State monitoring apparatus and state monitoring method

Abstract

This state monitoring apparatus (100) comprises: a storage device (150) that stores vibration data obtained from a sensor (20) that detects a physical quantity caused by vibration; and an arithmetic device (160) that receives the vibration data from the storage device (150) and performs abnormality determination. The arithmetic device (160) is configured to execute: a process for obtaining an envelope spectrum of a diagnosis target from vibration data of a bearing (12) to be diagnosed; a process for extracting diagnosis target data having an amplitude exceeding a first threshold value in the envelope spectrum of the diagnosis target; a process for calculating a comparison value of the diagnosis target from the amplitude of the extracted diagnosis target data; and a process for setting a value obtained by dividing the comparison value of the diagnosis target by a reference comparison value as a ratio of the comparison value, and determining that there is an abnormality if the ratio of the comparison value is equal to or greater than a preset second threshold value.

Description

Condition Monitoring Device and Condition Monitoring Method

[0001] The present disclosure relates to a condition monitoring device and a condition monitoring method.

[0002] Bearings are used in devices having rotating shafts such as machine tools and generators, and it is desirable to detect bearing abnormalities at an early stage from vibrations or the like.

[0003] For example, Japanese Unexamined Patent Application Publication No. 2023-106207 discloses a method for determining whether vibrations are caused by bearing abnormalities when the vibrations of a machine tool spindle exceed the normal range, and specifying the cause of the abnormalities when it is determined that the bearing is abnormal.

[0004] Also, for example, Japanese Patent No. 7146683 discloses a method for detecting bearing abnormalities without using bearing specifications by emphasizing vibration components proportional to the rotational frequency in the envelope spectrum and specifying the bearing damage site based on the peak position thereof.

[0005] Japanese Unexamined Patent Application Publication No. 2023-106207, Japanese Patent No. 7146683

[0006] Air-oil lubrication is the mainstream for bearings for ultra-high-speed spindles of machine tools, but the demand for grease lubrication is increasing from the viewpoints of reducing lubrication costs and environmental impact. In grease lubrication, it has been found that surface roughening occurs and progresses on the running track of the bearing raceway surface due to oil film breakage caused by grease deterioration. Surface roughening contributes to deterioration of the machining accuracy of the spindle and leads to defective products being produced by the machine tool, so abnormal detection at an early stage is important.

[0007] Compared with peeling, indentation, etc., the surface roughness at the initial stage of damage occurrence of surface roughening is small, and it is difficult to calculate the effective value from the vibration waveform and detect abnormalities using a threshold value.

[0008] For example, in the abnormality detection method described in Japanese Unexamined Patent Application Publication No. 2023-106207, it is difficult to detect surface roughening abnormalities because the damage frequency unique to the bearing is used when determining whether the vibrations are caused by the bearing.

[0009] Furthermore, since surface roughness occurs across the entire surface of the bearing raceway and rolling elements, vibrations caused by surface roughness do not exhibit periodicity. Therefore, methods that focus on the bearing damage frequency and envelope spectrum peaks, similar to those used for delamination and indentation, cannot be used to detect abnormalities.

[0010] For example, the method described in Japanese Patent Publication No. 7146683 does not use the bearing damage frequency, but it is difficult to detect surface roughness abnormalities because abnormalities cannot be diagnosed unless there is regularity in the peaks of the envelope spectrum.

[0011] This disclosure aims to provide a condition monitoring device that solves the above problems and is capable of detecting bearing surface roughness abnormalities at an early stage.

[0012] This disclosure relates to a bearing condition monitoring device for detecting surface roughness. The condition monitoring device includes a storage device that stores vibration data obtained from a sensor that detects physical quantities caused by vibration, and a computing device that receives vibration data from the storage device and performs abnormality determination. The computing device is configured to perform the following processes: obtaining the envelope spectrum of the bearing to be diagnosed from the vibration data of the bearing to be diagnosed; extracting data from the envelope spectrum of the bearing to be diagnosed whose amplitude exceeds a first threshold; calculating a comparison value of the diagnosed data from the amplitude of the extracted data; and determining an abnormality if the ratio of the comparison value of the diagnosed data is obtained by dividing the comparison value of the diagnosed data by a reference comparison value, and determining an abnormality if the ratio of the comparison value is equal to or greater than a preset second threshold.

