A clearance detection method and system for axle bearings

Abstract

The present application relates to the technical field of bearing clearance detection, and specifically to a clearance detection method and system for axle bearings, the method comprising: collecting the clearance upper limit data and clearance lower limit data of the bearing, constructing the clearance upper limit sequence and clearance lower limit sequence; obtaining the IMF component of the clearance upper limit sequence, and then obtaining the noise interference degree of each IMF component; obtaining the clearance upper limit subsequence and the clearance lower limit subsequence; obtaining the fluctuation similarity; obtaining the clearance trend comprehensive factor, and then obtaining the clearance consistency coefficient, and measuring the precise bearing clearance value. The present application improves the accuracy of axle bearing clearance detection.

Description

Technical Field

[0001] The present application relates to the technical field of bearing clearance detection, and in particular to a clearance detection method and system for axle bearings. Background Art

[0002] Bearings are key functional components of axles, and their operating status directly determines the safety, comfort and reliability of the entire vehicle. During the production process, the detection of bearing clearance is crucial. If the bearing clearance is too large, the number of rolling elements that bear the load at the same time will be reduced, and the load of a single rolling element will increase, thereby reducing the rotation accuracy of the bearing. If the bearing clearance is too small, the friction will increase, the heat generated will increase, the wear will be aggravated, and the service life of the bearing will be reduced. Therefore, it is necessary to accurately measure the bearing clearance.

[0003] Traditional bearing clearance measurement usually relies on precision testing instruments. These instruments measure the clearance size by driving the bearing to rotate. However, during measurement, changes in ambient temperature, load, and assembly may affect the measurement results, causing large fluctuations in clearance data, resulting in low reproducibility of axle bearing clearance test results, and the accuracy of the test also needs to be further improved. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of this application is to provide a clearance detection method and system for axle bearings. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present application provides a clearance detection method for an axle bearing, the method comprising the following steps:

[0006] Collect the upper limit data and lower limit data of the bearing clearance, and construct the upper limit sequence and lower limit sequence of the clearance;

[0007] The decomposition algorithm is used to obtain the IMF components of the clearance upper limit sequence; based on the abnormality of the data in the IMF components and the frequency domain characteristics of the IMF components, the noise interference degree of each IMF component is obtained;

[0008] Obtaining a clearance upper limit subsequence and a clearance lower limit subsequence based on the noise interference degree;

[0009] Based on the difference between the clearance upper limit subsequence and the clearance lower limit subsequence, the fluctuation similarity between each clearance upper limit subsequence and each clearance lower limit subsequence is obtained; based on the clearance upper limit subsequence and the trend change with the clearance upper limit subsequence, the clearance trend comprehensive factor between each clearance upper limit subsequence and each clearance lower limit subsequence is obtained;

[0010] Based on the fluctuation similarity and the clearance trend comprehensive factor, the clearance consistency coefficient between each pair of clearance upper limit subsequence and clearance lower limit subsequence is obtained;

[0011] Obtain precise bearing clearance values ​​based on the clearance consistency factor.

[0012] Furthermore, the method for obtaining the noise interference degree is:

[0013] Obtain the mutation clearness coefficient of each IMF component based on the abnormal degree of data in the IMF component;

[0014] The high-frequency influence intensity of each IMF component is obtained based on the frequency domain characteristics of the IMF component;

[0015] The product of the mutation distinctness coefficient and the high-frequency influence intensity is calculated as the noise interference degree of each IMF component.

[0016] Furthermore, the method for obtaining the mutation distinct coefficient is:

[0017] For the IMF component of the clearance upper limit sequence, the test algorithm is used to obtain the abnormal score of each data in the IMF component as the mutation significance coefficient of each data;

[0018] The mutation distinct coefficient is obtained according to the mutation significant coefficient. The mutation distinct coefficient of each IMF component is positively correlated with the mutation significant coefficient of all data in each IMF component.

