Asphalt mixture fatigue state sound emission early warning method based on multi-parameter fusion

Abstract

The invention discloses a multi-parameter fusion asphalt mixture fatigue state acoustic emission early warning method, and belongs to the technical field of asphalt mixture performance evaluation.The method comprises the steps that fatigue loading is conducted on an asphalt mixture through an acoustic emission test, acoustic emission test data in the testing process are obtained, and the acoustic emission test data are obtained; carrying out information entropy calculation on different parameters of the acoustic emission test data, and carrying out weighted sum calculation on key parameters determined by the information entropy to obtain comprehensive evaluation parameters; obtaining a dynamic threshold value through a kernel density estimation method according to the comprehensive evaluation parameters; and performing multi-parameter joint distribution calculation on the key parameters through a multi-parameter joint distribution model according to the dynamic threshold to obtain evaluation parameters, and evaluating the evaluation parameters according to the dynamic threshold to obtain early warning results of different levels. According to the invention, multi-dimensional and multi-threshold monitoring can be carried out, the fatigue state of the asphalt mixture can be identified more comprehensively and accurately, and a more reliable basis is provided for road maintenance decisions.

Description

Technical Field

[0001] The invention belongs to the technical field of performance evaluation of asphalt mixtures, and in particular relates to a multi-parameter fused asphalt mixture fatigue state acoustic emission early warning method. Background Art

[0002] With the continuous economic growth and rapid development of the transportation industry, asphalt mixture is one of the most commonly used materials in road construction. Its fatigue state directly affects the service life and safety of roads. Traditional asphalt mixture fatigue state monitoring methods mainly rely on macroscopic mechanical performance indicators, such as stiffness modulus, permanent deformation, etc. These methods often lag behind the actual occurrence of damage and it is difficult to detect early micro cracks in time. As a non-destructive testing method, acoustic emission technology can monitor the microscopic changes inside the material in real time, providing a new way for early damage identification.

[0003] However, the existing asphalt mixture fatigue state identification method based on acoustic emission technology still has the following problems:

[0004] 1. Most methods only use a single acoustic emission parameter for analysis, which is difficult to fully reflect the fatigue state of the material;

[0005] 2. Lack of effective multi-parameter fusion methods makes it difficult to fully utilize the multi-dimensional information of acoustic emission signals;

[0006] 3. The early warning mechanism is single and difficult to adapt to the complex road service environment and fatigue damage of varying degrees.

[0007] At present, with the development of intelligent transportation systems, the performance requirements for road construction materials are getting higher and higher. Real-time online monitoring of the fatigue state of asphalt mixture can not only optimize road maintenance plans and budgets, but also help improve road safety and extend service life. Therefore, the development of an early warning method for fatigue state of asphalt mixture based on fuzzy comprehensive evaluation of multivariate acoustic emission parameters has important theoretical value and application prospects. Summary of the invention

[0008] In order to solve the above technical problems, the present invention proposes a multi-parameter fusion asphalt mixture fatigue state acoustic emission early warning method, in which ring count, energy, duration and amplitude are selected as key parameters in the acoustic emission mechanical parameters for normalization, and the comprehensive evaluation parameter weights are obtained through fuzzy comprehensive evaluation calculation. The key parameters are fused through the weights to perform multi-dimensional and multi-threshold monitoring, which can more comprehensively and accurately identify the fatigue state of asphalt mixture, provide a more reliable basis for road maintenance decision-making, and solve the problems existing in the above-mentioned prior art.

[0009] To achieve the above object, the present invention provides a multi-parameter fusion asphalt mixture fatigue state acoustic emission early warning method, comprising:

[0010] Perform fatigue loading on the asphalt mixture through an acoustic emission test, obtain acoustic emission test data during the test, calculate information entropy values ​​for different parameters of the acoustic emission test data, and obtain key parameters according to the information entropy value calculation results;

[0011] Performing weighted sum calculation on the key parameters to obtain comprehensive evaluation parameters, wherein the weights in the weighted sum are obtained according to a fuzzy comprehensive evaluation method;

[0012] According to the comprehensive evaluation parameters, the dynamic threshold is obtained by the kernel density estimation method;

[0013] According to the dynamic threshold, the key parameters are calculated through a multi-parameter joint distribution model to obtain evaluation parameters, and the evaluation parameters are evaluated according to the dynamic threshold to obtain early warning results of different levels.

