Transmission device abnormality detection method and abnormality detection device

Abstract

To make it possible to detect with a small amount of calculation the occurrence of an abnormality in a transmission device such as a reduction gear or a driving unit that is provided in a robot and the like.SOLUTION: An abnormality detection device is for detecting an abnormality in a transmission device such as, for example, a reduction gear 38 driven by a motor 36. The abnormality detection device has: a bandpass filter 44 for performing bandpass filter processing on a detected value of the speed of the motor 36; and determination means (comparator 46) for determining presence or absence of the occurrence of an abnormality of the transmission device on the basis of a signal obtained by the bandpass filter processing. Between the bandpass filter 44 and the determination means, an RMS calculation unit 45 may be provided to calculate a root mean square (RMS) value of the signal output from the bandpass filter 44. The bandpass filter 44 is, for example, a fifth-order Butterworth filter.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a method and device for detecting an abnormality in a transmission device such as a reducer or a traveling drive unit used to transmit rotation from a motor to a load or the like. [Background technology]

[0002] In industrial robots, each axis is equipped with a servo motor, and the driving force is transmitted to each axis via a reducer attached to the servo motor. To move the robot horizontally in a straight line on a floor or other surface, a traveling drive unit with a rack-and-pinion structure is used, and the traveling drive unit is also driven by a servo motor. Both the reducer and the traveling drive unit are transmission devices that transmit the driving force from the servo motor to the target object. A transmission device is composed of, for example, a combination of gears. If an abnormality occurs, such as a loose bolt securing the gear to the shaft or a chipped tooth, vibrations may occur during high-speed rotation or, in robot applications, a decrease in positioning accuracy may occur. Therefore, there is a need for the ability to detect abnormalities in the transmission device early.

[0003] Patent Document 1 discloses a technology for determining whether a reducer is abnormal by acquiring multiple pieces of operational data from the operation of the reducer as current operational data, calculating a correlation coefficient between past operational data and the current operational data in an abnormal mode that indicates the type of abnormality in the reducer, and determining whether an abnormality exists and the cause of the abnormality. Examples of operational data that can be used include the average, maximum, amplitude, and standard deviation of the command current value to the servo motor, as well as the motor's rotation speed and rotation angle. This method requires statistical processing after acquiring a wide variety of operational data to achieve sufficient determination accuracy, resulting in a large overall calculation volume. Furthermore, collecting past operational data in an abnormal mode also requires a lot of effort.

[0004] Patent Document 2 discloses a technique for determining whether or not there are signs of failure in a reducer in a robot or the like, by determining the peak value of the amplitude of frequency components in a specific frequency band in the motor current during acceleration / deceleration of the servo motor, and determining whether or not there are signs of failure based on this peak value. The specific frequency band used for the determination is determined according to the natural frequency that causes the robot to vibrate due to resonance in the direction in which the gears of the reducer mesh, out of multiple mechanical natural frequencies of the robot. The frequency components in the specific frequency band are extracted using FFT (fast Fourier transform) calculations or a bandpass filter. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2019-18270 A [Patent Document 2] Patent Publication No. 2021-9041 Summary of the Invention [Problem to be solved by the invention]

[0006] The method described in Patent Document 2 detects the presence or absence of an abnormality in a reducer based on the motor current during acceleration and deceleration, and therefore has the problem that the detection sensitivity is easily affected by the servo gain setting value. Furthermore, if an FFT calculation is used to extract frequency components in a specific frequency band, the amount of calculation increases. If a bandpass filter is used to extract frequency components in a specific frequency band, the detection sensitivity is expected to vary depending on the configuration of the bandpass filter. For example, if the order of the bandpass filter is low, it will not be possible to sufficiently remove the influence of the DC component of the position command value. However, Patent Document 2 does not clarify the configuration of a bandpass filter suitable for detecting abnormalities in a reducer.

[0007] An object of the present invention is to provide a method and apparatus that can reliably detect the occurrence of an abnormality in a transmission device provided in a robot or the like with a small amount of calculation. [Means for solving the problem]

[0008] The abnormality detection method of the present invention is a method for detecting an abnormality in a transmission device driven by a motor, which detects the motor speed, performs band-pass filtering on the detected speed value, and determines whether or not an abnormality has occurred in the transmission device based on the signal obtained by the band-pass filtering. In the abnormality detection method of the present invention, by performing band-pass filtering on the detected motor speed value, it is possible to more reliably detect the occurrence of an abnormality in the transmission device, as will be described later. In the present invention, the transmission device is, for example, a reducer or a traveling drive unit connected to a motor.