[0019] Furthermore, the method for obtaining the high-frequency influence intensity is:

[0020] Use frequency domain conversion algorithm to obtain the frequency amplitude spectrum of each IMF component;

[0021] The product of each frequency and its corresponding amplitude in the frequency amplitude spectrum of the IMF component is calculated as the first amplitude-frequency product, and the sum of all first amplitude-frequency products in the IMF component is taken as the first amplitude-frequency sum value of the IMF component. The high-frequency impact intensity is obtained according to the first amplitude-frequency sum value, and the high-frequency impact intensity is positively correlated with the first amplitude-frequency sum value.

[0022] Further, the method for obtaining the clearance upper limit subsequence and the clearance upper limit subsequence is:

[0023] Arrange the noise interference degree of each IMF component from small to large to obtain a noise interference degree sequence, calculate the quartiles of the noise interference degree sequence, and use the set of IMF components corresponding to the noise interference degree greater than the third quartile as the denoised component set. Use the denoising algorithm to denoise the IMF components in the denoised component set to obtain the denoised IMF components;

[0024] For the clearance upper limit sequence, the IMF components corresponding to the noise interference degree less than or equal to the third quartile and all the denoised IMF components are reconstructed, and the reconstructed discrete signals are recorded as the clearance upper limit denoised signal sequence;

[0025] The same method as the clearance upper limit noise reduction signal sequence is used to obtain the clearance lower limit noise reduction signal sequence;

[0026] Evenly divide the clearance upper limit noise reduction signal sequence into a preset number of equal-length subsequences as each clearance upper limit subsequence;

[0027] The same method as the upper clearance limit subsequence is used to obtain the lower clearance limit subsequence.

[0028] Furthermore, the fluctuation similarity includes:

[0029] The degree of difference between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence is calculated as the fluctuation similarity between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence.

[0030] Further, the method for obtaining the comprehensive factor of the clearance trend is:

[0031] A test algorithm is used to obtain a trend test statistic of each clearance upper limit subsequence, and the absolute value of the trend test statistic is used as the statistical absolute value of each clearance upper limit subsequence;

[0032] The sum of the statistical absolute values ​​of all the clearance upper limit subsequences is calculated as the first sum, and the ratio of the statistical absolute value of each clearance upper limit subsequence to the first sum is used as the trend change weight of each clearance upper limit subsequence;

[0033] For each clearance upper limit subsequence, the product of the trend change weight of the clearance upper limit subsequence and the standard deviation of all elements in the clearance upper limit subsequence is used as the clearance upper limit trend factor of each clearance upper limit subsequence;

[0034] The same method as the upper clearance trend factor is used to obtain the lower clearance trend factor of each lower clearance subsequence in the lower clearance noise reduction signal sequence;

[0035] The sum of the upper limit clearance trend factor of each upper limit clearance subsequence in the upper limit clearance noise reduction signal sequence and the lower limit clearance trend factor of each lower limit clearance subsequence in the lower limit clearance noise reduction signal sequence is calculated as the comprehensive clearance trend factor between each upper limit clearance subsequence in the upper limit clearance noise reduction signal sequence and each lower limit clearance subsequence in the lower limit clearance noise reduction signal sequence.

[0036] Further, the clearance consistency coefficient is the inverse of the product of the fluctuation similarity and the clearance upper limit trend factor.

[0037] Further, obtaining the accurate bearing clearance value based on the clearance consistency coefficient includes:

[0038] Obtain a pair of clearance upper limit subsequence and clearance lower limit subsequence with the largest clearance consistency coefficient as a pair of accurate data subsequences;

[0039] The difference between the mean value of the elements in the upper clearance limit subsequence and the mean value of the elements in the lower clearance limit subsequence in the pair of accurate data subsequences is taken as the precise bearing clearance value.

[0040] In the second aspect, an embodiment of the present application also provides a clearance detection system for an axle bearing, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned clearance detection methods for an axle bearing when executing the computer program.

