JP7796913B2 - Sign determination device, processing method, and program - Google Patents

Description

This disclosure relates to a sign determination device, processing method, and program.

Patent Document 1 discloses a technique for diagnosing an abnormality in an electric motor that determines whether or not there is an abnormality in the electric motor based on a reference amplitude probability density function calculated from a reference sine wave signal waveform of the rated current of the electric motor and an inspection amplitude probability density function calculated from the current waveform when the electric motor is operating.

JP 2011-257362 A

Patent Document 1 discloses a technique for determining whether or not there is an abnormality in the bearing of a load driven by an electric motor. However, there is a demand for a technique that can detect abnormalities other than bearing abnormalities in addition to detecting abnormalities in bearings.
An object of the present disclosure is to provide a sign determination device, a processing method, and a program that are capable of appropriately determining a sign determination target.

The symptom determination device according to the present disclosure includes a measurement result acquisition unit that acquires measurement results of a current flowing through an electric motor, an analysis unit that performs frequency analysis on the measurement results and breaks them down into frequency components, and a prediction unit that determines whether there are signs of an abnormality in at least one of the electric motor and its load based on time series data of the frequency components, and if it is determined that there are signs of an abnormality, an identification unit that compares the frequency components with part information that associates the frequency components with the part where the abnormality will occur to identify the part, and further identifies the cause of the abnormality based on an image showing the time series data of the frequency components and cause information that associates the cause of the abnormality with the factor .

The processing method according to the present disclosure includes obtaining measurement results of a current flowing through an electric motor, performing frequency analysis on the measurement results to decompose them into frequency components, and determining whether there are signs of an abnormality in at least one of the electric motor and the load on the electric motor based on time series data of the frequency components. If it is determined that there are signs of an abnormality, the method includes comparing the frequency components with part information that associates the frequency components with the part where the abnormality occurs to identify the part, and further identifying the cause of the abnormality based on an image showing the time series data of the frequency components and cause information that associates the cause of the abnormality .

The program according to the present disclosure causes a computer to acquire measurement results of the current flowing through an electric motor, perform frequency analysis on the measurement results to decompose them into frequency components, and determine whether there are signs of an abnormality in at least one of the electric motor and the load on the electric motor based on time series data of the frequency components.If it is determined that there are signs of an abnormality, the program causes a computer to compare the frequency components with part information that associates the frequency components with the part where the abnormality occurs to identify the part, and further identify the cause of the abnormality based on an image showing the time series data of the frequency components and factor information that associates the cause of the abnormality .

According to at least one of the above aspects, it is possible to determine whether there are signs of an abnormality in at least one of the electric motor and the load of the electric motor.

1 is a diagram illustrating a configuration of a sign determination system according to a first embodiment. 1 is a schematic block diagram showing the configuration of a sign determination device according to a first embodiment. FIG. 2 is a diagram illustrating an example of an image according to the first embodiment. FIG. 10 is a diagram illustrating an example of cause information according to the first embodiment. FIG. 4 is a diagram illustrating an example of time information according to the first embodiment. 4 is a flowchart illustrating an example of the operation of the sign determination system according to the first embodiment. FIG. 10 is a diagram illustrating the configuration of a sign determination system according to a second embodiment. FIG. 11 is a first diagram illustrating an example of training data according to the second embodiment. FIG. 10 is a second diagram illustrating an example of training data according to the second embodiment. 10 is a flowchart illustrating an example of the operation of the sign determination system according to the second embodiment. FIG. 10 is a diagram illustrating the configuration of a sign determination system according to a fifth embodiment. FIG. 1 is a schematic block diagram illustrating the configuration of a computer according to at least one embodiment.

First Embodiment
<Configuration of the symptom determination system>
Hereinafter, the configuration of the sign determination system 1 according to the embodiment will be described in detail with reference to the drawings.
The symptom determination system 1 determines a symptom of an abnormality in a load 13 driven by an electric motor 11 .

FIG. 1 is a diagram showing the configuration of a sign determination system 1 according to the first embodiment.
The sign determination system 1 includes a power source 10 , an electric motor 11 , an electric wire 12 , a load 13 , a measuring instrument 16 , a converter 17 , and a sign determination device 100 .

A power source 10 supplies current to an electric motor 11 via an electric line 12 .
The electric motor 11 receives a current from the power source 10 via the electric wire 12. The electric motor 11 that receives the current rotates a shaft 14A provided in the electric motor 11, which in turn rotates a shaft 14B provided in the load 13.

Shaft 14B is rotated by electric motor 11. In other words, load 13 is driven by electric motor 11. Load 13 also includes bearings 15A and 15B that support shaft 14B. Lubricating oil is supplied to bearings 15A and 15B. The lubricating oil reduces friction between bearings 15A and 15B and shaft 14B.

Measuring instrument 16 measures the current flowing through electric wire 12. That is, measuring instrument 16 measures the current flowing through electric motor 11 via electric wire 12. An example of measuring instrument 16 is a current transducer (CT). Measuring instrument 16 measures the current and obtains an analog current waveform.

The converter 17 converts the analog current waveform acquired by the measuring instrument 16 into digital current data. The converter 17 transmits the converted digital current data to the sign determination device 100. In other words, the converter 17 transmits the measurement results of the current flowing through the electric motor 11 to the sign determination device 100.

Configuration of the sign determination device
The configuration of the sign determination device 100 will be described below.
The symptom determination device 100 determines whether or not there is a symptom of an abnormality in the bearings 15A and 15B of the load 13 driven by the electric motor 11.
An example of an abnormality in the bearings 15A and 15B is a change in the shape of the bearings 15A and 15B due to friction between the shaft 14B and the bearings 15A and 15B.

FIG. 2 is a schematic block diagram showing the configuration of the sign determination device 100.
The sign assessment device 100 includes a measurement result acquisition unit 101, an analysis unit 102, an image generation unit 103, a detection unit 104, a first assessment unit 105 (an example of a prediction unit), a calculation unit 106, a second assessment unit 107, an identification unit 108, an update unit 109, an output unit 110, and a memory unit 111.

The measurement result acquisition unit 101 acquires the digital current data transmitted from the converter 17. That is, the measurement result acquisition unit 101 acquires the measurement result of the current flowing through the electric motor 11. The measurement result is digital current data.
The analysis unit 102 uses FFT (Fast Fourier Transform) to decompose the measurement results acquired by the measurement result acquisition unit 101 into a plurality of frequency components.

The image generating unit 103 generates an image showing time-series data of each of a plurality of frequency components and time-series data of energy values associated with the frequency components. The image generated by the image generating unit 103 shows the frequency components as a color graph with time and frequency as axes, and shows energy values on the color graph. The color graph is represented by a set of points.
FIG. 3 shows an example of an image generated by the image generating unit 103. As shown in FIG.

