A method and system for testing the compatibility of bearing raceway surface quality and lubricating medium

Abstract

The application relates to a bearing raceway surface quality matching test method and system of a lubricating medium, belongs to the technical field of performance testing of mechanical basic parts, and also belongs to the field of tribology engineering. In particular, a test method based on multi-physical field signal fusion and Stribeck curve normalization modeling is provided, a matching relationship between surface roughness (Ra), waviness (Wa), texture direction angle (theta) and rheological properties of lubricating grease is quantified, a dynamic lubrication state regulation model is established, bearing manufacturing process optimization and lubricating grease selection are guided, and the problem of vibration noise fault testing of a new energy automobile driving motor bearing under a working condition of a rotating speed exceeding 10,000 r / min is solved. The method covers multiple technical links of surface topography measurement, lubricating grease performance testing, high-speed signal acquisition, feature extraction, modeling analysis and optimization regulation.

Description

Technical Field

[0001] This invention provides a method and system for testing the matching of bearing raceway surface quality and lubricating medium, belonging to the field of mechanical basic component performance testing technology; in particular, it relates to a method for testing the matching of bearing raceway surface quality and lubricating grease based on the Stribeck curve lubrication theory. Background Technology

[0002] As the "joints" of rotating machinery, the vibration performance of rolling bearings directly determines the precision, efficiency, lifespan, and reliability of the main equipment. With the rapid development of high-tech equipment such as new energy vehicles, high-speed electric spindles, and aero engines, the operating speed of bearings is constantly breaking through limits. The minor contradictions under the national standard vibration standard (1800 r / min) are drastically amplified under ultra-high speed conditions (speed ≥ 8000 r / min), the most prominent of which are abnormal vibration and noise problems, which seriously affect the quietness, ride comfort, and operational reliability of the entire machine.

[0003] The vibration and noise of bearings are not caused by a single factor, but rather by the dynamic rheological behavior of the micro-geometry of the bearing raceway working surface and the lubrication state at the contact interface under extreme shear conditions. This is a concentrated manifestation of the mismatch in complex interfacial interactions. In the enormous centrifugal force, shear force field, and temperature field generated by high-speed rotation, the lubrication state undergoes a significant "shear thinning" effect, with its apparent viscosity η and the lubrication film at the contact interface varying with the shear rate. The film thickness decreases sharply in a power-law manner. This directly leads to a significant reduction in the minimum oil film thickness calculated according to the elastohydrodynamic lubrication theory, with the film thickness ratio rapidly dropping from the safe zone to the mixed lubrication zone or even the boundary lubrication zone.

[0004] The deterioration of lubrication conditions triggers a chain reaction: in the mixed lubrication zone, the rolling elements and the micro-protrusions on the raceway surface engage in intermittent direct contact; in the boundary lubrication zone, continuous solid-solid contact occurs. This contact not only excites broadband mechanical vibrations and aerodynamic noise, but also induces a typical tribological instability phenomenon—stick-slip motion. The lubrication state at the contact interface periodically switches between static and dynamic friction, generating transient stress waves with frequencies as high as tens to hundreds of kHz (detectable by acoustic emission technology). Continuous stick-slip and micro-impact are the direct causes of early micro-pitting, wear, and even galling failure on the raceway surface, thus forming a vicious cycle of "poor lubrication → increased vibration → surface damage → further deterioration of lubrication conditions."

[0005] The raceway surface, serving as the "base" for carrying and transporting the lubricating medium, plays a decisive role in regulating the aforementioned processes due to its three-dimensional morphology. The surface morphology of the bearing raceway (wrinkle, roughness, surface topography) directly affects the microscopic oil storage capacity and oil film formation efficiency. However, pursuing excessively low roughness (Ra < 0.05 μm) to reduce friction may actually lead to worse noise performance at high speeds, as an overly smooth surface weakens the lubrication effect and is detrimental to the formation and maintenance of a stable oil film. Surface wrinkle reflects the mid-frequency profile component between roughness and shape error. Excessive surface wrinkle can periodically disturb the already thin lubricating oil film like "waves," easily triggering specific mid-to-low frequency "roaring" noises related to the wrinkle characteristic frequency. The surface texture direction angle θ (the angle between the machining texture direction and the rolling direction) regulates the flow, distribution, and replenishment of the lubricating medium in the contact area. When the texture direction is parallel to the rolling direction (θ≈0°), the grease is easily "scraped" away from the contact area quickly; when it is perpendicular to the rolling direction (θ≈90°), the grease has good oil retention but may increase frictional resistance. Therefore, by controlling the surface processing technology of bearing components, the surface roughness, surface waviness, and surface morphology can be kept within an appropriate range. This usually achieves the best balance between the transport, storage, and tribological properties of the lubricating medium at the bearing contact interface, resulting in optimal overall performance.

[0006] Therefore, in bearing vibration and noise control, the rheological properties of the lubricating medium and the surface morphology of the raceway are deeply coupled into an inseparable "tribological system." A high-performance lubricating medium may lose all its advantages if matched with an unsuitable surface morphology; conversely, a raceway with ultra-precision machining may perform worse than a conventional product if an incompatible lubricating medium is used. However, the current serious reality facing industry and academia is the complete lack of a standardized method and apparatus for scientifically, quantitatively, and online testing and evaluation of this "surface morphology-grease" system-level matching characteristics.

[0007] Chinese invention patent application CN104391046A discloses a method for determining the optimal grease amount for rolling bearings with acoustic emission. This method involves detecting the acoustic emission signal of the rolling bearing during operation and extracting its time-domain parameters. Then, by changing the amount of grease added, the method obtains the correspondence between different grease amounts and the extracted acoustic emission signal time-domain parameters. Next, by changing the rotational speed of the shaft, the method obtains the correspondence between the grease amount and the extracted acoustic emission signal time-domain parameters at different rotational speeds. Finally, by comparing the correspondence between the grease amount and the extracted acoustic emission signal time-domain parameters at different rotational speeds, the optimal grease addition range for the rolling bearing is obtained.

[0008] Chinese invention patent application CN105899945A discloses a viscosity estimation method based on demodulated acoustic emission. This method monitors at least one of the bearing's speed, load, and temperature; demodulates the acoustic emission signal from the acoustic emission sensor; calculates and stores the root mean square (RMS) value of the demodulated acoustic emission estimate; aligns the RMS value with one or more values ​​of speed, load, and temperature and stores the alignment value; and uses the aligned RMS value and at least one of speed, load, and temperature to model the viscosity and / or viscosity ratio from the alignment value. By doing so, the condition of the lubricated bearing is monitored, and multivariate regression analysis is used to estimate the lubricant viscosity and model the bearing's expected life.

[0009] Chinese invention patent application CN107702919A discloses a method for monitoring the lubrication status of sliding bearings based on acoustic emission. This method acquires acoustic emission signals from the sliding bearing under different operating conditions and faulty lubrication states. It extracts the signal feature values ​​of each acoustic emission signal, fuses these signal feature values ​​with process information to calculate corresponding fused feature values, and constructs a fault sample database. The method then monitors the acoustic emission signals in the sliding bearing under monitoring in real time, extracts the signal feature values ​​of the real-time monitored acoustic emission signals, calculates the corresponding fused feature values, matches the real-time calculated fused feature values ​​with the corresponding fused feature values ​​in the fault sample database, and determines the real-time lubrication status of the sliding bearing based on the matching results.

[0010] None of the above solutions disclose a method that can quantitatively evaluate the matching degree between bearing raceway surface quality and lubrication medium based on vibration and noise, and can be used to guide bearing optimization.

[0011] Currently, existing bearing vibration and noise evaluation standards are mainly based on the measurement of effective values ​​of vibration acceleration under medium- and low-speed conditions (such as 1800 r / min). This method cannot sensitively capture the characteristics of lubrication state transitions at high speeds, nor can it distinguish whether the vibration source originates from imbalance, misalignment, or lubrication failure. Furthermore, basic tribological studies often employ simplified models (such as ball-disc or four-ball testing machines), whose contact patterns, stress states, and kinematics differ significantly from actual bearings, resulting in low reliability when extrapolated to practical engineering applications. Surface morphology research and lubricant research are often pursued in parallel, with few instances of collaborative experiments at the bearing system level.

[0012] Chinese invention patent application CN121475689A discloses a method for evaluating the embedded lubrication state of rolling bearings. This method includes: concatenating the temperature signal of the rolling bearing with a frequency domain spectral amplitude sequence to generate a multi-source spectral fusion feature vector; using a random forest model to filter the mean features extracted from the temperature signal, and the time-domain and frequency-domain features extracted from the vibration signal, sound signal, and acoustic emission signal to obtain a sensitive feature set; inputting the sensitive feature set into a support vector machine model to output a first lubrication state evaluation result; inputting the multi-source spectral fusion feature vector into a MobileNet V2 model to output a second lubrication state evaluation result; and fusing the first and second lubrication state evaluation results using DS evidence theory to obtain the final lubrication state evaluation result for the rolling bearing. Although this scheme combines vibration, sound, and acoustic emission signals, it relies on a deep learning model, and the accuracy of the results depends heavily on the model design and training, making it difficult to guarantee accuracy.

[0013] Currently, there is no universally accepted mathematical model based on physical mechanisms that can integrate parameters such as surface roughness, waviness, and surface texture with the η (variable of the lubricating medium). The friction coefficient performance spectrum is quantitatively correlated with the vibration, noise, and stick-slip signals exhibited during bearing operation, forming a calculable "matching degree" index.

