JP2024008098A - Data processing method and data processing device - Google Patents

Abstract

An object of the present invention is to analyze waveform data and obtain appropriate analysis results.
[Solution] A data processing method executed by a data processing device uses a sensor provided in a machine equipped with an actuator that operates according to a control mode, which detects a physical quantity that changes according to the operation of the actuator. An acquisition process for acquiring output waveform data, and an extraction process for extracting feature quantities from a predetermined extraction target section of the waveform data, and the control mode of the actuator when the physical quantity is detected by the sensor. and an extraction step of varying the length of the extraction target section accordingly.
[Selection diagram] Figure 2

Description

The present disclosure relates to a data processing method and a data processing device.

Patent Document 1 describes a technology that uses a sensor to detect vibrations, noise, deflection of component parts, etc. that occur during the operation of a machine, and analyzes the waveform of the output signal of the sensor.

International Publication No. 2018/110337

Even if waveform data is analyzed in the same way when the control mode of the machine is position control mode and speed control mode, appropriate analysis results may not be obtained. Therefore, there is a need for a technique for analyzing waveform data and obtaining appropriate analysis results.

The present disclosure can be realized as the following forms.

(1) According to a first aspect of the present disclosure, a data processing method executed by a data processing device is provided. This data processing method acquires waveform data output from a sensor installed in a machine equipped with an actuator that operates according to a control mode, and that detects a physical quantity that changes according to the operation of the actuator. and an extraction step of extracting a feature amount from a predetermined extraction target section of the waveform data, the step of extracting the feature amount from a predetermined extraction target section of the waveform data, the step of extracting the feature amount according to the control mode of the actuator when the physical amount is detected by the sensor. and an extraction step of varying the length of the extraction target section.
According to this form of data processing method, it is possible to suppress the extraction of feature amounts from inappropriate extraction target sections, and therefore it is possible to suppress the inability to obtain appropriate analysis results of waveform data.
(2) In the data processing method of the above aspect, in the extraction step, the length of the extraction target section when the control mode is the position control mode is the length of the extraction target section when the control mode is not the position control mode. It may be shorter than the length of the section.
According to the data processing method of this form, when the control mode of the actuator is the position control mode, it is possible to suppress extraction of a feature amount from an inappropriate extraction target section.
(3) In the data processing method of the above aspect, in the extraction step, when the control mode is the position control mode, the operation of the actuator is started next from the timing at which the operation of the actuator is started. The section up to the timing may be set as the extraction target section.
According to the data processing method of this form, when the control mode of the actuator is the position control mode, it is possible to suppress extraction of a feature amount from an inappropriate extraction target section.
(4) In the data processing method of the above embodiment, in the extraction step, when the control mode is not the position control mode, the entire waveform data may be the extraction target section.
According to the data processing method of this form, when the control mode of the actuator is not the position control mode, it is possible to suppress extraction of a feature amount from an inappropriate extraction target section.
(5) The data processing method of the above embodiment may include a determination step of determining whether or not there is an abnormality in the machine using the feature amount.
According to this type of data processing method, it is possible to accurately determine whether there is an abnormality in the machine.
(6) According to the second aspect of the present disclosure, a data processing device is provided. This data processing device is a sensor installed in a machine equipped with an actuator that operates according to a control mode, and acquires waveform data output from the sensor that detects a physical quantity that changes according to the operation of the actuator. and an extraction unit that extracts a feature quantity from a predetermined extraction target section of the waveform data, the extraction unit extracting the feature quantity from a predetermined extraction target section of the waveform data, the extraction unit extracting the feature amount from a predetermined extraction target section of the waveform data, the extraction unit extracting the feature quantity according to the control mode of the actuator when the physical quantity is detected by the sensor. An extraction unit that varies the length of the extraction target section.
According to the data processing device of this form, it is possible to suppress extraction of a feature amount from an inappropriate extraction target section, and therefore it is possible to suppress the inability to obtain an appropriate analysis result of waveform data.
The present disclosure can also be implemented in various forms other than data processing methods and data processing devices. For example, it can be realized in the form of a feature extraction method, a feature extraction device, an anomaly detection method, an anomaly detection device, and the like.

