Diagnosis device and diagnosis system

Abstract

To accurately diagnose an abnormality of a diagnosis target.SOLUTION: A diagnosis device comprises: a data acquisition section that acquires vibration data representing vibration of a diagnosis target including an electric motor; a heat map generation section that performs frequency analysis on the vibration data and generates a heat map image in which a ratio of an amplitude intensity is a matrix for each combination of a rotation frequency of the electric motor and an amplitude intensity of a harmonic frequency from a result of the frequency analysis; a storage section that stores the heat map image generated from the vibration data when the diagnosis target is normal; and a diagnosis section that diagnoses presence or absence of an abnormality of the diagnosis target on the basis of the heat map image generated by the heat map generation section from the vibration data acquired by the data acquisition section and the heat map image stored in the storage section.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a diagnostic device and a diagnostic system. [Background technology]

[0002] Patent Documents 1 and 2 disclose inventions for diagnosing equipment abnormalities. The method disclosed in Patent Document 1 performs frequency analysis on a motor's current waveform to determine a spectrum level in a frequency range limited by filtering, and determines whether an abnormality exists based on the results of comparing the determined spectrum level with a judgment criterion. The system disclosed in Patent Document 2 calculates, as a first index, a correlation coefficient indicating the similarity between a first model constructed based on normal data indicating the frequency spectrum of vibration when the target device to be diagnosed is normal and diagnostic data indicating the frequency spectrum of the device to be diagnosed. It also calculates, as a second index, a Q statistic indicating the degree of abnormality of the target device based on the normal data and abnormal data indicating the frequency spectrum of vibration when the target device is abnormal, and the diagnostic data. It then determines whether the target device is abnormal based on the first index and the second index. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-90546 [Patent Document 2] Japanese Patent Publication No. 2022-14506 Summary of the Invention [Problem to be solved by the invention]

[0004] In the method of calculating frequency spectrum levels disclosed in Patent Document 1, if the frequency range used for judgment changes due to motor performance degradation or inverter drive, the limited frequency range must be changed, making diagnosis cumbersome. In the system disclosed in Patent Document 2, the correlation coefficient is calculated including frequency bands unrelated to abnormalities, so the correlation coefficient may be affected by noise, which may affect diagnosis. In addition, the system disclosed in Patent Document 2 also uses statistics, but only the overall characteristics of the frequency spectrum are known, so there is a risk of overlooking abnormalities.

[0005] The present invention has been made in view of the above, and has an object to accurately diagnose abnormalities in a diagnostic object. [Means for solving the problem]

[0006] A diagnostic device according to one aspect of the present invention includes a data acquisition unit that acquires vibration data representing vibrations of a diagnostic object including an electric motor; a heat map generation unit that performs frequency analysis on the vibration data and generates a heat map image in the form of a matrix of amplitude intensity ratios for each combination of amplitude intensity of the rotational frequency and harmonic frequencies of the electric motor from the results of the frequency analysis; a memory unit that stores the heat map image generated from the vibration data when the diagnostic object is normal; and a diagnostic unit that diagnoses the presence or absence of an abnormality in the diagnostic object based on the heat map image generated by the heat map generation unit from the vibration data acquired by the data acquisition unit and the heat map image stored in the memory unit.

[0007] In addition, in the diagnostic device according to the present invention, the memory unit may store the heat map image generated from the vibration data when the diagnostic object is abnormal for each type of abnormality, and the diagnostic unit may diagnose the type of abnormality of the diagnostic object based on the heat map image generated from the vibration data when the diagnostic object is abnormal.

[0008] In addition, in the diagnostic device according to the present invention, the diagnostic unit may diagnose the presence or absence of an abnormality in the object to be diagnosed based on the similarity between the heat map image generated by the heat map generation unit and the heat map image stored in the memory unit.