[0041] This application has at least the following beneficial effects:

[0042] This application performs EMD decomposition on the bearing clearance data, analyzes the high-frequency signal in each component, calculates the noise interference degree of each component, denoises the components significantly affected by noise, and then reconstructs each component, which has the beneficial effect of removing high-frequency noise. Based on the trend change characteristics of the divided subsequences, the clearance upper limit trend factor is calculated to judge the trend change degree of the subsequence. Combining the similarity relationship between different subsequences and the clearance upper limit trend factor, the clearance consistency coefficient is calculated to obtain the most accurate data virtualization in the bearing measurement. Its beneficial effect is that considering the factors such as friction and temperature changes in the bearing clearance measurement process, which are easy to interfere with the measurement data of the bearing clearance size, denoising is performed according to the relevant characteristics, thereby improving the accuracy of the axle bearing clearance detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 A flowchart of a clearance detection method for an axle bearing provided in accordance with an embodiment of the present application;

[0045] Figure 2A flow chart for obtaining a clearance consistency coefficient provided for one embodiment of the present application;

[0046] Figure 3 An IMF component frequency amplitude spectrum provided by one embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the clearance detection method and system for axle bearings proposed in accordance with the present application, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0048] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0049] The following is a detailed description of a specific scheme of a clearance detection method and system for axle bearings provided by the present application in conjunction with the accompanying drawings.

[0050] See also Figure 1 , which shows a flow chart of a clearance detection method for an axle bearing provided by an embodiment of the present application, the method comprising the following steps:

[0051] Step S1, obtaining detection data of axle bearing clearance.

[0052] The clearance is the distance that the inner ring and the outer ring of the bearing can move relative to each other during operation. This application collects the bearing clearance data to measure the radial clearance during operation. The intelligent sensing system includes a measuring instrument and a data reading system. This application sets the measuring instrument as a micrometer, and the measuring point is set on the inner ring of the bearing. After the measuring point is fixed, the bearing is rotated.

[0053] First, the data of the bearing rotating without load is collected. The intelligent sensing system reads the value of the micrometer in real time and records the data collected at this time as the lower limit of the clearance. Then, the load is applied to the bearing and the data collected at this time is recorded as the upper limit of the clearance.

[0054] In this embodiment, the time length for collecting the data of the clearance lower limit and the clearance upper limit is 1 minute, and the time interval for data collection is set to 0.1 second, thereby obtaining the clearance upper limit sequence and the clearance lower limit sequence when the bearing is running.

[0055] Step S2, perform EMD decomposition on the upper limit sequence of clearance, calculate the noise interference degree of each IMF based on the mutation data and high-frequency noise characteristics contained in the components; calculate the upper limit trend factor of clearance by analyzing the trend changes of the subsequence after removing part of the noise; calculate the clearance consistency coefficient based on the similarity and stability characteristics of the subsequences.

[0056] In the no-load state, the interaction force between the inner and outer rings and the rolling elements of the bearing is relatively small, resulting in a slightly larger internal clearance than in normal operation. When testing the bearing at this time, the smaller the movement of the bearing, the smaller the lower limit sequence of its internal clearance. After the load is applied, the interaction force between the inner and outer rings of the bearing and the rolling elements increases, making the internal clearance of the bearing smaller and the movement of the bearing larger. At this time, the measured upper limit sequence of clearance is correspondingly larger. The difference between the upper limit sequence of clearance and the lower limit sequence of clearance is the clearance measurement value of the bearing.

[0057] During the measurement process, the rotation of the bearing is not completely stable. For example, it is affected by the rough internal contact, vibration caused by friction, thermal expansion of the material caused by the increase in bearing temperature, etc., which may cause abnormal fluctuations in the measured clearance upper limit sequence and clearance lower limit sequence. The measurement accuracy of the bearing clearance is required to be high. These abnormal fluctuations will lead to inaccurate final calculation results. Therefore, it is necessary to remove noise from the measured data.

[0058] Taking the upper limit sequence of clearance as an example, the more abnormal fluctuations there are in the sequence, the more significant the change in amplitude. From the perspective of frequency domain, there are more high-frequency components in the sequence. However, analysis based only on the upper limit sequence value is likely to cause deviations in the measurement results.

