JPH0425908A - Detection of position for industrial robot - Google Patents

Abstract

PURPOSE:To eliminate the need of always synthesis-operating an absolute rotary angle signal and to reduce the burden of an operation so as to improve detection precision with simple constitution by detecting an absolute position through the use of information from a position detector after an absolute position initial value is set. CONSTITUTION:When power is supplied, a switch 11 is sequentially changed over to a resolver 4-side and a resolver 5-side by a command from an arithmetic unit 15, and position information from respective resolvers are outputted to the arithmetic unit 15 through an R-phase processing circuit 8 and an R/D converter 12. Then, respective digital values are stored in registers 15a and 15b. The arithmetic unit 15 operates the absolute position initial value at the time of power supply based on data stored in respective registers. The operated result is stored in a register 15e and the count values of the incremental pulses of A and B phases are added with the initial value of the register 15e in a accordance with the absolute rotary angle signal from the resolver 4, whereby it is outputted as position information.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a position detection method for detecting the absolute rotational position of a rotating shaft of an arm or the like in an industrial robot such as an arc welding robot.

[Prior Art] Position detection means for industrial robots generally include a multi-rotation absolute type and an incremental type that requires origin alignment.

Various types of multi-rotation absolute types have been proposed that are equipped with two absolute position detectors (detecting the absolute rotation angle within one rotation) such as resolvers and absolute encoders. For example, ■ Two absolute position detectors are installed across a reducer with a relatively large reduction ratio, one absolute position detector detects rough positions, and the other absolute position detector detects fine positions. However, there is a means for combining these detection results to detect the absolute position of multiple rotations, and for one of the two absolute position detectors.
The other method detects the rotation decelerated by a gear with a slight difference in the number of teeth, detects the number of rotations based on the amount of deviation in the detection results by two absolute position detectors, and obtains the absolute position of multiple rotations. be.

When such a multi-rotation absolute type is used to detect the angle of each axis of an industrial robot, based on its principle, the absolute angle of each axis can be obtained immediately after power is turned on.
Since the detection result is basically data equivalent to the angle of the motor shaft and its multiple rotations, it is possible to obtain a highly accurate angle of the robot arm, etc. that actually moves via the reducer. .

On the other hand, as for the incremental type, an incremental encoder is used, and a tl1 operation limit 1 limit switch (771 point limit switch) is installed.
When the power is turned on, the following origin alignment work is performed before position detection. In other words.

After the robot is moved to the operating limit, the limit switch is activated, and the robot is moved in the opposite direction, the counter is reset by the first Z pulse (generated once per rotation) output from the incremental encoder.

Thereafter, as the robot moves, the A pulse and B pulse having a phase difference of 9o'' output from the incremental encoder are counted up and down by a counter, and the count value is used as the robot's position information.

[Problem to be solved by the invention] However, although the latter incremental type position detection means has the same detection accuracy as the former multi-rotation absolute type, it requires a considerable amount of work to align the origin when the power is turned on. It has the disadvantage of requiring complicated work.

On the other hand, the former multi-rotation absolute type position detection means does not require any work to align the origin;
Since two absolute position detectors are used, processing to combine the respective detection results is required. In order to perform such synthesis processing, a circuit as shown in FIG. 6 or 7, for example, is configured. In Figures 6 and 7, 50a
, 50b are resolvers (#I! position detector) for detecting the absolute rotation angle before and after deceleration, respectively, 51a, 51b
are excitation circuits (drivers) for exciting the resolvers 50a and 50b, respectively; 52a and 52b are waveform processing circuits (around R phase processing; 53a and 53b are waveform processing circuits that receive detection signals from the resolvers 50a and 50b and perform waveform processing; Each waveform processing circuit 52a.

This is an R/D (resolver/digital) converter that receives the detection signal from 52b, converts it into digital data, and outputs bits.

