Signal processing device, diagnostic system, and signal processing method

Abstract

The objective of the present invention is to provide a signal processing device, a diagnostic system, and a signal processing method capable of improving the accuracy of a determination value used for determining a lubrication state in a rolling device. Provided is a signal processing device (3) for calculating a determination value to be used for determining a lubrication state in a rolling device on the basis of a measurement signal acquired by a sensor (2), the signal processing device (3) comprising: a first processing unit (321) that performs band-limiting and envelope-processing to the measurement signal and converts the measurement signal into a time-domain signal; a second processing unit (322) that converts the time-domain signal into a first frequency-domain signal; a third processing unit (323) that converts a logarithmic spectrum of the first frequency-domain signal into a first quefrency domain signal; a fourth processing unit (324) that generates a second quefrency domain signal in which a higher-order cepstrum domain of the first quefrency domain signal including the rotational frequency of the rolling device is set to a prescribed value; a fifth processing unit (325) that converts the second quefrency domain signal into a second frequency domain signal; and a determination value calculation processing unit (33) that calculates the determination value on the basis of the second frequency domain signal.

Description

Signal processing device, diagnostic system, and signal processing method

[0001] The present invention relates to a signal processing device, a diagnostic system, and a signal processing method.

[0002] As a diagnostic method for rolling devices such as rolling bearings, linear guide devices, ball screws, linear bearings, etc., a method has been disclosed in which a fast Fourier transform (FFT) is performed on measurement signals such as vibrations, sound waves, and acoustic emission (AE), thereby eliminating the influence of highly periodic components such as scratches that have occurred in specific locations, and is able to separate and evaluate the state of less periodic components such as bearing wear and lubricant deterioration (for example, Patent Document 1).

[0003] Japanese Patent Application Laid-Open No. 2020-148461

[0004] In the above-mentioned patent document, preset frequency components are uniformly excluded from the frequency spectrum to calculate a partial overall for abnormality diagnosis as a judgment value. As a result, weakly periodic components within the set frequency components are excluded from the partial overall judgment value, which may reduce the accuracy of diagnosing wear and deterioration of lubricating oil in lubricated parts of rolling devices.

[0005] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a signal processing device, a diagnostic system, and a signal processing method that can improve the accuracy of the judgment value used to judge the lubrication state in a rolling device.

[0006] In order to achieve the above object, a signal processing device according to one aspect of the present invention is a signal processing device that calculates a judgment value used to judge a lubrication state in a rolling device based on a measurement signal acquired by a sensor, and includes: a first processing unit that band-limits and envelope-processes the measurement signal to convert it into a time-domain signal; a second processing unit that converts the time-domain signal into a first frequency-domain signal; a third processing unit that converts a logarithmic spectrum of the first frequency-domain signal into a first quefrency-domain signal; a fourth processing unit that generates a second quefrency-domain signal in which a higher-order cepstrum domain including the rotational frequency of the rolling device of the first quefrency-domain signal is set to a predetermined value; a fifth processing unit that converts the second quefrency-domain signal into a second frequency-domain signal; and a judgment value calculation processing unit that calculates the judgment value based on the second frequency-domain signal.

[0007] With the above configuration, the second frequency domain signal is obtained in which the highly periodic components resulting from the rotational frequency of the rolling device are suppressed, thereby improving the accuracy of the determination value used to determine the lubrication state of the rolling device.

[0008] In a preferred embodiment of the signal processing device, the measurement signal may be a vibration signal acquired by a vibration sensor.

[0009] In a preferred embodiment of the signal processing device, the measurement signal may be an ultrasonic signal acquired by an ultrasonic sensor.

[0010] In a preferred embodiment of the signal processing device, the measurement signal may be an elastic wave signal acquired by an AE sensor.

[0011] In a preferred embodiment of the signal processing device, the fourth processing unit sets a high-order cepstrum region having a quefrency equal to or greater than 0.9 times the reciprocal of the rotation frequency of the rolling device as the predetermined value.

[0012] In a preferred embodiment of the signal processing device, the fourth processing unit sets the higher-order cepstrum domain to zero.

