A method for measuring the thickness of a lubricating film in real time and a system therefor

Abstract

This invention proposes a method and system for real-time measurement of lubricating film thickness. The method includes: acquiring an ultrasonic incident signal and a global reflected echo; sequentially performing digital filtering, mean elimination, and normalization on the ultrasonic incident signal and the global reflected echo; sequentially performing time shifting, amplitude adjustment, and weighted summation on the normalized ultrasonic incident signal to obtain the first theoretical lubricating film reflected echo under different time delays, calculating the Pearson correlation coefficient, obtaining the correlation coefficient matrix to determine the time delay parameter corresponding to the maximum correlation coefficient value, and achieving primary initial positioning; performing data interpolation on both the normalized ultrasonic incident signal and the global reflected echo, selecting a negative Pearson correlation coefficient as the objective function, determining the lubricating film hysteresis index corresponding to the optimal solution, and calculating the lubricating film thickness. The method proposed in this invention can solve the problem of balancing sensor limitations and accuracy efficiency.

Description

Technical Field

[0001] This invention relates to the field of measurement technology, and in particular to a method and system for real-time measurement of lubricating film thickness. Background Technology

[0002] Lubricating films play a crucial role in reducing friction, wear, and load transfer in mechanical systems, and their performance directly affects the operating efficiency and reliability of equipment. Therefore, accurate real-time measurement of lubricating film thickness is of significant engineering practical importance and academic research value for assessing the operating status of critical mechanical components, enabling early fault warning, and optimizing lubrication design.

[0003] Methods for measuring lubricating film thickness are typically based on electrical, optical, and acoustic principles. Electrical methods calculate film thickness by establishing a calibration curve between electrical parameters and the lubricating film thickness. Since 1946, resistance methods, voltage discharge methods, and capacitance methods have been proposed successively. These methods offer advantages such as simplicity and high spatial resolution, but also suffer from drawbacks such as insufficient stability, risk of electrical breakdown, and ambiguity in measurement point positioning. They are particularly unsuitable for water-lubricated bearings made of insulating composite materials in naval vessels. Eddy current methods are a novel electromagnetic measurement technique. However, their installation requires damaging the surfaces of the friction pair, thus affecting the dynamic performance of mechanical components, which limits their engineering applications.

[0004] Optical methods are among the most accurate techniques for measuring lubricating film thickness, including white light interferometry, light intensity methods, multi-beam interferometry, and fiber optic methods. These methods are based on optical theory and image processing techniques, calculating film thickness through spectral analysis, interference fringes, light intensity, or ordering. However, optical methods have stringent requirements regarding the light transmittance of the friction pair and the installation conditions of the measuring device. Therefore, they are mainly used in laboratory environments.

[0005] As a promising alternative technology, ultrasound provides a new approach for in-situ monitoring of lubricating film thickness due to its non-invasive and highly penetrating characteristics. Existing ultrasonic thickness measurement models include spring models, phase models, resonance models, and time-of-flight methods. Currently, the ultrasonic method has been widely applied in mechanical components such as rolling bearings, radial bearings, and thrust bearings. However, it still has the following unresolved problems: (1) Lack of a unified model. The frequency domain method requires switching the corresponding model for different film thickness ranges, which increases the complexity of the algorithm and potential errors; (2) Sensor limitations. To achieve a wide range of film thickness measurements, the frequency domain method requires the sensor to have sufficient bandwidth to connect different models. However, increasing the sensor bandwidth often leads to a decrease in its penetration and anti-interference ability, which is unacceptable for high-attenuation materials. In addition, the cost of customized sensors is high, which is not conducive to engineering applications; (3) The solution accuracy and computational efficiency are mutually exclusive. When the calculation time is a few microseconds, the resolution is only 7μm, while when the resolution is less than 0.1μm, the calculation time is as high as hundreds of microseconds.

[0006] To reduce monitoring costs and balance measurement accuracy and real-time performance, it is urgent to explore a novel method for film thickness identification with a unified measurement process. This requires solving two core challenges: first, establishing a unified model, namely, constructing an ultrasonic propagation model to characterize the lubricating film thickness; and second, proposing a film thickness calculation method that simultaneously ensures accuracy and efficiency. Summary of the Invention

[0007] The main objective of this invention is to provide a method and system for real-time measurement of lubricating film thickness, which aims to accurately and in real-time measure the thickness of the lubricating film.

