Rolling bearing fault frequency detection method, device and equipment and readable storage medium

Abstract

The invention provides a rolling bearing fault frequency detection method, device and equipment and a readable storage medium, and belongs to the field of industrial machinery. The method comprises the following steps: collecting a vibration signal of the rolling bearing in real time; performing data preprocessing on the vibration signal to obtain a processed vibration signal; performing feature enhancement on the processed vibration signal through adaptive frequency spectrum amplitude modulation to obtain a correction signal; performing envelope demodulation on the correction signal, and determining a fault frequency peak value according to an envelope demodulation result; according to the fault frequency peak value and the fault frequency characteristic, the fault type of the rolling bearing is determined, and the fault type comprises an outer ring fault, an inner ring fault and a ball fault. The rolling bearing fault diagnosis method and device can achieve the effect of improving the accuracy and efficiency of rolling bearing fault diagnosis.

Description

Technical Field

[0001] The present application relates to the field of industrial machinery, and in particular to a rolling bearing fault frequency detection method, device, equipment and readable storage medium. Background Art

[0002] Rolling bearings, as power transmission components, play a vital role in industrial machinery. However, due to a variety of factors, such as load fluctuations, lubrication failure, material fatigue, and manufacturing defects, rolling bearings can experience operational failures, leading to decreased mechanical equipment performance and potentially safety incidents. Downtime and repair costs associated with rolling bearing operational failures account for approximately 20% to 30% of total maintenance costs in industrial machinery, and rolling bearing failures account for over 40% of all mechanical failures in this sector. Therefore, rolling bearing fault detection plays a leading role in predictive maintenance and industrial safety management in industrial machinery.

[0003] Vibration signals are the most direct source of state observation signals. Due to their ease of acquisition and analysis, related technologies often collect vibration signals from rolling bearings during operation and perform rolling bearing fault diagnosis based on the analysis results. However, in the actual operation of rolling bearings, the operation of rolling bearings is easily affected by various factors such as noise, load, and speed changes, resulting in the collected vibration signals being insufficient to reflect the current operating state of the rolling bearings. Furthermore, in the rolling bearing fault diagnosis process, the data collection, data analysis, and fault diagnosis of vibration signals in related technologies are time-consuming, resulting in poor timeliness in rolling bearing diagnosis. Consequently, related technologies suffer from poor rolling bearing fault diagnosis accuracy and efficiency. Summary of the Invention

[0004] The purpose of this application is to provide a rolling bearing fault frequency detection method, device, equipment and readable storage medium, which can improve the accuracy and efficiency of rolling bearing fault diagnosis.

[0005] The embodiment of the present application is implemented as follows:

[0006] In a first aspect of an embodiment of the present application, a method for detecting rolling bearing fault frequency is provided, the method comprising:

[0007] Real-time collection of rolling bearing vibration signals;

[0008] Performing data preprocessing on the vibration signal to obtain a processed vibration signal;

[0009] The processed vibration signal is enhanced by adaptive spectrum amplitude modulation to obtain a corrected signal;

[0010] Perform envelope demodulation on the correction signal and determine the fault frequency peak value based on the envelope demodulation result;

[0011] The fault type of the rolling bearing is determined based on the fault frequency peak and fault frequency characteristics. The fault types include outer ring fault, inner ring fault and ball fault.

[0012] As a possible implementation method, data preprocessing is performed on the vibration signal to obtain a processed vibration signal, including:

[0013] Determine whether the vibration signal is a full life cycle signal. A full life cycle signal refers to the vibration signal of a rolling bearing going from normal operation to a faulty operation state.

[0014] If so, de-redundancy processing is performed on the vibration signal based on a sparse stacked autoencoder, and wavelet packet denoising is performed on the vibration signal after de-redundancy processing to obtain a processed vibration signal, where the wavelet packet denoising processing includes: wavelet packet decomposition and wavelet packet reconstruction;

[0015] If not, the vibration signal is subjected to wavelet packet noise reduction processing to obtain a processed vibration signal.

[0016] As a possible implementation method, wavelet packet noise reduction processing is performed on the vibration signal to obtain a processed vibration signal, including:

[0017] Perform first-level wavelet packet decomposition on the vibration signal to obtain the high-frequency signal and low-frequency signal corresponding to the vibration signal;

[0018] Performing secondary wavelet packet decomposition on the high-frequency signal and the low-frequency signal to obtain a plurality of sub-band signals, wherein each high-frequency signal and each low-frequency signal corresponds to two sub-band signals;

[0019] Calculate the frequency band energy of each sub-band signal based on the preset wavelet basis function;

[0020] determining a plurality of effective sub-band signals according to the frequency band energy of each sub-band signal;

[0021] The wavelet packet is reconstructed for each effective sub-band signal to obtain the processed vibration signal.

[0022] As a possible implementation method, the processed vibration signal is enhanced by adaptive spectrum amplitude modulation to obtain a corrected signal, including:

[0023] Performing Fourier transform on the processed vibration signal to obtain the phase and amplitude of the processed vibration signal;

[0024] Re-editing the amplitude to obtain a new amplitude with maximum grayscale energy;

[0025] A correction signal is generated based on the phase and the new amplitude.

[0026] As a possible implementation method, the amplitude is re-edited to obtain a new amplitude, including:

[0027] The amplitude is edited based on a preset weight range, and the grayscale value of the edited amplitude is scanned to obtain the grayscale energy corresponding to each preset weight value;

[0028] Determine the maximum grayscale energy from the grayscale energies corresponding to the preset weight values, and use the preset weight value corresponding to the maximum grayscale energy as the target weight value;

[0029] The amplitude corresponding to the target weight value is used as the new amplitude.

[0030] As a possible implementation method, envelope demodulation is performed on the correction signal, and the fault frequency peak is determined based on the envelope demodulation result, including:

[0031] Performing Hilbert transform on the corrected signal to obtain an analytical signal and an envelope signal of the corrected signal;

[0032] Determine the overall characteristics of the vibration signal based on the analytical signal, including: instantaneous amplitude and phase;

[0033] Determine the frequency domain characteristics of the vibration signal based on the envelope signal;

[0034] According to the frequency domain characteristics of the vibration signal, the fault frequency peak is determined.