[0014] In summary, the industry urgently needs an innovative testing method that can: 1) operate bearings under simulated real high-speed service conditions; 2) synchronously and with high fidelity acquire multi-source signals reflecting the system's macroscopic dynamics (vibration), microscopic interface behavior (stick-slip), and final acoustic output (noise); 3) combine precise offline surface morphology measurements with lubricating medium rheological data; 4) establish a normalized physical model for comprehensive evaluation of system matching based on Stribeck's lubrication theory; and 5) output intuitive matching quantification indicators and clear optimization directions. This is not only an engineering necessity for improving bearing product performance and overcoming the vibration and noise challenges of high-end equipment, but also a key to filling the theoretical gap in tribological system testing and evaluation. This invention is proposed against this backdrop. Summary of the Invention

[0015] The purpose of this invention is to provide a method and system for testing the matching of bearing raceway surface quality and lubrication medium, so as to solve the problem of how to quantitatively evaluate the matching degree of bearing raceway surface quality and lubrication medium.

[0016] To achieve the above objectives, the first aspect of the present invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium, comprising the following steps:

[0017] 1) Obtain the arithmetic mean height of the raceway surface of the bearing under test. Surface texture direction angle Arithmetic mean deviation of raceway circumferential waviness ;

[0018] 2) Obtain the test sample during high-speed operation testing: vibration characteristics obtained from fault characteristic frequencies and stick-slip characteristics obtained from acoustic emission signals; fill the bearing under test with the test lubricating medium to obtain the test sample;

[0019] 3) Calculate the normalized friction coefficient under operating conditions and comprehensive matching degree function :

[0020]

[0021] in, The average coefficient of friction is obtained by testing the friction performance of the lubricating medium under friction conditions close to high-speed operation test conditions, where U is the entrainment velocity in the rolling contact zone. To determine the apparent viscosity of the lubricating medium under high-speed testing, This represents the maximum Hertzian contact stress.

[0022] Equal to the lubrication potential parameter of the sample under test and the normalized vibration characteristic term The product of the speed severity factor With normalized stick-slip characteristic term The ratio of the products;

[0023] Lubrication potential parameters are 3D roughness adaptation function ripple influence function and texture direction adaptation function The product; exist It reaches its maximum value when it is within the set value, and decreases when it deviates from the set value; and Negative correlation; exist It is at its maximum when it is within the set degree, and decreases when it deviates from it; Negatively correlated with vibration characteristics Positively correlated with rotational speed; Positively correlated with sticky-slippery characteristics;

[0024] 4) smaller and The larger the value, the better the match between the bearing under test and the lubricating medium under test, and the matching test results of the sample under test can be obtained accordingly.

[0025] The first aspect of this invention provides a method for testing the matching of bearing raceway surface quality with lubricating medium. In one possible implementation, step 1) further obtains the average friction coefficient of the bearing steel balls under several friction conditions when rubbing the raceway under lubrication by the lubricating medium, obtained during friction performance testing of the lubricating medium under the test. In step 3), the average friction coefficient obtained under friction conditions close to the current test conditions during the friction performance test is used as... .

[0026] The first aspect of the present invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, the friction condition includes a combination of contact stress and sliding speed.

[0027] The first aspect of this invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, step 1) further involves obtaining the rheological relationship between apparent viscosity and shear rate obtained from rheological property testing of the lubricating medium under test; step 3) involves estimating the shear rate of the contact area at high-speed test rotation and substituting it into the rheological relationship to obtain... .

[0028] The first aspect of this invention provides a method for testing the surface quality of bearing raceways and the compatibility of lubricating media. In one possible implementation, in step 2), a radial vibration velocity signal is acquired at the midpoint of the axial direction perpendicular to the outer cylindrical surface of the outer ring of the bearing sample under test, and the acceleration signal is obtained by differentiation. The acceleration signal is then high-pass filtered to retain the high-frequency resonance band, and then Hilbert transform is performed to obtain the envelope signal. The envelope signal is then subjected to FFT to obtain the envelope spectrum. Based on the bearing geometric parameters and real-time rotational speed, the theoretical fault characteristic frequency is calculated. From the envelope spectrum, the characteristic frequency and its amplitude at the second and third harmonics are extracted to construct a vibration feature vector containing the vibration characteristics.

[0029] The first aspect of this invention provides a method for testing the matching of bearing raceway surface quality and lubrication medium. In one possible implementation, in step 2), an acoustic emission sensor is installed on the outer ring end face of the bearing sample to collect the original voltage signal. The original voltage signal is bandpass filtered to focus on the high-frequency stress wave components generated by stick-slip and micro-impact. The bandpass filtered voltage signal is subjected to continuous wavelet transform to obtain its time-frequency energy distribution. The total energy of wavelet coefficients in several preset different characteristic frequency bands and the number of times the energy exceeds the set energy threshold in the CWT time-frequency graph are used as the stick-slip feature to form a stick-slip feature vector.

[0030] The first aspect of this invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, step 3) involves:

[0031] The set value is 1.0;

[0032] );

[0033] ;

[0034] .

[0035] The first aspect of this invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, step 3) involves... ),in For the i-th component of the vibration feature vector, It is its corresponding weighting coefficient.

[0036] The first aspect of this invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, step 3) involves... ,in It is the stick-slip feature vector The j-th component, Its corresponding weighting coefficient.

[0037] The first aspect of this invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, step 4) involves...

[0038] when Less than the calibration value and Greater than the calibration value If so, it is determined that the bearing under test and the lubricating medium under test are in an optimal matching state;

[0039] when Greater than the calibration value Less than the calibration value ,and Less than the calibration value Greater than the calibration value If the bearing under test and the lubricating medium under test are in a good matching state, then it is determined that the bearing under test and the lubricating medium under test are in a good matching state.

[0040] when Greater than the calibration value Less than the calibration value ,or Less than the calibration value Greater than the calibration value If so, it is determined that the bearing under test and the lubricating medium under test are in a critical matching state;

[0041] when Greater than the calibration value and Less than the calibration value If the bearing under test and the lubricating medium under test are mismatched, then it is determined that the bearing under test and the lubricating medium under test are in a mismatched state.

[0042] in, , .

[0043] The first aspect of this invention provides a method for testing the matching of bearing raceway surface quality and lubrication medium. In one possible implementation, this method further includes the following optimization and adjustment process for the sample to be tested:

[0044] A) Develop optimization strategies for test samples in critical matching and mismatch states. Optimization strategies include adjusting the surface parameters of the bearing raceway and changing the lubrication medium.

[0045] B) Prepare the test sample again according to the optimization strategy;

[0046] C) Repeat steps 1) to 4). If the optimization verification results show that the sample to be tested has reached the optimized matching state or the good matching state in Zone II, then the optimization is successful and the process ends.

[0047] If the optimization verification results show that the sample to be tested is still in a critical matching state or a mismatch state, then the optimization strategy should be revised and the process should return to step B.

[0048] The first aspect of this invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium. In one possible implementation, the weighting coefficients are determined through the following steps:

[0049] a) Collect a historical test database containing a wide range of bearing raceway surface quality and lubrication medium combinations and their final performance;

[0050] b) For each sample in the historical test database, extract its vibration characteristics and stick-slip characteristics;

[0051] c) Using vibration characteristics and stick-slip characteristics as input features and final performance as the target variable, a regression model is trained using machine learning algorithms; the model can provide a score of the importance of each input feature for predicting the final performance.

[0052] d) Normalize the obtained importance scores and use them as the initial values ​​for the weight coefficients of the corresponding input features.

[0053] The first aspect of this invention provides a method for testing the matching of bearing raceway surface quality and lubrication medium. In one possible implementation, step 2) involves processing the acquired noise signal to obtain abnormal sound characteristics during high-speed operation testing of the sample under test; if based on... and If the bearing under test and the lubricating medium under test are determined to be in a good matching state, and are located within a set boundary region close to the critical matching state, while the abnormal noise characteristics exceed a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a critical matching state; if based on and If the bearing under test and the lubricating medium under test are determined to be in a critical matching state and are located in a set boundary region close to the good matching state within the critical matching state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a good matching state.

[0054] If based on and If the bearing under test and the lubricating medium under test are determined to be in a good matching state, and are located within a set boundary region of the good matching state close to the optimized matching state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in an optimized matching state; if based on and If the bearing under test and the lubricating medium under test are determined to be in a critical matching state and are located in a set boundary region close to the mismatch state within the critical matching state, and the abnormal noise characteristics are greater than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a mismatch state.

[0055] If based on and If the bearing under test and the lubricating medium under test are determined to be in an optimized matching state, and are located within a set boundary region close to a good matching state within the optimized matching state, and the abnormal noise characteristics exceed a set threshold, then the bearing under test and the lubricating medium under test are corrected to a good matching state; if based on and If it is determined that the bearing under test and the lubricating medium under test are in a mismatched state, and are located in a set boundary region close to the critical matching state within the mismatched state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a critical matching state.

[0056] A second aspect of the present invention provides a bearing raceway surface quality and lubrication medium matching test system, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the following method steps.

[0057] 1) Obtain the arithmetic mean height of the raceway surface of the bearing under test. Surface texture direction angle Arithmetic mean deviation of raceway circumferential waviness ;

[0058] 2) Obtain the test sample during high-speed operation testing: vibration characteristics obtained from fault characteristic frequencies and stick-slip characteristics obtained from acoustic emission signals; fill the bearing under test with the test lubricating medium to obtain the test sample;

[0059] 3) Calculate the normalized friction coefficient under operating conditions and comprehensive matching degree function :

[0060]

[0061] in, The average coefficient of friction is obtained by testing the friction performance of the lubricating medium under friction conditions close to high-speed operation test conditions, where U is the entrainment velocity in the rolling contact zone. To determine the apparent viscosity of the lubricating medium under high-speed testing, This represents the maximum Hertzian contact stress.