[0008] To achieve the above objectives, the present invention provides a method for real-time measurement of lubricating film thickness, comprising the following steps: Step S1: Collect the reference signal of the first test block-air interface as the ultrasonic incident signal. The first test block is a bearing test block. Step S2: Drop lubricating water between the first test block and the second test block, and collect the signal of the ultrasonic wave after multiple reflections and superpositions at the upper and lower interfaces of the lubricating film as the overall reflected echo. The second test block is a shaft test block or a shaft sleeve test block. Step S3: Perform digital filtering, mean elimination, and normalization on the ultrasonic incident signal and the overall reflected echo in sequence; Step S4: After performing time shift, amplitude adjustment and weighted summation on the normalized ultrasonic incident signal, the first theoretical lubricating film reflected echo under different time delays is obtained. The Pearson correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo is calculated for different time delays. The correlation coefficient matrix is ​​obtained to determine the time delay parameter corresponding to the maximum correlation coefficient value, thereby achieving primary initial positioning. Step S5: Perform data interpolation on both the normalized ultrasonic incident signal and the overall reflected echo. Select a negative Pearson correlation coefficient as the objective function. Optimize the objective function based on the particle swarm optimization algorithm to determine the lubricating film hysteresis index corresponding to the optimal solution. Calculate the lubricating film thickness based on the lubricating film hysteresis index corresponding to the optimal solution to complete the precise positioning of the lubricating film.

[0009] Preferably, a signal recovery algorithm is used to interpolate the normalized ultrasonic incident signal and the overall reflected echo to improve the solution resolution; after selecting a negative Pearson correlation coefficient as the objective function, a local solution space is constructed with the time delay parameter corresponding to the determined maximum correlation coefficient value as the center.

[0010] Preferably, before the step of acquiring the reference signal at the first test block-air interface as the ultrasonic incident signal, the method further includes: The density and sound velocity of the three media—the first test block, the second test block, and the lubricating film—were obtained.

[0011] Preferably, when constructing the local solution space, it is randomly generated within the upper and lower bounds of the parameters to be solved. N There are 4 particles, each with a position representing a candidate parameter. All particles form the initial solution space for subsequent optimization search.

[0012] Preferably, when a negative Pearson correlation coefficient is selected as the objective function, the interpolated ultrasonic incident signal is sequentially time-shifted, amplitude-adjusted, and weighted summed to obtain the second theoretical lubricating film reflection echo under different time delays, and the Pearson correlation coefficient between the second theoretical lubricating film reflection echo corresponding to different time delays and the overall reflection echo after interpolation is calculated.

[0013] Preferably, the first theoretical lubricating film reflected echo is obtained by multiplying the coefficient matrix by the signal shift matrix, based on the normalized ultrasonic incident signal, using the theoretical lubricating film reflected echo as the theoretical lubricating film reflected echo.

[0014] Preferably, when obtaining the correlation coefficient matrix, a main loop is constructed with the signal delay index as the iteration variable. In each loop, the incident signal and the lubricating film reflected echo are equal to the coefficient matrix multiplied by the signal shift matrix to generate the first theoretical lubricating film reflected echo. The correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo obtained in step S3 is calculated. When the loop ends, the time delay parameter corresponding to the maximum correlation coefficient value is calculated.

[0015] Preferably, the specific steps for calculating the Pearson correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo corresponding to different time delays, obtaining the correlation coefficient matrix to determine the time delay parameter corresponding to the maximum correlation coefficient value, and realizing the first-level initial positioning include: Initialize the iteration counter and set the current iteration number. k The value is set to 1, and the maximum number of iterations is preset to control the algorithm termination condition; The first theoretical lubricating film reflection echo is constructed by weighted summation of the ultrasonic incident signal; Calculate the first k The correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo obtained in step S3 under each cycle is updated. when k If the number of iterations is less than or equal to the preset maximum number, return to the step of constructing the first theoretical lubricating film reflection echo using the weighted summed ultrasonic incident signal; when k If the number of iterations exceeds the preset maximum number, determine the time delay parameter corresponding to the maximum correlation coefficient value in the correlation coefficient matrix, and end the iteration loop.

