Rolling bearing replacement timing prediction method and rolling bearing replacement timing prediction device

Abstract

To provide a method and a device for predicting the replacement time of a rolling bearing capable of more accurately predicting the replacement time.SOLUTION: In the present invention, a frequency spectrum of vibration data representing vibration of a rolling bearing is obtained for each sampling period and stored in a storage unit 6 in association with the sampling period, and a specific peak that does not appear in a normal state is detected from the frequency spectrum within a first frequency range including a theoretical frequency at which a peak occurs in an abnormal state. When the specific peak is detected, a specific peak is detected as a re-detected specific peak within a second frequency range narrower than the first frequency range including the specific peak with respect to one or a plurality of frequency spectra stored in the storage part 6 before the sampling period, and the replacement time of the rolling bearing is predicted based on the specific peak and the re-detected specific peak.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method and an apparatus for predicting when to replace a rolling bearing. [Background technology]

[0002] A rolling bearing is a device that supports a load by placing rolling elements, such as balls or rollers, between two components (a shaft and a raceway), and is installed in devices with a variety of applications that include a rotating body. These rolling bearings can suffer from abnormalities, such as wear (wear and scratches), fatigue due to deformation, and fusion due to pressure, which can impede smooth rolling and cause malfunctions in the device. For this reason, it is desirable to be able to predict the remaining life of a rolling bearing and estimate when it should be replaced. For example, Patent Document 1 discloses an abnormality detection device for a rolling bearing that predicts the remaining life of the bearing.

[0003] The rolling bearing abnormality detection device disclosed in Patent Document 1 comprises a vibration sensor that detects the vibration frequency of the rolling bearing, a bandpass filter that passes the output signal of the vibration sensor with bandpass characteristics selected according to the contact frequency between the inner ring, outer ring, and balls calculated on the basis of the geometric dimensions of the bearing being measured, a conversion device that frequency-converts the time-axis signal from the filter, and a calculation device that calculates a spectral intensity corresponding to the contact frequency based on the output signal of the device and calculates a deterioration index for each of the inner ring, outer ring, and balls from the ratio of this calculated value to a reference spectral intensity, stores the increase over time of the calculated spectral intensity, and predicts the remaining life of the rolling bearing from the rate of this increase. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Utility Model Application Publication No. 2-140451 Summary of the Invention [Problem to be solved by the invention]

[0005] The rolling bearing anomaly detection device disclosed in Patent Document 1 calculates the spectral intensity at the contact frequency between the inner ring, outer ring, and balls calculated based on geometric dimensions, i.e., at the theoretical frequency that produces a peak on the frequency spectrum when an abnormality occurs, and detects abnormalities and predicts the remaining life based on this calculated spectral intensity. However, in reality, when an abnormality occurs, a peak does not necessarily occur at the theoretical frequency, and a peak may occur at a frequency that deviates from the theoretical frequency. In such cases, even if the spectral intensity is calculated based on the theoretical frequency, the peak at the time of abnormality cannot be detected, and the remaining life may not be predicted. Furthermore, Patent Document 1 does not describe the relationship between the rate of increase in spectral intensity and the remaining life, making it unclear how to estimate the remaining life from the rate of increase in spectral intensity. Therefore, there is room for improvement in the remaining life prediction disclosed in Patent Document 1, i.e., the prediction of the replacement time.

[0006] The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a method and device for predicting the replacement time of a rolling bearing that can more accurately predict the replacement time of a rolling bearing. [Means for solving the problem]

[0007] After extensive investigation, the inventors have found that the above object can be achieved by the present invention described below. That is, a method for predicting the replacement time of a rolling bearing according to one aspect of the present invention comprises a spectrum processing step of acquiring vibration data representing vibrations occurring in a rolling bearing for a predetermined sampling period at predetermined acquisition intervals, determining a frequency spectrum of the vibration data for that sampling period, and storing the determined frequency spectrum in a storage unit in association with that sampling period; a peak detection step of detecting, from the frequency spectrum, a specific peak that does not appear when the rolling bearing is normal, within a predetermined first frequency range that includes a theoretical frequency that will produce a peak on the frequency spectrum when an abnormality occurs; and, if a specific peak is detected in the peak detection step, detecting the specific peak. the peak re-detection process detects, as a re-detected specific peak, a specific peak that does not appear when the rolling bearing is normal, within a predetermined second frequency range that is narrower than the first frequency range and that includes the frequency of the specific peak detected in the peak detection process, for one or more frequency spectra stored in the storage unit before the sampling period when the specific peak is detected; and a replacement time prediction process that predicts the replacement time of the rolling bearing based on the sampling period when the specific peak is detected in the peak detection process and the amplitude of the detected specific peak, and the sampling period when the re-detected specific peak is detected in the peak re-detection process and the amplitude of the detected re-detected specific peak. Preferably, in the above-mentioned method for predicting when to replace a rolling bearing, the replacement time prediction step obtains, in a coordinate space having time and amplitude as coordinate axes, a fitting curve that best fits a midpoint of a sampling period when a specific peak is detected in the peak detection step and the amplitude of the detected specific peak, and a midpoint of a sampling period when a re-detected specific peak is detected in the peak re-detection step and the amplitude of the re-detected specific peak, and determines, as the replacement time, a time point at which the obtained fitting curve intersects with a predetermined judgment amplitude value for determining the replacement time. Preferably, in the above-mentioned method for predicting when to replace a rolling bearing, the first frequency range is set so that the theoretical frequency is its midpoint.

[0008] Typically, when a rolling bearing in a normal state experiences some kind of fault that impedes smooth rolling, the fault progresses and the amplitude of the peak increases, after which the peak can be identified and its frequency can be identified. Meanwhile, peaks before identification are presumed to occur around the identified frequency. The rolling bearing replacement time prediction method detects specific peaks in a first frequency range that includes the theoretical frequency, thereby enabling more reliable detection of specific peaks. Furthermore, the method redetects specific peaks by narrowing the range to a second frequency range that is narrower than the first frequency range. This eliminates from detection peaks that are included in the first frequency range but are not caused by the fault, enabling more accurate detection of specific peaks caused by the fault.

[0009] On the other hand, the remaining lifespan ((the end point of the sampling period in the frequency spectrum (vibration data) when the specific peak is detected) + (remaining lifespan at that time) = (replacement time)), which is the time until the rolling bearing needs to be replaced after the fault is identified, depends on the progression of the fault (tendency of change over time, trend). If the fault progresses slowly, the remaining lifespan will be longer, and conversely, if the fault progresses quickly, the remaining lifespan will be shorter. The above-mentioned rolling bearing replacement time prediction method detects not only the specific peak but also the previous redetected specific peak, and predicts the replacement time based on the sampling period and amplitude of the detected specific peak when the specific peak is detected in the peak detection process, and the sampling period and amplitude of the detected redetected specific peak when the redetected specific peak is detected in the peak redetection process. Therefore, the replacement time is predicted taking into account the progression of the fault, thereby enabling more accurate prediction of the replacement time. Furthermore, when the above-mentioned rolling bearing replacement time prediction method predicts the replacement time using multiple redetected specific peaks, it can predict the replacement time more accurately than when predicting the replacement time using a single redetected specific peak.