[0009] A diagnostic system according to one aspect of the present invention comprises a diagnostic device having a diagnostic object including an electric motor, a sensor that detects vibrations in the diagnostic object and outputs vibration data representing the detected vibrations, a data acquisition unit that acquires the vibration data output by the sensor, a heat map generation unit that performs frequency analysis on the vibration data and generates a heat map image in which the ratios of amplitude intensities are expressed as a matrix for each combination of amplitude intensities of the rotational frequency and harmonic frequencies of the electric motor from the results of the frequency analysis, a memory unit that stores the heat map image generated from the vibration data when the diagnostic object is normal, and a diagnostic unit that diagnoses the presence or absence of an abnormality in the diagnostic object based on the heat map image generated by the heat map generation unit from the vibration data acquired by the data acquisition unit and the heat map image stored in the memory unit. [Effects of the Invention]

[0010] The present invention provides an effect of enabling an abnormality in a diagnostic object to be diagnosed with high accuracy. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 is a diagram showing the configuration of a diagnostic system according to an embodiment. [Figure 2] FIG. 2 is a block diagram showing the configuration of the diagnostic device according to the embodiment. [Figure 3] FIG. 3 is a functional block diagram showing functions realized by the diagnostic device. [Figure 4] FIG. 4 is a flowchart showing the flow of processing executed by the diagnostic device. [Figure 5] FIG. 5 is a diagram illustrating an example of an amplitude spectrum. [Figure 6] FIG. 6 is a diagram illustrating an example of a heat map image. [Figure 7] FIG. 7 is a flowchart showing the flow of processing executed by the diagnostic device. [Figure 8] FIG. 8 is a functional block diagram showing functions realized by a diagnostic device according to a modified example. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments described below. In addition, in the description of the drawings, the same or corresponding elements are appropriately designated by the same reference numerals.

[0013] FIG. 1 is a diagram showing the configuration of a diagnostic system 1 according to an embodiment of the present invention. The diagnostic system 1 is a system that diagnoses whether or not an abnormality has occurred in a diagnostic object 2. The diagnostic object 2 includes an electric motor 30 having a rotating shaft 30a, a first bearing 41, a second bearing 42, and a load 50. The electric motor 30 rotates the rotating shaft 30a using supplied power, thereby driving a load 50 connected to the rotating shaft 30a. The first bearing 41 and the second bearing 42 are bearings that support the rotating shaft 30a. The first bearing 41 is installed on the electric motor 30 side of the rotating shaft 30a, and the second bearing 42 is installed on the load 50 side of the rotating shaft 30a. The load 50 is, for example, a pump, and is driven by the rotating shaft 30a. Note that the load 50 is not limited to a pump and may be, for example, a blower, a compressor, or a conveying device.

[0014] The first sensor 31 and the second sensor 32 are sensors that detect vibration. The first sensor 31 is installed on the first bearing 41, and the second sensor 32 is installed on the second bearing 42. The first sensor 31 and the second sensor 32 detect, for example, acceleration as a physical quantity that represents vibration. The first sensor 31 and the second sensor 32 record the acceleration detected over a predetermined time period as vibration data and transmit the recorded vibration data to the data collecting device 20, for example, via wireless communication. This predetermined time period is, for example, one hour, but is not limited to one hour and may be less than one hour or more than one hour. The first sensor 31 and the second sensor 32 may also transmit the vibration data to the data collecting device 20 via a wired connection.

[0015] The communication network NW is a communication network configured from the Internet network, a mobile phone network, etc. The data collection device 20 is a device that collects vibration data. The data collection device 20 transmits the vibration data transmitted from the first sensor 31 and the second sensor 32 to the diagnostic device 10 via the communication network NW. Note that the first sensor 31 and the second sensor 32 may transmit the vibration data to the diagnostic device 10 via the communication network NW without going through the data collection device 20.

[0016] The diagnostic device 10 is a device that provides a cloud service. The diagnostic device 10 provides a service of diagnosing a diagnostic object 2 using data transmitted from a data collection device 20 via a communication network NW. The diagnostic device 10 may be on-premise. The terminal device 11 is, for example, a thin client, and is a device for operating the diagnostic device 10.

