System and method for testing wave velocity in concrete pressure-bearing state based on machine learning

Abstract

The invention discloses a system and a method for testing a wave velocity in a concrete pressure-bearing state based on machine learning, and relates to the field of wave velocity measurement. Whether the parameters of the sound wave measurement transducer need to be adjusted is judged by judging whether the noise level data in the received sound wave waveform image meets the requirement, so that the accuracy of measuring the time data of the ultrasonic wave passing through the concrete is improved; by constructing a waveform image mapping model, a mapping tool is provided for obtaining waveform image noise level data corresponding to collected historical received sound wave waveform image data; by constructing the final sound wave waveform noise level mapping equation, a mapping model is provided for subsequently adjusting the test parameter data of the sound wave test transducer and the sound wave receiving amplification factor data during the current wave velocity test, so that the wave velocity test experiment does not need to be carried out after each adjustment, and the operability is improved.

Description

Technical Field

[0001] This invention belongs to the field of wave velocity measurement, and more specifically, relates to a wave velocity testing system and method for concrete under compressive conditions based on machine learning. Background Technology

[0002] There is a quantitative relationship between the strength of concrete and its acoustic wave velocity. The wave velocity of concrete specimens under no-stress conditions can be independently tested using an ultrasonic device. However, concrete structures bear loads and are primarily under compression. In the initial stage of compression, the internal pores of concrete decrease, becoming denser, and the wave velocity increases. In the middle stage of compression, the tensile stress inside the concrete gradually increases, and cracks begin to appear at the contact points between coarse and fine aggregates, reaching the maximum wave velocity. In the later stage of compression, the internal cracks gradually enlarge, forming through cracks, and the wave velocity decreases.

[0003] Because the relevant parameters of the wave velocity measurement transducer were not precisely and reasonably set before the experiment, there may be a lot of noise data in the ultrasonic waveform image received during the wave velocity test. This may affect the process of measuring the time data of ultrasonic waves passing through concrete, resulting in inaccurate passage time data and consequently inaccurate wave velocity calculation results in concrete. Summary of the Invention

[0004] To address the problems in related technologies, this invention proposes a machine learning-based wave velocity testing system and method for concrete under compressive conditions, in order to overcome the aforementioned technical problems existing in the prior art.

[0005] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution:

[0006] This invention relates to a machine learning-based method for testing wave velocity in concrete under compressive conditions, comprising the following steps:

[0007] S1. Collect multiple sets of historical transducer experimental parameter data, acoustic wave amplification factor data, concrete property parameter data, and received acoustic wave waveform image data during historical wave velocity tests to obtain historical transducer parameter data matrix, historical acoustic wave amplification factor dataset, historical concrete property parameter data matrix, and historical received waveform image dataset; collect multiple sets of existing acoustic wave waveform image data and corresponding waveform noise level data to obtain existing acoustic wave waveform image dataset and existing waveform noise level dataset.

[0008] S2. Construct a waveform image mapping model using the existing acoustic waveform image dataset and the existing waveform noise level dataset, and map the noise level data corresponding to the historical received waveform image dataset to obtain the historical received waveform image noise level dataset.

[0009] S3. The final acoustic waveform noise level mapping equation is constructed using the historical transducer parameter data matrix, the historical acoustic amplification factor dataset, the historical concrete property parameter data matrix, and the historical received waveform image noise level dataset.

[0010] S4. Adjust the test parameters of the acoustic transducer and the acoustic receiving amplification factor data during the current wave velocity test in accordance with the final acoustic waveform noise level mapping equation. After the adjustment is completed, conduct the wave velocity test of concrete under pressure.

[0011] By collecting multiple sets of historical wave velocity test transducer experimental parameter data, acoustic wave receiving amplification factor data, concrete property parameter data, and received acoustic wave waveform image data, data support was provided for the subsequent construction of the final acoustic wave waveform noise level mapping equation. By collecting multiple sets of existing acoustic wave waveform image data and corresponding waveform noise level data, a data foundation was provided for the subsequent construction of the waveform image mapping model. The constructed waveform image mapping model provided a mapping tool for obtaining the waveform image noise level data corresponding to the collected historical received acoustic wave waveform image data. By mapping the noise level data corresponding to the historical received waveform image dataset, label data was provided for the subsequent construction of the final acoustic wave waveform noise level mapping equation. Finally, the constructed final acoustic wave waveform noise level mapping equation provided a mapping model for subsequent adjustments to the experimental parameter data of the acoustic transducer and the acoustic wave receiving amplification factor data during the current wave velocity test, thus eliminating the need for wave velocity testing after each adjustment and improving operability.

