SYNCING DEVICE AND STORAGE MEDIUM

Abstract

Synchronization device (1), comprising: a machine data acquisition unit (11) for acquiring time series information about the drive of a machine (15) as machine data, which is information that is acquired at a time specified by initial time information; a measurement data acquisition unit (12) for acquiring time series information about a state of the machine (15) as measurement data, which is information that includes the intensity of a noise or vibration and is acquired at a time specified by second time information that differs from the first time information; a correlation calculation unit (13) for calculating a correlation time difference, which is a time difference when the strength of the correlation between an absolute value of the machine data and a frequency component of the measurement data is at its maximum, based on the machine data, the measurement data and a time difference when one of the machine data and the measurement data is offset in a positive or negative direction of a time axis; and an output unit (14) for synchronized data for outputting the machine data synchronized with the measurement data based on the correlation time difference as synchronized machine data and for outputting the measurement data synchronized with the machine data based on the correlation time difference as synchronized measurement data, wherein the correlation calculation unit (33) includes a peak extraction unit (337) to obtain from the measurement data a frequency at which a frequency spectrum of the measurement data exhibits peak values, and The strength of the correlation is calculated based on a temporal change in a frequency component where the frequency spectrum of the measurement data shows peak values, and the machine data.

Description

field of technology

[0001] The present disclosure relates to a synchronization device that synchronously outputs a plurality of time-series data pieces and a storage medium. State of the art

[0002] It is known that, generally, in a device containing a power source such as a motor, the drive noise of the motor or machine, as well as the object being driven by the motor, contains a great deal of information regarding the states of the power source and the driven object. Accordingly, there is a need for a technique in which, to determine the state of a device, measurement data—that is, data about the drive noise or vibration of a motor or machine measured by a sensor or similar device—is acquired while machine data, such as the position, speed, or torque of the motor or machine, is acquired, and these data points are analyzed synchronously. However, since a system for acquiring machine data and a system for acquiring measurement data are fundamentally different systems, a certain degree of inventiveness is required.Against this technical background, the following patent literature 1 discloses a technique in which a time at which a given feature occurs is extracted from the respective machine data and measurement data measured by different systems, thereby synchronizing these two pieces of data with each other.

[0003] Patent literature 2 relates to a device comprising a vibration exciter that sets a test piece into vibration, a control device for the vibration exciter, a load measuring device for measuring the reaction force exerted by the vibration exciter, and a vibration response calculation device that performs vibration response calculations based on the output signals of the load measuring device and / or internally stored input signals, as well as a vibration signal generation device that generates a vibration signal based on the calculation results, and which is also provided with a displacement measuring device for measuring the displacement exerted by the vibration exciter.

[0004] Patent literature 3 relates to a process analysis system in which the receiving device receives manually entered data. The symbolisation device classifies and symbolises the data entered via the receiving device. The shift-time search device analyzes correlations while shifting the time-series data classified and symbolised by the symbolisation device and the time-series data stored in the storage device, searching for the shift time that yields the strongest correlation. The response model generation device describes the relationship between the time-series data using a response model that utilizes the shift time searched for by the shift-time search device.

[0005] Patent literature 4 relates to a machine system comprising a sensor unit that periodically detects an acceleration in the region of the tip of a movable component, a data acquisition unit that obtains first time series data of the acceleration in the region of the tip of the movable component corresponding to a sensor signal received via a wireless signal path, a data calculation unit that calculates second time series data corresponding to the first time series data based on a drive command from the motor, a delay time calculation unit that calculates a delay time of the first time series data relative to the second time series data according to a degree of correlation between the first time series data and the second time series data, and a correction unit that corrects the first time series data based on the delay time. List of quotations Patent literature Patent literature 1: JP 2019 - 219 725 A Patent literature 2: JP H10 - 281 925 A Patent literature 3: JP 2007 - 323 504 A Patent literature 4: DE 10 2015 002 192 A1 Summary of the invention; Technical task

[0006] In the method described in patent literature 1, a point in time at which the volume of a machining noise exceeds a threshold, or a point in time at which the acceleration of a tool's vibration exceeds a threshold, is cited as an example of the point in time at which a predetermined feature occurs. However, the points in time at which these features occur vary depending on the driving conditions, such as machine configuration, the type of tool used for machining, the material of the workpiece, and the machining pattern. Therefore, the problem with the method described in patent literature 1 is that it is necessary to redefine the point in time at which a feature occurs depending on these driving conditions.This means that the method described in patent literature 1 has the problem that adjustment work is required to prepare a measurement before the measurement, and therefore a synchronization process cannot be carried out efficiently.

[0007] The present disclosure was made in light of the above statements, and one of its aims is to obtain a synchronization device capable of efficiently carrying out a synchronization process. Solution to the task

[0008] To solve the problem described above and achieve the objective, a synchronization device according to claim 1 and a computer-readable storage medium according to claim 6 are provided. The advantageous embodiments are defined in dependent claims. Advantageous effects of the invention

[0009] The synchronization device according to the present disclosure achieves the effect that it is possible to carry out a synchronization process efficiently. Brief description of the drawings Fig. Figure 1 is a block diagram showing an example of a functional configuration of a synchronization device according to a first embodiment. Fig. Figure 2 is a view that represents an example configuration of a drive system comprising a drive device with a synchronization device function according to the first embodiment. Fig. Figure 3 is a block diagram that shows an example of a functional configuration of the drive device according to the one described in Figure 3. Fig. 2 shows the first embodiment. Fig. Figure 4 is a flowchart that illustrates an example of a processing flow of a synchronization process in the first embodiment. Fig. 5 is a waveform diagram showing an example of machine data used to describe Fig. 4 can be used. Fig. Figure 6 is a waveform diagram showing an example of measurement data used to describe Fig. 4 can be used. Fig. Figure 7 is a block diagram showing an example of a functional configuration of a correlation calculation unit in the first embodiment. Fig. 8 is a flowchart that shows an example of a processing flow of a [product / service] in [a specific context]. Fig. The correlation calculation shown in Figure 4 is demonstrated. Fig. Figure 9 is a waveform diagram used to describe a process of outputting synchronized data in the first embodiment. Fig. Figure 10 is a block diagram that shows an example of a functional configuration of a synchronization device according to a second embodiment. Fig. Figure 11 is a block diagram that shows an example of a functional configuration of a correlation calculation unit in the second embodiment. Fig. Figure 12 is a view showing an example configuration of a data analysis system that includes a data analysis device with a synchronization device function according to the second embodiment. Fig. Figure 13 is a block diagram that shows an example of a functional configuration of the in Fig. The data analysis device shown in Figure 12 is shown. Fig. Figure 14 is a block diagram showing an example of a functional configuration of a visualization terminal according to the second embodiment. Fig. Figure 15 is a diagram that represents a display example of the visualization terminal based on an analysis result of the data analysis device in the second embodiment. Fig. Figure 16 is a block diagram showing an example of a functional configuration of a synchronization device according to a third embodiment. Fig. Figure 17 is a block diagram showing an example of a functional configuration of a correlation calculation unit in the third embodiment. Fig. Figure 18 is a view that represents an example configuration of a simulation system that includes a simulation terminal with a synchronization device function according to the third embodiment. Fig. Figure 19 is a block diagram showing an example of a functional configuration of the simulation terminal according to the third embodiment. Description of embodiments

[0010] A synchronization device and a storage medium according to each embodiment of the present disclosure will below be described in detail with reference to the accompanying drawings. First embodiment.