FIG. 2 is an explanatory diagram schematically showing the configuration of a data processing device. A flowchart showing the details of a data processing method. An explanatory diagram showing an example of waveform data. FIG. 4 is an explanatory diagram showing an example of a determination result by a determination unit.

A. First embodiment:
FIG. 1 is an explanatory diagram schematically showing the configuration of a data processing apparatus 100 in the first embodiment. The data processing device 100 includes a CPU 101, a memory 102, an input/output interface 103, and an internal bus 104. The CPU 101, memory 102, and input/output interface 103 are connected via an internal bus 104 so that they can communicate in both directions. A sensor 250 provided in the machine 200 is connected to the input/output interface 103. In this embodiment, in addition to the sensor 250, an input device 105 and a display device 106 are connected to the input/output interface 103. The input device 105 is, for example, a keyboard or a mouse, and the display device 106 is, for example, a liquid crystal display. Note that the input device 105 and the display device 106 may be integrated as a touch panel.

The CPU 101 functions as an acquisition unit 110, an extraction unit 120, and a determination unit 130 by executing a computer program stored in advance in the memory 102. The acquisition unit 110 acquires waveform data output from the sensor 250. The extraction unit 120 extracts feature amounts from the extraction target section of the waveform data. The determination unit 130 determines whether or not there is an abnormality in the machine 200 using the feature amount, and causes the display device 106 to display the determination result.

Machine 200 includes an actuator 210, a control device 220 that controls actuator 210, and a sensor 250. In this embodiment, the machine 200 is a processing machine such as a machining center that processes a workpiece, and the actuator 210 is a servo motor that moves a table to which the workpiece is fixed. The actuator 210 and the control device 220 are configured to be operable in a position control mode in which the position of the table is set as a target value. Note that the machine 200 is not limited to a processing machine, and may be, for example, a transport device that transports a workpiece. The actuator 210 is not limited to a servo motor, and may be, for example, a stepping motor or a power cylinder such as an electric cylinder.

The sensor 250 detects a physical quantity that changes depending on the operation of the actuator 210. In this embodiment, the sensor 250 is a torque sensor that detects the torque generated by the servo motor that is the actuator 210. The sensor 250 is not limited to a torque sensor, and may be, for example, a current sensor that detects the current supplied to the actuator 210. The number of sensors 250 may be one or more.

FIG. 2 is a flowchart showing the details of the data processing method executed by the data processing device 100. FIG. 3 is an explanatory diagram showing an example of waveform data Wt. FIG. 4 is an explanatory diagram showing an example of the determination result by the determination unit 130. In this embodiment, the data processing method shown in FIG. 2 is repeatedly executed by the data processing device 100 at predetermined time intervals while the machine 200 is powered on.

First, in step S110, the acquisition unit 110 acquires waveform data Wt output from the sensor 250 that detects the torque of the servo motor, which is the actuator 210. In this embodiment, the machine 200 is provided with a rotation speed sensor that detects the rotation speed of the servo motor in addition to a sensor 250 that detects the torque of the servo motor, and the acquisition unit 110 acquires torque waveform data Wt. In addition, the waveform data Ws output from the rotational speed sensor is acquired.

In step S120, the extraction unit 120 determines whether the control mode of the actuator 210 is the position control mode. In this embodiment, the control mode of the actuator 210 is input by the user via the input device 105. Note that the process in step S120 may be executed prior to the process in step S110.