[0009] In the anomaly detection method of the present invention, the band-pass filtering is performed by a fifth-order Butterworth filter. do. When a third-order Butterworth filter is used, it becomes susceptible to the influence of unnecessary frequency components, and when a seventh-order or higher Butterworth filter is used, it becomes difficult to implement and the amount of calculation increases. Therefore, bandpass filtering using a fifth-order Butterworth filter is optimal.

[0010] In the anomaly detection method of the present invention, it is preferable to determine whether an anomaly has occurred by calculating the root mean square or root mean square of the signal obtained by band-pass filtering and comparing it with a threshold value. By performing the determination in this manner, it is possible to reliably detect an anomaly while reducing the amount of calculation required for the determination.

[0011] In the abnormality detection method of the present invention, the motor speed can be detected from a signal obtained from an encoder connected to the motor. If the motor is subject to servo control, the motor is generally equipped with an encoder that detects the motor's position, so the motor speed can be easily obtained by, for example, performing differential processing on the signal from the encoder.

[0012] The abnormality detection device of the present invention is an abnormality detection device that detects an abnormality in a transmission device driven by a motor, and includes a band-pass filter that performs band-pass filtering on the detected value of the motor speed, and a determination means that determines whether or not an abnormality has occurred in the transmission device based on the signal output from the band-pass filter. By providing the band-pass filter and performing band-pass filtering on the detected value of the motor speed, the abnormality detection device of the present invention can more reliably detect the occurrence of an abnormality in the transmission device, as will be described later.

[0013] In the anomaly detection device of the present invention, the bandpass filter is a bandpass filter consisting of a fifth-order Butterworth filter. When a third-order Butterworth filter is used, it becomes susceptible to the influence of unnecessary frequency components, and when a seventh-order or higher Butterworth filter is used, it becomes difficult to implement and the amount of calculation increases. Therefore, a bandpass filter consisting of a fifth-order Butterworth filter is optimal.

[0014] In the anomaly detection device of the present invention, it is preferable to further provide a calculation means for calculating the root mean square or root mean square of the signal output from the band-pass filter, and to have the determination means determine whether an anomaly has occurred by comparing the value obtained by the calculation means with a threshold value. By comparing the root mean square or root mean square with the threshold value, an anomaly can be reliably detected while reducing the amount of calculation required to determine whether an anomaly has occurred.

[0015] The abnormality detection device of the present invention may further include speed calculation means for detecting the motor speed from a signal obtained from an encoder connected to the motor. If the motor is subject to servo control, the motor is generally equipped with an encoder for detecting the motor position, and therefore the motor speed can be easily obtained by providing speed calculation means for performing, for example, differential processing on the signal from the encoder.

[0016] The abnormality detection device of the present invention may be provided inside a control device, such as a robot controller, that executes servo control of a motor based on an externally provided command value. By providing the abnormality detection device inside the control device, it becomes possible to detect the occurrence of an abnormality in the transmission device in conjunction with the servo control of the motor by the control device. [Effects of the Invention]

[0017] According to the present invention, occurrence of an abnormality in a transmission device provided in a robot or the like can be reliably detected with a small amount of calculation. [Brief explanation of the drawings]

[0018] [Figure 1] 1A is a schematic diagram showing the overall configuration of a robot to which an anomaly detection method according to the present invention is applied, and FIG. 1B is a front view of the robot. [Figure 2] (a) and (b) are side and top views of the robot, respectively. [Figure 3] 1 is a block diagram showing a configuration of a control device including an abnormality detection device according to an embodiment of the present invention; [Figure 4] FIG. 4 is a waveform diagram illustrating detection of an abnormality in the reducer. [Figure 5] FIG. 4 is a waveform diagram illustrating detection of an abnormality in the reducer. [Figure 6] FIG. 10 is a waveform diagram illustrating detection of an abnormality in the traveling drive unit. [Figure 7] FIG. 10 is a waveform diagram illustrating detection of an abnormality in the traveling drive unit. [Figure 8] FIG. 10 is a waveform diagram illustrating detection of an abnormality in the traveling drive unit. DETAILED DESCRIPTION OF THE INVENTION

[0019] Next, an embodiment of the present invention will be described with reference to the drawings. The present invention relates to detecting an abnormality in a transmission device driven by a motor. The transmission device is, for example, a reducer or a traveling drive unit, and is incorporated into an industrial robot. First, a robot, which is an example of equipment into which a transmission device is incorporated, will be described with reference to FIGS. 1 and 2. In FIG. 1, (a) is a schematic diagram showing the overall configuration of the robot 10, and (b) is a front view of the robot 10. In FIG. 2, (a) and (b) are a side view and a plan view of the robot 10, respectively.