[0013] The condition monitoring device disclosed herein makes it possible to detect bearing surface roughness abnormalities at an early stage, which were previously difficult to detect early.

[0014] This is a block diagram showing the configuration of the condition monitoring device according to this embodiment. This is a block diagram showing the main parts of the hardware configuration of the arithmetic unit 160. This is a flowchart for explaining the reference value calculation process. This is a diagram showing the relationship between the envelope spectrum and the reference value R. This is a flowchart for explaining the process of calculating reference values ​​for each frequency band. This is a diagram for explaining the reference values ​​calculated for each frequency band. This is a flowchart for explaining the process of calculating reference values ​​from multiple vibration waveforms. This is a flowchart for explaining the comparison value calculation process. This is a diagram for explaining the extraction process in step S33. This is a flowchart for explaining the process of calculating comparison values ​​for each frequency band. This is a diagram for explaining the extraction process in step S43. This is a flowchart for explaining the process of calculating comparison values ​​when there are multiple reference value data. This is a flowchart for explaining the abnormality diagnosis process. This is a flowchart for explaining the abnormality diagnosis process when comparison values ​​are calculated for each frequency band. This is a flowchart for explaining the abnormality diagnosis process when a first threshold and a comparison value are calculated individually for multiple vibration signals. This is a diagram for explaining an example of the relationship between the ratio of the reference value and the comparison value and the second threshold.

[0015] Embodiments of the present invention will be described in detail below with reference to the drawings. In the drawings, identical or corresponding parts are denoted by the same reference numerals, and their descriptions will not be repeated.

[0016] [Basic Configuration of the Condition Monitoring Device] Figure 1 is a block diagram showing the configuration of the condition monitoring device according to this embodiment. Referring to Figure 1, the condition monitoring device 100 receives signals from vibration sensors 20 installed on the device under test 10, monitors the state of the device under test 10, and detects abnormalities. The device under test 10 is equipment including rotating machinery installed in, for example, a factory or power plant. The rotating machinery includes a bearing 12. The bearing 12 is, for example, a rolling bearing. The bearing 12 includes an inner ring 16 fitted onto a rotating shaft 19, an outer ring 14 fixed to the device under test 10, a plurality of rolling elements 18 arranged between the inner ring and the outer ring, and a cage (not shown) that maintains the spacing between the rolling elements 18.

[0017] Sensor 20 can detect abnormal vibrations that occur during rotation. In this embodiment, acceleration is used as an example of the monitored quantity, but sensor 20 may be an acoustic sensor, an ultrasonic sensor, or an AE sensor in addition to an acceleration sensor. These sensors detect physical quantities caused by vibration and output vibration signals that represent the vibration.

[0018] The status monitoring device 100 includes an amplifier 110, a filter 120, an A / D converter 130, a data acquisition unit 140, a storage device 150, an arithmetic unit 160, and a display unit 170.

[0019] The voltage waveform (hereinafter referred to as the vibration voltage waveform) contained in the vibration signal output by the sensor 20 installed on the rotating machine is amplified by the amplifier 110, and the filter 120 performs bandpass filtering (passband 1 kHz to 10 kHz) to remove signals in the frequency band unnecessary for analysis and allow only the necessary frequency band to pass through. The A / D converter 130 receives the output signal from the amplifier 110. The data acquisition unit 140 receives the digital signal from the A / D converter 130 and records the measurement data in the storage device 150. The storage device 150 can be a hard disk, various types of memory, etc. The arithmetic unit 160 reads the measured data from the storage device 150, performs envelope processing and FFT analysis, and after the FFT analysis, calculates the reference value R, comparison value CA, etc., which will be explained later.

[0020] The calculation unit 160 determines whether or not there is an abnormality in the bearing of the device under test 10. If the calculation unit 160 determines whether or not there is an abnormality, it displays the determination result on the display unit 170.

[0021] Figure 2 is a block diagram showing the main components of the hardware configuration of the arithmetic unit 160. Referring to Figure 2, the arithmetic unit 160 includes an input unit 161, an interface (I / F) unit 162, a CPU (Central Processing Unit) 163, a RAM (Random Access Memory) 164, a ROM (Read Only Memory) 165, and an output unit 166.