[0035] As a possible implementation method, the fault type of the rolling bearing is determined based on the fault frequency peak and fault frequency characteristics. The fault types include outer ring fault, inner ring fault and ball fault, including:

[0036] Determine the rolling bearing fault frequency characteristics based on the rolling bearing's rotational speed, rolling element diameter, bearing diameter, contact angle, number of gear teeth, and correction factor;

[0037] The fault type of the rolling bearing is determined based on the matching degree between the fault frequency peak and the fault frequency characteristic.

[0038] According to a second aspect of the embodiments of the present application, a rolling bearing fault frequency detection device is provided, the device comprising:

[0039] Acquisition module, used for collecting vibration signals of rolling bearings in real time;

[0040] A processing module, used for performing data preprocessing on the vibration signal to obtain a processed vibration signal;

[0041] The processing module is further used to perform adaptive spectrum amplitude modulation processing on the processed vibration signal to obtain a corrected signal;

[0042] a determination module, configured to perform envelope demodulation on the correction signal and determine the fault frequency peak value according to the envelope demodulation result;

[0043] The determination module is further used to determine the fault type of the rolling bearing according to the fault frequency peak value. The fault types include: outer ring fault, inner ring fault and ball fault.

[0044] In a third aspect of an embodiment of the present application, a computer device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the rolling bearing fault frequency detection method described in the first aspect above is implemented.

[0045] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores a computer program. When the computer program is executed by a processor, the rolling bearing fault frequency detection method described in the first aspect is implemented.

[0046] The beneficial effects of the embodiments of the present application include:

[0047] An embodiment of the present application provides a rolling bearing fault frequency detection method. The method collects rolling bearing vibration signals in real time and performs data preprocessing, such as de-redundancy and noise reduction, on the rolling bearing vibration signals to obtain a processed vibration signal. The processed vibration signal is then feature-enhanced using adaptive spectrum amplitude modulation to obtain a corrected signal, and the corrected signal is envelope demodulated to obtain an envelope demodulation result. The fault frequency peak is determined based on the envelope demodulation result, and the type of rolling bearing fault is determined based on a comparison of the fault frequency peak with a theoretical fault frequency characteristic. Data preprocessing can significantly reduce the number of non-fault vibration signals, thereby significantly saving fault detection time. Data preprocessing can also significantly reduce the noise in the vibration signal and significantly suppress spectral smearing of the fault frequency. Furthermore, feature enhancement using adaptive spectrum amplitude modulation enhances the fault frequency characteristics and suppresses the non-fault frequency characteristics, making the fault frequency characteristics easier to identify and the fault diagnosis more reliable. This improves the accuracy and efficiency of rolling bearing fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 A flow chart of a first rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0050] Figure 2 A flow chart of a second rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0051] Figure 3 A flowchart of a third rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0052] Figure 4 A wavelet packet decomposition tree diagram of a vibration signal provided in an embodiment of the present application;

[0053] Figure 5 A histogram of frequency band energy distribution of a sub-band signal provided in an embodiment of the present application;

[0054] Figure 6 A flowchart of a fourth rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0055] Figure 7 A flow chart of adaptive spectrum amplitude modulation of a vibration signal provided in an embodiment of the present application;

[0056] Figure 8 A flowchart of a fifth rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0057] Figure 9 A schematic diagram of an amplitude editing method provided in an embodiment of the present application;

[0058] Figure 10 A schematic diagram of a standardized amplitude provided in an embodiment of the present application;

[0059] Figure 11 A flowchart of a sixth rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0060] Figure 12 A flowchart of a seventh rolling bearing fault frequency detection method provided in an embodiment of the present application;

[0061] Figure 13 This is a table of geometric parameters of an existing deep groove ball bearing;

[0062] Figure 14 A schematic diagram of a bearing fault detection result provided in an embodiment of the present application;

[0063] Figure 15 This is a schematic diagram of an existing bearing fault detection result;

[0064] Figure 16 is an existing bearing data set information table;

[0065] Figure 17 Another schematic diagram of bearing fault detection results provided in an embodiment of the present application;

[0066] Figure 18 This is another schematic diagram of existing bearing fault detection results;

[0067] Figure 19 A schematic structural diagram of a rolling bearing fault frequency detection device provided in an embodiment of the present application;

[0068] Figure 20 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0069] To make the objectives, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Generally, the components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0070] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for protection, but merely represents selected embodiments of the present application. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments in the present application without creative work are within the scope of protection of the present application.

[0071] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0072] Currently, rolling bearing fault diagnosis is often performed by collecting vibration signals from rolling bearings during operation and analyzing the results. However, this approach is susceptible to various factors, such as noise, load, and speed fluctuations, resulting in the collected vibration signals being insufficient to reflect the current operating status of the rolling bearing. Furthermore, the processing steps involved in vibration signal data collection, analysis, and fault diagnosis are time-consuming, resulting in poor timeliness in rolling bearing diagnosis. This, in turn, leads to poor accuracy and efficiency in rolling bearing fault diagnosis.

[0073] To this end, an embodiment of the present application provides a rolling bearing fault frequency detection method. The method collects the rolling bearing's vibration signal in real time and performs data preprocessing on the vibration signal to obtain a processed vibration signal. The processed vibration signal is then enhanced using adaptive spectrum amplitude modulation to obtain a corrected signal. The corrected signal is then envelope demodulated, and the fault frequency peak is determined based on the envelope demodulation result. The rolling bearing fault type is determined based on a comparison between the fault frequency peak and the fault frequency characteristics. This method improves the accuracy and efficiency of rolling bearing fault diagnosis.

[0074] The rolling bearing fault frequency detection method provided in the embodiment of the present application is explained in detail below.

[0075] Figure 1 For the flow chart of the rolling bearing fault frequency detection method provided in this application, see Figure 1 The rolling bearing fault frequency detection method provided in the embodiment of the present application is applied to computer equipment. The rolling bearing fault frequency detection method provided in the embodiment of the present application includes:

[0076] S101. Collect vibration signals of rolling bearings in real time.