[0020] The illustrated robot 10 is configured as a horizontal articulated robot intended to transport a workpiece 50 such as a glass substrate. In particular, the robot 10 is configured as a so-called double-hand robot equipped with two hands 13A and 13B, each of which holds a workpiece 50 (not shown in FIG. 1). The robot 10 is electrically connected via a cable to a control device (robot controller) 40, which receives external operation commands for the robot 10 and drives and controls the robot 10 based on these operation commands. As will be described later, the control device 40 is equipped with an abnormality detection device 42 (see FIG. 3) that detects abnormalities in the transmission device within the robot 10. Note that FIG. 1(a) shows the horizontal articulated mechanism of the robot 10 in a raised state, while FIGS. 1(b) and 2(a) and (b) show the horizontal articulated mechanism in a lowered state.

[0021] The robot 10 includes a base 22 that can move on a pair of parallel rails 21 that are linearly arranged on a floor surface FL (see FIG. 1(b)), a rotating table 23 that is arranged on the base 22 and rotates in a horizontal plane around a rotation axis 31, and an elevating mechanism 24 that is arranged upright relative to the rotating table 23. The rails 21 are, for example, rack rails that mesh with a pinion that is arranged on the base 22. The pinion is mechanically connected to a travel motor that is arranged in the base 22. Driving the travel motor rotates the pinion, causing the robot 10 to move horizontally along the rails 21. A gear for transmitting driving force may be arranged between the travel motor and the pinion, and these gears and pinion form a travel drive unit of the robot 10. A cover 25 that covers the rails 21 is attached.

[0022] In order to rotate the rotating table 23 around the rotation axis 31 relative to the base 22, a motor also referred to as a θ-axis motor is provided on the base 22 in addition to the travel motor. A reducer is attached to the θ-axis motor, and the rotating table 23, which is a load on the θ-axis motor, is driven by the θ-axis motor via the reducer. The lifting mechanism 24 includes a fixed part 24A attached to the rotating table 23 and a moving part 24B that moves up and down relative to the fixed part 24A by a motor. A motor that lifts and lowers the moving part 24B is provided on the fixed part 24A, and this motor is also attached with a reducer, and the moving part 24B is driven by the lifting motor via the reducer to move up and down.

[0023] Moving unit 24B is provided with arm support units 26A and 26B extending horizontally, each holding a horizontal multi-joint mechanism, and each arm support unit 26A and 26B has a horizontal multi-joint mechanism attached to its tip. The upper horizontal multi-joint mechanism includes a first arm 11A attached to arm support unit 26A and rotatable in a horizontal plane around a common axis 32, and a second arm 12A attached to the tip of first arm 11A and rotatable in a horizontal plane around axis 33A, with hand 13A attached to the tip of second arm 12A. Similarly, the lower horizontal multi-joint mechanism includes a first arm 11B attached to arm support unit 26B and rotatable in a horizontal plane around common axis 32, and a second arm 12B attached to the tip of first arm 11B and rotatable in a horizontal plane around axis 33B, with hand 13B attached to the tip of second arm 12B.

[0024] The hands 13A and 13B are fork-shaped with multiple rod-like members arranged in parallel so that they can transport the plate-shaped workpiece 50 while maintaining it in a horizontal position by holding it from below. That is, the hands 13A and 13B are designed to hold the workpiece 50 on the hands 13A and 13B. The hands 13A and 13B move forward or backward relative to the workpiece 50 when taking out the workpiece 50 stored in a load lock chamber or a cassette, etc., and holding it on the hands 13A and 13B, or when storing the held workpiece 50 in a load lock chamber, etc., and the direction in which the hands 13A and 13B move forward or backward is parallel to the direction in which the rod-like members extend.