[0010] In another aspect, in the above-mentioned rolling bearing replacement time prediction device, when the peak redetection step detects the redetected specific peak in a past direction for the multiple frequency spectra, the peak detection step and the specific peak detected in the peak detection step are regarded as the initial peak redetection step and the initial redetected specific peak, respectively, for each of the multiple frequency spectra, and the redetected specific peak is detected in the second frequency range that includes the redetected specific peak detected in the previous peak redetection step for that frequency spectrum. Preferably, in the above-mentioned rolling bearing replacement time prediction method, the second frequency range is set so that the frequency of the redetected specific peak is its center frequency. Preferably, in the above-mentioned rolling bearing replacement time prediction method, the peak redetection step detects the redetected specific peak continuously and sequentially in a past direction for the multiple frequency spectra. Preferably, in the above-mentioned rolling bearing replacement time prediction method, the peak redetection step detects the redetected specific peak discontinuously in a past direction for the multiple frequency spectra.

[0011] This rolling bearing replacement time prediction method detects the redetected specific peak in the second frequency range that includes the redetected specific peak detected in the previous peak redetection process, so that the redetected specific peak can be detected more reliably even if the frequency of the redetected specific peak changes over time.

[0012] In another aspect, in the above-mentioned rolling bearing replacement timing prediction method, the peak detection step further detects, for each sampling period after the sampling period in which the specific peak is detected, the specific peak in the second frequency range that includes the specific peak detected in the peak detection step immediately before the sampling period in question, for the frequency spectrum stored in the memory unit for that sampling period.

[0013] This method for predicting the time to replace a rolling bearing detects a specific peak in a second frequency range that includes the specific peak detected in the previous peak detection process, so that the specific peak can be more reliably detected after detection even if the frequency of the specific peak changes over time after detection.

[0014] In another aspect, in the above-mentioned rolling bearing replacement timing prediction method, the length of the sampling period is set based on the second frequency range, and the narrower the second frequency range, the longer the sampling period is set.

[0015] In this rolling bearing replacement timing prediction method, peaks can be detected from frequency spectrum data with a frequency resolution suitable for the second frequency range.

[0016] According to another aspect of the present invention, a rolling bearing replacement timing prediction device includes a vibration detection sensor that acquires vibration data representing vibrations occurring in a rolling bearing, a storage unit, and a control processing unit that performs spectrum processing, peak detection processing, peak re-detection processing, and replacement timing prediction processing, wherein the spectrum processing acquires vibration data for a predetermined sampling period at predetermined acquisition intervals, calculates a frequency spectrum of the vibration data for the sampling period for the vibration data, and stores the calculated frequency spectrum in the storage unit in association with the sampling period, and the peak detection processing detects a specific peak that does not appear when the rolling bearing is normal from the frequency spectrum within a predetermined first frequency range that includes a theoretical frequency that will produce a peak on the frequency spectrum when an abnormality occurs. The peak re-detection process is a process in which, when a specific peak is detected in the peak detection process, a specific peak that does not appear when the rolling bearing is normal is detected as a re-detected specific peak, within a predetermined second frequency range narrower than the first frequency range and including the frequency of the specific peak detected in the peak detection process, for one or more frequency spectra stored in the storage unit prior to the sampling period when the specific peak is detected, and the replacement time prediction process is a process in which the replacement time of the rolling bearing is predicted based on the sampling period when a specific peak is detected in the peak detection process and the amplitude of the detected specific peak, and the sampling period when a re-detected specific peak is detected in the peak re-detection process and the amplitude of the detected re-detected specific peak.

[0017] Such a rolling bearing replacement timing prediction device can predict the replacement timing of a rolling bearing with higher accuracy. [Effects of the Invention]

[0018] The method and device for predicting the replacement time of a rolling bearing according to the present invention detect a specific peak in a first frequency range that includes the theoretical frequency, thereby enabling more reliable detection of the specific peak, and re-detecting the specific peak by narrowing the range to a second frequency range that is narrower than the first frequency range, thereby excluding from the detection target any confusing peaks that are included in the first frequency range but are not attributable to the occurrence of an obstruction that hinders the smooth rotation of the rolling bearing, thereby enabling more accurate detection of the specific peak. The method and device for predicting the replacement time of a rolling bearing described above detect not only the specific peak but also the previous re-detected specific peak, and predicts the replacement time based on the sampling period when the specific peak is detected and the amplitude of the detected specific peak, and the sampling period when the re-detected specific peak is detected and the amplitude of the re-detected specific peak, thereby predicting the replacement time taking into account the progression of the obstruction, thereby enabling more accurate prediction of the replacement time. [Brief explanation of the drawings]

[0019] [Figure 1] 1 is a block diagram showing the configuration of a rolling bearing replacement timing prediction device according to an embodiment; [Figure 2] FIG. 10 is a diagram for explaining a predetermined time length in vibration data. [Figure 3] FIG. 1 is a diagram illustrating a mechanical facility equipped with a rolling bearing. [Figure 4] FIG. 3 is a diagram for explaining first and second frequency ranges. [Figure 5] FIG. 10 is a diagram for explaining how to predict the replacement time. [Figure 6] 10 is a flowchart showing the operation of the rolling bearing replacement time predicting device before a specific peak is detected. [Figure 7] 10 is a flowchart showing the operation of the rolling bearing replacement time predicting device after the specific peak is detected. [Figure 8] FIG. 10 is a block diagram showing the configuration of a rolling bearing replacement timing prediction device according to a modified example of the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, one or more embodiments of the present invention will be described with reference to the drawings. However, the scope of the invention is not limited to the disclosed embodiments. In addition, components with the same reference numerals in each drawing indicate the same components, and their description will be omitted as appropriate. In this specification, when referring to a general term, a reference numeral without a subscript is used, and when referring to an individual component, a reference numeral with a subscript is used.

[0021] FIG. 1 is a block diagram showing the configuration of a rolling bearing replacement timing prediction device according to an embodiment. FIG. 2 is a diagram illustrating a predetermined time length in vibration data. FIG. 3 is a diagram illustrating mechanical equipment equipped with a rolling bearing. FIG. 4 is a diagram illustrating first and second frequency ranges. FIG. 4A shows the first frequency range, and FIG. 4B shows the second frequency range. The horizontal axis in each of FIGS. 4A and 4B represents frequency [Hz], and the vertical axis represents amplitude. FIG. 5 is a diagram illustrating how to predict replacement timing. The horizontal axis in FIG. 5 represents time (elapsed time), and the vertical axis represents amplitude.