[0077] Alternatively, vibration sensors are currently used to collect vibration signals of rolling bearings. Vibration signals are signals generated when rolling bearings are running. Information such as the amplitude, frequency, and phase of the vibration signals can accurately reflect the mechanical operating status of the rolling bearings. The vibration signals of rolling bearings can be used as information for identifying the mechanical operating status of rolling bearings and for diagnosing and analyzing mechanical faults.

[0078] S102: Perform data preprocessing on the vibration signal to obtain a processed vibration signal.

[0079] Optionally, the vibration signal is preprocessed to obtain a vibration signal after noise reduction and redundancy removal. The processed vibration signal has a lower signal-to-noise ratio than the original vibration signal, and the processed vibration signal more closely reflects the fault characteristics of the rolling bearing. The original vibration signal refers to the vibration signal directly collected by the vibration sensor from the rolling bearing.

[0080] S103 , performing feature enhancement on the processed vibration signal through adaptive spectrum amplitude modulation to obtain a corrected signal.

[0081] Optionally, the processed vibration signal is enhanced through adaptive frequency amplitude modulation to produce a modified signal. The modified signal is the vibration signal obtained by amplitude modulating the original vibration signal of the rolling bearing. It is worth noting that after feature enhancement through adaptive frequency amplitude modulation (ASAM), the characteristic frequency of rolling bearing faults is more easily detected.

[0082] It is worth noting that spectrum amplitude modulation enhances the fault characteristic frequency part of the vibration signal, while the non-fault characteristic frequency part is suppressed, making the fault characteristics in the vibration signal of the rolling bearing easier to identify and the fault frequency detection of the rolling bearing easier.

[0083] S104: Perform envelope demodulation on the corrected signal, and determine the fault frequency peak value according to the envelope demodulation result.

[0084] Optionally, the corrected signal is analyzed using envelope demodulation technology to obtain an envelope demodulation result corresponding to the corrected signal. The envelope demodulation result includes the analyzed signal and the envelope signal. Based on the envelope demodulation result, the fault frequency peak of the rolling bearing vibration signal can be determined. The fault frequency peak indicates the maximum fault frequency in the frequency band obtained after envelope demodulation of the corrected signal.

[0085] S105. Determine the rolling bearing fault type according to the fault frequency peak and the fault frequency characteristics. The fault types include outer ring fault, inner ring fault, and ball fault.

[0086] Optionally, the type of rolling bearing fault is determined based on a comparison of the fault frequency peak value with a theoretically calculated fault frequency characteristic, wherein the fault frequency characteristic is a theoretical characteristic value calculated based on a theoretical formula, and the fault frequency peak value is a fault frequency characteristic obtained through the rolling bearing fault detection provided by this application.

[0087] In an embodiment of the present application, a vibration signal of a rolling bearing is collected in real time and preprocessed by performing data preprocessing such as de-redundancy and noise reduction to obtain a processed vibration signal. The processed vibration signal is then feature-enhanced using adaptive spectrum amplitude modulation to obtain a corrected signal, and the corrected signal is envelope demodulated to obtain an envelope demodulation result. The fault frequency peak is determined based on the envelope demodulation result, and the type of rolling bearing fault is determined based on a comparison of the fault frequency peak with a theoretical fault frequency characteristic. Data preprocessing can significantly reduce the number of non-fault vibration signals, thereby significantly saving fault detection time. Data preprocessing can also significantly reduce the noise in the vibration signal and significantly suppress spectral smearing of the fault frequency. Furthermore, feature enhancement using adaptive spectrum amplitude modulation enhances the fault frequency characteristics and suppresses the non-fault frequency characteristics, making the fault frequency characteristics easier to identify and the fault diagnosis results more reliable. This improves the accuracy and efficiency of rolling bearing fault diagnosis.

[0088] In an optional embodiment, see Figure 2 The operation of step S102 may specifically be:

[0089] S201. Determine whether the vibration signal is a full life cycle signal. The full life cycle signal refers to a vibration signal when the rolling bearing enters a faulty operation state from a normal operation state.

[0090] Optionally, vibration signals are classified and processed based on their data sampling type. Vibration signals are primarily divided into full-lifecycle signals and non-full-lifecycle signals. Full-lifecycle signals refer to vibration signals generated during the entire period from when the rolling bearing operates normally to when it mechanically fails, until the rolling bearing ceases to operate. Non-full-lifecycle signals refer to fault vibration signals generated from the onset of a mechanical failure. In other words, non-full-lifecycle signals are vibration signals directly related to the mechanical failure of the rolling bearing, and wavelet packet noise reduction processing can be performed directly on the fault-related vibration signals.

[0091] Optionally, the full life cycle signal contains a large amount of redundant information due to the long signal acquisition interval, that is, the full life cycle signal contains a large amount of vibration signals generated by the normal operation of the rolling bearing. The vibration signals generated by normal operation are regarded as non-essential feature signals, and these non-essential feature signals need to be processed by dimensionality reduction to retain the vibration signals after the mechanical failure occurs to the greatest extent.

[0092] S202: If yes, perform redundancy removal processing on the vibration signal based on the sparse stacked autoencoder, and perform wavelet packet denoising processing on the vibration signal after redundancy removal processing to obtain a processed vibration signal, wherein the wavelet packet denoising processing includes wavelet packet decomposition and wavelet packet reconstruction.

[0093] Optionally, if it is determined that the original vibration signal belongs to a full life cycle signal, that is, the original vibration signal not only includes the vibration signal generated by the mechanical failure of the rolling bearing, but also includes the vibration signal generated during the normal operation of the rolling bearing, the original vibration signal of the rolling bearing is de-redundantly processed by a sparse stacked autoencoder to achieve efficient reconstruction of the low-resolution fault signal of the rolling bearing, which can improve the fault characteristics in the original vibration signal of the rolling bearing while improving the efficiency of vibration signal processing and analysis.

[0094] Optionally, de-redundancy processing refers to deleting or compressing vibration signals that are irrelevant to mechanical faults in the original vibration signals of the rolling bearing, that is, removing the vibration signals generated during the normal operation of the rolling bearing, so as to make the fault characteristics of the vibration signal more obvious. It is worth noting that this application only uses the compressed sensing method to compress the original vibration signals of the rolling bearing throughout its life cycle, so as to reduce the interference of non-fault vibration signals in the full life cycle signals on the fault frequency detection, and realize the self-encoding of the vibration signals by compressed sensing.