[0022] The CPU 163 implements a state monitoring method, which will be described in detail later, by executing various programs stored in the ROM 165. The RAM 164 is used as a work area by the CPU 163. The ROM 165 records a program that includes each step of the flowchart (described later) showing the procedure for the state monitoring method. The input unit 161 is a means for reading data from an external source such as a keyboard, mouse, recording medium, or communication device. The output unit 166 is a means for outputting the calculation results of the CPU 163 to a display, recording medium, or communication device.

[0023] The condition monitoring device 100 shown in this embodiment detects surface roughness occurring in the bearing being monitored. The condition monitoring device 100 performs a process consisting of three steps, which are described below: 1) calculation of a reference value, 2) calculation of a comparison value, and 3) abnormality diagnosis.

[0024] 1) Calculation process of reference value First, a vibration waveform is obtained using a bearing in a normal state. The reference value is calculated from the vibration waveform in the normal state.

[0025] Figure 3 is a flowchart illustrating the calculation process for the reference value. This process is mainly performed by the arithmetic unit 160 in the condition monitoring device 100. First, in step S1, the arithmetic unit 160 acquires a vibration waveform from the storage device 150 using a bearing in a normal state. Next, in step S2, the arithmetic unit 160 performs filtering on the acquired vibration waveform in a normal state. For filtering, for example, a high-pass filter that allows waveforms with frequencies of 4 kHz or higher to pass through can be used.

[0026] Next, in steps S3 and S4, the arithmetic unit 160 performs envelope processing on the filtered vibration waveform and then obtains the envelope spectrum using a Fast Fourier Transform (FFT). Hereafter, the magnitude of the peak at each frequency in the envelope spectrum will be expressed as amplitude.

[0027] Next, the condition monitoring device 100 calculates the effective value of the amplitude from the envelope spectrum obtained in step S4, for example, using the calculation formula shown in equation (1) below, and sets it as the reference value R.

[0028]

[0029] Note that the i-th data point represents the i-th peak in the envelope spectrum, and its amplitude is the magnitude of the peak. In step S5, the condition monitoring device 100 calculates a single reference value R for the entire frequency band. Figure 4 shows the relationship between the envelope spectrum and the reference value R.

[0030] Then, in step S6, the condition monitoring device 100 stores reference value data in a database, which is linked to the calculated reference value R, the envelope spectrum, and the method for calculating the reference value R.

[0031] In addition to using the calculation formula in equation (1), the reference value may also be calculated using basic statistics such as the mean and median of the amplitude.

[0032] Furthermore, regarding the calculation of the reference value, in addition to calculating a single reference value R for the entire frequency band of the envelope spectrum as shown in Figures 3 and 4, the maximum frequency of the envelope spectrum may be divided into fixed intervals, and a reference value may be calculated for each frequency band.

[0033] Figure 5 is a flowchart illustrating the process of calculating reference values ​​for each frequency band. In Figure 5, steps S11 to S14 are the same as steps S1 to S4 in Figure 3, so the explanation will not be repeated.

[0034] In step S15, the condition monitoring device 100 divides the envelope spectrum into multiple bands by dividing it at regular frequency intervals. Then, in step S16, the condition monitoring device 100 calculates one reference value for each of the divided frequency bands by the same process as in step S5 of Figure 3.

[0035] Then, in step S17, the condition monitoring device 100 stores reference value data in a database, which is linked to the calculated reference values ​​for each frequency band, the envelope spectrum, and the method for calculating the reference values.

[0036] Figure 6 is a diagram illustrating the reference values ​​calculated for each frequency band. In the example shown in Figure 6, the entire frequency band of the envelope spectrum is divided into five bands: BA, BB, BC, BD, and BE. For each of these bands, reference values ​​RA, RB, RC, RD, and RE are calculated, respectively.

[0037] It should be noted that multiple vibration waveforms representing normal conditions may be obtained. For example, when acquiring vibration waveforms once an hour for a wind turbine, the obtained waveforms will vary slightly depending on the operating conditions at the time of acquisition. In such cases, where multiple vibration waveforms are obtained for the same bearing, it is desirable to be able to accurately determine bearing abnormalities.