[0025] In this robot, the horizontal articulated mechanism is configured such that the hands 13A and 13B move forward and backward in a linear motion perpendicular to the direction of extension of the arm support 26, using link mechanisms incorporated into the first arms 11A and 11B and the second arms 12A and 12B. That is, both hands 13A and 13B move forward and backward in the same direction. Movement of the tips of the hands 13A and 13B away from the central axis 32 is forward motion, and movement in the opposite direction to the forward motion is backward motion. The first arms 11A and 11B and the second arms 12A and 12B perform bending motion as a whole. To maintain the orientation of the hands 13A and 13B in a constant horizontal plane, the hands 13A and 13B are attached to the tips of the second arms 12A and 12B so that they can rotate in a horizontal plane around wrist axes 34A and 34B, respectively. In the upper horizontal multi-joint mechanism, the first arm 11A and the second arm 12A are driven by a motor provided in the arm support unit 26 via a reducer, and the hand 13A moves in a direction perpendicular to the extension direction of the arm support unit 26 while maintaining its orientation. Similarly, in the lower horizontal multi-joint mechanism, the first arm 11B and the second arm 12B are driven by a motor provided in the arm support unit 26 via a reducer, and the hand 13B moves in a direction perpendicular to the extension direction of the arm support unit 26 while maintaining its orientation. In this robot, the hands 13A and 13B can be moved forward and backward independently. In the following description, the forward and backward movement of the hands 13A and 13B by the link mechanisms incorporated in the first arms 11A and 11B and the second arms 12A and 12B is referred to as arm extension and retraction movement.

[0026] Ultimately, the movements of the robot 10 can be divided into horizontal movement along the rail 21 (this is referred to as movement of the X-axis or travel axis), rotation of the lifting mechanism 24 relative to the base 22 around the rotation axis 31 (this is referred to as movement of the θ-axis or pivot axis), forward and backward movement of the hands 13A, 13B, i.e., extension and contraction movement of the arms (this is referred to as movement of the R-axis), and raising and lowering of the arm support part 26 by the lifting mechanism 24 (this is referred to as movement of the Z-axis), each of which is driven by a motor for each axis provided on the robot 10. The robot 10 executes an operation of moving only one of these axes or an operation of moving two or more axes simultaneously, under drive control from the control device 40 in response to a command corresponding to the movement operation.

[0027] The following describes the transportation of a workpiece 50, such as a glass substrate, using a robot 10. Consider the case where two load-lock chambers are provided along a rail 21 and the workpiece 50 is transported between these load-lock chambers. First, with both hands 13A and 13B retracted and the first arms 11A and 11B and the second arms 12A and 12B folded, i.e., with the arms retracted, the robot 10 is moved along the rail 21 to a position facing the unloading load-lock chamber. At the same time, the lifting mechanism 24 raises and lowers the hand 13A to a height corresponding to the load-lock chamber. The arm is then extended to advance the hand 13A into the unloading load-lock chamber, and the hand 13A is then slightly raised to place the workpiece 50 on the hand 13A. The arm is then retracted to retract the hand 13A, and the hand 13A, along with the workpiece 50, is withdrawn from the load-lock chamber. Next, with the workpiece 50 loaded, the robot unit 10 is moved along the rail 21 to a position facing the load-lock chamber on the loading side. At the same time, the lifting mechanism 24 raises and lowers the arm support unit 26A so that the height of the hand 13A corresponds to the height of the load-lock chamber. Then, with the workpiece 50 loaded, the hand 13A is advanced to enter the second cassette, and the hand 13A is slightly lowered to place the workpiece 50 in the load-lock chamber on the loading side. The hand 13A is then retracted to withdraw the hand 13A from the load-lock chamber. The workpiece 50 is transported through the above operations. If the load-lock chamber on the unloading side is located on one side of the rail 21 and the load-lock chamber on the other side, in addition to the above operations, a rotation operation around the rotation axis 31 is also performed.

[0028] In the robot 10 described above, reducers are attached to the θ-axis motor and the motors of each horizontal articulated mechanism, and these axes are driven by the motors via the reducers. A traveling drive unit is connected to the traveling motor, and by driving the traveling motor, the robot 10 moves along the rails 21. The reducer and traveling drive unit are transmission devices connected to the motor. A mechanism for detecting the occurrence of an abnormality in the transmission device 10 provided in the robot 10 will be described below. Figure 3 is a block diagram showing the configuration of a control device 40 connected to the robot 10.