[0022] The rolling bearing replacement time prediction device 1000 in the embodiment includes, for example, a vibration detection sensor 1, a control processing unit 2, an input unit 3, an output unit 4, an interface unit (IF unit) 5, and a memory unit 6, as shown in FIG.

[0023] The vibration detection sensor 1 is connected to the control processing unit 2 and is a device that acquires vibration data that represents vibrations occurring in the rolling bearing in accordance with the control of the control processing unit 2. In this embodiment, a plurality of vibration data acquired over a predetermined sampling period at predetermined acquisition intervals by the vibration detection sensor that detects vibrations occurring in the rolling bearing is stored in the storage unit 6.

[0024] As described below, vibration data is transformed from time space to frequency space using a fast Fourier transform. In this case, the frequency resolution of the frequency spectrum depends on the number of data points used in the fast Fourier transform. The greater the number of data points, the higher the frequency resolution, and the corresponding amount of vibration data is required for the predetermined sampling period. Here, if the sampling interval is SP, the number of data points is Nfft, and the length of the predetermined sampling period is TW, then TW = SP × Nfft. In this embodiment, the length of the predetermined sampling period TW is set based on the second frequency range FW2 described below. The narrower the second frequency range, the longer the sampling period is set. To detect significant peaks in the second frequency range ±Δfw2, for example, as shown in FIG. 2, the second frequency range ±Δfw2 (= 2 × Δfw2) needs to be divided into four or more parts. Therefore, the frequency resolution is 2 × Δfw2 / 4 = Δfw2 / 2 [Hz] or greater, and therefore the length of the predetermined sampling period TW is equal to or greater than its reciprocal, 2 / Δfw2 [seconds]. For example, if Δfw2 is set to 0.02 Hz and the sampling interval SP is set to 0.2 ms (= 0.0002 s), the length TW of the predetermined sampling period is TW = 2 / 0.02 = 100 s or more, and the number of data points Nfft is Nfft = 100 / 0.0002 = 500,000 or more. Since fast Fourier transforms generally handle powers of 2, the minimum number exceeding 500,000 is 2^19 = 524,288, and the length TW of the predetermined sampling period is TW = 0.0002 × 524,288 = 104.8576 s.

[0025] One or more vibration detection sensors 1 are installed in a device to be monitored, such as a mechanical facility, that includes a rolling bearing. The mechanical facility is an example of a device that includes a rolling bearing, and may be any facility that includes a rolling bearing. For example, the mechanical facility M is a reducer M shown in FIG. 3, which generally includes first through third rolling bearings BE-1 through BE-3, first and second rotating shafts AX-1 and AX-2, first and second gears GA-1 and GA-2, and a housing (not shown) that accommodates the first through third rolling bearings BE-1 through BE-3, the first and second rotating shafts AX-1 and AX-2, and the first and second gears GA-1 and GA-2. The first rotating shaft AX-1 is fixed to the first gear GA-1 and is the rotating shaft of the first gear GA-1, and is supported by the first rolling bearing BE-1. The second rotating shaft AX-2 is fixed to the second gear GA-2, is the rotating shaft of this second gear GA-1, and is supported by the second and third rolling bearings BE-2 and BE-3. The first gear GA-1 and the second gear GA-2 are in mesh with each other, and for example, the rotational force caused by the rotation of the first rotating shaft AX-1 is transmitted to the second rotating shaft AX-2 via the first and second gears GA-1 and GA-2, causing the second rotating shaft AX-2 to rotate.

[0026] For the reducer M configured as described above, the vibration detection sensor 1 includes three vibration detection sensors, namely, first to third vibration detection sensors 1-1 to 1-3. The first to third vibration detection sensors 1-1 to 1-3 are respectively arranged on the outer peripheries of the first to third rolling bearings BE-1 to BE-3. The vibration detection sensors 1 (1-1 to 1-3) are not limited to the rolling bearings BE, and may be arranged, for example, on the housing. In short, the first to third vibration detection sensors 1-1 to 1-3 are arranged at locations where vibrations caused by the rolling bearings BE propagate. The first to third vibration detection sensors 1-1 to 1-3 are, for example, acceleration sensors or AE (Acoustic Emission) sensors, and appropriate sensors are used depending on the frequency of vibrations occurring in the monitored object. In this embodiment, the first to third vibration detection sensors 1-1 to 1-3 output detection results to the control processing unit 2.

[0027] In this embodiment, for example, when a predetermined sampling period starts, the control processing unit 2 instructs the first through third vibration detection sensors 1-1 to 1-3 to acquire the vibration data, and in response to this instruction, the first through third vibration detection sensors 1-1 to 1-3 detect vibrations at each sampling timing according to a predetermined sampling interval and output vibration data for the length TW of the predetermined sampling period to the control processing unit 2. The control processing unit 2 stores and saves the vibration data in the storage unit 6 in association with the sampling period (for example, the start time of the sampling period).

[0028] The input unit 3 is connected to the control processing unit 2 and is a device that inputs various commands, such as a command to start replacement timing prediction, and various data required to operate the rolling bearing replacement timing prediction device 1000, such as the name of the machinery and equipment to be monitored, to the rolling bearing replacement timing prediction device 1000, and is, for example, a plurality of input switches to which predetermined functions are assigned, a keyboard, a mouse, etc. The output unit 4 is connected to the control processing unit 2 and is a device that outputs the commands, data, replacement timing, etc. input from the input unit 3 in accordance with the control of the control processing unit 2, and is, for example, a display device such as a CRT display, liquid crystal display, or organic EL display, or a printing device such as a printer.

[0029] The input unit 3 and the output unit 4 may form a so-called touch panel. In this touch panel, the input unit 3 is a position input device, such as a resistive or capacitive type, that detects and inputs an operation position, and the output unit 4 is a display device. In this touch panel, the position input device is provided on the display surface of the display device, and one or more input content candidates that can be input are displayed on the display device. When a user touches the display position showing the input content they want to input, the position is detected by the position input device, and the display content displayed at the detected position is input to the rolling bearing replacement timing prediction device 1000 as the user's operation input content. With such a touch panel, the user can easily intuitively understand the input operation, thereby providing a rolling bearing replacement timing prediction device 1000 that is easy for the user to use.

[0030] The IF unit 5 is connected to the control processing unit 2 and is a circuit that inputs and outputs data to and from external devices under the control of the control processing unit 2, and is, for example, an interface circuit for RS-232C, which is a serial communication method, an interface circuit using the Bluetooth (registered trademark) standard, an interface circuit for infrared communication such as the IrDA (Infrared Data Association) standard, and an interface circuit using the USB (Universal Serial Bus) standard. The IF unit 5 is also a circuit that communicates with external devices, and may be, for example, a data communication card or a communication interface circuit conforming to the IEEE802.11 standard or the like.

