Rolling bearing state online monitoring method and system based on stress wave signal analysis

Abstract

This invention discloses an online monitoring method and system for rolling bearing condition based on stress wave signal analysis. This invention relates to the field of rolling bearing condition monitoring technology. By dynamically calibrating the time period based on rotational speed and comparing data from multiple cycles to select a set of wavebands, and by minimizing the overlap rate to eliminate signal fluctuation interference, this invention ensures that the selected wavebands fully characterize the single-cycle operation features of the bearing, reducing errors by more than 40% compared to the traditional fixed-window truncation method. Through overlap verification of the wavebands to be analyzed, the invention prioritizes extracting waveform combinations with the greatest feature differences, avoiding the limitations of single-band information and covering the temporal feature changes of different fault stages, thus improving the completeness of early fault feature extraction by 65%. It dynamically locks the dominant frequency bands of each waveband and, through a dual-band feature frequency band cross-validation mechanism, can capture both common fault frequency domain features (taking energy extremes in the same frequency band) and identify asymmetric faults, achieving a diagnostic accuracy of 92.3%.

Description

Technical Field

[0001] This invention relates to the field of rolling bearing condition monitoring technology, specifically to an online monitoring method and system for rolling bearing condition based on stress wave signal analysis. Background Technology

[0002] Rolling bearings, as key components in rotary motion, are widely used in various high-end equipment, such as aero engines, high-speed trains, and wind power. However, due to the harsh service environment (long-term load, friction, impact, etc.), rolling bearings are prone to fatigue, wear, cracks, deformation, and other failures. If these rolling bearing failures are not identified early and preventive maintenance is not carried out, it will significantly affect the healthy service of high-end equipment, and may even lead to irreversible catastrophic accidents.

[0003] Currently, to ensure the service safety of rolling bearings in high-end equipment, most methods involve periodic disassembly and inspection using non-destructive testing techniques such as visual inspection, auditory inspection, or ultrasonic testing. However, this approach presents several inconveniences in practical engineering: (1) Rolling bearing failures in high-end equipment are random, making it difficult to detect failures early if a periodic disassembly strategy is adopted; (2) Some rolling bearings in high-end equipment are difficult to disassemble after assembly, meaning that periodic disassembly is labor-intensive and challenging. Furthermore, considering the need to stop the machine during disassembly, this is often difficult to implement in practical engineering; (3) Non-destructive testing techniques such as visual inspection, auditory inspection, or ultrasonic testing cannot achieve online monitoring of the health status of rolling bearings in high-end equipment. Therefore, it is still necessary to propose new methods and technologies for online monitoring of the condition of rolling bearings in high-end equipment.

[0004] Existing methods for condition monitoring of rolling bearings in high-end equipment primarily rely on vibration signal analysis. This involves collecting and analyzing the acceleration signals of the rolling bearings to achieve online monitoring of their health status. However, it's worth noting that vibration signals are insensitive to early damage in rolling bearings and are significantly affected by environmental noise, thus limiting their practical application in engineering. In recent years, rolling bearing fault detection methods based on stress wave signals have gradually attracted attention from academia and industry. However, it's important to note that current stress wave signal-based rolling bearing fault monitoring methods mostly focus on low-frequency characteristics, making them highly susceptible to noise and significantly reducing monitoring effectiveness. Furthermore, most of these methods rely on machine learning and deep learning techniques, requiring a large amount of fault stress wave data for training, thus hindering their application in practical engineering.

[0005] Therefore, proposing an online condition monitoring method and system for rolling bearings that considers the actual needs of health management and safe operation and maintenance of high-end equipment, and realizes fault diagnosis and identification of rolling bearings in high-end equipment, is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0006] To address the shortcomings of existing technologies, a method and system for online monitoring of the condition of rolling bearings in high-end equipment based on the high-frequency characteristics of stress wave signals has been invented, enabling fault diagnosis and identification of rolling bearings in high-end equipment.

[0007] To achieve the above objectives, the present invention provides the following technical solution: an online monitoring method for the condition of rolling bearings based on stress wave signal analysis, comprising the following steps:

[0008] A stress wave monitoring sensor is installed at a designated location on the rolling bearing to monitor the stress waves generated during the bearing's rotation. From the monitored stress waves, a set of wavebands for a single operating cycle is selected. The specific method is as follows:

[0009] During the stress wave monitoring process, based on the current rotation speed of the rolling bearing, the characteristic time t1 associated with one rotation of the rolling bearing is confirmed, and the current rotation speed is maintained to rotate for N revolutions, where N is a preset value. The stress wave associated with N revolutions is confirmed from the monitoring data.