[0014] Optionally, fatigue loading is performed on the asphalt mixture through an acoustic emission test system; wherein the acoustic emission test system includes an acoustic emission sensor, a data collector, a data processing unit and a fatigue loading device, wherein the acoustic emission sensor is arranged on the surface of the asphalt mixture, and the asphalt mixture is arranged inside the fatigue loading device, the acoustic emission sensor, the data collector and the data processing unit are connected in sequence, the acoustic emission test data collected by the acoustic emission sensor is obtained through the data collector, and the acoustic emission test data is processed and analyzed by the data processing unit.

[0015] Optionally, the process of calculating information entropy values ​​for different parameters in the acoustic emission test data includes:

[0016] Acquire acoustic emission test data of different groups in different time periods, calculate the information entropy values ​​of different parameters in the acoustic emission test data, and obtain the information entropy of different parameters of different groups in different time periods;

[0017] The information entropy of different parameters of different groups in the same time period is averaged to obtain the information entropy of different parameters in the corresponding time period. The key parameters are obtained by analyzing the information entropy of different parameters in different time periods, including ring count, energy, duration and amplitude.

[0018] Optionally, before weighting and calculating the key parameters, the step includes:

[0019] The key parameters are normalized respectively to obtain normalized key parameters, and subsequent weighted sum calculations are performed using the key parameters.

[0020] Optionally, the process of obtaining comprehensive parameters includes:

[0021] According to the key parameters, the indicator weights corresponding to the key parameters are calculated by the fuzzy comprehensive evaluation method, and the weights are subjected to a consistency check. When the consistency check is met, the indicator weights are used as the weights in the weighted sum, otherwise the indicator weights are adjusted until they meet the consistency check.

[0022] Optionally, the process of obtaining the dynamic threshold includes:

[0023] Through the sliding window method, the comprehensive evaluation parameters in the corresponding time window are obtained, and the calculation threshold is determined according to different percentiles of the comprehensive evaluation parameters in the corresponding time window. The calculation threshold is weighted and calculated with the old threshold to obtain a new threshold, i.e., a dynamic threshold, wherein the new threshold is updated at fixed intervals, and the old threshold is the new threshold before the update.

[0024] Optionally, the process of calculating key parameters through a multi-parameter joint distribution model includes:

[0025] Construct a multi-parameter joint distribution model:

[0026] C(u1,u2,…,u n )=P(Z1≤Φ -1 (u1),Z2≤Φ -1 (u2),…,Z n ≤Φ -1 (u n ))

[0027] Among them, C(u1,u2,…,u n ) is the joint probability of each combination of acoustic emission parameters; u1,u2,…,u n are the marginal cumulative distribution functions of various acoustic emission parameters (such as ring count, energy, duration, amplitude, etc.). The marginal distribution function u is transformed into i Convert to the probability value Φ under the standard normal distribution -1 (u i ). Then, we use the joint probability distribution P(Z1, Z2, …, Z n ), where Z i It follows the standard normal distribution under the correlation matrix R, capturing the dependency structure between different AE parameters. The correlation matrix R describes the correlation between these AE parameters.

[0028] Marginal distribution fitting is performed on different key parameters respectively to obtain different cumulative distribution functions, and the different cumulative distribution functions are calculated through a multi-parameter joint distribution model to obtain corresponding evaluation parameters.

[0029] Optionally, the process of evaluating the evaluation parameters includes:

[0030] When the evaluation parameter exceeds a certain threshold value in the dynamic threshold value, a warning of the corresponding level is triggered, wherein the warning state, i.e., the fatigue state, includes a stable stage, a linear accumulation stage, a rapid development stage, and a critical failure stage.

