Bearing fault diagnosis method and system based on data analysis

Abstract

The invention discloses a bearing fault diagnosis method and system based on data analysis, and particularly relates to the technical field of bearing fault diagnos.The method comprises the steps that a preliminary available coefficient is compared with a preset preliminary available coefficient threshold value, and the preliminary available condition of data of a vibration signal, collected by a vibration sensor, of a target bearing is judged; comparing the final available coefficient with a preset final available coefficient threshold value, if the final available coefficient is smaller than the preset final available coefficient threshold value, judging the final available condition of the data, and preprocessing the uploaded vibration signal of the final available data to obtain the final available data; the method comprises the steps of preprocessing a vibration signal of a target bearing, obtaining a vibration spectrum image set of the target bearing according to the preprocessed vibration signal, and performing fault diagnosis on the target bearing according to a fault type corresponding to a preset bearing vibration fault and the vibration spectrum image set, so that it can be determined that the collected data related to the bearing vibration fault is accurate; therefore, the diagnosis result of the current bearing vibration fault diagnosis system is relatively accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of bearing fault diagnosis, and in particular to a bearing fault diagnosis method and system based on data analysis. Background Art

[0002] Bearing vibration failure refers to abnormal vibration caused by various reasons during the operation of the bearing. This vibration can be detected and diagnosed by vibration sensors or other monitoring equipment installed on the bearing. Bearing vibration failure will increase the friction and wear of the bearing and shorten its service life. In addition, vibration may also cause damage or instability to surrounding equipment or structures, thereby affecting the operating stability of the entire production system. In severe cases, vibration failure may also cause sudden equipment shutdown, resulting in production interruption and additional maintenance costs. Therefore, it is crucial to diagnose bearing vibration failure in a timely manner and take timely treatment measures to ensure safe operation of equipment and production efficiency.

[0003] Normally, when diagnosing bearing vibration faults, the information about bearing vibration collected by the vibration sensor is basically transmitted to the bearing vibration fault diagnosis system. The bearing vibration fault diagnosis system analyzes the received information to determine whether the bearing vibration is faulty, and then takes corresponding measures in a timely manner according to the actual bearing vibration fault situation; However, in the above method, when using a bearing vibration fault diagnosis system to diagnose bearing vibration faults, fault diagnosis is often performed on the basis that the collected data related to the bearing vibration fault is accurate. If fault diagnosis is still performed when there is an error in the collected data related to the bearing vibration fault, it may cause problems in the diagnosis results of the current bearing vibration fault diagnosis system, resulting in errors in the final bearing vibration fault diagnosis results. Summary of the invention

[0004] The purpose of the present invention is to solve the problem that when using a bearing vibration fault diagnosis system to diagnose bearing vibration faults, fault diagnosis is often performed only on the basis that the collected data related to the bearing vibration fault is accurate. If fault diagnosis is still performed when there is an error in the collected data related to the bearing vibration fault, it may cause problems in the diagnosis results of the current bearing vibration fault diagnosis system, resulting in errors in the final bearing vibration fault diagnosis results. A bearing fault diagnosis method and system based on data analysis are proposed.

[0005] In a first aspect of the present invention, a bearing fault diagnosis method based on data analysis is first proposed, the method comprising: Obtain the current sensor position information and electromagnetic interference information of the target bearing, extract the preliminary available key features of the sensor data from the sensor position information and the electromagnetic interference information, analyze the key features, and determine the preliminary availability of the data; For the initially available data, obtain the data transmission information transmitted to the bearing vibration fault diagnosis system, and extract the final available key features of the sensor data from the data transmission information to determine the final availability of the data; For the final available data, the uploaded vibration signals are preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signals; The corresponding characteristic frequency is extracted according to the fault type corresponding to the preset bearing vibration fault, and the fault diagnosis of the target bearing is performed according to its corresponding characteristic frequency and vibration spectrum atlas.

[0006] Optionally, extracting preliminary available key features about the sensor data from the sensor location information and the electromagnetic interference information, and analyzing the key features to determine preliminary availability of the data includes: Generate a sensor position deviation index through sensor position information, generate an electromagnetic interference deviation index through electromagnetic interference information, and judge the preliminary availability of data based on the sensor position deviation index and the electromagnetic interference deviation index; The method for obtaining the sensor position deviation index is: Acquire the current actual position image of each vibration sensor in the target bearing to obtain an actual position image set; Overlap each image in the actual position image set with the corresponding image in the preset position image set, calculate the linear movement distance of the image position of each sensor, and obtain the linear movement distance of the position of the sensor; The linear movement distance of each sensor position in the target bearing is added to obtain the sensor position deviation index.