[0013] In a preferred embodiment of the signal processing device, the fourth processing unit may generate the second quefrency domain signal using a preset rotation frequency of the rolling device.

[0014] In a preferred embodiment of the signal processing device, the decision value calculation processing unit may use a partial overall of the second frequency domain signal as the decision value.

[0015] In a preferred embodiment of the signal processing device, the decision value calculation processing unit may calculate the partial overall by band-limiting the second frequency domain signal.

[0016] As a desirable aspect of the signal processing device, the judgment value calculation processing unit may be configured to use any one of the maximum value, effective value, crest factor, kurtosis, and skewness of the time domain signal obtained by IFFT processing of the second frequency domain signal as the judgment value.

[0017] A diagnostic system according to one aspect of the present invention comprises a signal processing device according to any one of claims 1 to 7 and a pump that supplies lubricant to the rolling device, wherein the signal processing device further comprises a judgment processing unit that judges the judgment value using a threshold value to judge the lubrication state of the rolling device, and when the judgment processing unit judges the lubrication state of the rolling device to be normal and then judges the lubrication state of the rolling device to be abnormal, it operates the pump, and when the lubrication state of the rolling device is judged to be abnormal two times in a row, it notifies that the lubrication state of the rolling device is abnormal.

[0018] With the above configuration, it is possible to maintain a good lubricated state of the rolling device and prevent an abnormality from being notified simply due to a lack of lubricant, thereby reducing the burden of managing the rolling device.

[0019] A signal processing method according to one aspect of the present invention is a signal processing method for calculating a judgment value used to judge a lubrication state in a rolling device based on a measurement signal acquired by a sensor, the method comprising: a first processing step of band-limiting and envelope-processing the measurement signal to convert it into a time-domain signal; a second processing step of converting the time-domain signal into a first frequency-domain signal; a third processing step of converting a logarithmic spectrum of the first frequency-domain signal into a first quefrency-domain signal; a fourth processing step of generating a second quefrency-domain signal in which a higher-order cepstrum domain including the rotational frequency of the rolling device of the first quefrency-domain signal is set to a predetermined value; a fifth processing step of converting the second quefrency-domain signal into a second frequency-domain signal; and a judgment value calculation processing step of calculating the judgment value based on the second frequency-domain signal.

[0020] With the above configuration, the second frequency domain signal is obtained in which the highly periodic components resulting from the rotational frequency of the rolling device are suppressed, thereby improving the accuracy of the determination value used to determine the lubrication state of the rolling device.

[0021] In a preferred embodiment of the signal processing method, the measurement signal may be a vibration signal acquired by a vibration sensor.

[0022] In a preferred embodiment of the signal processing method, the measurement signal may be an ultrasonic signal acquired by an ultrasonic sensor.

[0023] In a preferred embodiment of the signal processing method, the measurement signal may be an elastic wave signal acquired by an AE sensor.

[0024] As a desirable aspect of the signal processing method, in the fourth processing step, it is preferable that a high-order cepstrum region having a quefrency equal to or greater than 0.9 times the reciprocal of the rotation frequency of the rolling device is set as the predetermined value.

[0025] As a desirable aspect of the signal processing method, in the fourth processing step, the higher-order cepstrum domain is preferably set to zero.

[0026] As a desirable aspect of the signal processing method, in the fourth processing step, the second quefrency domain signal may be generated using a preset rotation frequency of the rolling device.

[0027] As a desirable aspect of the signal processing method, in the decision value calculation processing step, a partial overall of the second frequency domain signal may be used as the decision value.

[0028] As a desirable aspect of the signal processing method, in the judgment value calculation processing step, the second frequency domain signal may be band-limited to calculate the partial overall.

[0029] As a desirable aspect of the signal processing method, in the judgment value calculation processing step, the judgment value may be any one of the maximum value, effective value, crest factor, kurtosis, and skewness of the time domain signal obtained by IFFT processing the second frequency domain signal.

