A method for identifying a sliding bearing body fault and a misaligned gap fault

Abstract

The application discloses a kind of sliding bearing body fault and gap inappropriate fault identification method, the method is first to the vibration acceleration instantaneous signal of sliding bearing and save sampling;After Fourier spectrum analysis is carried out to vibration acceleration instantaneous signal, and according to whether the spectrum peak in Fourier spectrum near 2 times frequency, 3 times frequency exists to judge whether sliding bearing exists body fault or gap inappropriate;If there is, further to the vibration acceleration instantaneous signal of double spectrum analysis is carried out, and according to the coupling of high frequency component in double spectrum with the frequency component of rotation to the body fault and gap inappropriate of sliding bearing is identified.The application is identified to the body fault and gap inappropriate of sliding bearing by double spectrum analysis to vibration signal, according to the coupling of high frequency component in double spectrum with the frequency component of rotation to the body fault and gap inappropriate of sliding bearing is identified, can quickly, effectively identify sliding bearing body fault and gap inappropriate fault, greatly improve bearing repair work efficiency.

Description

Technical Field

[0001] This invention relates to a method for identifying faults and improper clearances in sliding bearings by analyzing their vibration signals, belonging to the field of diagnostic technology. Background Technology

[0002] Sliding bearings are key components of large electric motors, bearing the crucial task of supporting the rotor and achieving the coordination between static and rotating states. Because they directly involve rotation and the resulting mechanical wear, coupled with harsh operating environments, poor installation, or inadequate lubrication, sliding bearing failures are difficult to avoid. Therefore, bearing fault diagnosis is an essential procedure.

[0003] Vibration signal analysis-based bearing fault diagnosis is currently the most widely used diagnostic method. This method is based on the premise that bearing faults cause characteristic vibrations in the bearing system, intensifying these vibrations and manifesting them as components at certain characteristic frequencies in the vibration signal; moreover, different types of bearing faults result in different frequencies of characteristic vibrations. Therefore, this method can be implemented by monitoring the time-domain parameters of the vibration signal and analyzing its spectrum. The time-domain parameters of the vibration signal mainly include peak-to-peak vibration acceleration and effective vibration velocity. The occurrence of a bearing fault will cause these parameters to increase, which can be used to roughly determine whether a bearing fault has occurred. Further analysis of the vibration signal spectrum allows for more precise fault diagnosis, such as determining the fault type.

[0004] Due to the prevalent and complex background noise interference in engineering sites, as well as the weak and nonlinear coupling characteristics of bearing fault features, bearing fault diagnosis methods based on vibration signal analysis still face severe challenges, and many researchers are continuously working to improve them. For example, bispectral analysis is introduced to suppress noise and detect the phase coupling of nonlinear vibration signals, demodulating the bearing characteristic signals modulated at high frequencies, thereby significantly improving the performance of bearing fault diagnosis.

[0005] Body failure is a common type of fault in sliding bearings. This type of failure is caused by wear, fatigue, corrosion, etc., resulting in cracks, scratches, or foreign object embedding on the bearing body. Improper clearance is another common type of fault in sliding bearings. This refers to improper installation causing excessive or insufficient top clearance between the journal and bearing bush, as well as uneven clearance between the left and right sides, leading to a certain degree of imbalance. Both types of faults will induce characteristic vibrations at rotational frequencies. Therefore, these two types of faults can be diagnosed through spectral analysis of the vibration signals. In fact, spectral analysis of vibration signals can be used to diagnose these two types of sliding bearing faults.

[0006] Identifying both body faults and improper clearance in sliding bearings can greatly facilitate on-site maintenance. For example, if a body fault is confirmed, the bearing can be disassembled, repaired, and reassembled; if improper clearance is confirmed, only clearance adjustment is needed without disassembling the bearing. Clearly, this significantly improves the efficiency of bearing maintenance and has practical engineering value. However, since both types of faults in sliding bearings induce characteristic vibrations at rotational frequencies, current fault diagnosis methods based on vibration signal characteristic analysis cannot identify them. Therefore, effectively identifying body faults and improper clearance in sliding bearings has become a challenge for relevant personnel. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by providing a method for identifying sliding bearing body faults and improper clearance faults, so as to quickly and effectively identify sliding bearing body faults and improper clearance faults, thereby improving the efficiency and accuracy of bearing maintenance.