[0007] Optionally, the method for obtaining the electromagnetic interference deviation index is: Obtain the electromagnetic interference signal around the current target bearing when it is working, preprocess it, obtain the effective electromagnetic interference signal, and mark it as ; use Wavelet-assisted electromagnetic interference signal , perform wavelet transformation, the formula is:

[0008] In the formula, are the wavelet coefficients, and is the scale parameter and translation parameter, which is used to control the scaling and translation of the wavelet function. is the mother wavelet function, is its complex conjugate; Using wavelet coefficients Calculate the current electromagnetic interference index , the calculation formula is: ,in, Represents the energy of wavelet coefficients; Calculate the difference between the current electromagnetic interference index and the preset electromagnetic interference index threshold to obtain the current electromagnetic interference deviation index; The sensor position deviation index and the electromagnetic interference deviation index are weightedly summed to obtain a preliminary availability coefficient, and the preliminary availability coefficient is compared with a preset preliminary availability coefficient threshold. If the preliminary availability coefficient is less than the preset preliminary availability coefficient threshold, it is determined that the data collected by the sensor on the target bearing vibration fault diagnosis is preliminarily available.

[0009] Optionally, extracting key features of the final availability of the sensor data from the data transmission information, and determining the final availability of the data includes: The data transmission frequency consistency index and the data transmission completeness index are generated from the data transmission information, and the final availability of the data is judged according to the data transmission frequency consistency index and the data transmission completeness index, including: The method for obtaining the data transmission frequency consistency index is: The time corresponding to the data most recently uploaded to the bearing vibration fault diagnosis system by each vibration sensor in the target bearing is used to obtain a data upload time series; Calculate the absolute difference between any two data in the uploaded time series, and collect all the calculated absolute differences to obtain an absolute difference set; The data in the absolute difference set that is greater than the preset difference threshold is marked as 1, and the ratio of the number of 1s in the absolute difference set to the total number of data is calculated to obtain the data transmission frequency consistency index.

[0010] Optionally, the method for obtaining the data transmission completeness index is: Mark the data of each vibration sensor in the target bearing that was recently uploaded to the bearing vibration fault diagnosis system accordingly, and record it as an issued mark; Obtain all the marks of the data of the vibration sensor received by the bearing vibration fault diagnosis system, record them as receiving marks, and make one-to-one correspondence between the sending marks and the receiving marks, obtain the number of mismatches between the sending marks and the receiving marks, record them as the number of missing marks; Calculate the ratio of the number of missing tags to the number of sent tags to obtain the data transmission completeness index; The data transmission frequency consistency index and the data transmission completeness index are weightedly summed to obtain the final availability coefficient, and the final availability coefficient is compared with the preset final availability coefficient threshold. If the final availability coefficient is less than the preset final availability coefficient threshold, it is determined that the data related to the target bearing vibration fault diagnosis obtained by the bearing vibration fault diagnosis system is finally available.

[0011] Optionally, for the finally available data, the uploaded vibration signal is preprocessed, and a vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signal, including: Preprocess the corresponding vibration signal uploaded by each vibration sensor to obtain the time domain signal of the vibration sensor , for the time domain signal Perform fast Fourier transform to convert it into a complex signal in the frequency domain ,in Indicates frequency; Calculate the amplitude spectrum of a vibration signal , Represents the amplitude of the frequency domain signal; The frequency The amplitude of the corresponding vibration signal is plotted into a spectrum diagram, in which the horizontal axis represents the frequency and the vertical axis represents the amplitude of the vibration signal, and the spectrum diagram is used as the spectrum diagram of the sensor; Obtain the frequency spectra corresponding to all vibration sensors to obtain a frequency spectra set.

[0012] Optionally, extracting a corresponding characteristic frequency according to a fault type corresponding to a preset bearing vibration fault, and diagnosing a fault of a target bearing according to the corresponding characteristic frequency and a vibration spectrum atlas includes: The fault types corresponding to the preset bearing vibration faults include inner ring fault, outer ring fault, rolling element fault and spatial frequency fault; The calculation formula of the characteristic frequency corresponding to the inner race fault is: , where is the characteristic frequency corresponding to the inner race fault, is the diameter of the inner ring of the bearing; is the diameter of the bearing outer ring; is the rotational frequency of the bearing; The calculation formula of the characteristic frequency corresponding to the outer race fault is: , where is the characteristic frequency corresponding to the outer race fault, is the diameter of the inner ring of the bearing; is the diameter of the bearing outer ring; is the rotational frequency of the bearing; The calculation formula for the characteristic frequency corresponding to rolling element failure is: , where is the characteristic frequency corresponding to the rolling element fault, is the number of rolling elements in the bearing, is the rotational frequency of the bearing; The calculation formula of the characteristic frequency corresponding to the spatial frequency fault is: , where is the characteristic frequency corresponding to the spatial frequency fault, is the number of rods of the cage, is the rotation frequency of the cage; Fault diagnosis of the target bearing is performed based on the characteristic frequencies and vibration spectrum atlases corresponding to inner race faults, outer race faults, rolling element faults and spatial frequency faults.