[0030] A desirable aspect of the signal processing method further includes a judgment processing step in which the judgment value is judged using a threshold value to judge the lubrication state of the rolling device, and in the judgment processing step, if the lubrication state of the rolling device is judged to be normal after which it is judged to be abnormal, lubricant is supplied to the rolling device, and if the lubrication state of the rolling device is judged to be abnormal two times in a row, it is notified that the lubrication state of the rolling device is abnormal.

[0031] With the above configuration, it is possible to maintain a good lubricated state of the rolling device and prevent an abnormality from being notified simply due to a lack of lubricant, thereby reducing the burden of managing the rolling device.

[0032] According to the present invention, a signal processing device, a diagnostic system, and a signal processing method are provided that can improve the accuracy of the judgment value used to judge the lubrication state in a rolling device.

[0033] FIG. 1 is a diagram illustrating an example of a diagnostic system according to a first embodiment. FIG. 2 is a diagram illustrating an example of a block configuration of a signal processing device according to the first embodiment. FIG. 3 is a flowchart illustrating an example of diagnostic processing in the signal processing device according to the first embodiment. FIG. 4 is a schematic diagram illustrating an example of a measurement signal. FIG. 5 is a schematic diagram illustrating an example of a measurement signal after band limitation. FIG. 6 is a schematic diagram illustrating an example of a time domain signal after envelope processing. FIG. 7 is a schematic diagram illustrating an example of a first frequency domain signal. FIG. 8 is a schematic diagram illustrating an example of a first quefrency domain signal. FIG. 9 is a schematic diagram illustrating an example of a second quefrency domain signal. FIG. 10 is a schematic diagram illustrating an example of a second frequency domain signal. FIG. 11 is a diagram illustrating an example of a block configuration of a signal processing device according to a second embodiment. FIG. 12 is a diagram illustrating an example of a diagnostic system according to a third embodiment. FIG. 13 is a diagram illustrating an example of a block configuration of a signal processing device according to the third embodiment. FIG. 14 is a diagram illustrating an example of a diagnostic system according to a fourth embodiment. FIG. 15 is a diagram illustrating an example of a block configuration of a signal processing device according to the fourth embodiment. FIG. 16 is a flowchart illustrating an example of diagnostic processing in the signal processing device according to the fourth embodiment.

[0034] Hereinafter, modes for carrying out the invention (hereinafter referred to as embodiments) will be described in detail with reference to the drawings. Note that the present invention is not limited to the following embodiments. Furthermore, the components in the following embodiments include those that can be easily imagined by a person skilled in the art, those that are substantially the same, and those that are within the so-called equivalent range. Furthermore, the components disclosed in the following embodiments can be combined as appropriate.

[0035] (Embodiment 1) Fig. 1 is a diagram showing an example of a diagnostic system according to embodiment 1. In the example shown in Fig. 1 , a signal processing device 3 according to embodiment 1 is applied to a diagnostic system 1 for a rolling bearing (hereinafter also simply referred to as "bearing") 11 incorporated in a mechanical facility 10.

[0036] Examples of the mechanical equipment 10 include a railway vehicle, a machine tool, a wind power generation device, an elevator device, etc. In the example shown in Fig. 1 , a signal processing device 3 determines the lubrication state of a bearing 11 based on a measurement signal acquired by a sensor 2 provided on the mechanical equipment 10.

[0037] The application range of the signal processing device 3 according to the present disclosure is not limited to the bearing diagnostic system as shown in FIG. 1 , but can be widely applied to rolling devices such as rolling bearings, linear guide devices, ball screws, and linear bearings.

[0038] An example of the sensor 2 is an acceleration sensor such as an acceleration pickup. The sensor 2 may be installed at any location where it can detect vibrations that occur as the bearing 11 rotates. In addition to an acceleration sensor, other sensors that can be used as the sensor 2 include an ultrasonic sensor, an AE (Acoustic Emission) sensor, and a shock pulse sensor. Sensors that can detect acceleration, velocity, strain, stress, displacement, etc., and equivalently detect vibrations and convert them into electrical signals can also be used as appropriate.