[0008] The problem described in this invention is solved by the following technical solution:

[0009] A method for identifying body faults and improper clearance faults in sliding bearings is disclosed. The method first samples and saves the instantaneous vibration acceleration signal of the sliding bearing; then, it performs Fourier spectrum analysis on the instantaneous vibration acceleration signal, and determines whether the sliding bearing has a body fault or improper clearance based on the presence of spectral peaks near the rotational frequency and its second and third harmonics in the Fourier spectrum; if present, it further performs bispectral analysis on the instantaneous vibration acceleration signal, and identifies the body fault and improper clearance of the sliding bearing based on the coupling between the rotational frequency component and the high-frequency component in the bispectrum.

[0010] The above-mentioned method for identifying sliding bearing body faults and improper clearance faults includes the following steps:

[0011] a. Using a sampling frequency of 10kHz, measure the instantaneous vibration acceleration signal of the sliding bearing for 10s, denoted as A(t), where t represents time;

[0012] b. Calculate the rotational frequency f of the sliding bearing. r = n / 60, where n is the current rotational speed of the sliding bearing, in r / min;

[0013] c. Calculate the Fourier transform of A(t), denoted as X(f), where f represents the frequency, and its value is 0, Δf, 2Δf, 3Δf, ..., Nyquist frequency (here, Δf represents the frequency resolution, which is the reciprocal of the sampling duration when the sampling frequency is fixed, and is 0.1Hz in this invention; the Nyquist frequency is half of the sampling frequency, and is 5kHz in this invention);

[0014] d. Calculate the Fourier power spectrum P(f):

[0015] P(f)=E[X(f)X * (f)]

[0016] Where E represents the expected value; X * (f) denotes the complex conjugate of X(f);

[0017] e. Examine the frequency shift f in the P(f) spectrum. r Check for spectral peaks near its 2nd and 3rd harmonics, and determine whether the sliding bearing has a structural fault or improper clearance using the following methods:

[0018] If the frequency f in the P(f) spectrum r If spectral peaks are found near the second and third harmonics, it is determined that the sliding bearing has a structural fault or improper clearance, and subsequent steps are taken to identify the two; otherwise, it is determined that the sliding bearing does not have a structural fault or improper clearance.

[0019] f. Calculate the bispectral B(f1,f2) of A(t):

[0020] B(f1,f2)=E[X(f1)X(f2)X * (f1+f2)]

[0021] Where f1 and f2 represent two frequencies, both of which take values ​​of 0, Δf, 2Δf, 3Δf, ..., Nyquist frequencies. The bispectrum characterizes the degree of correlation between the two frequency components f1 and f2 of A(t). That is, if there is coupling between the two frequency components f1 and f2, the degree of correlation is high, and the bispectrum will have a non-zero value. The higher the amplitude of the two frequency components f1 and f2 and the tighter the coupling, the larger the bispectral value will be, and obvious spectral peaks will appear in the bispectral graph of the two frequency coordinates.

[0022] g. Examine the B(f1,f2) spectrum for the presence of spectral peaks corresponding to the coupling between the frequency conversion component and the high-frequency component in the [2, 5] kHz band, and then identify the main body fault or improper gap as follows:

[0023] If the B(f1,f2) spectrum contains a series of spectral peaks corresponding to the coupling between the frequency conversion component and the high-frequency component in the [2,5]kHz frequency band, and the corresponding pseudo power, i.e., the value of B(f1,f2), is greater than the predetermined pseudo power threshold, then the fault is judged to be a body fault; if the corresponding pseudo power is less than or equal to the predetermined pseudo power threshold, then the gap is judged to be improper.