[0062] Equal to the lubrication potential parameter of the sample under test and the normalized vibration characteristic term The product of the speed severity factor With normalized stick-slip characteristic term The ratio of the products;

[0063] Lubrication potential parameters are 3D roughness adaptation function ripple influence function and texture direction adaptation function The product; exist It reaches its maximum value when it is within the set value, and decreases when it deviates from the set value; and Negative correlation; exist It is at its maximum when it is within the set degree, and decreases when it deviates from it; Negatively correlated with vibration characteristics Positively correlated with rotational speed; Positively correlated with sticky-slippery characteristics;

[0064] 4) smaller and The larger the value, the better the match between the bearing under test and the lubricating medium under test, and the matching test results of the sample under test can be obtained accordingly.

[0065] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching testing system. In one possible implementation, step 1) further obtains the average friction coefficient of the bearing steel balls under several friction conditions when rubbing the raceway under lubrication by the lubricating medium, obtained during friction performance testing of the lubricating medium under test; in step 3), the average friction coefficient obtained under friction conditions close to the current test conditions during friction performance testing is used as... .

[0066] The second aspect of the present invention provides a bearing raceway surface quality and lubrication medium matching test system, wherein in one possible implementation, the friction condition is a combination of contact stress and sliding speed.

[0067] The second aspect of this invention provides a bearing raceway surface quality and lubricant medium matching test system. In one possible implementation, step 1) further obtains the rheological relationship between apparent viscosity and shear rate obtained from rheological property testing of the lubricant medium under test; step 3) estimates the contact area shear rate at high-speed test rotation speed and substitutes it into the rheological relationship to obtain... .

[0068] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system. In one possible implementation, in step 2), a radial vibration velocity signal is acquired at the axial midpoint perpendicular to the outer cylindrical surface of the outer ring of the bearing sample under test, and the acceleration signal is obtained by differentiation; the acceleration signal is high-pass filtered to retain the high-frequency resonance band, and then Hilbert transform is performed to obtain the envelope signal; the envelope signal is FFTed to obtain the envelope spectrum; based on the bearing geometric parameters and real-time rotation speed, the theoretical fault characteristic frequency is calculated, and the characteristic frequency and its amplitude at the second and third harmonics are extracted from the envelope spectrum to construct a vibration feature vector containing the vibration characteristics.

[0069] The second aspect of the present invention provides a bearing raceway surface quality and lubrication medium matching test system. In one possible implementation, in step 2), an acoustic emission sensor is installed on the outer ring end face of the bearing sample to collect the original voltage signal. The original voltage signal is bandpass filtered to focus on the high-frequency stress wave components generated by stick-slip and micro-contact. The bandpass filtered voltage signal is subjected to continuous wavelet transform to obtain its time-frequency energy distribution. The total energy of the wavelet coefficients in several preset different characteristic frequency bands and the number of times the energy exceeds the set energy threshold in the CWT time-frequency diagram are used as the stick-slip feature to form a stick-slip feature vector.

[0070] The second aspect of the present invention provides a bearing raceway surface quality and lubrication medium matching test system, in one possible implementation, wherein in step 3):

[0071] The set value is 1.0;

[0072] );

[0073] ;

[0074] .

[0075] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system, in one possible implementation, in step 3), ),in For the i-th component of the vibration feature vector, It is its corresponding weighting coefficient.

[0076] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system, in one possible implementation, in step 3), ,in It is the stick-slip feature vector The j-th component, Its corresponding weighting coefficient.

[0077] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system, in one possible implementation, in step 4),

[0078] when Less than the calibration value and Greater than the calibration value If so, it is determined that the bearing under test and the lubricating medium under test are in an optimal matching state;

[0079] when Greater than the calibration value Less than the calibration value ,and Less than the calibration value Greater than the calibration value If the bearing under test and the lubricating medium under test are in a good matching state, then it is determined that the bearing under test and the lubricating medium under test are in a good matching state.

[0080] when Greater than the calibration value Less than the calibration value ,or Less than the calibration value Greater than the calibration value If so, it is determined that the bearing under test and the lubricating medium under test are in a critical matching state;

[0081] when Greater than the calibration value and Less than the calibration value If the bearing under test and the lubricating medium under test are mismatched, then it is determined that the bearing under test and the lubricating medium under test are in a mismatched state.

[0082] in, , .

[0083] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system, which, in one possible implementation, further includes the following sample optimization and adjustment process:

[0084] A) Develop optimization strategies for test samples in critical matching and mismatch states. Optimization strategies include adjusting the surface parameters of the bearing raceway and changing the lubrication medium.

[0085] B) Prepare the test sample again according to the optimization strategy;

[0086] C) Repeat steps 1) to 4). If the optimization verification results show that the sample to be tested has reached the optimized matching state or the good matching state in Zone II, then the optimization is successful and the process ends.

[0087] If the optimization verification results show that the sample to be tested is still in a critical matching state or a mismatch state, then the optimization strategy should be revised and the process should return to step B.

[0088] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system, in one possible implementation, determining the weighting coefficients through the following steps:

[0089] a) Collect a historical test database containing a wide range of bearing raceway surface quality and lubrication medium combinations and their final performance;

[0090] b) For each sample in the historical test database, extract its vibration characteristics and stick-slip characteristics;

[0091] c) Using vibration characteristics and stick-slip characteristics as input features and final performance as the target variable, a regression model is trained using machine learning algorithms; the model can provide a score of the importance of each input feature for predicting the final performance.

[0092] d) Normalize the obtained importance scores and use them as the initial values ​​for the weight coefficients of the corresponding input features.

[0093] The second aspect of this invention provides a bearing raceway surface quality and lubrication medium matching test system. In one possible implementation, step 2) involves processing the acquired noise signal to obtain abnormal sound characteristics during high-speed operation testing of the sample under test; if based on... and If the bearing under test and the lubricating medium under test are determined to be in a good matching state, and are located within a set boundary region close to the critical matching state, while the abnormal noise characteristics exceed a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a critical matching state; if based on and If the bearing under test and the lubricating medium under test are determined to be in a critical matching state and are located in a set boundary region close to the good matching state within the critical matching state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a good matching state.

[0094] If based on and If the bearing under test and the lubricating medium under test are determined to be in a good matching state, and are located within a set boundary region of the good matching state close to the optimized matching state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in an optimized matching state; if based on and If the bearing under test and the lubricating medium under test are determined to be in a critical matching state and are located in a set boundary region close to the mismatch state within the critical matching state, and the abnormal noise characteristics are greater than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a mismatch state.

[0095] If based on and If the bearing under test and the lubricating medium under test are determined to be in an optimized matching state, and are located within a set boundary region close to a good matching state within the optimized matching state, and the abnormal noise characteristics exceed a set threshold, then the bearing under test and the lubricating medium under test are corrected to a good matching state; if based on and If it is determined that the bearing under test and the lubricating medium under test are in a mismatched state, and are located in a set boundary region close to the critical matching state within the mismatched state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a critical matching state.

[0096] The beneficial effects of this invention are as follows: This invention pioneers a system-level quantitative matching degree test method and corresponding system, and for the first time organically integrates bearing raceway surface morphology, high-speed performance of lubricating medium, and real-time multi-physics field operating signals through a physical model based on Stribeck theory, creatively defining a quantifiable matching degree function. and normalized coefficient of friction This represents a qualitative leap from qualitative, experience-based judgment to quantitative, scientific evaluation.

[0097] Further, online accurate diagnosis of lubrication status was achieved: the constructed - The two-dimensional diagnostic map can intuitively and accurately divide the real-time operating status of the bearing into four distinct regions, and the diagnostic conclusions provide judgments and modification suggestions, with a diagnostic depth far exceeding that of traditional vibration analysis. Attached Figure Description

[0098] Figure 1 This is a schematic diagram of the bearing raceway surface quality and lubrication medium matching test method of the present invention;

[0099] Figure 2 This is a flowchart of step S100: digital precision measurement of bearing raceway surface morphology characteristic parameters and establishment of a digital feature library of bearing raceway surface morphology.

[0100] Figure 3 This is a flowchart of step S200, which involves high-speed shear rheology and tribological property spectrum testing of lubricating media.

[0101] Figure 4 This is a flowchart of step S300: High-speed operation test of bearing-lubricating medium system and synchronous acquisition of multi-physics field signals;

[0102] Figure 5(a) is a flowchart of step S401 in the multi-source signal processing and state-sensitive feature vector extraction in the high-speed operation test of step S400;

[0103] Figure 5(b) is a flowchart of step S402 in the multi-source signal processing and state-sensitive feature vector extraction in the high-speed operation test step S400;

[0104] Figure 5(c) is a flowchart of step S403 in the multi-source signal processing and state-sensitive feature vector extraction in the high-speed operation test of step S400;

[0105] Figure 6(a) is a flowchart of step S501 in the method of system matching degree modeling and lubrication state accurate diagnosis based on Stribeck theory in step S500.

[0106] Figure 6(b) is a flowchart of step S502 in the method of system matching degree modeling and lubrication state accurate diagnosis based on Stribeck theory in step S500.

[0107] Figure 6(c) is a flowchart of step S503 in the method of system matching degree modeling and lubrication state accurate diagnosis based on Stribeck theory in step S500.

[0108] Figure 7(a) is a flowchart of step S601 in the formulation and verification of matching optimization control strategy in step S600;

[0109] Figure 7(b) is a flowchart of step S602 in step S600, which is the formulation and verification of the matching optimization control strategy.