For example, the analysis unit 102 decomposes the measurement result into two frequency components. A color graph of the two decomposed frequency components is shown in Fig. 3. Assume that an abnormality occurs in the bearing 15 of the load 13 at time T2.
The frequency values for the two energy values are constant band values until time T1. If there is no abnormality in the load 13, there is no change in the current flowing through the motor 11, and therefore the values of the two frequency components obtained by decomposing the current measurement results are constant band values until time T1.
The value of the frequency component after time T2 changes by Z from the value of the constant band shown before time T1. This is because, when an abnormality occurs in bearing 15 at time T2 and the measurement result of the current flowing through electric motor 11 is analyzed, the frequency component has a different value from that at time T1.
In FIG. 3, the horizontal axis indicates frequency and the vertical axis indicates time, but the horizontal axis may indicate time and the vertical axis may indicate frequency.

On the color graph of two frequency components, the energy value of each frequency component is shown by color. Examples of colors include blue, green, yellow, and red. Low energy values are shown in blue, and as the energy value increases, the colors are green, yellow, and red, in that order.
The energy value of the frequency component indicating the state of the bearing has a value above a certain level, and therefore is shown in the same color (for example, red) before time T1, from T1 to T2, and after T2.
The image generating unit 103 may display colors not only on the color graph but also in areas other than the color graph. For example, the image generating unit 103 may display areas other than the color graph in a dark blue color in the generated image, thereby indicating that the energy value at the corresponding frequency is zero or close to zero.

The detection unit 104 detects the change value of the frequency component based on the image generated by the image generation unit 103. For example, if the image is a graph like the one in Figure 3, the detection unit 104 detects Z as the change value.

The first determination unit 105 compares the change value detected by the detection unit 104 with a preset first threshold value to determine whether there are signs of an abnormality in the bearing 15. For example, if the change value detected by the detection unit 104 is equal to or greater than the first threshold value, the first determination unit 105 determines that there are signs of an abnormality in the bearing 15. Furthermore, if the change value detected by the detection unit 104 is not equal to or greater than the first threshold value, the first determination unit 105 determines that there are no signs of an abnormality in the bearing 15. If an abnormality occurs in the bearing 15, the value of the frequency component changes, so by comparing the change value with a certain threshold value, it is possible to determine whether there are signs of an abnormality in the bearing 15.

The calculation unit 106 calculates the similarity based on the feature amount of the image associated with the cause of the abnormality in the bearing 15 and the feature amount of the image generated by the image generation unit 103. Examples of the cause include poor lubrication, poor installation, intrusion of foreign matter, rust, and insufficient clearance.
A user of the sign determination system 1 records in advance factor information, which is information associating factors with images, in the storage unit 111. The image is an image showing time-series data of frequency components and energy values.
4 shows an example of an image in the factor information. For example, a user of the sign assessment system 1 records factor information in which the image A in FIG. 4 is associated with factor A in the storage unit 111. The user of the sign assessment system 1 also records factor information in which the image B in FIG. 4 is associated with factor B in the storage unit 111.

For example, the calculation unit 106 extracts image features associated with factors, such as those shown in FIG. 4 , using a convolutional neural network (CNN) technique. The calculation unit 106 also extracts feature features of the images generated by the image generation unit 103 using a convolutional neural network technique. The calculation unit 106 calculates similarities between the extracted feature features. The calculation unit 106 calculates similarities between feature features of multiple images associated with factors in the factor information and feature features of the image generated by the image generation unit 103. That is, the calculation unit 106 calculates multiple similarities between each of the images in the multiple pieces of factor information.
The operation of the calculation unit 106 can be performed in two ways: first, based on known factors and known images before the user of the sign assessment system 1 uses the system; and second, based on newly occurring factors and newly occurring images when the user of the sign assessment system 1 uses the system.

When the first determination unit 105 determines that there is a sign, the second determination unit 107 determines whether the similarity calculated by the calculation unit 106 is equal to or greater than a preset second threshold. When the calculation unit 106 calculates multiple similarities, the second determination unit 107 determines whether each of the multiple similarities is equal to or greater than the second threshold.

If the second determination unit 107 determines that the similarity is equal to or greater than the second threshold, the identification unit 108 identifies a factor associated with the similarity calculated by the calculation unit 106. For example, if the similarity with the image A in Fig. 4 is equal to or greater than the second threshold, the second determination unit 107 determines that the similarity is equal to or greater than the second threshold. The identification unit 108 identifies a factor associated with the image A in Fig. 4.
Furthermore, it is also possible to identify the area when the second judgment unit 107 judges that the value is equal to or greater than the second threshold, and it is also possible to identify the area when at least one of the first judgment unit 105 or the second judgment unit 107 judges that there are signs or that the value is equal to or greater than the second threshold.

Furthermore, when the first determination unit 105 determines that there is a sign of an abnormality, the identification unit 108 identifies the abnormality by comparing the frequency component with the part information in which the frequency component is associated with the part where the abnormality occurs. Examples of the abnormality include the inner ring of the bearing 15, the outer ring of the bearing 15, and the ball between the inner ring and the outer ring of the bearing 15.
A user of the sign assessment system 1 records body part information in advance in the storage unit 111. Note that the user of the sign assessment system 1 may also record newly generated body part information in the storage unit 111 when using the sign assessment system 1.

For example, suppose the value of the frequency component in Figure 3 changes by Z from F1, and the first determination unit 105 determines that there is a sign of an abnormality in bearing 15. In this case, the identification unit 108 compares the frequency component value F1 with the part information. If the part information associates the frequency component value F1 with the inner ring of bearing 15, the identification unit 108 identifies the inner ring of bearing 15 as the part where an abnormality will occur.

The identifying unit 108 also compares the identified cause with time information that associates the cause with the time until the abnormality occurs, and further identifies the time.
The user of the sign assessment system 1 records time information in advance in the storage unit 111. Note that the user of the sign assessment system 1 may also record newly generated time information in the storage unit 111 when using the sign assessment system 1.

5 is a diagram illustrating an example of time information according to the first embodiment. For example, assume that the identification unit 108 has identified a cause A. As illustrated in FIG. 5 , if the cause A is associated with a time T3 until an abnormality occurs in the time information, the identification unit 108 identifies the time T3.
The identifying unit 108 may identify the time when the abnormality occurs instead of the time T3 until the abnormality occurs.

If the second determination unit 107 determines that the similarity is not equal to or greater than the second threshold, the update unit 109 accepts an input from the outside and updates the factor information.
For example, suppose it is determined that the similarity of the image generated by the image generation unit 103 to the image in the factor information is not equal to or greater than the second threshold. The output unit 110 outputs the image generated by the image generation unit 103 to a display device (not shown) included in the sign determination system 1. A user of the sign determination system 1 checks the output image through the display device. The user of the sign determination system 1 identifies the cause of the abnormality in the bearing 15 using a device other than the sign determination device 100. The user of the sign determination system 1 associates the newly identified factor with the output image and inputs this as factor information to the sign determination system 1. The update unit 109 accepts the input and updates the factor information.
This allows the factor information to be updated even for factors that the second determination unit 107 cannot determine, thereby increasing the number of factors that the second determination unit 107 can determine.

The output unit 110 outputs the content identified by the identification unit 108 to a notification device included in the sign determination system 1. Examples of the notification device include a display device and a speaker. The signals output by the output unit 110 include a signal representing an image and a signal related to sound.
For example, the output unit 110 outputs to a display device the cause, location, and time identified by the identification unit 108. The user of the symptom assessment system 1 can check the display on the display device the cause of the abnormality occurring in the bearing 15, the location where the abnormality will occur in the bearing 15, and the time until the abnormality occurs in the bearing 15. In this way, the output unit 110 outputs a signal showing an image to the display device, so that the user can easily understand the content identified by the identification unit 108.