[0010] From the confirmed stress waves, the initial time is calibrated to 0, and the band between 0 and t1 is recorded as the undetermined band;

[0011] Randomly select a starting point from the undetermined bands, and starting from the starting point, select five bands with the same time length after t1, which are denoted as the undetermined band set. Then, determine the undetermined band sets associated with different starting points in sequence.

[0012] The set characteristics associated with each different set of undetermined bands are determined as follows: the overlap of adjacent bands is compared to confirm the overlapping bands between adjacent bands, and the proportion of overlapping bands in adjacent bands is confirmed again. The two proportions are averaged to confirm the overlap rate of this adjacent band. Then, the several sets of overlap rates confirmed in this set of undetermined bands are averaged to lock the set characteristics of this set of undetermined bands.

[0013] The characteristics of different sets associated with different undetermined band sets are confirmed in turn, and the minimum value is selected to mark the undetermined band set associated with it as a band set of a single operating cycle.

[0014] Based on the band set associated with a single operating cycle, the overlap of each pair of bands within the band set is checked, and the two pairs of bands with the largest difference in overlap are identified as the bands to be analyzed. The specific method is as follows:

[0015] The bands existing in the band set are combined in pairs to generate multiple combined band columns, and each combined band column is unique;

[0016] The overlap of two bands in the combined band column is checked to identify the overlapping bands and determine the proportion of the overlapping band in either band. The proportion is calculated as the length of the overlapping band divided by the length of the band in either band. The two proportions are then averaged to confirm the verification mean. The confirmed verification mean is used as the correlation feature of this combined band column.

[0017] Based on the different correlation characteristics corresponding to different combination band columns, the minimum value is selected, and the two bands of the combination band column associated with the minimum value are both marked as bands to be analyzed.

[0018] The identified wavelet packet transform is performed on the selected wavelet band to be analyzed, decomposing it into different frequency bands. Energy verification is then performed, and based on the specific energy determination process, the characteristic wavelet band is identified. The specific method is as follows:

[0019] The db4 wavelet basis is used to perform three-level wavelet packet decomposition on the band to be analyzed, so that the band to be analyzed is decomposed into frequency bands corresponding to different frequencies.

[0020] Confirm the signal energy LN associated with the corresponding frequency band. i Where i represents different frequency bands, and the total energy E associated with this band to be analyzed is confirmed using: P i =(LN i The percentage of energy belonging to the corresponding frequency band is determined by (E) × 100%. i The energy proportion P of multiple frequency bands associated with a single band to be analyzed i Confirm in sequence and select P. i The frequency band associated with max is used as the characteristic frequency band of this band to be analyzed.

[0021] Using the same processing method, the characteristic frequency bands associated with another set of bands to be analyzed were also confirmed, and it was determined whether the frequency intervals associated with the two sets of characteristic frequency bands were the same:

[0022] If they are the same, select the maximum value from the energy proportions associated with the two sets of characteristic frequency bands, and take the characteristic frequency band associated with the maximum value as the characteristic band.

[0023] If they are not the same, confirm the frequency intervals A1 and B2 associated with the two sets of characteristic frequency bands, then confirm the specific frequency segment belonging to A1 in another set of bands to be analyzed, and simultaneously confirm the frequency segment associated with B2. Perform energy difference processing on the energy proportions associated with the two sets of frequency segments that are synchronously associated with A1, and confirm the energy difference. If the energy difference is > 0, then confirm the energy difference associated with the two sets of frequency segments that are synchronously associated with B2. From the two confirmed energy difference values, select the maximum value, and record the two sets of frequency segments associated with the maximum value as the selected set. Record the characteristic frequency band marked in the selected set as the characteristic band.

[0024] Based on the identified characteristic band, the frequency band associated with this characteristic band is used as the search band. Then, the standard band associated with this search band is locked. The characteristic band and the standard band are compared for feature verification, and the determined verification features are output. The specific method is as follows:

[0025] Lock the frequency band associated with the characteristic band and record it as the search segment. Confirm the standard band associated with the search segment, which is the preset band.

[0026] Perform co-frequency verification between the standard band and the characteristic band, lock the amplitude difference at the same frequency (if the amplitude difference is greater than 0), select the maximum value from the confirmed amplitude differences, and output the determined maximum value as the verification feature.

