JPH1039908A - Failure prediction method for industrial robots - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To provide a failure prediction method for an industrial robot that can perform diagnosis at any time even during operation of a production line. SOLUTION: Data of each axis i constituting a robot mechanism is taken in from a servo system for controlling a servomotor, and a load torque in consideration of interference of other axes is used by a robot itself and a mass model of the load and data of each axis i. after calculating the T _o, to calculate the load torque T _o and the angular velocity ω _i and the load side of the work rate W _o is the product of. Moreover, calculating the driving torque T _i by the data of the drive current I _i of the servo motor for driving each axis i uptake than the servo system, multiplying the torque constant k _i of the servo motor corresponding thereto to the driving current I _i After that, the drive-side power W _i , which is the product of the drive torque T _i and the angular velocity ω _i , is calculated. Then, a ratio of the work rate (W _i / W _o) from work rate W _i of work rate W _o and the drive side of the calculated load side, the robot mechanism by comparing the preset determination value this It is determined whether or not the part has deteriorated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for predicting a failure of a robot mechanism of an industrial robot driven by a servomotor without depending on the contents of the operation during operation of a production line.

[0002]

2. Description of the Related Art In an industrial robot, the gears and the like constituting a driving force transmission system of a robot mechanism mainly composed of an arm and a driving motor thereof are worn due to long-term use, thereby adversely affecting the operation and control accuracy. Will come.
Generally, such phenomena cannot be easily predicted from the outside,
In addition, since it is apt to be left unattended until an adverse effect appears, in the past, countermeasures were often taken after a situation where a gear or the like was damaged occurred.

In order to deal with this, for example, Japanese Patent Application Laid-Open
In -123105, when the robot is in a normal state,
For example, at the time of new introduction of the robot body or at the end of regular maintenance, the robot is actually operated according to a preset reference operation pattern, the operation data at this time is measured in advance as reference data, and this reference data is It is recorded in the storage device of the control device. Then, after a predetermined operation time has elapsed, the robot is actually operated again in the same manner as the reference operation pattern, and the operation data at this time is measured. And the difference is diagnosed based on the difference. As this operation data,
In -123105, since a servo control deviation (deviation) is used, it is said that there is no need to provide a dedicated sensor for diagnosing the presence or absence of a failure, and a mechanical system such as an arm driving mechanism and a driving mechanism It is said that there is an advantage that it is possible to comprehensively diagnose a control system such as a servo system for controlling the system.

[0004]

However, Japanese Patent Application Laid-Open Publication No.
The method of 3-123105 has the following problems. The first point is that the operation data in the normal state of the robot is measured for each robot as reference data, and this reference data needs to be stored. The second point is that the diagnosis cannot be performed during the operation of the production line because the robot must be operated with a reference operation pattern in the diagnosis process. The third point is
It is difficult to improve the accuracy of diagnosis for a failure that is specific to an actual operation pattern during the operation of the production line.

The present invention has been made to solve these problems, and predicts a failure of an industrial robot driven by a servomotor without depending on the contents of the operation during operation of a production line. Providing a method, in particular, without previously measuring operation data in an operation pattern that is a reference in a normal state of the robot as reference data, and can perform a diagnosis at any time even during operation of a production line, It is an object of the present invention to provide a method for predicting a failure of an industrial robot capable of performing a highly accurate diagnosis even for a failure specialized in an actual operation pattern during the operation of a production line.

[0006]

In order to achieve the above object, according to the present invention, a power based on the actual position (angle), angular velocity, and angular acceleration of each drive shaft calculated from a mass point model, that is, the power, load side of the work rate W _o, and more work rate W _i of the driving side is calculated from the operation command to the drive shaft, it obtains the ratio (W _o / W _i) or the difference (W _o -W _i), By comparing this with the determination value, the presence or absence of the deterioration of the robot mechanism and its level are evaluated.

Generally, the actual performance of an operation command to a drive shaft is referred to as followability. Particularly, the quality of the followability during acceleration / deceleration greatly affects the performance of the robot mechanism. As the deterioration of the robot mechanism progresses due to aging, the gears and the like that constitute the driving force transmission system are worn, thereby deteriorating the responsiveness particularly during acceleration and deceleration. When the followability during acceleration / deceleration deteriorates, the power W _i on the drive side is reduced.
The difference between the load side work rate W _o comes appear larger. In general, during acceleration, the power W _i on the drive side appears larger than the power W _o on the load side, and conversely, during deceleration, the power W _o on the load side becomes higher than the power W _i on the drive side. Appears to grow larger.