[0011] Fig. Figure 1 is a block diagram showing an example of a functional configuration of a synchronization device 1 according to a first embodiment. The synchronization device 1 according to the first embodiment comprises a machine data acquisition unit 11, a measurement data acquisition unit 12, a correlation calculation unit 13, and an output unit 14 for synchronized data.

[0012] The machine data acquisition unit 11 acquires time-series information about the operation of a machine connected to the synchronization device 1 (not shown). The machine data is managed based on initial time information. This initial time information specifies the acquisition time of the machine data. The machine data is information about the operation of a machine driven by a power source such as a motor. Examples of machine data include a command value relating to the angle of rotation, position, speed, current, and thrust of a motor connected to the machine, or measured values ​​thereof. The machine data acquisition unit 11 acquires at least one type of machine data.For example, machine data concerning the angle of rotation and machine data concerning the position are different types of machine data, and pieces of machine data concerning the angle of rotation taken at different times are the same type of machine data. When acquiring a multitude of machine data types, each of these types is linked to every other type by the initial time information, and each type of machine data is acquired with a predefined sampling period.

[0013] The measurement data acquisition unit 12 acquires time-series information about the state of the machine connected to the synchronization device 1. The measurement data is managed based on secondary time information. This secondary time information specifies the acquisition time of the measurement data and is acquired along with the measurement data itself. The measurement data is information obtained by measuring the state of a machine driven by a power source such as a motor, either directly from the machine or from a sensor attached to the machine. Examples of measurement data include the machine's drive noise, the position of a moving part of the machine, vibration acceleration in the moving part of the machine, force and pressure absorbed by the moving part of the machine, and a captured moving image of a drive state of the machine.The measurement data can be a calculated value, which is determined using one or more measured values, instead of directly using a measurement taken by the sensor. The measurement data acquisition unit 12 acquires at least one type of measurement data. For example, the measurement data about the machine's drive noise and the measurement data about the position of the machine's moving part are different types of measurement data, and segments of the measurement data about the machine's drive noise at different times are the same type of measurement data. The measurement data is acquired with a predefined sampling period. The sampling period of the measurement data can be a period that differs from the sampling period of the machine data.

[0014] The correlation calculation unit 13 calculates a correlation score based on the machine data acquired by the machine data acquisition unit 11 and the measurement data acquired by the measurement data acquisition unit 12. The correlation score indicates the degree of correlation between the machine data and the measurement data. The correlation score is calculated based on one type of time series data from the machine data and one type of time series data from the measurement data. Any combination of the two data sets used to calculate the correlation score can be used, as long as a correlation exists between them. Each of the machine data and measurement data examples described above can be combined in any way.

[0015] Furthermore, the correlation calculation unit 13 adjusts the time difference if either the machine data or the measurement data is shifted in a positive or negative direction along a time axis, and calculates the correlation score based on this adjusted time difference. This means that by changing the time difference, a variety of correlation scores can be obtained. Additionally, the correlation calculation unit 13 calculates a correlation score representing the maximum correlation strength between the machine data and the measurement data, and calculates a correlation time difference, which is the time difference at which the correlation score is obtained.

[0016] The output unit 14 for synchronized data synchronizes the machine data with the measurement data based on the correlation time difference calculated by the correlation calculation unit 13. The output unit 14 for synchronized data determines, as synchronized machine data, the machine data synchronized with the measurement data based on the correlation time difference and outputs the synchronized machine data to the outside of the synchronization device 1. Furthermore, the output unit 14 for synchronized data determines, as synchronized measurement data, the measurement data synchronized with the machine data based on the correlation time difference and outputs the synchronized measurement data to the outside of the synchronization device 1.Segments of machine data and segments of measurement data can be correlated using the initial and subsequent time information, as well as information about the correlation time difference. Therefore, a synchronization process between different data types can also be performed based on the correlation time difference information calculated by the correlation calculation unit 13.

[0017] Fig. Figure 2 is a view that represents an example configuration of a drive system 500 comprising a drive device 510 with a function of the synchronization device 1 according to the first embodiment. Fig. The drive system 500 comprises a blower 15 and the drive device 510, which drives the blower 15. The drive device 510 comprises a control device 16, the measurement data acquisition unit 12, and an inspection terminal 18. The blower 15 comprises an impeller 15a and a motor 15b connected to the impeller 15a. The blower 15 and the inspection terminal 18 are electrically connected to the control device 16.

[0018] The control device 16 contains an inverter (not shown). The control device 16 drives the motor 15b by supplying an electrical signal to the motor 15b. The impeller 15a is set in motion by the motor 15b and rotates at a predetermined speed. The rotational speed of the impeller 15a changes depending on the electrical signal output by the control device 16, and simultaneously, a noise caused by blowing air also changes.

[0019] The data acquisition unit 12 measures the sound pressure of a noise emitted by the blower 15 and / or the control device 16. A microphone is one example of a data acquisition unit 12. The data acquisition unit 12 can be installed at a predetermined fixed position in the configuration of the drive system 500 or at any position according to the configuration of the drive system 500. The data acquisition unit 12 is preferably installed near the blower 15 or the control device 16, each of which represents a source of the sound to be measured. When using a microphone as the data acquisition unit 12, it is desirable for the microphone to be directional. In such a case, it is possible to reduce the mixing of ambient noise with the measurement data to be acquired.

[0020] The inspection terminal 18 is an end device that provides an inspector of the blower 15 with the information necessary for the inspection. The inspection terminal 18 includes a cable connection 21 and is connected to the control device 16 via a detachable cable 19a. The inspection terminal 18 receives control data from the control device 16 via cable 19a. Examples of devices suitable for the inspection terminal 18 are a smartphone, a tablet, and a laptop computer.

[0021] The inspection terminal 18 is connected to the measurement data acquisition unit 12 via a cable 19b. The inspection terminal 18 acquires the sound pressure data measured by the measurement data acquisition unit 12.

[0022] Cables 19a and 19b can be used for data exchange between information terminals, such as a USB (Universal Serial Bus) cable, a LAN (Local Area Network) cable, and an SPI (Serial Peripheral Interface) communication cable. The inspection terminal 18 includes a connector of the required standard, which depends on the cable used for the connection. The connection between the inspection terminal 18 and the measurement data acquisition unit 12, and the connection between the inspection terminal 18 and the control device 16, can be established wirelessly without the use of a cable.

[0023] Furthermore, a microphone built into the inspection terminal 18 can be used as a measurement data acquisition unit 12. In this case, only the inspection terminal 18 and the cable 19a connecting the control device 16 and the inspection terminal 18 are required at the time of inspection, thus reducing the number of accessories. Consequently, the inspection can be carried out more easily.

[0024] Fig. Figure 3 is a block diagram showing an example of a functional configuration of the drive device 510 according to the one described in Fig. Figure 2 shows the first embodiment. As in Fig. As shown in Figure 3, the drive device 510 comprises the measurement data acquisition unit 12, the control device 16 and the inspection terminal 18. The control device 16 comprises a control unit 161, and the inspection terminal 18 comprises a control unit 181, a storage unit 182 and a monitor 183.

[0025] The control unit 161 is a component that integrally controls the entire control device 16. The control unit 161 comprises a motor drive unit 162 and the machine data acquisition unit 11. The motor drive unit 162 is a component that generates an electrical signal to control the drive state of a motor connected to the control device 16 based on a command from a user of the drive device 510 or a higher-level controller.

[0026] The machine data acquisition unit 11 is connected to the motor drive unit 162 and acquires data about the drive via the control device 16 in time series. The machine data acquisition unit 11 stores machine data collected in time series together with the initial time information managed by the control device 16. The measurement data acquisition unit 12 stores measurement data collected in time series together with the secondary time information.