[0017] FIG. 2 is a block diagram showing the configuration of the diagnostic device 10. The diagnostic device 10 includes a processor 101, a memory 102, a storage 103, and a communication I / F 104, all connected via a bus 105. The memory 102 is, for example, a random access memory (RAM), and is configured as a volatile memory or a nonvolatile memory. The memory 102 serves as a workspace for the processor 101 to perform arithmetic processing, and stores the results of the arithmetic processing performed by the processor 101. The storage 103, which is an example of a storage unit, is configured with a read-only memory (ROM) and an auxiliary storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The ROM of the storage 103 stores programs used by the processor 101 to perform arithmetic processing. The auxiliary storage device of the storage 103 stores data used by the processor 101 to perform arithmetic processing. The communication I / F 104 includes a communication module that communicates information via wired or wireless communication. The communication I / F 104 communicates with the data collection device 20 and the terminal device 11 via the communication network NW.

[0018] The processor 101 is, for example, a CPU (Central Processing Unit), which reads a program from the storage 103 and executes it using the memory 102 as a work space. The processor 101 executes the program, thereby realizing the functions of the diagnostic device 10.

[0019] 3 is a functional block diagram showing functions realized in the diagnostic device 10 by the processor 101 executing a program. The data acquisition unit 101a acquires vibration data transmitted by the data collection device 20. The heat map generation unit 101b generates a heat map image in the form of a matrix of amplitude intensity ratios for each combination of amplitude intensity of the rotational frequency of the electric motor 30 and the harmonic frequencies obtained by performing a fast Fourier transform on the vibration data acquired by the data acquisition unit 101a. The diagnostic unit 101c diagnoses the presence or absence of an abnormality in the diagnostic object 2 based on the heat map image generated by the heat map generation unit 101b from the vibration data acquired by the data acquisition unit 101a and the heat map image generated from vibration data when the diagnostic object 2 is normal, which is stored in the storage 103.

[0020] Next, an example of the operation of the embodiment will be described. FIG. 4 is a flowchart showing the flow of a process for generating a reference heat map image used for diagnosing the diagnosis object 2. When the diagnosis device 10 receives vibration data transmitted from the data collection device 20, it stores the received vibration data in the storage 103. The operator operates the terminal device 11 to select vibration data from the first sensor 31 and vibration data from the second sensor 32 when the diagnosis object 2 is operating normally, and instructs the diagnosis device 10 to generate a reference heat map image. Upon receiving this instruction, the diagnosis device 10 performs a fast Fourier transform on each of the vibration data obtained by the first sensor 31 and the vibration data obtained by the second sensor 32 (step S101). By performing a fast Fourier transform on the vibration data, an amplitude spectrum of the first bearing 41 and an amplitude spectrum of the second bearing 42 are obtained.

[0021] Next, the diagnostic device 10 obtains, from the amplitude spectrum obtained in step S101, the amplitude intensity at the rotation frequency of the electric motor 30 and the amplitude intensity at harmonic frequencies of the rotation frequency (step S102). FIG. 5 is a diagram showing an example of the amplitude spectrum obtained in step S101. In FIG. 5, 1X is the amplitude intensity at the rotation frequency, and 2X to 6X are the amplitude intensities at frequencies two to six times the rotation frequency. The rotation frequency of the electric motor 30 is stored in advance in the diagnostic device 10. When obtaining the amplitude intensity of the rotation frequency from the amplitude spectrum, the diagnostic device 10 obtains the peak value within a predetermined band including the rotation frequency as the amplitude intensity, and also obtains the peak value within a predetermined band including the harmonic frequency for each harmonic frequency as the amplitude intensity. For example, this predetermined band is a frequency band from 50% to 100% of the rotation frequency for the rotation frequency, and a frequency band from 50% to 100% of the harmonic frequency for the harmonic frequencies. In addition to the amplitude intensity of frequencies that are n times (n is an integer) the rotation frequency, the diagnostic device 10 may also acquire the amplitude intensity of frequencies between the rotation frequency and the harmonic frequency, such as 0.5 times, 1.5 times, 2.5 times, 3.5 times, 4.5 times, and 5.5 times the rotation frequency.