Therefore, in the present invention, the power W _i on the drive side is obtained.
And by comparing the preset determination value ratio or difference between the load side of the work rate W _o, and to determine the follow-up of the good or bad, thereby evaluating the level and the presence or absence of degradation of the robot mechanism It was decided to. When the deterioration of the robot mechanism is small, the difference between the power W _i on the drive side and the power W _o on the load side should be small, and the power ratio (W
judgment value in using _o / W _i) is the lower limit value 1-beta as well as its upper limit 1 + beta if the allowable value and beta, the ratio of the work rate (W _o / W _i) is the range If it is within the range, the deterioration of the robot mechanism is determined to be small. On the other hand, when the difference in power (W _o −W _i ) is used, if the allowable value is β, + β is set to the upper limit and −β is set to the lower limit. _{If o−} W _i ) is within this range, it is determined that the deterioration of the robot mechanism is small.

The power W _i on the drive side and the power W on the load side
_As a method of calculating _{o, when} a servomotor is used as the operating mechanism of the drive shaft, the power W _i on the drive side is calculated from the drive current I _i supplied to the servomotor, and the power W _i on the load side is calculated. _o is the actual position (angle data) of the drive shaft detected by an encoder (position detector) attached to or attached to the servomotor, and angular velocity data and angular acceleration data obtained based on this angle data. Is calculated by the following formula.

That is, in the invention according to the first aspect, each data of the angle θ _i , the angular velocity ω _i , and the angular acceleration α _i of each drive shaft i constituting the robot mechanism is fetched from the servo system for controlling the servo motor. , a motion equation relating to the mass point model of the robot mechanical unit including the load, and calculates an angle theta _i, the angular velocity omega _i, and the angular acceleration alpha _i and the load torque T _o, and thereafter, the load torque T _o and the angular velocity omega _i to calculated the load side of the work rate W _o is the product. Also, each drive shaft i
The data of the drive current I _i of the servo motor for driving the uptake from the servo system, the driving torque T _i by multiplying the torque constant k _i of the servo motor to the drive current I _i
Then, the power W _i on the driving side, which is the product of the driving torque T _i and the angular velocity ω _i , is calculated. Then, a power ratio (W _o / W _i ) is determined from the calculated load-side power W _o and drive-side power W _i, and the power ratio (W W / W _i ) is obtained.
_o / W _i ) is compared with a predetermined judgment value to evaluate whether or not the robot mechanism has deteriorated and its level (claim 1).

With the above configuration, the data of the angle θ _i , the angular velocity ω _i , and the angular acceleration α _i of each drive shaft i constituting the robot mechanism are fetched from a servo system for controlling a servo motor. calculating a work rate W _o from these data servomotor or load side, whereas the data of the drive current I _i of the servo motor for driving the drive shafts i, similarly uptake than servo system of the robot from the data The power W _i on the driving side is calculated, and the ratio of the two powers (W
_o / W _i ) is compared with a predetermined judgment value, so that the power that changes every moment according to the actual operation pattern is used for the judgment, so that the robot can be constantly operated even when the production line is operating. It is possible to evaluate whether or not the mechanical section has deteriorated and its level. Since the difference between the work rate W _i on the load side of the work rate W _o and the driving side in the deceleration time as the degree of degradation of the robot mechanism is large increases, the ratio of the work rate (W _o / W _i) is 1 The closer to, the lower the level of deterioration of the robot mechanism, and conversely, the further away from 1, the higher the level of deterioration of the robot mechanism.

According to the first aspect of the present invention, the power ratio (W _o / W) is calculated based on the power W _o on the load side and the power W _i on the drive side.
_i ) and instead of comparing this power ratio (W _o / W _i ) with a predetermined determination value, the power ratio W _o on the load side and the power W _i on the drive side are used instead. to calculate the difference between the work rate (W _o -W _i), the difference in the work rate (W _o -
W _i ) may be compared with a predetermined judgment value (claim 2). Since the difference between the work rate W _i on the load side of the work rate W _o and the driving side in the deceleration time as the degree of degradation of the robot mechanism is large becomes large as described above, the difference in work rate (W _o -W _The closer the value of _i ) is to 0, the lower the level of deterioration of the robot mechanism is. The farther from _i is, the higher the level of deterioration of the robot mechanism is.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a block diagram showing a servo system for controlling the robot drive unit of the present invention. In the figure, 1 is a command position calculation unit, 2 is a position loop, 3 is a speed loop, 4 is a current loop, 5 is a servo motor, etc. Is a motor for driving the robot, 6 is an encoder, and 7 is a deterioration diagnosis calculation unit. S in the figure is a Laplace operator, which indicates differentiation.