In the circuit shown in FIG. 6, the R/D converter 53a,
53b is directly input to the CPU 54, and the CPU 54 inputs 2
Multi-rotation absolute data is obtained by performing arithmetic processing such as multiplication 9 judgment 2 summation on the detected values from the resolvers 50a and 50b. In addition to significantly increasing the burden on the CPU 54, there are also problems such as an increase in the number of signal lines for inputting data to the CPU 54.

On the other hand, in the circuit shown in FIG.
The load on the CPU 54 is reduced by interposing a converter 55 having a function of 5-composite arithmetic processing between the converter 3b and the CPU 54.

The number of cores of the signal line between the CPU 54 and the converter 55 increases.

As described above, in the multi-rotation absolute type position detection means up to 29, the number of signal lines between the CPU 54 and the CPU 54 is large, and in some cases, there is a time lag in data update when serial transfer is used. 5.
4 or the converter 55 is burdened.

In addition, in principle, the highest bit of the signal line has a motor rotation speed of about 100 revolutions, and if noise or the like were to be mixed in, there was a risk that the robot would suddenly run out of control or the control would become unstable.

The present invention aims to solve these problems, and while retaining the greatest feature of the conventional multi-rotation absolute type, which is that absolute position can be obtained immediately after power is turned on, the present invention is also partially compatible with the incremental type. The purpose of the present invention is to provide a method for detecting the position of an industrial robot with a simple configuration and high detection accuracy, with a small computational burden.

[Means for Solving the Problems] In order to achieve the above object, the method for detecting the position of an industrial robot according to the present invention (claim 1) includes transmitting absolute rotation angle signals to the before deceleration side and the after deceleration side of the deceleration mechanism, respectively. Detecting the absolute position of a component of an industrial robot equipped with first and second position detectors, the absolute position initial value of the component being detected from the first and second position detector. After that, the absolute position of the component is detected by integrating the incremental pulses corresponding to the absolute rotation angle signal from the first position detector to the initial absolute position value. It is characterized by

Further, in the method for detecting the position of an industrial robot (claim 2) of the present invention, in the method of claim 1, the first
The absolute rotation angle signal from the position detector is digitized by a resolver/digital converter, outputted as position information, and converted into incremental pulses by a resolver/pulse converter.

Furthermore, in the method for detecting the position of an industrial robot (claim 3) of the present invention, in the method of claim 1, the first position detector, which is an absolute encoder, receives the incremental pulse in addition to the absolute rotation angle signal. It is characterized by having the function of generating.

[Function] In the method for detecting the position of an industrial robot according to the present invention (claim 1), first, when the power is turned on, coarse position information (rotation speed information) is obtained based on the absolute rotation angle signal from the second position detector. ) is obtained, and dense position information (angle information within one rotation) is obtained based on the absolute rotation angle signal from the first position detector.
is obtained. Then, absolute position initial values of the constituent members at the time of power-on are synthesized from these positional information. After this, when the industrial robot enters the normal operating state, the incremental pulses generated according to the absolute rotation angle signal from the first position detector are integrated, and the number of pulses is added to the initial absolute position value. Then, the absolute position of the component is detected. That is, immediately after the power is turned on, the absolute position of the component is detected by a multi-rotation absolute type detection method, and thereafter, the detection result is used as an initial value, and an incremental type detection method is adopted.

Furthermore, in the method for detecting the position of an industrial robot according to the present invention (claim 2), the resolver serving as the first position detector obtains an absolute rotation angle signal before the deceleration mechanism, and the detection result is transmitted to the resolver/digital converter. digitized by Furthermore, by converting the absolute rotation angle signal from the first resolver into pulses using the resolver/pulse converter, incremental pulses corresponding to the absolute rotation angle signal from the first resolver can be obtained.

Furthermore, in the method for detecting the position of an industrial robot according to the present invention (claim 3), in addition to obtaining an absolute rotation angle signal in front of the reduction gear using the absolute encoder serving as the first position detector, the absolute rotation angle signal from the first absolute encoder is Incremental pulses are obtained according to the absolute rotation angle signal obtained by this encoder.