[0110] Figure 8 This is a flowchart of the bearing raceway surface quality and lubrication medium matching test method of the present invention;

[0111] Figure 9 This is a flowchart of the bearing raceway surface quality and lubrication medium matching test method with optimization and verification stages of the present invention;

[0112] Figure 10 This is a flowchart of the bearing raceway surface quality and lubrication medium matching test method with graded boundary correction based on abnormal sound characteristics, according to the present invention.

[0113] Figure 11(a) is a three-dimensional view of the test device for the surface quality of bearing raceway and the matching characteristics of lubricating grease in the soundproof chamber;

[0114] Figure 11(b) is a top view of the test device for the surface quality of bearing raceways and the matching characteristics of grease in a soundproof chamber;

[0115] Figure 12 This is a schematic diagram of the bearing raceway surface quality and lubrication medium matching test system of the present invention.

[0116] Figures 11(a) and 11(b) include: support base I, support base II, base III, pneumatic loading device IV, pneumatic loading device V, acoustic emission sensor 2, laser vibration sensor 3, test bearing 4, fixing ring 5, drive shaft 6, condenser microphone 7, fixed tip 8; drive motor 9, worktable 10, coupling 11, magnetorheological damper 12, coupling 13. Detailed Implementation

[0117] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0118] The purpose of this invention is to provide a systematic, quantitative, and engineering-practical method for testing the surface quality of bearing raceways and the matching performance of lubricating grease, thereby completely solving the problem that existing technologies cannot perform online, quantitative, and condition-level diagnostic evaluation of the "surface-lubricating medium" system. To achieve this objective, the overall approach of this invention is to first characterize the bearing raceway surface morphology using precise offline parameters, and then combine this with test signals from a multi-signal bearing vibration and noise test bench to ultimately propose a fusion diagnostic model for bearing surface morphology and vibration / noise. First, high-precision offline measurements are performed on the bearing raceway surface and candidate lubricating media (taking lubricating grease as an example) to establish a basic performance database. Then, on a designed high-speed bearing test bench (a test device for testing the matching characteristics of bearing raceway surface quality and lubricating grease), the bearing system is run under simulated operating conditions, and vibration, acoustic emission (stick-slip), and noise signals are collected simultaneously. Finally, based on Stribeck curve theory, a normalized matching function model that integrates offline basic parameters with online dynamic signal characteristics is constructed. By calculating the function value and its derived normalized friction coefficient, the real-time lubrication state of the bearing system is accurately diagnosed and quantitatively rated for matching performance.

[0119] To achieve the above objectives, the present invention provides the following technical solution:

[0120] The bearing raceway surface quality and grease matching characteristic testing device, as shown in Figures 11(a) and 11(b), includes support base I and II, base III, worktable 10, drive motor 9, magnetorheological damper 12, pneumatic loading devices IV and V, acoustic emission sensor 2, torque sensor, acceleration sensor (using laser vibration sensor 3), and microphone (using condenser microphone 7), etc. The entire testing device is located in a semi-anechoic chamber. The worktable 10 is equipped with drive motor 9 and fixed center 8. Drive motor 9 and fixed center 8 are fixed in the middle via coupling 11 and shaft connector 13, respectively, so that drive shaft 6 rotates with drive motor. Magnetorheological damper 12 is provided between drive motor 9 and drive shaft 6. The inner ring of test bearing 4 (sample to be tested) is assembled on drive shaft 6, and further fixed on both sides by fixing rings 5. Rubber shock-absorbing pads are placed between base III and worktable 10 to buffer vibrations on worktable 10. The drive motor 9 is fixed to the workbench 10 by bolts. The drive motor speed can reach up to 20,000 r / min, realizing high-speed operation testing of the test bearing 4.

[0121] At the start of the bearing vibration test, the test bearing 4 is interference-fitted onto the drive shaft 6. The outer ring of the test bearing is clamped and fixed by pneumatic loading devices IV and V, while the inner ring is connected to one side of the magnetorheological damper 12. The other side of the magnetorheological damper 12 is connected to the main shaft of the drive motor 9. After the drive motor 9 rotates, the sensors simultaneously measure the bearing vibration-related signals. The sensors include a laser vibration sensor 3, an acoustic emission sensor 2, and a high-sensitivity capacitive microphone 7.

[0122] The laser vibration sensor 3 is used to test the peak value of high-frequency vibration acceleration and low-frequency vibration velocity of bearing components. The detection range covers the full frequency band characteristics of bearing vibration. Through spectrum analysis and comparative analysis of bearing component vibration fault frequencies, the specific components causing abnormal bearing vibration signals can be identified.

[0123] The acoustic emission sensor 2 is bonded to the cantilever using adhesive material and installed in the acoustic emission sensor mounting bracket via mounting slots in the crossbeam and top plate. Movement in both horizontal and vertical directions brings the acoustic emission sensor into contact with the outer ring surface of the test bearing. When the lubricating medium does not match the bearing raceway surface morphology, the lubricating film on the contact surface is unevenly loaded. Under high-speed operation, the bearing contact surface lubrication state exhibits a solid-solid contact condition. In this case, high-frequency frictional stick-slip signals generated at the bearing's internal interface due to factors such as bearing surface machining defects, insufficient machining accuracy, or mismatch between the machined surface quality and the lubricating grease can be captured by the acoustic emission sensor 2. Through spectrum analysis, the specific location of the abnormal stick-slip signal in the bearing's operating state can be identified.

[0124] Simultaneously, a laser vibration sensor detects the vibration signal of the bearing's outer ring. A high-sensitivity condenser microphone 3 is directly mounted on the test bearing stage, 10mm away from the bearing, for synchronous testing of directional sound pressure abnormalities generated by the ball bearing at high speeds.

[0125] Method 1 for testing the compatibility of bearing raceway surface quality with lubricating medium:

[0126] The overall process of the bearing raceway surface quality and grease matching test method based on Stribeck curve lubrication theory is as follows: Figure 1 As shown, the steps include the following.

[0127] Step S100: Digital precision measurement of bearing raceway surface morphology characteristic parameters and establishment of a digital feature library of bearing raceway surface morphology.

[0128] Before assembling the bearing to be tested, non-contact three-dimensional surface morphology measurement technology is used to perform full-circumferential digital scanning and feature extraction on the working surfaces of the inner and outer raceways.

[0129] Specifically, such as Figure 2 As shown, step S100 includes:

[0130] S101: Using a white light interferometer, measure the three-dimensional morphology of the raceway surface of the bearing under test according to the ISO-25178 series standards. Select at least eight measurement areas evenly distributed along the raceway circumference, each area not less than 2mm × 2mm. Extract the key amplitude parameter from the three-dimensional morphology data: arithmetic mean height. (3D roughness, unit) and surface texture orientation angle Report the arithmetic mean height of each measurement area. average (as the arithmetic mean height of the raceway surface) ) and standard deviation Simultaneously, the arithmetic mean roughness of the one-dimensional profile parameters is extracted. For supplementary reference.

[0131] S102: Using a contact profilometer, perform a long-stroke (≥50mm) cross-sectional profile scan along the circumference of the raceway of the bearing under test. Select according to ISO-4287 standard. =2.5 , =0.8mm A filter bank with a diameter of 8mm was used to separate roughness, waviness, and shape profile. Key parameters were calculated from the waviness profile: the arithmetic mean deviation of waviness. (unit and average ripple wavelength (Unit: mm). Measured at at least six uniformly distributed circumferential sections, report the average of the arithmetic mean deviations of the waviness at each circumferential section. (As the arithmetic mean deviation of raceway circumferential waviness) ) and standard deviation .

[0132] S103: The above measurements ( , , The morphological data of the raceway of the bearing under test is stored in the surface morphology digital feature library, along with its three-dimensional morphological map, two-dimensional contour curve and power spectral density analysis map, and stored in the structured database.

[0133] Step S200: High-speed shear rheology and tribological property spectrum test of lubricating medium.

[0134] Under the harsh shear and pressure environment of the high-speed contact area of ​​the simulated bearing, a comprehensive rheological and tribological performance test was conducted on the lubricating medium to obtain the "high-speed performance lubrication data spectrum".

[0135] Specifically, such as Figure 3As shown, step S200 includes:

[0136] S201: Rheological property testing, using a strain-controlled rotary rheometer with a cone-plate rotor measurement system. For the lubricating medium under test, a steady-state shear rate scan is performed in a constant temperature environment (e.g., 25°C, 80°C), ranging from... Coverage Record the complete flow profile (reflecting apparent viscosity). With shear rate (The relationship between them). Focus on and record in The apparent viscosity values ​​of the lubricating medium under the conditions of high-speed operation (corresponding to, for example, automotive bearings) are denoted as follows: , , (Unit: mPa·s). The zero-shear viscosity of the lubricating medium under test was obtained by fitting the curve using the Bingham model. and infinite shear viscosity Key parameters, etc.

[0137] As another embodiment, the above shear rate data points can be adjusted according to needs or actual application scenarios, such as increasing or decreasing shear rate data points and changing the data values ​​of shear rate data points.

[0138] S202: Friction performance test, using a high-frequency reciprocating friction testing machine. The upper sample is... 10mm GCr15 bearing steel balls (preferably the same as the bearing rollers to be tested), with a disc of the same material as the lower sample. Two sets of simulated operating conditions are set up:

[0139] Operating Condition A (High Stress / Medium Speed): Normal load Fn = 100 N, frequency f = 50 Hz, stroke s = 1 mm, average sliding speed m / s, temperature T=80°C.

[0140] Operating Condition B (Medium Stress / High Speed): Normal load Fn = 50 N, frequency f = 200 Hz, stroke s = 0.5 mm, average sliding speed ≈0.2 m / s, temperature T=80°C.