[0024] The above-mentioned method for identifying sliding bearing body faults and improper clearance faults involves calculating the bispectral B(f1,f2) of A(t) by limiting f1 to [fr -5,f r The range is +5]Hz, while f2 is limited to the range [2, 5]kHz.

[0025] The above-mentioned method for identifying sliding bearing body faults and improper clearance faults involves sampling the instantaneous vibration acceleration signal of the sliding bearing at a sampling frequency of 10kHz and a sampling duration of 10s.

[0026] In the above-mentioned method for identifying sliding bearing body faults and improper clearance faults, the pseudo-power threshold is set to 4 times the maximum value of the bispectral value of the vibration acceleration signal of the sliding bearing under healthy conditions.

[0027] In the above-mentioned method for identifying sliding bearing body faults and improper clearance faults, the pseudo-power threshold is set to 1.5 × 10⁻⁶. 10 .

[0028] This invention uses bispectral analysis of vibration signals to identify structural faults and improper clearances in sliding bearings based on the coupling between the rotational frequency and high-frequency components in the bispectral analysis. This allows for rapid and effective identification of structural faults and improper clearances in sliding bearings, significantly improving the efficiency of bearing maintenance. A specific test analysis example is provided in the embodiments, namely, testing a YKKK1000-8W induced draft fan motor. The test results show that the method of this invention exhibits significant and easily identifiable differences, indicating that identifying structural faults and improper clearances in sliding bearings based on bispectral analysis is entirely feasible. Attached Figure Description

[0029] The invention will now be described in further detail with reference to the accompanying drawings.

[0030] Figure 1 This is a flowchart of the present invention;

[0031] Figure 2 The time-domain waveform of the horizontal vibration acceleration signal when the sliding bearing has a body fault;

[0032] Figure 3 The Fourier spectrum of the horizontal vibration acceleration signal when the sliding bearing has a body fault;

[0033] Figure 4 The bispectral waveform of the horizontal vibration acceleration signal when the sliding bearing has a body fault;

[0034] Figure 5 A photograph of a sliding bearing body failure (bearing shell wear);

[0035] Figure 6 The time-domain waveform of the horizontal vibration acceleration signal when there is an improper clearance in the sliding bearing;

[0036] Figure 7 The Fourier spectrum of the horizontal vibration acceleration signal when there is an improper clearance in the sliding bearing;

[0037] Figure 8 The bispectral representation of the horizontal vibration acceleration signal when there is an improper clearance in the sliding bearing;

[0038] Figures 9A-9B These are the layout diagrams of the measuring points in the horizontal and axial directions (X in the diagram represents the fixed position of the sensor).

[0039] The symbols in the text are as follows: A(t) is the instantaneous signal of the vibration acceleration of the sliding bearing, and t represents time; f r Let be the rotational frequency of the sliding bearing; n be the current rotational speed of the sliding bearing, in r / min; X(f) is the Fourier transform of A(t), where f represents the frequency; Δf represents the frequency resolution; P(f) is the Fourier power spectrum; E represents the expected value; X * (f) represents the conjugate complex number of X(f); B(f1,f2) is the bispectral representation of the vibration acceleration signal, where f1 and f2 represent two frequencies. Detailed Implementation

[0040] This invention targets sliding bearings. Based on the Fourier spectrum of the vibration signal (this invention uses the vibration acceleration signal), and by performing bispectral analysis, it examines the rotational frequency component and its coupling with the high-frequency component, thereby providing a method to identify sliding bearing body faults and improper clearances.