If it is determined in step S120 that the control mode of the actuator 210 is the position control mode, the extraction unit 120 determines in step S130 whether the operation of the actuator 210 is stopped or not. The stop flag data representing the waveform data Wt is acquired from the control device 220, and in step S135, the stop flag data is used to determine an extraction target section from which the feature quantity is to be extracted from the waveform data Wt. Specifically, the extraction unit 120 divides the waveform data Wt into a plurality of sections using the timing at which the operation of the actuator 210 is started as a boundary, and calculates the next operation of the actuator 210 from the timing at which the operation of the actuator 210 starts. Each section up to the timing when the process starts is determined as an extraction target section. As shown in FIG. 3, the first operation of the actuator 210 is started at time t1 and then stopped, the second operation is started and stopped at time t2, and the third operation is started at time t3. If the operation is stopped after the operation is performed and the operation is not started thereafter, the extraction unit 120 determines the interval S1 from time t1 to time t2 as the first extraction target interval, and determines the interval S2 from time t2 to time t3 as the first extraction target interval. This is determined as the second extraction target section. In this embodiment, the extraction unit 120 also determines the section S3 from time t3 to the end of the waveform data Wt as an extraction target section.

If it is determined in step S120 that the mode is not the position control mode, the extraction unit 120 determines the entire waveform data Wt S0 as the extraction target section in step S140. That is, the extraction unit 120 varies the length of the extraction target section along the time axis depending on the control mode of the actuator 210. Specifically, the extraction unit 120 makes the length of the extraction target section when the control mode of the actuator 210 is the position control mode shorter than the length of the extraction target section when the control mode is not the position control mode. There is. Note that control modes other than the position control mode include, for example, a speed control mode in which the actuator 210 is controlled using the speed of the portion moved by the actuator 210 as a target value, and a speed control mode in which the actuator 210 is controlled using the acceleration of the portion moved by the actuator 210 as a target value. There is an acceleration control mode to control.

In step S150, the extraction unit 120 extracts a plurality of feature amounts from the extraction target section determined in step S135 or step S140. FIG. 4 shows an example of feature amounts extracted by the extraction unit 120 in the position control mode. The torque amplitude during stop represents the maximum value of the torque amplitude while the rotation of the actuator 210 is stopped. The positive acceleration torque represents the maximum absolute value of positive torque during rotational acceleration, and the negative acceleration torque represents the maximum absolute value of negative torque during rotational acceleration. The deceleration torque represents the maximum absolute value of positive torque during rotational deceleration, and the negative side deceleration torque represents the maximum absolute value of negative torque during rotational deceleration. The operating torque amplitude represents the maximum value of the torque amplitude during rotational acceleration and rotational deceleration. The damped vibration represents the maximum value of the amplitude of the damped vibration during constant speed rotation and while the rotation is stopped, and the continuous vibration represents the maximum value of the torque amplitude during constant speed rotation. The power running energy is calculated by dividing the sum of the torques in the part where the direction of rotation and the torque direction match, excluding the part where the rotation is stopped, by the number of samples. It is calculated by dividing the sum of the torques in the portion where the direction of rotation and the direction of torque do not match, except for , by the number of samples. The maximum torque represents the maximum value of torque, and the minimum torque represents the minimum value of torque. The average loss is the value obtained by subtracting the sum of the torques in the parts where the direction of rotation and the torque do not match from the sum of the torques in the parts where the direction of rotation and the direction of torque match, excluding the part where rotation is stopped. is calculated by dividing by the number of samples. Note that the number of feature amounts extracted by the extraction unit 120 may be one instead of a plurality. When the mode is not the position control mode, the extraction unit 120 may extract the load factor in one cycle and the load factor in a short time as the feature quantities. The load factor in one cycle represents the effective torque in one operation cycle of the actuator 210, and the short-time load factor is the effective torque in the time window when sliding the time window of a predetermined width in the extraction target section. represents the maximum value of

In step S160, the determination unit 130 determines whether or not there is an abnormality in the machine 200 using the feature amount. In this embodiment, the memory 102 stores a threshold value for each feature amount in advance, and the determination unit 130 determines whether the machine 200 is abnormal by comparing the threshold value stored in the memory 102 with the feature amount. Determine the presence or absence. The determining unit 130 determines that there is an abnormality when one or more feature amounts are equal to or greater than the threshold value, and determines that there is no abnormality when all the feature amounts are less than the threshold value. Note that the same threshold value or different threshold values may be used when it is determined in step S120 that the mode is the position control mode and when it is determined that the mode is not the position control mode.