[0016] Preferably, the lubricating film thickness is calculated based on the lubricating film hysteresis index corresponding to the optimal solution. h hour ， Lubricating film thickness h Calculate using the following formula: h = k fine c 2 T s / 2 L ; in, c 2 represents the propagation speed of ultrasound in the lubricating film. L The interpolation factor. T s Sampling time, k fine This is the lubrication film hysteresis index corresponding to the optimal solution calculated based on the particle swarm optimizer.

[0017] The present invention also proposes a system for real-time measurement of lubricating film thickness, comprising: An ultrasonic transducer is used to emit ultrasonic waves and receive reflected echoes. The signal acquisition unit is used to acquire the ultrasonic incident signal and the overall reflected echo; The processor is used to execute the above-described method for real-time measurement of lubricating film thickness in order to calculate the lubricating film thickness.

[0018] The method for real-time measurement of lubricating film thickness proposed in this invention has the following beneficial effects: (1) Compared with the traditional ultrasonic frequency domain thickness measurement method, the present invention can achieve a wide range of film thickness measurement from 1 to 500 μm without model switching. The working principle is simple and the measurement range is wide. (2) Real-time measurement is possible. The thickness of the lubricating film is a time variable. However, existing ultrasonic film thickness measurement methods are difficult to monitor in real time due to excessive calculation time. This invention is based on the hierarchical solution idea, which has high measurement accuracy and short measurement time, and effectively solves the contradiction between calculation accuracy and efficiency. (3) No sensor limitation. Traditional methods require customized ultrasonic transducers with a large frequency domain working bandwidth, while the present invention directly calculates the lubricating film thickness based on the time domain waveform, without sensor waveform bandwidth limitation. (4) This invention is particularly suitable for engineering applications. It can be easily integrated into hardware without feature recognition and is compatible with low-cost commercial sensors (<500 RMB), thereby significantly reducing costs. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the method for real-time measurement of lubricating film thickness according to the present invention. Figure 2 This is a schematic diagram of the three-layer ultrasonic propagation model in the method for real-time measurement of lubricating film thickness of the present invention; Figure 3 This is a flowchart illustrating a specific embodiment of the method for real-time measurement of lubricating film thickness according to the present invention; Figure 4 This is a time-domain solution curve of the method for real-time measurement of lubricating film thickness according to the present invention; Figure 5 This is a schematic diagram showing the calculation results of the method for real-time measurement of lubricating film thickness according to the present invention; Figure 6 This is a schematic diagram illustrating the calculation time of the method for real-time measurement of lubricating film thickness according to the present invention.

[0020] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This invention proposes a method for real-time measurement of lubricating film thickness.

[0023] This method for real-time measurement of lubricating film thickness uses a three-layer ultrasonic propagation model (such as...). Figure 2Based on the model shown, the bearing and shaft are simplified to materials with parallel upper and lower interfaces and smooth surfaces. In the model, media 1, 2 and 3 represent the bearing, lubricating film and metal shaft / sleeve, respectively.

[0024] An incident wave is emitted into the system via an ultrasonic sensor. I When the signal propagates to interface 1, part of the signal is reflected and part is transmitted to the lubricating film, resulting in multiple reflections and transmissions between interfaces 1 and 2. Due to the limitations of the bearing structure, the ultrasonic testing mode is typically a reflection mode. Therefore, the signal received by the transducer is... P 1, P 2, ..., P n Meanwhile, the micron-level thickness of the lubricating film results in short time intervals between the initial echo and multiple echoes from the liquid film. Therefore, echoes that cannot be completely separated in the time domain are superimposed into a single reflected echo. P .

[0025] Assuming the incident wave is I(t) Then each reflected echo can be represented as: (1) in, n The number of ultrasonic echoes; t Represents a time variable; τ This indicates the time it takes for the signal to travel one round trip within the water film. Furthermore, R ij and T ij ( i , j =1, 2, 3) represent the ultrasonic waves originating from the medium. i Incident on medium j The sound pressure reflection coefficient and sound pressure transmission coefficient are expressed as: (2) in, Z i, Z j They are media i and j The acoustic impedance is equal to the density of the medium. ρ Speed ​​of sound c The product of.