[0041] During testing, vibration acceleration signals are measured, including horizontal, vertical, and axial vibration acceleration signals. This invention uses a Lans LC0104T piezoelectric vibration acceleration signal sensor to measure the vibration acceleration signal. This type of sensor is widely used in industrial settings. Regarding the selection of measurement points, the selection scheme for pedestal bearings is given in "GB 10068—2008 Mechanical Vibration Measurement, Evaluation and Limitation of Motors with Shaft Center Height of 56mm and Above". If the vibration acceleration signal sensor is arranged on a horizontal measurement point, that is, the measuring surface of the vibration acceleration signal sensor is in contact with the horizontal measurement point, then the measured vibration acceleration signal is in the horizontal direction; if the vibration acceleration signal sensor is arranged on an axial measurement point, that is, the measuring surface of the vibration acceleration signal sensor is in contact with the axial measurement point, then the measured vibration acceleration signal is in the axial direction.

[0042] When a sliding bearing experiences a failure due to cracks, scratches, or embedded foreign objects, it will induce characteristic vibrations at the rotational frequency. Furthermore, with each rotor revolution, the journal impacts the failure point, and this impact signal covers a wide frequency range (from low frequencies to high frequencies of several kHz), inevitably covering the natural frequency of a component in the shaft system, thus triggering resonance. This will extend the bearing vibration to higher frequencies (several kHz). Therefore, the vibration characteristics of a sliding bearing failure manifest in two ways: first, it induces characteristic vibrations at the rotational frequency; second, the impact induces resonance, extending the vibration to higher frequencies, essentially representing the coupling of the rotational frequency component of the vibration with certain high-frequency (several kHz) components.

[0043] In contrast, while improper clearance will also cause characteristic vibrations at the rotational frequency, it will not result in impact or resonance. Therefore, the vibration characteristics of improper clearance in sliding bearings will not extend to higher frequencies; that is, there will be no coupling between the rotational frequency component and the high-frequency component.

[0044] Clearly, the presence of coupling between the rotational frequency component and the high-frequency component can be used as a criterion for identifying faults in the sliding bearing body and improper clearance.

[0045] Bispectral analysis, as a higher-order spectral analysis, can detect the coupling, Gaussianity, and nonlinearity of signals, which is beyond the reach of traditional Fourier spectral analysis.

[0046] For a vibration signal A(t) (t represents time), assuming its Fourier transform is X(f) (f represents frequency), then the Fourier power spectrum of A(t) is:

[0047] P(f)=E[X(f)X * (f)] (1)

[0048] Where P(f) actually represents the pseudopower of the component with frequency f in the vibration signal A(t); E represents the expected value; X * (f) denotes the conjugate complex number of X(f). The bispectral representation of A(t) is defined as...

[0049] B(f1,f2)=E[X(f1)X(f2)X * (f1+f2)] (2)

[0050] Clearly, the bispectrum characterizes the correlation and coupling degree between the two frequency components of A(t)—component f1 and component f2. It should be noted that when calculating the bispectrum, the values ​​of f1 and f2 are 0, Δf, 2Δf, 3Δf, ..., the Nyquist frequency. Therefore, the bispectrum of A(t) actually characterizes the correlation and coupling degree between all pairs of frequency components of A(t). For example, if f1 is Δf and f2 is 2Δf, then the corresponding bispectrum B(f1, f2) represents the correlation and coupling degree between the Δf frequency component and the 2Δf frequency component of A(t). The pseudo-power in this invention is merely a representation of the degree of "correlation" and has no unit.

[0051] Regarding the vibration signal of a sliding bearing in a large motor, if there is correlation or coupling between certain frequency components, the bispectral values ​​will be non-zero. Moreover, the larger the amplitude of these frequency components and the tighter the correlation or coupling, the larger the bispectral values ​​will be, and they will be reflected as obvious spectral peaks in the bispectral graph.

[0052] Based on the vibration characteristics of sliding bearing failures, it can be seen that: for a sliding bearing body failure, a spectral peak will appear in the bispectral diagram of its vibration signal, and this spectral peak corresponds to the coupling of the rotational frequency component and a certain high-frequency component; while for improper clearance, a spectral peak corresponding to the coupling of the rotational frequency component and a certain high-frequency component will not appear in the bispectral diagram of its vibration signal.

[0053] Based on this, faults in the sliding bearing body and improper clearance can be identified.