In step S170, the determination unit 130 causes the display device 106 to display the determination result. FIG. 4 shows an example of the determination results. In the example of FIG. 4, since the torque amplitude during stop is equal to or greater than the threshold value, it is determined that there is an abnormality in the machine 200. For example, if the actuator 210 is operating in the position control mode and the table moved by the actuator 210 is stopped before reaching the target position due to a foreign object being caught, the actuator 210 is operated in the position control mode to move the table to the target position. However, since the rotation of the actuator 210 is inhibited by the foreign matter being bitten, the torque amplitude increases during the stop. Therefore, if the torque amplitude during the stop is equal to or greater than the threshold value, there is a possibility that a foreign object has been caught.

Thereafter, the data processing device 100 ends this process. Note that step S110 of the data processing method is referred to as an acquisition step, steps S120 to S150 are referred to as an extraction step, step S160 is referred to as a determination step, and step S170 is referred to as a determination result output step. I may call you.

According to the data processing device 100 in this embodiment described above, the extraction unit 120 varies the length of the feature extraction target section according to the control mode of the actuator 210. Specifically, the extraction unit 120 sets the length of the extraction target section when it is determined that the control mode of the actuator 210 is the position control mode to be longer than the length of the extraction target section when it is determined that the control mode of the actuator 210 is not the position control mode. shorten. Therefore, according to the control mode of the actuator 210, the feature amount can be extracted from the extraction target section suitable for analyzing the waveform data Wt. Therefore, it is possible to prevent the waveform data Wt from being properly analyzed due to the feature amount being extracted from an inappropriate extraction target section. In order to appropriately analyze the waveform data Wt, in the position control mode, each section of the waveform data Wt corresponding to a period in which each positioning operation by the machine 200 is executed is set as an extraction target section and feature quantities are extracted. In modes other than the position control mode, it is preferable to extract the feature amount using the entire waveform data Wt as the extraction target section. In the present embodiment, when the extraction unit 120 determines that the control mode of the actuator 210 is the position control mode, the extraction unit 120 selects the next operation of the actuator 210 from the timing when the operation of the actuator 210 is started in the waveform data Wt. Each section up to the timing at which is started is determined to be an extraction target section, and if it is determined that the position control mode is not, the entire waveform data Wt is determined to be an extraction target section. Therefore, the waveform data Wt can be appropriately analyzed using the feature amount extracted from the extraction target section. Furthermore, in this embodiment, the extraction section 120 automatically determines the extraction target section instead of the user determining the extraction target section, so there is no possibility that the feature amount will be extracted from an inappropriate extraction target section due to human error. can be prevented.

Furthermore, in the present embodiment, the determination unit 130 determines whether or not there is an abnormality in the machine 200 using the feature amount extracted from the extraction target section. Therefore, it is possible to accurately determine whether there is an abnormality in the machine 200 using appropriately extracted feature amounts. Further, since the determination unit 130 displays the determination result on the display device 106, the user can inspect the machine 200 and replace parts after understanding the presence or absence of an abnormality to prevent breakdowns of the machine 200. be able to.

B. Other embodiments:
(B1) In the data processing device 100 of the first embodiment described above, the extraction unit 120 extracts the length of the extraction target section when the control mode of the actuator 210 is the position control mode, and when the control mode is not the position control mode. is shorter than the length of the extraction target section. In contrast, the extraction unit 120 makes the length of the extraction target section when the control mode of the actuator 210 is the position control mode longer than the length of the extraction target section when the control mode is not the position control mode. Good too.

(B2) In the data processing device 100 of the first embodiment described above, when the control mode of the actuator 210 is the position control mode, the extracting unit 120 extracts from the waveform data Wt when the operation of the actuator 210 is started. The section from the timing to the next timing when the operation of the actuator 210 is started is set as the extraction target section. On the other hand, when the control mode of the actuator 210 is the position control mode, the extracting unit 120 extracts from the waveform data Wt the timing at which the operation of the actuator 210 is stopped next. The section up to the timing may be set as the extraction target section.