[0039] The signal processing device 3 receives as input the measurement signals acquired by the sensors 2 (for example, vibration signals acquired by an acceleration sensor such as an acceleration pickup, ultrasonic signals acquired by an ultrasonic sensor, and elastic wave signals acquired by an AE sensor).

[0040] Examples of judgment values ​​used in diagnosing the bearing 11 include the partial overall power spectrum (POA) of the measurement signal acquired by the sensor 2 converted into a frequency domain signal, or scalar values ​​such as the maximum value, root mean square (RMS), peak value, kurtosis, and skewness of the time domain signal that can be acquired by further converting the frequency domain signal into a time domain signal.

[0041] Vibration components caused by the lubrication state of the bearing 11, such as the amount of lubricant in the bearing 11, the degree of deterioration, and the load (preload state) of the bearing 11, are relatively small compared to periodic vibration components that occur when, for example, damage occurs to the raceway surfaces or rolling elements 113 of the inner ring 111 or outer ring 112 of the bearing 11. Therefore, in order to determine the lubrication state of the bearing 11 with high accuracy, it is necessary to suppress the highly periodic vibration components that occur with the rotation of the bearing 11. Below, a configuration of a signal processing device 3 according to embodiment 1 that is capable of suppressing the highly periodic vibration components that occur with the rotation of the bearing 11, and specific examples of signal processing in the signal processing device 3 according to embodiment 1 will be described with reference to each of FIGS. 2 to 10 .

[0042] 2 is a diagram illustrating an example of a block configuration of a signal processing device according to embodiment 1. As illustrated in FIG. 2, the signal processing device 3 according to embodiment 1 includes an AD conversion unit 31, a first processing unit 321, a second processing unit 322, a third processing unit 323, a fourth processing unit 324, a fifth processing unit 325, a judgment value calculation processing unit 33, and a storage unit 34. The first processing unit 321, the second processing unit 322, the third processing unit 323, the fourth processing unit 324, the fifth processing unit 325, and the judgment value calculation processing unit 33 are components that can be realized by processing of an arithmetic processing device such as a CPU (Central Processing Unit) or an MPU (Micro Processing Unit). The storage unit 34 is configured by a storage device such as a ROM (Read Only Memory) or a RAM (Random Access Memory).

[0043] 3 is a flowchart showing an example of a diagnostic process in the signal processing device according to embodiment 1. First, the signal processing device 3 acquires a measurement signal input from the sensor 2 (step S101).

[0044] The AD converter 31 performs AD conversion processing on the measurement signal input from the sensor 2 in step S101 to convert it into digital data (step S102). Fig. 4 is a schematic diagram showing an example of the measurement signal.

[0045] The first processing unit 321 performs band limiting processing (step S103) and envelope processing (envelope processing) on ​​the measurement signal converted into digital data by the AD conversion unit 31 to convert it into a time domain signal (step S104). Fig. 5 is a schematic diagram showing an example of the measurement signal after band limiting. Fig. 6 is a schematic diagram showing an example of the time domain signal after envelope processing. Figs. 5 and 6 show an example in which a T-period vibration component appears in the time domain signal.

[0046] The second processing unit 322 performs a fast Fourier transform (FFT) on the time-domain signal after the envelope processing to convert it into a first frequency-domain signal including signal strength for each frequency (step S105). Fig. 7 is a schematic diagram showing an example of the first frequency-domain signal.

[0047] Zfn shown in Fig. 7 indicates the rotational frequency component of the bearing 11. 2Zfn shown in Fig. 7 indicates the second harmonic component of the rotational frequency of the bearing 11. 3Zfn shown in Fig. 7 indicates the third harmonic component of the rotational frequency of the bearing 11. 4Zfn shown in Fig. 7 indicates the fourth harmonic component of the rotational frequency of the bearing 11. These highly periodic frequency components are assumed to be frequency components caused by the rotation of the bearing 11.

[0048] The third processing unit 323 performs a logarithmic transformation on the first frequency domain signal after the FFT processing (step S106), and then performs an inverse fast Fourier transform (IFFT) on the logarithmic spectrum after the logarithmic transformation processing to convert it into a first quefrency domain signal (step S107). Fig. 8 is a schematic diagram showing an example of the first quefrency domain signal.