[0026] P This can be obtained by superimposing all reflected echoes in the time domain. In practice, ultrasonic signals are continuous signals generated by the periodic charging and discharging of a set of capacitors. To obtain a processable signal, it is necessary to discretize the signal using a sampling device. A series of discrete values ​​are extracted from the continuous signal using a periodically approximating ideal sampling unit impulse sequence. Therefore, PThe discretized expression is: ; (3) in, i For discrete signals; T s The sampling time is equal to the reciprocal of the sampling frequency; k This represents the lubrication film hysteresis index. In practice, signals are usually stored and processed in index form, represented as... P [ i ].

[0027] In the method for real-time measurement of lubricating film thickness of this invention, the ultrasonic grading algorithm includes a first-level coarse positioning and a second-level fine positioning. In the coarse positioning stage, multiple reflections from the lubricating layer can be simulated by time-shifting and amplitude adjustment of the incident signal. Since the lubricating film is typically thin, ultrasonic propagation attenuation can be approximately ignored. Therefore, by obtaining the acoustic impedance of the three-layer model, the sound pressure loss for each ultrasonic reflection and transmission can be obtained. First, the incident wave is acquired. I [ i [This is followed by a description of a loop that covers all lubricant film thicknesses, and then adjustments to the hysteresis index.] k Determine the number of echoes and construct a simulated overall reflection echo. S coar [ i ]. S [ i The symbol represents the general method for calculating reflected echoes, applicable to both coarse and fine positioning. S coar and S fine These represent the basis for the coarse and fine positioning stages, respectively. S [ i The calculated reflected echo. In each loop, the calculation... S coar [ i [and reality] P [ i Cross-correlation coefficient: (4) in, N This represents the number of data points in the time domain.

[0028] Get C coar [ k The index corresponding to the maximum value is stored as... k coar This completes the first-level coarse positioning of the film thickness.

[0029] In the fine localization stage, the proposed algorithm transforms the iterative exhaustive solution method into a metaheuristic optimization problem. It utilizes a signal recovery algorithm to... I [ i ]and P [ i Interpolation is used to improve solution resolution. Subsequently, a negative Pearson correlation coefficient is chosen as the optimization objective function. (5) in, N * This indicates the number of data points in the time domain after interpolation. i* Indicates the index of the interpolated signal. S fine [ i* ]and P [ i* ] represent the interpolated simulated echo and the actual echo, respectively.

[0030] An improved particle swarm optimization (PSO) was selected to optimize the non-gradient function with boundary constraints in equation (5), with the optimization variables being: x The search space is constrained within a space defined by... k coar Centered on the sample, extending upwards and downwards by 1.5 original sampling points, this can be represented as: (6) in, lb and ub These represent the lower boundary and the upper boundary, respectively. L This is the interpolation factor.

[0031] Finally, the optimized optimal solution is denoted as k fine The formula for calculating the lubricating film thickness is as follows: (7) To avoid the interpretation overhead of processing element by element, and to effectively utilize the parallel computing capabilities of computers with single instruction multiple data streams, the simulated echo... S coar [ i ]and S fine [ i* The construction process of [ ] is vectorized. Memory is pre-allocated, and the coefficient matrix is ​​constructed. V : (8) in, A and B They are respectively equal to T 12 R 23T 21 and R 21 R 23 , m The number of simulated reflected echoes, It represents the Hadamardi (or Hadama) stack.

[0032] Subsequently, a shift matrix is ​​established. D : (9) in d This indicates a lagging index, which is used for coarse positioning. d =k, for precise positioning d =round( k ).