[0029] The control device 40 drives and controls the motor 36 of each axis of the robot 10 based on input command values, such as position command values. The motor 36 is equipped with an encoder 37 that detects the rotational position of the motor 36. A reducer 38 is attached to the rotating shaft of the motor 36, and a load 39 is mechanically connected to the motor 36 via the reducer 38. A traveling drive unit may be used instead of the reducer 38. The control device 40 is equipped with a control unit 41 that servo-controls the motor 36 by receiving a signal indicating the position of the motor 36 as feedback from the encoder 37, and an abnormality detection device 42 according to an embodiment of the present invention that detects abnormalities in the transmission device. While the figure shows one motor 36, one control unit 41, and one abnormality detection device 42, the robot 10 is equipped with multiple motors 36, and therefore the control device 40 is equipped with a control unit 41 and an abnormality detection device 42 for each motor 36 it controls. Therefore, the control device 40 drives and controls the motor 36 and detects abnormalities in the reducer 38 for each axis. The control unit 41 uses a standard servo control mechanism for controlling the motors of each axis of an industrial robot.

[0030] The abnormality detection device 42 includes a speed calculation unit 43 that receives a signal indicating the position of the motor 36 from the encoder 37 and performs a differential operation on the signal to continuously calculate the speed of the motor 36, a band-pass filter 44 that performs band-pass filtering on the signal indicating the speed value calculated by the speed calculation unit 43, and an RMS calculation unit 45 that calculates a root mean square (RMS) value from the speed value after band-pass filtering. The abnormality detection device 42 also includes a comparator 46 that compares the RMS value calculated by the RMS calculation unit 45 with a threshold value, and an alarm processing unit 47 that determines that an abnormality has occurred in the reducer 38 and outputs an alarm when the RMS value exceeds the threshold value. Control devices used to control robots are generally implemented by microprocessors or the like that operate using software, so the abnormality detection device 42 in this embodiment is also implemented by digital signal processing technology that uses, for example, a microprocessor or the like.

[0031] The bandpass filter 44 is configured, for example, by cascading a lowpass filter 44L and a highpass filter 44H. The optimum passband of the bandpass filter 44 varies depending on the motor 36, the reducer 38, and the size and configuration of the robot 10 itself, but is, for example, 3 to 10 Hz. In this case, the cutoff frequency of the lowpass filter 44L is set to 10 Hz, and the cutoff frequency of the highpass filter 44H is set to 3 Hz. It is preferable that both the lowpass filter 44L and the highpass filter 44H are fifth-order Butterworth filters, in which case the bandpass filter 44 is also configured as a fifth-order Butterworth filter. If the transmission device is a rank-and-pinion type traveling drive unit, a bandpass filter 44 with a passband of, for example, 3 to 20 Hz is used, and in this case, it is also preferable that the bandpass filter 44 is a fifth-order Butterworth filter.

[0032] The RMS calculation unit 45 includes a square calculation unit (Sq) 45A that squares the velocity value processed by the band-pass filter 44, an integrating unit (Sum) 45A that integrates the value calculated by the square calculation unit 45 over a predetermined period of time, and a square root calculation unit (Sqrt) 45C that finds the square root of the value integrated by the integrating unit 45. Since digital signal processing is assumed here, the integrating unit 45 integrates, for each sampling period of the signal, the values ​​calculated by the square calculation unit 45 over a predetermined number of sampling periods immediately preceding it, and outputs the value as the RMS value.

[0033] The detection of an abnormality in the reducer 38 by the anomaly detection device 42 of this embodiment will be described below based on an example using FIG. 4. Using the robot 10 described with reference to FIGS. 1 and 2, the motor 36 was driven to rotate the turntable 23 around the rotation axis 31 in both cases where the reducer 38 connected to the θ-axis motor 36 was normal (normal product) and where it was abnormal (abnormal product). In FIG. 4, (a-1) shows the speed of the motor 36 detected for the normal product, and (b-1) shows the speed of the motor 36 detected for the abnormal product. The speed command for the motor 36 was to drive the motor 36 at the maximum speed specified for the θ-axis in the robot 10. In FIG. 4 and subsequent figures, the speed is shown in speed units used internally by the control device 40. The speed value data obtained in this manner was subjected to bandpass filtering using a band filter 44 with a passband of 3 to 10 Hz. The results for the normal product are shown in (a-2) and (a-3). (a-2) shows the results of processing with a third-order Butterworth filter, and (a-3) shows the results of processing with a fifth-order Butterworth filter. Similarly, for defective products, (b-2) shows the results of processing with a third-order Butterworth filter, and (b-3) shows the results of processing with a fifth-order Butterworth filter. In these figures, the gray line labeled "FILTERED" shows the values ​​obtained by processing with the band-pass filter 44, and the black line labeled "RMS" shows the results obtained by further processing with the RMS processing unit 45, i.e., the RMS value.