[Embodiment of the Invention] Hereinafter, a device to which a position detection method for an industrial robot as an embodiment of the present invention is applied will be explained with reference to the drawings. Fig. 1 is a block diagram showing the circuit configuration thereof, and Fig. 2 is a FIG. 3 is a block diagram showing a modified example of the circuit configuration, and FIGS. 4 and 5 are sectional views of essential parts showing modified examples of the position detector arrangement. be.

First, a partial configuration of an industrial robot to which the method of the present invention is applied will be explained with reference to FIG. In Figure 2,
1 is an arm (component) of an industrial robot, 1a is a rotating shaft of the arm 1, 2 is a speed reduction mechanism such as a harmonic reducer connected to one end of the rotating shaft 1a, and 3 is a speed reduction mechanism connected to the arm via this speed reduction mechanism 2. 1, 3a is the rotating shaft of this drive motor 3, 4 and 5 are respectively provided between the deceleration mechanism 2, and the absolute rotation angle signals of the deceleration before deceleration side and the deceleration after side of the deceleration mechanism 2 are converted into a finite angle ( For example 180° or 36
The first resolver 4 is connected to the rotation shaft 3a of the drive motor 3, and the first resolver 4 is connected to the rotation shaft 3a of the drive motor 3, and the absolute It detects the rotation angle (dense position information), and the second
The resolver 5 is connected to the rotating shaft 1a protruding from the arm 1 to the side opposite to the reducer 2, detects the absolute rotation angle after deceleration of the reducer mechanism 2, and obtains information regarding the rotational speed of the rotation m3a (
rough location information).

The apparatus of this embodiment is constructed as shown in FIG. 1 in order to detect the absolute position based on the absolute rotation angle signals from the two resolvers 4 and 5 provided in this manner. In FIG. 1, reference numerals 6 and 7 are a common CO8 excitation circuit and a SIN excitation circuit that are always connected to the resolvers 4 and 5 to simultaneously excite these resolvers 4 and 5, and these CO8 excitation circuits 6. Resistors 9 and 10 are interposed between the SIN excitation circuit 7 and the resolver 5, respectively, and the impedance of the excitation phase of the second resolver 5 is set higher than the impedance of the excitation phase of the first resolver 4. There is.

Reference numeral 8 denotes a common R sum processing circuit that receives the absolute rotation angle signals from the resolvers 4 and 5 and performs waveform processing.
or 5 in sequence. 12 is 1 connected to the R sum processing circuit 8
This R/D converter 12 is composed of R/D converters (hereinafter referred to as R/D converters).
After processing, the absolute rotation angle signals from each resolver 4 and 5 are digitized and output in bits.

Furthermore, 13 is a resolver/pulse converter (hereinafter referred to as an R/P converter) which is attached to the R/D converter 12. The angle signal is converted into a two-phase pulse with a phase difference of 90' (A
pulse and B pulse). 14 is a pulse counter that counts up and down the A pulse and B pulse from the R/P converter 13, and 15 is an absolute position based on data (position information) from the R/D converter 12 and the pulse counter 14 as described below. Arithmetic units (CPU) that perform arithmetic processing and switching control of the switch 11, 15a to 15e are registers provided in the arithmetic unit 15, and the registers 15a, 15
b are resolvers 4 and 5 from the R/D converter 12, respectively.
It stores the absolute rotation angle signal (digital value) of
A general-purpose register, RAM, etc. in the arithmetic unit 15 are used. The registers 15c and 15d are for storing the absolute rotation angle signals (digital values) from the resolvers 4 and 5 at that time, respectively, with the basic posture of the robot (e.g., a vertically or horizontally determined angle) as a mechanical reference. A backed up version, ROM, etc. is used. The register 15e stores an initial absolute position value calculated as described later, and is constituted by a RAM or the like.

In this embodiment, the apparatus configured as described above detects the absolute position of the rotating shaft 3a of the motor 3 (that is, the position of the arm l) in the following manner.