[0013] Optionally, the fault diagnosis of the target bearing according to the characteristic frequencies and vibration spectrum atlases corresponding to the inner race fault, the outer race fault, the rolling element fault and the spatial frequency fault includes: For each spectrum graph in the vibration spectrum graph set, the energy values ​​of the characteristic frequencies and their multiples corresponding to the inner race fault, the outer race fault, the rolling element fault and the spatial frequency fault in the spectrum graph are obtained respectively to obtain several target energy values; All target energy values ​​in each spectrum graph are removed, the remaining energy values ​​are extracted, and the average value of the remaining energy values ​​is calculated to obtain the average energy value; Compare each target energy value with the average energy value. If the target energy value is greater than the average energy value, the vibration fault type of the target bearing is recorded as the vibration fault type. If the target energy value is not greater than the average energy value, it means that the target bearing does not have this vibration fault type; Obtain the bearing vibration fault types corresponding to all the spectrum graphs in the vibration spectrum graph set, and obtain all the bearing vibration fault types of the target bearing.

[0014] In a second aspect of the present invention, a bearing fault diagnosis system based on data analysis is also provided, the system comprising: Preliminary judgment module: obtains the current sensor position information and electromagnetic interference information of the target bearing, extracts the preliminary available key features of the sensor data from the sensor position information and electromagnetic interference information, analyzes the key features, and judges the preliminary availability of the data; Final judgment module: for the initially available data, obtain the data transmission information transmitted to the bearing vibration fault diagnosis system, extract the final available key features of the sensor data from the data transmission information, and judge the final availability of the data; Spectrum atlas module: For the final available data, the uploaded vibration signal is preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signal; Vibration fault diagnosis module: extract the corresponding characteristic frequency according to the fault type corresponding to the preset bearing vibration fault, and perform fault diagnosis on the target bearing according to its corresponding characteristic frequency and vibration spectrum atlas.

[0015] Beneficial effects of the present invention: The present invention proposes a bearing fault diagnosis method and system based on data analysis. When using a bearing vibration fault diagnosis system to diagnose bearing vibration faults, it can determine that the collected data related to the bearing vibration fault is accurate, so that the diagnosis result of the current bearing vibration fault diagnosis system is more accurate, and the probability of error in the final bearing vibration fault diagnosis result is lower. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The present invention will be further described below in conjunction with the accompanying drawings.

[0017] Figure 1 A flow chart of a bearing fault diagnosis method based on data analysis provided by an embodiment of the present invention; Figure 2 A framework diagram of bearing fault diagnosis based on data analysis provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0020] The embodiment of the present invention provides a bearing fault diagnosis method based on data analysis. Figure 1 , Figure 1 A flow chart of a bearing fault diagnosis method based on data analysis provided by an embodiment of the present invention. The method comprises the following steps: Obtain the current sensor position information and electromagnetic interference information of the target bearing, extract the preliminary available key features of the sensor data from the sensor position information and the electromagnetic interference information, analyze the key features, and determine the preliminary availability of the data; For the initially available data, obtain the data transmission information transmitted to the bearing vibration fault diagnosis system, and extract the final available key features of the sensor data from the data transmission information to determine the final availability of the data; For the final available data, the uploaded vibration signals are preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signals; The corresponding characteristic frequency is extracted according to the fault type corresponding to the preset bearing vibration fault, and the fault diagnosis of the target bearing is performed according to its corresponding characteristic frequency and vibration spectrum atlas.

[0021] Based on a bearing fault diagnosis method based on data analysis provided by an embodiment of the present invention, through the above-mentioned method, when using a bearing vibration fault diagnosis system to diagnose bearing vibration faults, it can be determined that the collected data related to the bearing vibration fault is accurate, so that the diagnosis result of the current bearing vibration fault diagnosis system is more accurate, and the probability of error in the final bearing vibration fault diagnosis result is lower.

[0022] It should be noted that the target bearing is a bearing that needs to be diagnosed for vibration fault.

[0023] In one embodiment, extracting preliminary available key features of sensor data from the sensor location information and the electromagnetic interference information, and analyzing the key features to determine preliminary availability of the data includes: Generate a sensor position deviation index through sensor position information, generate an electromagnetic interference deviation index through electromagnetic interference information, and judge the preliminary availability of data based on the sensor position deviation index and the electromagnetic interference deviation index; The method for obtaining the sensor position deviation index is: Acquire the current actual position image of each vibration sensor in the target bearing to obtain an actual position image set; Overlap each image in the actual position image set with the corresponding image in the preset position image set, calculate the linear movement distance of the image position of each sensor, and obtain the linear movement distance of the position of the sensor; The linear movement distance of each sensor position in the target bearing is added to obtain the sensor position deviation index.

[0024] In one implementation, the above method can accurately obtain the actual position movement degree of all sensors in the target bearing, and the obtained position deviation index of the sensor is more accurate; It should be noted that the preset position image set is an image set presented on an image based on the position of the vibration sensor determined when the vibration sensor on the actual target bearing can effectively collect data on the target bearing vibration fault diagnosis, which can be specifically determined by professional personnel; in addition, the positions of all vibration sensors in the target bearing can be obtained by high-precision camera equipment or other methods; It should be noted that the greater the position deviation of the vibration sensor, the greater the inaccuracy of the data collected for bearing vibration fault diagnosis, because the offset of the vibration sensor position will cause the characteristics of the vibration signal it receives to change, and it cannot accurately reflect the actual working state of the target bearing. This distortion will directly affect the analysis and interpretation of the data by the vibration fault diagnosis system.