[0031] The memory unit 6 is connected to the control processing unit 2 and is a circuit that stores various predetermined programs and various predetermined data under the control of the control processing unit 2. The various predetermined programs include, for example, a control processing program, which includes a control program, a spectrum processing program, a peak detection program, a peak redetection program, and a replacement time prediction program. The control program controls each of the units 1, 3-6 of the rolling bearing replacement time prediction device 1000 according to the function of each unit. The spectrum processing program acquires multiple different vibration data sets for multiple different sampling periods using the vibration detection sensor 1, calculates a frequency spectrum of the vibration data for each of the multiple sampling periods, and stores the calculated frequency spectrum in the memory unit 6 in association with the sampling period. The peak detection program detects a specific peak that does not appear when the rolling bearing is normal, within a predetermined first frequency range that includes a theoretical frequency that produces a peak on the frequency spectrum when an abnormality occurs, from the frequency spectrum. The peak re-detection program is a program that, when a specific peak is detected by the peak detection program, detects as a re-detected specific peak a specific peak that does not appear when the rolling bearing is normal, within a predetermined second frequency range narrower than the first frequency range and that includes the frequency of the specific peak detected by the peak detection program, for one or more frequency spectra stored in storage unit 6 before the sampling period when the specific peak is detected. The replacement time prediction program is a program that predicts the replacement time of the rolling bearing based on the sampling period when a specific peak is detected by the peak detection program and the amplitude of the detected specific peak, and the sampling period when a re-detected specific peak is detected by the peak re-detection program and the amplitude of the detected re-detected specific peak.The various types of predetermined data include data necessary for executing each of these programs, such as a frequency spectrum associated with the sampling period, a theoretical frequency, a first frequency range, a second frequency range, a specific peak, a redetected specific peak, and a predicted replacement time. The storage unit 6 includes, for example, a nonvolatile storage element such as a read-only memory (ROM) or a rewritable nonvolatile storage element such as an electrically erasable programmable read-only memory (EEPROM). The storage unit 6 also includes a random access memory (RAM) that serves as a working memory for the control processing unit 2 and stores data generated during execution of the predetermined programs. The storage unit 6 may also include a hard disk drive or a solid-state drive (SSD) capable of storing relatively large amounts of data.

[0032] The control processing unit 2 is a circuit that controls each of the units 1, 3 to 6 of the rolling bearing replacement timing prediction device 1000 according to the function of each unit, and predicts the replacement time for the rolling bearing BE. The control processing unit 2 is configured to include, for example, a CPU (Central Processing Unit) and its peripheral circuits. By executing the control processing program, the control processing unit 2 is functionally configured to include a control unit 21, spectrum processing unit 22, peak detection unit 3, peak re-detection unit 24, and timing prediction unit 25.

[0033] The control unit 21 controls each of the units 1, 3 to 6 of the rolling bearing replacement timing predicting device 1000 in accordance with the function of each unit, and controls the rolling bearing replacement timing predicting device 1000 as a whole.

[0034] The spectrum processing unit 22 executes spectrum processing to acquire vibration data for each of a plurality of different sampling periods using the vibration detection sensor 1, by calculating a frequency spectrum of the vibration data for each of the plurality of sampling periods, and storing the calculated frequency spectrum in the storage unit 6 in association with the sampling period. More specifically, in the spectrum processing, at the start timing of a sampling period having a preset acquisition interval, each of the first to third vibration detection sensors 1-1 to 1-3 acquires vibration data for the length TW of the predetermined sampling period, and stores the vibration data in the storage unit 6 as the vibration data for the current sampling period. Next, the spectrum acquisition unit 22 converts the vibration data in time space acquired by the first vibration detection sensor 1-1 into vibration data in frequency space using, for example, a fast Fourier transform (FFT), thereby calculating the frequency spectrum of the vibration data, and stores the frequency spectrum in association with the sampling period and the first vibration detection sensor 1-1 (for example, an identifier (sensor ID) of the first vibration detection sensor 1-1) in the storage unit 6. Next, similarly, the spectrum acquisition unit 22 converts the vibration data in time space acquired by the second vibration detection sensor 1-2 into vibration data in frequency space using the FFT, thereby obtaining a frequency spectrum of the vibration data, and stores the frequency spectrum in the storage unit 6 in association with the sampling period and the second vibration detection sensor 1-2 (for example, the identifier (sensor ID) of the second vibration detection sensor 1-2). Similarly, the spectrum acquisition unit 22 converts the vibration data in time space acquired by the third vibration detection sensor 1-3 into vibration data in frequency space using the FFT, thereby obtaining a frequency spectrum of the vibration data, and stores the frequency spectrum in association with the sampling period and the third vibration detection sensor 1-3 (for example, the identifier (sensor ID) of the third vibration detection sensor 1-3) in association with the sampling period. This process is repeatedly executed at the start timing of a sampling period at a preset acquisition interval. The acquisition interval is set in advance as appropriate depending on, for example, the target for which an abnormality is to be detected. Since the life of a rolling bearing can be predicted from, for example, the load and rotation speed, the acquisition interval is set as appropriate depending on the life of the rolling bearing.For example, if the monitoring target is a rolling bearing that supports a shaft that is used relatively hard, such as a shaft that is constantly operated and rotates at high speed, the acquisition interval is set to a relatively short time length, such as one hour or one day, or if the monitoring target is a rolling bearing that supports a shaft that is used relatively softly, such as a shaft that rotates relatively slowly, the acquisition interval is set to a relatively long time length, such as one month or six months. The start timing of the sampling period is represented, for example, by a serial number from the start of acquisition of the vibration data, the time of the start timing, etc.

[0035] When storing the frequency spectrum in the storage unit 6, the peak detection unit 23 executes a peak detection process to detect, from the frequency spectrum, a specific peak that does not appear when the rolling bearing is normal, within a predetermined first frequency range that includes a theoretical frequency that will produce a peak on the frequency spectrum when an abnormality occurs. When the peak detection unit 23 detects a specific peak, it stores in the storage unit 6 the sampling period, amplitude, and frequency of the detected specific peak.

[0036] The theoretical frequency ft, which produces a peak on the frequency spectrum when an abnormality occurs, is publicly known and varies depending on the location of the rolling bearing damage (bearing damage), as shown in Table 1 below, for example. The locations of the bearing damage include, for example, the inner ring, outer ring, rolling elements, and cage. Here, fti is the theoretical frequency when bearing damage occurs on the inner ring, fto is the theoretical frequency when bearing damage occurs on the outer ring, ftb is the theoretical frequency when bearing damage occurs on the rolling elements, and ftm is the theoretical frequency when bearing damage occurs on the cage. d is the diameter of the rolling element, D is the pitch circle diameter of the rolling element, Z is the number of rolling elements, and α is the contact angle.