[0095] Optionally, the vibration signal after de-redundancy processing is subjected to wavelet packet noise reduction processing to reduce noise interference in the rolling bearing vibration signal, thereby improving the signal-to-noise ratio of the vibration signal. The wavelet packet noise reduction processing includes wavelet packet decomposition and wavelet packet reconstruction. Wavelet packet decomposition is to perform multi-layer decomposition on the vibration signal after de-redundancy processing to obtain vibration signals in each frequency band and delete vibration signals that are significantly affected by noise. Wavelet packet reconstruction is to reconstruct the sub-band signals after wavelet packet decomposition and noise reduction processing. The vibration signal obtained after wavelet packet reconstruction has a higher signal-to-noise ratio.

[0096] It is worth noting that since the vibration signal of the rolling bearing is very stable and contains a large number of fault characteristics, the high-frequency signal and the low-frequency signal in the vibration signal are decomposed at the same time through wavelet packet decomposition, that is, the vibration signal is processed through low-pass filter and high-pass filter.

[0097] Optionally, using a sparse stacked autoencoder (SSAE) to perform compressed sensing dimensionality reduction on the rolling bearing's original vibration signal can significantly reduce the rolling bearing's vibration signal and the number of events used in the simulation, significantly improving the efficiency of rolling bearing fault diagnosis. The sparse stacked autoencoder is a deep learning model that combines the advantages of both stacked and sparse autoencoders. By training multiple autoencoders layer by layer and adding sparsity constraints to each autoencoder's hidden layer, it extracts more representative and robust features.

[0098] Optionally, wavelet packet noise reduction processing can be used to greatly reduce the noise portion of the vibration signal and suppress the spectrum smearing phenomenon to a great extent.

[0099] S203: If not, perform wavelet packet noise reduction processing on the vibration signal to obtain a processed vibration signal.

[0100] Optionally, the original vibration signal of the rolling bearing is decomposed by wavelet packet according to the scale and frequency band of the vibration signal, and the sub-band signal after wavelet packet decomposition is reconstructed by analyzing the wavelet packet coefficients and energy distribution in different frequency bands to obtain a reconstructed vibration signal.

[0101] Optionally, the reconstructed signal is finally subjected to adaptive spectrum amplitude modulation (ASAM), and then frequency fault detection of the rolling bearing is achieved through envelope detection and peak detection of the frequency band.

[0102] In an optional embodiment, see Figure 3 The operation of "performing wavelet packet noise reduction processing on the vibration signal to obtain a processed vibration signal" in the above step S202 may specifically be:

[0103] S301 , performing first-level wavelet packet decomposition on the vibration signal to obtain a high-frequency signal and a low-frequency signal corresponding to the vibration signal.

[0104] Optionally, since the vibration signal of a rolling bearing is non-stationary, it contains a large number of important fault frequency characteristics. Performing wavelet packet noise reduction on the vibration signal of the rolling bearing can reduce the interference of noise in the vibration signal, thereby improving the signal-to-noise ratio of the vibration signal, thereby making the detection results more accurate. It is worth noting that wavelet packet noise reduction can simultaneously process both high-frequency and low-frequency signals in the vibration signal of the rolling bearing. Wavelet packet noise reduction is equivalent to high-pass filtering and low-pass filtering, and performs noise reduction on the vibration signal of the rolling bearing at both high-frequency and low-frequency levels.

[0105] Optionally, after performing first-level wavelet packet decomposition on the vibration signal of the rolling bearing, a high-frequency signal and a low-frequency signal corresponding to the vibration signal can be obtained, and the high-frequency signal and the low-frequency signal are divided according to the frequency of the vibration signal.

[0106] S302 : Perform secondary wavelet packet decomposition on the high-frequency signal and the low-frequency signal to obtain a plurality of sub-frequency band signals, wherein each high-frequency signal and each low-frequency signal corresponds to two sub-frequency band signals.

[0107] Optionally, by further performing a two-level wavelet packet decomposition on the high-frequency and low-frequency signals of the vibration signal, the sub-band signals corresponding to the high-frequency signal and the sub-band signals corresponding to the low-frequency signal can be obtained. It is worth noting that as the level of wavelet packet decomposition increases, information on the rolling bearing vibration signal across all frequency bands can be obtained.

[0108] Alternatively, the wavelet packet decomposition coefficients can be solved according to the following formulas (1) and (2), which are as follows:

[0109]

[0110] in, represents the j-th level wavelet packet coefficient, represents the j+1th level wavelet packet coefficient, h 0,2l-k and h l,2l-k Both represent wavelet packet decomposition coefficients.

[0111] S303 : Calculate the frequency band energy of each sub-band signal based on a preset wavelet basis function.

[0112] Optionally, the preset wavelet basis function may be a demy wavelet basis function, through which the energy coefficient of each sub-band signal after wavelet packet decomposition may be calculated, and the band energy of each sub-band signal may be determined based on the energy coefficient of each sub-band signal.

[0113] S304: Determine a plurality of effective sub-band signals according to the frequency band energy of each sub-band signal.

[0114] Optionally, multiple sub-band signals with higher band energies are selected from the band energies of the sub-band signals of the vibration signal as valid sub-band signals, and sub-band signals with lower band energies are discarded. Valid sub-band signals are sub-bands with relatively higher band energies.

[0115] S305 , performing wavelet packet reconstruction on each effective sub-band signal to obtain a processed vibration signal.

[0116] Optionally, a vibration signal with a higher signal-to-noise ratio can be obtained by performing wavelet packet reconstruction based on the effective sub-band signal.

[0117] Optionally, wavelet packet reconstruction is performed on the sub-band signal after wavelet packet decomposition according to the following formula (3) to obtain a valid vibration signal. Formula (3) is as follows:

[0118]

[0119] Among them, g 0,l-2k and g 1,l-2k Both represent wavelet packet reconstruction coefficients.