[0049] The fourth processing unit 324 multiplies the reciprocal 1 / Zfn of the rotational frequency component of the bearing 11 by a predetermined coefficient (e.g., 0.9) to generate a second quefrency domain signal in which a high-order cepstrum domain equal to or greater than 1 / Zfn × 0.9 is set to a predetermined value (e.g., zero) (step S108). Fig. 9 is a schematic diagram showing an example of the second quefrency domain signal.

[0050] The coefficient by which the reciprocal 1 / Zfn of the rotational frequency component of the bearing 11 is multiplied is not limited to 0.9. For example, the coefficient may be a value that allows setting of a region containing the reciprocal 1 / Zfn of the rotational frequency component of the bearing 11 and having high cepstrum intensity extending in the quefrency direction (horizontal axis direction). The coefficient by which the reciprocal 1 / Zfn of the rotational frequency component of the bearing 11 is multiplied is preferably a value around 0.9. Specifically, for example, if the coefficient by which the reciprocal 1 / Zfn of the rotational frequency component of the bearing 11 is multiplied is set to be around 1 or more, the reciprocal component of the rotational frequency component of the bearing 11 cannot be sufficiently suppressed. Furthermore, for example, if the coefficient by which the reciprocal 1 / Zfn of the rotational frequency component of the bearing 11 is multiplied is set to be around 0.8 or less, low-periodic components resulting from the lubrication state of the bearing 11 are removed, which may reduce the accuracy of the determination value used to determine the lubrication state of the bearing 11.

[0051] Specifically, the fourth processing unit 324 performs a liftering process (lifter process) that sets, for example, a high-order cepstrum region of the first quefrency domain signal equal to or greater than 1 / Zfn×0.9 to zero. Note that the coefficient by which 1 / Zfn is multiplied is not limited to 0.9. Furthermore, the high-order cepstrum region is not limited to zero. Specifically, the coefficient by which 1 / Zfn is multiplied may be, for example, a Logarithmic Amplitude Spectrum (LAS) value corresponding to 1 / Zfn×(a predetermined coefficient) of the first quefrency domain signal, or a sufficiently small value equal to or less than the LAS value corresponding to 1 / Zfn×(a predetermined coefficient) of the first quefrency domain signal.

[0052] In the configuration of the signal processing device 3 according to the first embodiment, the rotational frequency of the bearing 11 is a specified value, which is stored in the storage unit 34. The fourth processing unit 324 reads out the specified value Zfn of the rotational frequency of the bearing 11 stored in the storage unit 34, and executes the processing of step S108.

[0053] The fifth processing unit 325 performs FFT processing on the second quefrency domain signal to convert it into a second frequency domain signal (step S109). Fig. 10 is a schematic diagram showing an example of the second frequency domain signal.

[0054] As a result, as shown in Figure 10, a second frequency domain signal is obtained in which the highly periodic frequency components that are assumed to be frequency components caused by the rotation of the bearing 11 and that appeared in the first frequency domain signal shown in Figure 7 are suppressed.

[0055] The judgment value calculation processing unit 33 calculates a judgment value used to judge the lubrication state of the bearing 11 using the second frequency domain signal in which the highly periodic frequency components are suppressed (step S110).

[0056] Specifically, the judgment value calculation processing unit 33 calculates, for example, a partial overall (POA) of the second frequency domain signal as the judgment value used to judge the lubrication state of the bearing 11. At this time, for example, in order to remove disturbance noise and the like, the second frequency domain signal may be band-limited to calculate the partial overall.

[0057] Alternatively, the judgment value calculation processing unit 33, for example, calculates one of the maximum value, root mean square (RMS), peak value, kurtosis, skewness, etc. of the time domain signal obtained by further IFFT processing the second frequency domain signal, and sets this as the judgment value to be used in determining the lubrication state of the bearing 11.

[0058] This makes it possible to obtain a judgment value in which the highly periodic vibration component caused by the rotation of the bearing 11 is suppressed.