The storage unit 111 stores the cause information, body part information, and time information recorded by the user of the sign determination system 1. An example of the storage unit 111 is a hard disk.

<<Operation of the Premonition Judgment System>>
The operation of the sign determination system 1 will be described below.
FIG. 6 is a flowchart showing the operation of the sign determination system 1.

The measuring instrument 16 measures the current and acquires an analog current waveform (step S1).
The converter 17 converts the analog current waveform acquired in step S1 into digital data (step S2).

The measurement result acquisition unit 101 acquires the measurement result, which is digital data, from the converter 17 (step S3).
The analysis unit 102 resolves the measurement results acquired in step S3 into frequency components by FFT (step S4).

The image generating unit 103 generates an image of a color graph showing the time-series data of the frequency components and the energy values (step S5).
The detector 104 detects the change value of the frequency component based on the image generated in step S5 (step S6).

The first determination unit 105 compares the change value detected in step S6 with a first threshold value and determines whether there is a sign of an abnormality in the bearing 15 (step S7).
If it is determined that there is no sign of abnormality (step S7: NO), the sign determination system 1 returns to step S1 and performs the operations from step S1.
On the other hand, if it is determined that there are signs of an abnormality (step S7: YES), the calculation unit 106 calculates the similarity between the image features in the cause information and the image features generated in step S5 (step S8).

The second determination unit 107 determines whether the similarity calculated in step S8 is equal to or greater than a second threshold value (step S9).
If it is determined that the similarity is equal to or greater than the second threshold (step S9: YES), the identifying unit 108 identifies factors related to the similarity (step S10).
The identifying unit 108 identifies a portion of the bearing 15 where an abnormality occurs (step S11). The identifying unit 108 also identifies the time until the abnormality occurs in the bearing 15 (step S12).

The output unit 110 outputs the content identified by the identification unit 108 to the notification device (step S13).
The notification device displays the content identified by the identification unit 108 to the user of the sign determination system 1 (step S14).

On the other hand, if it is determined that the similarity is less than the second threshold (step S9: NO), the output unit 110 outputs the image generated in step S5 to the display device (step S15). Then, the identification unit 108 identifies the part of the bearing 15 where the abnormality occurs (step S11).

Based on the display output in step S15, the user of the sign assessment system 1 identifies the cause of the abnormality using a device separate from the sign assessment device 100. In step S16, the user of the sign assessment system 1 inputs new factor information related to the identified factor into the sign assessment system 1.
The update unit 109 updates the cause information recorded in the storage unit 111 with the cause information input in step S17.

<Actions and Effects>
The symptom determination device 100 according to the present disclosure includes a measurement result acquisition unit 101 that acquires measurement results of the current flowing through the electric motor 11, an analysis unit 102 that performs frequency analysis on the measurement results and resolves them into frequency components, an image generation unit 103 that generates an image showing time series data of the frequency components, and a first determination unit 105 that determines, based on the image, whether or not there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11.

The symptom assessment device 100 can determine whether there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11, based on an image showing the frequency components of the electric motor 11's current. Furthermore, the symptom assessment device 100 can determine whether there are signs of an abnormality from a long distance, even if it is not installed near the bearing 15, where vibrations occur and temperature changes affect it.

In addition, the analysis unit 102 of the sign determination device 100 decomposes the signal into multiple frequency components using FFT, and the image shows the time series data for each of the multiple frequency components.

The symptom assessment device 100 can determine whether there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11 based on an image showing multiple frequency components resolved by FFT.

The image of the sign assessment device 100 further shows time-series data of energy values associated with frequency components, and the device is equipped with a detection unit 104 that detects change values of the frequency components based on the image, and a first assessment unit 105 that compares the change value with a pre-set first threshold value to determine whether or not a sign is present.

The symptom assessment device 100 can determine whether there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11 based on an image showing frequency components and energy values.

In addition, the image of the sign determination device 100 shows the frequency versus energy value as a color graph with time and frequency as axes, and the energy value is shown by color on the color graph.

The symptom assessment device 100 can determine whether there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11 based on a color graph image showing frequency components and energy values.

The symptom assessment device 100 also includes an identification unit 108 that, when it is determined that a symptom exists, identifies the part by comparing the frequency component with part information that associates the frequency component with the part where the abnormality occurs.

The symptom assessment device 100 identifies the location where an abnormality will occur based on the location information. This allows the user of the symptom assessment device 100 to identify the location where an abnormality will occur in the bearing 15.

Furthermore, if the identification unit 108 of the sign assessment device 100 determines that a sign exists, it further identifies the cause based on the cause information that associates the image with the cause of the abnormality.

The symptom assessment device 100 identifies the cause of the abnormality based on the cause information. This allows the user of the symptom assessment device 100 to identify the cause of the abnormality in the bearing 15.

The sign assessment device 100 also includes a calculation unit 106 that calculates similarity based on the image feature amounts associated with the factors and the feature amounts of the image generated by the image generation unit 103, and a second assessment unit 107 that, when it is determined that a sign exists, determines whether the similarity is equal to or greater than a predetermined second threshold, and an identification unit 108 identifies the factor related to the calculated similarity when it is determined that the similarity is equal to or greater than the second threshold.

The symptom assessment device 100 calculates similarity based on the image features and identifies the cause of the abnormality. This allows the user of the symptom assessment device 100 to identify the cause of the abnormality in the bearing 15.

The sign determination device 100 also includes an update unit 109 that accepts input from outside and updates the factor information when it is determined that the similarity is not equal to or greater than the second threshold.

If the sign assessment device 100 is unable to identify a cause, it accepts input from outside and updates the cause information. This allows the sign assessment device 100 to identify more causes.

In addition, the identification unit 108 of the sign assessment device 100 compares the identified cause with time information that associates the cause with the time until the abnormality occurs, and further identifies the time.

The symptom assessment device 100 determines the time until an abnormality occurs based on the time information. This allows the user of the symptom assessment device 100 to determine the time until an abnormality occurs in the bearing 15.

The sign determination device 100 also includes an output unit 110 that outputs the content identified by the identification unit 108.

The sign assessment device 100 outputs the cause, area, time, and other details identified by the identification unit 108. This allows the user of the sign assessment device 100 to confirm the cause, area, time, and other details identified by the identification unit 108.

The symptom determination method disclosed herein includes obtaining measurement results of the current flowing through the electric motor 11, performing frequency analysis on the measurement results to resolve them into frequency components, generating an image showing time-series data of the frequency components, and determining, based on the image, whether there are any signs of an abnormality in the bearings 15 of the load 13 driven by the electric motor 11.

By using the sign determination method, a user can determine whether there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11, based on an image showing the frequency components of the electric motor 11's current.

The program disclosed herein causes a computer to acquire measurement results of the current flowing through the electric motor 11, perform frequency analysis on the measurement results to resolve them into frequency components, generate an image showing time-series data of the frequency components, and determine, based on the image, whether there are any signs of an abnormality in the bearings 15 of the load 13 driven by the electric motor 11.