[0022] Next, the diagnostic device 10 creates a matrix of amplitude intensity ratios for pairs of amplitude intensity at the rotational frequency and amplitude intensity at the harmonic frequency, and generates heat map images for each of the first bearing 41 and the second bearing 42 based on the magnitude of the ratio (step S103). FIG. 6 is a diagram showing an example of a heat map image. In the matrix of the heat map image, the amplitude intensity of the columns is the numerator and the amplitude intensity of the rows is the denominator. For example, the ratio in the first row and second column is (2X amplitude intensity / 1X amplitude intensity), and the ratio in the first row and third column is (3X amplitude intensity / 1X amplitude intensity). For example, the diagnostic device 10 generates a heat map image by representing this ratio as black when it is 0 and white when it is 5 or greater, with 256 grayscale levels from 0 to 5. Note that the range of the ratio when creating the grayscale is not limited to 0 to 5 and may be other ranges. Furthermore, if the diagnostic device 10 acquires amplitude intensities at frequencies 0.5, 1.5, 2.5, 3.5, 4.5, and 5.5 times the rotation frequency in addition to the amplitude intensities at frequencies n times the rotation frequency, the ratios of the amplitude intensities at these frequencies may also be added to the heat map image.

[0023] Next, the diagnostic device 10 stores the generated heat map image as a reference heat map image when the diagnostic object 2 is normal (step S104). Note that the diagnostic device 10 may generate a heat map image for each of a plurality of vibration data from the plurality of vibration data when the diagnostic object 2 is normal, and may perform known machine learning on the generated plurality of heat map images to average the generated heat map image and store the averaged heat map image as the reference heat map image when the diagnostic object 2 is normal.

[0024] Next, an example of operation when diagnosing the diagnosis object 2 will be described. Fig. 7 is a flowchart showing the flow of processing when the diagnosis device 10 diagnoses the diagnosis object 2. When the operator operates the terminal device 11 to instruct the diagnosis device 10 to diagnose the diagnosis object 2, the diagnosis device 10 starts the processing shown in Fig. 7. Upon receiving this instruction, the diagnosis device 10 receives the vibration data transmitted from the data collection device 20, and then performs a fast Fourier transform on each of the vibration data obtained by the first sensor 31 and the vibration data obtained by the second sensor 32 (step S201). By performing a fast Fourier transform on the vibration data, the amplitude spectrum of the first bearing 41 and the amplitude spectrum of the second bearing 42 are obtained.

[0025] Next, the diagnostic device 10 acquires the amplitude intensity at the rotational frequency of the electric motor 30 and the amplitude intensity at the harmonic frequencies of the rotational frequency from the amplitude spectrum obtained in step S201 (step S202). After completing the process of step S202, the diagnostic device 10 creates a matrix of amplitude intensity ratios for pairs of amplitude intensity at the rotational frequency and amplitude intensity at the harmonic frequencies, and generates heat map images of each of the first bearing 41 and the second bearing 42 based on the magnitude of the ratio values ​​(step S203).

[0026] Next, the diagnostic device 10 diagnoses the diagnostic object 2 using the heat map image stored in step S104 and the heat map image generated in step S203 (step S204). Specifically, the diagnostic device 10 calculates the similarity between the reference heat map image and the heat map image generated in step S203. The similarity calculated here is an example of an abnormality score. This similarity is, for example, the sum of squared differences for each pixel between the reference heat map image and the heat map image generated in step S203. Note that the similarity is not limited to the sum of squared differences, and may be EMD (Earth Mover's Distance) or cosine similarity.

[0027] The diagnostic device 10 diagnoses the diagnostic object 2 based on, for example, the calculated similarity and a predetermined threshold. For example, when the diagnostic device 10 calculates the similarity using a sum of squared differences, if the similarity is equal to or greater than a predetermined threshold, the diagnostic device 10 diagnoses that an abnormality has occurred in the diagnostic object 2, and if the similarity is less than the predetermined threshold, the diagnostic device 10 diagnoses that an abnormality has occurred in the diagnostic object 2. For example, when the diagnostic device 10 calculates the similarity using cosine similarity ... the diagnostic device 10 diagnoses that an abnormality has occurred in the diagnostic object 2. For example, when the diagnostic device 10 diagnoses that an abnormality has occurred in the heat map image obtained from the vibration data of the first sensor 31, the diagnostic device 10 diagnoses that an abnormality has occurred in the first bearing 41, and if the diagnostic device 10 diagnoses that an abnormality has occurred in the heat map image obtained from the vibration data of the second sensor 32, the diagnostic device 10 diagnoses that an abnormality has occurred in the second bearing 42.