(B3) In the data processing device 100 of the first embodiment described above, the extraction unit 120 sets the entire waveform data Wt as an extraction target section when the control mode of the actuator 210 is not the position control mode. On the other hand, when the control mode of the actuator 210 is not the position control mode, the extraction unit 120 may set some sections of the waveform data Wt as extraction target sections.

(B4) The data processing device 100 of the first embodiment described above includes a determination unit 130 that determines whether or not there is an abnormality in the machine 200 using the feature amount extracted by the extraction unit 120. On the other hand, the determination unit 130 may not be provided. In this case, the user may determine whether there is an abnormality in the machine 200 using the feature amount extracted by the extraction unit 120.

(B5) The display device 106 is connected to the data processing device 100 of the first embodiment described above, and the determination unit 130 causes the display device 106 to display the determination result as to whether or not there is an abnormality in the machine 200. On the other hand, a warning lamp or a warning buzzer may be connected to the data processing device 100. In this case, the determination unit 130 may activate a warning lamp or a warning buzzer when determining that there is an abnormality in the machine 200.

(B6) In the first embodiment described above, the data processing method is executed to detect an abnormality in the machine 200 while the machine 200 is powered on. In contrast, the data processing method may be performed, for example, after the machine 200 has been powered off, in order to investigate the cause of the anomaly. In this case, for example, while the power of the machine 200 is turned on, the waveform data Wt output from the sensor 250 is stored in the memory 102 of the data processing device 100, and the acquisition unit 110 performs the steps of the data acquisition method. In S110, the waveform data Wt stored in the memory 102 may be acquired.

The present disclosure is not limited to the embodiments described above, and can be implemented in various configurations without departing from the spirit thereof. For example, the technical features in the embodiments corresponding to the technical features in each form described in the summary column of the invention may be used to solve some or all of the above-mentioned problems, or to achieve one of the above-mentioned effects. In order to achieve some or all of the above, it is possible to replace or combine them as appropriate. Further, unless the technical feature is described as essential in this specification, it can be deleted as appropriate.

100... data processing device, 101... CPU, 102... memory, 103... input/output interface, 104... internal bus, 105... input device, 106... display device, 110... acquisition section, 120... extraction section, 130... determination section, 200... Machine, 210... Actuator, 220... Control device, 250... Sensor

Claims (6)

A data processing method performed by a data processing device, the method comprising:
an acquisition step of acquiring waveform data output from a sensor installed in a machine equipped with an actuator that operates according to a control mode and that detects a physical quantity that changes according to the operation of the actuator;
an extraction step of extracting a feature amount from a predetermined extraction target section of the waveform data, the feature amount being extracted from a predetermined extraction target section according to the control mode of the actuator when the physical quantity is detected by the sensor; an extraction process that varies the length of the
A data processing method comprising: The data processing method according to claim 1,
In the extraction step, the length of the extraction target section when the control mode is a position control mode is shorter than the length of the extraction target section when the control mode is not the position control mode. . The data processing method according to claim 2,
In the extraction step, when the control mode is the position control mode, an interval from a timing when the operation of the actuator starts to a timing when the operation of the actuator next starts is set as the extraction target interval. , data processing methods. 4. The data processing method according to claim 3,
In the extraction step, when the control mode is not the position control mode, the entire waveform data is set as the extraction target section. The data processing method according to claim 1,
A data processing method, comprising a determination step of determining whether or not there is an abnormality in the machine using the feature amount. A data processing device,
an acquisition unit that acquires waveform data output from a sensor that is installed in a machine equipped with an actuator that operates according to a control mode and that detects a physical quantity that changes according to the operation of the actuator;
an extraction unit that extracts a feature amount from a predetermined extraction target section of the waveform data, the feature amount being extracted from a predetermined extraction target section in accordance with the control mode of the actuator when the physical quantity is detected by the sensor; an extraction part with different lengths;
A data processing device comprising:

Images