By running the program, a user of the program can determine whether there are signs of an abnormality in the bearing 15 of the load 13 driven by the electric motor 11, based on an image showing the frequency components of the electric motor 11's current.

Second Embodiment
<Configuration of the symptom determination system>
Hereinafter, the configuration of the sign determination system 1 according to the embodiment will be described in detail with reference to the drawings.
The symptom assessment system 1 according to the first embodiment assesses symptoms of abnormalities in the bearings 15A and 15B of the load 13. The symptom assessment system 1 according to the second embodiment assesses symptoms of abnormalities including at least one of the bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation.

The configuration of the sign determination system 1 according to the second embodiment is similar to the configuration of the sign determination system 1 according to the first embodiment shown in Figure 1, so a description of each component will be omitted. From here on, the same components will be denoted by the same reference numerals in the description of this embodiment.

Configuration of the sign determination device
The configuration of the sign determination device 100 will be described below.
The symptom determination device 100 determines whether there are any signs of abnormality including at least one of bearing 15, breakage of the rotor bar in the electric motor 11, change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation.

The configuration of the sign assessment device 100 is similar to that of the sign assessment device 100 according to the first embodiment shown in Figure 2, so a description of each component will be omitted and only the different components will be described. Furthermore, the same reference numerals will be used hereafter to describe this embodiment.

When the image generated by the image generating unit 103 indicates signs of an abnormality, the frequency components change significantly, as in the image shown in Figure 3, similar to signs of an abnormality in the bearing 15 according to the first embodiment.

For example, the analysis unit 102 decomposes the measurement results into two frequency components. Assume that the image generated by the image generation unit 103 shows a color graph of the two decomposed frequency components shown in Fig. 3, and that an abnormality occurs at time T2, including at least one of bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation.
In this case, as in the first embodiment, the value of the frequency component after time T2 changes by Z from the value of the fixed band shown before time T1. This is because at time T2, an abnormality occurs including at least one of bearing 15, breakage of a rotor bar in electric motor 11, a change in tension of the belt connecting electric motor 11 and load 13, and cavitation, and when the measurement results of the current flowing through electric motor 11 are resolved, the frequency component has a different value from that at time T1.

The energy value of the frequency component indicating the state of bearing 15, breakage of the rotor bar in electric motor 11, and change in tension of the belt connecting electric motor 11 and load 13 has a value equal to or greater than a certain level. If there is an abnormality including at least one of bearing 15, breakage of the rotor bar in electric motor 11, and change in tension of the belt connecting electric motor 11 and load 13 before time T1, between T1 and T2, or after T2, the energy value of this frequency component may change.
Cavitation is identified by an increase in the energy values of multiple frequency components. When cavitation occurs at time T1, it can be identified by an increase in the energy values of multiple frequency components after time T1 (e.g., from blue to yellow).

The detection unit 104 detects the change value of the frequency component based on the image generated by the image generation unit 103. For example, if the image is a graph like the one in Figure 3, the detection unit 104 detects Z as the change value.

The first determination unit 105 compares the change value detected by the detection unit 104 with a predetermined first threshold value and determines whether there are signs of an abnormality including at least one of the bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation. For example, if the change value detected by the detection unit 104 is equal to or greater than the first threshold value, the first determination unit 105 determines that there are signs of an abnormality including at least one of the bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation. Furthermore, if the change value detected by the detection unit 104 is not equal to or greater than the first threshold value, the first determination unit 105 determines that there are no signs of an abnormality including at least one of the bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation. If an abnormality occurs, including at least one of bearing 15, breakage of a rotor bar in electric motor 11, a change in tension in the belt connecting electric motor 11 and load 13, or cavitation, the value of the frequency component will change. Therefore, by comparing the change value with a certain threshold, it is possible to determine whether there are signs of an abnormality, including at least one of bearing 15, breakage of a rotor bar in electric motor 11, a change in tension in the belt connecting electric motor 11 and load 13, or cavitation.

The calculation unit 106 calculates the similarity based on the image feature values associated with the causes of abnormalities, including at least one of the bearing 15, broken rotor bars in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and the load 13, and cavitation, and the image feature values generated by the image generation unit 103. Examples of causes include poor lubrication, improper installation, foreign matter intrusion, rust, and insufficient clearance for the bearing 15; aging deterioration for broken rotor bars in the electric motor 11 and changes in tension in the belt connecting the electric motor 11 and the load 13; and an abnormality in the pump (not shown) operated by the electric motor 11 for cavitation.

Similar to the first embodiment, when the first determination unit 105 determines that there is a sign of an abnormality, the identification unit 108 identifies the abnormality by comparing the frequency component with the part information that associates the frequency component with the part where the abnormality occurs. Examples of the above parts include the inner ring of the bearing 15, the outer ring of the bearing 15, and the ball between the inner and outer rings of the bearing 15 for the bearing 15, the rotor bar for a broken rotor bar in the electric motor 11, the belt for a change in tension of the belt connecting the electric motor 11 and the load 13, and the pump operated by the electric motor 11 for cavitation.

For example, suppose the value of the frequency component in Figure 3 changes by Z from F1, and the first determination unit 105 determines that there is a sign of an abnormality regarding breakage of a rotor bar in the electric motor 11. In this case, the identification unit 108 compares the frequency component value F1 with the part information. If the frequency component value F1 and the rotor bar are associated with the part information, the identification unit 108 identifies the rotor bar as the part where an abnormality will occur.

As in the first embodiment, when the second determination unit 107 determines that the similarity is not equal to or greater than the second threshold, the update unit 109 receives an input from the outside and updates the factor information.
For example, suppose it is determined that the similarity of the image generated by the image generation unit 103 to the image in the cause information is not equal to or greater than the second threshold. If this determination is made, the user of the sign assessment system 1 can use a device other than the sign assessment device 100 to identify the cause of the abnormality, which includes at least one of the bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation.

As in the first embodiment, the output unit 110 outputs the details identified by the identification unit 108 to the notification device included in the sign determination system 1 .
For example, the output unit 110 outputs the cause, location, and time identified by the identification unit 108 to a display device. A user of the symptom assessment system 1 can check, through the display on the display device, the cause of an abnormality, including at least one of the bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension of a belt connecting the electric motor 11 and the load 13, and cavitation, the location where the abnormality will occur, and the time until the abnormality will occur. In this way, the output unit 110 outputs a signal showing an image to the display device, so that the user can easily understand the content identified by the identification unit 108.

<<Operation of the Premonition Judgment System>>
The operation of the sign determination system 1 will be described below.
The flowchart showing the operation of the sign assessment system 1 in this case is similar to the flowchart shown in Figure 6, which shows the operation of the sign assessment system 1 according to the first embodiment, so we will omit the explanation of each processing step and only explain the different processing steps.

In step S7, the first judgment unit 105 compares the change value detected in step S6 with the first threshold value and determines whether there are any signs of abnormality including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation.
In step S11, which follows step S10, the identifying unit 108 identifies a location where an abnormality occurs, including at least one of bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation. In step S12, the identifying unit 108 identifies the time until an abnormality occurs, including at least one of bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension of the belt connecting the electric motor 11 and the load 13, and cavitation.