[0027] In the inspection terminal 18, the control unit 181 is a component that integrally controls the entire inspection terminal 18. The control unit 181 comprises a synchronization unit 10, an anomaly diagnostic unit 184, a display unit 185, a machine data communication unit 186, and a measurement data communication unit 187. The synchronization unit 10 comprises the correlation calculation unit 13 and the output unit 14 for synchronized data. During the configuration of Fig. 3. The functions of the machine data acquisition unit 11 and the measurement data acquisition unit 12 are located outside the inspection terminal 18. Consequently, the machine data acquisition unit 11 and the measurement data acquisition unit 12 are separate from the synchronization unit 10. On the other hand, the synchronization unit 10 contains the correlation calculation unit 13 and the output unit 14 for synchronized data and therefore has the function of outputting synchronized machine data and synchronized measurement data. Therefore, the synchronization unit 10 can be considered a component corresponding to the synchronization device 1.

[0028] Synchronization unit 10 is connected to machine data acquisition unit 11 via machine data communication unit 186. Machine data communication unit 186 acquires machine data from machine data acquisition unit 11 through communication. Synchronization unit 10 acquires the machine data from machine data communication unit 186 at a requested interval and stores the machine data in storage unit 182. Synchronization unit 10 is also connected to measurement data acquisition unit 12 via measurement data communication unit 187. Measurement data communication unit 187 acquires measurement data from measurement data acquisition unit 12 through communication. Synchronization unit 10 acquires the measurement data from measurement data communication unit 187 at a requested interval and stores the measurement data in storage unit 182.

[0029] The first set of time information, managed by the control device 16, and the second set of time information, managed by the measurement data acquisition unit 12, are managed by different systems and are therefore distinct pieces of time information. With regard to the machine data and the measurement data, which are based on these different pieces of time information, the synchronization unit 10 performs a synchronization process, as described later, based on the information about the time difference calculated by the correlation calculation unit 13. The machine data and measurement data that have undergone the synchronization process are output by the output unit 14 as synchronized machine data and synchronized measurement data, respectively.

[0030] The anomaly diagnostic unit 184 diagnoses a condition of the blower 15 or the control device 16 using synchronized machine data and synchronized measurement data. For example, if the sound pressure level of the blower 15's drive noise (i.e., the measurement data) does not change, even though the speed of the motor 15b (i.e., the machine data) has changed from standstill to a user-commanded speed, an anomaly in the blower 15 can be diagnosed. Although the blower 15 has been given a command to start blowing, no blowing noise is heard and no blowing occurs, thus enabling such a diagnosis.The diagnosis by the anomaly diagnostic unit 184 only needs to be an anomaly diagnosis using the synchronized machine data and the synchronized measurement data, and it is sufficient to perform one necessary diagnosis depending on a device configuration or drive pattern.

[0031] The display unit 185 visualizes and displays a diagnostic result from the anomaly diagnosis performed by the anomaly diagnosis unit 184. The synchronized machine data and the synchronized measurement data, synchronized by the synchronization unit 10, can be displayed by the display unit 185 as the basis for the diagnostic result, using a means such as a graph. Displaying the synchronized machine data and the synchronized measurement data allows the inspector to further understand the cause of the diagnostic result. Ideally, the synchronized machine data and the synchronized measurement data should be arranged simultaneously in a vertical or horizontal direction and displayed as parallel or superimposed graphs. If such a display is possible, the inspector can verify the machine's condition in a more easily understandable way.

[0032] The storage unit 182 is connected to the control unit 181, stores machine data or measurement data in response to a request from the control unit 181, and, if necessary, stores an anomaly diagnostic result or the like. The storage unit 182 comprises read-only memory (ROM) or working memory (RAM) and can be further expanded by the use of flash memory, a hard disk drive (HDD), or a solid-state drive (SSD).

[0033] Monitor 183 is controlled by display unit 185. Monitor 183 displays the anomaly diagnosis result received from anomaly diagnosis unit 184 to the inspector. An example of Monitor 183 is a liquid crystal display device. Monitor 183 only needs to be able to communicate the diagnosis result to the inspector and can be implemented using a light-emitting diode (LED), an audio output device, or the like.

[0034] The respective functions of control units 161 and 181 can be implemented by software. In such cases, a program constituting the software is installed on a computer that executes the functions of control units 161 and 181. There is no restriction on software implementation of the respective functions of control units 161 and 181. The respective functions of control units 161 and 181 can also be implemented by an electronic circuit such as an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a complex programmable logic device (CPLD).

[0035] Next, the synchronization process in the first embodiment will be described with further reference to Fig. 4 to 9 in addition to Fig. 2 and Fig. 3 described. Fig. Figure 4 is a flowchart that illustrates an example of a processing flow of the synchronization process in the first embodiment. Fig. 5 is a waveform diagram showing an example of machine data used to describe Fig. 4 can be used. Fig. Figure 6 is a waveform diagram showing an example of measurement data used to describe Fig. 4 can be used. Fig. Figure 7 is a block diagram showing an example of a functional configuration of the correlation calculation unit 13 in the first embodiment. Fig. 8 is a flowchart that shows an example of a processing flow of a [product / service] in [a specific context]. Fig. The correlation calculation shown in Figure 4 is demonstrated. Fig. Figure 9 is a waveform diagram used to describe a process of outputting synchronized data in the first embodiment.

[0036] First, the machine data acquisition unit 11 records machine data in step S11. Fig. Figure 5 shows, as an example of machine data, current data indicating the strength of a current supplied by the motor drive unit 162 to the motor 15b. The horizontal axis in Fig. 5 represents time. Fig. Figure 5 shows a waveform in a case where the blower 15 is switched on at time t0, operates with a constant current for a period of time, and is then switched off at time t1. At time t1 and thereafter, a waveform is shown in a case where the blower 15 operates with a constant current, with the polarity of the current alternating. The times t0 and t1 are the initial time information described above.

[0037] The power data is sampled by the machine data acquisition unit 11. In the first embodiment, a sampling frequency of 100 Hz is assumed for the power data. The sampling frequency does not necessarily have to be 100 Hz, and any desired sampling frequency can be set depending on data accuracy, hardware performance, and the like. The machine data acquired in step S11 is entered into the synchronization unit 10 via the machine data communication unit 186.

[0038] In step S12, the measurement data acquisition unit records 12 measurement data points. Fig. Figure 6 shows, as an example of the measurement data measured by the measurement data acquisition unit 12, sound pressure data representing a drive noise of the blower 15 when the blower 15 is driven by the current waveform in Fig. 5 is driven. The horizontal axis in Fig. 6 represents time. The sound pressure data in Fig. 6 will be completed in the same time period as in Fig. 5 is recorded, but the recording times T0 and T1 are based on the second time information managed by the measurement data acquisition unit 12, and therefore times other than those in Fig. 5 is indicated.

[0039] In the first embodiment, a sampling frequency of 48 kHz is assumed for the measurement data. The sampling frequency does not necessarily have to be 48 kHz, and any desired sampling frequency can be set depending on data accuracy, hardware performance, and the like. The measurement data acquired in step S12 is entered into the synchronization unit 10 via the measurement data communication unit 187.

[0040] The processes of step S11 and step S12 can be performed in reverse order or simultaneously. Furthermore, in the processes of steps S11 and S12, the data to be collected does not necessarily include all data pieces to be synchronized; only some data can be collected. The data to be collected at this point need only consist of data from a section containing data collected at the same time. By limiting the data to be collected to a few data pieces, it is possible to reduce the computational effort of a process performed by the correlation calculation unit 13, which will be described later.