[0025] In one embodiment, the electromagnetic interference deviation index is obtained by: Obtain the electromagnetic interference signal around the current target bearing when it is working, preprocess it, obtain the effective electromagnetic interference signal, and mark it as ; use Wavelet-assisted electromagnetic interference signal , perform wavelet transformation, the formula is:

[0026] In the formula, are the wavelet coefficients, and is the scale parameter and translation parameter, which is used to control the scaling and translation of the wavelet function. is the mother wavelet function, is its complex conjugate; Using wavelet coefficients Calculate the current electromagnetic interference index , the calculation formula is: ,in, Represents the energy of wavelet coefficients; The difference between the current electromagnetic interference index and the preset electromagnetic interference index threshold is calculated to obtain the current electromagnetic interference deviation index.

[0027] It should be noted that the preset electromagnetic interference index refers to the baseline value or threshold of electromagnetic interference in the environment surrounding the target bearing under ideal or normal working conditions. This preset value is predetermined by historical data, experimental data or expert experience, and is used to compare the current actual electromagnetic interference situation to determine whether the degree of electromagnetic interference in the current environment is within an acceptable range. In addition, when the current target bearing is working, the surrounding electromagnetic interference signal is obtained through a high-precision electromagnetic sensor installed near the target bearing.

[0028] It should be noted that the larger the electromagnetic interference deviation index, the more serious the impact on the signal collected by the vibration sensor will be, and the signal may even be unable to accurately reflect the actual vibration of the target bearing. Even if proper preprocessing, such as filtering and denoising, is performed, it may still be impossible to completely eliminate the impact of electromagnetic interference. The larger the electromagnetic interference deviation index, the more serious the impact on the signal collected by the vibration sensor is.

[0029] In one embodiment, judging the preliminary availability of data according to the sensor position deviation index and the electromagnetic interference deviation index includes: The sensor position deviation index and the electromagnetic interference deviation index are normalized to obtain the preliminary availability coefficient, which is calculated by the following formula:

[0030] In the formula, is the preliminary availability coefficient, and They are the sensor position deviation index and the electromagnetic interference deviation index, and are preset proportional coefficients of the sensor position deviation index and the electromagnetic interference deviation index, respectively, and and All are greater than 0; The preliminary availability coefficient is compared with a preset preliminary availability coefficient threshold value, and if the preliminary availability coefficient is less than the preset preliminary availability coefficient threshold value, it is determined that the data collected by the sensor and related to the target bearing vibration fault diagnosis are preliminary available; If the preliminary availability coefficient is not less than the preset preliminary availability coefficient threshold, it is determined that the data collected by the sensor related to the target bearing vibration fault diagnosis is preliminarily unavailable, and a first-level alarm is issued.

[0031] It should be noted that the preset preliminary availability coefficient threshold is set by professionals based on actual conditions and is not specifically limited; When the preliminary availability coefficient is not less than the preset preliminary availability coefficient threshold, it is determined that the data collected by the sensor related to the vibration fault diagnosis of the target bearing is preliminarily unavailable, and a first-level alarm is issued to remind the staff that the data collected by the vibration sensor for the vibration fault diagnosis of the target bearing is unavailable and needs to be checked in time and corresponding measures taken.

[0032] It should be noted that through the above method, the preliminary availability of the data collected by the sensor related to the target bearing vibration fault diagnosis can be systematically determined. This method combines the sensor position deviation index and the electromagnetic interference deviation index, and performs a comprehensive evaluation based on the preset proportional coefficient, which effectively improves the accuracy and reliability of data evaluation.

[0033] In one embodiment, extracting key features of the final availability of sensor data from the data transmission information and determining the final availability of the data includes: The data transmission frequency consistency index and the data transmission completeness index are generated from the data transmission information, and the final availability of the data is judged according to the data transmission frequency consistency index and the data transmission completeness index, including: The method for obtaining the data transmission frequency consistency index is: The time corresponding to the data most recently uploaded to the bearing vibration fault diagnosis system by each vibration sensor in the target bearing is used to obtain a data upload time series; Calculate the absolute difference between any two data in the uploaded time series, and collect all the calculated absolute differences to obtain an absolute difference set; The data in the absolute difference set that is greater than the preset difference threshold is marked as 1, and the ratio of the number of 1s in the absolute difference set to the total number of data is calculated to obtain the data transmission frequency consistency index.

[0034] It should be noted that the preset difference threshold is set by professionals based on actual conditions and is not specifically limited; It should be noted that in the process of collecting the vibration signal of the target bearing, the frequency uploaded by each vibration sensor should be consistent every time it uploads data to the bearing vibration fault diagnosis system. If the frequency is inconsistent, it is very likely that the vibration signal of the target bearing obtained by the bearing vibration fault diagnosis system will be inaccurate and cannot represent the actual vibration condition of the current target bearing. The reasons are: Data loss or duplication: If the frequency of sensor uploads is inconsistent, the system may lack data in certain time periods or duplicate data may appear in the same time period, thus affecting the continuous monitoring and fault diagnosis of bearing vibration.