[0038] Figure 7 is a flowchart illustrating the process of calculating a reference value from multiple vibration waveforms. Steps S21 to S26 in the flowchart of Figure 7 are the same as steps S1 to S6 in the flowchart of Figure 3, but performed individually for multiple vibration waveforms (No. 1 to No. X), so the explanation will not be repeated here.

[0039] In this way, in step S26 as well, the condition monitoring device 100 stores in the database the envelope spectrum obtained in S24, the reference value calculated in step S25, and the method for calculating the reference value, linked together.

[0040] 2) Calculation process of comparison values ​​Next, we will explain the process of calculating the comparison values ​​used for anomaly detection. Figure 8 is a flowchart illustrating the process of calculating comparison values.

[0041] First, in step S31, the state monitoring device 100 acquires the reference value data stored in the database of the storage device 150. Subsequently, the state monitoring device 100 sets the first threshold value TH1 based on the reference value R included in the reference value data acquired in step S31. For example, a real multiple (e.g., 2 to 5 times) of the reference value R can be set as the first threshold value TH1.

[0042] Next, in step S33, only the data exceeding the set first threshold value TH1 is extracted from the envelope spectrum included in the acquired reference value data.

[0043] FIG. 9 is a diagram for explaining the extraction process in step S33. As shown in FIG. 9, a common first threshold value TH1 is set for the entire region with respect to the amplitude of the envelope spectrum. In the extraction process of step S33, data whose amplitude exceeds the first threshold value TH1 (data indicated by black circles in FIG. 9) is extracted from the envelope spectrum.

[0044] Returning again to FIG. 8, further in step S34, a comparison value CA is calculated. The comparison value CA is a value obtained from the amplitude and is a value for comparison with the data obtained from the bearing to be diagnosed. For the data extracted in step S33, for example, the effective value of the amplitude can be calculated from the calculation formula shown in the following formula (2) to obtain the comparison value CA.

[0045]

[0046] The i-th data is the data indicating the i-th peak among the data extracted from the envelope spectrum, and its amplitude is the magnitude of the peak.

[0047] The comparison value CA can also be expressed as the integrated value of the extracted part of the envelope spectrum. In addition to using formula (2), basic statistics such as the average value and median value of the amplitude for the extracted data may be used for the comparison value CA.

[0048] Next, the condition monitoring device 100 stores comparison value data in a database, which is obtained by linking the first threshold value TH1 set in step S32 with the normal state comparison value CA calculated in step S34 to the reference value data.

[0049] Furthermore, regarding the calculation of the first threshold TH1 and the comparison value CA, in addition to calculating a single reference value for the entire frequency band of the envelope spectrum as shown in Figures 3 and 4, the maximum frequency of the envelope spectrum may be divided into fixed intervals and calculated separately for each frequency band.

[0050] Figure 10 is a flowchart illustrating the process of calculating comparison values ​​for each frequency band. In Figure 10, steps S41 to S45 are the same as steps S31 to S35 in Figure 8, but performed separately for each frequency band.

[0051] In other words, when using reference value data calculated for each frequency band, a first threshold is set for each frequency band (S42), and only data exceeding the first threshold is extracted for each frequency band (S43).

[0052] Figure 11 is a diagram illustrating the extraction process in step S43. As shown in Figure 11, a first threshold TH1A is set for band BA with respect to the amplitude of the envelope spectrum, a first threshold TH1B is set for band BB, and a first threshold TH1C is set for band BC. In the extraction process of step S43, data exceeding the corresponding first thresholds TH1A, TH1B, and TH1C for each band BA, BB, and BC (data shown as black circles in Figure 11) is extracted.

[0053] Then, using the extracted data, the comparison values ​​CAA, CAB, and CAC are calculated for each frequency band using the aforementioned formula (2) (S44). Finally, the comparison value data, which links the reference value data with the first thresholds TH1A, TH1B, and TH1C set in step S42 and the normal state comparison values ​​CAA, CAB, and CAC calculated in step S44, is saved in the database for each frequency band (S45).