First, when the power is turned on, the switch 11 is sequentially switched to the resolver 4 side/resolver 5 side by a command from the operation bag W15, and the R sum processing circuit 8 and the R/D converter 12
Then, the position information (absolute rotation angle signal) from each resolver 4, 5 is output as a parallel digital value to the arithmetic bag M15, and each digital value is input to the register 15a, 1.
5b. Here, for example, each resolver 4.5 is an LX resolver (one that generates one cycle of signals in one rotation), and the R/D conversion M12 has 4096 pulses (
12 bits)/1 period. In addition, when the robot's basic posture 1, for example, vertically and horizontally, is set as a mechanical standard, the absolute rotation angle signals (digital values) from each resolver 4 and 5 are P xot P za
(This work is normally performed by the manufacturer at the time of initial adjustment of the robot).

Now, when the power is turned on, the switch 11 is first switched to the resolver 4 side, the absolute rotation angle signal from the resolver 4 at that time is stored in the register 15a as a digital value P1, and then the switch 11 is switched to the resolver 5 side. is switched, and the absolute rotation angle signal from the resolver 5 at that time is stored in the register 15b as a digital value P2.

Based on the data stored in the registers 15a to 15d, the arithmetic unit i15 calculates the initial absolute position value at power-on as follows.

That is, the difference between the position information P2 from the second resolver 5 and the reference position information P2o is multiplied by the number N equivalent to the reduction ratio of the reduction mechanism 2 to obtain the number 4 equivalent to one cycle of the first resolver 4.
096, and the integer part is taken as the number of rotations or the number of cycles (coarse position information; upper bits of position information) of the second resolver 5. On the other hand, position information P from the first resolver 4
1 and the reference position information P1° is defined as position information within one rotation or one period (dense position information; lower bits of position information). This can be expressed as the following equation (1).

f((P2-P2O)XN/4096)X4096+(
P, −P, , ) −(1) However, f() represents the integer part of the numerical value in ().

In addition, p14p□. If the power is turned on nearby, f
(Pz Pz.)XN/
The decimal part of 4096 is corrected so that it almost matches (Pyo-P, .)74096. For example, the decimal part above is 0°9
5 and (pl-p□.)/4096 is 0.05, then as equation (1), f ((P2-P2o)X
N/4096)X4096+1 is used.

Now, the upper bits and lower bits of the position information obtained as described above are used as the initial value (absolute position initial value) of the position information when the power is turned on, and the register 15e of the arithmetic unit 15 is used.
After the first resolver 4 is stored in the R/D converter 12, the switch 11 is connected to the resolver 4 so that the first resolver 4 is always connected.
Switch to the side.

An R/P converter 13 attached to the R/D converter 12 outputs two-phase A and B incremental pulses in response to the absolute rotation angle signal from the resolver 4. The incremental pulses are constantly counted up and down by the pulse counter 14, and the count value and the initial value stored in the register 15e are summed by the arithmetic unit 15, and the summed result is output as position information.

Here, in this embodiment, the case where incremental pulses are obtained by the R/P converter 13 is explained, but
Without using the R/P converter 13 or pulse counter 14, the third
As shown in the figure, a counter 14A is provided between the R/D converter 12 and the arithmetic unit 15, and the digital value from the first resolver 4 is used as it is, and only the upper bits are processed by the counter 14A using carry and borrow signals. It may be configured to count.

In this way, according to this embodiment, immediately after the power is turned on, two
After creating an initial absolute position value using a multi-rotation absolute type inspection method using two resolvers 4 and 5, an incremental type inspection method is adopted from then on, and the dense position from the first resolver 4 is created. Since the absolute position is detected by adding the incremental pulses created based on the information to the initial value, the
There is no need to perform synthetic calculation processing on the absolute rotation angle signals from the two resolvers 4 and 5, the calculation load on the calculation device 15, etc. is reduced, and the detection accuracy can be greatly improved with a simple configuration.