[0059] This embodiment uses the EMD (Empirical Mode Decomposition) empirical mode decomposition algorithm to obtain the characteristic representation of each frequency. The input of the algorithm is the clearance upper limit sequence, the decomposition layer number is set to 8, and the output is the IMF component of the clearance upper limit sequence obtained by EMD decomposition. Among them, the EMD empirical mode decomposition algorithm is a well-known technology, and the specific process will not be repeated.

[0060] Different IMF components have different frequency characteristics. For the clearance upper limit sequence, its key signals are concentrated in the low frequency band. As the frequency increases, it contains fewer key signals and more noise. Therefore, it is necessary to denoise the first few layers of IMF components. However, the number of layers of IMF components that need to be denoised is uncertain. The following is an analysis of the IMF components to obtain the denoising level.

[0061] First, this embodiment uses a sliding t-test algorithm to perform mutation detection on each IMF component. The input of the algorithm is the IMF component, and the output is the abnormal score of each data in the IMF component as the mutation significance coefficient of each data. The sliding t-test is a well-known technology, and the specific process will not be repeated.

[0062] In order to further analyze the high-frequency signals of the IMF components, the Fourier transform algorithm is used to calculate the frequency amplitude spectrum of each IMF component. The more high-frequency signals there are in the IMF component, the larger the amplitude of the high frequency in the corresponding frequency amplitude spectrum will be. Thus, the noise interference degree of each IMF component is calculated. Among them, the frequency amplitude spectrum of the IMF component is as follows: Figure 3 shown.

[0063] The method for obtaining the mutation distinctness coefficient is as follows: the mutation distinctness coefficient of each IMF component is positively correlated with the mutation significance coefficient of all data in each IMF component.

[0064] Specifically, the present application proposes an embodiment in which the sum of the mutation significance coefficients of all data in the i-th IMF component is used as the mutation distinct coefficient of the i-th IMF component. The larger the mutation distinct coefficient is, the more mutation points may exist in the component.

[0065] The method for obtaining the high-frequency influence intensity is as follows: for the i-th IMF component, the product of each frequency and its corresponding amplitude in the frequency amplitude spectrum of the IMF component is calculated as the first amplitude-frequency product, and the sum of all the first amplitude-frequency products in the IMF component is taken as the first amplitude-frequency sum value of the IMF component. The high-frequency influence intensity is positively correlated with the first amplitude-frequency sum value.

[0066] Specifically, the present application proposes an embodiment in which a calculation result of a logarithmic function with the number 2 as the base and the first amplitude-frequency sum as a real number is used as the high-frequency influence intensity of the IMF component.

[0067] It should be noted that the positive correlation relationship can be an idempotent relationship of multiplication, addition, or exponential function, which can be determined based on actual conditions and will not be elaborated in this application.

[0068] Specifically, the product of the mutation distinctness coefficient and the high-frequency influence intensity is calculated as the noise interference degree of each IMF component.

[0069] The greater the high-frequency influence, the more common the high-frequency signal in this component. The greater the noise interference, the more obvious the noise signal in the signal, which can easily cause deviation in the final clearance measurement result and need to be denoised.

[0070] Since the noise interference degree better reflects the amount of noise contained in the IMF component, the greater the noise interference degree, the more noise is contained in the IMF component. Therefore, denoising is performed on several IMF components with larger noise interference degrees.

[0071] Specifically, the noise interference degree of each IMF component is arranged from small to large to obtain a noise interference degree sequence, the quartiles of the noise interference degree sequence are calculated, and the set of IMF components corresponding to the noise interference degree greater than the third quartile is used as the denoising component set, and the IMF components in the denoising component set are denoised using a wavelet transform denoising algorithm to obtain the denoised IMF components. Wavelet denoising is a well-known technology, and the specific process will not be repeated.

[0072] Further, all denoised IMF components are reconstructed with other IMF components, and the reconstructed discrete signal is recorded as the clearance upper limit denoised signal sequence. The process of reconstructing IMF is a well-known technology and will not be described in detail here.