[0141] As another embodiment, operating conditions A and B can be adjusted according to needs or actual application scenarios, such as adding or removing operating conditions and changing the condition parameters in the operating conditions. Adding more subdivided operating conditions can obtain the average friction coefficient measured under conditions closer to the test operating conditions in high-speed operation testing.

[0142] The operating time for each set of working conditions shall not be less than 30 minutes, and the average friction coefficient during the steady-state phase (the last 10 minutes) shall be recorded. and its standard deviation . The value characterizes the stability of the friction process and is directly related to the stick-slip tendency. The average friction coefficient obtained from operating condition A is... The standard deviation of the average friction coefficient is The average friction coefficient obtained from operating condition B is: The standard deviation of the average friction coefficient is .

[0143] S203: Will ( , , , , , , The "high-speed performance lubrication data spectrum" (i.e., lubrication data) of the lubrication medium to be tested is stored in the lubrication database.

[0144] Step S300: High-speed operation test of bearing-lubricating medium system and synchronous acquisition of multi-physics field signals.

[0145] The bearing to be tested, having completed the S100 measurement, is filled with the test lubricating medium that has completed the S200 test (hereinafter, the test sample refers to the bearing filled with the test lubricating medium), and installed on the designed high-speed bearing test bench (bearing raceway surface quality and grease matching characteristic testing device). It is then run under the set operating conditions, and three types of signals are collected simultaneously. For example... Figure 4 As shown, the specific steps are as follows.

[0146] S301: The high-speed bearing test bench must have a constant and stable drive capability of ≥10000 r / min, integrate a magnetorheological damper to isolate high-frequency vibration of the drive motor, and be equipped with a high-precision spindle, a programmable radial loading system, and a temperature control system. The high-speed bearing test bench with the bearing to be tested installed is placed in a semi-anechoic chamber as the entire test unit, with a background noise level ≤28dB.

[0147] S302: Configured with a multi-sensor integration and data synchronous acquisition system.

[0148] 1) Vibration signal channel: A non-contact laser Doppler vibration meter is used, with the laser beam vertically focused on the axial midpoint of the outer cylindrical surface of the bearing outer ring under test. Sampling frequency. The system is set to 102.4 kHz with a resolution of 16 bits. The radial vibration velocity signal v(t) is continuously acquired, and the acceleration signal a(t) is obtained through digital differentiation.

[0149] 2) Acoustic emission signal channel: A wideband resonant acoustic emission sensor is used, which is tightly mounted to the outer ring end face of the bearing sample under test using a special magnetic clamp and silicone grease coupling agent. The signal is then amplified by a 40dB preamplifier and sampled at the specified frequency. =2MHz, acquiring raw voltage signal .

[0150] 3) Noise signal channel: A 1 / 2-inch pre-polarized measuring microphone is used, with its diaphragm center distance from the outer ring surface of the bearing sample 100.0±1.0 mm, directly aligned with the bearing's central axis. After calibration using a sound calibrator, the signal is transmitted via... The sound pressure signal p(t) was acquired at a sampling rate of 44.1 kHz.

[0151] Synchronization and timing: All sensor signals are connected to the same high-performance data acquisition device, and the sampling clock is provided by the same high-stability temperature-controlled crystal oscillator to ensure strict synchronization between multiple channels with a synchronization error of less than 1μs.

[0152] S303: Test Condition Matrix and Process.

[0153] Install the bearing to be tested and apply the set radial load (e.g., 300 N).

[0154] Perform a stepped speed increase test: Sequentially set the spindle speed of the drive motor to... =8000 r / min, =12000 r / min, =16000 r / min, =20000 r / min. At each speed point, the drive motor is started until the spindle reaches the target speed and runs stably for 10 minutes. Then, the three-channel signals are synchronously acquired, and the acquisition time is T. acq = 65.536 seconds (for easy 2) N Point FFT).

[0155] Life test: At a typical high speed (e.g., 18000 r / min), it runs continuously for 120 hours, during which data is automatically collected for 30 seconds every hour to observe long-term matching stability.

[0156] Step S400: Multi-source signal processing and state-sensitive feature vector extraction during high-speed operation testing.

[0157] The acquired high-dimensional raw time-domain signals are preprocessed, transformed, and feature-mined to extract key indicators that characterize the bearing's operating state, especially the lubrication state and surface interaction characteristics. As shown in Figures 5(a), 5(b), and 5(c), the specific steps are as follows.

[0158] S401: As shown in Figure 5(a), vibration signal processing and feature extraction.

[0159] a) Preprocessing: Butterworth bandpass filtering (passband: 500Hz~10kHz) is applied to the acceleration signal a(t) to suppress low-frequency rigid body motion and high-frequency electronic noise.

[0160] b) Time-domain characteristics: Calculate the effective value of the filtered acceleration signal a(t). Peak and kurtosis .

[0161] c) Frequency domain characteristics:

[0162] Perform a Fast Fourier Transform (FFT) on the filtered acceleration signal a(t) to obtain the linear spectrum. .

[0163] Demodulation analysis of the original acceleration signal a(t): First, high-pass filtering (cutoff frequency > 500Hz) is used to retain the high-frequency resonance band, and then Hilbert transform is used to obtain the envelope signal e(t). The envelope spectrum E(f) is obtained by performing FFT on the envelope signal e(t).

[0164] Feature vector construction: Based on the bearing geometry and real-time rotational speed, the theoretical fault characteristic frequencies are calculated: inner ring pass frequency (BPFI), outer ring pass frequency (BPFO), rolling element rotation frequency (BSF), and cage rotation frequency (FTF). From the envelope spectrum E(f), the amplitudes of the above characteristic frequencies and their second and third harmonics are extracted to construct a 6-dimensional vibration feature vector. = [E(BPFI), E(2 BPFI), E(BPFO), E(2 BPFO), E(BSF), E(FTF)] T .

[0165] S402: As shown in Figure 5(b), acoustic emission signal processing and stick-slip feature extraction:

[0166] a) Preprocessing and filtering: processing the original voltage signal Perform bandpass filtering from 100kHz to 400kHz to obtain To focus on the high-frequency stress wave components generated by stickiness and micro-touch.

[0167] b) Time-frequency analysis: for Perform continuous wavelet transform (CWT), using Morlet wavelet as the mother wavelet, to obtain its time-frequency energy distribution CWT(t, f).

[0168] c) Feature vector construction :

[0169] Define two characteristic frequency bands: =[120, 180] kHz, =[250, 320] kHz.

[0170] Calculate the total energy of the wavelet coefficients within each frequency band:

[0171] , .

[0172] Set an energy threshold and count the number of "burst" events exceeding this threshold in the CWT time-frequency graph. .

[0173] Constructing a 3D stick-slip eigenvector: Logarithmic processing is used to compress the dynamic range of data.

[0174] S403: As shown in Figure 5(c), noise signal processing and abnormal sound feature extraction.

[0175] a) Preprocessing: The sound pressure signal p(t) is subjected to A-weighted digital filtering to obtain... Calculate A-weighted sound pressure level .

[0176] b) Separation of heterophonic components: Given the vibration a(t) and noise Originating from a common internal stimulus source, blind source separation technology is employed. A fast independent component analysis is performed to obtain two independent source signal estimates, s1(t) and s2(t). Typically, one of the components... It contains components that are strongly correlated with the characteristic vibration frequency of the bearing and is considered a "source of abnormal noise".

[0177] c) Feature vector construction : For the separated Perform an FFT to obtain the spectrum S(f). Find the peak value with the largest amplitude in the range of 0–5 kHz of S(f) and record its frequency. and amplitude Simultaneous calculation Root mean square (RMS) value The 3D anomaly feature vector is constructed as follows: .

[0178] Step S500: System matching degree modeling and accurate lubrication state diagnosis method based on Stribeck theory.

[0179] This is the core innovation of the present invention, which aims to establish a mathematical model with clear physical meaning and quantifiable characteristics to comprehensively evaluate the matching degree of the "surface morphology-lubricating medium" system under specific high-speed operating conditions. As shown in Figures 6(a), 6(b), and 6(c), the specific steps are as follows.

[0180] S501: As shown in Figure 6(a), calculate the normalized friction coefficient under the working condition. .

[0181] Introducing the core idea of ​​Stribeck's lubrication theory, a friction coefficient decoupled from specific rotational speed and load magnitude is defined. It is used to purely characterize the lubrication mechanism.

[0182]

[0183] In the formula:

[0184] The average friction coefficient is obtained from step S202 under the conditions closest to the current test conditions (contact stress, sliding speed). (such as working condition B) ).

[0185] U: Rolling contact zone speed (m / s). For tests where the inner ring rotates and the outer ring is fixed, let... , The bearing pitch circle diameter (mm) is given. The current test speed (r / min).

[0186] : Apparent viscosity at the current test speed; at the current test speed N i Below, the estimated shear rate in the contact area ( The minimum film thickness estimated based on the elastohydrodynamic lubrication theory (EHL theory) is used here. The value is interpolated from the rheological curve obtained in step S201 to obtain the corresponding apparent viscosity. (Pa·s).

[0187] Maximum Hertzian contact stress (Pa) is calculated based on bearing type, radial load, and material parameters.

[0188] It is a dimensionless number. The smaller the value, the more the system tends to be in a low-friction hydrodynamic lubrication or elastohydrodynamic lubrication state.

[0189] S502: As shown in Figure 6(b), construct the comprehensive matching degree function. .

[0190] This invention proposes a comprehensive evaluation index—the matching degree function. Its mathematical model is as follows:

[0191]

[0192] Molecular part: Characterized by the product of the system's "lubrication potential" and "operational smoothness".