[0035] Inaccurate data timing: The timing information of the vibration signal is crucial for fault diagnosis. Inconsistent frequencies will make the timing information of the data unreliable and unable to accurately reflect the actual change trend and periodic characteristics of the bearing vibration.

[0036] Therefore, maintaining the frequency consistency of vibration sensor data upload is crucial to ensure that the bearing vibration fault diagnosis system obtains accurate and reliable vibration signal data.

[0037] In one embodiment, a method for obtaining a data transmission completeness index is as follows: Mark the data of each vibration sensor in the target bearing that was recently uploaded to the bearing vibration fault diagnosis system accordingly, and record it as an issued mark; Obtain all the marks of the data of the vibration sensor received by the bearing vibration fault diagnosis system, record them as receiving marks, and make one-to-one correspondence between the sending marks and the receiving marks, obtain the number of mismatches between the sending marks and the receiving marks, record them as the number of missing marks; The ratio of the number of missing tags to the number of sent tags is calculated to obtain the data transmission completeness index.

[0038] It should be noted that when the bearing vibration fault diagnosis system obtains the data transmitted by each vibration sensor, the received data should be consistent with the data transmitted by the sensor. If they are inconsistent, it is very likely that the vibration signal of the target bearing obtained by the bearing vibration fault diagnosis system will be inaccurate and cannot represent the actual vibration condition of the current target bearing. The reason is that if the vibration signal collected by the sensor cannot be accurately transmitted to the fault diagnosis system, it may cause data loss or incompleteness. Such data loss may make the system unable to obtain complete vibration signal information, thereby affecting the accurate assessment of the bearing health status.

[0039] In one embodiment, judging the final availability of data according to the data transmission frequency consistency index and the data transmission completeness index includes: The data transmission frequency consistency index and the data transmission completeness index are normalized to obtain the final availability coefficient, which is calculated using the following formula:

[0040] In the formula, is the final available coefficient, and They are the data transmission frequency consistency index and the data transmission completeness index, and are preset proportional coefficients of the data transmission frequency consistency index and the data transmission completeness index, respectively, and and All are greater than 0; The final availability coefficient is compared with a preset final availability coefficient threshold value. If the final availability coefficient is less than the preset final availability coefficient threshold value, it is determined that the data related to the target bearing vibration fault diagnosis acquired by the bearing vibration fault diagnosis system is finally available. If the final availability coefficient is not less than the preset final availability coefficient threshold, it is determined that the data related to the target bearing vibration fault diagnosis acquired by the bearing vibration fault diagnosis system is ultimately unavailable, and a secondary alarm is issued.

[0041] It should be noted that the preset final availability coefficient threshold is set by professionals based on actual conditions and is not specifically limited; When the final availability coefficient is not less than the preset final availability coefficient threshold, it is determined that the data related to the target bearing vibration fault diagnosis obtained by the bearing vibration fault diagnosis system is ultimately unavailable, and an alarm is issued to remind the staff that the data related to the target bearing vibration fault diagnosis obtained by the bearing vibration fault diagnosis system is unavailable and needs to be checked in time and corresponding measures taken.

[0042] It should be noted that, through the above method, the final availability of the data can be effectively evaluated based on the frequency consistency and integrity of the sensor data, so as to carry out bearing vibration fault diagnosis and maintenance work in a timely and accurate manner.

[0043] In one embodiment, for the finally available data, the uploaded vibration signal is preprocessed, and a vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signal, including: Preprocess the corresponding vibration signal uploaded by each vibration sensor to obtain the time domain signal of the vibration sensor , for the time domain signal Perform fast Fourier transform to convert it into a complex signal in the frequency domain ,in Indicates frequency; Calculate the amplitude spectrum of a vibration signal , Represents the amplitude of the frequency domain signal; The frequency The amplitude of the corresponding vibration signal is plotted into a spectrum diagram, in which the horizontal axis represents the frequency and the vertical axis represents the amplitude of the vibration signal, and the spectrum diagram is used as the spectrum diagram of the sensor; Obtain the frequency spectra corresponding to all vibration sensors to obtain a frequency spectra set.

[0044] It should be noted that vibration signal preprocessing is to optimize the signal for subsequent analysis or diagnosis. The following are common vibration signal preprocessing steps: a. Data cleaning and preprocessing: Removal of outliers: Detect and remove outliers that may be caused by sensor errors or abnormal equipment operation; Interpolation processing: fill in the gaps caused by intermittent sensors or data loss to ensure the continuity of the time series; b. Noise removal and filtering: Time domain filtering: Use filters to remove high-frequency noise or low-frequency interference in vibration signals, for example: Moving average filtering: Smooth the signal and retain the trend of the signal; Median filtering: removes outliers and impulse noise and retains the center position of the signal; Wavelet transform: extract signal components of different scales and remove noise; Frequency domain filtering: select appropriate filters according to the spectrum characteristics, such as: low-pass filtering: remove high-frequency components; High-pass filter: remove low-frequency components; Bandpass filtering: limits the frequency components of a signal to a specific frequency band; c. Feature extraction Time domain features: Extract statistical features of vibration signals in the time domain, such as mean, variance, peak value, etc.