[0041] In step S13, the correlation calculation unit 13 performs a correlation calculation using the machine data acquired in the process of step S11 and the measurement data acquired in the process of step S12. The correlation calculation unit 13 calculates a correlation score, which indicates the degree of correlation between the machine data and the measurement data, and calculates a correlation value at which the degree of correlation is maximal, as well as a correlation time difference based on a number of calculated correlation scores. Further details of the correlation calculation are described later.

[0042] In step S14, output unit 14 for synchronized data outputs synchronized data based on a result of the correlation calculation performed in step S13. The synchronized data described here are the synchronized machine data and synchronized measurement data described above.

[0043] Next, the process is described in step S13 in the flowchart of Fig. 4 described. The process of step S13 is carried out by the correlation calculation unit 13 according to the flowchart of Fig. 8 carried out. As in Fig. As shown in Figure 7, the correlation calculation unit 13 comprises a sampling adjustment unit 131, a normalization unit 132, a correlation assessment value calculation unit 133, a correlation value calculation unit 134 and a correlation time difference determination unit 135.

[0044] In step S131 of Fig. 8. The correlation calculation unit 13 performs a process in which the sampling frequency of the machine data and the sampling frequency of the measurement data are made equal. This process is controlled by the sampling adjustment unit 131. Fig. 7 carried out.

[0045] As described above, in this example the sampling frequency of the current data (machine data) is 100 Hz, and the sampling frequency of the sound pressure data (measurement data) is 48 kHz. Therefore, a period of 480 samples is defined as the decimation period for the sound pressure data, and a data segment is extracted from the sound pressure data for each decimation period. This process reduces the measurement data to 100 Hz.

[0046] In the process described above, a decimation frequency, which is the reciprocal of the decimation period, is set to 100 Hz. This 100 Hz is the greatest common divisor (GCD) of the sampling frequency of the machine data and the sampling frequency of the measurement data, but the decimation frequency can be a value smaller than the GCD. By setting the decimation frequency to a value smaller than the GCD, the amount of computation performed by the correlation calculation unit 13 after a decimation process can be reduced.

[0047] As described above, the scanning unit 131 performs the decimation process depending on the sampling frequency of the machine data and the sampling frequency of the measurement data, generating machine data and measurement data that are aligned to have the same sampling period. The sampling period described here is a decimation period. The scanning unit 131 can perform an interpolation process instead of, or in conjunction with, the decimation process to generate machine data and measurement data that are aligned to have the same sampling frequency.

[0048] Furthermore, the sampling unit 131 can perform a filtering process using a digital filter on the machine data and / or the measurement data prior to the decimation process. Examples of digital filters include a low-pass filter, a high-pass filter, and a band-pass filter. Performing the filtering process reduces the influence of aliasing noise that arises during the decimation or interpolation process. Additionally, the filtering process reduces the influence of noise generated when the machine data acquisition unit 11 and the measurement data acquisition unit 12 acquire their respective data segments.

[0049] In step S132 of Fig. 8. Normalization Unit 132 normalizes the machine data and the measurement data, which are adjusted to have the same sampling period. Normalization Unit 132 subtracts an offset component from each of the machine data and measurement data to average them to zero. Furthermore, Normalization Unit 132 performs scaling between the machine data and the measurement data by multiplying the machine data and / or the measurement data by a specific numerical value.

[0050] The normalization performed by normalization unit 132 reduces the impact of errors arising from the fact that machine data and measurement data are different data types when calculating a correlation score. Specifically, the machine data and measurement data may have different dimensions. The processes of step S131 and step S132 can be performed in reverse order. That is, the process with normalization unit 132 can precede the process with the scanning unit 131. However, if the scanning unit 131 process is performed first, the computational workload can be reduced. Therefore, in cases where reducing the computational load is desirable, the scanning unit 131 process should preferably be performed first.

[0051] Next, in step S133, the following is calculated: Fig. 8. The correlation rating calculation unit 133 calculates a correlation rating that indicates the degree of correlation between the machine data and the measurement data. In the next step, S134, the correlation rating calculation unit 134 calculates a multitude of correlation ratings between the machine data and the measurement data and calculates a correlation value based on this multitude of correlation ratings. That is, the correlation rating calculation unit 134 calculates a correlation rating with the maximum correlation strength among the multitude of correlation ratings. Furthermore, in step S135, the correlation time difference determination unit 135 determines a correlation time difference based on the correlation value. The correlation time difference is information about the time difference when a correlation value with the maximum correlation strength is obtained.

[0052] Next, a procedure for determining the correlation time difference is described in more detail. First, one of the machine data and one of the measurement data are selected as reference data. It is desirable to use data stored for a shorter period as the reference data. For simplicity, the reference data will be referred to as "first data" below, and data that is not the reference data will be referred to as "second data."

[0053] Next, a correlation score is calculated between the data obtained by shifting the second set of data by a sampling period Δt and the first set of data. This calculation yields a correlation score sequence with a multitude of correlation scores. If the correlation score is denoted by V, the correlation score sequence can be expressed using the integers M and N and the sampling period Δt as follows. {V(Δt×(−M)),…,V(Δt×(−1)),V(Δt×0),V(Δt×1),…,V(Δt×N)}

[0054] In the correlation score sequence above, the integers M and N are coefficients that determine the maximum time difference between a time point shifted in the negative direction of the time axis and a time point shifted in the positive direction of the time axis. The maximum time difference can be determined from the sampling times of the respective data segments. In a case where the maximum time difference is predetermined by the device configuration, its value can be used.

[0055] On a lower page of Fig. 9 are the ones in Fig. The sound pressure data shown in section 6 are presented as the first data. Furthermore, the following data is shown on the top side of... Fig. 9 the in Fig. The current data shown in Figure 5 is represented as a second set of data, and it is a waveform obtained by shifting the entire waveform of the current data in the negative direction of the time axis by the time Δt×n. That is to say, in the waveform of Fig. 9 the start of the current data, based on the sound pressure data as reference data, precedes the time Δt×n.

[0056] Regarding the correlation score, which is calculated for the first data points as reference data and the second, time-shifted data points, a scoring index is used to determine the similarity between the two time-series data points. Specific examples of the scoring index are a correlation coefficient, covariance, absolute error, squared error, and Mahalanobis distance. For a section that, as a result of the time shift Δt×n, contains only one data point from the two time-series data points, e.g., for section A in Fig. 9, a calculated value set to zero.

[0057] Furthermore, if the data points used to calculate the correlation score have negative values, the calculation can be performed by determining the absolute values ​​of the respective data points. Similarly, if, when comparing correlation scores, the degree of influence of an absolute value is high and the degree of influence of a phase is low, a suitable correlation assessment can be performed by determining and comparing the absolute values ​​of the respective data points. In particular, if a frequency component of the sound pressure data is used as measurement data, the frequency component of the sound pressure data does not contain a phase, making it desirable to use an absolute value of the machine data in the correlation calculation.

[0058] Next, a correlation score with the maximum correlation strength in the calculated correlation score sequence is set as the correlation value, and a time difference when the correlation score is obtained is set as the correlation time difference. The correlation score with the maximum correlation strength is a value at which the correlation score is either maximum or minimum. Whether the correlation score is maximum or minimum depends on what is used as the correlation score.