[0193] As above, this is the effective viscosity of the grease under the current operating conditions, which is a direct reflection of its basic lubrication capability.

[0194] : Three-dimensional roughness fitting function. Obtained by fitting a large amount of bench test data: This function measures the three-dimensional average height (i.e., the arithmetic mean height of the raceway surface obtained in step S101). =0.10 The maximum value of 1.0 is reached at this time. The further away from this optimal value, the greater the deviation. The smaller the value, the lower the match between the surface roughness and the current grease.

[0195] : Ripple effect function. ). Arithmetic mean deviation of waviness The larger the value, the smaller the function value, reflecting that excessive ripple weakens the system's matching performance.

[0196] : Texture orientation adaptation function. This function is applied to the surface texture direction angle. = The maximum value of 1.25 is achieved at 0° or 90°, and it drops to 0.75 at 0° or 90°, indicating that the 45° texture direction has the best overall fit.

[0197] These parameters constitute lubrication potential.

[0198] : Normalized vibration characteristic term. ),in It is the vibration feature vector extracted in step S401. The i-th component, It is its corresponding weight coefficient (determined through feature importance analysis). The design makes the vibration smaller ( The smaller the value, the larger its contribution to The greater the positive impact, the better.

[0199] The denominator is the product of the "severity" of the operating condition and the "microscopic instability" of the interface.

[0200] : Speed ​​severity factor. This factor is greater than 1 and positively correlated with speed; it can be further designed to increase non-linearly with increasing speed (a speed increase beyond linearity), for example... This demonstrates the challenge that high speed poses to system stability.

[0201] : Normalized stick-slip characteristic term. ,in It is the stick-slip feature vector extracted in step S402. The j-th component, Assign weights to them. It is directly positively correlated with the intensity of the interface's sticky and slippery activity.

[0202] function The physical meaning of its value: its level reflects the overall performance under given high-speed and harsh operating conditions. Below, the current "surface morphology-lubricating medium" combination ( , , , To what extent can it suppress interfacial micro-instability activities? And maintain the smooth operation of macroscopic dynamics. . The higher the value, the better the system matching.

[0203] S503: As shown in Figure 6(c), accurate diagnosis of lubrication condition and matching degree classification.

[0204] Calculated The x-axis is... Using the vertical axis as the ordinate, a two-dimensional state diagnostic map is constructed. Through cluster analysis of a large amount of experimental data, the following diagnostic boundaries are established:

[0205] Zone I: Optimized matching zone (hydrodynamic lubrication state);

[0206] Criterion: ≤0.004 )and ≥1.5 );

[0207] Status description: The system is in an excellent hydrodynamic lubrication state with extremely low friction, no significant stick-slip, and extremely low vibration and noise levels, which is an ideal design and operation target.

[0208] Zone II: Good matching zone (mixed lubrication state);

[0209] Criterion: 0.004 < ≤0.008 And 1.0 ( )≤ <1.5;

[0210] Status description: The system is in a stable mixed lubrication state, with low levels of friction and stick-slip activity, acceptable vibration and noise, good matching, and potential for further optimization.

[0211] Zone III: Critical matching zone (unstable lubrication condition);

[0212] Criterion: 0.008 < ≤0.012 ) or 0.8 ( )≤ <1.0;

[0213] Status description: The system is in a critical state of transition from mixed lubrication to boundary lubrication. Stick-slip activity begins to intensify, intermittent abnormal noise may occur, the matching is at a critical level, and there is a risk of abnormal noise.

[0214] Zone IV: Mismatch zone (boundary lubrication condition);

[0215] Criterion: 0.012 < and <0.8;

[0216] Status description: The system is in a poor boundary lubrication state, with high friction, severe stick-slip, high vibration and noise levels, and a high risk of rapid wear or early failure. The "surface morphology-lubricating medium" combination is severely mismatched.

[0217] Step S600: Formulating and verifying matching optimization control strategies.

[0218] Based on the diagnostic results of S503, output quantization optimization is performed. As shown in Figures 7(a) and 7(b), the specific steps are as follows.

[0219] S601: As shown in Figure 7(a), optimize the strategy library.

[0220] If the diagnosis is zone IV (mismatch):

[0221] like Too high Extremely low; replacement of grease is strongly recommended.

[0222] Target selection strategy: The value is more than 50% higher than that of existing lubricating media, and (The average friction coefficient obtained under operating condition B) is less than 0.08 for high-performance lubricating media; if the three-dimensional roughness fitting function or ripple effect function The function value is significantly low (<0.7), and surface processing optimization is recommended. For example, if the arithmetic mean height... <0.06 It is recommended to use micro-shot peening process to improve the particle size to 0.09–0.11. If the arithmetic mean deviation of the waviness >0.18 More precise ultra-fine grinding is required.

[0223] If the diagnosis result is Zone III (borderline):

[0224] It is recommended to perform fine-tuning of the parameters. For example, based on the existing grease, try to adjust its arithmetic average height. To 0.10 Adjust, or change the texture direction Adjust to 45°. Alternatively, try using a different brand from the same series but with a different average friction coefficient standard deviation. Smaller (more stable friction) lubricant grades.

[0225] If the diagnosis result is Zone II (good), the decision to perform lean optimization can be made based on specific application requirements (such as pursuing ultimate silence).

[0226] S602: As shown in Figure 7(b), the optimization effect verification experiment.

[0227] 1. Prepare new bearing samples (adjust surfaces) or select new lubricating media in accordance with the strategy established in S601.

[0228] 2. For the optimized new combination, repeat the test process from S100 to S500.

[0229] 3. Comparative Analysis: Focus on the changes before and after optimization under the same critical operating condition (e.g., N=18000 r / min). This mainly includes: and Values ​​and their positional migration on the diagnostic spectrum; effective value of vibration acceleration The percentage decrease; weighted sound pressure level The decrease value (dB); acoustic emission stick-slip characteristic energy The percentage decrease.

[0230] Success criteria: The optimized state point should migrate from at least region IV or III to region II, ideally entering region I. Simultaneously, the aforementioned key performance indicators should show significant and consistent improvement (e.g., Decrease >25%, Decrease >2.5dB Value increase >30%.

[0231] As another implementation, a signal feature adaptive weight determination algorithm for the above method is also provided, specifically including:

[0232] Step Z01: Collect a historical test database containing a wide range of "surface-lubricating media" combinations and their final performance (lifetime, noise subjective evaluation).

[0233] Step Z02: For each sample in the database, extract its 6-dimensional vibration feature vector. 3D stick-slip eigenvectors 3D heterophonic feature vector .

[0234] Step Z03: Using vibration characteristics and stick-slip characteristics as input features, and the final performance as the target variable, train a regression model using a machine learning algorithm (such as random forest or gradient boosting decision tree). This model can provide a score for the importance of each input feature in predicting the final performance.

[0235] Step Z04: Normalize the obtained importance scores and use them as the vibration feature weights in step S502. and stick-slip feature weights The initial value. This makes the matching function... The construction can adaptively focus on the signal features that have the greatest impact on the final performance of the system.

[0236] This invention pioneers a system-level quantitative matching degree testing method: for the first time, it organically integrates bearing raceway surface morphology, high-speed performance of lubricating medium, and real-time multiphysics operating signals through a physical model based on Stribeck theory, and creatively defines a quantifiable matching degree function. and normalized coefficient of friction This represents a qualitative leap from qualitative, empirical judgment to quantitative, scientific evaluation. It enables precise online diagnosis of lubrication conditions: the constructed... - The two-dimensional diagnostic map can intuitively and accurately divide the real-time operating status of the bearing into four distinct regions, and the diagnostic conclusions provide judgments and modification suggestions, with a diagnostic depth far exceeding that of traditional vibration analysis.

[0237] The method of this invention forms a complete closed loop of "testing-diagnosis-optimization": This method not only evaluates the compatibility of existing combinations but also provides specific and quantitative optimization strategies (adjusting surface parameters or changing lubricating media) based on diagnostic results. Through verification experiments, it forms a closed loop, significantly shortening the collaborative R&D cycle for bearing products and lubricating greases. It boasts high technical integration and comprehensive information dimensions: deeply integrating knowledge from multiple disciplines such as precision surface metrology, rheology, acoustics, vibration engineering, signal processing, and tribology, it collects vibration, acoustic emission, and noise signals, comprehensively perceiving the system's macroscopic dynamic response, microscopic interface behavior, and final acoustic output, ensuring comprehensive information acquisition. This method can be directly applied to the process R&D departments of bearing manufacturers, the formulation evaluation laboratories of lubricating media suppliers, and supplier access testing for high-end OEMs. Bearing systems optimized using this method can operate stably in a low-vibration and low-noise state, significantly improving the quality of OEM products and user experience, and extending bearing life, demonstrating significant engineering and economic value.

[0238] In the above implementation, the abnormal sound feature vector, as a type of noise perceptible to the user, is used only as a reference and is not used for normalizing the friction coefficient under operating conditions. and comprehensive matching degree function Therefore, the calculation can be performed without extracting the variant sound feature vector. As another implementation, after extracting the variant sound feature vector, the variant sound vector can also serve the following functions:

[0239] Because it is difficult to accurately extract the characteristic components caused by poor "surface-grease" matching from vibration signals, abnormal noise features are further introduced to solve this problem. The abnormal noise components extracted through blind source separation can be compared with the characteristic vibration frequencies, thereby accurately pinpointing the noise source directly related to the lubrication state. In this embodiment, although the abnormal noise features are not explicitly included in the ξ formula, they indirectly determine the system's matching rating by influencing the final grading judgment in the following ways: The ξ value provides the theoretical matching degree, while the abnormal noise features provide a "verification" of the actual acoustic performance; the combination of the two makes the diagnosis more comprehensive. Abnormal noise features act as a bridge connecting the lubrication state (microscopic interface behavior) and user-perceptible noise (macroscopic acoustic output). Their verification and correction role in diagnosis and grading ensures the robustness and engineering practicality of the method of this invention.