[0120] Figure 4 This application provides a wavelet packet decomposition tree diagram of a vibration signal, see Figure 4 In this embodiment, the original vibration signal is subjected to three-layer wavelet packet decomposition. After the first layer of decomposition, a low-frequency vibration signal and a high-frequency vibration signal are obtained. In the second layer of decomposition, the high-frequency vibration signal and the low-frequency vibration signal are further decomposed. As the wavelet packet decomposition level increases, the wavelet packet decomposition can obtain information about the original vibration signal across all frequency bands. The energy of the sub-bands after the three-layer wavelet packet decomposition accounts for only 1 / 4 of the energy of the original vibration signal's frequency bands.

[0121] Figure 5 This is a histogram of the frequency band energy distribution of a sub-band signal provided by this application, see Figure 5 In an embodiment of the present application, a three-layer wavelet packet decomposition and reconstruction is performed on the vibration signal after compression and dimensionality reduction. The present application calculates the frequency band energy coefficient of the decomposed sub-band signal through the dmey wavelet basis function, finds the part with higher energy distribution from the frequency band energy coefficient distribution diagram, discards the part with lower frequency band energy, and performs wavelet packet reconstruction based on the sub-band signal after noise reduction processing.

[0122] In an optional embodiment, see Figure 6 The operation of step S103 may specifically be:

[0123] S601 . Perform Fourier transform on the processed vibration signal to obtain the phase and amplitude of the processed vibration signal.

[0124] Optionally, the vibration signal obtained after wavelet packet reconstruction is subjected to Fourier transform to obtain the phase and amplitude of the vibration signal. The reconstructed vibration signal can retain the image information of the original vibration signal.

[0125] Optionally, the following formula (4) is used to characterize the Fourier series expression of the full life cycle signal with a mean value of 0. Formula (4) is as follows:

[0126]

[0127] Where D represents the Fourier coefficient, q represents the length of the periodic signal x, and x(t) represents the full life cycle signal.

[0128] Optionally, the full life cycle signal x(t) is subjected to a Hilbert transform based on the following formula (5), thereby obtaining the integral characteristic of the full life cycle signal x(t). Formula (5) is as follows:

[0129]

[0130] Optionally, the phase of each element of the Fourier series in the above formula (5) is shifted by 90°. The phase of formula (5) is shifted by 90° to become formula (6), which is as follows:

[0131]

[0132] Optionally, the phase and amplitude of the vibration signal can be obtained based on the following formula (7), which is as follows:

[0133]

[0134] Where FT is the Fourier transform, and x(f) is the frequency domain representation of the full life cycle signal x(t).

[0135] S602: Re-edit the amplitude to obtain a new amplitude with maximum grayscale energy.

[0136] Optionally, the maximum grayscale energy refers to the amplitude with the best amplitude modulation effect, and the amplitude of the vibration signal is edited based on the weight value to obtain a new amplitude.

[0137] S603: Generate a correction signal according to the phase and the new amplitude.

[0138] Optionally, the edited amplitude and phase are re-computed and Fourier transformed to obtain a modified signal. The modified signal only changes the amplitude of the original vibration signal, while the phase of the original vibration signal remains unchanged, thereby enhancing the fault characteristic frequency of the vibration signal.

[0139] Optionally, the new amplitude and phase obtained by editing are combined, and the combined vibration signal is subjected to Fourier transform to obtain a correction signal. The correction signal is determined based on the following formula (8), which is as follows:

[0140]

[0141] Among them, x m represents the correction signal, IFT is the inverse Fourier transform, and n represents the preset weight value.

[0142] Alternatively, REBs diagnosis identifies rolling bearing faults by analyzing the characteristic frequencies in the bearing's vibration signal. REBs diagnosis can pinpoint the repetitive pulse frequency (RPM) caused by bearing defects. The core of REBs diagnosis is detecting periodic impulse signals caused by defects in bearing components such as the inner and outer rings, rolling elements, or cages.

[0143] Figure 7 This application provides a flow chart of a vibration signal adaptive spectrum amplitude modulation, see Figure 7 , the vibration signal obtained after wavelet packet reconstruction is used as the input signal, the input signal is Fourier transformed to obtain the phase and amplitude of the vibration signal, the optimal amplitude modulation weight value is determined based on the grayscale image recognition of the edited amplitude, the amplitude is modulated according to the optimal amplitude modulation weight to obtain a new amplitude, the new amplitude and the original phase are integrated, and the integrated signal is iteratively Fourier calculated to obtain a corrected signal, and the corrected signal is envelope demodulated to obtain the fault frequency peak.

[0144] In an optional embodiment, see Figure 8 The operation of step S602 may specifically be:

[0145] S801 : Edit the amplitude based on a preset weight range, and perform grayscale value scanning on the edited amplitude to obtain grayscale energy corresponding to each preset weight value.

[0146] Optionally, the preset weight range is a selection range of amplitude weight values pre-set by the user, and the amplitude weight n satisfies -0.5<=n<=1.5. This value range is only used as an example and this application does not make any specific limitations on this.

[0147] Optionally, the original amplitude of the vibration signal is edited based on a preset weight range to obtain multiple amplitude images. Image recognition is performed on each amplitude image to obtain a grayscale value corresponding to each amplitude image. The grayscale value is used to represent the grayscale energy of the amplitude image. The grayscale value refers to the brightness value of each pixel in the amplitude image, and the grayscale value generally ranges from 0 to 255.

[0148] S802 : Determine a maximum grayscale energy from the grayscale energies corresponding to the preset weight values, and use the preset weight value corresponding to the maximum grayscale energy as a target weight value.

[0149] Optionally, the amplitude of the original vibration signal is re-edited based on each preset weight value to obtain an amplitude image corresponding to each preset weight value, and image recognition and grayscale value conversion are performed on each amplitude image to obtain the grayscale energy of the amplitude image after re-editing each preset weight value.

[0150] Optionally, the maximum grayscale energy is determined from the grayscale energy of the amplitude image after re-editing the preset weight values, that is, the grayscale energy peak is determined, and the preset weight value corresponding to the grayscale energy peak is used as the target weight value. The preset weight value refers to any weight value within a preset weight range, and the target weight value refers to the preset weight value that achieves the best amplitude modulation effect when re-editing the amplitude of the original vibration signal. The target weight value can be any preset weight value within the preset weight range, and this application does not impose specific limitations on this.