In addition, in this embodiment, a common R/D converter 12 is used and switched for the two resolvers 4 and 5, and the excitation circuits 6 and 7 and the R sum processing circuit 8 are also used for the two resolvers 4 and 5. 5, it is no longer necessary to provide an R/D converter, an excitation circuit, an R sum processing circuit, etc. for each resolver 4, 5, and the electrical circuit required for the device is halved compared to the conventional one. In addition, the signal cable between the robot and the control panel used to require 3 pairs per resolver, but now requires only 4 pairs with 2 resolvers. , 5, it is sufficient to provide almost all the electric circuits required for one resolver, which greatly simplifies the device configuration and significantly reduces equipment costs.

By the way, by switching the two resolvers 4 and 5, the R/D
When connecting to the conversion $12 (R-sum processing circuit 8) or the excitation circuit 6.7, as in this embodiment, the excitation circuit 6.7 can be left connected to the two resolvers 4 and 5. The number of switches can be saved. Furthermore, dividing into two resolvers 4 and 5 on the robot side is extremely advantageous as it saves signal cables between the robot and the control panel.

At this time, unless the excitation circuits 6 and 7 are quite strong, the impedance of the excitation phase of the resolvers 4 and 5 will cause S
This will disturb the IN curve, COS curve, and their phases. In terms of circuitry, there is usually a volume or the like for adjusting this, and the impedance of this excitation phase usually changes depending on the position of the detector. When there is one resolver, the total error is considered in consideration of this impedance change, but when there are two resolvers as in this embodiment, the error may be doubled.

Therefore, in this embodiment, the second resolver 4 is
By increasing the winding impedance of the excitation phase of the first resolver 5 by about 2 to 3 times, and by interposing resistors 9 and 10 in series with the excitation phase of the second resolver 5, the excitation of the first resolver 4 is increased. Disturbances are kept to a minimum. As a result, the absolute rotation angle signal (fine position information) from the first resolver 4, which is directly involved in motor control and tends to include rotational irregularities and position errors, is protected from changes in the impedance of the second resolver 5, and is Detection accuracy can be obtained. As mentioned above, when a resistor 9.10 (or inductance) is inserted in the excitation phase, the output (detector) of the second resolver 5 becomes smaller, so it is necessary to adjust this by adjusting the transformation ratio, etc. You can also do this.

In the above embodiment, the case where the resolvers 4 and 5 are used as the two position detectors has been described, but an absolute encoder may also be used.

In this case, the excitation circuits 6 and 7 and the R/D converter 12 are not necessary (some kind of converter may be necessary depending on the signal format such as the Gray code). In this case, ■ one of the first absolute encoders that detects dense position information.
Create a carry-borrow signal at the rotation boundary, and use this to create the upper bit, or ■ to the first absolute encoder.

By providing a function of simultaneously generating incremental pulses in addition to the absolute rotation angle signal and counting these incremental pulses with a pulse counter, the same effects as in the above embodiment can be obtained.

Furthermore, in the above embodiment, as shown in FIG. 2, a case has been described in which the second resolver 5 is directly connected to the rotating shaft 1a of the arm 1, but the resolvers 4, 5 and the drive motor 3 are The arrangement may be as shown in FIG. 4 or FIG. 5. In these Figures 4-5, the resolver 4-5 is shown at the part where the lever (component) 24 for driving the robot's upper arm (not shown) is rotated via the link mechanism (not shown). An example of the arrangement of the drive motor 3 will be explained.

In the arrangement example shown in FIG. 4, in order to rotate the lever 24 around the Sl axis, the lever 24 is connected to the output shaft 2f of the reduction mechanism (harmonic reduction gear) 2 by a bolt 24a, and The input shaft 2d of the pulley 18. It is connected to the drive motor 3 via a timing belt 17 and a pulley 16.