The control signal output from the command position calculation unit 1 is transmitted to a position loop 2, a speed loop 3, and a current loop 4, converted into a current signal in the current loop 4, and output from the current loop 4. Drives the motor 5. The angle θ _i of the drive shaft, which is the position data detected by the encoder 6, is calculated by the Laplace operator S to obtain the angular velocity ω _i
And angular acceleration α _i , and the angle θ _i is
The angular velocities ω _i are respectively fed back to the speed loop 3 and also transmitted to the deterioration diagnosis calculation unit 7 as data necessary for deterioration diagnosis. The current output from the current loop 4 is detected by a current detector provided on a transmission path between the current loop 4 and the motor 5, and the detected current I _i
Is returned to the current loop 4 and is also transmitted to the deterioration diagnosis operation unit 7 as data necessary for the deterioration diagnosis. As described above, the block diagram of the servo system shown in FIG. 2 is different from the conventional one in that the deterioration diagnosis operation unit 7 and the data transmission path to the deterioration diagnosis operation unit 7 are added.

FIG. 1 is a flow chart showing a flow of processing performed by the deterioration diagnosis calculation unit 7 shown in FIG. 2 to evaluate the presence or absence of deterioration of the robot mechanism and its level. In step 11, as described with reference to FIG. 2, the angle θ _i of each drive shaft detected by the encoder 6, and the angular velocity ω _i and the angular acceleration α _i converted by the Laplace operator S based on the angle θ _i. Is taken into the deterioration diagnosis operation unit 7.

In step 12, a motion equation relating to the mass model of the robot mechanism including the load and step 11
In captured angle theta _i, calculates an angular velocity omega _i, and the angular acceleration alpha _i and the load torque T _o. here,
It will be described more specifically the process of calculating the load torque T _o. FIG. 3 shows an example of a mass model in a vertical articulated robot having a six-axis configuration. In the figure, reference numeral 21 denotes a mass model of the load, and reference numerals 22 to 26 denote mass models of the second to sixth drive shafts. Reference numerals 31 to 36 denote first to sixth drive shafts. With respect to this mass model, the torque acting on the load side of each drive shaft, that is, the load torque, is calculated by obtaining each term of the equation of motion represented by the equation (1).

[0017]

(Equation 1)

In the equation (1), the left side is a torque matrix representing the load torque acting on each axis. The first term on the right side represents torque due to inertial force, the second term represents torque due to centrifugal force and Coriolis force, and the third term represents torque due to unbalance force. Since the first to third terms are calculated from the angle θ _i , angular velocity ω _i , and angular acceleration α _{i of} each drive shaft, the load torque obtained as the sum of the first to third terms is calculated from the other axes. It takes into account the reaction force received. In addition,
Equation (1) is a general equation of motion for a mass model of a robot, and is not an equation specific to the present invention.

[0019] In step 13, by multiplying the angular velocity omega _i to the load torque T _o calculated in step 12, calculates the load side of the work rate W _o.

Meanwhile, in parallel with the calculation of the load side of the work rate W _o in steps 12 and 13 described above, it calculates the work rate W _i of the drive side at step 14 and 15.
In step 14, by multiplying the transmission line captures the current I _i for each drive shaft detected by the current detector provided on the torque constant k _i of the motor between the current loop 4 and the motor 5, the drive The torque acting on the drive side of the shaft, that is, the drive torque _Ti is calculated. In step 15, the driving torque T _i calculated in step 14 is added to the angular velocity ω.
By multiplying by _i , the power W _i on the driving side is calculated.

In step 16, the power ratio (W _o) is calculated based on the load-side power W _o calculated in step 13 and the driving-side power W _i calculated in step 15.
/ W _i ). In step 17, step 16
By comparing the power ratio (W _o / W _i ) calculated in the above with a predetermined determination value, the presence or absence of the deterioration of the robot mechanism and the level thereof are determined. Specifically, 1 + β is set as the upper limit of the determination value and 1−β
Is the lower limit, and if the power ratio (W _o / W _i ) is within this range, it is determined that the deterioration of the robot mechanism is small. Here, β is a value set as an allowable value of deterioration.

[0022] In addition, the magnitude of the ratio of the work rate (W _o / W _i), also can evaluate the level of deterioration, the ratio of the work rate (W _o /
The closer the W _i ) is to 1, the lower the level of degradation, and conversely, the farther away from 1, the greater the level of degradation. Normally, the level of this deterioration is small when a new robot body is newly introduced or when regular maintenance is completed, and gradually increases with use. Therefore, if the allowable value β is set to a value smaller than a value that causes trouble in robot operation, Further, the occurrence of a failure can be predicted before the failure due to aging occurs in the robot mechanism.