On the other hand, in step S11, which follows step S15, the identification unit 108 identifies the location where an abnormality occurs, including at least one of the bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

In this embodiment, the analysis unit 102 decomposes the signal into a plurality of frequency components, and a configuration using the frequency components has been described. However, it is possible to obtain a peak frequency component from the frequency components decomposed by the analysis unit 102, and adopt a configuration using the peak frequency component.
At this time, a threshold value is set for the frequency component that becomes a peak, and the frequency component that becomes a peak when the threshold value is exceeded is acquired.
Furthermore, by identifying the energy value of the peak frequency component by color, it is possible to determine whether or not there is a sign of an abnormality even if a peak occurs in an unexpected frequency component.

<Actions and Effects>
The symptom assessment device 100 according to the present disclosure includes a measurement result acquisition unit 101 that acquires measurement results of the current flowing through the electric motor 11, an analysis unit 102 that performs frequency analysis on the measurement results and breaks them down into frequency components, an image generation unit 103 that generates an image showing time-series data of the frequency components, and a first assessment unit 105 that determines, based on the image, whether or not there are signs of an abnormality including at least one of bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension of a belt connecting the electric motor 11 and load 13, and cavitation.

The symptom assessment device 100 can determine whether there are any signs of abnormality, including at least one of the following: bearing 15, broken rotor bars in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and the load 13, and cavitation, based on an image showing the frequency components of the electric motor 11's current. Furthermore, the symptom assessment device 100 can determine whether the above signs exist from a distance, even if it is not installed near the electric motor 11 or the load 13.

In addition, the analysis unit 102 of the sign determination device 100 decomposes the signal into multiple frequency components using FFT, and the image shows the time series data for each of the multiple frequency components.

Based on an image showing multiple frequency components decomposed by FFT, the symptom assessment device 100 can determine whether there are any signs of abnormality, including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

The image of the sign assessment device 100 further shows time-series data of energy values associated with frequency components, and the device is equipped with a detection unit 104 that detects change values of the frequency components based on the image, and a first assessment unit 105 that compares the change value with a pre-set first threshold value to determine whether or not a sign is present.

Based on images showing frequency components and energy values, the symptom assessment device 100 can determine whether there are any signs of abnormality, including at least one of the following: bearing 15, broken rotor bars in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and load 13, and cavitation.

In addition, the image of the sign determination device 100 shows the frequency versus energy value as a color graph with time and frequency as axes, and the energy value is shown by color on the color graph.

Based on a color graph image showing frequency components and energy values, the symptom assessment device 100 can determine whether there are any signs of abnormality, including at least one of the following: bearing 15, broken rotor bars in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and load 13, and cavitation.

The symptom assessment device 100 also includes an identification unit 108 that, when it is determined that a symptom exists, identifies the part by comparing the frequency component with part information that associates the frequency component with the part where the abnormality occurs.

The symptom assessment device 100 identifies the location where an abnormality will occur based on the location information. This allows the user of the symptom assessment device 100 to identify the location where an abnormality will occur, including at least one of the following: bearing 15, broken rotor bar in the electric motor 11, change in tension in the belt connecting the electric motor 11 and load 13, and cavitation.

Furthermore, if the identification unit 108 of the sign assessment device 100 determines that a sign exists, it further identifies the cause based on the cause information that associates the image with the cause of the abnormality.

The symptom assessment device 100 identifies the cause of the abnormality based on the cause information. This allows the user of the symptom assessment device 100 to identify the cause of the abnormality, including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

The sign assessment device 100 also includes a calculation unit 106 that calculates similarity based on the image feature amounts associated with the factors and the feature amounts of the image generated by the image generation unit 103, and a second assessment unit 107 that, when it is determined that a sign exists, determines whether the similarity is equal to or greater than a predetermined second threshold, and an identification unit 108 identifies the factor related to the calculated similarity when it is determined that the similarity is equal to or greater than the second threshold.

The symptom assessment device 100 calculates similarity based on the image features and identifies the cause of the abnormality. This allows the user of the symptom assessment device 100 to identify the cause of the abnormality, including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and load 13, and cavitation.

The sign determination device 100 also includes an update unit 109 that accepts input from outside and updates the factor information when it is determined that the similarity is not equal to or greater than the second threshold.

If the sign assessment device 100 is unable to identify a cause, it accepts input from outside and updates the cause information. This allows the sign assessment device 100 to identify more causes.

In addition, the identification unit 108 of the sign assessment device 100 compares the identified cause with time information that associates the cause with the time until the abnormality occurs, and further identifies the time.

The symptom assessment device 100 determines the time until an abnormality occurs based on the time information. This allows the user of the symptom assessment device 100 to determine the time until an abnormality occurs, including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

The sign determination device 100 also includes an output unit 110 that outputs the content identified by the identification unit 108.

The sign assessment device 100 outputs the cause, area, time, and other details identified by the identification unit 108. This allows the user of the sign assessment device 100 to confirm the cause, area, time, and other details identified by the identification unit 108.

The sign determination method disclosed herein includes obtaining measurement results of the current flowing through the electric motor 11, performing frequency analysis on the measurement results to resolve them into frequency components, generating an image showing time-series data of the frequency components, and determining, based on the image, whether there are signs of an abnormality including at least one of bearing 15, breakage of a rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and load 13, and cavitation.

By using the sign determination method, a user can determine whether there are signs of abnormality, including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation, based on an image showing the frequency components of the electric motor 11's current.

The program disclosed herein causes a computer to acquire measurement results of the current flowing through the electric motor 11, perform frequency analysis on the measurement results to decompose them into frequency components, generate an image showing time-series data of the frequency components, and determine, based on the image, whether there are signs of abnormality including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

By running the program, a user of the program can determine whether there are any signs of abnormality, including at least one of the following: bearing 15, broken rotor bars in the motor 11, changes in tension in the belt connecting the motor 11 and the load 13, and cavitation, based on an image showing the frequency components of the current in the motor 11.

Third Embodiment
The sign determination device 100 according to the third embodiment will be described below.
In the sign assessment device 100 according to the first and second embodiments, the detection unit 104 detects a change in frequency components based on the generated image, and the first assessment unit 105 determines that a sign of abnormality exists when the change detected by the detection unit 104 is equal to or greater than a first threshold. The calculation unit 106 extracts features of the image generated by the image generation unit 103 using a convolutional neural network technique and calculates a similarity between the extracted features. The second assessment unit 107, when the first assessment unit 105 determines that a sign exists, determines whether the similarity calculated by the calculation unit 106 is equal to or greater than a predetermined second threshold. If the second assessment unit 107 determines that the similarity is equal to or greater than the second threshold, the identification unit 108 identifies a factor related to the similarity calculated by the calculation unit 106. However, in the symptom assessment device 100 according to the third embodiment, the processing unit 112 described below determines whether there are any signs of an abnormality, and if there are any signs of an abnormality, determines the time until the abnormality occurs, the location where the abnormality occurred, and the cause of the abnormality.

