case of an important text.When the key with the approximate coordinate value is first pressed and released from °C, the value of the displacement that continues to increase gradually within a range that does not exceed the maximum displacement is calculated, and when the key is released, Then, find the displacement that continues to be gradually decreased until it becomes zero, and add or subtract it from the orthogonal coordinate value (, 4) found one cycle ago depending on the direction of movement, and use it as the current orthogonal target coordinate value. There is. The difference between this orthogonal target value converted into the encoder equivalent value of each axis of the robot and the current value read from the actual output from the pulse encoder PE via the counter 7 is the current command pulse for each axis. By outputting the number to the D/A 6, a speed command is given to the servo driver to drive the robot.

In other words, when teaching is performed by manually moving the robot, the closing process 2UC,
20d, and the microprocessor (B) 12 to indicate that the processing keys 20a to 20f have been pressed.
A detailed explanation of the process of lighting up the LEDC light emitting diodes 21σ to 211 above the keys 20a to 20f pressed via f'l0t7y will be omitted.

It starts moving while accelerating, and when you release the key, it gradually decelerates,
The robot will stop after repeating several cycles until the displacement at the time of target value calculation becomes zero, so it is extremely difficult to stop the robot with high precision as you will have to bring your face close to the robot's hands and watch the robot's movements while trying to determine the position. It becomes a difficult task.

Next, the content of the present invention will be described and its effects will be clarified. First, after stopping the robot near the point to be positioned using the teaching method described above, by pressing the OUT key 106 of the posture information input device section 10, the OUT key processing routine is called by command interpretation processing.

Then, RAM22 indicates the point at which the robot is currently stopped.
The content of the orthogonal coordinate value (,4) within (X value.

The microprocessor (A) 4 uses ASCII code to convert the Y value, Z value, α value, β value, Y value) into posture information values.
When converted into data and outputted as serial I/p 8K, the microprocessor CBNz of the teaching pendant lul+ formats and writes the transmitted data to the RAM 24, and the written data is read out and displayed on the keyboard display I/P. The contents are displayed on the posture information display unit 9 from F16. Looking at this display state, for example, move the robot +2 more distances in the X direction,
is 0 degrees and β is -1.5 degrees, and since the hand is not horizontal, if it is determined that if β is set to 0 degrees, that point will be the target position, then 7 of the posture information human power device unit 10 - When the N key 1oα is pressed, the LN key processing routine is called and the key waits.

Then, key in the values of posture information X and β that you want to change in sequence, and finally press the CR key 1od. Microprocessor (A
) 4 stores these data that are sent sequentially, and when the CR code is read, it is considered that the input is completed, and the data is decoded to retrieve the attitude information CY, Z, α, γ that was not input.
) converts the data obtained at the time of display, and newly input data (X, β) converts the data from ASCII code to orthogonal coordinate values (B). Then, according to the flow shown in FIG. 5, the difference between the orthogonal coordinate values CE') and CA) is calculated, and the above-mentioned arm motion processing routine is used to determine that the difference is within the range in which motion control is possible.

The robot is moved and positioned using the input coordinate values as the target coordinate values. If there is an input outside the controllable range, an error message is displayed on the posture information display section 9 and the next process is awaited. Then, when the movement to the target point is completed, by pressing the teaching key 20α, the coordinate values of the point currently stopped are taught to the storage unit 22 as the position coordinate values.

Thereafter, manual movement, posture information input and output operations, and teaching are repeated in the same manner to perform multi-point teaching work.

According to this embodiment, most of the assembly work involves inserting parts with the hand horizontally or at a right angle.
, Y, and Z position dimensions are known in advance, making it easy to input posture information, and unlike traditional manual movement teaching, the VC face can be brought close to the mouth and pot and the direction and position can be fine-tuned many times while relying on visual inspection. There is no need to make adjustments, allowing for efficient, highly accurate, and safe teaching.

xpn73″M25 is a grodarum for controlling the microprocessor (B) 12.

As explained above, according to the present invention, when trying to determine the position and posture of the robot's hands with high precision, the robot is moved in positive and negative directions many times while relying on visual inspection, unlike conventional teaching. without having to adjust
This can be done by simply arranging the displacement amount of position and posture at once and setting it from the manual device, improving the efficiency of teaching work.
This has the effect of enabling more accurate positioning.

[Brief explanation of the drawing]

FIG. 1 is a hardware configuration diagram of the present invention, FIG. 2 is a plan view showing the external appearance of the teaching box shown in FIG. 1, and FIG.
The figure shows a flowchart of a robot teaching method according to the present invention, including a routine for controlling the robot to move to a target point by displaying posture information and inputting posture information, and FIG. 4 shows the orthogonal target value calculation process shown in FIG. 3. FIG. 5 is a flowchart showing the input point orthogonal target value calculation process shown in FIG. 3. FIG. 5 is the robot coordinate system XY
Z and hand coordinate system XM. It is a diagram showing the relationship between Yx and ZJF. 1... Robot I/control device 2... a-bot mechanism section 6
... Teaching box 4 ... Microprocessor (A) 5 ... Pass line 6 ... Router 7 ... Counter 8 ... Serial I/p9 ... Attitude information input device section 10 ... Attitude information input device section 11...LED 12...Microprocessor (B) 13...Keyboard display I/F 14...Rea software generator 1/-ta 15...Level changer 16...ACIA 17 ...PI''?51
8...Bus Line No. 40 Procedural amendment (method) Indication of the case 1982 Patent Application No. 61049 Title of the invention Person who amends the robot teaching method Relationship with the case Patent applicant Name of person Title +5
101 IT Co., Ltd. New type science manufactured by Hitachi
Contents of human correction

Claims (1)

[Claims] In a teaching method for a robot controlled by a point-to-point method, a command is given to move the robot to an arbitrary position and stop it, and the posture information of the stopped robot is captured into memory. is displayed on the display device in a Cartesian coordinate system, and based on the robot posture information displayed in this Cartesian coordinate system, commands are given in the Cartesian coordinate system to correct the robot to a predetermined posture, and the corrected robot is A robot teaching method characterized by storing posture information in a memory and teaching the robot.