[0059] The signal processing device 3 uses the determination value calculated by the processes of steps S101 to S110 to determine the lubrication state of the bearing 11 (step S111). After that, the process returns to step S101, and the processes of steps S101 to S111 are repeatedly executed.

[0060] The diagnostic processing in the signal processing device 3 according to the first embodiment described above enables highly accurate diagnosis in which the highly periodic vibration components associated with the rotation of the bearing 11 are suppressed when determining the lubrication state of the bearing 11.

[0061] 11 is a diagram showing an example of a block configuration of a signal processing device according to embodiment 2. Here, configurations that are different from those of embodiment 1 described above will be described in detail, and detailed descriptions of configurations that are the same as those of embodiment 1 described above may be omitted.

[0062] In the configuration of the signal processing device 3a according to the second embodiment shown in FIG. 11, the fourth processing unit 324a extracts the rotational frequency component Zfn of the bearing 11 from the first frequency domain signal (see FIG. 7) generated by the second processing unit 322a, and executes the processing of step S108 of the diagnostic processing shown in FIG. 3.

[0063] This allows for higher accuracy in removing the rotational frequency component of the bearing 11 than in the configuration according to the first embodiment.

[0064] (Embodiment 3) Fig. 12 is a diagram showing an example of a diagnostic system according to embodiment 3. Fig. 13 is a diagram showing an example of a block configuration of a signal processing device according to embodiment 3. Here, configurations that are different from those of the above-described embodiments 1 and 2 will be described in detail, and detailed description of configurations that are similar to those of the above-described embodiments 1 and 2 may be omitted.

[0065] In the diagnostic system 1a shown in FIG. 12, the rotational frequency Zfn is input from the bearing 11 to a signal processing device 3b according to the third embodiment.

[0066] In the configuration of the signal processing device 3b according to the third embodiment shown in FIG. 13, the fourth processing unit 324b executes the processing of step S108 of the diagnostic processing shown in FIG. 3 using the rotational frequency Zfn input from the rotation sensor 4 provided on the bearing 11.

[0067] As a result, similar to the configuration according to the second embodiment, the precision of removing the rotational frequency component of the bearing 11 can be improved compared to the configuration according to the first embodiment.

[0068] (Fourth embodiment) Fig. 14 is a diagram showing an example of a diagnostic system according to a fourth embodiment. Fig. 15 is a diagram showing an example of a block configuration of a signal processing device according to the fourth embodiment. Fig. 16 is a flowchart showing an example of diagnostic processing in the signal processing device according to the fourth embodiment. Here, configurations and processing that are different from those of the first embodiment described above will be described in detail, and detailed descriptions of configurations and processing that are similar to those of the first embodiment described above may be omitted.

[0069] In the example shown in FIG. 14 , the signal processing device 3c according to the fourth embodiment, like the signal processing device 3 according to the first embodiment, is applied to a diagnosis system 1b for a bearing 11 incorporated in a mechanical facility 10 such as a railway vehicle, a machine tool, a wind power generation device, an elevator device, etc.

[0070] As shown in Fig. 14, the diagnostic system 1b according to the fourth embodiment further includes a pump 200 that supplies lubricant to the bearing 11, and a notification device 300 that notifies of an abnormality in the bearing 11. Moreover, as shown in Fig. 15, the signal processing device 3c according to the fourth embodiment further includes a determination processing unit 35 that determines the lubrication state of the bearing 11. A specific example of the bearing diagnostic processing in the signal processing device 3c will be described below with reference to Fig. 16.

[0071] The determination processing unit 35 uses the determination value calculated by the processes of steps S101 to S110 to perform a process of determining the lubrication state of the bearing 11 (hereinafter also simply referred to as "bearing determination process"). Specifically, the determination processing unit 35 performs a process of comparing the determination value S calculated by the determination value calculation processing unit 33 with a predetermined determination threshold value Sth (step S121). The determination threshold value Sth is stored in advance in, for example, the storage unit 34a.

[0072] More specifically, the determination processing unit 35 determines whether the determination value S calculated by the determination value calculation processing unit 33 is equal to or greater than the determination threshold value Sth. If the determination value S is less than the determination threshold value Sth (step S121; No), the flag value Flag is set to "0" (step S122), and the process returns to step S101. The flag value Flag is temporarily stored in, for example, the storage unit 34a.