As shown in FIG. 7, the sign determination device 100 includes a measurement result acquisition unit 101, an analysis unit 102, a processing unit 112 (an example of a processing unit), an update unit 109, an output unit 110, and a memory unit 111.

The processing unit 112 predicts whether there are signs of an abnormality, the time until the abnormality actually occurs, the location where the abnormality occurs, and the cause of the abnormality, based on each piece of time series data decomposed into multiple frequency components by the analysis unit 102 using FFT. The processing unit 112 predicts whether there are signs of an abnormality, the time until the abnormality actually occurs, the location where the abnormality occurs, and the cause of the abnormality, for example, by using a trained model (e.g., a convolutional neural network) whose parameters are determined using training data, which is a type of machine learning. Here, we will explain the trained models that the processing unit 112 uses for each prediction.

The trained model used by the processing unit 112 to predict signs of abnormality in the bearing 15 is referred to as the first trained model. The trained model used by the processing unit 112 to predict signs of abnormality including at least one of the bearing 15, breakage of a rotor bar in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and the load 13, and cavitation is referred to as the second trained model. In the following explanation, to make it easier to understand how parameters are determined, the first trained model that predicts signs of abnormality in the bearing 15 is cited as a specific example of a trained model.

(First trained model)
First, the first trained model will be described. The processing unit 112 predicts signs of abnormality in the bearing 15 based on each piece of time-series data decomposed into multiple frequency components by FFT performed by the analysis unit 102. Here, a trained model will be described for the case where the processing unit 112 predicts signs of abnormality in the bearing 15 based on each piece of time-series data decomposed into multiple frequency components by FFT performed by the analysis unit 102.

In this case, each piece of time series data decomposed into multiple frequency components by FFT performed by the analysis unit 102 becomes one piece of input data. Furthermore, the time until an abnormality occurs for that input data, the location of the abnormality, and the cause of the abnormality become one piece of output data. A combination of input data and the output data corresponding to that input data becomes one piece of training data. For example, before the symptom assessment device 100 predicts an abnormality, the output data (i.e., data indicating the time until the abnormality actually occurs, the location of the abnormality, and the cause of the abnormality) is identified for each piece of input data, which is time series data decomposed into multiple frequency components by FFT performed by the analysis unit 102 when used by another device to predict an abnormality. Alternatively, for example, by conducting experiments or simulations, the output data (i.e., data indicating the time until the abnormality actually occurs for that time series data, the location of the abnormality, and the cause of the abnormality) is identified for each piece of input data, which is time series data decomposed into multiple frequency components by FFT performed by the analysis unit 102. In this way, training data consisting of multiple data sets combining input data and output data can be prepared. Training data is data used to determine parameter values in a learning model where the parameter values have not yet been determined.

Figure 8 is a diagram showing an example of training data. One set of data consists of input data, which is time-series data decomposed into multiple frequency components by FFT performed by the analysis unit 102, and output data for that input data (i.e., data indicating the time until an abnormality actually occurs, the location where the abnormality occurred, and the cause of the abnormality). In the example shown in Figure 8, the training data includes 10,000 sets of data.

For example, consider the case where parameters in a learning model are determined using training data consisting of 10,000 sets of data, as shown in Figure 8. In this case, the training data is divided into training data, evaluation data, and test data. Examples of the ratios of training data, evaluation data, and test data include 70%, 15%, 15%, 95%, 2.5%, and 2.5%. For example, suppose the training data, data #1 to #10,000, is divided into training data (data #1 to #7000), evaluation data (data #7001 to #8500), and 15% test data (data #8501 to #10,000). In this case, data #1, which is the training data, is input into a convolutional neural network, which is a learning model. The convolutional neural network outputs either the absence of an anomaly, the time until an anomaly actually occurs, the location of the anomaly, and the cause of the anomaly. Training data input data is input into the convolutional neural network, and each time the convolutional neural network outputs information indicating the absence of an abnormality, the time until an abnormality actually occurs, the location of the abnormality, and the cause of the abnormality (in this case, each time each piece of data #1 to #7000 is input into the convolutional neural network), the parameters indicating the weighting of the data connections between nodes are changed (i.e., the convolutional neural network model is changed) by performing backpropagation, for example, in response to the output. In this way, training data is input into the neural network and the parameters are adjusted.

Next, the input data of the evaluation data (data #7001 to #8500) is input in order to the convolutional neural network whose parameters have been changed using the training data. Depending on the input evaluation data, the convolutional neural network outputs either that no abnormality has occurred, or the time until the abnormality actually occurs, the location where the abnormality occurred, and the cause of the abnormality. Here, if the data output by the convolutional neural network differs from the output data associated with the input data in Figure 8, the parameters are changed so that the output of the convolutional neural network becomes the output data associated with the input data in Figure 8. The convolutional neural network (i.e., the learning model) whose parameters have been determined in this way is the first trained model.

Next, as a final check, the test data (data #8501 to #10000) are input sequentially to the convolutional neural network of the first trained model. Depending on the input test data, the convolutional neural network of the trained model outputs either that no abnormality has occurred, or the time until the abnormality actually occurs, the location where the abnormality occurred, and the cause of the abnormality. If the output data output by the convolutional neural network of the trained model for all test data matches the output data associated with the input data in Figure 8, the convolutional neural network of the first trained model is the desired model. Furthermore, if the output data output by the convolutional neural network of the first trained model for even one piece of test data does not match the output data associated with the input data in Figure 8, the parameters of the trained model are determined using new training data. The above-mentioned determination of the parameters of the trained model is repeated until a first trained model with the desired parameters is obtained. When a first trained model with the desired parameters is obtained, the first trained model is recorded in the memory unit 111.

(Second trained model)
Next, the second trained model will be described. FIG. 9 is a diagram showing an example of training data. The parameters of the second trained model can be determined in the same manner as the first trained model described above. However, as shown in FIG. 9, abnormalities can occur not only in the bearings but also in other locations such as broken rotor bars in the electric motor 11, changes in tension in the belt connecting the electric motor 11 and the load 13, and cavitation, and there are multiple causes of abnormalities.

<<Operation of the Premonition Judgment System>>
The operation of the sign determination system 1 will be described below.
FIG. 10 is a flowchart showing the operation of the sign determination system 1.

The measuring instrument 16 measures the current and acquires an analog current waveform (step S1).
The converter 17 converts the analog current waveform acquired in step S1 into digital data (step S2).

The measurement result acquisition unit 101 acquires the measurement result, which is digital data, from the converter 17 (step S3).
The analysis unit 102 resolves the measurement results acquired in step S3 into frequency components by FFT (step S4).

The processing unit 112 inputs each piece of time series data decomposed into multiple frequency components by the analysis unit 102 using FFT into the trained model (step S21). The trained model of the processing unit 112 outputs the time until the abnormality actually occurs, the location where the abnormality occurred, and the cause of the abnormality (step S22).

The output unit 110 outputs the information identified by the processing unit 112 (i.e., the time until the abnormality actually occurs, the location where the abnormality occurred, and the cause of the abnormality) to the alarm device (step S23).
The notification device notifies the user of the sign determination system 1 of the content identified by the processing unit 112 (step S24).