[0028] Next, the diagnostic device 10 outputs the diagnostic result to the terminal device 11 (step S205). The operator can know the presence or absence of an abnormality in the diagnostic object 2 from the diagnostic result displayed on the terminal device 11. The diagnostic device 10 may also output a heat map image used for the diagnosis to the terminal device 11.

[0029] In this embodiment, the heat map image represents the relationship between the rotational frequency and the harmonic frequency. For example, misalignment of the rotating shaft 30a with respect to the center of rotation of the load 50, an abnormality in the first bearing 41, or an abnormality in the second bearing 42 is reflected in the heat map image, allowing for diagnosis of abnormalities from the heat map image. For example, the diagnostic device 10 stores a heat map image for misalignment and a heat map image for a bearing abnormality. When the diagnostic device 10 diagnoses an abnormality in the diagnostic object 2, it may calculate the similarity between the stored heat map image for misalignment and the heat map image generated in step S203, and diagnose the abnormality as misalignment if the calculated similarity satisfies a predetermined threshold. Alternatively, it may calculate the similarity between the stored heat map image for a bearing abnormality and the heat map image generated in step S203, and diagnose the abnormality in the bearing if the calculated similarity satisfies a predetermined threshold.

[0030] In this embodiment, when an abnormality occurs in the diagnosis object 2, the heat map image has a different shade than when the diagnosis object 2 is normal, making the abnormality visible and allowing the operator to easily recognize the occurrence of the abnormality. In addition, in this embodiment, peak values ​​within a predetermined band including the rotation frequency and a predetermined band including the harmonic frequency are acquired as amplitude intensity, so that diagnosis of the diagnosis object 2 can be performed without being affected even if the rotation frequency of the electric motor 30 deviates.

[0031] Furthermore, if there is an abnormality in the diagnostic object 2, the heat map image will represent the characteristics of the abnormality in the diagnostic object 2, and even if the model number of the electric motor 30 is different, if the type of abnormality is the same, the heat map image will be a similar image, so it can be used as information to identify the type of abnormality, allowing the type of abnormality to be accurately identified while reducing the cost of collecting data.

[0032] [Variations] Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments and can be implemented in various other forms. For example, the above-described embodiments may be modified as follows to implement the present invention. The above-described embodiments and the following modifications may be combined with each other. The present invention also includes configurations in which the components of the above-described embodiments and modifications are appropriately combined. Furthermore, further effects and modifications can be easily derived by those skilled in the art. Therefore, the broader aspects of the present invention are not limited to the above-described embodiments and modifications, and various modifications are possible.

[0033] In the above-described embodiment, the presence or absence of an abnormality in the diagnostic object 2 is determined based on the similarity, but the scores used to determine the presence or absence of an abnormality are not limited to the scores in the embodiment. For example, the diagnostic device 10 according to the modified example may calculate abnormality scores using the first to third methods in addition to the scores in the embodiment, and determine the presence or absence of an abnormality using the calculated scores.

[0034] The first method stores the results of a fast Fourier transform of vibration data when the diagnosis target 2 is normal, and determines the sum of absolute differences between the stored transform results and the results of a fast Fourier transform of vibration data newly received from the data collecting device 20 as the abnormality score. The second method, for example, calculates the difference for each matrix element between a reference heat map image and a heat map image generated from newly received vibration data, and determines the sum of the calculated differences as the abnormality score. The third method stores the results of a fast Fourier transform of vibration data when the diagnosis target 2 is normal, calculates the Mahalanobis distance between the stored transform results and the results of a fast Fourier transform of vibration data newly received from the data collecting device 20, and determines the calculated result as the abnormality score. Note that when the rotation speed of the electric motor 30 is controlled by an inverter in multiple patterns, the diagnosis device 10 may calculate an abnormality score for each pattern using each of the first to third methods in addition to the abnormality score of the embodiment, and determine the minimum of the abnormality scores calculated for each pattern as the abnormality score for each method.