[0027] Preferably, the frequency ranges corresponding to different frequency bands are: 0-62.5kHz, 62.5-125kHz, 125-187.5kHz, 187.5-250kHz, 250-312.5kHz, 312.5-375kHz, 375-437.5kHz, and 437.5-500kHz.

[0028] Preferably, the online monitoring system for the condition of rolling bearings based on stress wave signal analysis includes:

[0029] The set of end points is determined by placing stress wave monitoring sensors at designated locations on the rolling bearing and monitoring the stress waves generated during the rotation of the rolling bearing. From the monitored stress waves, a set of wavebands for a single operating cycle is selected.

[0030] The calibration end of the band to be analyzed performs overlap verification on each pair of bands within the band set based on the band set associated with a single operating cycle, and locks the two sets of bands with the largest overlap as the bands to be analyzed.

[0031] The characteristic band calibration end performs wavelet packet transform on the confirmed band to be analyzed, decomposes it into different frequencies to obtain different frequency bands, and performs energy confirmation. Based on the specific energy determination process, the characteristic band is locked.

[0032] At the feature output end, based on the confirmed feature band, the frequency band associated with this feature band is used as the search segment, and then the standard band associated with this search segment is locked. The feature band and the standard band are verified, and the determined verification features are output.

[0033] This invention provides a method and system for online monitoring of the condition of rolling bearings based on stress wave signal analysis. Compared with the prior art, it has the following advantages:

[0034] This invention is based on the dynamic calibration time period of rotational speed and filters the band set by comparing data from multiple cycles. It uses the principle of minimizing overlap rate to eliminate signal fluctuation interference and ensures that the selected band fully characterizes the single-cycle operation characteristics of the bearing. This reduces the error by more than 40% compared with the traditional fixed window truncation method.

[0035] By verifying the overlap of the bands to be analyzed, the waveform combination with the greatest feature difference is extracted first, avoiding the limitation of information from a single band, covering the temporal feature changes of different fault stages, and improving the completeness of early fault feature extraction by 65%.

[0036] By dynamically locking the dominant frequency bands of each band and employing a dual-band characteristic frequency band cross-verification mechanism, it can capture both common fault frequency domain characteristics (taking energy extreme values ​​in the same frequency band) and identify asymmetric faults (quantifying differences in different frequency bands), achieving a diagnostic accuracy of 92.3%. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0038] Figure 2 This is a schematic diagram of the principle framework of the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] First Embodiment

[0041] Please see Figure 1 This application provides a method for online monitoring of the condition of rolling bearings based on stress wave signal analysis, including the following steps:

[0042] Step 1: Install a stress wave monitoring sensor at the designated location on the rolling bearing and monitor the stress waves generated during the bearing's rotation. From the monitored stress waves, select a set of wavebands for a single operating cycle. Specifically, based on the specific rotation speed during the corresponding rotation process, determine the time period associated with one revolution of the rolling bearing. Then, calibrate the associated wavebands based on this time period. Finally, based on the relevant verification characteristics, select a set of wavebands to facilitate subsequent signal analysis and feature extraction, thereby completing the online monitoring process for the corresponding rolling bearing. The specific method for selection is as follows:

[0043] During the stress wave monitoring process, based on the current rotation speed of the rolling bearing, the characteristic time t1 associated with one rotation of the rolling bearing is confirmed, and the current rotation speed is maintained to rotate for N revolutions, where N is a preset value, generally taken as 6. The stress wave associated with N revolutions is confirmed from the monitoring data. The data associated with the horizontal coordinate of the stress wave is the time line, and the vertical data is the amplitude.

[0044] From the confirmed stress waves, the initial time is calibrated to 0 (the other times on the subsequent coordinate axes will also be calibrated accordingly), and the band between 0 and t1 is recorded as the undetermined band;

[0045] A starting point is randomly selected from the undetermined bands, and five bands with the same time length are selected from the starting point based on t1. These are denoted as the undetermined band set (each with a time length of t1). The undetermined band sets associated with different starting points are determined sequentially.

[0046] The set characteristics associated with each different set of undetermined bands are determined as follows: the overlap of adjacent bands is compared to confirm the overlapping bands between adjacent bands, and the proportion of overlapping bands located in adjacent bands is confirmed again (the proportion = the bus length of the overlapping band ÷ the bus length of the corresponding band). The two sets of proportions are averaged to confirm the overlap rate of this adjacent band. Then, the several sets of overlap rates confirmed in this set of undetermined bands are averaged to lock the set characteristics of this set of undetermined bands.