<Actions and Effects>
In the symptom assessment device 100 according to the present disclosure, the processing unit 112 predicts whether or not there are signs of an abnormality, the time until the abnormality actually occurs, the location where the abnormality will occur, and the cause of the abnormality, based on each piece of time series data that has been decomposed into multiple frequency components by FFT performed by the analysis unit 102.

Based on the prediction results from the processing unit 112, the symptom assessment device 100 can determine whether there are any signs of abnormality, including at least one of the following: bearing 15, breakage of the rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

Fourth Embodiment
The sign determination device 100 according to the fourth embodiment will be described below.
The symptom assessment device 100 may provide, as the cause information, information that associates a cause of an abnormality in the bearing 15, an image, and a countermeasure for the cause. This allows the user of the symptom assessment device 100 to check not only the cause but also an appropriate countermeasure for the cause.

In addition, the update unit 109 of the sign determination device 100 may accept input from outside and update the time information or body part information.

Fifth Embodiment
The sign determination device 100 according to the fifth embodiment will be described below.
11 is a diagram showing an example of a sign assessment system 1 according to the fifth embodiment. In the sign assessment system 1, the sign assessment device 100 may be configured as a system capable of communicating with a terminal device 200 having a request unit 2001. In this case, the request unit 2001 of the terminal device 200 may be configured to issue a request to the sign assessment device 100 to assess signs of abnormalities in the electric motor 11 and the load 13 of the electric motor 11.
Furthermore, the terminal device 200 may have the function of the output unit 110, and may be configured to output the abnormality sign determination result of the sign determination device 100 as the aforementioned notification device.

<Computer Configuration>
FIG. 12 is a schematic block diagram illustrating the configuration of a computer according to at least one embodiment.
The computer 1100 includes a processor 1110 , a main memory 1120 , storage 1130 , and an interface 1140 .
The above-described sign assessment device 100 is implemented in a computer 1100. The operations of the above-described processing units are stored in the form of a program in a storage 1130. The processor 1110 reads the program from the storage 1130, loads it into the main memory 1120, and executes the above-described processing in accordance with the program. The processor 1110 also allocates storage areas in the main memory 1120 corresponding to the above-described storage units in accordance with the program.

The program may be for realizing some of the functions to be performed by computer 1100. For example, the program may be for performing functions in combination with other programs already stored in storage 1130 or in combination with other programs implemented in other devices. In other embodiments, computer 1100 may be equipped with a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or instead of the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions implemented by the processor 1110 may be implemented by the integrated circuit.

Examples of storage 1130 include a magnetic disk, a magneto-optical disk, and semiconductor memory. Storage 1130 may be internal media directly connected to the bus of computer 1100, or external media connected to the computer via interface 1140 or a communication line. Furthermore, if this program is distributed to computer 1100 via a communication line, computer 1100 that receives the program may deploy the program in main memory 1120 and execute the above-mentioned processing. In at least one embodiment, storage 1130 is a non-transitory, tangible storage medium.

The program may also be one that realizes some of the functions described above. Furthermore, the program may be a so-called differential file (differential program) that realizes the functions described above in combination with other programs already stored in storage 1130.

<Additional Notes>
The sign determination device 100 described in each embodiment can be understood, for example, as follows.

(1) The symptom determination device 100 according to the present disclosure includes a measurement result acquisition unit 101 that acquires measurement results of the current flowing through the electric motor 11, an analysis unit 102 that performs frequency analysis on the measurement results and resolves them into frequency components, and a prediction unit 105 that determines whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11 based on time-series data of the frequency components.

The symptom assessment device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11 based on the frequency components of the current in the electric motor 11. Furthermore, the symptom assessment device 100 can determine whether there are signs of an abnormality from a long distance, even if it is not installed near the electric motor 11 and the load 13, which are affected by vibrations, temperature changes, etc.

(2) Furthermore, the analysis unit 102 of the symptom determination device 100 decomposes the signal into multiple frequency components using FFT, and the prediction units (105, 112) determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on the time series data of each of the multiple frequency components.

The symptom determination device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on multiple frequency components decomposed by FFT.

(3) Furthermore, the signs of abnormality in the electric motor 11 determined by the prediction unit (105, 112) of the sign determination device 100 include at least one of breakage of a rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

The symptom assessment device 100 can determine whether there are any signs of abnormality, including at least one of the following: a broken rotor bar in the electric motor 11, a change in tension in the belt connecting the electric motor 11 and the load 13, and cavitation.

(4) Furthermore, the prediction unit (112) of the sign determination device 100 includes a trained model whose parameters are determined using training data, in which time series data of frequency components is used as input data and the determination result of whether or not there is a sign of an abnormality in the input data is used as output data.

The symptom determination device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 using a trained model whose parameters are determined using training data.

(5) The sign determination device 100 also includes an image generation unit 103 that generates an image showing time-series data of frequency components, and the prediction unit 105 determines whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on the image.

The symptom assessment device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on an image showing the frequency components of the electric motor 11's current. The symptom assessment device 100 can determine whether there are signs of an abnormality from a long distance, even if it is not installed near the electric motor 11 and the load 13, which are affected by vibrations, temperature changes, etc.

(6) Furthermore, the analysis unit 102 of the sign determination device 100 decomposes the signal into multiple frequency components using FFT, and the image represents the time series data for each of the multiple frequency components.

The symptom determination device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on an image showing multiple frequency components decomposed by FFT.

(7) The image of the sign assessment device 100 further shows time-series data of energy values associated with frequency components, and the device is equipped with a detection unit 104 that detects change values of the frequency components based on the image, and a prediction unit 105 that compares the change value with a preset first threshold value to determine whether or not a sign is present.

The symptom assessment device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on an image showing frequency components and energy values.

(8) Furthermore, the image of the sign determination device 100 shows frequency components as a color graph with time and frequency as axes, and energy values are indicated by color on the color graph.

The symptom determination device 100 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 based on a color graph image showing frequency components and energy values.

(9) Furthermore, the sign assessment device 100 includes an identification unit 108 that, when it is determined that a sign exists, identifies the part by comparing the frequency component with part information that associates the frequency component with the part where the abnormality occurs.

The symptom assessment device 100 identifies the location where an abnormality will occur based on the location information. This allows the user of the symptom assessment device 100 to identify the location where an abnormality will occur in at least one of the electric motor 11 and the load 13.

(10) Furthermore, if it is determined that a sign exists, the identification unit 108 of the sign assessment device 100 further identifies the cause based on the cause information that associates the image with the cause of the abnormality.

The symptom assessment device 100 identifies the cause of the abnormality based on the cause information. This allows the user of the symptom assessment device 100 to identify the cause of the abnormality in the bearing 15.

(11) The sign assessment device 100 also includes a calculation unit 106 that calculates a similarity based on the image feature associated with the factor and the feature of the image generated by the image generation unit 103, and a second assessment unit 107 that, when it is determined that a sign exists, determines whether the similarity is equal to or greater than a predetermined second threshold, and the identification unit 108 identifies the factor related to the calculated similarity when it is determined that the similarity is equal to or greater than the second threshold.