[0034] The results shown in Figure 4 show that while the speed data of the motor 36 itself cannot effectively distinguish between normal and abnormal conditions of the reducer, bandpass filtering clearly reveals the difference between the two, with the abnormal product exhibiting larger amplitude, especially the maximum amplitude. Comparing the RMS values, the peak value was larger for the abnormal product. Setting a threshold value that takes into account the normal RMS value makes it possible to detect abnormalities in the reducer 38. For normal products, both the amplitude of the values ​​after bandpass filtering and the peak value of the RMS value were smaller when a fifth-order bandpass filter was used. This suggests that a fifth-order Butterworth filter is preferable to a third-order filter for detecting abnormalities in the reducer 38. While a seventh-order or higher-order Butterworth filter could be used, it would be impractical due to increased overshoot, increased implementation costs, and reduced accuracy due to calculation errors.

[0035] The bandpass filter 44 will now be further examined. To investigate the fluctuations in the speed of the motor 36, an FFT (fast Fourier transform) can be performed on the signal indicating the speed. However, using the signal directly would result in a waveform that is very difficult to see. Therefore, the speed signal was processed using a bandpass filter with a passband frequency range of 2 to 20 Hz, a fifth-order Butterworth filter, and an FFT was performed on the filtered speed signal to obtain a spectrum. In Figure 5, (a-1) and (a-2) respectively show the filtered speed value and the spectrum obtained by FFT for a normal product. Similarly, (b-1) and (b-2) in Figure 5 respectively show the filtered speed value and spectrum for an abnormal product. In the normal product, the intensity of frequency components above 2 Hz is low. In contrast, in the abnormal product, peaks were observed in the frequency range from slightly below 4 Hz to slightly above 5 Hz, as well as in the 8 to 10 kHz range corresponding to the second harmonic of that frequency range. These peaks are thought to represent frequency components specific to abnormal products, and it can be seen that in order to detect whether or not an abnormality has occurred in the reducer 38 of this robot 10, it is sufficient to use a bandpass filter 44 with a passband of 3 to 10 Hz.

[0036] Furthermore, for the defective products, the spectra shown in Fig. 5(c-1) and (c-2) were obtained by applying FFT to the velocity values ​​after band-pass filtering shown in Fig. 4(b-2) and (b-3). Fig. 5(c-1) shows the spectrum obtained by processing with a third-order Butterworth filter. In this spectrum, in addition to peaks specific to defective products, peaks due to unnecessary frequency components exist in the range below 2 Hz. In contrast, the spectrum obtained by processing with a fifth-order Butterworth filter shown in Fig. 5(c-2) did not show peaks of unnecessary frequency components located in the range below 2 Hz. This shows that it is preferable to use a fifth-order Butterworth filter rather than a third-order filter as the band-pass filter 44 in order to prevent the influence of peaks due to unnecessary frequency components. In addition, in the velocity and RMS value results after filtering using a third-order Butterworth filter for an abnormal product, as shown in Figure 4 (b-2), unnatural fluctuations are seen between approximately 1 second and approximately 3 seconds. This is also thought to be the influence of the unnecessary frequency components mentioned above that could not be completely removed by band-pass filter processing. In recent robots, processing such as limiting jerk (also known as jerk or jerk) is sometimes performed to suppress manipulator vibration, and in such cases, it is thought that a low-order band-pass filter cannot remove the DC component of the jerk. It is thought that this DC component that could not be removed becomes an unnecessary frequency component in the spectrum obtained from the third-order Butterworth filter.

[0037] The robot 10 shown in FIGS. 1 and 2 moves horizontally along a rail 21 using a traveling drive unit including a pinion mounted on a base 22. Here, we describe an example in which the robot 10 shown in FIGS. 1 and 2 moves back and forth along the rail 21, and an abnormality in the traveling drive unit of the robot 10 is detected using the abnormality detection device 42 of this embodiment. In this case, the traveling drive unit is provided instead of the reducer 38 in the block diagram shown in FIG. 4. The abnormal product was a traveling drive unit with loose bolts securing the gears. A fifth-order Butterworth filter with a passband of 3 to 20 Hz was used as the bandpass filter 44. The results are shown in FIG. 6. Similar to (a-3) or (b-3) in FIG. 4, FIG. 6 shows the velocity values ​​after bandpass filtering and the RMS values ​​obtained by processing the velocity values ​​after bandpass filtering with the RMS processing unit 45. In FIG. 6, (a) shows the results for a normal product, and (b) shows the results for an abnormal product.