[0045] Frequency domain features: Use methods such as Fourier transform or wavelet transform to convert the vibration signal into the frequency domain and extract the frequency components and spectral energy.

[0046] Time-frequency domain features: Combine time domain and frequency domain information, such as wavelet packet analysis, to extract the time-varying characteristics of the signal.

[0047] d. Data normalization and standardization Normalization: Scaling data to a uniform range of values ​​to eliminate differences in values ​​between different sensors or devices.

[0048] Standardization: Adjust the mean and variance of the data to give it uniform statistical properties, facilitating the application and comparison of subsequent algorithms.

[0049] 3. Data Quality Assessment Data integrity check: Ensure the integrity of the preprocessed data, including data length, continuity, and missing value processing.

[0050] Signal quality assessment: Analyze the quality of the pre-processed signal to confirm whether it is suitable for subsequent vibration analysis and fault diagnosis.

[0051] Through the above preprocessing steps, the vibration signal can be effectively prepared for subsequent spectrum analysis, time domain analysis, pattern recognition or fault diagnosis, helping engineers and technicians to more accurately understand the operating status and health of the mechanical system. Other preprocessing methods can also be selected according to actual conditions and are not limited.

[0052] In one embodiment, extracting the corresponding characteristic frequency according to the fault type corresponding to the preset bearing vibration fault, and diagnosing the fault of the target bearing according to the corresponding characteristic frequency and the vibration spectrum atlas includes: The fault types corresponding to the preset bearing vibration faults include inner ring fault, outer ring fault, rolling element fault and spatial frequency fault; The calculation formula of the characteristic frequency corresponding to the inner race fault is: , where is the characteristic frequency corresponding to the inner race fault, is the diameter of the inner ring of the bearing; is the diameter of the bearing outer ring; is the rotational frequency of the bearing; The calculation formula of the characteristic frequency corresponding to the outer race fault is: , where is the characteristic frequency corresponding to the outer race fault, is the diameter of the inner ring of the bearing; is the diameter of the bearing outer ring; is the rotational frequency of the bearing; The calculation formula for the characteristic frequency corresponding to rolling element failure is: , where is the characteristic frequency corresponding to the rolling element fault, is the number of rolling elements in the bearing, is the rotational frequency of the bearing; The calculation formula of the characteristic frequency corresponding to the spatial frequency fault is: , where is the characteristic frequency corresponding to the spatial frequency fault, is the number of rods of the cage, is the rotation frequency of the cage; Fault diagnosis of the target bearing is performed based on the characteristic frequencies and vibration spectrum atlases corresponding to inner race faults, outer race faults, rolling element faults and spatial frequency faults.

[0053] It should be noted that inner race faults are usually manifested in the vibration spectrum as the energy peaks of the fundamental frequency (BPFI) of the ball bearing and its multiples. Outer race faults are usually manifested in the vibration spectrum as the energy peaks of the outer race fundamental frequency (BPFO) of the ball bearing and its multiples. Rolling element faults are usually manifested as the spinning frequency of the rolling element between the inner and outer races of the bearing. Spatial frequency is the vibration frequency caused by uneven cage or fault.

[0054] It should be noted that the bearing inner ring diameter is usually provided by the bearing specifications or manufacturer, or obtained by measuring the diameter of the bearing inner ring; the bearing outer ring diameter is also provided by the bearing specifications or manufacturer, or obtained by measuring the diameter of the bearing outer ring; the bearing rotation frequency is provided by measuring the bearing speed or by the working parameters of the equipment; the number of rolling elements is obtained by the design or specifications of the bearing, usually determined by the bearing model; the number of cage rods is obtained by the design or specifications of the bearing, usually determined by the bearing model.

[0055] In one embodiment, fault diagnosis of a target bearing according to characteristic frequencies and vibration spectrum atlases corresponding to inner race faults, outer race faults, rolling element faults, and spatial frequency faults includes: For each spectrum graph in the vibration spectrum graph set, the energy values ​​of the characteristic frequencies and their multiples corresponding to the inner race fault, the outer race fault, the rolling element fault and the spatial frequency fault in the spectrum graph are obtained respectively to obtain several target energy values; All target energy values ​​in each spectrum graph are removed, the remaining energy values ​​are extracted, and the average value of the remaining energy values ​​is calculated to obtain the average energy value; Compare each target energy value with the average energy value. If the target energy value is greater than the average energy value, the vibration fault type of the target bearing is recorded as the vibration fault type. If the target energy value is not greater than the average energy value, it means that the target bearing does not have this vibration fault type; Obtain the bearing vibration fault types corresponding to all the spectrum graphs in the vibration spectrum graph set, and obtain all the bearing vibration fault types of the target bearing.

[0056] It should be noted that the target bearing vibration fault type diagnosed by calculating the average energy value and comparing it with the target energy value can provide a more comprehensive, robust and accurate bearing vibration fault diagnosis. This method combines the advantages of overall spectrum analysis and can better support the implementation of predictive maintenance and equipment health management strategies.