[0031] Compared with the prior art, the present invention has the following advantages and technical effects:

[0032] Therefore, in order to solve the above problems, the present invention provides a method for identifying and warning the fatigue state of asphalt mixture based on multivariate acoustic emission parameters. This method, combined with the relevant technical means adopted, has the following advantages: 1. Multi-parameter fusion: four key parameters, namely ring count, energy, duration and amplitude, are selected to comprehensively reflect the fatigue state of asphalt mixture; 2. Fuzzy comprehensive evaluation: the fuzzy comprehensive evaluation method is used to determine the weight of each parameter to improve the accuracy and reliability of identification; 3. Multi-level early warning mechanism: a multi-level threshold early warning system of single indicators and comprehensive indicators is established to adapt to different degrees of fatigue damage; 4. Real-time monitoring: combined with the adjustment of the dynamic threshold of the fatigue loading process and the multi-level threshold early warning system, the real-time identification of the fatigue state of asphalt mixture is realized.

[0033] This method based on multivariate acoustic emission parameters and multi-level early warning mechanism can achieve accurate identification and timely early warning of asphalt mixture fatigue state. This method does not rely on complex machine learning algorithms, but through statistical analysis and comprehensive evaluation of acoustic emission parameters, combined with a multi-level early warning mechanism, it provides a scientific and reliable basis for road maintenance decision-making, which helps to realize intelligent and refined management of roads. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0035] Figure 1 A schematic diagram of an acoustic emission test system according to an embodiment of the present invention;

[0036] Figure 2 is a flowchart of an embodiment of the present invention;

[0037] Figure 3 It is an information entropy diagram of each acoustic emission parameter in an embodiment of the present invention;

[0038] Figure 4 It is a comprehensive parameter timing characteristic diagram of an embodiment of the present invention;

[0039] Figure 5 A kernel density estimation curve and a threshold adjustment diagram according to an embodiment of the present invention;

[0040] Among them, 1-acoustic emission sensor, 2-fatigue loading device, 3-data collector, 4-data processing unit. DETAILED DESCRIPTION

[0041] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0042] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0043] The present invention provides a multi-parameter fusion asphalt mixture fatigue state acoustic emission early warning method, which relates to the field of asphalt mixture damage mechanics and crack propagation monitoring. The method builds a test system to conduct an acoustic emission test on an asphalt mixture specimen, and calculates the information entropy of each parameter of the acoustic emission signal during the test. According to the information entropy, ring count, energy, duration, and amplitude are selected as key parameters to screen out effective events. A fuzzy comprehensive evaluation method is used to obtain a comprehensive evaluation parameter, and the fatigue damage evolution law of the asphalt mixture specimen in the acoustic emission test is calculated based on the evolution characterization of the parameter.

[0044] Specifically, the technical solution of the present invention is: ring count, energy, duration, and amplitude are selected as key parameters from the acoustic emission mechanical parameters; each acoustic emission event corresponds to a corresponding parameter, and the weights of different key parameters are obtained by fuzzy comprehensive evaluation method, and the comprehensive evaluation parameters are obtained by weighting and calculating the key parameter information; the fatigue damage evolution law of asphalt mixture specimens in the acoustic emission test is calculated based on the evolution characterization of the comprehensive evaluation parameters. Multiple normalized parameters are integrated into a comprehensive index to reflect the overall fatigue state of the asphalt mixture.