[0054] Furthermore, as shown in Figure 7, if multiple reference value data are acquired, a first threshold value is set individually for each reference value data. Figure 12 is a flowchart illustrating the process of calculating a comparison value when there are multiple reference value data. In Figure 12, the processes in steps S51 to S55 are the same as the processes in steps S31 to S35 in Figure 8, but performed separately for each reference value data.

[0055] Specifically, a first threshold value is individually set for the envelope spectrum contained in each of the reference value data (No. 1 to No. X) (S52), data exceeding the individually set first threshold value is extracted (S53), and a comparison value is individually calculated for each reference value data (S54). Finally, comparison value data, which links the first threshold value set in step S52 and the normal state comparison value calculated in step S54 to the reference value data, is saved in the database for each reference value data (S55).

[0056] 3) Once the abnormality diagnosis processing standard value and the comparison value have been calculated, the condition monitoring device 100 then performs a process to diagnose whether or not the target bearing is abnormal.

[0057] Figure 13 is a flowchart illustrating the abnormality diagnosis process. First, in steps S101 to S104, the condition monitoring device 100 obtains an envelope spectrum from the vibration waveform obtained from the bearing to be diagnosed using the same method as in steps S1 to S4 in Figure 3.

[0058] Then, in step S105, the condition monitoring device 100 retrieves comparison data of normal bearings that have been previously stored in the database. Next, in step S106, the condition monitoring device 100 extracts only the data (peak points of the envelope spectrum) that exceed the first threshold obtained from the database from the envelope spectrum to be diagnosed. In step S107, the condition monitoring device 100 calculates a comparison value (hereinafter referred to as the comparison value of the diagnostic target) for the data extracted in step S106.

[0059] In step S108, the condition monitoring device 100 calculates the ratio of the comparison values. Here, the comparison value corresponds to the value obtained by integrating the amplitudes, as can be seen from equation (2). The ratio of the comparison values ​​is the value obtained by dividing the comparison value of the item to be diagnosed by the comparison value of the normal state. In step S109, the condition monitoring device 100 sets a threshold value (hereinafter referred to as the second threshold value) for determining the ratio of the comparison values. For example, the threshold value for the ratio of the comparison values ​​can be a real multiple (2 to 5 times) of the comparison value of the normal state.

[0060] Then, in step S110, the condition monitoring device 100 determines whether the ratio of the comparison values ​​is equal to or greater than the second threshold. If the ratio of the comparison values ​​is equal to or greater than the second threshold (YES in S110), the condition monitoring device 100 determines in step S111 that the condition of the bearing to be diagnosed is abnormal. On the other hand, if the ratio of the comparison values ​​is less than the second threshold (NO in S110), the condition monitoring device 100 determines in step S112 that the condition of the bearing to be diagnosed is normal.

[0061] Furthermore, as shown in Figures 10 and 11, when the maximum frequency of the envelope spectrum is divided into fixed intervals and the first threshold and comparison values ​​are calculated for each frequency band, an anomaly diagnosis is performed as follows.

[0062] Figure 14 is a flowchart illustrating the anomaly diagnosis process when comparison values ​​are calculated for each frequency band. In Figure 14, steps S121 to S132 are the same as steps S101 to S112 in Figure 13, but performed for each frequency band.

[0063] Similarly, when using comparison value data calculated for each frequency band, only the data exceeding the threshold obtained in step S125 is extracted for each frequency band (S125, S126). Then, comparison values ​​are calculated for each frequency band using the extracted data (S127). After that, the ratio of the comparison values ​​is calculated for each frequency band (S128), and a second threshold for the ratio of the comparison values ​​is set for each frequency band (S129).

[0064] The condition monitoring device 100 determines that the condition of the bearing to be diagnosed is abnormal if there is a frequency band in which the ratio of the comparison values ​​is equal to or greater than the second threshold (YES in S130) (S131). On the other hand, if there is no frequency band in which the ratio of the comparison values ​​is equal to or greater than the second threshold (NO in S130), the condition monitoring device 100 determines that the condition of the bearing to be diagnosed is normal (S132).

[0065] Furthermore, as shown in Figure 12, when the first threshold and comparison value are calculated individually for multiple vibration signals, an anomaly diagnosis is performed as follows.