[0057] Based on the same inventive concept, the embodiment of the present invention also provides a bearing fault diagnosis system based on data analysis. Figure 2 , Figure 2 A schematic diagram of a bearing fault diagnosis system based on data analysis provided by an embodiment of the present invention includes: Preliminary judgment module: obtains the current sensor position information and electromagnetic interference information of the target bearing, extracts the preliminary available key features of the sensor data from the sensor position information and electromagnetic interference information, analyzes the key features, and judges the preliminary availability of the data; Final judgment module: for the initially available data, obtain the data transmission information transmitted to the bearing vibration fault diagnosis system, extract the final available key features of the sensor data from the data transmission information, and judge the final availability of the data; Spectrum atlas module: For the final available data, the uploaded vibration signal is preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signal; Vibration fault diagnosis module: extract the corresponding characteristic frequency according to the fault type corresponding to the preset bearing vibration fault, and perform fault diagnosis on the target bearing according to its corresponding characteristic frequency and vibration spectrum atlas.

[0058] Based on a bearing fault diagnosis system based on data analysis provided by an embodiment of the present invention, through the above-mentioned method, when using a bearing vibration fault diagnosis system to diagnose bearing vibration faults, it can be determined that the collected data related to the bearing vibration fault is accurate, so that the diagnosis result of the current bearing vibration fault diagnosis system is more accurate, and the probability of error in the final bearing vibration fault diagnosis result is lower.

[0059] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A bearing fault diagnosis method based on data analysis, characterized in that: The following steps are involved: Obtain the current sensor position information and electromagnetic interference information of the target bearing, extract the preliminary available key features of the sensor data from the sensor position information and the electromagnetic interference information, analyze the key features, and determine the preliminary availability of the data; For the initially available data, obtain the data transmission information transmitted to the bearing vibration fault diagnosis system, and extract the final available key features of the sensor data from the data transmission information to determine the final availability of the data; For the final available data, the uploaded vibration signals are preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signals; Extract the corresponding characteristic frequency according to the fault type corresponding to the preset bearing vibration fault, and perform fault diagnosis on the target bearing according to its corresponding characteristic frequency and vibration spectrum atlas; The preliminary availability of the judgment data includes: Generate a sensor position deviation index through sensor position information, generate an electromagnetic interference deviation index through electromagnetic interference information, and judge the preliminary availability of data based on the sensor position deviation index and the electromagnetic interference deviation index; The method for obtaining the sensor position deviation index is: Acquire the current actual position image of each vibration sensor in the target bearing to obtain an actual position image set; Overlap each image in the actual position image set with the corresponding image in the preset position image set, calculate the linear movement distance of the image position of each sensor, and obtain the linear movement distance of the position of the sensor; Add the linear movement distance of each sensor in the target bearing to obtain the sensor position deviation index; The method for obtaining the electromagnetic interference deviation index is: Obtain the electromagnetic interference signal around the current target bearing when it is working, preprocess it, obtain the effective electromagnetic interference signal, and mark it as ; use Wavelet-assisted electromagnetic interference signal , perform wavelet transformation, the formula is: ； In the formula, are the wavelet coefficients, and is the scale parameter and translation parameter, which is used to control the scaling and translation of the wavelet function. is the mother wavelet function, is its complex conjugate; Using wavelet coefficients Calculate the current electromagnetic interference index , the calculation formula is: ,in, Represents the energy of wavelet coefficients; Calculate the difference between the current electromagnetic interference index and the preset electromagnetic interference index threshold to obtain the current electromagnetic interference deviation index; The sensor position deviation index and the electromagnetic interference deviation index are weighted and summed to obtain the preliminary usable coefficient. The preliminary availability coefficient is compared with a preset preliminary availability coefficient threshold value. If the preliminary availability coefficient is less than the preset preliminary availability coefficient threshold value, it is determined that the data collected by the sensor related to the target bearing vibration fault diagnosis is preliminary available.

2. A bearing fault diagnosis method based on data analysis according to claim 1, characterized in that: Extracting the key features of the final availability of sensor data from the data transmission information to determine the final availability of the data includes: The data transmission frequency consistency index and the data transmission completeness index are generated from the data transmission information, and the final availability of the data is judged according to the data transmission frequency consistency index and the data transmission completeness index, including: The method for obtaining the data transmission frequency consistency index is: The time corresponding to the data most recently uploaded to the bearing vibration fault diagnosis system by each vibration sensor in the target bearing is used to obtain a data upload time series; Calculate the absolute difference between any two data in the uploaded time series, and collect all the calculated absolute differences to obtain an absolute difference set; The data in the absolute difference set that is greater than a preset difference threshold is marked as 1, and the ratio of the number of 1s in the absolute difference set to the total number of data is calculated to obtain the data transmission frequency consistency index.