[0033] Based on this, combined D The signal shift matrix H can be obtained by performing a cyclic shift. The analog signal can be calculated according to the following formula: (10) Reference Figure 1 In this preferred embodiment, a method for real-time measurement of lubricating film thickness includes the following steps: Step S1: Collect the reference signal of the first test block-air interface as the ultrasonic incident signal. The first test block is a bearing test block. Step S2: Drop lubricating water between the first test block and the second test block, and collect the signal of the ultrasonic wave after multiple reflections and superpositions at the upper and lower interfaces of the lubricating film as the overall reflected echo. The second test block is a shaft test block or a shaft sleeve test block. Step S3: Perform digital filtering, mean elimination, and normalization on the ultrasonic incident signal and the overall reflected echo in sequence; Step S4: After performing time shift, amplitude adjustment and weighted summation on the normalized ultrasonic incident signal, the first theoretical lubricating film reflected echo under different time delays is obtained. The Pearson correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo is calculated for different time delays. The correlation coefficient matrix is ​​obtained to determine the time delay parameter corresponding to the maximum correlation coefficient value, thereby achieving primary initial positioning. Step S5: Perform data interpolation on both the normalized ultrasonic incident signal and the overall reflected echo. Select a negative Pearson correlation coefficient as the objective function. Optimize the objective function based on the particle swarm optimization algorithm to determine the lubricating film hysteresis index corresponding to the optimal solution. Calculate the lubricating film thickness based on the lubricating film hysteresis index corresponding to the optimal solution to complete the precise positioning of the lubricating film.

[0034] In step S4, the first theoretical lubricating film reflected echo is obtained by multiplying the coefficient matrix by the signal shift matrix, based on the normalized ultrasonic incident signal, since the theoretical lubricating film reflected echo is equal to the coefficient matrix multiplied by the signal shift matrix.

[0035] When obtaining the correlation coefficient matrix, a main loop is constructed with the signal delay index as the iteration variable. In each loop, the incident signal and the lubricating film reflected echo are used to generate the first theoretical lubricating film reflected echo by multiplying the coefficient matrix by the signal shift matrix. The first theoretical lubricating film reflected echo (i.e., the analog signal, such as...) is then calculated. Figure 3 The correlation coefficient between the total reflected echo (as shown) and the normalized overall reflected echo (i.e., the actual signal) obtained in step S3 is calculated. The loop ends and the time delay parameter corresponding to the maximum correlation coefficient value is calculated.

[0036] Specifically, in step S5, a signal recovery algorithm is used to interpolate the normalized ultrasonic incident signal and the overall reflected echo to improve the solution resolution.

[0037] After selecting a negative Pearson correlation coefficient as the objective function, a local solution space is constructed centered on the time delay parameter corresponding to the determined maximum correlation coefficient value. When constructing the local solution space, it is randomly generated within the upper and lower bounds of the parameter to be solved. N There are 4 particles, each with a position representing a candidate parameter. All particles form the initial solution space for subsequent optimization search.

[0038] Furthermore, the procedure prior to step S1 includes: The density and sound velocity of the three media—the first test block, the second test block, and the lubricating film—were obtained.

[0039] In step S5, when a negative Pearson correlation coefficient is selected as the objective function, the interpolated ultrasonic incident signal is sequentially time-shifted, amplitude-adjusted, and weighted summed to obtain the second theoretical lubricating film reflection echo under different time delays. The Pearson correlation coefficient between the second theoretical lubricating film reflection echo and the interpolated overall reflection echo is then calculated.

[0040] Specifically, the Pearson correlation coefficients of the first theoretical lubricating film reflected echo and the normalized overall reflected echo corresponding to different time delays are calculated. The correlation coefficient matrix is ​​then obtained to determine the time delay parameter corresponding to the maximum correlation coefficient value. The specific steps for achieving primary initial positioning include: Initialize the iteration counter and set the current iteration number. k The value is set to 1, and the maximum number of iterations is preset to control the algorithm termination condition; The first theoretical lubricating film reflection echo is constructed by weighted summation of the ultrasonic incident signal; Calculate the first kThe correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo obtained in step S3 under each cycle is updated. when k If the number of iterations is less than or equal to the preset maximum number, return to the step of constructing the first theoretical lubricating film reflection echo using the weighted summed ultrasonic incident signal; when k If the number of iterations exceeds the preset maximum number, determine the time delay parameter corresponding to the maximum correlation coefficient value in the correlation coefficient matrix, and end the iteration loop.

[0041] The lubricating film thickness is calculated based on the lubricating film hysteresis index corresponding to the optimal solution. h hour ， Lubricating film thickness h Calculate using the following formula: h = k fine c 2 T s / 2 L ; in, c 2 represents the propagation speed of ultrasound in the lubricating film. L The interpolation factor. T s Sampling time, k fine This is the lubrication film hysteresis index corresponding to the optimal solution calculated based on the particle swarm optimizer.