[0151] It is worth noting that when the preset weight value n is 1, the new amplitude is obtained by editing based on the preset weight value, and the correction signal obtained by reconstructing based on the new amplitude is the original vibration signal; when the preset weight value n is 0, the new amplitude is obtained by editing based on the preset weight value, and the correction signal obtained by reconstructing based on the new amplitude is the vibration signal reconstructed by the wavelet packet.

[0152] S803: Use the amplitude corresponding to the target weight value as the new amplitude.

[0153] Optionally, the amplitude edited based on the target weight value is used as the new amplitude of the vibration signal.

[0154] In an optional embodiment, see Figure 9 In the rolling bearing fault frequency detection method provided in the embodiments of this application, the amplitude editing process is achieved through image grayscale value recognition. Image recognition is first performed on the amplitude image to obtain the grayscale values corresponding to the amplitude image. The grayscale value abscissa data is calculated, and the weight value that optimizes the amplitude modulation effect is adaptively selected as the amplitude modulation weight. The higher the grayscale energy of the amplitude image, the better the amplitude modulation effect.

[0155] In an optional embodiment, see Figure 10 , the new correction signal is envelope demodulated by envelope demodulation technology to find the bearing fault frequency peak, and the bearing fault frequency peak is compared with the theoretical fault frequency calculated by the formula to determine the fault type to which the rolling bearing fault frequency belongs, thereby realizing rolling bearing fault detection.

[0156] In an optional embodiment, see Figure 11 The operation of step S104 may specifically be:

[0157] S1101. Perform Hilbert transform on the corrected signal to obtain an analytical signal and an envelope signal of the corrected signal.

[0158] Optionally, the modified signal is subjected to Hilbert transform based on the following formulas (9) and (10) to obtain an analytical signal A and an envelope signal SES of the modified signal. Formulas (9) and (10) are as follows:

[0159] A{x m (t,n)}=x m (t,n)+j×H{x m (t,n)}(9)

[0160] SES{x m (t,n)}=|FT{|A{x m (t,n)}| 2}|(10)

[0161] Optionally, the envelope signal SES is converted into an LSES signal based on the following formula (11). The LSES signal has a more robust characteristic than the envelope signal SES. Formula (11) is as follows:

[0162] LSES{x(t)}=|FT{log(|A{x m (t,n)}| 2 +ε)}|(11)

[0163] Optionally, the square envelope or norm of the analytical signal is calculated based on the following formula (12). The analytical signal is the envelope signal of the single-sideband modulated signal. The analytical signal represents the spectrum of one side of the modulated signal in the frequency domain. Formula (12) is as follows:

[0164]

[0165] Optionally, the following formula (13) is a variation of the above formula (12), and the formula (13) is as follows:

[0166]

[0167] Optionally, the correction signal is expressed by the following formula (14), which is as follows:

[0168]

[0169] Optionally, the analytical signal of the corrected signal is calculated by the following formula (15), which is as follows:

[0170]

[0171] Optionally, the envelope signal SES of the correction signal is calculated by the following formula (16), which is as follows:

[0172]

[0173] in,

[0174] Optionally, q-1 and δ(k) are defined based on the following formula (17), which is as follows:

[0175]

[0176] S1102. Determine overall characteristics of the vibration signal based on the analytical signal, where the overall characteristics include: instantaneous amplitude and phase.

[0177] Optionally, the overall feature is used to characterize the instantaneous amplitude and phase of the vibration signal.

[0178] S1103: Determine the frequency domain characteristics of the vibration signal according to the envelope signal.

[0179] S1104. Determine the fault frequency peak value according to the frequency domain characteristics of the vibration signal.

[0180] In an optional embodiment, see Figure 12 The operation of step S105 may specifically be:

[0181] S1201. Determine the fault frequency characteristics of the rolling bearing according to the rotational speed of the rolling bearing, the diameter of the rolling element, the bearing diameter, the contact angle, the number of gear teeth, and the correction coefficient.

[0182] Optionally, the fault frequency characteristic is generally determined by analyzing the fault characteristic frequency and the frequency multiplication, wherein the frequency multiplication is an integer multiple of the fault characteristic frequency, and the maximum fault frequency peak is obtained by calculation based on the frequency multiplication.

[0183] Optionally, the outer ring fault frequency of the rolling bearing is calculated by the following formula (18), which is as follows:

[0184]

[0185] Alternatively, the inner ring fault frequency of the rolling bearing is calculated by the following formula (19), which is as follows:

[0186]

[0187] Alternatively, the ball failure frequency of the rolling bearing is calculated by the following formula (20), which is as follows:

[0188]

[0189] Wherein, N represents the rotational speed of the rolling shaft, d represents the diameter of the rolling element of the rolling bearing, D represents the bearing diameter of the rolling bearing, β represents the contact angle of the rolling bearing, and k represents the correction coefficient of the rolling bearing.

[0190] Optionally, the fault frequency characteristic refers to a theoretical fault frequency value calculated by formula (18), formula (19) and formula (20).

[0191] S1202. Determine the rolling bearing fault type based on the matching degree between the fault frequency peak and the fault frequency characteristic.

[0192] Optionally, based on the comparison result of the fault frequency peak and the fault frequency feature, it is determined which theoretical fault frequency feature the fault frequency peak is close to, and then the type of the rolling bearing fault is determined.

[0193] The following compares the fault diagnosis effects of the prior art fast spectral kurtosis method - FK and the rolling bearing fault frequency detection method provided by the present application with reference to the accompanying drawings.

[0194] In an optional embodiment, see Figure 13 The deep groove ball bearing geometric parameter table used in the embodiments of this application represents inner race fault data for the drive-end bearing of the rolling bearing. The deep groove ball bearing aggregate parameter represents the non-full life cycle signal of the rolling bearing, which is derived from signals collected after a rolling bearing failure. Based on the rolling bearing's geometric parameters, it can be calculated that when the rolling bearing's experimental rotational speed is 1750 rpm, the characteristic frequency of the rolling bearing's inner race fault is 157.46 Hz, and the rolling bearing's rotational frequency is 58.32 Hz.