[0037] [Table 1]

[0038] When a rolling bearing (e.g., an unused rolling bearing) experiences some kind of obstacle that prevents smooth rolling, the frequency of a peak that appears in the frequency spectrum does not actually coincide with the theoretical frequency due to factors such as the rolling bearing's dimensional tolerances and deformation due to load. Therefore, in order to detect unknown peaks that appear in the frequency spectrum, a predetermined first frequency range including the theoretical frequency ft is appropriately set in advance. For example, as shown in FIG. 4A, the first frequency range FW1 is set so that the theoretical frequency ft is its central frequency (ft - Δfw1 ≦ FW1 ≦ ft + Δfw1, where Δfw1 is, for example, approximately 1 to 5% of ft).

[0039] In detecting a specific peak (a peak that appears due to the fault) that does not appear when the rolling bearing is normal, for example, if a peak is detected in a first frequency range FW1 and a peak (a harmonic) also exists at a frequency that is an integer multiple of the frequency of the detected peak (for example, twice the frequency or three times the frequency), the detected peak is determined to be a specific peak and it is determined that the specific peak has been detected. If no peak exists at a frequency that is an integer multiple of the frequency of the detected peak, it is determined that the detected peak is not a specific peak and it is determined that the specific peak has not been detected. For example, in Figure 4A, peaks PK1 and PK2 are detected in the first frequency range FW1, and no peaks exist at integer multiples of the frequency fp1 of peak PK1 that are characteristic of bearing damage vibration, while peaks exist at integer multiples of the frequency fp2 of peak PK2. In this case, peak PK1 is determined not to be a specific peak, while peak PK2 is determined to be a specific peak and it is determined that the specific peak has been detected. In this embodiment, multiple first to third vibration detection sensors 1-1 to 1-3 are used, and therefore, if a specific peak is detected at a common frequency in the frequency spectra of two or more of the first to third vibration detection sensors 1-1 to 1-3, it is ultimately determined that a specific peak has been detected.

[0040] When a specific peak is detected by the peak detection unit 23, the peak re-detection unit 24 executes a peak re-detection process to detect, as a re-detected specific peak, a specific peak that does not appear under normal conditions and that includes the frequency of the specific peak detected by the peak detection unit 23 and is within a predetermined second frequency range FW2 narrower than the first frequency range FW1, for one or more frequency spectra stored in the storage unit 6 before the sampling period when the specific peak is detected. The peak re-detection unit 24 stores the sampling period, amplitude, and frequency of the re-detected specific peak in the storage unit 6.

[0041] Since the purpose is to detect how the fault progresses (tendency of change over time, trend), the peak re-detection unit 24 may detect the specific peak as a re-detected specific peak for one frequency spectrum stored in the storage unit 6 before the sampling period of the frequency spectrum in which the peak detection unit 23 detected the specific peak, or may detect the specific peak as a re-detected specific peak for some frequency spectra (a plurality of frequency spectra less than all) of all frequency spectra stored in the storage unit 6 before the sampling period, or may detect the specific peak as a re-detected specific peak for all frequency spectra stored in the storage unit 6 before the sampling period. When detecting the specific peak as a re-detected specific peak for a plurality of frequency spectra, the peak re-detection unit 24 may, for example, detect the re-detected specific peak consecutively in the past direction for the plurality of frequency spectra, or, for example, the peak re-detection unit 24 may detect the re-detected specific peak non-continuously in the past direction for the plurality of frequency spectra. In this embodiment, the peak re-detection unit 24 successively retrieves a predetermined number of frequency spectra stored in the memory unit 6 in association with a predetermined number of sampling periods prior to the sampling period, in the past direction, and detects the re-detected specific peaks successively in the past direction from the retrieved frequency spectra.

[0042] Typically, a rolling bearing in a normal state experiences some kind of obstacle that impedes smooth rolling. The obstacle progresses, causing the peak amplitude to increase, and the peak can then be identified. Therefore, before the peak is detected by the peak detector 23, the frequency of the specific peak is unknown. Therefore, the first frequency range FW1 must be set relatively wide. Meanwhile, the specific peak before identification is presumed to occur around the identified frequency. Therefore, a predetermined second frequency range that includes the frequency of the specific peak and is narrower than the first frequency range FW1 is appropriately set in advance. For example, as shown in FIG. 4B, the second frequency range FW2 is set so that the frequency fp2 of the specific peak PK2 is its center frequency (fp2 - Δfw2 ≦ FW2 ≦ fp2 + Δfw2, where Δfw2 is, for example, approximately 0.1 to 0.5% of the theoretical frequency ft).

[0043] In the detection of the re-detected specific peak, the peak having the maximum amplitude in the second frequency range FW2 and having peaks at integer multiples of that frequency is determined as the re-detected specific peak. For example, in FIG. 4B, peak PK2' having the maximum amplitude in the second frequency range FW2 and having peaks at integer multiples of that frequency is determined as the re-detected specific peak.

[0044] The second frequency range FW2 may be the same when detecting a re-detected specific peak from each of a plurality of frequency spectra. However, in this embodiment, when detecting the re-detected specific peak in the past direction for the plurality of frequency spectra, the peak re-detector 24 regards the peak detection process and the specific peak detected in the peak detection process as the first peak re-detection process and the first re-detected specific peak, respectively, for each of the plurality of frequency spectra, and detects the re-detected specific peak in the second frequency range FW2 that includes the re-detected specific peak detected in the previous peak re-detection process for that frequency spectrum. For example, the re-detected specific peak can be detected even if there is a shift in the frequency of the re-detected specific peak due to wear, etc.

[0045] Furthermore, after a specific peak is detected for the first time, the peak detection unit 23 may further detect the specific peak in the second frequency range FW2 that includes the first detected specific peak PK2 for each sampling period after the sampling period when the specific peak is detected, for the frequency spectrum stored in the storage unit 6 for that sampling period. However, in this embodiment, the peak detection unit 23 further detects the specific peak in the second frequency range FW2 that includes the specific peak detected in the peak detection process immediately before that sampling period for the frequency spectrum stored in the storage unit 6 for that sampling period after the sampling period when the specific peak is detected. For example, the specific peak can be detected even if there is a shift in the frequency of the specific peak due to wear or the like.

[0046] The timing prediction unit 25 executes a replacement timing prediction process to predict the replacement timing of the rolling bearing based on the sampling period when a specific peak is detected by the peak detection unit 23 and the amplitude of the detected specific peak, and the sampling period when a re-detected specific peak is detected by the peak re-detection unit 24 and the amplitude of the re-detected specific peak.