[0059] The correlation time difference is determined by the procedure described above. In the first embodiment, the correlation time difference is obtained by calculating the time difference at which the correlation score is at its maximum or minimum. However, the correlation calculation unit 13 only needs to determine the correlation time difference that is the time difference at which the correlation is strongest, and could calculate the correlation time difference using a procedure that does not determine the correlation score. For example, the correlation time difference can be obtained by calculating a phase difference using a Fourier transform or by using a convolutional neural network. The output unit 14 for synchronized data can output the synchronized machine data and the synchronized measurement data based on the correlation time difference.Synchronized machine data is machine data that is synchronized with the measurement data. Synchronized measurement data is measurement data that is synchronized with the machine data. Since the sound pressure data is used as reference data in the example above, the sound pressure data and the current data can be synchronized by offsetting the current data relative to the reference data by the correlation time difference.

[0060] The synchronization device according to the first embodiment can synchronize machine data and measurement data based on different time information pieces acquired by different systems, based on the calculation of the correlation between the two data pieces. Therefore, the synchronization device can output these two data pieces synchronously, even if the two data pieces to be subjected to the synchronization process are based on different time information pieces acquired by different systems.

[0061] Furthermore, according to the synchronization device of the first embodiment, synchronization is performed by calculating a correlation, so that it is not necessary to define a feature that serves as a reference for synchronization in the synchronization device. In particular, the synchronization process can be carried out from a first operating time, since it is not necessary to store machine data at normal time or measurement data at normal time for synchronization.

[0062] The correlation between the machine data and the measurement data used in the synchronization device according to the first embodiment is generic. Therefore, a machine incorporating the synchronization device according to the first embodiment can synchronously output a variety of time-series data segments. For example, the relationship between a motor's current value and the intensity of drive noise generated by the motor is such that when current flows through the motor, the motor is driven and thus the drive noise increases, and when no current flows through the motor, the motor stops and thus the drive noise decreases. This relationship does not depend on the motor's drive pattern.Therefore, for many machines, including motors, even if the configuration or drive pattern of the respective machines is changed, a large number of time series data pieces can be output synchronously.

[0063] The synchronization device according to the first embodiment performs a synchronization process by calculating a correlation between machine data and measurement data that differ from each other and were acquired by different systems. Therefore, the inspection terminal with the integrated synchronization device according to the first embodiment can diagnose a blower anomaly based on the synchronized machine data and the synchronized measurement data output by the synchronization device. Consequently, the inspection terminal according to the first embodiment can diagnose an anomaly using machine data and measurement data acquired by different devices without requiring a high-speed synchronization network.The inspection terminal according to the first embodiment can diagnose an anomaly without performing a reset for synchronization, even if a blower drive pattern or a blower mechanical configuration is changed.

[0064] Furthermore, according to the first embodiment, the inspection terminal allows a microphone to be positioned near a location that could be a contributing factor to an anomalous noise from the device. This enables the microphone to record the noise and display sound pressure data of the noise in synchronization with machine data at the time the noise occurred. Thus, if an anomalous noise occurs in the blower, the user can visualize both the machine data and the noise. The inspector can then assess the cause of the noise by closely observing the relationship between the machine data and the noise. This is made possible, in particular, by the simultaneous and parallel display of the machine data and the measurement data, so that they appear to be aligned in the same direction.If the images are aligned vertically or horizontally, it is possible to represent the state of the machine in a more easily understandable way.

[0065] Furthermore, according to the first embodiment, the inspection terminal can perform a synchronization process between input machine data and measurement data. Consequently, even if the inspection terminal is subsequently connected to an existing device, it is possible to achieve a similar effect to one achieved when the inspection terminal is installed at the time of delivery. In particular, integrating the measurement data acquisition unit into the inspection terminal facilitates the visualization of the mechanical condition of the existing device.

[0066] Furthermore, according to the first embodiment, the inspection terminal can perform the inspection of a large number of devices with a single terminal by connecting the inspection terminal to the devices sequentially via cables. This reduces the number of inspection terminals required.

[0067] The synchronization device according to the first embodiment can include the scanning adjustment unit. The scanning adjustment performed by the scanning adjustment unit makes it possible to calculate a correlation evaluation value in order to carry out a synchronization process even in a case where the machine data and the measurement data have different scanning periods.

[0068] Furthermore, the synchronization device according to the first embodiment can include the normalization unit. The normalization of the data, which is performed by the normalization unit before the calculation of the correlation score, makes it possible to reduce the influence of an error caused in a case where the machine data and the measurement data are in different units of measurement. Second embodiment.

[0069] Fig. Figure 10 is a block diagram showing an example of a functional configuration of a synchronization device 2 according to a second embodiment. Compared to the configuration of the synchronization device 1 according to the first embodiment, which is described in Figure 10, the configuration of the synchronization device 2 is shown in Figure 10. Fig. As shown in Figure 1, in the synchronization device 2 according to the second embodiment, the correlation calculation unit 13 is replaced by a correlation calculation unit 23. In the second embodiment, a temporal change in a frequency spectrum is used as the data to be synchronized. Therefore, the correlation calculation unit 23 calculates a correlation after a frequency transformation has been performed on the data. Other components are the same as or correspond to those shown in Figure 1. Fig. 1 shown. Identical or corresponding components are shown with the same reference symbols as in . Fig. 1 is designated, and redundant descriptions of the same are omitted.

[0070] Fig. Figure 11 is a block diagram showing an example of a functional configuration of the correlation calculation unit 23 in the second embodiment. Compared to the configuration of the correlation calculation unit 13 in the first embodiment, which is shown in Fig. As shown in Figure 7, the correlation calculation unit 23 in the second embodiment includes a frequency transformation unit 234, a frequency selection unit 235, and an envelope processing unit 236 between the sampling adjustment unit 131 and the normalization unit 132. Other components are the same as or correspond to those shown in Figure 7. Fig. 7 shown. Identical or corresponding components are indicated with the same reference symbols as in Fig. 7 is designated, and redundant descriptions of the same are omitted.

[0071] In Fig. 11. The sampling unit 131 performs the decimation process depending on the sampling frequency of the machine data and the sampling frequency of the measurement data, generating machine data and measurement data that are aligned to have the same sampling period. Of the machine data and measurement data aligned to have the same sampling period, the data that are not to be subjected to a frequency transformation process are fed into the normalization unit 132. Conversely, the data that are to be subjected to the frequency transformation process are fed into the frequency transformation unit 234.

[0072] The frequency transformation unit 234 performs a process of transforming at least a portion of the time series data from the machine data or the measurement data into a time series frequency spectrum. A fast short-time Fourier transform (STFFT), a wavelet transform, a discrete cosine transform, a cepstrum, or the like can be used as a method for performing the transformation into such a time series frequency spectrum. The time series frequency spectrum transformed by the frequency transformation unit 234 can be used for display on a display unit 287 described later.

[0073] The frequency selection unit 235 selects a frequency spectrum within a predefined frequency range from the frequency spectrum of time series calculated by the frequency transformation unit 234 and uses this frequency spectrum as machine data or measurement data for a correlation calculation. The frequency range for selection ideally includes a frequency related to the operation of a mechanical device. Examples include a resonant frequency of the mechanical device, a rotational frequency of the device, and a frequency that is an integer multiple of these frequencies. The frequency range for selection can be predefined or determined from the obtained machine data or measurement data.

[0074] In a case where the wavelet transform or the discrete cosine transform is used in the frequency transform unit 234, the process with the frequency transform unit 234 and the process with the frequency selection unit 235 can be performed in reverse order. In a case where the process with the frequency selection unit 235 is performed first, the calculated frequency spectrum of time series cannot be used for display, but the calculation of the frequency transform can be restricted to the frequency range for selection, thus reducing the computational scope.