[0195] In an optional embodiment, see Figure 14 In the bearing fault detection results obtained by performing bearing fault detection based on the rolling bearing fault frequency detection method provided in the embodiment of the present application, the rolling bearing rotation frequency fr = rpm × n ÷ 60 = 58.3 Hz and the rolling bearing inner ring fault frequency BPFI = 0.6 × Z × fr ÷ n = 157.46 Hz are both clearly visible. The bearing fault detection results obtained by the rolling bearing fault frequency detection method provided in the embodiment of the present application are similar to the theoretical calculation results, which can verify the reliability and effectiveness of the rolling bearing fault frequency detection method provided in the embodiment of the present application. However, the rolling bearing fault frequency detection method provided in the embodiment of the present application still has certain defects, such as the frequency multiplication of BPFI is also severely compressed, and its frequency amplitude is relatively small.

[0196] In an optional embodiment, see Figure 15 The rolling bearing fault detection results obtained by the prior art fast spectral kurtosis method for bearing fault detection show that the two largest rolling bearing rotation frequencies are fr = 58.32 Hz and BPFI = 157.46 Hz, respectively. However, in the entire envelope spectrum, there are still many interference frequencies affecting the axis frequency of the rolling shaft. The rolling bearing fault detection results of the prior art are not as clear as the rolling bearing fault detection results obtained by the rolling bearing fault detection method provided in the embodiments of the present application. This shows that the demodulation frequency band with the maximum kurtosis value determined by the prior art fast spectral kurtosis method may be a frequency band containing impulse noise rather than a characteristic frequency band of rolling bearing faults.

[0197] In an optional embodiment, see Figure 16 The bearing dataset used in this embodiment includes outer ring crack failures of rolling bearings. This bearing dataset represents a full lifecycle bearing failure dataset. Because the bearing load is applied horizontally, the vibration signal in this direction can contain more degradation information about the rolling bearing. Therefore, the vibration signals acquired in this embodiment are all horizontal vibration signals of the rolling bearing.

[0198] In an optional embodiment, see Figure 17 and Figure 18 The bearing fault detection results obtained by using the rolling bearing fault frequency detection method provided by the embodiment of the present application can clearly display not only the bearing fault characteristic frequency, but also the operating shaft frequency of the rolling bearing. However, the rolling bearing fault detection results obtained by using the existing fast spectral kurtosis method suffer from the aliasing problem of the shaft frequency and other noise signals, which greatly reduces the accuracy of the bearing fault detection results. Therefore, the rolling bearing fault detection method provided by the embodiment of the present application has higher reliability than the existing technology.

[0199] The following describes the apparatus, device, and computer-readable storage medium used to implement the rolling bearing fault frequency detection method provided in this application. The specific implementation process and technical effects are described above and will not be repeated below.

[0200] Figure 19 This is a schematic diagram of the structure of a rolling bearing fault frequency detection device provided by an embodiment of the present application, see Figure 19 , the device comprises:

[0201] Acquisition module 1901, used for collecting vibration signals of rolling bearings in real time;

[0202] The processing module 1902 is used to perform data preprocessing on the vibration signal to obtain a processed vibration signal;

[0203] The processing module 1902 is further configured to perform adaptive spectrum amplitude modulation processing on the processed vibration signal to obtain a corrected signal;

[0204] The determination module 1903 is configured to perform envelope demodulation on the correction signal and determine the fault frequency peak value according to the envelope demodulation result;

[0205] The determination module 1903 is further configured to determine the fault type of the rolling bearing according to the fault frequency peak value. The fault types include outer ring fault, inner ring fault and ball fault.

[0206] In an optional implementation, the processing module 1902 may be specifically configured to:

[0207] Determine whether the vibration signal is a full life cycle signal. A full life cycle signal refers to the vibration signal of a rolling bearing going from normal operation to a faulty operation state.

[0208] If so, de-redundancy processing is performed on the vibration signal based on a sparse stacked autoencoder, and wavelet packet denoising is performed on the vibration signal after de-redundancy processing to obtain a processed vibration signal, where the wavelet packet denoising processing includes: wavelet packet decomposition and wavelet packet reconstruction;

[0209] If not, the vibration signal is subjected to wavelet packet noise reduction processing to obtain a processed vibration signal.

[0210] In an optional implementation, the processing module 1902 may further be configured to:

[0211] Perform first-level wavelet packet decomposition on the vibration signal to obtain the high-frequency signal and low-frequency signal corresponding to the vibration signal;

[0212] Performing secondary wavelet packet decomposition on the high-frequency signal and the low-frequency signal to obtain a plurality of sub-band signals, wherein each high-frequency signal and each low-frequency signal corresponds to two sub-band signals;

[0213] Calculate the frequency band energy of each sub-band signal based on the preset wavelet basis function;

[0214] determining a plurality of effective sub-band signals according to the frequency band energy of each sub-band signal;

[0215] The wavelet packet is reconstructed for each effective sub-band signal to obtain the processed vibration signal.

[0216] In an optional implementation, the processing module 1902 may further be configured to:

[0217] Performing Fourier transform on the processed vibration signal to obtain the phase and amplitude of the processed vibration signal;

[0218] Re-editing the amplitude to obtain a new amplitude with maximum grayscale energy;

[0219] A correction signal is generated based on the phase and the new amplitude.

[0220] In an optional implementation, the processing module 1902 may further be configured to:

[0221] The amplitude is edited based on a preset weight range, and the grayscale value of the edited amplitude is scanned to obtain the grayscale energy corresponding to each preset weight value;

[0222] Determine the maximum grayscale energy from the grayscale energies corresponding to the preset weight values, and use the preset weight value corresponding to the maximum grayscale energy as the target weight value;

[0223] The amplitude corresponding to the target weight value is used as the new amplitude.

[0224] In an optional implementation, the determination module 1903 may be specifically configured to:

[0225] Performing Hilbert transform on the corrected signal to obtain an analytical signal and an envelope signal of the corrected signal;

[0226] Determine the overall characteristics of the vibration signal based on the analytical signal, including: instantaneous amplitude and phase;

[0227] Determine the frequency domain characteristics of the vibration signal based on the envelope signal;

[0228] According to the frequency domain characteristics of the vibration signal, the fault frequency peak is determined.

[0229] In an optional implementation, the determining module 1903 may further be configured to:

[0230] Determine the rolling bearing fault frequency characteristics based on the rolling bearing's rotational speed, rolling element diameter, bearing diameter, contact angle, number of gear teeth, and correction factor;

[0231] The fault type of the rolling bearing is determined based on the matching degree between the fault frequency peak and the fault frequency characteristic.