[0038] When the traveling drive unit causes the robot 10 to move back and forth along the rail 21, the robot 10 accelerates in one direction to a predetermined speed, then moves at a constant speed, then decelerates and stops, then accelerates in the opposite direction to the predetermined speed, moves at a slower speed, and then decelerates again and stops. The SP value represents how close the predetermined speed is to the maximum speed specified for the horizontal movement of the robot 10. If SP=100%, the predetermined speed during the reciprocating movement is the same as the specified maximum speed, and if SP=50%, the predetermined speed during the reciprocating movement is 50% of the specified maximum speed. Figure 6 shows the results for horizontal movement of the robot 10 using different SP values.

[0039] As shown in Figure 6, for a normal product, the peak value of the velocity value after band-pass filtering is small, and the RMS value only fluctuates slightly around 0 regardless of the SP value. In contrast, for an abnormal product, the velocity value after band-pass filtering fluctuates greatly, the RMS value pulsates, and the peak of the RMS value increases as the SP value increases. This demonstrates that the anomaly detection device 42 of this embodiment can also detect anomalies in the traveling drive unit. To detect anomalies, it is preferable to drive the traveling drive unit, which is driven by the traveling motor, as fast as possible within the maximum speed range specified for that traveling drive unit. Even when detecting anomalies in the reducers of each axis of a robot rather than the traveling drive unit, it is considered preferable to drive that axis as fast as possible within the maximum speed range specified for that axis.

[0040] FIG. 6 , as explained above, shows an example of detecting an abnormality in a traveling drive unit based on the speed of the traveling motor. For comparison, a case where an abnormality in a power transmission is detected using a quantity other than motor speed will be described. FIG. 7 shows the results of detecting an abnormality based on the torque command value for the traveling motor in the traveling drive unit of the robot 10 shown in FIGS. 1 and 2 . The torque command value is obtained from the control unit 41 of the control device 40. The results shown in FIG. 7 were obtained in the same manner as in FIG. 6 , except that the torque command value, instead of the speed, is input to the band-pass filter 44 of the abnormality detection device 42. The torque command value is shown in units used internally by the control device 40. The torque command value for the traveling motor is proportional to the current supplied to the traveling motor. The figure shows XY coordinate values ​​(X is time, Y is the torque command value) at a representative point for the torque command value after band-pass filtering. In the results shown in FIG. 7 , the amplitude of the torque command value after band-pass filtering was larger in the abnormal product, and pulsation was observed in the RMS value in the abnormal product. However, even in normal products, when the SP value is large (for example, SP = 100%), pulsation is observed in the RMS value. For this reason, when attempting to determine whether or not there is an abnormality based on the torque command value, it is thought that setting the threshold value for the determination becomes more difficult than when determining whether or not there is an abnormality based on the speed, and the risk of erroneous determination increases.

[0041] FIG. 8 shows the results of detecting an abnormality based on the position deviation of the travel motor in the travel drive unit of the robot 10 shown in FIGS. 1 and 2. The position deviation is the deviation between the position command value given to the control device 40 and the actual position of the motor 36 obtained by the encoder 38, and is obtained from the control unit 41. The results shown in FIG. 8 were obtained in the same manner as in FIG. 6, except that the position deviation, instead of the velocity, is input to the band-pass filter 44 of the abnormality detection device 42. The position deviation value is shown in units used internally by the control device 40. In the figure, the XY coordinate values ​​(X is time, Y is position deviation) of representative points for the position deviation value after band-pass filtering are shown. The results shown in FIG. 8 also showed pulsation in the RMS value for an abnormal product. However, even in normal products, pulsation in the RMS value was observed when the SP value was large (e.g., SP = 100%). Therefore, when attempting to determine the presence or absence of an abnormality based on the position deviation, it is considered that setting the threshold for determination is more difficult than when determining the presence or absence of an abnormality based on the velocity, and the risk of erroneous determination increases. The above results relate to the traveling drive unit, but it is believed that for reducers as well, by performing bandpass filter processing on the motor speed value, it will be possible to detect abnormalities more accurately than when performing bandpass filter processing on the torque command value for the motor or the position deviation of the motor.