By the way, at step 16-17, and calculates the ratio of the load side of the work rate W _o and the drive side of the work rate W _i from work rate (W _o / W _i), the ratio of the work rate (W _o /
W _i ) is compared with a predetermined judgment value. Instead, the difference in power (W _o −W _i ) from the power W _o on the load side and the power W _i on the drive side is used instead. is calculated, it may be compared with a preset determination value this difference in work rate (W _o -W _i). Specifically, + β is set as the upper limit value of the determination value and −β is set as the lower limit value. If the difference in the power (W _o −W _i ) is within this range, the deterioration of the robot mechanism is determined to be small. I do. Similarly, the magnitude of the difference between the work rate (W _o -W _i), the level of degradation can also be evaluated, the difference in work rate (W _o -W _i) is higher the level of degradation is small near 0, conversely The further away from zero, the greater the level of degradation.

Since the power ratio and the power difference are based on the power W _i on the driving side, which is input to the robot mechanism, (W _o / W _i ) and (W _o −), respectively. W _i ), but need not be limited to this. That is, the ratio of power (W _o / W _i ) is obtained by exchanging the numerator and denominator (W _i / W _i / W _i ).
W _o ), and similarly, the difference in power (W _o)
-W _i) may be compared with the determination value as (W _i -W _o).

As described above, according to the present embodiment, FIG.
The diagnostic process described in (1) can be performed at any time during normal operation of the production line, so it is necessary to measure the operation data in the normal state of the robot as reference data for each individual robot as in the conventional technology. Disappears,
There is no need to stop the production line for failure diagnosis. In addition, since fault diagnosis can be performed during normal production line operation, it is possible to perform high-precision diagnosis even for failures that are specific to actual operation patterns during production line operation. Now you can. Further, the diagnosis of the prior art is a regular diagnosis method, whereas the diagnosis of the present embodiment is a so-called constant monitoring method in which the diagnosis is performed even during the operation of the production line. Even a sudden malfunction can be immediately detected.

[0026]

According to the present invention, in a method for predicting a failure in a robot mechanism driven by a servomotor, an angle θ _i of each drive shaft i constituting the robot mechanism is calculated.
Angular velocity omega _i, and the respective data of the angular acceleration alpha _i, uptake from the servo system for controlling the servo motor, to calculate the work rate W _o of the servo motor or load side from these data, while driving the respective drive shafts i the data of the drive current I _i of the servo motor, similarly uptake than the servo system, to calculate the work rate W _i of the drive side of the robot from the data, both work rate ratio (W _i / W _o) or difference ( since W _i -W _o) it was set to be compared with a preset determination value, by utilizing the determined work rate that changes every moment depending on the actual operation pattern, a reference in a normal state of the robot Diagnosis can be performed at any time during operation of the production line without previously measuring the operation data in the operation pattern as reference data. It has become possible to perform diagnosis in high precision the failure as specializing in the operation pattern.

[Brief description of the drawings]

FIG. 1 is a flowchart showing a procedure for diagnosing a robot mechanism in one embodiment of the present invention.

FIG. 2 is a block diagram of a servo system that controls a robot driving unit according to an embodiment of the present invention.

FIG. 3 is a diagram illustrating an example of a mass model in a vertical articulated robot having a six-axis configuration.

[Explanation of symbols]

 5 Servo motor

Claims (2)

[Claims] 1. A method for predicting a failure in a robot mechanism driven by a servomotor, wherein the data of the angle θ _i , the angular velocity ω _i , and the angular acceleration α _i of each drive shaft i constituting the robot mechanism are transferred to the servo motor. uptake from servo system for controlling to calculate the motion equation relating mass point model of the robot mechanical unit including the load, the angle theta _i, the angular velocity omega _i, and by the angular acceleration alpha _i load torque T _o, and thereafter, the calculating the load torque T _o and the angular velocity omega _i and the load side of the work rate W _o is the product of the uptake from the servo system data of the drive current I _i of the servo motor for driving the drive shafts i, the drive current I _i is the torque constant k _i of the servomotor
Calculating a driving torque T _i by multiplying, thereafter, calculates the drive torque T _i and the angular velocity omega _i and work rate W _i of the driving side is the product of the load side of the work rate W _o and drive The power ratio (W _o / W _i ) is calculated from the power W _i on the side of the robot, and the ratio of the power (W _o / W _i ) is compared with a predetermined judgment value to thereby obtain the robot mechanism section. A method for predicting the failure of an industrial robot, wherein the presence or absence of deterioration of the robot is determined. 2. The method of claim 1, to calculate the ratio of the load side of the work rate W _o and the drive side of the work rate W _i from work rate (W _o / W _i), the ratio of the work rate (W _o / Instead of comparing W _i ) with a predetermined judgment value, the difference in power (W _o −W) from the power W _o on the load side and the power W _i on the drive side.
_i) calculates, failure prediction method of an industrial robot, characterized in that so as to compare the difference of the work rate (W _o -W _i) preset determination value.