[0073] If the determination value S is equal to or greater than the determination threshold value Sth (step S121; Yes), the determination processor 35 then determines whether the flag value Flag is "1" (step S123). If the flag value Flag is "0" (step S123; No), the determination processor 35 sets the flag value Flag to "1" (step S124) and outputs a control command (hereinafter also referred to as a "lubricant supply command") to the pump 200 to instruct it to supply lubricant to the bearing 11 (step S125).

[0074] The pump 200 supplies lubricant to the bearing 11 based on a lubricant supply command output from the signal processing device 3c.

[0075] If the flag value Flag is "1" (step S123; Yes), this indicates that the judgment value S again became equal to or greater than the judgment threshold value Sth in the previous bearing judgment process immediately after lubricant was supplied to the bearing 11 in step S125 (step S121; Yes). In this case, the judgment processing unit 35 sets the flag value Flag to "0" (step S126), outputs a control command (hereinafter also referred to as an "abnormality notification command") to the alarm device 300 indicating that an abnormality has occurred in the bearing 11 (step S127), and ends the diagnosis process.

[0076] Based on the abnormality notification command output from the signal processing device 3c, the notification device 300 issues an alarm indicating that an abnormality has occurred in the bearing 11. This makes it possible to notify the manager of the mechanical equipment 10 that an abnormality has occurred in the bearing 11 provided in the mechanical equipment 10.

[0077] In the diagnosis process according to the fourth embodiment, if the flag value is "0" and an abnormality is determined in the bearing determination process (step S121; Yes), lubricant is supplied to the bearing 11 (step S125). This allows the bearing 11 to be kept well lubricated.

[0078] Furthermore, even though lubricant has been supplied to the bearing 11 in step S125, if an abnormality is determined again in the bearing determination process when the flag value is "1" (step S121; Yes), it is assumed that an abnormality has occurred in the bearing 11, and the manager of the machinery equipment 10 is notified of the abnormality in the bearing 11 (step S127). This makes it possible to prevent an abnormality being notified simply because of a lack of lubricant.

[0079] In this way, the bearing diagnosis process according to the fourth embodiment can reduce the management burden on the mechanical equipment 10.

[0080] REFERENCE SIGNS LIST 1, 1a, 1b Diagnosis system 2 Sensor 3, 3a, 3b, 3c Signal processing device 4 Rotation sensor 10 Machine equipment 11 Bearing 31 AD conversion unit 321 First processing unit 322 Second processing unit 323 Third processing unit 324 Fourth processing unit 325 Fifth processing unit 33 Judgment value calculation processing unit 34, 34a Storage unit 35 Judgment processing unit 111 Inner ring 112 Outer ring 113 Rolling element 200 Pump 300 Notification device

Claims

1. A signal processing device that calculates a judgment value used to judge a lubrication state of a rolling device based on a measurement signal acquired by a sensor, comprising: a first processing unit that band-limits and envelope-processes the measurement signal to convert it into a time-domain signal; a second processing unit that converts the time-domain signal into a first frequency-domain signal; a third processing unit that converts a logarithmic spectrum of the first frequency-domain signal into a first quefrency-domain signal; a fourth processing unit that generates a second quefrency-domain signal in which a higher-order cepstrum domain including the rotational frequency of the rolling device of the first quefrency-domain signal is set to a predetermined value; a fifth processing unit that converts the second quefrency-domain signal into a second frequency-domain signal; and a judgment value calculation processing unit that calculates the judgment value based on the second frequency-domain signal.

2. The signal processing device according to claim 1, wherein the measurement signal is a vibration signal acquired by a vibration sensor.

3. The signal processing device according to claim 1, wherein the measurement signal is an ultrasonic signal acquired by an ultrasonic sensor.

4. The signal processing device according to claim 1, wherein the measurement signal is an elastic wave signal acquired by an AE sensor.