[0042] In the above-described embodiment, the speed of the motor driving the transmission is calculated and band-pass filtering is performed on the motor speed value, thereby enabling reliable detection of abnormalities, such as failures, in the transmission device. Furthermore, by performing RMS processing or the like, abnormalities can be easily detected simply by setting a threshold value. While it is possible to detect abnormalities in the transmission device using an acceleration sensor or the like, in this embodiment, abnormalities are detected based on values ​​obtained during servo control of the motor without using an acceleration sensor or the like, thereby reducing the cost of installing an acceleration sensor or the like. Because the abnormality detection device 42 of this embodiment can be implemented by software executed by a microprocessor, it is highly compatible with general robot controllers (i.e., control devices 40), which are also implemented by running software on a microprocessor. Therefore, it can be easily incorporated into existing control devices 40 by simply modifying the software. Furthermore, such robot controllers (control devices 40) often already have a mechanism or algorithm for calculating the speed of the motor 36 from the output signal of the encoder 38. In such cases, there is no need to separately provide a speed calculation unit 43 in the abnormality detection device 42; the motor speed can be obtained from the existing mechanism or algorithm.

[0043] In the above explanation, the abnormality detection device 42 performs band-pass filtering on the speed values, then calculates the root mean square (RMS) value of the speed values ​​and compares it with a threshold value, and determines whether an abnormality exists based on the magnitude relationship between the RMS value and the threshold value. If the only determination of whether an abnormality exists is based on a comparison with a threshold value, the square root calculation unit 45C of the RMS calculation unit 45 is not necessarily required, and a configuration in which the output of the integrating unit 45B is compared with the threshold value can also be used. In this case, the root mean square of the speed values ​​is compared with the threshold value, and the threshold value used here is one for the root mean square value rather than the RMS value. In other words, the abnormality detection device 42 determines whether an abnormality has occurred in the transmission by calculating the root mean square or root mean square from the signal obtained by band-pass filtering and comparing it with a threshold value. [Explanation of symbols]

[0044] 10...Robot; 11A, 11B...First arm; 12A, 12B...Second arm; 13A, 13B...Hand; 21...Rail; 22...Base; 23...Rotating table; 24...Lifting mechanism; 24A...Fixed part; 24B...Moving part; 25...Cover; 26A, 26B...Arm support part; 31...Rotating axis; 32...Common axis; 33A, 33B...Axis; 34A, 34B...Wrist axis; 36...Motor; 3 7 Encoder; 38... Reducer; 39... Load; 40... Control device; 41... Control unit; 42... Abnormality detection device; 43... Speed ​​calculation unit; 44... Band pass filter; 44L... Low pass filter; 44H... High pass filter; 45... RMS calculation unit; 45A... Square calculation unit; 45B... Integration unit; 45C... Square root calculation unit; 46... Comparator; 47... Alarm processing unit; 50... Work.

Claims

1. 1. A method for detecting an abnormality in a transmission device driven by a motor, comprising: Detecting the speed of the motor; The detected velocity value is subjected to band-pass filtering using a fifth-order Butterworth filter; An abnormality detection method for determining whether or not an abnormality has occurred in the transmission device based on the signal obtained by the bandpass filter processing.

2. 2. The anomaly detection method according to claim 1, wherein the presence or absence of the anomaly is determined by calculating the root mean square or root mean square of the signal obtained by the band-pass filtering process and comparing it with a threshold value.

3. 3. The abnormality detection method according to claim 1, wherein the speed of the motor is detected from a signal obtained from an encoder connected to the motor.

4. An abnormality detection device for detecting an abnormality in a transmission device driven by a motor, a band-pass filter that performs band-pass filtering on the detected value of the motor speed; a determination means for determining whether or not an abnormality has occurred in the transmission device based on the signal output from the bandpass filter; and The anomaly detection device, wherein the band-pass filter is a fifth-order Butterworth filter.

5. further comprising a calculation means for calculating the root mean square or root mean square of the signal output from the band pass filter, 5. The abnormality detection device according to claim 4, wherein the determining means determines whether or not the abnormality has occurred by comparing the value obtained by the calculating means with a threshold value.

6. 6. The abnormality detection device according to claim 4, further comprising speed calculation means for detecting the speed of said motor from a signal obtained from an encoder connected to said motor.

7. 7. The abnormality detection device according to claim 4, wherein the abnormality detection device is provided inside a control device that executes servo control of the motor based on a command value given from an external source.

Images