[0075] The envelope processing unit 236 detects an envelope for time-series data of a frequency spectrum in the frequency domain for selection. Examples of envelope methods include a low-pass filter method and a Hilbert transform. The envelope processing unit 236 can be omitted.

[0076] Fig. Figure 12 is a view showing an example configuration of a data analysis system 520, which includes a data analysis device 281 with a function of the synchronization device 2 according to the second embodiment. Fig. The data analysis system 520 comprises a four-axis robot 25, a controller 261, servo drives 262 to 265, an accelerometer 271, a camera 272, a logger 291, a database server 293, the data analysis device 281, and a visualization terminal 282. The data analysis device 281 and the visualization terminal 282 are connected to the logger 291 via the database server 293. The database server 293 can be a cloud server located in a network 292.

[0077] The four-axis robot 25 is a device that performs tasks based on a command from the controller 261. The four-axis robot 25 comprises four motors (not shown) that drive the respective axes of the four axes, as well as mechanical elements that transmit force to the respective axes. Although an actual four-axis robot comprises many parts, in Fig. 12 For the sake of simplicity, only some components are shown.

[0078] The controller 261 is connected to the four-axis robot 25 via the servo drives 262 to 265 and integrally controls the drive of the four-axis robot 25. The controller 261 manages the initial timing information described above. Furthermore, the controller 261 is responsible for issuing a command to an axis as required, based on this initial timing information, in order to drive the four-axis robot 25 in a work sequence intended by a commander.

[0079] The servo drives 262 to 265 are connected to the respective motors of the four-axis robot 25 and, in accordance with a command from the controller 261, generate electrical signals which serve as commands to drive the four-axis robot 25 and apply the electrical signals to the motors of the respective axes.

[0080] The accelerometer 271 is attached to a moving part of an arm of the four-axis robot 25 and measures vibrations in time series that are generated in the four-axis robot 25 by the drive of the motors of the respective axes. A three-axis accelerometer can be used as the accelerometer 271.

[0081] Camera 272 is an industrial camera that captures the workflow of the four-axis robot 25 by recording the robot intermittently or continuously. Camera 272 manages the second timing information described above. Based on this second timing information, camera 272 captures the workflow of the four-axis robot 25 along with its operating noise and stores the captured image and sound data as video data. This video data is then transmitted to logger 291.

[0082] The logger 291 is connected to the accelerometer 271 via a cable 274 and to the camera 272 via a cable 275. The logger 291 manages third-party time information. The logger 291 accesses the respective connected devices sequentially or simultaneously, records rotational speed data of the respective axes in the four-axis robot 25 as machine data together with time information, and transmits the rotational speed data and the time information to the database server 293. In addition, the logger 291 records a measurement value from the accelerometer 271 and the moving image data from the camera 272 as measurement data together with the time information of the data and transmits the data and the time information to the database server 293.

[0083] In the configuration of Fig. Since accelerometer 271 does not have time information, logger 291 appends the third time information to the measurement data of accelerometer 271 and transmits the measurement data with the appended third time information to database server 293.

[0084] Database server 293 stores the machine data acquired by logger 291, along with the measurement data and the associated time information, in the database. Database server 293 can be implemented using a relational database management system (RDBMS) or a non-structured query language (not only SQL). When storing the data, database server 293 stores the data along with the associated time information, which is managed by database server 293.

[0085] The data analysis device 281 is a device that analyzes a state of the four-axis robot 25 using data collected in the database about the four-axis robot 25 in order to perform a learning process. Fig. Figure 13 is a block diagram that shows an example of a functional configuration of the in Fig. Figure 12 shows the data analysis device 281. The data analysis device 281 comprises an input unit 283, a database communication unit 284, the synchronization device 2, a data analysis unit 285 and an output unit 286.

[0086] The input unit 283 captures machine data and measurement data intended for data analysis. The user defines the machine data and measurement data to be analyzed via the input unit 283. In addition to specifying the type of data to be analyzed, a range for the data acquisition time can also be defined. By defining the data acquisition time, the amount of data to be analyzed can be reduced, thus shortening the processing time.

[0087] The database communication unit 284 submits a query to the database server 293 using a tool such as SQL, specifying the machine and measurement data to be analyzed, and captures the required data fragments. If the data acquisition time range is defined, data fragments for analysis are captured based on the fourth set of time information managed by the database server 293. The database server 293 can select data fragments accordingly by appending time information common to all data fragments. If the acquisition time range is defined, data can be captured by specifying a range larger than the designated time.Consequently, even in a case where some data is lost due to the correlation time difference determined at the time of the synchronization process, data encompassing the designated acquisition time range can be provided to the user without missing parts if the acquisition time range is set in anticipation of such a case.

[0088] The synchronization device 2 can perform a synchronization process by combining the rotational speed data based on the first time information, the motion image data based on the second time information, and the acceleration data based on the third time information. In a case where the synchronization is performed using the rotational speed data based on the first time information as machine data, with regard to the machine data at time t0, a maximum of i=1,2,3,4 |ωi (t0)| with the maximum absolute value of the rotational speed among the rotational speeds ω1(t0), ω2(t0), ω3(t0), and ω4(t0) of the four axes at time t0 can be used as machine data at time t0. Consequently, for example, when performing a correlation calculation with audio data in motion data as measurement data, it can be prevented that an unsuitable correlation weighting value is calculated due to the influence of drive noise generated by the operation of another axis. In general, it is assumed that the drive noise is generated when at least one motor is driven. By selecting the maximum value among the values ​​of the four axes, it is therefore possible to calculate the correlation weighting value taking into account the influence of the drive noise generated by the operation of another axis.

[0089] The data analysis unit 285 analyzes the operation of the four-axis robot 25 using synchronized machine data and synchronized measurement data. The data analysis unit 285 performs data analysis on the four-axis robot 25 for the purpose of anomaly diagnosis or predictive maintenance. Existing methods such as machine learning or statistical analysis can be used for these analyses. The results of the analysis performed by the data analysis unit 285 are output or displayed via the output unit 286.

[0090] The visualization terminal 282 is an end device for visualizing and displaying the status of the four-axis robot 25. Examples of the visualization terminal 282 include a display, a smartphone, a tablet terminal, and a laptop computer. The visualization terminal 282 displays the status of the four-axis robot 25 using data about the four-axis robot 25 collected in the database server 293 and an analysis result from the data analysis device 281.

[0091] Fig. Figure 14 is a block diagram showing an example of a functional configuration of the visualization terminal 282 according to the second embodiment. Fig. 14 are components with functions that they include in Fig. The 13 components shown have in common the same reference symbols as those in Fig. 13 used terms are designated, and redundant descriptions of the same are omitted.

[0092] The visualization terminal 282 comprises the input unit 283, the database communication unit 284, the synchronization device 2 and the display unit 287.

[0093] The synchronization device 2 acquires the synchronized machine data and the synchronized measurement data stored in the database server 293 via the database communication unit 284. The synchronization device 2 outputs the synchronized machine data and the synchronized measurement data, and the display unit 287 displays the synchronized machine data and the synchronized measurement data output by the synchronization device 2.