[0047] When the peak re-detection unit 24 detects one re-detected specific peak using one frequency spectrum, the timing prediction unit 25 predicts the replacement time based on the specific peak detected by the peak detection unit 23 and the one re-detected specific peak detected by the peak re-detection unit 24. When the peak re-detection unit 24 detects multiple re-detected specific peaks using multiple frequency spectrums, the timing prediction unit 25 predicts the replacement time based on the specific peak detected by the peak detection unit 23 and the multiple re-detected specific peaks detected by the peak re-detection unit 24. In the sampling period after the peak detection unit 23 detects a specific peak for the first time, the timing prediction unit 25 predicts the replacement time based on the specific peak detected for the first time, the specific peak detected during that sampling period, and one or more re-detected specific peaks. The timing prediction unit 25 outputs this predicted replacement time to the output unit 4.

[0048] For example, in the example shown in FIG. 5, when the peak detection unit 23 detects a specific peak indicated by a ● at timing TD0 of the sampling period (for example, the central time TD0 as a representative time point of the sampling period), the peak re-detection unit 24 detects a specific peak indicated by a ● at timing TD -1 For the frequency spectrum, a specific peak is detected again within the second frequency range FW2 including the specific peak ●, and is indicated by, for example, × at timing TD -1 The peak re-detection unit 24 detects the re-detection specific peak at the timing TD. Note that the four frequency spectra are extracted and detected going back in time. -1 The timing before TD -2 For the frequency spectrum of -1 The re-detected specific peak is detected within the second frequency range FW2 including the re-detected specific peak x at the timing TD -2 Then, the peak re-detection unit 24 detects the specific peak at the timing TD -2 The timing before TD -3 For the frequency spectrum of -2 In the example shown in FIG. 5, the redetected specific peak is detected within the second frequency range FW2 including the redetected specific peak x at the timing TD -3 Then, the peak re-detection unit 24 detects a peak at the timing TD -3 The timing before TD -4 For the frequency spectrum of -3 In the example shown in FIG. 5, the redetected specific peak is detected within the second frequency range FW2 including the redetected specific peak x at the timing TD -4 The timing prediction unit 25 detects a specific peak ● at timing TD0 and a specific peak ● at timing TD -1 Redetection of specific peaks at × and timing TD -2 The replacement time is predicted based on the re-detected specific peak x.

[0049] For example, the timing prediction unit 25 obtains a fitting curve that best fits the center point and amplitude of the sampling period at the specific peak and the center point and amplitude of the sampling period at the re-detected specific peak in a coordinate space with time and amplitude as coordinate axes, and determines the time at which the obtained fitting curve intersects with a predetermined judgment amplitude value for determining the replacement time as the replacement time. In the example shown in Figure 5, the timing prediction unit 25 obtains the timing TD0 at the specific peak ●, its amplitude, and timing TD -1 Redetection of a specific peak at the timing TD -1 and its amplitude and timing TD -2 Redetection of a specific peak at the timing TD -2 A fitting curve α that best fits the amplitude is determined, and the time ESα at which the fitting curve α intersects with a predetermined judgment amplitude value Th for determining the time for replacement is determined as the time for replacement ESα.

[0050] Timing TD after timing TD0 +1 Then, the peak detection unit 23 detects the timing TD +1 For the frequency spectrum of the timing TD0, within the second frequency range FW2 including the specific peak ● at the timing TD +1 A specific peak at, for example, the timing TD is detected and indicated by an x. +1 The timing predictor 25 then detects the specific peaks ● at timing TD0 and TD +1 Specific peaks at × and each timing TD -1 , T.D. -2 The replacement time is predicted based on each re-detected specific peak x.

[0051] Alternatively, for example, when the peak detector 23 detects a specific peak indicated by a black circle at the timing TD0, the peak re-detector 24 detects a specific peak indicated by a black circle at the timing TD1 immediately before the timing TD0. -1For the frequency spectrum, a specific peak is detected again within the second frequency range FW2 including the specific peak ●, and the specific peak is detected at timing TD -1 Then, the peak re-detection unit 24 detects the specific peak at the timing TD -1 The timing before TD -2 For the frequency spectrum of -1 The re-detected specific peak is detected within the second frequency range FW2 including the re-detected specific peak ○ at the timing TD -2 Then, the peak re-detection unit 24 detects the specific peak at the timing TD -2 The timing before TD -3 For the frequency spectrum of -2 The re-detected specific peak is detected within the second frequency range FW2 including the re-detected specific peak ○ at the timing TD -3 Then, the peak re-detection unit 24 detects the specific peak at the timing TD -3 The timing before TD -4 For the frequency spectrum of -3 The re-detected specific peak is detected within the second frequency range FW2 including the re-detected specific peak ○ at the timing TD -4 The re-detection specific peak at timing TD0 is detected. Then, the timing prediction unit 25 detects the specific peak ● at timing TD -1 Redetect specific peaks at × and each timing TD -1 , T.D. -2 , T.D. -3 , T.D. -4 The timing for replacement is predicted based on each re-detected specific peak x at each time. For example, in the same manner as described above, the timing prediction unit 25 predicts the timing TD0 and its amplitude at the specific peak ●, and each timing TD -1 , T.D. -2 , T.D. -3 , T.D. -4 Each re-detected specific peak ○ at each timing TD -1 , T.D. -2 , T.D. -3 , T.D.-4 A fitting curve β that best fits each of these amplitudes is determined, and the time ESβ at which this fitting curve β intersects with the threshold amplitude value Th is determined as the replacement time ESβ.

[0052] Timing TD after timing TD0 +1 Then, the peak detection unit 23 detects the timing TD +1 For the frequency spectrum of the timing TD0, within the second frequency range FW2 including the specific peak ● at the timing TD +1 A specific peak is detected at the timing TD, for example, indicated by a circle. +1 The timing predictor 25 then detects the specific peaks ● at timing TD0 and TD +1 Specific peaks ○ and each timing TD -1 , T.D. -2 , T.D. -3 , T.D. -4 The replacement time is predicted based on each re-detected specific peak ○.

[0053] Timing TD after timing TD0 +2 Then, the peak detection unit 23 detects the timing TD +2 For the frequency spectrum of +1 Within the second frequency range FW2 including the specific peak ○ at the timing TD +2 A specific peak is detected at the timing TD, for example, indicated by a circle. +2 The timing prediction unit 25 then detects the specific peak ● at the timing TD0, the specific peak ● at each timing TD +1 , T.D. +2 Each specific peak ○ and each timing TD -1 , T.D. -2 , T.D. -3 , T.D. -4 The replacement time is predicted based on each re-detected specific peak ○.

[0054] Similarly, timing TD after timing TD0 +3Then, the peak detection unit 23 detects, at the timing TD +3 a specific peak ○ within the second frequency range FW2 including the specific peak at the timing TD +2 in the frequency spectrum at the timing TD +3 and detects the specific peak at the timing TD +3 for example, indicated by ○. Then, the timing prediction unit 25 uses the specific peak ● at the timing TD0, each timing TD +1 , TD +2 , TD +3 each specific peak ○ at each timing TD -1 , TD -2 , TD -3 , TD -4 and each re-detection specific peak ○ at each of these timings TD to predict the replacement timing.