5. The signal processing device according to claim 1, wherein the fourth processing unit sets a high-order cepstrum region equal to or greater than a quefrency obtained by multiplying the reciprocal of the rotation frequency of the rolling device by 0.9 as the predetermined value.

6. The signal processing device according to claim 1, wherein the fourth processing unit sets the higher-order cepstrum region to zero.

7. The signal processing device according to claim 1, wherein the fourth processing unit generates the second quefrency domain signal using a preset rotation frequency of the rolling device.

8. The signal processing device according to any one of claims 1 to 7, wherein the judgment value calculation processing unit uses a partial overall of the second frequency domain signal as the judgment value.

9. The signal processing device according to claim 8, wherein the judgment value calculation processing unit calculates the partial overall by band-limiting the second frequency domain signal.

10. The signal processing device according to any one of claims 1 to 7, wherein the judgment value calculation processing unit sets the judgment value to one of the maximum value, effective value, crest factor, kurtosis, and skewness of a time domain signal obtained by IFFT processing of the second frequency domain signal.

11. A diagnostic system comprising: a signal processing device according to any one of claims 1 to 7; and a pump for supplying lubricant to the rolling device, wherein the signal processing device further comprises a judgment processing unit that judges the lubrication state of the rolling device by making a threshold judgment on the judgment value, and wherein the judgment processing unit operates the pump when it judges the lubrication state of the rolling device to be normal and then judges the lubrication state of the rolling device to be abnormal, and when it judges the lubrication state of the rolling device to be abnormal two times in a row, reports that the lubrication state of the rolling device is abnormal.

12. A signal processing method for calculating a judgment value used to judge a lubrication state of a rolling device based on a measurement signal acquired by a sensor, the signal processing method comprising: a first processing step of band-limiting and envelope-processing the measurement signal to convert it into a time domain signal; a second processing step of converting the time domain signal into a first frequency domain signal; a third processing step of converting a logarithmic spectrum of the first frequency domain signal into a first quefrency domain signal; a fourth processing step of generating a second quefrency domain signal in which a higher-order cepstrum domain including the rotational frequency of the rolling device of the first quefrency domain signal is set to a predetermined value; a fifth processing step of converting the second quefrency domain signal into a second frequency domain signal; and a judgment value calculation processing step of calculating the judgment value based on the second frequency domain signal.

13. The signal processing method according to claim 12, wherein the measurement signal is a vibration signal acquired by a vibration sensor.

14. The signal processing method according to claim 12, wherein the measurement signal is an ultrasonic signal acquired by an ultrasonic sensor.

15. The signal processing method according to claim 12, wherein the measurement signal is an elastic wave signal acquired by an AE sensor.

16. A signal processing method according to claim 12, wherein in the fourth processing step, a high-order cepstrum region equal to or greater than a quefrency equal to 0.9 times the reciprocal of the rotational frequency of the rolling device is set as the predetermined value.

17. A signal processing method according to claim 12, wherein in said fourth processing step, said high-order cepstrum domains are set to zero.

18. The signal processing method according to claim 12, wherein in the fourth processing step, the second quefrency domain signal is generated using a preset rotation frequency of the rolling device.

19. A signal processing method according to any one of claims 12 to 18, wherein in the judgment value calculation processing step, a partial overall of the second frequency domain signal is set as the judgment value.

20. The signal processing method according to claim 19, wherein in said judgment value calculation processing step, said second frequency domain signal is band-limited to calculate a partial overall.

21. A signal processing method according to any one of claims 12 to 18, wherein in the judgment value calculation processing step, the judgment value is one of the maximum value, effective value, crest factor, kurtosis, and skewness of a time domain signal obtained by IFFT processing of the second frequency domain signal.

22. A signal processing method according to any one of claims 12 to 18, further comprising a judgment processing step of judging the judgment value against a threshold value to judge the lubrication state of the rolling device, wherein in the judgment processing step, if the lubrication state of the rolling device is judged to be normal after being judged to be abnormal, lubricant is supplied to the rolling device, and if the lubrication state of the rolling device is judged to be abnormal two times in a row, a signal is issued that the lubrication state of the rolling device is abnormal.

Images