[0240] In this embodiment, the heterophonic feature vector = This is key information extracted from noise signals through blind source separation; it directly reflects the abnormal components in the audible noise radiated by the bearing during high-speed operation. Abnormal sound characteristics and vibration characteristics. Slippery characteristics Together, they constitute a multi-dimensional sensing system, playing the following roles in lubrication condition diagnosis and matching degree grading:

[0241] 1. The relationship between abnormal noise characteristics and lubrication status.

[0242] Frequency correlation: Dominant frequency of heterophonic components The frequency closely matches a characteristic fault frequency of the bearing (such as the outer ring pass frequency BPFO and the inner ring pass frequency BPFI) or its harmonics. When the lubrication condition deteriorates (entering mixed lubrication or boundary lubrication), the rolling elements impact the raceway micro-protrusions, and the excited vibration energy propagates through the structure and radiates as noise, causing... A significant peak appears at this point, therefore, It directly indicates the bearing component corresponding to the source of the abnormal sound.

[0243] Amplitude and Energy: (Peak amplitude) and The root mean square value of the separated source signal quantifies the intensity of the abnormal noise. The more severe the poor lubrication (the more intense the stick-slip activity and the greater the vibration and impact), the higher the abnormal noise energy and the more prominent the audible noise. Experiments show that when the system degenerates from the hydrodynamic lubrication zone (Zone I) to the mixed lubrication zone (Zone II), the abnormal noise energy can increase by 3 to 5 times; when it enters the boundary lubrication zone (Zone IV), the abnormal noise energy can increase by more than 10 times.

[0244] 2. The relationship between abnormal sound characteristics and diagnosis and grading.

[0245] The abnormal sound features were not directly substituted into the formula for the matching degree function ξ (to avoid overly complex functions), but they play a key role as a "diagnostic verification indicator" and a "grading correction factor" in the following aspects:

[0246] Diagnostic verification: When vibration characteristics When an increase in the amplitude of a certain fault frequency (such as BPFO) is detected, if the abnormal sound characteristics... In If the frequency matches this value, it confirms that the fault frequency does indeed radiate audible noise, thus ruling out sensor interference or misjudgment due to vibration of non-bearing structures. This dual verification of "vibration and noise" greatly improves the reliability of the diagnosis.

[0247] Hierarchical boundary correction, such as Figure 10 As shown: Based on In the two-dimensional diagnostic atlas, the zoning boundaries are determined based on statistical analysis of a large amount of experimental data. However, in practical applications, samples near certain boundaries may require rating adjustments due to prominent abnormalities. For example:

[0248] If a sample It falls in Zone II (good matching zone), but has unusual sound characteristics. If the value exceeds a certain empirical threshold (e.g., 0.15mV), it indicates that although macroscopic vibration and stick-slip are still acceptable, significant audible noise has been generated, possibly indicating early uneven lubrication or localized contact problems. In this case, the state should be adjusted to Zone III (critical matching). Conversely, if a sample's... It falls on the edge of Zone III, but has very low heterophonic characteristics (e.g., (<0.05mV), indicating that its actual performance is better than the calculated value, it can be classified into Zone II. In the optimization and verification phase, in addition to comparison... and In addition, the reduction in abnormal noise energy is an important indicator for directly measuring the noise reduction effect. For example, in the optimization case, the noise reduction of 8.4 dB directly confirms the improvement in matching.

[0249] Implementation Method 2 for Testing the Matching of Bearing Raceway Surface Quality and Lubricating Medium:

[0250] like Figure 8 As shown, the present invention provides a method for testing the compatibility of bearing raceway surface quality with lubricating medium, comprising the following steps:

[0251] 1) Obtain the arithmetic mean height of the raceway surface of the bearing under test. Surface texture direction angle Arithmetic mean deviation of raceway circumferential waviness ;

[0252] To obtain the average coefficient of friction of the bearing steel balls rubbing against the raceway under several different combinations of contact stress and sliding speed under the friction performance test of the lubricating medium under the test. ;

[0253] Obtain the rheological relationship between apparent viscosity and shear rate obtained from rheological property testing of the lubricating medium under test. ;

[0254] 2) Obtain the vibration characteristics of the sample under test during high-speed operation testing, based on the fault characteristic frequency. (Vibration), stick-slip characteristics obtained based on acoustic emission signals (Slippery) and abnormal sound characteristics obtained from noise signal processing (Noise); The bearing to be tested is filled with the lubricating medium to be tested to obtain the test sample;

[0255] 3) Calculate the normalized friction coefficient under operating conditions and comprehensive matching degree function :

[0256]

[0257] in, The average coefficient of friction is given under friction conditions (contact stress and sliding speed) that are close to those of high-speed operation testing, where U is the entrainment speed in the rolling contact zone. To estimate the corresponding contact area shear rate at high-speed test rotation speed. Afterwards, according to The apparent viscosity of the lubricating medium to be tested was obtained. This represents the maximum Hertzian contact stress.

[0258] Equal to the lubrication potential parameter of the sample under test and the normalized vibration characteristic term The product of the speed severity factor With normalized stick-slip characteristic term The ratio of the products;

[0259] Lubrication potential parameters are 3D roughness adaptation function ripple influence function and texture direction adaptation function The product; exist It reaches its maximum value when it is within the set value, and decreases when it deviates from the set value; and Negative correlation; exist It is at its maximum when it is within the set degree, and decreases when it deviates from it; Negatively correlated with vibration characteristics Positively correlated with rotational speed; Positively correlated with sticky-slippery characteristics;

[0260] 4) smaller and The larger the value, the better the match between the bearing under test and the lubricating medium under test, and the matching test results of the sample under test can be obtained accordingly.

[0261] In one possible implementation, in step 2), a radial vibration velocity signal is acquired at the midpoint of the axial direction perpendicular to the outer cylindrical surface of the outer ring of the bearing sample under test, and the acceleration signal is obtained by differentiation; the acceleration signal is high-pass filtered to retain the high-frequency resonance band, and then Hilbert transform is performed to obtain the envelope signal; the envelope signal is FFTed to obtain the envelope spectrum; based on the bearing geometric parameters and real-time rotational speed, the theoretical fault characteristic frequency is calculated, and the characteristic frequency and its amplitude at the second and third harmonics are extracted from the envelope spectrum to construct a vibration feature vector containing the vibration characteristics.

[0262] In one possible implementation, in step 2), an acoustic emission sensor is installed on the outer ring end face of the bearing of the sample to be tested to collect the original voltage signal. The original voltage signal is bandpass filtered to focus on the high-frequency stress wave components generated by stick-slip and micro-contact. The bandpass filtered voltage signal is subjected to continuous wavelet transform to obtain its time-frequency energy distribution. The total energy of the wavelet coefficients in several preset different characteristic frequency bands and the number of times the energy exceeds the set energy threshold in the CWT time-frequency diagram are used as the stick-slip feature to form the stick-slip feature vector.

[0263] In one possible implementation, step 3) involves:

[0264] The set value is 1.0;

[0265] );

[0266] ;

[0267] .

[0268] In one possible implementation, in step 3), ),in For the i-th component of the vibration feature vector, It is its corresponding weighting coefficient.

[0269] In one possible implementation, in step 3), ,in It is the stick-slip feature vector The j-th component, Its corresponding weighting coefficient.

[0270] In one possible implementation, in step 4),

[0271] when Less than or equal to the calibration value and Greater than or equal to the calibration value If so, it is determined that the bearing under test and the lubrication medium under test are in optimal matching (hydrodynamic lubrication state).

[0272] when Greater than the calibration value Less than or equal to the calibration value ,and Less than the calibration value Greater than or equal to the calibration value If the bearing under test and the lubricating medium under test are well matched (mixed lubrication state), then it is determined that the bearing under test and the lubricating medium under test are in good matching (mixed lubrication state).

[0273] when Greater than the calibration value Less than or equal to the calibration value ,or Less than the calibration value Greater than or equal to the calibration value If the bearing under test and the lubrication medium under test are in a critical match (unstable lubrication state), then it is determined that the bearing under test and the lubrication medium under test are in a critical match (unstable lubrication state).

[0274] when Greater than the calibration value and Less than the calibration value If the bearing under test and the lubricating medium under test are mismatched (boundary lubrication state), then it is determined that the bearing under test and the lubricating medium under test are in a mismatch (boundary lubrication state).

[0275] in, , .

[0276] In one possible implementation, the weighting coefficients are determined through the following steps:

[0277] a) Collect a historical test database containing a wide range of bearing raceway surface quality and lubrication medium combinations and their final performance;

[0278] b) For each sample in the historical test database, extract its vibration characteristics and stick-slip characteristics;

[0279] c) Using vibration characteristics and stick-slip characteristics as input features and final performance as the target variable, a regression model is trained using machine learning algorithms; the model can provide a score of the importance of each input feature for predicting the final performance.

[0280] d) Normalize the obtained importance scores and use them as the initial values ​​for the weight coefficients of the corresponding input features.

[0281] In one possible implementation, such as Figure 9 As shown, the following steps were used to optimize and adjust the test samples with unstable lubrication and boundary lubrication states:

[0282] A) Develop optimization strategies for test samples with unstable lubrication conditions and boundary lubrication conditions. Optimization strategies include adjusting the surface parameters of the raceway of the bearing to be tested and replacing the grease (lubricating medium).