[0073] The clearance upper limit noise reduction signal sequence removes some high-frequency noise in the clearance upper limit sequence, and it is necessary to further remove the trend noise characteristics in the clearance upper limit sequence. The trend noise characteristics may be caused by factors such as temperature changes, bearing types, and load instability during the detection process of the bearing clearance. In addition, this trend feature is easily reflected in the short-term local sequence.

[0074] Furthermore, the clearance upper limit noise reduction signal sequence is evenly divided into a preset number of equal-length subsequences as each clearance upper limit subsequence. In this embodiment, the preset number is 6.

[0075] A trend analysis is performed on each upper limit clearance subsequence. If the upper limit clearance subsequence has an obvious trend, it is considered that the data in the upper limit clearance subsequence is likely to affect the accuracy of the final measurement result.

[0076] The Mann-Kendall test algorithm is used to calculate the trend change degree of each clearance upper limit subsequence. This algorithm has the advantage of not being disturbed by a small number of outliers.

[0077] The input of the Mann-Kendall test algorithm is the clearance upper limit subsequence, and the output is the trend test statistic of each clearance upper limit subsequence. The absolute value of the trend test statistic is recorded as the statistical absolute value. The larger the statistical absolute value, the more obvious the trend of the clearance upper limit subsequence. The Mann-Kendall test algorithm is a well-known technology and will not be described in detail here.

[0078] The sum of the statistical absolute values ​​of all upper limit clearance subsequences is calculated as the first sum, and the ratio of the statistical absolute value corresponding to the jth upper limit clearance subsequence to the first sum is taken as the trend change weight of the jth upper limit clearance subsequence. The smaller the weight, the less the jth upper limit clearance subsequence is affected by the trend.

[0079] The product of the trend change weight of the jth clearance upper limit subsequence and the standard deviation of all elements in the jth clearance upper limit subsequence is used as the clearance upper limit trend factor of the jth clearance upper limit subsequence. The standard deviation reflects the degree of discreteness of the data. The smaller the calculated clearance upper limit trend factor is, the smaller the trend change degree of the subsequence is.

[0080] Through the above calculation, the upper limit clearance trend factor of each upper limit clearance subsequence in the upper limit clearance noise reduction signal sequence is obtained. For the lower limit clearance sequence, the same steps can be used to obtain the lower limit clearance trend factor of each lower limit clearance subsequence in the lower limit clearance noise reduction signal sequence.

[0081] When calculating the bearing clearance, since there is still unstable data in the clearance upper limit noise reduction signal sequence or the clearance lower limit noise reduction signal sequence, this embodiment selects any subsequence from the clearance upper limit noise reduction signal sequence and the clearance lower limit noise reduction signal sequence and calculates the clearance consistency coefficient, and obtains the subsequence pair for calculating the accurate bearing clearance value according to the clearance consistency coefficient.

[0082] The subsequences in the upper limit noise reduction signal sequence or the lower limit noise reduction signal sequence generally fluctuate around a certain value. If the similarity of the subsequences is higher, it means that the environmental conditions during the bearing clearance measurement are relatively stable, and the consistency of the clearance results is stronger.

[0083] Specifically, the process of obtaining the clearance consistency coefficient is as follows:

[0084] The clearance consistency coefficient is obtained based on the comprehensive factors of fluctuation similarity and clearance trend.

[0085] The method for obtaining the fluctuation similarity is: calculating the difference between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence as the fluctuation similarity between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence. The smaller the fluctuation similarity, the higher the similarity of the environmental conditions of the subsequences.

[0086] Specifically, the present application proposes an embodiment of calculating the DTW distance between the upper limit gap subsequence and the lower limit gap subsequence as the difference degree between the upper limit gap subsequence and the lower limit gap subsequence.