[0042] by Figure 3 The following example illustrates the specific process of this calculation method.

[0043] 1. Start the ultrasound grading algorithm and initialize system parameters and variables.

[0044] 2. Acquire ultrasonic reference signal and water film reflection signal Specifically, the ultrasonic reference signal is the ultrasonic reflection signal at the air interface of medium 1; the water film reflection signal is the overall reflected echo signal collected when a liquid water film exists in mediums 1 and 3. Both serve as the basic input data for subsequent signal processing.

[0045] 3. Perform FIR filtering, mean removal, and normalization respectively. Specifically, a finite impulse response (FIR) bandpass filter is used to suppress out-of-band noise and retain effective frequency band components; a mean-removal operation is used to eliminate DC offset in the signal; normalization scales the signal amplitude to the [-1, 1] range to eliminate matching deviation caused by differences in system gain, thereby improving the stability and accuracy of subsequent correlation calculations.

[0046] 4. Data interpolation ( L (times) processing Specifically, an upsampling method is used to increase the signal time axis resolution by a factor of L. Data interpolation improves the accuracy of waveform details, which helps to more accurately estimate the echo delay in subsequent calculations, thereby improving the thickness inversion resolution.

[0047] 5. Obtain the interpolated signal, that is, obtain... P [ i* ]、 I [ i* ] Specifically, P [ i* ] is for interpolation L The water film reflection signal after multiplication, I [ i* ] is for interpolation L The interpolated ultrasound reference signal is then used. After interpolation, the data is stored in the variable area, awaiting subsequent calls from the algorithm.

[0048] 6. Start the iterative loop and set... k =1 Specifically, initialize the iteration counter and set the current iteration number. k The value is set to 1, and the maximum number of iterations is preset to control the algorithm's termination condition.

[0049] 7. Through S [ i ]= V * H To calculate the reflected echo of the first theoretical lubricating film Specifically, within each cycle, a first theoretical water film reflection signal is constructed using an ultrasonic reference signal and formula (10). V The coefficient matrix, H This is the signal shift matrix.

[0050] Step S8: Calculate the cross-correlation coefficient. C coar [ k ]=corr( S coar [ i ], P [ i ]) Specifically, calculate the first k Theoretical water film reflection signal under one cycle S coar [ i [Compared with measured water film reflection signal] P [ i The cross-correlation coefficients between [ ] are calculated and stored in [ ] Ccoar Among the variables, since the water film thickness directly affects the echo delay and waveform shape, the closer the candidate parameter is to the true value, the more similar the waveforms are, and the larger the cross-correlation coefficient is.

[0051] 9. Determine the current iteration number k ≤Maximum number of iterations Specifically, determine the current iteration number. k Is it less than or equal to the preset maximum number of iterations? If less than, then let... k = k +1, and return to step 7 to continue the next round of calculation; otherwise, terminate the iteration.

[0052] 10. k coar =Max( C coar ) Specifically, calculate array C coar The index corresponding to the maximum value is denoted as . k coar .

[0053] 11. Construct the solution space Specifically, input k coar The solution space is constructed using formula (6). Within this space, solutions are randomly generated within the upper and lower bounds of the parameters to be solved. N There are 4 particles, each with a position representing a candidate parameter. All particles form the initial solution space for subsequent optimization search.

[0054] 12. Through S [i]= V * H To construct the second theoretical water film reflection echo Specifically, input P [ i* ]and I [ i* The second theoretical water film reflection echo is constructed using formula (10).

[0055] 13. Construct the target function corr([ S fine [ i* ], P [ i* ]) Specifically, the cross-correlation coefficient is used as the objective function to measure the waveform similarity between the theoretical and measured signals. This objective function is then used in subsequent optimization algorithms.

[0056] 14: Use the particle swarm optimizer to optimize and solve the objective function.

[0057] 15. Through the formula h = k fine c 2T s / 2 L To calculate the thickness of the lubricating film Specifically, obtain the water film hysteresis index corresponding to the optimal solution calculated by the particle swarm optimizer. k fine Combined with the propagation speed of ultrasound in water film c 2. Interpolation factor L and sampling time T s Calculate the water film thickness h This formula originates from the ultrasonic time-of-flight method, where the denominator 2 L The time scale scaling caused by interpolation was taken into account to ensure consistency in thickness units. The final output... h This is the result of the water film thickness calculation.