[0066] Figure 15 is a flowchart illustrating the abnormality diagnosis process when a first threshold and comparison value are calculated individually for multiple vibration signals. In Figure 15, steps S141 to S144 are the same as steps S101 to S104 in Figure 13.

[0067] Next, multiple comparison value data (No. 1 to No. X) are obtained, and data exceeding the threshold of the obtained comparison value data are individually extracted (S145, S146). The comparison values ​​corresponding to No. 1 to No. X are individually calculated using the aforementioned formula (2) (S147). Subsequently, the comparison values ​​of No. 1 to No. X obtained in step S147 (let's call them CA1 to CAX) are divided by the comparison values ​​corresponding to No. 1 to No. X obtained from the database in step S145 (let's call them CAt1 to CAtX) to individually calculate the ratio of the comparison values ​​corresponding to No. 1 to No. X that are the subject of diagnosis (CA1 / CAT1, CA2 / CAT2, ..., CAX / CATX) (S148).

[0068] In step S149, the mean, median, and other basic statistics of the ratios of the X comparison values ​​(CA1 / CAT1, CA2 / CAT2, ..., CAX / CATX) calculated individually in S148 are used as the final ratios of the comparison values.

[0069] The condition monitoring device 100 determines that the condition of the bearing to be diagnosed is abnormal if the ratio of the comparison values ​​is equal to or greater than the second threshold (YES in S151) (S152). On the other hand, if the ratio of the comparison values ​​is not equal to or greater than the second threshold (NO in S151), the condition monitoring device 100 determines that the condition of the bearing to be diagnosed is normal (S153).

[0070] As described above, in this embodiment, an envelope spectrum is obtained by filtering the vibration waveform of a bearing in a normal state, and a first threshold value is set based on the integrated value or statistical quantity of the amplitude of the envelope spectrum. For both the envelope spectrum in the normal state and the envelope spectrum to be diagnosed, data with amplitudes exceeding the first threshold value are extracted, and an integrated amplitude value is obtained by integrating the amplitudes of the extracted data. By comparing and evaluating the integrated amplitude value in the normal state and the integrated amplitude value in the bearing to be diagnosed, the occurrence of surface roughness is detected. As part of this comparison and evaluation, the ratio of the comparison values ​​(= integrated amplitude value in the bearing to be diagnosed / integrated amplitude value in the normal state) is calculated, and if the ratio of the comparison values ​​exceeds the second threshold value, it is determined that there is a surface roughness abnormality.

[0071] The inventors of this application conducted an experiment to determine whether a normal bearing and a bearing with surface roughness could be distinguished using the diagnostic method of this disclosure. First, reference values ​​and comparison values ​​were calculated from the vibration signals obtained from a normal bearing, and then comparison values ​​were calculated from the vibration signals obtained from the bearing to be diagnosed (with surface roughness). Finally, the ratio of the comparison values ​​was calculated from the two obtained comparison values.

[0072] The test conditions for acquiring the vibration signal were as follows: • Test bearing: Angular contact ball bearing • Surface roughness: Inner and outer rings of the bearing • Rotational speed: 1000 min －１ - Sampling frequency: 25.6 kHz The analysis conditions for analyzing the vibration signal were as follows: - Filter: High-pass filter 4 kHz - Method for calculating the reference value: RMS value of the envelope waveform - First threshold: 2 to 5 times the reference value - Method for calculating the comparison value: RMS value of the amplitude of the envelope spectral data that exceeded the first threshold Figure 16 is a diagram illustrating an example of the relationship between the ratio of the reference value and the comparison value and the second threshold.

[0073] When the first threshold value TH1 was set to twice the reference value R, the comparison value was 7.4 for a normal bearing and 46.2 for the bearing under diagnosis. In this case, the ratio of the comparison values ​​was 6.3.

[0074] When the first threshold value TH1 was set to three times the reference value R, the comparison value was 4.5 for a normal bearing and 38.1 for the bearing under diagnosis. In this case, the ratio of the comparison values ​​was 8.4.

[0075] When the first threshold value TH1 was set to four times the reference value R, the comparative value was 2.9 for a normal bearing and 31.8 for the bearing under diagnosis. In this case, the ratio of the comparative values ​​was 11.0.