[0055] In the example shown in FIG. 5, in the case of the specific peak ● and the re-detection specific peak ×, the progress of the failure is relatively fast, the remaining life from the timing TD0 is relatively short, and the replacement timing ESα reflecting the progress of this failure (change trend over time, trend) is predicted. On the other hand, in the case of the specific peak ● and the re-detection specific peak ○, the progress of the failure is relatively slow, the remaining life from the timing TD0 is relatively long, and the replacement timing ESβ reflecting the progress of this failure is predicted (ESα < ESβ). Thus, even if the timing TD0 at which the specific peak ● was detected is the same, if the progress of the failure is different, the remaining life is different, and therefore the replacement timing is also different. The rolling bearing replacement timing prediction device 1000 in the embodiment can predict the replacement timing in consideration of such a progress of the failure.

[0056] These control processing unit 2, input unit 3, output unit 4, IF unit 5, and storage unit 6 can be configured by, for example, a computer such as a desktop type or a notebook type.

[0057] Next, the operation of this embodiment will be described. Fig. 6 is a flowchart showing the operation of the rolling bearing replacement timing prediction device before the specific peak is detected. Fig. 7 is a flowchart showing the operation of the rolling bearing replacement timing prediction device after the specific peak is detected.

[0058] When the power is turned on, the rolling bearing replacement timing prediction device 1000 configured as described above initializes the necessary parts and starts operation. By executing the control processing program, the control processing unit 2 is functionally configured to include a control unit 21, a spectrum processing unit 22, a peak detection unit 23, a peak re-detection unit 24, and a timing prediction unit 25.

[0059] When the operation starts and the timing for starting the sampling period arrives, as shown in FIG. 6, the rolling bearing replacement time prediction device 1000 first acquires from the memory unit 6 the vibration data for the length TW of the sampling period acquired by the first to third vibration detection sensors 1-1 to 1-3 using the spectrum processing unit 22 of the control processing unit 2, calculates the frequency spectrum of the vibration data for the current sampling period (S11), and stores this calculated frequency spectrum in the memory unit 6 in association with the current sampling period (S12).

[0060] Next, the rolling bearing replacement timing prediction device 1000 detects a specific peak within a predetermined first frequency range including the theoretical frequency from the frequency spectrum for the current sampling period using the peak detection unit 23 of the control processing unit 2 (S13). If this detection results in no specific peak being detected (NO (absent)), the rolling bearing replacement timing prediction device 1000 ends this processing for the current sampling period. On the other hand, if the above determination results in a specific peak being detected (YES (present)), the rolling bearing replacement timing prediction device 1000 then executes processing S14.

[0061] In this process S14, the rolling bearing replacement time predicting device 1000 causes the peak detecting section 23 of the control processing section 2 to store the detected specific peak (its sampling period, amplitude and frequency) in the storage section 6.

[0062] Next, the rolling bearing replacement time prediction device 1000 detects the specific peak as a re-detected specific peak using the peak re-detection unit 24 of the control processing unit 2, and stores this detected re-detected specific peak (its sampling period, amplitude and frequency) in the memory unit 6 (S15).

[0063] Next, the rolling bearing replacement timing predicting device 1000 predicts the replacement timing of the rolling bearing BE by the timing predicting section 25 of the control processing section 2 based on the specific peak and the redetected specific peak (S16).

[0064] Then, the rolling bearing replacement time prediction device 1000 causes the timing prediction unit 25 to output this predicted replacement time to the output unit 4 (S17), and ends this processing for the current sampling period. Note that the replacement time may be output to an external device via the IF unit 5 as necessary.

[0065] On the other hand, when the sampling period begins after the specific peak has been detected, in Figure 7, the rolling bearing replacement time prediction device 1000 first uses the spectrum processing unit 22 of the control processing unit 2 to determine the frequency spectrum of the vibration data for the current sampling period (S21), as in the above-mentioned process S11, and then stores this determined frequency spectrum in the memory unit 6 in association with the current sampling period (S22), as in the above-mentioned process S12.

[0066] Next, the rolling bearing replacement time prediction device 1000 detects a specific peak in the current sampling period using the peak detection unit 23 of the control processing unit 2, and stores this detected specific peak (its sampling period and amplitude) in the memory unit 6 (S23).

[0067] Next, the rolling bearing replacement timing prediction device 1000 causes the timing prediction unit 25 of the control processing unit 2 to predict the replacement timing of the rolling bearing BE based on the specific peaks detected so far and the redetected specific peak (S24).

[0068] Then, the rolling bearing replacement time prediction device 1000 causes the timing prediction unit 25 to output this predicted replacement time to the output unit 4 (S25), similar to the above-mentioned process S17, and ends this process for the current sampling period.

[0069] As described above, the rolling bearing replacement time prediction device 1000 in the embodiment and the rolling bearing replacement time prediction method implemented therein detects specific peaks in a first frequency range that includes the logical frequency, thereby enabling more reliable detection of specific peaks, and re-detects specific peaks by narrowing down to a second frequency range that is narrower than the first frequency range, thereby excluding from the detection target confusing peaks that are included in the first frequency range but are not caused by the occurrence of the fault, enabling more accurate detection of specific peaks caused by the occurrence of the fault.

[0070] After identification, the remaining lifespan ((end point of the sampling period in the frequency spectrum (vibration data) when the specific peak was detected) + (remaining lifespan at that time) = (replacement time)), which is the time until the rolling bearing needs to be replaced, depends on how the fault progresses (tendency of change over time, trend). The rolling bearing replacement time prediction device 1000 and rolling bearing replacement time prediction method detect not only the specific peak but also the re-detected specific peak that precedes it, and predicts the replacement time based on the sampling period when the specific peak was detected and the amplitude of the detected specific peak, and the sampling period when the re-detected specific peak was detected and the amplitude of the detected re-detected specific peak. Since the replacement time is predicted taking into account how the fault progresses, it can predict the replacement time more accurately. Furthermore, when predicting the replacement time using multiple re-detected specific peaks, the rolling bearing replacement time prediction device 1000 and rolling bearing replacement time prediction method can predict the replacement time more accurately than when predicting the replacement time using a single re-detected specific peak.

[0071] The rolling bearing replacement time prediction device 1000 and the rolling bearing replacement time prediction method detect a redetected specific peak in a second frequency range that includes the redetected specific peak detected in the previous peak redetection process, and therefore can more reliably detect the redetected specific peak even if the frequency of the redetected specific peak changes over time.