[0045] To achieve the above object, the present invention provides a multi-parameter fusion asphalt mixture fatigue state acoustic emission early warning method, the method comprising the following steps:

[0046] Step 1. Build an acoustic emission test system for monitoring the fatigue state of asphalt mixture. The system includes an acoustic emission sensor 1, a data collector 3, a data processing unit 4 and a fatigue loading device 2. The acoustic emission sensor 1 is set on the asphalt mixture to be tested, the fatigue loading device 2 performs fatigue loading on the asphalt mixture to be tested, the data collector 3 collects the relevant signals collected by the sensor, and processes the relevant signals through the data processing unit 4. The structure of the acoustic emission test system is as follows: Figure 1 As shown;

[0047] Step 2. Perform continuous acoustic emission testing on the asphalt mixture according to step 1, record the acoustic emission signals during fatigue loading at different times, i.e., acoustic emission test data, and obtain and establish an acoustic emission parameter table at different times and in different groups based on the acoustic emission signals;

[0048] Step 3. By calculating and analyzing the information entropy value of each parameter in the acoustic emission signal during the asphalt mixture loading process, four key parameters, namely, ring count N, energy E, duration D, and amplitude A, are selected from different parameters according to the information entropy value to participate in subsequent related calculations and evaluations;

[0049] Step 4. Normalize the ring count, energy, duration, and amplitude parameters;

[0050] Step 5. Establish a multi-parameter fusion evaluation equation and use weighted summation calculation for different key parameters to obtain comprehensive evaluation parameters;

[0051] Step 6. Calculate the weight of each parameter index in the weighted sum calculation in step 5 by fuzzy comprehensive evaluation method;

[0052] Step 7. Calculate the comprehensive evaluation parameters based on the weights in step 6, draw a time series feature diagram, and analyze the evolution law;

[0053] Step 8. Determine the dynamic threshold of each acoustic emission parameter and adjust the dynamic threshold by kernel density estimation;

[0054] Step 9. Combine the parameter thresholds in step 8 and establish a multi-level early warning mechanism based on the multi-parameter joint distribution model;

[0055] Step 10. Based on the multi-level early warning mechanism and comprehensive analysis results, a warning classification model for fatigue status of asphalt mixture is constructed.

[0056] The present invention provides an asphalt mixture fatigue state identification method based on multivariate acoustic emission parameters. By real-time monitoring of the acoustic emission activity of the asphalt mixture and calculating the information entropy value of each acoustic emission parameter, the ring count, energy, duration, and amplitude in the acoustic emission signal are used as key parameters according to the information entropy value to screen out effective events in the entire acoustic emission mechanical test process. The evolution law of the fatigue state of the asphalt mixture during the acoustic emission test is calculated by the evolution characterization of effective events, providing technical support for non-destructive monitoring of intelligent roads.

[0057] The above technical solution of the present invention is described in detail with reference to the accompanying drawings:

[0058] In order to clearly illustrate the technical features of this patent, this patent is elaborated in detail below through a specific implementation method and in combination with its accompanying drawings.

[0059] Taking the fatigue acoustic emission test of a 380*50*63mm standard asphalt mixture four-point bending beam as an example, the crack extension of the asphalt mixture specimen under fatigue loading was obtained, such as Figure 1 As shown, the specific steps include:

[0060] Step 1: A 380*50*63mm standard asphalt mixture four-point bending beam specimen is fatigue loaded using a fatigue loading device 2, and multiple acoustic emission sensors 1 are pasted on the surface of the asphalt mixture specimen. The acoustic emission sensor 1 is connected to the signal amplifier through a data line, and then connected to the acoustic emission signal acquisition analyzer, i.e., the data collector 3, and the computer host, i.e., the data processing unit 4, through a data transmission line. When crack propagation occurs in the asphalt mixture specimen during fatigue loading, the acoustic emission signal is transmitted to the computer host, and the acoustic signal can be converted into a digital signal through signal conversion, as shown in the parameter table in Table 1.

[0061] Step 2: Record the acoustic emission test data during fatigue loading of the asphalt mixture specimen, as shown in Table 1 below. Table 1 is the acoustic emission test data of the asphalt mixture fatigue loading test. The serial number in the table is the number of the acoustic emission event corresponding to each acoustic emission signal, i.e., the acoustic emission test data, and the acoustic emission parameters in the table refer to the various parameters of the signal waveform when each acoustic emission event occurs.