3. A bearing fault diagnosis method based on data analysis according to claim 2, characterized in that: The method for obtaining the data transmission completeness index is: Mark the data of each vibration sensor in the target bearing that was recently uploaded to the bearing vibration fault diagnosis system accordingly, and record it as an issued mark; Obtain all the marks of the data of the vibration sensor received by the bearing vibration fault diagnosis system, record them as receiving marks, and make one-to-one correspondence between the sending marks and the receiving marks, obtain the number of mismatches between the sending marks and the receiving marks, record them as the number of missing marks; Calculate the ratio of the number of missing tags to the number of sent tags to obtain the data transmission completeness index; The data transmission frequency consistency index and the data transmission completeness index are weightedly summed to obtain the final availability coefficient, and the final availability coefficient is compared with the preset final availability coefficient threshold. If the final availability coefficient is less than the preset final availability coefficient threshold, it is determined that the data related to the target bearing vibration fault diagnosis obtained by the bearing vibration fault diagnosis system is finally available.

4. A bearing fault diagnosis method based on data analysis according to claim 1, characterized in that: For the final available data, the uploaded vibration signal is preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signal, including: Preprocess the corresponding vibration signal uploaded by each vibration sensor to obtain the time domain signal of the vibration sensor , for the time domain signal Perform fast Fourier transform to convert it into a complex signal in the frequency domain ,in Indicates frequency; Calculate the amplitude spectrum of a vibration signal , Represents the amplitude of the frequency domain signal; The frequency The amplitude of the corresponding vibration signal is plotted into a spectrum diagram, in which the horizontal axis represents the frequency and the vertical axis represents the amplitude of the vibration signal, and the spectrum diagram is used as the spectrum diagram of the sensor; Obtain the frequency spectra corresponding to all vibration sensors to obtain a frequency spectra set.

5. The bearing fault diagnosis method based on data analysis according to claim 1 is characterized in that: According to the fault type corresponding to the preset bearing vibration fault, the corresponding characteristic frequency is extracted, and the fault diagnosis of the target bearing is performed according to the corresponding characteristic frequency and vibration spectrum atlas, including: The fault types corresponding to the preset bearing vibration faults include inner ring fault, outer ring fault, rolling element fault and spatial frequency fault; The calculation formula of the characteristic frequency corresponding to the inner race fault is: , where is the characteristic frequency corresponding to the inner race fault, is the diameter of the inner ring of the bearing; is the diameter of the bearing outer ring; is the rotational frequency of the bearing; The calculation formula of the characteristic frequency corresponding to the outer race fault is: , where is the characteristic frequency corresponding to the outer race fault, is the diameter of the inner ring of the bearing; is the diameter of the bearing outer ring; is the rotational frequency of the bearing; The calculation formula for the characteristic frequency corresponding to rolling element failure is: , where is the characteristic frequency corresponding to the rolling element fault, is the number of rolling elements in the bearing, is the rotational frequency of the bearing; The calculation formula of the characteristic frequency corresponding to the spatial frequency fault is: , where is the characteristic frequency corresponding to the spatial frequency fault, is the number of rods of the cage, is the rotation frequency of the cage; Fault diagnosis of the target bearing is performed based on the characteristic frequencies and vibration spectrum atlases corresponding to inner race faults, outer race faults, rolling element faults and spatial frequency faults.

6. A bearing fault diagnosis method based on data analysis according to claim 5, characterized in that: The fault diagnosis of the target bearing based on the characteristic frequencies and vibration spectrum atlases corresponding to the inner ring fault, outer ring fault, rolling element fault and spatial frequency fault includes: For each spectrum graph in the vibration spectrum graph set, the energy values ​​of the characteristic frequencies and their multiples corresponding to the inner race fault, the outer race fault, the rolling element fault and the spatial frequency fault in the spectrum graph are obtained respectively to obtain several target energy values; All target energy values ​​in each spectrum graph are removed, the remaining energy values ​​are extracted, and the average value of the remaining energy values ​​is calculated to obtain the average energy value; Compare each target energy value with the average energy value. If the target energy value is greater than the average energy value, the vibration fault type of the target bearing is recorded as the vibration fault type. If the target energy value is not greater than the average energy value, it means that the target bearing does not have this vibration fault type; Obtain the bearing vibration fault types corresponding to all the spectrum graphs in the vibration spectrum graph set, and obtain all the bearing vibration fault types of the target bearing.

7. A bearing fault diagnosis system based on data analysis, used to implement a bearing fault diagnosis method based on data analysis as claimed in any one of claims 1 to 6, characterized in that: include: Preliminary judgment module: obtains the current sensor position information and electromagnetic interference information of the target bearing, extracts the preliminary available key features of the sensor data from the sensor position information and electromagnetic interference information, analyzes the key features, and judges the preliminary availability of the data; Final judgment module: for the initially available data, obtain the data transmission information transmitted to the bearing vibration fault diagnosis system, extract the final available key features of the sensor data from the data transmission information, and judge the final availability of the data; Spectrum atlas module: For the final available data, the uploaded vibration signal is preprocessed, and the vibration spectrum atlas of the target bearing is obtained according to the preprocessed vibration signal; Vibration fault diagnosis module: extract the corresponding characteristic frequency according to the fault type corresponding to the preset bearing vibration fault, and perform fault diagnosis on the target bearing according to its corresponding characteristic frequency and vibration spectrum atlas.

Images