[0087] The method for obtaining the clearance trend comprehensive factor is as follows: the clearance upper limit trend factor of each clearance upper limit subsequence in the clearance upper limit noise reduction signal sequence is calculated, and the sum of the clearance lower limit trend factor of each clearance lower limit subsequence in the clearance lower limit noise reduction signal sequence is calculated, and the sum of the clearance upper limit trend factor of each clearance upper limit subsequence in the clearance upper limit noise reduction signal sequence and the clearance lower limit subsequence in the clearance lower limit noise reduction signal sequence is calculated as the clearance trend comprehensive factor between each clearance upper limit subsequence in the clearance upper limit noise reduction signal sequence and each clearance lower limit subsequence in the clearance lower limit noise reduction signal sequence. The smaller the clearance trend comprehensive factor is, the smaller the influence of the trend change of the subsequence is.

[0088] The clearance consistency coefficient is negatively correlated with the fluctuation similarity and the clearance upper limit trend factor.

[0089] It should be noted that the negative correlation relationship can be a subtraction relationship, a division relationship, etc., which can be determined according to actual conditions and will not be elaborated in this application.

[0090] Furthermore, the inverse of the product of the fluctuation similarity and the clearance upper limit trend factor is taken as the clearance consistency coefficient. The clearance consistency coefficient acquisition flow chart is as follows: Figure 2 shown.

[0091] Furthermore, the larger the obtained clearance consistency coefficient is, the more stable the clearance upper limit subsequences in the clearance upper limit noise reduction signal sequence and the clearance lower limit subsequences in the clearance lower limit noise reduction signal sequence are, and the more accurate the data in the subsequences are.

[0092] Step S3, calculating the bearing clearance data according to the subsequence obtained above.

[0093] A pair of upper clearance limit subsequence and lower clearance limit subsequence with the largest clearance consistency coefficient is obtained as a pair of accurate data subsequences. This pair of upper clearance limit subsequence and lower clearance limit subsequence has high data accuracy. The difference between the mean of the elements in the upper clearance limit subsequence and the mean of the elements in the lower clearance limit subsequence in the pair of accurate data subsequences is taken as the precise bearing clearance value to complete the clearance detection of the axle bearing.

[0094] Based on the same inventive concept as the above method, an embodiment of the present application also provides a clearance detection system for an axle bearing, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of any one of the above-mentioned clearance detection methods for axle bearings are implemented.

[0095] It should be noted that the above sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0096] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0097] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present application should be included in the protection scope of the present application.