[0062] Table 1

[0063]

[0064] Step 3: Select 8 different time periods of the fatigue process of asphalt mixture, take 10 groups of acoustic emission test data in each time period, calculate the information entropy value of each parameter of these 10 groups of data respectively, and average the calculation results. The average is the information entropy of each parameter in this time period, and analyze the evolution law of the information entropy of each parameter. In the above process, the correspondence and evolution characteristics of the information entropy evolution law of the parameters and the change trend of the original parameters are analyzed. It is found that the information entropy evolution law of the four parameters of ring count CNTS, energy E, duration D and amplitude A has a clear correspondence with the change trend of the original parameters, and can highlight the evolution characteristics of the original parameters. This shows that the amount of information contained in these four parameters can more accurately reflect the fatigue state of asphalt mixture in different time periods. Therefore, the four parameters of ring count CNTS, energy E, duration D and amplitude A are selected as key parameters, such as Figure 3 The formula for calculating the information entropy of each parameter is:

[0065]

[0066] X={x1,x2,…,x n} is an acoustic emission parameter data, i represents the corresponding information number, H(X) represents the entropy value of the random variable X, P(x i ) indicates that the random variable X takes the value x i The probability of n is the number of possible values, log b Represents the logarithm with base b.

[0067] Step 4: In view of the differences in the distribution dimensions of the key parameters, in order to ensure the uniform contribution of these key parameters to the distance metric in the comprehensive evaluation, the key parameter information of the ring count, energy, duration, and amplitude parameters is normalized to standardize the data within the range of [0,1]. The normalization calculation formula is:

[0068]

[0069] Where: X n is the value after normalization calculation, X is the original value, and X min and X max are the minimum and maximum values ​​under this feature respectively;

[0070] Step 5: In order to integrate the information of multiple key acoustic emission parameters, a more comprehensive description of fatigue state is provided. This method not only considers the importance of a single parameter, but also reflects the interaction between parameters. Therefore, the evaluation equation of multi-parameter fusion is established and the weighted summation is used for calculation. The comprehensive parameter C = w1N + w2E + w3D + w4A, where C is the comprehensive evaluation parameter, N, E, D, A are the normalized ring count, energy, duration and amplitude, respectively, and w1, w2, w3, w4 are the corresponding weights.

[0071] Step 6: Calculate the weights of each key parameter index by fuzzy comprehensive evaluation method. As shown in Table 2, the judgment matrix of each key parameter of acoustic emission is used to calculate λ max =4.26, CR=0.09<1, the matrix consistency test passed.

[0072] Table 2

[0073]

[0074] Step 7: Calculate the comprehensive evaluation parameter C, C = 0.12 × N + 0.18 × E + 0.43 × D + 0.27 × A, and draw the comprehensive parameter time series characteristic diagram, such as Figure 4 shown.

[0075] Step 8: Dynamic threshold determination method.

[0076] In the process of determining the threshold, a dynamic threshold method based on non-parametric statistics is adopted. Taking the comprehensive evaluation parameter C in a certain time period as an example, the 75% quantile Q1 is calculated to be 0.05, the 90% quantile Q2 = 0.07, and the 95% quantile Q3 = 0.10. Therefore, the initial threshold is set as: first-level warning: C>Q1 = 0.05, second-level warning: C>Q2 = 0.07, and third-level warning: C>Q3 = 0.10, as shown in Table 3, which shows the warning thresholds for different time periods.

[0077] The probability density function of the data is obtained using the kernel density estimation method, and the threshold is fine-tuned based on the characteristics of the kernel density estimation curve. The results are as follows: Figure 5 As shown. In order to adapt to the dynamic changes in data distribution, a 6-minute sliding time window is used, and the statistics are updated every 3 minutes. The calculated threshold is the threshold calculated by the sliding time window, the old threshold is the threshold used for evaluation before fine-tuning, and the new threshold is weighted averaged with the old threshold by the exponentially weighted moving average method:

[0078] New threshold = α × calculated threshold + (1-α) × old threshold

[0079] Where: α is the smoothing coefficient, which is 0.3.