[0035] Next, the diagnostic device 10 normalizes the calculated abnormality scores. Specifically, for multiple vibration data when the diagnostic object 2 is normal, the diagnostic device 10 calculates an abnormality score for each of the method of the embodiment and the first to third methods, and sets the average of the calculated abnormality scores as the reference value used for normalization. Next, the diagnostic device 10 normalizes the abnormality scores calculated for each of the method of the embodiment and the first to third methods from the newly transmitted vibration data by dividing them by the reference value. The diagnostic device 10 sets the maximum of the normalized abnormality scores as the final abnormality score, and compares the abnormality score with a threshold value to determine whether or not an abnormality exists.

[0036] In the present invention, the data acquisition unit 101a may process the acquired vibration data, and the heat map generation unit 101b may generate a heat map image using the processed data. FIG. 8 is a functional block diagram showing functions implemented by the diagnostic device 10 according to the modified example. The diagnostic device 10 according to the modified example includes a data processing unit 101d that processes the vibration data. For example, the data processing unit 101d first deletes data such as missing values ​​in the vibration data and outliers immediately after the start of recording the vibration data. Next, the data processing unit 101d performs unsupervised clustering processing, for example, by the k-means method, on the vibration data from which the missing values ​​and outliers have been deleted so that the number of clusters reaches a predetermined number (for example, 10), and removes vibration data when the electric motor 30 is stopped.

[0037] Next, the data processing unit 101d performs windowing and fast Fourier transform on the clustered vibration data to obtain an amplitude spectrum for each of the vibration data. The data processing unit 101d also removes, for example, a portion of a low-frequency band below the rotational frequency of the electric motor 30 from each of the obtained amplitude spectra. The heat map generation unit 101b generates a reference heat map image from the amplitude spectrum obtained when the diagnostic object 2 is operating normally, from the plurality of amplitude spectra from which the low-frequency band has been removed. [Explanation of symbols]

[0038] 1 Diagnostic System 2. Diagnostic Targets 10 Diagnostic equipment 11 Terminal equipment 20 Data Collection Equipment 30 Electric motor 30a Rotating shaft 31 First sensor 32 Second sensor 41 First bearing 42 Second bearing 101a Data acquisition section 101b Heat map generation section 101c Diagnostic Department

Claims

1. a data acquisition unit that acquires vibration data representing vibrations of an object to be diagnosed, the vibration data representing vibrations of an object to be diagnosed, the electric motor; a heat map generation unit that performs frequency analysis on the vibration data and generates a heat map image based on the result of the frequency analysis, in which ratios of amplitude intensities are expressed as a matrix for each combination of amplitude intensities of the rotational frequency of the motor and harmonic frequencies; a storage unit that stores the heat map image generated from the vibration data when the diagnostic object is normal; a diagnosis unit that diagnoses whether or not there is an abnormality in the diagnosis target based on a heat map image generated by the heat map generation unit from the vibration data acquired by the data acquisition unit and a heat map image stored in the storage unit; A diagnostic device comprising:

2. the storage unit stores the heat map image generated from the vibration data when the diagnostic target is abnormal for each type of abnormality; The diagnosis unit diagnoses the type of abnormality of the diagnostic object based on the heat map image generated from the vibration data when the diagnostic object is abnormal. The diagnostic device of claim 1 .

3. The diagnosis unit diagnoses whether or not there is an abnormality in the diagnosis target based on a similarity between the heat map image generated by the heat map generation unit and the heat map image stored in the storage unit. The diagnostic device of claim 1 .

4. a diagnostic target including an electric motor; a sensor that detects vibrations of the diagnosis target and outputs vibration data representing the detected vibrations; a data acquisition unit that acquires vibration data output by the sensor; a heat map generation unit that performs frequency analysis on the vibration data and generates a heat map image based on the result of the frequency analysis, in which ratios of amplitude intensities are expressed as a matrix for each combination of amplitude intensities of the rotational frequency of the motor and harmonic frequencies; a storage unit that stores the heat map image generated from the vibration data when the diagnostic object is normal; a diagnosis unit that diagnoses whether or not there is an abnormality in the diagnosis target based on a heat map image generated by the heat map generation unit from the vibration data acquired by the data acquisition unit and a heat map image stored in the storage unit; a diagnostic device having A diagnostic system comprising:

Images