[0232] The above-mentioned device is used to execute the method provided in the above-mentioned embodiment. Its implementation principle and technical effect are similar and will not be repeated here.

[0233] The above modules can be one or more integrated circuits configured to implement the above methods, such as one or more application-specific integrated circuits (ASICs), one or more microprocessors, or one or more field programmable gate arrays (FPGAs). For another example, when a module is implemented by scheduling program code through a processing element, the processing element can be a general-purpose processor, such as a central processing unit (CPU) or other processor that can call program code. For another example, these modules can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0234] Figure 20This is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. Figure 20 The computer device includes: a memory 2001 and a processor 2002. The memory 2001 stores a computer program that can be run on the processor 2002. When the processor 2002 executes the computer program, the steps in any of the above method embodiments are implemented.

[0235] An embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0236] Optionally, the present application also provides a program product, such as a computer-readable storage medium, comprising a program, which, when executed by a processor, is used to execute any of the above-mentioned rolling bearing fault frequency detection method embodiments.

[0237] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0238] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0239] In addition, the functional units in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional units.

[0240] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor (English: processor) to perform some steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (English: Read-Only Memory, abbreviated: ROM), a random access memory (English: Random Access Memory, abbreviated: RAM), a magnetic disk or an optical disk, and other media that can store program code.

[0241] The above are only specific embodiments of the present application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0242] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A rolling bearing fault frequency detection method, characterized in that: The method comprises: Real-time collection of rolling bearing vibration signals; performing data preprocessing on the vibration signal to obtain a processed vibration signal; The processed vibration signal is enhanced by adaptive spectrum amplitude modulation to obtain a corrected signal; performing envelope demodulation on the correction signal, and determining a fault frequency peak value according to the envelope demodulation result; The fault type of the rolling bearing is determined according to the fault frequency peak and the fault frequency characteristics, and the fault types include: outer ring fault, inner ring fault and ball fault.

2. The rolling bearing fault frequency detection method according to claim 1, characterized in that: The performing data preprocessing on the vibration signal to obtain a processed vibration signal includes: Determining whether the vibration signal is a full life cycle signal, wherein the full life cycle signal refers to a vibration signal of the rolling bearing entering a faulty operation state from a normal operation state; If so, performing redundancy removal processing on the vibration signal based on a sparse stacked autoencoder, and performing wavelet packet denoising processing on the vibration signal after the redundancy removal processing to obtain a processed vibration signal, wherein the wavelet packet denoising processing includes: wavelet packet decomposition and wavelet packet reconstruction; If not, wavelet packet noise reduction processing is performed on the vibration signal to obtain a processed vibration signal.

3. The rolling bearing fault frequency detection method according to claim 2, characterized in that: The performing wavelet packet noise reduction processing on the vibration signal to obtain a processed vibration signal includes: Performing a first-level wavelet packet decomposition on the vibration signal to obtain a high-frequency signal and a low-frequency signal corresponding to the vibration signal; Performing secondary wavelet packet decomposition on the high-frequency signal and the low-frequency signal to obtain a plurality of sub-frequency band signals, wherein each high-frequency signal and each low-frequency signal corresponds to two sub-frequency band signals; Calculating the frequency band energy of each of the sub-band signals based on a preset wavelet basis function; determining a plurality of effective sub-band signals according to the frequency band energy of each of the sub-band signals; Wavelet packet reconstruction is performed on each of the effective sub-band signals to obtain a processed vibration signal.

4. The rolling bearing fault frequency detection method according to claim 1, characterized in that: The method of enhancing the characteristics of the processed vibration signal by adaptive spectrum amplitude modulation to obtain a corrected signal includes: Performing Fourier transform on the processed vibration signal to obtain the phase and amplitude of the processed vibration signal; Re-editing the amplitude to obtain a new amplitude with maximum grayscale energy; The correction signal is generated according to the phase and the new amplitude.

5. The rolling bearing fault frequency detection method according to claim 4, characterized in that: The amplitude is re-edited to obtain a new amplitude, including: Editing the amplitude based on a preset weight range, and performing grayscale value scanning on the edited amplitude to obtain grayscale energy corresponding to each preset weight value; Determine the maximum grayscale energy from the grayscale energies corresponding to the preset weight values, and use the preset weight value corresponding to the maximum grayscale energy as the target weight value; The amplitude corresponding to the target weight value is used as the new amplitude.

6. The rolling bearing fault frequency detection method according to claim 1, characterized in that: The performing envelope demodulation on the correction signal and determining the fault frequency peak value according to the envelope demodulation result includes: performing a Hilbert transform on the correction signal to obtain an analytical signal and an envelope signal of the correction signal; Determining overall characteristics of the vibration signal based on the analytical signal, the overall characteristics including: instantaneous amplitude and phase; determining frequency domain characteristics of the vibration signal according to the envelope signal; A fault frequency peak is determined according to the frequency domain characteristics of the vibration signal.

7. The rolling bearing fault frequency detection method according to claim 1, characterized in that: The method of determining the rolling bearing fault type according to the fault frequency peak and the fault frequency characteristics includes: outer ring fault, inner ring fault and ball fault, including: Determining a fault frequency characteristic of the rolling bearing according to the rotational speed of the rolling bearing, the diameter of the rolling element, the bearing diameter, the contact angle, the number of gear teeth, and the correction coefficient; The fault type of the rolling bearing is determined according to the matching degree between the fault frequency peak and the fault frequency characteristic.

8. A rolling bearing fault frequency detection device, characterized in that: The device comprises: Acquisition module, used to collect vibration signals of rolling bearings in real time; a processing module, configured to perform data preprocessing on the vibration signal to obtain a processed vibration signal; The processing module is further used to perform adaptive spectrum amplitude modulation processing on the processed vibration signal to obtain a corrected signal; a determination module, configured to perform envelope demodulation on the correction signal and determine a fault frequency peak value according to the envelope demodulation result; The determination module is further configured to determine the fault type of the rolling bearing according to the fault frequency peak value, wherein the fault types include outer ring fault, inner ring fault and ball fault.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the steps of the method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

Images