[0054] This invention first samples and saves the instantaneous vibration acceleration signal of the sliding bearing (sampling frequency 10kHz, sampling duration 10s). Then, it performs Fourier spectrum analysis on the instantaneous vibration acceleration signal and examines whether there are spectral peaks near the rotational frequency and its second and third harmonics in the Fourier spectrum, thereby determining whether there is a fault in the sliding bearing itself or improper clearance. If so, it further performs bispectral analysis on the instantaneous vibration acceleration signal and examines the coupling between the rotational frequency component and the high-frequency component in the bispectrum, thereby identifying the fault in the sliding bearing itself and improper clearance. If there is a series of spectral peaks in the bispectrum corresponding to the coupling between the rotational frequency component and the high-frequency component, it indicates that there is a large-scale coupling between the rotational frequency component and the high-frequency component, and the coupling is strong, so the fault can be determined to be a fault in the bearing itself; otherwise, it is improper clearance.

[0055] The above-mentioned method for identifying sliding bearing body faults and improper clearances mainly includes the following steps:

[0056] a. The instantaneous vibration acceleration signal of the sliding bearing is measured at a sampling frequency of 10kHz for a duration of 10s and denoted as A(t) (t represents time and can be measured using a conventional vibration acceleration measuring device);

[0057] b. Calculate the rotational frequency f r =n / 60, where n is the current rotational speed (in r / min, which can be measured using a conventional rotational speed measuring device);

[0058] c. Calculate the Fourier transform of A(t), denoted as X(f);

[0059] d. Calculate the Fourier power spectrum P(f) according to equation (1);

[0060] e. Examine the frequency shift f in the P(f) spectrum. r The presence of spectral peaks near the second and third harmonics of the sliding bearing can be used to determine whether there is a fault in the bearing body or improper clearance.

[0061] If the frequency f in the P(f) spectrum r If spectral peaks are present near the second and third harmonics, it can be determined that the sliding bearing has a structural fault or improper clearance, and subsequent steps can be taken to identify these two issues. Otherwise, it can be determined that the sliding bearing does not have a structural fault or improper clearance.

[0062] f. Calculate the bispectral B(f1,f2) of the vibration acceleration signal according to equation (2);

[0063] It should be noted that calculating the bispectral density of vibration acceleration signals across the entire frequency range from 0 to the Nyquist frequency would involve an extremely high computational load. The Nyquist frequency is half the sampling frequency, which is 5 kHz in this invention. Therefore, based on the vibration characteristics of sliding bearing failures and the purpose of identifying inherent faults and improper clearances, f1 is limited to [f r -5,f r The range is +5]Hz, while f2 is limited to the range [2, 5]kHz.

[0064] g. Examine the B(f1,f2) spectrum to see if there are spectral peaks corresponding to the coupling between the frequency conversion component and the high-frequency component in the [2,5]kHz frequency band, so as to identify the main body fault or improper gap.

[0065] If the B(f1,f2) spectrum contains a series of spectral peaks corresponding to the coupling between the frequency conversion component and the high-frequency component in the [2,5] kHz band, it indicates that the frequency conversion component and the high-frequency component in the [2,5] kHz band have a large-scale coupling, and this coupling is strong (the corresponding pseudo-power is greater than a predetermined pseudo-power threshold, which is 1.5 × 10 in this invention). 10 If the frequency conversion component has only a very small coupling range with the high-frequency component in the [2, 5] kHz frequency band, and it is a weak coupling (the corresponding pseudo power is less than or equal to the predetermined pseudo power threshold), it can be judged as an improper gap. Regarding the predetermined pseudo power threshold (1.5 × 10⁻⁶ in this invention),...10 The value was determined by measuring and examining the maximum value of the bispectral vibration acceleration signal of the sliding bearing under healthy conditions and taking a margin of 3.

[0066] The most significant innovation of this invention is as follows: by performing bispectral analysis on the instantaneous vibration acceleration signal of the sliding bearing, accurate identification of faults in the sliding bearing body and improper clearance can be achieved, thereby effectively improving the efficiency of on-site maintenance work. Based on current literature search results, this is the first such invention.