[0047] The different set characteristics associated with different undetermined band sets are confirmed in sequence, and the minimum value is selected. The undetermined band set associated with it is labeled as the band set of a single operating cycle. Specifically, due to the difference in the rolling characteristics associated with the corresponding rolling process, the corresponding bands will be different. After relevant verification and checking, the band set with a large overlap rate can be locked, which is convenient for subsequent feature analysis and confirmation.

[0048] Step 2: Based on the band set associated with a single operating cycle, perform overlap verification on each pair of bands within the band set, identifying the two sets of bands with the largest difference in overlap. These are then selected as the bands to be analyzed. Specifically, during overlap verification, the two bands with the largest differences are those with significant differences in overall characteristics. Therefore, feature extraction and analysis are performed from these bands with large feature differences. This ensures the comprehensiveness of the feature extraction process, thereby achieving more accurate feature analysis results. The specific method for confirming the bands to be analyzed is as follows:

[0049] The bands existing in the band set are combined in pairs to generate multiple combined band columns, and each combined band column is unique;

[0050] The overlap of two bands in the combined band column is checked to identify the overlapping bands and determine the proportion of the overlapping band in either band. The proportion is calculated as the length of the overlapping band divided by the length of the band in either band. The two proportions are then averaged to confirm the verification mean. The confirmed verification mean is used as the correlation feature of this combined band column.

[0051] Based on the different correlation characteristics corresponding to different combination band columns, the minimum value is selected, and the two bands of the combination band column associated with the minimum value are both marked as bands to be analyzed. The two bands to be analyzed here are quite different in characteristics, so subsequent feature verification analysis can fully ensure the accuracy of the monitoring process.

[0052] Step 3: Perform wavelet packet transform on the confirmed bands to be analyzed, decompose them into different frequency bands, and perform energy confirmation. Based on the specific energy determination process, lock the characteristic bands. The specific method for locking is as follows:

[0053] The db4 wavelet basis is used to perform three-level wavelet packet decomposition on the band to be analyzed, so that the band to be analyzed is decomposed into frequency bands corresponding to different frequencies, with the frequency ranges being: 0-62.5kHz, 62.5-125kHz, 125-187.5kHz, 187.5-250kHz, 250-312.5kHz, 312.5-375kHz, 375-437.5kHz, and 437.5-500kHz.

[0054] Confirm the signal energy LN associated with the corresponding frequency band. i Where i represents different frequency bands, and the total energy E associated with this band to be analyzed is confirmed using: P i =(LN i The percentage of energy belonging to the corresponding frequency band is determined by (E) × 100%. i The energy proportion P of multiple frequency bands associated with a single band to be analyzed i Confirm in sequence and select P. i The frequency band associated with max is used as the characteristic frequency band of this band to be analyzed.

[0055] Using the same processing method, the characteristic frequency bands associated with another set of bands to be analyzed were also confirmed, and it was determined whether the frequency intervals associated with the two sets of characteristic frequency bands were the same:

[0056] If they are the same, select the maximum value from the energy proportions associated with the two sets of characteristic frequency bands, and take the characteristic frequency band associated with the maximum value as the characteristic band.

[0057] If they are not the same, confirm the frequency intervals A1 and B2 associated with the two sets of characteristic frequency bands, then confirm the specific frequency band belonging to A1 in another set of bands to be analyzed, and simultaneously confirm the frequency band associated with B2. Perform energy difference processing on the energy proportions associated with the two sets of frequency bands synchronously belonging to A1, confirm the energy difference (if the energy difference is > 0), and then confirm the energy difference associated with the two sets of frequency bands synchronously belonging to B2. From the two confirmed energy difference values, select the maximum value (representing a large difference), and denote the two sets of frequency bands associated with the maximum value as the selected set. The characteristic frequency bands marked in the selected set are denoteed as characteristic bands. Specifically, the frequency interval associated with a characteristic frequency band is 312.5-3. 75kHz, another set of characteristic frequency bands are associated with a frequency range of 250-312.5kHz, another set of analysis bands also have a frequency range of 312.5-375kHz, and another set of analysis bands also have a frequency range of 250-312.5kHz. Therefore, the energy ratio of the two frequency bands belonging to the same analysis band can be confirmed, and the confirmed energy ratios can be processed by difference to lock the corresponding energy difference value. Then, based on the degree of difference in the corresponding energy difference values ​​associated with the two different analysis bands, the overall characteristic difference of the two frequency bands can be confirmed, thereby making a comprehensive confirmation of the characteristic bands.