[0058] Figure 5 This is a specific implementation result of the present invention. The proposed method exhibits good agreement with the reference film thickness across all film thicknesses. As the film thickness increases from 1 μm to 10 μm, the error gradually decreases from -28.7% to 1.68%. For film thicknesses between 7 and 60 μm, the relative error of the proposed ultrasonic grading algorithm is less than 4%. With further increases in film thickness, the maximum relative error of the ultrasonic grading algorithm is only 4.37%. These results demonstrate that the proposed ultrasonic grading algorithm has a large measurement range and high accuracy.

[0059] Figure 6 This is a specific calculation time result for this invention. The average calculation time for a standard computer processing film thicknesses of 1–500 μm is 39.7 ms. For thinner films, the calculation time varies considerably, ranging from a maximum of 49.97 ms to a minimum of only 30.82 ms. As the thickness increases, the standard deviation decreases. For film thicknesses greater than 300 μm, the calculation time stabilizes at 39 ms, with a standard deviation within 5 ms. In contrast, across the entire film thickness range, the calculation time for a high-performance computer fluctuates between 7.44 ms and 12.06 ms, with an average of 9.45 ms. These results demonstrate that this method has a shorter calculation time and strong potential for real-time monitoring.

[0060] The method for real-time measurement of lubricating film thickness proposed in this invention has the following beneficial effects: (1) Compared with traditional ultrasonic frequency domain thickness measurement methods, the present invention can achieve film thickness measurement in a wide range of 1~500μm without model switching. The working principle is simple and the measurement range is wide. (2) Real-time measurement is possible. The thickness of the lubricating film is a time variable. However, existing ultrasonic film thickness measurement methods are difficult to monitor in real time due to excessive calculation time. This invention is based on the hierarchical solution idea, which has high measurement accuracy and short measurement time, and effectively solves the contradiction between calculation accuracy and efficiency. (3) No sensor limitation. Traditional methods require customized ultrasonic transducers with a large frequency domain working bandwidth, while the present invention directly calculates the lubricating film thickness based on the time domain waveform, without sensor waveform bandwidth limitation. (4) This invention is particularly suitable for engineering applications. It can be easily integrated into hardware without feature recognition and is compatible with low-cost commercial sensors (<500 RMB), thereby significantly reducing costs.

[0061] The present invention also proposes a system for real-time measurement of lubricating film thickness.

[0062] In this preferred embodiment, a system for real-time measurement of lubricating film thickness includes: An ultrasonic transducer is used to emit ultrasonic waves and receive reflected echoes. The signal acquisition unit is used to acquire the ultrasonic incident signal and the overall reflected echo; The processor is used to execute the above-described method for real-time measurement of lubricating film thickness in order to calculate the lubricating film thickness.

[0063] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for real-time measurement of lubricating film thickness, characterized in that, Includes the following steps: The reference signal of the first test block-air interface was collected as the ultrasonic incident signal. The first test block was a bearing test block. Lubricating water was dripped between the first and second test blocks, and the signal of the ultrasonic wave after multiple reflections and superpositions at the upper and lower interfaces of the lubricating film was collected as the overall reflected echo. The second test block was a shaft test block or a shaft sleeve test block. The ultrasonic incident signal and the overall reflected echo were sequentially digitally filtered, mean-eliminating, and normalized. After performing time shifting, amplitude adjustment and weighted summation on the normalized ultrasonic incident signal, the first theoretical lubricating film reflected echo under different time delays is obtained. The Pearson correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo is calculated. The correlation coefficient matrix is ​​obtained to determine the time delay parameter corresponding to the maximum correlation coefficient value, so as to achieve primary initial positioning. Data interpolation is performed on both the normalized ultrasonic incident signal and the overall reflected echo. A negative Pearson correlation coefficient is selected as the objective function. The objective function is optimized based on the particle swarm optimization algorithm to determine the lubricating film hysteresis index corresponding to the optimal solution. The lubricating film thickness is calculated based on the lubricating film hysteresis index corresponding to the optimal solution to complete the precise positioning of the lubricating film.