[0067] The following is a specific example of test analysis:

[0068] A YKKK1000-8W induced draft fan motor was tested. Its rated capacity is 8500kW, rated voltage is 10kV, rated current is 582A, and rated speed is 746r / min. This motor uses sliding bearings.

[0069] This invention adopts Figure 1 The process shown identifies body faults and improper clearances in sliding bearings.

[0070] Figure 2 This is the time-domain waveform of the horizontal vibration acceleration signal of the sliding bearing. The corresponding rotational speed n and rotational frequency are 747.6 r / min and 12.46 Hz, respectively. Figure 3 and Figure 4 These are the corresponding Fourier spectra and bispectral spectra, respectively.

[0071] according to Figure 3 The Fourier spectrum shows that the spectral peaks are located near the rotational frequency and its harmonics, therefore it can be determined that the bearing failure may be due to a fault in the bearing itself or improper clearance. Figure 4 The bispectral analysis reveals a wide-ranging coupling between the rotational frequency (12.46 Hz) component of the vibrational acceleration and the components within the [2.54.5] kHz frequency band, with the strongest coupling occurring at approximately 3.2 kHz (corresponding to a pseudo-power of approximately 10 × 10⁻⁶). 10 Therefore, it was determined that the bearing failure was due to a fundamental structural fault. After disassembly and reassembly of the sliding bearing, significant wear and metal fragments were found on the bearing bush. Figure 5 As shown.

[0072] Subsequently, on-site technicians performed fine grinding on the worn bearing and completed its reassembly. After the induced draft fan motor was restarted, the instantaneous vibration acceleration signal in the horizontal direction of the sliding bearing was measured again. Figure 6 The waveform in its time domain is given. The corresponding rotational speed n and rotational frequency are 748.2 r / min and 12.47 Hz, respectively. Figure 7 and Figure 8 These are the corresponding Fourier spectra and bispectral spectra, respectively. According to... Figure 7The Fourier spectrum shows that the spectral peaks are located near the rotational frequency and its harmonics, therefore it can be determined that the fault may be a problem with the component itself or improper spacing. However, by comparison... Figure 7 and Figure 3 It can be seen that these spectral peaks are significantly lower than before the fault repair, which verifies the quality of the repair. Furthermore, according to... Figure 8 The bispectral analysis shows that the rotational frequency (12.47 Hz) component of the vibrational acceleration is coupled only to the high-frequency component around 4 kHz, and the coupling is weak (the corresponding pseudo-power is approximately 0.8 × 10⁻⁶). 10 Therefore, it is determined that the bearing failure was due to improper clearance.

[0073] It should be noted that by measuring and examining samples of vibration acceleration signals of sliding bearings under healthy conditions, and performing bispectral analysis, the threshold for determining strong and weak coupling was pre-determined based on the maximum value and a margin of 3. In this invention, it is 1.5 × 10⁻⁶. 10 Strong coupling refers to a pseudo-power greater than the pseudo-power threshold of 1.5 × 10⁻⁶. 10 Weak coupling refers to pseudo-power that is less than or equal to the pseudo-power threshold of 1.5 × 10⁻⁶. 10 .

[0074] Furthermore, in practical engineering, a certain degree of installation error is unavoidable, making it impossible to achieve 100% perfect clearance. In practice, slight clearance discrepancies are acceptable and have minimal impact on the normal operation of the motor. During a subsequent temporary shutdown, the clearance of the non-drive end sliding bearing was measured. Its top clearance was 0.31 mm, while the left and right side clearances were 0.38 mm and 0.41 mm respectively, all within the allowable range.