The sign assessment device 100 calculates similarity based on the image features and identifies the cause of the abnormality. This allows the user of the sign assessment device 100 to identify the cause of the abnormality in at least one of the electric motor 11 and the load 13.

(12) The sign determination device 100 also includes an update unit 109 that receives input from outside and updates the cause information when it is determined that the similarity is not equal to or greater than the second threshold.

If the sign assessment device 100 is unable to identify a cause, it accepts input from outside and updates the cause information. This allows the sign assessment device 100 to identify more causes.

(13) Furthermore, the identification unit 108 of the sign determination device 100 compares the identified cause with time information that associates the cause with the time until the abnormality occurs, and further identifies the time.

The symptom assessment device 100 determines the time until an abnormality occurs based on the time information. This allows the user of the symptom assessment device 100 to determine the time until an abnormality occurs in at least one of the electric motor 11 and the load 13.

(14) The sign determination device 100 also includes an output unit 110 that outputs the content identified by the identification unit 108.

The sign assessment device 100 outputs the cause, area, time, and other details identified by the identification unit 108. This allows the user of the sign assessment device 100 to confirm the cause, area, time, and other details identified by the identification unit 108.

(15) The symptom assessment system 1 is also an abnormality symptom assessment system comprising a symptom assessment device 100 capable of communicating with a terminal device 200. The terminal device 200 includes a request unit 2001 that requests an abnormality symptom assessment for the electric motor 11 and the load 13 of the electric motor 11. The symptom assessment device 100 includes a measurement result acquisition unit 101 that acquires measurement results of the current flowing through the electric motor 11 in response to a request from the terminal device 200, an analysis unit 102 that performs frequency analysis on the measurement results and breaks them down into frequency components, and a prediction unit 105 that determines whether there is a symptom of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11 based on time-series data of the frequency components.

The symptom assessment system 1 can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11 based on the frequency components of the current in the electric motor 11. Furthermore, the symptom assessment device 100 can determine whether there are signs of an abnormality from a long distance, even if it is not installed near the electric motor 11 and the load 13, which are affected by vibrations, temperature changes, etc.

(16) The sign determination method disclosed herein includes obtaining measurement results of the current flowing through the electric motor 11, performing frequency analysis on the measurement results to resolve them into frequency components, and determining whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11 based on time-series data of the frequency components.

By using the sign determination method, a user can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11, based on the frequency components of the current in the electric motor 11. Furthermore, by using the sign determination method, a user can determine whether there are signs of an abnormality from a distance, even if the device is not installed near the electric motor 11 and the load 13, which are affected by vibrations, temperature changes, etc.

(17) The program disclosed herein causes a computer to acquire measurement results of the current flowing through the electric motor 11, perform frequency analysis on the measurement results to resolve them into frequency components, and determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11 based on the time-series data of the frequency components.

By having a computer execute the program, a user of the program can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11, based on the frequency component of the current in the electric motor 11. Furthermore, by having a computer execute the program, a user of the program can determine whether there are signs of an abnormality from a distance, even if the computer is not located near the electric motor 11 and the load 13, which are affected by vibrations, temperature changes, etc.

(18) The sign assessment method disclosed herein includes a request step in which the terminal device 200 requests an abnormality sign assessment for the electric motor 11 and the load of the electric motor 11; a measurement result acquisition step in which the sign assessment device 100, which is capable of communicating with the terminal device 200, acquires measurement results of the current flowing through the electric motor 11 in response to the request from the terminal device 200; an analysis step in which the measurement results are frequency analyzed and broken down into frequency components; and a prediction step in which, based on time-series data of the frequency components, it is determined whether or not there is an abnormality sign in at least one of the electric motor 11 and the load 13 of the electric motor 11.

By using the sign determination method, a user can determine whether there are signs of an abnormality in at least one of the electric motor 11 and the load 13 of the electric motor 11, based on the frequency components of the current in the electric motor 11. Furthermore, by using the sign determination method, a user can determine whether there are signs of an abnormality from a distance, even if the device is not installed near the electric motor 11 and the load 13, which are affected by vibrations, temperature changes, etc.

REFERENCE SIGNS 1: Sign determination system 10: Power source 11: Electric motor 12: Electric wire 13: Load 14: Shaft 15: Bearing 16: Measuring instrument 17: Converter 100: Sign determination device 101: Measurement result acquisition unit 102: Analysis unit 103: Image generation unit 104: Detection unit 105: First determination unit 106: Calculation unit 107: Second determination unit 108: Identification unit 109: Update unit 110: Output unit 111: Storage unit 112: Processing unit 200: Terminal device 1100: Computer 1110: Processor 1120: Main memory 1130: Storage 1140: Interface 2001: Request unit

Claims (7)

a measurement result acquisition unit that acquires a measurement result of a current flowing through the electric motor;
an analysis unit that performs frequency analysis on the measurement result and resolves it into frequency components;
a prediction unit that determines whether or not there is a sign of an abnormality in at least one of the electric motor and a load of the electric motor based on the time series data of the frequency components;
Equipped with
an identification unit that, when it is determined that the sign exists, compares the frequency component with part information in which the frequency component is associated with a part in which the abnormality occurs to identify the part , and further identifies the cause of the abnormality based on an image showing time-series data of the frequency component and cause information in which the cause of the abnormality is associated with the cause ;
An indication determination device comprising: a calculation unit that calculates a similarity based on the feature amount of the image associated with the factor and the feature amount of the image;
a second determination unit that, when it is determined that the sign exists, determines whether the similarity is equal to or greater than a second threshold value that is set in advance;
the specifying unit specifies the factor related to the calculated similarity when it is determined that the similarity is equal to or greater than the second threshold.
The sign determination device according to claim 1 . an update unit that receives an input from an external device and updates the factor information when it is determined that the similarity is not equal to or greater than the second threshold;
The sign determination device according to claim 2 , further comprising: the identifying unit compares the identified cause with time information that associates the cause with the time until the abnormality occurs, and further identifies the time.
The sign determination device according to claim 1 . an output unit that outputs the content identified by the identification unit;
The sign determination device according to claim 1 , comprising: obtaining a measurement of the current through the electric motor;
performing frequency analysis on the measurement results to resolve them into frequency components;
determining whether or not there is a sign of an abnormality in at least one of the electric motor and a load of the electric motor based on the time series data of the frequency components;
Including,
If it is determined that the sign exists, comparing the frequency component with part information in which the frequency component is associated with a part where the abnormality occurs, to identify the part, and further identifying the cause of the abnormality based on an image showing time-series data of the frequency component and cause information in which the cause of the abnormality is associated with the image ;
A processing method comprising: On the computer,
obtaining a measurement of the current through the electric motor;
performing frequency analysis on the measurement results to resolve them into frequency components;
determining whether or not there is a sign of an abnormality in at least one of the electric motor and a load of the electric motor based on the time series data of the frequency components;
Execute
If it is determined that the sign exists, comparing the frequency component with part information in which the frequency component is associated with a part where the abnormality occurs, to identify the part, and further identifying the cause of the abnormality based on an image showing time-series data of the frequency component and cause information in which the cause of the abnormality is associated with the image ;
A program that executes the following.