[0094] Fig. Figure 15 is a diagram showing a display example of the visualization terminal 282 based on the analysis result of the data analysis device 281 in the second embodiment. On the lower side of Fig. Figure 15 shows the rotational speed data of a first axis in the four-axis robot 25, which are synchronized machine data. Furthermore, the upper side of Fig. 15. A result of the frequency transformation of the audio data into the moving image data, which are synchronized measurement data, is displayed as a diagram. Time point 0, time t0, and time t1 are time points after synchronization and are time points common to the synchronized machine data and the synchronized measurement data. As in Fig. As shown in Figure 15, it is possible to present the machine's condition in a more easily understandable way by displaying the machine data and the measurement data simultaneously and in parallel, so that they are aligned in a straight line in the same direction, i.e., vertically or horizontally. In particular, the simultaneous display of the machine data and the distribution of the frequency components of the sound data facilitates the assessment of a factor contributing to an anomalous noise or the like, should it occur.

[0095] In the second embodiment, the synchronization device 2 is arranged in the data analysis device 281 or the visualization terminal 282, but can also be arranged in the logger 291 or between the logger 291 and the database server 293. In this case, the database server 293 stores the synchronized machine data and measurement data in the database, thereby eliminating the need for a synchronization process performed by the data analysis device 281 or the visualization terminal 282. This reduces the load on the process and speeds up the response to user actions.

[0096] The synchronization device according to the second embodiment performs a frequency transformation of machine data and / or measurement data to obtain a frequency spectrum of time series and performs a correlation calculation. Therefore, the data to be subjected to the correlation calculation can be limited to data over a specific frequency, thus improving synchronization accuracy. In particular, by performing the correlation calculation with a frequency that represents the operation of a machine, for example, a resonant frequency of a mechanical device, it is possible to carry out a synchronization process with higher accuracy.

[0097] Furthermore, according to the second embodiment, the synchronization device performs an envelope method during the frequency transformation. This makes it possible to reduce the influence of errors in the measurement data during the correlation calculation.

[0098] Furthermore, according to the second embodiment, the synchronization device can synchronize moving image data with measured values ​​from the accelerometer using sound data. Therefore, by using a synchronized moving image, the user can easily identify a phenomenon that occurred throughout the entire device at the time of an accelerometer anomaly.

[0099] The data analysis device according to the second embodiment stores the rotational speed data of the four axes, the measured values ​​of the accelerometer installed on the arm, and the motion image data during the drive in the database server, along with their respective time information. This eliminates the need for a configuration to synchronize the states of the four-axis robot over time, and drive data can be accumulated with a simpler configuration. In particular, a data accumulation environment can be easily integrated into an existing system.

[0100] Furthermore, according to the second embodiment, the data analysis device can synchronize the machine data with the measurement data via the synchronization devices provided in the analysis device and the visualization terminal. This allows the analysis and visualization to be performed on the basis of highly accurate data when the database is used for analysis, visualization, and the like. For example, the user can gain an overview of the data by using an existing method for integrating and displaying the data collected in the database. By using the synchronized machine data and the synchronized measurement data, a clearer causal relationship is established with respect to a phenomenon resulting from a combination of the respective data pieces. In addition, the user can subject the data collected in the database to analysis, machine learning, and the like.The use of synchronized data makes it possible to include the causal relationship of the phenomenon resulting from a combination of data in the analysis and learning, thus obtaining a more accurate result. Third embodiment.

[0101] Fig. Figure 16 is a block diagram showing an example of a functional configuration of a synchronization device 3 according to a third embodiment. Compared to the configuration of the synchronization device 1 according to the first embodiment, which is described in Figure 16, the configuration of the synchronization device 3 is shown in Figure 16. Fig. As shown in Figure 1, in the synchronization device 3 according to the third embodiment, the correlation calculation unit 13 is replaced by a correlation calculation unit 33. In the third embodiment, similar to the second embodiment, a temporal change in a frequency spectrum is used as the data to be synchronized. Therefore, the correlation calculation unit 33 calculates a correlation after a frequency transformation has been performed on the data. In addition, the correlation calculation unit 33 automatically calculates a frequency used for a correlation calculation by peak extraction. Other components are the same as or correspond to those shown in Figure 1. Fig. 1 shown. Identical or corresponding components are shown with the same reference symbols as in . Fig. 1 is designated, and redundant descriptions of the same are omitted.

[0102] Fig. Figure 17 is a block diagram showing an example of a functional configuration of the correlation calculation unit 33 in the third embodiment. Compared to the configuration of the correlation calculation unit 23 in the second embodiment, which is described in Figure 17, the following applies: Fig. As shown in Figure 11, in the third embodiment of the correlation calculation unit 33, a peak extraction unit 337 is provided between the frequency transformation unit 234 and the frequency selection unit 235. Furthermore, the normalization unit 132 is replaced by a normalization unit 332, the correlation rating calculation unit 133 is replaced by a correlation rating calculation unit 333, the correlation value calculation unit 134 is replaced by a correlation value calculation unit 334, and the correlation time difference determination unit 135 is replaced by a correlation time difference determination unit 335. Other components are the same as or correspond to those shown in Figure 11. Fig. 11 shown. Identical or corresponding components are indicated with the same reference numerals as in Fig. 11 is designated, and redundant descriptions of the same are omitted.

[0103] In Fig. 17. The peak extraction unit 337 extracts, with respect to machine data or measurement data to be subjected to frequency transformation, one or more frequencies at which a frequency spectrum exhibits peak values. This frequency is referred to here as the "peak frequency." Regarding the peak frequency, a frequency spectrum of time series, which is the result of a transformation performed by the frequency transformation unit 234, is averaged in one time direction and considered as an average power for each frequency. A frequency at which such an average power reaches a peak value can be used as the peak frequency. Alternatively, a peak value of a frequency spectrum calculated by performing a fast Fourier transform (FFT) on the entire data set can be calculated as the peak frequency.In a case where the results of the frequency transformation unit 234 are not used in the calculation of the peak frequency, the process with the peak extraction unit 337 and the process with the frequency transformation unit 234 can be carried out in reverse order.

[0104] The normalization unit 332 normalizes the machine data and the measurement data, which are adjusted to have the same sampling period, in a similar way to the normalization unit 132. However, in a case where two or more frequencies are extracted by the peak extraction unit 337, a different coefficient is multiplied for each frequency to make all scalings the same.

[0105] The correlation rating calculation unit 333 and the correlation rating calculation unit 334 calculate a correlation rating and a correlation value in a similar manner to the correlation rating calculation unit 133 and the correlation value calculation unit 134. The correlation time difference determination unit 335 determines a correlation time difference in a similar manner to the correlation time difference determination unit 135. However, in a case where two or more frequencies are designated by the peak extraction unit 337, the correlation rating calculation unit 333 calculates a correlation rating sequence for each frequency. The correlation value calculation unit 334 calculates as its correlation value a correlation rating with the maximum correlation strength among the correlation ratings of all frequencies.Then the correlation time difference determination unit 335 determines a time difference corresponding to the correlation value as the correlation time difference.

[0106] The frequency selected by the peak extraction unit 337 or the frequency selection unit 235 can be stored in a memory of the synchronization device 3. Therefore, by selecting the frequency stored in the memory, the processes carried out by the peak extraction unit 337 and the frequency selection unit 235 can be skipped in a second and subsequent process.

[0107] Fig. Figure 18 is a view that represents an example configuration of a simulation system 530, which includes a simulation terminal 38 with a function of the synchronization device 3 according to the third embodiment. Fig. 18 are components with functions that they include in Fig. The 12 components shown have in common the same reference symbols as those in Fig. The 12 used terms are designated, and redundant descriptions of the same are omitted.

[0108] In Fig. 18 The simulation system 530 includes the four-axis robot 25, the controller 261, the servo drives 262 to 265, the accelerometer 271, the logger 291 and the simulation terminal 38.