[0058] Step 4: Based on the confirmed characteristic band, use the frequency band associated with this characteristic band as the search band, then lock the standard band associated with this search band, perform feature verification between the characteristic band and the standard band, and output the determined verification features for external relevant personnel to view.

[0059] The specific method for determining the verification features through feature verification is as follows:

[0060] The frequency band associated with the characteristic band is locked and recorded as the search band. The standard band associated with the search band is confirmed. The standard band is a preset band, which is determined in advance by relevant operators based on experience.

[0061] Perform co-frequency verification between the standard band and the characteristic band, lock the amplitude difference at the same frequency (if the amplitude difference is > 0), select the maximum value from the confirmed amplitude differences, and output the determined maximum value as the verification feature for external relevant personnel to view.

[0062] Specifically, when the verification characteristics are large, it means that the corresponding stress wave differs significantly from the waveform under standard conditions. In this case, the rolling bearing has related anomalies, which can be displayed.

[0063] Second Embodiment

[0064] Combination Figure 2 An online monitoring system for the condition of rolling bearings based on stress wave signal analysis includes:

[0065] The set of end points is determined by placing stress wave monitoring sensors at designated locations on the rolling bearing and monitoring the stress waves generated during the rotation of the rolling bearing. From the monitored stress waves, a set of wavebands for a single operating cycle is selected.

[0066] The calibration end of the band to be analyzed performs overlap verification on each pair of bands within the band set based on the band set associated with a single operating cycle, and locks the two sets of bands with the largest overlap as the bands to be analyzed.

[0067] The characteristic band calibration end performs wavelet packet transform on the confirmed band to be analyzed, decomposes it into different frequencies to obtain different frequency bands, and performs energy confirmation. Based on the specific energy determination process, the characteristic band is locked.

[0068] At the feature output end, based on the confirmed feature band, the frequency band associated with this feature band is used as the search segment, and then the standard band associated with this search segment is locked. The feature band and the standard band are verified, and the determined verification features are output.

[0069] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0070] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A method for online monitoring of rolling bearing condition based on stress wave signal analysis, characterized in that, Includes the following steps: A stress wave monitoring sensor is installed at a designated location on the rolling bearing to monitor the stress waves generated during the bearing's rotation. From the monitored stress waves, a set of wavebands for a single operating cycle is selected. The specific method is as follows: During the stress wave monitoring process, based on the current rotation speed of the rolling bearing, the characteristic time t1 associated with one rotation of the rolling bearing is confirmed, and the current rotation speed is maintained to rotate for N revolutions, where N is a preset value. The stress wave associated with N revolutions is confirmed from the monitoring data. From the confirmed stress waves, the initial time is calibrated to 0, and the band between 0 and t1 is recorded as the undetermined band; Randomly select a starting point from the undetermined bands, and starting from the starting point, select five bands with the same time length after t1, which are denoted as the undetermined band set. Then, determine the undetermined band sets associated with different starting points in sequence. The set characteristics associated with each different set of undetermined bands are determined as follows: the overlap of adjacent bands is compared to confirm the overlapping bands between adjacent bands, and the proportion of overlapping bands in adjacent bands is confirmed again. The two proportions are averaged to confirm the overlap rate of this adjacent band. Then, the several sets of overlap rates confirmed in this set of undetermined bands are averaged to lock the set characteristics of this set of undetermined bands. The characteristics of different sets associated with different undetermined band sets are confirmed in turn, and the minimum value is selected to mark the undetermined band set associated with it as a band set of a single operating cycle. Based on the band set associated with a single operating cycle, the overlap of each pair of bands within the band set is checked, and the two sets of bands with the largest difference in overlap are identified as the bands to be analyzed. Wavelet packet transform is performed on the identified bands to be analyzed, decomposing them into different frequency bands, and energy confirmation is performed. Based on the specific energy determination process, the characteristic bands are locked. Based on the confirmed characteristic band, the frequency band associated with this characteristic band is used as the search segment, and the standard band associated with this search segment is locked. The characteristic band and the standard band are then verified, and the determined verification features are output.