[0076] When the first threshold value TH1 was set to five times the reference value R, the comparison value for a normal bearing was 2.0, while the comparison value for the bearing under diagnosis was 26.8. In this case, the ratio of the comparison values ​​was 13.6.

[0077] In each of the above cases, the ratio of the obtained comparative values ​​is shown in Figure 16. For a normal bearing, the ratio of the comparative values ​​is 1.0, and is therefore indicated by a circle in Figure 16. On the other hand, bearings that are being diagnosed and have surface roughness correspond to the tips of the bar graphs and are therefore indicated by triangles in Figure 16.

[0078] As shown by the dashed line in Figure 16, by setting the second threshold value applied to the ratio of the comparison values ​​to 2 to 5, it is possible to distinguish bearings with surface roughness from normal bearings.

[0079] As explained above, in this embodiment, since anomaly diagnosis is performed using comparative values ​​and basic statistics as reference values ​​in the envelope spectrum, it is less susceptible to the influence of the presence or absence of peaks or the magnitude of their amplitudes in specific frequency bands. Therefore, anomalies can be detected even in surface roughness where no peaks are observed in a specific frequency band (bearing-specific damage frequency) in the envelope spectrum.

[0080] Furthermore, by using only amplitude data exceeding a threshold value set based on a reference value calculated from the amplitude in the envelope spectrum, the bearing condition can be accurately evaluated with less susceptibility to noise. Therefore, abnormalities can be detected even in the initial stages of surface roughness where the generated vibrations are small.

[0081] [Note] The exemplary embodiments described above will be understood by those skilled in the art to be specific examples of the following embodiments.

[0082] (1) This disclosure relates to a condition monitoring device 100 for detecting surface roughness of a bearing 12. The condition monitoring device 100 includes a storage device 150 that stores vibration data obtained from a sensor 20 that detects physical quantities caused by vibration, and a calculation device 160 that receives vibration data from the storage device 150 and performs abnormality determination. The calculation device 160 is configured to perform the following processes: obtaining the envelope spectrum of the bearing 12 to be diagnosed from the vibration data of the bearing 12 to be diagnosed; extracting data from the envelope spectrum of the bearing to be diagnosed in which the amplitude exceeds a first threshold; calculating a comparison value of the data to be diagnosed from the amplitude of the extracted data to be diagnosed; and determining that there is an abnormality if the ratio of the comparison value of the data to be diagnosed is obtained by dividing the comparison value of the data to be diagnosed by a reference comparison value, and if the ratio of the comparison value is equal to or greater than a preset second threshold.

[0083] (2) In the condition monitoring device described in paragraph 1, the calculation device 160 is configured to perform the following processes: obtaining an envelope spectrum of a normal state from the vibration signal of a bearing in a normal state; extracting normal data in the envelope spectrum of a normal state in which the amplitude exceeds a first threshold; and calculating a reference comparison value CA from the amplitude of the extracted normal data.

[0084] (3) In the state monitoring device described in paragraph 2, the calculation device 160 is configured to set a first threshold value R calculated from the amplitude of the envelope spectrum in a normal state.

[0085] (4) In the status monitoring device described in paragraph 3, the calculation device 160 is configured to set a real multiple of the reference value R as the first threshold and a real multiple of the reference comparison value CA as the second threshold.

[0086] (5) In the condition monitoring device described in paragraph 3, the calculation device 160 is configured to store reference value data in the storage device 150, which includes either the envelope spectrum corresponding to a bearing in a normal state, a reference value R, a method for calculating the reference value R, a first threshold value TH1, or a reference comparison value CA, and to acquire the reference value data from the storage device 150 and use it for abnormality diagnosis.

[0087] (6) In the state monitoring device described in paragraph 2, the calculation device 160 is configured to set one first threshold value TH1 for the entire frequency band of the envelope spectrum, as shown in Figure 9.

[0088] (7) In the state monitoring device described in paragraph 2, the calculation device 160 is configured to divide the bandwidth of the envelope spectrum into a plurality of frequency bands BA-BC, as shown in Figure 11, and to set first threshold values ​​TH1A-TH1C for each of the divided frequency bands BA-BE.