[0109] The simulation terminal 38 is connected to the controller 261 and the logger 291. The simulation terminal 38 is an end device that performs a drive simulation on the four-axis robot 25. The drive simulation can be performed based on command data and operating data such as the position, speed, and current of each motor of the four-axis robot 25, which are based on the first set of time information, and on acceleration data of the four-axis robot 25, which are based on the third set of time information. The command data or the operating data such as the position, speed, and current of each motor are acquired from the controller 261. The acceleration data of the four-axis robot 25 are acquired from the logger 291. A display, a smartphone, a tablet, or a laptop computer can be used as the simulation terminal 38.

[0110] Fig. Figure 19 is a block diagram showing an example of a functional configuration of the simulation terminal 38 according to the third embodiment. Fig. 19 The simulation terminal 38 comprises a command operation data communication unit 381, an acceleration data communication unit 382, ​​the synchronization device 3, a simulation unit 383 and a display unit 384.

[0111] In the simulation terminal 38, the synchronization device 3 performs a correlation calculation between command data or operating data, representing the operation of a machine, and acceleration data, representing a state of the machine, in order to carry out a synchronization process. The simulation unit 383 performs a drive simulation of the four-axis robot 25 using synchronized machine data and synchronized measurement data output by the synchronization device 3. A result of the simulation by the simulation unit 383 is displayed to the user by the display unit 384.

[0112] The synchronization device according to the third embodiment extracts one or more frequencies at which the frequencies exhibit peak values, normalizes the extracted frequencies so that the scales are equal, and then calculates a correlation time difference at which the correlation strength is maximum. This makes it possible, for example, to perform a highly accurate synchronization process between machine data and sound data without designating a frequency, even for a device where the peak frequency of a drive noise generated by mechanical resonance or the like is unknown. In a case where a peak value is calculated for a device with an unknown peak frequency, using an FFT or the like to determine a peak value for a noise in a specific section could erroneously identify a peak frequency that corresponds to stationary noises such as, for example,A noise is attributable to a surrounding device, which poses a problem. To resolve this problem, the synchronization device according to the third embodiment extracts one or more frequencies at which the frequency spectra exhibit peak values ​​and performs a correlation calculation between the extracted frequency data and the machine data representing the operation of a machine. This allows ambient noise, which is irrelevant to the operation of the machine, to be eliminated, and thus the peak frequency of the drive noise can be reliably determined.

[0113] Furthermore, according to the third embodiment, the simulation terminal can synchronously input command and operating data acquired from the controller, as well as acceleration data acquired from the accelerometer, into the simulation. This allows an estimated accelerometer reading obtained through the simulation to be compared with an actual measured value, thus validating the simulation model. Since the simulation model can be validated, it is easy to update it. During validation and updates, a data synchronization process is performed by the synchronization device, so the simulation unit only needs to run a simulation without the synchronization process. Consequently, it is possible to perform an accurate simulation with a simpler device.Furthermore, in a case where a function of a digital twin is implemented by the simulation terminal, it is also possible to easily perform the validation of the digital twin.

[0114] The configurations described in the above embodiments are merely examples and can be combined with other known technologies; the embodiments can be combined with one another, and some of the configurations can be omitted or modified without deviating from their core. List of reference symbols

[0115] 1, 2, 3 Synchronization device; 10 Synchronization unit; 11 Machine data acquisition unit; 12 Measurement data acquisition unit; 13, 23, 33 Correlation calculation unit; 14 Output unit for synchronized data; 15 Blower; 15a Impeller; 15b Motor; 16 Control device; 18 Inspection terminal; 19a, 19b, 274, 275 Cable; 21 Cable connection; 25 Four-axis robot; 38 Simulation terminal; 131 Scanning adjustment unit; 132, 332 Normalization unit; 133, 333 Correlation evaluation value calculation unit; 134, 334 Correlation value calculation unit; 135, 335 Correlation time difference determination unit; 161, 181 Control unit; 162 Motor drive unit; 182 Storage unit; 183 Monitor; 184 Anomaly diagnostic unit; 185, 287, 384 Display unit; 186 Machine data communication unit; 187 Measurement data communication unit; 234 Frequency transformation unit; 235 Frequency selection unit; 236 Envelope processing unit; 261 Control unit; 262 to 265 Servo drive;271 Accelerometer; 272 Camera; 281 Data analysis device; 282 Visualization terminal; 283 Input unit; 284 Database communication unit; 285 Data analysis unit; 286 Output unit; 291 Logger; 292 Network; 293 Database server; 337 Peak extraction unit; 381 Command operation data communication unit; 382 Acceleration data communication unit; 383 Simulation unit; 500 Drive system; 510 Drive device; 520 Data analysis system; 530 Simulation system.

Claims

[1] Synchronization device (1), comprising: a machine data acquisition unit (11) for acquiring time series information about the drive of a machine (15) as machine data, which is information that is acquired at a time specified by initial time information; a measurement data acquisition unit (12) for acquiring time series information about a state of the machine (15) as measurement data, which is information that includes the intensity of a noise or vibration and is acquired at a time specified by second time information that differs from the first time information; a correlation calculation unit (13) for calculating a correlation time difference, which is a time difference when the strength of the correlation between an absolute value of the machine data and a frequency component of the measurement data is at its maximum, based on the machine data, the measurement data and a time difference when one of the machine data and the measurement data is offset in a positive or negative direction of a time axis; and an output unit (14) for synchronized data for outputting the machine data synchronized with the measurement data based on the correlation time difference as synchronized machine data and for outputting the measurement data synchronized with the machine data based on the correlation time difference as synchronized measurement data, wherein the correlation calculation unit (33) includes a peak extraction unit (337) to obtain from the measurement data a frequency at which a frequency spectrum of the measurement data exhibits peak values, and The strength of the correlation is calculated based on a temporal change in a frequency component where the frequency spectrum of the measurement data shows peak values, and the machine data. [2] Synchronization device (1) according to claim 1, wherein the correlation calculation unit (13) includes a normalization unit (132) to normalize the machine data and the measurement data used to calculate a correlation value. [3] Synchronization device (1) according to claim 1 or 2, wherein the machine (15) is a machine driven by a motor (15b). [4] Synchronization device (1) according to claim 1 or 2, wherein the machine (15) is a machine driven by a motor (15b), and The machine data are data about a rotation angle, a speed, a current or a force of the motor. [5] Computer-readable storage medium that stores a program that causes a computer to perform a function of the synchronization device according to any one of claims 1 to 4. [6] Computer-readable storage medium that stores a program which causes a computer to execute processes that include: a first step (S11) of capturing time series information about the drive of a machine (15) as machine data, which is information that is captured at a time specified by initial time information; a second step (S12) of acquiring time series information about a state of the machine (15) as measurement data, which is information that includes the strength of a noise or vibration and is acquired at a time specified by second time information that differs from the first time information; a third step (S134) of calculating a correlation value with maximum correlation strength between an absolute value of the machine data and a frequency component of the measurement data based on the machine data, the measurement data and a time difference, when one of the machine data and the measurement data is offset in a positive or negative direction of a time axis; a fourth step (S135) of calculating a correlation time difference, which is a time difference when the correlation value is obtained; and a fifth step (S14) of determining the machine data synchronized with the measurement data as synchronized machine data based on the correlation time difference and determining the measurement data synchronized with the machine data as synchronized measurement data based on the correlation time difference, wherein the third step (S134) receives a peak extraction step to obtain a frequency from the measurement data at which a frequency spectrum of the measurement data exhibits peak values, and The strength of the correlation is calculated based on a temporal change in a frequency component where the frequency spectrum of the measurement data shows peak values, and the machine data.

Images