[0283] B) Prepare a new sample for testing according to the optimized strategy; if the surface parameters are adjusted, prepare a new bearing sample, refill the lubricating medium, and prepare the sample for testing. If the grease is replaced, clean it thoroughly and replace it again.

[0284] C) Repeat steps 1) to 4) to verify the optimization effect. If the optimization verification results show that the sample under test has reached the dynamic pressure lubrication state in Zone I or the mixed lubrication state in Zone II, the optimization is successful and the final report is output.

[0285] If the optimization verification results show that the sample to be tested is still in an unstable lubrication state in Zone III or a boundary lubrication state in Zone II, then the optimization strategy is revised and the process returns to step B) to prepare the sample to be tested again according to the optimization strategy and continue testing and verification.

[0286] This invention belongs to the interdisciplinary technical field of high-end equipment manufacturing and tribological testing, specifically relating to a testing method for the matching characteristics of raceway surface quality and lubricating grease in high-speed ball bearings. In particular, it is a testing method based on multi-physics signal fusion and Stribeck curve normalization modeling. By quantifying the matching relationship between surface roughness (Ra), waviness (Wa), texture orientation angle (θ), and lubricating grease rheological properties, a dynamic lubrication state control model is established to guide the optimization of bearing manufacturing processes and grease selection, solving the vibration and noise fault testing problem of drive motor bearings in new energy vehicles operating at speeds exceeding 10,000 r / min. This method encompasses multiple technical aspects, including surface morphology measurement, lubricating grease performance testing, high-speed signal acquisition, feature extraction, modeling analysis, and optimization control.

[0287] Implementation method of bearing raceway surface quality and lubrication medium matching test system:

[0288] The present invention discloses a bearing raceway surface quality and lubrication medium matching test system, the schematic diagram of which is shown below. Figure 12 As shown, it specifically includes a memory, a processor, a system bus, and a computer program stored in the memory. The processor and the memory communicate and exchange data with each other through the system bus. The processor executes the computer program to implement the steps of the bearing raceway surface quality and lubrication medium matching test method of the present invention.

[0289] The steps of the bearing raceway surface quality and lubrication medium matching test method of the present invention have been described sufficiently clearly in the embodiments of the bearing raceway surface quality and lubrication medium matching test method, and will not be repeated here.

Claims

1. A method for testing the compatibility of bearing raceway surface quality with lubricating medium, characterized in that, Includes the following steps: 1) Obtain the arithmetic mean height of the raceway surface of the bearing under test. Surface texture direction angle Arithmetic mean deviation of raceway circumferential waviness ; 2) Obtain the vibration characteristics of the sample under test during high-speed operation testing based on the fault characteristic frequency and the stick-slip characteristics based on the acoustic emission signal; The bearing to be tested is filled with the lubricating medium to be tested to obtain the test sample; 3) Calculate the normalized friction coefficient under operating conditions and comprehensive matching degree function : in, The average coefficient of friction is obtained by testing the friction performance of the lubricating medium under friction conditions close to high-speed operation test conditions, where U is the entrainment velocity in the rolling contact zone. To determine the apparent viscosity of the lubricating medium under high-speed testing, This represents the maximum Hertzian contact stress. Equal to the lubrication potential parameter of the sample under test and the normalized vibration characteristic term The product of the speed severity factor With normalized stick-slip characteristic term The ratio of the products; Lubrication potential parameters are 3D roughness adaptation function ripple influence function and texture direction adaptation function The product; exist It reaches its maximum value when it is within the set value, and decreases when it deviates from the set value; and Negative correlation; exist It is at its maximum when it is within the set degree, and decreases when it deviates from it; Negatively correlated with vibration characteristics Positively correlated with rotational speed; Positively correlated with sticky-slippery characteristics; 4) smaller and The larger the value, the better the match between the bearing under test and the lubricating medium under test, and the matching test results of the sample under test can be obtained accordingly.

2. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 1, characterized in that, In step 1), the average friction coefficient of the bearing steel balls under several friction conditions obtained during the friction performance test of the lubricating medium under the test is also obtained; in step 3), the average friction coefficient obtained under friction conditions close to the current test conditions during the friction performance test is used as... .

3. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 2, characterized in that, The friction conditions are a combination of contact stress and sliding speed.

4. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 1, characterized in that, In step 1), the rheological relationship between apparent viscosity and shear rate obtained from the rheological property test of the lubricating medium under test is also obtained; in step 3), the shear rate of the contact area at the high-speed test rotation speed is estimated and substituted into the rheological relationship to obtain... .

5. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 1, characterized in that, In step 2), a radial vibration velocity signal is acquired at the midpoint of the axial direction perpendicular to the outer cylindrical surface of the outer ring of the bearing sample under test, and the acceleration signal is obtained by differentiation. The acceleration signal is then high-pass filtered to retain the high-frequency resonance band, and then Hilbert transform is performed to obtain the envelope signal. The envelope signal is then subjected to FFT to obtain the envelope spectrum. Based on the bearing geometric parameters and real-time rotational speed, the theoretical fault characteristic frequency is calculated. From the envelope spectrum, the characteristic frequency and the amplitude at its second and third harmonics are extracted to construct a vibration feature vector containing the vibration characteristics.

6. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 1, characterized in that, In step 2), an acoustic emission sensor is installed on the outer ring end face of the bearing of the sample to be tested to collect the original voltage signal. The original voltage signal is bandpass filtered to focus on the high-frequency stress wave component generated by stick-slip and micro-contact. The voltage signal after bandpass filtering is subjected to continuous wavelet transform to obtain its time-frequency energy distribution. The total energy of the wavelet coefficients in several preset different characteristic frequency bands and the number of times the energy exceeds the set energy threshold in the CWT time-frequency diagram are used as the stick-slip feature to form the stick-slip feature vector.

7. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 1, characterized in that, In step 3): The set value is 1.0; )； ； 。 8. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 5, characterized in that, In step 3), ),in For the i-th component of the vibration feature vector, It is its corresponding weighting coefficient.

9. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 6, characterized in that, In step 3), ,in It is the stick-slip feature vector The j-th component, Its corresponding weighting coefficient.

10. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 1, characterized in that, In step 4): when Less than the calibration value and Greater than the calibration value If so, it is determined that the bearing under test and the lubricating medium under test are in an optimal matching state; when Greater than the calibration value Less than the calibration value ,and Less than the calibration value Greater than the calibration value If the bearing under test and the lubricating medium under test are in a good matching state, then it is determined that the bearing under test and the lubricating medium under test are in a good matching state. when Greater than the calibration value Less than the calibration value ,or Less than the calibration value Greater than the calibration value If so, it is determined that the bearing under test and the lubricating medium under test are in a critical matching state; when Greater than the calibration value and Less than the calibration value If the bearing under test and the lubricating medium under test are mismatched, then it is determined that the bearing under test and the lubricating medium under test are in a mismatched state. in, , .

11. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 10, characterized in that, It also includes the following sample optimization and adjustment process: A) Develop optimization strategies for test samples in critical matching and mismatch states. Optimization strategies include adjusting the surface parameters of the bearing raceway and changing the lubrication medium. B) Prepare the test sample again according to the optimization strategy; C) Repeat steps 1) to 4). If the optimization verification results show that the sample to be tested has reached the optimized matching state or the good matching state in Zone II, then the optimization is successful and the process ends. If the optimization verification results show that the sample to be tested is still in a critical matching state or a mismatch state, then the optimization strategy should be revised and the process should return to step B.

12. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 8 or 9, characterized in that, The weighting coefficients are determined by the following steps: a) Collect a historical test database containing a wide range of bearing raceway surface quality and lubrication medium combinations and their final performance; b) For each sample in the historical test database, extract its vibration characteristics and stick-slip characteristics; c) Using vibration characteristics and stick-slip characteristics as input features and final performance as the target variable, a regression model is trained using machine learning algorithms; the model can provide a score of the importance of each input feature for predicting the final performance. d) Normalize the obtained importance scores and use them as the initial values ​​for the weight coefficients of the corresponding input features.

13. The method for testing the matching of bearing raceway surface quality and lubrication medium according to claim 10, characterized in that, In step 2), during the high-speed operation test of the sample to be tested, the abnormal sound characteristics are obtained by processing the collected noise signal. If based on and If the bearing under test and the lubricating medium under test are determined to be in a good matching state, and are located within a set boundary region close to the critical matching state, while the abnormal noise characteristics exceed a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a critical matching state; if based on and If the bearing under test and the lubricating medium under test are determined to be in a critical matching state and are located in a set boundary region close to the good matching state within the critical matching state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a good matching state. If based on and If the bearing under test and the lubricating medium under test are determined to be in a good matching state, and are located within a set boundary region of the good matching state close to the optimized matching state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in an optimized matching state; if based on and If the bearing under test and the lubricating medium under test are determined to be in a critical matching state and are located in a set boundary region close to the mismatch state within the critical matching state, and the abnormal noise characteristics are greater than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a mismatch state. If based on and If the bearing under test and the lubricating medium under test are determined to be in an optimized matching state, and are located within a set boundary region close to a good matching state within the optimized matching state, and the abnormal noise characteristics exceed a set threshold, then the bearing under test and the lubricating medium under test are corrected to a good matching state; if based on and If it is determined that the bearing under test and the lubricating medium under test are in a mismatched state, and are located in a set boundary region close to the critical matching state within the mismatched state, and the abnormal noise characteristics are less than a set threshold, then the bearing under test and the lubricating medium under test are corrected to be in a critical matching state.

14. A bearing raceway surface quality and lubrication medium matching test system, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the bearing raceway surface quality and lubrication medium matching test method according to any one of claims 1 to 13.

Images