Here, the pulley 18 is attached to the input shaft 2d, the pulley 16 is attached to the rotating shaft 3a of the drive motor 3, and the timing belt 17 is wound around these pulleys 16 and 18. As a result, the rotational driving force of the active motor 3 is applied to the pulley 16. Timing belt 17 and pulley 18
It is transmitted to the deceleration mechanism 2 via the deceleration mechanism 2, and after being decelerated by the deceleration mechanism 2, it is transmitted to the lever 24 through the output shaft 2f. The drive motor 3 is attached to the fixed side (frame) 39 of the robot, and the drive motor 3 is attached slightly in the vertical direction in FIG. 4 with respect to the fixed side 39 in order to adjust the tension of the timing belt 17. It can be moved freely and fixed at a position where the tension of the timing belt 17 is appropriate.

The drive motor 3 arranged in this way has its rotating shaft 3a.
A first resolver 4 is attached that generates an absolute rotation angle signal (fine position information on the input side of the deceleration mechanism 2) within a finite angle range.

Further, the speed reduction mechanism 2 includes a circular spline 2a, a wave generator 2b, and a flex spline 2C.
The circular spline 2a is connected to the bolt 3.
8 is attached to the fixed side 39, and the flex spline 2c is attached to the output shaft 2f. Further, the input shaft 2d of the reduction mechanism 2 is supported on the fixed side 39 by a bearing 35, and the output shaft 2f of the reduction mechanism 2 is supported on the fixed side 39 and the arm (lower arm) 23 by bearings 36 and 37, respectively. It is pivotally supported.

The input shaft 2d of the deceleration mechanism 2, which is provided coaxially with the output shaft 2f of the deceleration mechanism 2, has a hollow portion 2e formed along its axis. Further, one end side of a detection shaft 19 is fixed to the output shaft 2f by a bolt 19a, and this detection shaft 19 is guided to the input side of the deceleration mechanism 2 through the hollow part 2e of the input shaft 2d. The other end of the input shaft 19 is exposed outside the input shaft 2d.

The other end of the detection shaft 19, which is exposed to the outside, is connected via a coupling 34 to a rotor that generates an absolute rotation angle signal (coarse positional information on the output side of the deceleration mechanism 2) of the detection shaft 19 within a finite angle range. Two resolvers 5 are attached. This resolver 5 is fixed to a fixed side 39 by a resolver fixing fitting 33.

With such an arrangement, the lever 24 is connected to the deceleration mechanism 2.
The receiver is rotatably driven by the rotational driving force of the drive motor 3 decelerated by the output shaft 2f, and the upper arm is driven to swing through a link mechanism connected to the lever 24.

At this time, the rotation of the lever 24 is caused by the rotation of the output shaft 2 of the deceleration mechanism 2.
f and the detection shaft 19 to the outside (input side) of the input shaft 2d of the deceleration mechanism 2, and is transmitted to the second resolver 5 attached to the detection shaft 19. This results in
As the lever 24 rotates (swings the upper arm), coarse position information (absolute rotation angle signal) on the output side of the speed reduction mechanism 2 is obtained by the resolver 5. Further, detailed positional information on the input side of the speed reduction mechanism 2 is obtained by the resolver 4 in the same manner as in the above embodiment. Processing based on the position information obtained in this manner is performed in exactly the same manner as in the above embodiment.

According to the arrangement described above, the lengths of the reduction mechanism 2 and the drive motor 3 in the axial direction can be significantly shortened, especially compared to the arrangement shown in FIG. 2, and the device can be configured more compactly. There are advantages.

Note that the reference numeral 40 in FIG. 4 indicates an oil seal.

On the other hand, in the arrangement example shown in FIG. 5, unlike the arrangement shown in FIG.
are directly connected to each other, and furthermore, a rotating shaft 3a of this drive motor 3
and the rotating shaft 4a of the first resolver 4.

Therefore, in this arrangement example, the rotating shaft 3a of the drive motor 3 and the rotating shaft 4a of the first resolver 4 are provided along the axes of the rotating shafts 3a, 4a so as to communicate with the hollow portion 2e of the input shaft 2d. Hollow parts 3b and 4b are respectively formed in the detection shaft 19, one end of which is fixed to the output shaft 2f, and the detection shaft 19 is connected to the input shaft 2d.
Hollow portion 2e and hollow portions 3b, 4 of rotating shafts 3a, 4a
b to the outside of the first resolver 4 (deceleration mechanism 2
(input side), and is exposed to the other end of the detection shaft 19 or to the outside of the resolver 4.