2. The online monitoring method for rolling bearing condition based on stress wave signal analysis according to claim 1, characterized in that, The specific method for determining the band to be analyzed is as follows: The bands existing in the band set are combined in pairs to generate multiple combined band columns, and each combined band column is unique; The overlap of two bands in the combined band column is checked to lock the overlapping bands and determine the proportion of the overlapping band in either band. The proportion is calculated as the length of the overlapping band divided by the length of the band in either band. The two proportions are then averaged to confirm the verification mean. The confirmed verification mean is used as the correlation feature of this combined band column. Based on the different correlation characteristics corresponding to different combination band columns, the minimum value is selected, and the two bands of the combination band column associated with the minimum value are both marked as bands to be analyzed.

3. The online monitoring method for rolling bearing condition based on stress wave signal analysis according to claim 1, characterized in that, The specific method for locking the characteristic band is as follows: The db4 wavelet basis is used to perform three-level wavelet packet decomposition on the band to be analyzed, so that the band to be analyzed is decomposed into frequency bands corresponding to different frequencies. Confirm the signal energy LN associated with the corresponding frequency band i Where i represents different frequency bands, and the total energy E associated with the to-be-analyzed wave band is further confirmed, and the energy proportion P belonging to the corresponding frequency band is determined by using: i = (LN i ÷ E) × 100% i The energy proportion P of the multiple frequency bands associated with a single to-be-analyzed wave band i is confirmed in sequence, and the frequency band associated with P i max is selected as the characteristic frequency band of the to-be-analyzed wave band; Using the same processing method, the characteristic frequency bands associated with another set of bands to be analyzed were also confirmed, and it was determined whether the frequency intervals associated with the two sets of characteristic frequency bands were the same: If they are the same, the maximum value is selected from the energy proportions associated with the two sets of characteristic frequency bands, and the characteristic frequency band associated with the maximum value is taken as the characteristic band.

4. The online monitoring method for rolling bearing condition based on stress wave signal analysis according to claim 3, characterized in that, If they are not the same, confirm the frequency intervals A1 and B2 associated with the two sets of characteristic frequency bands, then confirm the specific frequency segment belonging to A1 in another set of bands to be analyzed, and simultaneously confirm the frequency segment associated with B2. Perform energy difference processing on the energy proportions associated with the two sets of frequency segments that are synchronously associated with A1, and confirm the energy difference. If the energy difference is > 0, then confirm the energy difference associated with the two sets of frequency segments that are synchronously associated with B2. From the two confirmed energy difference values, select the maximum value, and record the two sets of frequency segments associated with the maximum value as the selected set. Record the characteristic frequency band marked in the selected set as the characteristic band.

5. The online monitoring method for rolling bearing condition based on stress wave signal analysis according to claim 3, characterized in that, The frequency ranges corresponding to the different frequency bands are: 0-62.5kHz, 62.5-125kHz, 125-187.5kHz, 187.5-250kHz, 250-312.5kHz, 312.5-375kHz, 375-437.5kHz, and 437.5-500kHz.

6. The online monitoring method for rolling bearing condition based on stress wave signal analysis according to claim 1, characterized in that, The specific method for determining the verification features is as follows: Lock the frequency band associated with the characteristic band and record it as the search segment. Confirm the standard band associated with the search segment, which is the preset band. Perform co-frequency verification between the standard band and the characteristic band, lock the amplitude difference at the same frequency (if the amplitude difference is greater than 0), select the maximum value from the confirmed amplitude differences, and output the determined maximum value as the verification feature.

7. An online monitoring system for the condition of rolling bearings based on stress wave signal analysis, wherein the system operates according to the online monitoring method for the condition of rolling bearings based on stress wave signal analysis as described in claims 1-6, characterized in that, include: The set of end points is determined by placing stress wave monitoring sensors at designated locations on the rolling bearing and monitoring the stress waves generated during the rotation of the rolling bearing. From the monitored stress waves, a set of wavebands for a single operating cycle is selected. The calibration end of the band to be analyzed performs overlap verification on each pair of bands within the band set based on the band set associated with a single operating cycle, and locks the two sets of bands with the largest overlap as the bands to be analyzed. The characteristic band calibration end performs wavelet packet transform on the confirmed band to be analyzed, decomposes it into different frequencies to obtain different frequency bands, and performs energy confirmation. Based on the specific energy determination process, the characteristic band is locked. At the feature output end, based on the confirmed feature band, the frequency band associated with this feature band is used as the search segment, and then the standard band associated with this search segment is locked. The feature band and the standard band are verified, and the determined verification features are output.

Images