[0089] (8) In other aspects, this disclosure relates to a condition monitoring method for detecting surface roughness of a bearing 12 by analyzing vibration data obtained from a sensor 20 that detects physical quantities caused by vibration of the bearing 12 and outputs vibration signals. The condition monitoring method comprises the steps of: obtaining the envelope spectrum of the bearing to be diagnosed from the vibration data of the bearing to be diagnosed (S103, S104); extracting data from the envelope spectrum of the bearing to be diagnosed in which the amplitude exceeds a first threshold (S106); calculating a comparison value of the data to be diagnosed from the amplitude of the extracted data to be diagnosed (S107); and determining that the ratio of the comparison value of the data to be diagnosed is an abnormality if the ratio of the comparison value is equal to or greater than a preset second threshold (S108-S112).

[0090] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the present invention is indicated by the claims rather than by the description of the embodiments above, and all modifications within the meaning and scope of the claims are intended to be included.

[0091] 10 Test device, 12 Bearing, 14 Outer ring, 16 Inner ring, 18 Rolling element, 19 Rotating shaft, 20 Sensor, 100 Condition monitoring device, 110 Amplifier, 120 Filter, 130 Converter, 140 Data acquisition unit, 150 Storage device, 160 Arithmetic unit, 161 Input unit, 162 I / F unit, 164 RAM, 165 ROM, 166 Output unit, 170 Display unit.

Claims

1. A condition monitoring device for detecting surface roughness of a bearing, comprising: a storage device that stores vibration data obtained from a sensor that detects a physical quantity caused by vibration; and a computing device that receives the vibration data from the storage device and performs abnormality determination, wherein the computing device is configured to perform the following: a process of obtaining the envelope spectrum of the bearing to be diagnosed from the vibration data of the bearing to be diagnosed; a process of extracting data from the envelope spectrum of the bearing to be diagnosed in which the amplitude exceeds a first threshold; a process of calculating a comparison value of the diagnosed data from the amplitude of the extracted data; and a process of dividing the comparison value of the diagnosed data by a reference comparison value to obtain the ratio of the comparison values, and determining that there is an abnormality if the ratio of the comparison values ​​is equal to or greater than a preset second threshold.

2. The state monitoring device according to claim 1, wherein the calculation device is configured to perform the following: a process of obtaining an envelope spectrum of a normal state from vibration data of a bearing in a normal state; a process of extracting normal data in the envelope spectrum of the normal state in which the amplitude exceeds the first threshold; and a process of calculating the reference comparison value from the amplitude of the extracted normal data.

3. The state monitoring device according to claim 2, wherein the calculation device is configured to set the first threshold value based on a reference value calculated from the amplitude of the envelope spectrum in the normal state.

4. The state monitoring device according to claim 3, wherein the calculation device is configured to set a real multiple of the reference value as the first threshold value and a real multiple of the comparison value of the reference as the second threshold value.

5. The state monitoring device according to claim 3, wherein the calculation device is configured to store reference value data including any of the following: an envelope spectrum corresponding to the bearing in a normal state, the reference value, a method for calculating the reference value, a first threshold value, or a comparison value of the reference, in the storage device, and to acquire the reference value data from the storage device and use it for abnormality diagnosis.

6. The state monitoring device according to claim 2, wherein the calculation device is configured to set one first threshold for the entire frequency band of the envelope spectrum.

7. The state monitoring device according to claim 2, wherein the calculation device is configured to divide the bandwidth of the envelope spectrum into a plurality of frequency bands and to set the first threshold for each of the divided frequency bands.

8. A condition monitoring method for detecting bearing surface roughness by analyzing vibration data obtained from a sensor that detects a physical quantity caused by bearing vibration and outputs a vibration signal, comprising the steps of: obtaining the envelope spectrum of the bearing to be diagnosed from the vibration data of the bearing to be diagnosed; extracting data from the envelope spectrum of the bearing to be diagnosed in which the amplitude exceeds a first threshold; calculating a comparison value of the bearing to be diagnosed from the amplitude of the extracted data; and determining that the ratio of the comparison value of the bearing to be diagnosed is obtained by dividing the comparison value of the bearing to be diagnosed by a reference comparison value, and determining that there is an abnormality if the ratio of the comparison value is equal to or greater than a preset second threshold.

Images