[0072] The rolling bearing replacement time prediction device 1000 and rolling bearing replacement time prediction method detect a specific peak in a second frequency range that includes the specific peak detected in the previous peak detection process, and therefore can more reliably detect the specific peak even if the frequency of the specific peak changes over time.The rolling bearing replacement time prediction device 1000 and rolling bearing replacement time prediction method predict the replacement time using a specific peak in a sampling period after the sampling period in which the specific peak was detected, and can therefore more accurately predict the replacement time.

[0073] In the above-described embodiment, a data logger 7 may be provided between the vibration detection sensor 1 and the control processing unit 2, and the spectrum processing unit 22 may be provided in this data logger 7.

[0074] 8 is a block diagram showing the configuration of a rolling bearing replacement timing prediction device according to a modified example of the embodiment. A rolling bearing replacement timing prediction device 1000a according to this modified example includes, for example, a vibration detection sensor 1, a data logger 7, a control processing unit 2, an input unit 3, an output unit 4, an IF unit 5, and a storage unit 6, as shown in FIG.

[0075] The vibration detection sensor 1, control processing unit 2, input unit 3, output unit 4, IF unit 5 and memory unit 6 in rolling bearing replacement time prediction device 1000a are similar to the vibration detection sensor 1, control processing unit 2, input unit 3, output unit 4, IF unit 5 and memory unit 6 in rolling bearing replacement time prediction device 1000, except that some functions of control processing unit 2 and some functions of memory unit 6 have been moved to data logger 7, and therefore their description will be omitted. The control processing unit 2 functionally comprises a control unit 21, a peak detection unit 23, a peak re-detection unit 24 and a timing prediction unit 25, and the spectrum processing function has been moved to data logger 7. The function of storing vibration data and its frequency spectrum has been moved to data logger 7.

[0076] The data logger 7 is connected to both the vibration detection sensor 1 and the control processing unit 2, and is configured with, for example, a computer, and includes a data logger control processing unit 71 and a data logger storage unit 72. The data logger control processing unit 71 is configured with a CPU and its peripheral circuits, and functionally includes a spectrum processing unit 22 that functions in the same manner as described above, except that data is stored in the data logger storage unit 72 instead of the storage unit 6. The data logger storage unit 72 is configured with a ROM, The data logger 7 is provided with an EEPROM, RAM, a hard disk drive, etc., and stores the vibration data and frequency spectrum. In response to a request from the control processing unit 2, the data logger 7 outputs the frequency spectrum of the vibration data to the control processing unit 2.

[0077] In order to express the present invention, the present invention has been properly and sufficiently described above through the embodiments with reference to the drawings, but it should be recognized that those skilled in the art can easily change and / or improve the above-mentioned embodiments. Therefore, unless the changes or improvements made by those skilled in the art are at a level that causes departure from the scope of the claims described in the claims, such changes or improvements are interpreted as being included in the scope of the claims. [Explanation of symbols]

[0078] BE (BE-1 to BE-3) Rolling bearings (1st to 3rd rolling bearings) FW1 First frequency range FW2 Second frequency range 1000 Rolling bearing replacement timing prediction device 1 (1-1 to 1-3) Vibration detection sensor (first to third vibration detection sensor) 2. Control processing section 6 Memory section 7 Data Logger 21 Control Unit 22 Spectral Processing Section 23 Peak detector 24 Peak redetection section 25 Timing Prediction Department

Claims

1. a spectrum processing step of acquiring vibration data representing vibrations occurring in the rolling bearing at predetermined acquisition intervals for a predetermined sampling period, determining a frequency spectrum of the vibration data for the sampling period, and storing the determined frequency spectrum in a storage unit in association with the sampling period; a peak detection step of detecting, from the frequency spectrum, a specific peak that does not appear when the rolling bearing is normal, within a predetermined first frequency range that includes a theoretical frequency that causes a peak on the frequency spectrum when an abnormality occurs; a peak re-detection step of, when a specific peak is detected in the peak detection step, re-detecting a specific peak that does not appear when the rolling bearing is normal, within a predetermined second frequency range narrower than the first frequency range and that includes the frequency of the specific peak detected in the peak detection step, for one or more frequency spectra stored in the storage unit prior to the sampling period when the specific peak is detected; a replacement time prediction step for predicting the replacement time of the rolling bearing based on a sampling period when a specific peak is detected in the peak detection step and an amplitude of the detected specific peak, and a sampling period when a re-detected specific peak is detected in the peak re-detection step and an amplitude of the re-detected specific peak, A method for predicting when to replace rolling bearings.

2. In the peak re-detection step, when the re-detected specific peak is detected in a past direction for the plurality of frequency spectra, the peak detection step and the specific peak detected in the peak detection step are regarded as the first peak re-detection step and the first re-detected specific peak, respectively, for each of the plurality of frequency spectra, and the re-detected specific peak is detected in the second frequency range including the re-detected specific peak detected in the previous peak re-detection step for the frequency spectrum. The method for predicting the replacement time of a rolling bearing according to claim 1.

3. The peak detection step further includes detecting, for each sampling period after the sampling period in which the specific peak is detected, the specific peak in the second frequency range including the specific peak detected in the peak detection step immediately before the sampling period in question, for the frequency spectrum stored in the storage unit in the sampling period in question. The method for predicting the replacement time of a rolling bearing according to claim 1.

4. The length of the sampling period is set based on the second frequency range; The narrower the second frequency range, the longer the sampling period is set. The method for predicting the replacement time of a rolling bearing according to claim 1.

5. a vibration detection sensor that acquires vibration data representing vibrations occurring in the rolling bearing; A memory unit; a control processing unit that performs spectrum processing, peak detection processing, peak re-detection processing, and replacement time prediction processing; the spectrum processing is a process of acquiring vibration data for a predetermined sampling period at predetermined acquisition intervals, determining a frequency spectrum of the vibration data for the sampling period, and storing the determined frequency spectrum in the storage unit in association with the sampling period; the peak detection process is a process of detecting, from the frequency spectrum, a specific peak that does not appear when the rolling bearing is normal, within a predetermined first frequency range that includes a theoretical frequency that causes a peak on the frequency spectrum when an abnormality occurs, the peak re-detection process is a process for detecting, when a specific peak is detected in the peak detection process, a specific peak that does not appear when the rolling bearing is normal, as a re-detected specific peak, within a predetermined second frequency range narrower than the first frequency range and including the frequency of the specific peak detected in the peak detection process, for one or more frequency spectra stored in the storage unit prior to the sampling period when the specific peak is detected; the replacement time prediction process is a process for predicting the replacement time of the rolling bearing based on a sampling period when a specific peak is detected in the peak detection process and an amplitude of the detected specific peak, and a sampling period when a re-detected specific peak is detected in the peak re-detection process and an amplitude of the re-detected specific peak. Rolling bearing replacement timing prediction device.

Images