[0080] Table 3

[0081]

[0082] Step 9: Establish a multi-level early warning mechanism

[0083] For the multi-parameter combination trigger condition, the present invention adopts the joint probability density function to describe the complex dependency structure between multiple acoustic emission parameters, which can effectively capture nonlinear correlation and is suitable for describing the dependency relationship between multi-dimensional random variables. The Gaussian Copula function family is used to construct a multi-parameter joint distribution model, which is defined as follows:

[0084] C(u1,u2,…,u n )=P(Z1≤Φ -1 (u1),Z2≤Φ -1 (u2),…,Z n ≤Φ -1 (u n ))

[0085] Among them, C(u1,u2,…,u n ) is the joint probability of each combination of acoustic emission parameters; u1,u2,…,u n are the marginal cumulative distribution functions of various acoustic emission parameters (such as ring count, energy, duration, amplitude, etc.). The marginal distribution function u is transformed into iConvert to the probability value Φ under the standard normal distribution -1 (u i ). Then, we use the joint probability distribution P(Z1, Z2, …, Z n ), where Z i It follows the standard normal distribution under the correlation matrix R, capturing the dependency structure between different AE parameters. The correlation matrix R describes the correlation between these AE parameters.

[0086] The marginal distribution is fitted for each acoustic emission parameter (ring count N, energy E, duration D, amplitude A) to obtain its cumulative distribution function F i (x i );Use the maximum likelihood estimation method to calculate the correlation matrix R of the multi-parameter joint distribution model;Construct the joint distribution function F(x1,x2,...,x n )=C(F1(x1),F2(x2),...,F n (x n )); Combined with the multi-parameter combined warning threshold calculated in step 8, when C(u1,u2,...,u n ) exceeds the threshold, triggering an early warning of the corresponding level.

[0087] Step 10: Based on the multi-level early warning mechanism and comprehensive analysis results, a fatigue state classification model for asphalt mixture is constructed.

[0088] Based on the warning trigger and duration, fatigue status is divided into the following four levels:

[0089] a) Stable stage: No warning is triggered or the first-level warning is triggered only occasionally. (Occasionally: within a 30-minute period, the warning is triggered no more than 3 times, and the time interval between each trigger is greater than 10 minutes)

[0090] b) Linear accumulation stage: Level 1 warning is triggered frequently, and Level 2 warning is triggered occasionally. (Frequent: within a 30-minute period, the warning is triggered 5 times or more, and the time interval between each trigger is less than 5 minutes)

[0091] c) Rapid development stage: The second-level warning is continuously triggered, and the third-level warning begins to appear. (Continuous: Within 1 hour, the interval between each triggering of the warning signal is less than 3 minutes, and the signal exceeds 70% of the time period within 1 hour)

[0092] d) Critical failure stage: Level 3 warning is frequently triggered, and the comprehensive parameter C shows a sharp upward trend.

[0093] If you need to further adjust the above triggering criteria, you can optimize the settings based on actual needs, experimental data or historical experience, which will not be elaborated here.

[0094] This method can accurately evaluate the fatigue state of asphalt mixture and provide a basis for timely detection of early road damage. Through accurate monitoring of the fatigue state of asphalt mixture, it is possible to guide the optimization decision of road maintenance work and determine the best maintenance time and method, thereby extending the service life of the road and improving maintenance efficiency. The present invention provides key technical support for realizing intelligent road non-destructive monitoring and scientific road maintenance, and helps to establish a more efficient and economical road maintenance system.