2. The method for real-time measurement of lubricating film thickness as described in claim 1, characterized in that, The signal recovery algorithm is used to interpolate the normalized ultrasonic incident signal and the overall reflected echo to improve the solution resolution. After selecting the negative Pearson correlation coefficient as the objective function, a local solution space is constructed with the time delay parameter corresponding to the determined maximum correlation coefficient value as the center.

3. The method for real-time measurement of lubricating film thickness as described in claim 1, characterized in that, Before the step of acquiring the reference signal from the first test block-air interface as the ultrasonic incident signal, the following steps are also included: The density and sound velocity of the three media—the first test block, the second test block, and the lubricating film—were obtained.

4. The method for real-time measurement of lubricating film thickness as described in claim 3, characterized in that, When constructing the local solution space, it is randomly generated within the upper and lower bounds of the parameters to be solved. N There are 4 particles, each with a position representing a candidate parameter. All particles form the initial solution space for subsequent optimization search.

5. The method for real-time measurement of lubricating film thickness as described in claim 1, characterized in that, When a negative Pearson correlation coefficient is selected as the objective function, the interpolated ultrasonic incident signal is sequentially time-shifted, amplitude-adjusted, and weighted summed to obtain the second theoretical lubricating film reflection echo under different time delays. The Pearson correlation coefficient between the second theoretical lubricating film reflection echo and the interpolated overall reflection echo is then calculated.

6. The method for real-time measurement of lubricating film thickness as described in claim 1, characterized in that, The theoretical lubricating film reflected echo is equal to the coefficient matrix multiplied by the signal shift matrix. The first theoretical lubricating film reflected echo is obtained based on the normalized ultrasonic incident signal.

7. The method for real-time measurement of lubricating film thickness as described in claim 1, characterized in that, When obtaining the correlation coefficient matrix, a main loop is constructed with the signal delay index as the iteration variable. In each loop, the incident signal and the lubricating film reflected echo are equal to the coefficient matrix multiplied by the signal shift matrix to generate the first theoretical lubricating film reflected echo. The correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo obtained in step S3 is calculated. When the loop ends, the time delay parameter corresponding to the maximum correlation coefficient value is calculated.

8. The method for real-time measurement of lubricating film thickness as described in any one of claims 1 to 7, characterized in that, The specific steps for calculating the Pearson correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo corresponding to different time delays, obtaining the correlation coefficient matrix, and determining the time delay parameter corresponding to the maximum correlation coefficient value to achieve primary initial positioning include: Initialize the iteration counter and set the current iteration number. k The value is set to 1, and the maximum number of iterations is preset to control the algorithm termination condition; The first theoretical lubricating film reflection echo is constructed by weighted summation of the ultrasonic incident signal; Calculate the first k The correlation coefficient between the first theoretical lubricating film reflected echo and the normalized overall reflected echo obtained in step S3 under each cycle is updated. when k If the number of iterations is less than or equal to the preset maximum number, return to the step of constructing the first theoretical lubricating film reflection echo using the weighted summed ultrasonic incident signal; when k If the number of iterations exceeds the preset maximum number, determine the time delay parameter corresponding to the maximum correlation coefficient value in the correlation coefficient matrix, and end the iteration loop.

9. The method for real-time measurement of lubricating film thickness as described in claim 1, characterized in that, The lubricating film thickness is calculated based on the lubricating film hysteresis index corresponding to the optimal solution. h hour ， Lubricating film thickness h Calculate using the following formula: h = k fine c 2 T s / 2 L ； in, c 2 represents the propagation speed of ultrasound in the lubricating film. L The interpolation factor. T s Sampling time, k fine This is the lubrication film hysteresis index corresponding to the optimal solution calculated based on the particle swarm optimizer.

10. A system for real-time measurement of lubricating film thickness, characterized in that, include: An ultrasonic transducer is used to emit ultrasonic waves and receive reflected echoes. The signal acquisition unit is used to acquire the ultrasonic incident signal and the overall reflected echo; A processor for performing the method for real-time measurement of lubricating film thickness as described in any one of claims 1 to 9, to calculate the lubricating film thickness.

Images