Claims

1. A clearance detection method for axle bearings, characterized in that: The method comprises the following steps: Collect the upper limit data and lower limit data of the bearing clearance, and construct the upper limit sequence and lower limit sequence of the clearance; The decomposition algorithm is used to obtain the IMF components of the clearance upper limit sequence; based on the abnormality of the data in the IMF components and the frequency domain characteristics of the IMF components, the noise interference degree of each IMF component is obtained; Obtaining a clearance upper limit subsequence and a clearance lower limit subsequence based on the noise interference degree; Based on the difference between the clearance upper limit subsequence and the clearance lower limit subsequence, the fluctuation similarity between each clearance upper limit subsequence and each clearance lower limit subsequence is obtained; based on the clearance upper limit subsequence and the trend change with the clearance upper limit subsequence, the clearance trend comprehensive factor between each clearance upper limit subsequence and each clearance lower limit subsequence is obtained; Based on the fluctuation similarity and the clearance trend comprehensive factor, the clearance consistency coefficient between each pair of clearance upper limit subsequence and clearance lower limit subsequence is obtained; Obtain accurate bearing clearance values ​​based on the clearance consistency coefficient; The method for obtaining the clearance upper limit subsequence and the clearance upper limit subsequence is: Arrange the noise interference degree of each IMF component from small to large to obtain a noise interference degree sequence, calculate the quartiles of the noise interference degree sequence, and use the set of IMF components corresponding to the noise interference degree greater than the third quartile as the denoised component set. Use the denoising algorithm to denoise the IMF components in the denoised component set to obtain the denoised IMF components; For the clearance upper limit sequence, the IMF components corresponding to the noise interference degree less than or equal to the third quartile and all the denoised IMF components are reconstructed, and the reconstructed discrete signals are recorded as the clearance upper limit denoised signal sequence; The same method as the clearance upper limit noise reduction signal sequence is used to obtain the clearance lower limit noise reduction signal sequence; Evenly divide the clearance upper limit noise reduction signal sequence into a preset number of equal-length subsequences as each clearance upper limit subsequence; The same method as the upper limit subsequence of the clearance is used to obtain the lower limit subsequence of the clearance; The method for obtaining the clearance trend comprehensive factor is: A test algorithm is used to obtain a trend test statistic of each clearance upper limit subsequence, and the absolute value of the trend test statistic is used as the statistical absolute value of each clearance upper limit subsequence; The sum of the statistical absolute values ​​of all the clearance upper limit subsequences is calculated as the first sum, and the ratio of the statistical absolute value of each clearance upper limit subsequence to the first sum is used as the trend change weight of each clearance upper limit subsequence; For each clearance upper limit subsequence, the product of the trend change weight of the clearance upper limit subsequence and the standard deviation of all elements in the clearance upper limit subsequence is used as the clearance upper limit trend factor of each clearance upper limit subsequence; The same method as the upper clearance trend factor is used to obtain the lower clearance trend factor of each lower clearance subsequence in the lower clearance noise reduction signal sequence; Calculate the sum of the upper clearance trend factor of each upper clearance subsequence in the upper clearance noise reduction signal sequence and the lower clearance trend factor of each lower clearance subsequence in the lower clearance noise reduction signal sequence as the clearance trend comprehensive factor between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence; The clearance consistency coefficient is the inverse of the product of the fluctuation similarity and the clearance upper limit trend factor; The method of obtaining an accurate bearing clearance value based on the clearance consistency coefficient includes: Obtain a pair of clearance upper limit subsequence and clearance lower limit subsequence with the largest clearance consistency coefficient as a pair of accurate data subsequences; The difference between the mean value of the elements in the upper clearance limit subsequence and the mean value of the elements in the lower clearance limit subsequence in the pair of accurate data subsequences is taken as the precise bearing clearance value.

2. A clearance detection method for axle bearings as claimed in claim 1, characterized in that: The method for obtaining the noise interference degree is: Obtain the mutation clearness coefficient of each IMF component based on the abnormal degree of data in the IMF component; The high-frequency influence intensity of each IMF component is obtained based on the frequency domain characteristics of the IMF component; The product of the mutation distinctness coefficient and the high-frequency influence intensity is calculated as the noise interference degree of each IMF component.

3. A clearance detection method for axle bearings as claimed in claim 2, characterized in that: The method for obtaining the mutation distinct coefficient is: For the IMF component of the clearance upper limit sequence, the test algorithm is used to obtain the abnormal score of each data in the IMF component as the mutation significance coefficient of each data; The mutation distinct coefficient is obtained according to the mutation significant coefficient. The mutation distinct coefficient of each IMF component is positively correlated with the mutation significant coefficient of all data in each IMF component.

4. A clearance detection method for axle bearings as claimed in claim 2, characterized in that: The method for obtaining the high-frequency influence intensity is: Use frequency domain conversion algorithm to obtain the frequency amplitude spectrum of each IMF component; The product of each frequency and its corresponding amplitude in the frequency amplitude spectrum of the IMF component is calculated as the first amplitude-frequency product, and the sum of all first amplitude-frequency products in the IMF component is taken as the first amplitude-frequency sum value of the IMF component. The high-frequency impact intensity is obtained according to the first amplitude-frequency sum value, and the high-frequency impact intensity is positively correlated with the first amplitude-frequency sum value.

5. A clearance detection method for axle bearings as claimed in claim 1, characterized in that: The fluctuation similarity includes: The degree of difference between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence is calculated as the fluctuation similarity between each upper clearance subsequence in the upper clearance noise reduction signal sequence and each lower clearance subsequence in the lower clearance noise reduction signal sequence.

6. A clearance detection system for axle bearings, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the clearance detection method for an axle bearing as described in any one of claims 1 to 5 are implemented.

Images