[0075] To deepen understanding, Figure 4 and Figure 8 Compare them. Figure 4 The bispectral representation of vibration acceleration signal in a sliding bearing under conditions of inherent failure is characterized by a wide-ranging and strong coupling between the rotational frequency component and components within the [2.5 4.5] kHz frequency band (corresponding to a pseudo-power of approximately 10 × 10⁻⁶). 10 ). Figure 8 The bispectral representation of vibration acceleration signal under improper sliding bearing clearance is characterized by the fact that the rotational frequency component is only coupled to a very small range of high-frequency components around 4kHz, and this coupling is weak (the corresponding pseudo-power is only about 0.5 × 10⁻⁶). 10 (This figure represents only 5% of the cases where the bearing body itself is faulty). Clearly, there are significant and easily identifiable differences between the bispectral analyses of the two scenarios: bearing body faults and improper clearance. This demonstrates that identifying bearing body faults and improper clearance based on bispectral analysis is indeed feasible.

Claims

1. A method for identifying sliding bearing body faults and improper clearance faults, characterized in that, The method first samples and saves the instantaneous vibration acceleration signal of the sliding bearing; then, it performs Fourier spectrum analysis on the instantaneous vibration acceleration signal, and judges whether there is a fault in the sliding bearing or improper clearance based on whether there are spectral peaks near the rotational frequency and its second and third harmonics in the Fourier spectrum; if so, it further performs bispectral analysis on the instantaneous vibration acceleration signal, and identifies the fault in the sliding bearing and improper clearance based on the coupling between the rotational frequency component and the high-frequency component in the bispectrum. The method includes the following steps: a. The instantaneous vibration acceleration signal of the sliding bearing for a certain duration is measured at a certain sampling frequency and denoted as A(t), where t represents time; b. Calculate the rotational frequency of the sliding bearing fr = n / 60, where n is the current rotational speed of the sliding bearing in r / min; c. Calculate the Fourier transform of A(t), denoted as X(f), where f represents the frequency; d. Calculate the Fourier power spectrum P(f): P(f)=E[X(f)X * (f)] Where E represents the expected value; X * (f) denotes the complex conjugate of X(f); e. Examine the rotational frequency fr and its 2nd and 3rd harmonics in the P(f) spectrum for spectral peaks, and determine whether the sliding bearing has a body fault or improper clearance in the following manner: If there are spectral peaks near the rotation frequency fr and its second and third harmonics in the P(f) spectrum, it is determined that the sliding bearing has a body fault or improper clearance, and subsequent steps are taken to identify the two; otherwise, it is determined that the sliding bearing does not have a body fault or improper clearance. f. Calculate the bispectral B(f1, f2) of the instantaneous vibration acceleration signal: B(f1, f2) = E[X(f1)X(f2)X... * [(f1+f2)], where f1 and f2 represent two frequencies; f1 is limited to the range of [fr-5, fr+5] Hz, and f2 is limited to the range of [2, 5] kHz; g. Examine the B(f1,f2) spectrum for the presence of spectral peaks corresponding to the coupling between the frequency conversion component and the high-frequency component in the [2,5]kHz band, and then identify the main body fault or improper gap in the following manner: If the B(f1,f2) spectrum contains a series of spectral peaks corresponding to the coupling between the frequency conversion component and the high-frequency component in the [2,5]kHz frequency band, and the corresponding pseudo power is greater than the predetermined pseudo power threshold, then it is judged as a body fault; if the corresponding pseudo power is less than or equal to the predetermined pseudo power threshold, then it is judged as an improper gap.

2. The method for identifying sliding bearing body faults and improper clearance faults according to claim 1, characterized in that, When sampling the instantaneous vibration acceleration signal of the sliding bearing, the sampling frequency is 10kHz and the sampling duration is 10s.

3. The method for identifying sliding bearing body faults and improper clearance faults according to claim 1, characterized in that, The pseudo-power threshold is set to four times the maximum value of the bispectral value of the instantaneous vibration acceleration signal of the sliding bearing under healthy conditions.

4. The method for identifying sliding bearing body faults and improper clearance faults according to claim 1, characterized in that, The pseudo-power threshold is set to 1.5 × 10⁻⁶. 10 .

Images