[0095] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A multi-parameter fusion asphalt mixture fatigue state acoustic emission early warning method, characterized in that: include: Perform fatigue loading on the asphalt mixture through an acoustic emission test, obtain acoustic emission test data during the test, calculate information entropy values ​​for different parameters of the acoustic emission test data, and obtain key parameters according to the information entropy value calculation results; Performing weighted sum calculation on the key parameters to obtain comprehensive evaluation parameters, wherein the weights in the weighted sum are obtained according to a fuzzy comprehensive evaluation method; According to the comprehensive evaluation parameters, the dynamic threshold is obtained by the kernel density estimation method; According to the dynamic threshold, the key parameters are calculated through a multi-parameter joint distribution model to obtain evaluation parameters, and the evaluation parameters are evaluated according to the dynamic threshold to obtain early warning results of different levels.

2. The method according to claim 1, characterized in that: The asphalt mixture is fatigue loaded by an acoustic emission test system; wherein the acoustic emission test system comprises an acoustic emission sensor, a data collector, a data processing unit and a fatigue loading device, wherein the acoustic emission sensor is arranged on the surface of the asphalt mixture, and the asphalt mixture is arranged inside the fatigue loading device, the acoustic emission sensor, the data collector and the data processing unit are connected in sequence, the acoustic emission test data collected by the acoustic emission sensor is obtained by the data collector, and the acoustic emission test data is processed and analyzed by the data processing unit.

3. The method according to claim 1, characterized in that: The process of calculating the information entropy value of different parameters in the acoustic emission test data includes: Acquire acoustic emission test data of different groups in different time periods, calculate the information entropy values ​​of different parameters in the acoustic emission test data, and obtain the information entropy of different parameters of different groups in different time periods; The information entropy of different parameters of different groups in the same time period is averaged to obtain the information entropy of different parameters in the corresponding time period. The key parameters are obtained by analyzing the information entropy of different parameters in different time periods, including ring count, energy, duration and amplitude.

4. The method according to claim 1, characterized in that The key parameters are weighted and calculated before: The key parameters are normalized respectively to obtain normalized key parameters, and subsequent weighted sum calculations are performed based on the key parameters.

5. The method according to claim 1, characterized in that The process of obtaining comprehensive parameters includes: According to the key parameters, the indicator weights corresponding to the key parameters are calculated by the fuzzy comprehensive evaluation method, and the weights are subjected to a consistency check. When the consistency check is met, the indicator weights are used as the weights in the weighted sum, otherwise the indicator weights are adjusted until they meet the consistency check.

6. The method according to claim 1, characterized in that The process of obtaining the dynamic threshold includes: Through the sliding window method, the comprehensive evaluation parameters in the corresponding time window are obtained, and the calculation threshold is determined according to different percentiles of the comprehensive evaluation parameters in the corresponding time window. The calculation threshold is weighted and calculated with the old threshold to obtain a new threshold, i.e., a dynamic threshold, wherein the new threshold is updated at fixed intervals, and the old threshold is the new threshold before the update.

7. The method according to claim 1, characterized in that The process of calculating key parameters through the multi-parameter joint distribution model includes: Construct a multi-parameter joint distribution model: C(u1,u2,…,u n )=P(Z1≤Φ -1 (u1),Z2≤Φ -1 (u2),…,Z n ≤Φ -1 (u n )) Among them, C(u1,u2,…,u n ) is the joint probability of each acoustic emission parameter, i.e., the key parameters combined; u1,u2,…,u n are the marginal cumulative distribution functions of each acoustic emission parameter, Φ is the standard normal distribution function, P(Z1, Z2, …, Z n ) is the multidimensional normal distribution function, where Z i It obeys the standard normal distribution under the correlation matrix R, where the correlation matrix R is a matrix describing the correlation between the acoustic emission parameters; Marginal distribution fitting is performed on different key parameters respectively to obtain different cumulative distribution functions, and the different cumulative distribution functions are calculated through a multi-parameter joint distribution model to obtain corresponding evaluation parameters.

8. The method according to claim 1, characterized in that The process of evaluating the evaluation parameters includes: When the evaluation parameter exceeds a certain threshold value in the dynamic threshold value, a warning of the corresponding level is triggered, wherein the warning state, i.e., the fatigue state, includes a stable stage, a linear accumulation stage, a rapid development stage, and a critical failure stage.

Images