The second resolver 5 is attached to the other end of the detection shaft 19 exposed to the outside via a coupling 34. Note that this resolver 5 has resolver fixing fittings 33.
is fixed to the outside of the first resolver 4.

With such an arrangement, position information can be obtained in a manner similar to that shown in FIG. Further, according to the arrangement shown in FIG. 5, the entire device can be constructed slimly, which is particularly advantageous when the present invention is applied to the wrist of a robot.

[Effects of the Invention] As detailed above, according to the method for detecting the position of an industrial robot (claims 1 to 3) of the present invention, two position detectors (
(resolver, absolute encoder) to create an initial absolute position value, and after that, the absolute position is detected by summing the initial value with the incremental pulses created based on the detailed position information from the first position detector. Since it is configured to do this, there is no need to constantly perform synthetic calculation processing on the absolute rotation angle signal, reducing the calculation load and significantly improving detection accuracy with a simple configuration.

[Brief explanation of drawings]

Figures 1 to 5 show a method for detecting the position of an industrial robot as an embodiment of the present invention. Figure 1 is a block diagram showing its circuit configuration, and Figure 2 shows its specific application part. A schematic configuration diagram, FIG. 3 is a block diagram showing a modified example of the circuit configuration, FIGS. 4 and 5 are sectional views of essential parts showing modified examples of the position detector arrangement, and FIGS. 6 and 7 are Both are block diagrams showing a synthesis processing circuit in a conventional position detection means using two absolute position detectors. In the figure, 1-arm (component member), 1a-rotating shaft,
2-Reduction mechanism, 2a-Circular spline, 2b-Wave generator, 2cm flex spline, 2
d-manpower shaft, 2e-hollow part, 2f-power shaft, 3-drive motor, 3a-rotation shaft, 3b-hollow part, 4-first resolver (first position detector), 4a-rotation shaft , 4b-hollow part, 5-second resolver (second position detector), 6-CO
3 excitation circuit, 7-5IN excitation circuit, 8-R sum processing circuit,
9.10--Resistor, 11-Switch (switching means). 12-Resolver/Digital (R/D) converter, 1
3-Resolver/Pulse (R/P) converter, 14°1
．． 4 A, --- Pulse counter, 15- Arithmetic unit (C
PU), 15 a to 15 e - register, 16 - pulley, 17 - timing belt, 18 - pulley, 19 detection axis, 19a - bolt, 23 - arm (lower arm), 24 -
Lever (component), 24 a-bolt, 33 resolver fixing metal fittings, 34-coupling, 35-37-bearing, 38-bolt, 39-fixed side (frame), 40-
Oil seal. Figure 2 Figure 4 Figure 6 Figure 7 Figure 1b

Claims (3)

[Claims] (1) A position detection method for detecting the absolute position of a component of an industrial robot, which is provided with first and second position detectors that generate absolute rotation angle signals on the before and after deceleration sides of a deceleration mechanism, respectively. Then, the absolute position initial value of the component is determined from the position information based on the absolute rotation angle signals from the first and second position detectors, and thereafter, the absolute rotation value from the first position detector is determined. A method for detecting a position of an industrial robot, characterized in that the absolute position of the component is detected by adding up incremental pulses corresponding to an angle signal to the initial value of the absolute position. (2) The absolute rotation angle signal from the first position detector, which is a resolver, is digitized by a resolver/digital converter, outputted as position information, and converted into the incremental pulse by the resolver/pulse converter. The method for detecting the position of an industrial robot according to claim 1. (3) The position of the industrial robot according to claim 1, wherein the first position detector, which is an absolute encoder, has a function of generating the incremental pulses in addition to the absolute rotation angle signal. Detection method.