[0012] Preferably, step S1 includes the following steps:

[0013] S11. Set various test parameter types for the acoustic wave test transducer to obtain the transducer test parameter type set. a 1i This indicates the type of the i-th test parameter set for the acoustic wave test transducer. This indicates the total number of test parameter types set for the acoustic wave test transducer; then, various property parameters of concrete are set to obtain the concrete property parameter type set. a 2i This indicates the type of the i-th attribute parameter of the concrete. This indicates the total number of types of concrete property parameters that are set.

[0014] S12. Based on the transducer experimental parameter type set and the concrete property parameter type set, when historically using acoustic wave testing transducers with different parameters to conduct wave velocity tests on concrete with different property parameters, the transducer experimental parameter data, acoustic wave receiving amplification factor data, concrete property parameter data, and received acoustic wave waveform image data of the corresponding acoustic wave testing transducer are collected to obtain a historical transducer parameter data matrix. Historical sound wave amplification dataset Historical concrete property parameter data matrix and historical received waveform image dataset , , These represent the acoustic amplification factor data and the corresponding received acoustic waveform image data collected during the i-th historical wave velocity test of concrete, respectively. This indicates the total number of historical wave velocity tests conducted on concrete. , They are as follows:

[0015] ; ;

[0016] in, , These represent the experimental parameter data of the j-th type of transducer and the concrete property parameter data collected during the i-th historical wave velocity test of concrete, respectively.

[0017] S13. Collect multiple sets of existing acoustic waveform image data and corresponding waveform noise level data to obtain an existing acoustic waveform image dataset. and existing waveform noise level datasets b 1i b 2i These represent the i-th existing acoustic waveform image data and the corresponding waveform noise level data, respectively. This indicates the total number of existing acoustic waveform image data collected. The total number of existing acoustic waveform image data collected is the same as the total number of corresponding waveform noise level data.

[0018] The transducer experimental parameters include frequency, power, amplitude, operating voltage, impedance, coupling coefficient, transmission power, duty cycle, and operating frequency; the concrete property parameters include cement density, water density, sand density, aggregate density, and compressive strength. When measuring the wave velocity of internal ultrasonic waves in concrete under different pressure conditions, the time of the ultrasonic signal is often measured to calculate the wave velocity; therefore, the quality of the ultrasonic waveform image data is crucial. This solution collects multiple sets of existing acoustic waveform image data and corresponding waveform noise level data, providing data support for the subsequent construction of a waveform image mapping model, thus providing data support for obtaining the noise level data of each acoustic waveform image.

[0019] Preferably, step S2 includes the following steps:

[0020] S21. Construct a waveform image mapping model using the existing acoustic waveform image dataset and the existing waveform noise level dataset;

[0021] S22. Input the historical received waveform image dataset into the waveform image mapping model for mapping to obtain the historical received waveform image noise level dataset. , This represents the noise level data of the received acoustic waveform image data collected during historical wave velocity tests on concrete.

[0022] By mapping the historical received waveform image dataset, the waveform image data is transformed into digital data, reducing the complexity of the data and making it easier to construct the final acoustic waveform noise level mapping equation based on the data.

[0023] Preferably, step S3 includes the following steps:

[0024] S31. Construct the initial acoustic waveform noise level mapping equation; as follows.

[0025] ;

[0026] in, The dependent variable of the initial acoustic waveform noise level mapping equation represents the received waveform image noise level data; ceil represents the rounding function. Let be the i-th independent variable in the initial acoustic waveform noise level mapping equation, representing the transducer parameter data of the i-th type. for The coefficients of the independent variables; The first equation of the initial acoustic waveform noise level mapping equation There are *i* independent variables, representing the concrete property parameter data for the *i*th type. for The coefficients of the independent variables; The first equation of the initial sound wave waveform noise level mapping equation There are 1 independent variable, representing the sound wave amplification factor data. for The coefficients of the independent variables; This is the bias of the initial acoustic waveform noise level mapping equation;

[0027] S32. Substitute the historical transducer parameter data matrix, the historical acoustic amplification dataset, and the historical concrete property parameter data matrix into the initial acoustic waveform noise level mapping equation to perform mapping, thereby obtaining the historical acoustic waveform image noise level mapping dataset. c i This refers to the historical acoustic waveform image noise level mapping data obtained by inputting the data in the i-th row of the historical transducer parameter data matrix and the data in the historical concrete property parameter data matrix, as well as the i-th data in the historical acoustic amplification factor dataset, into the initial acoustic waveform noise level mapping equation.

[0028] S33. Set a mapping error threshold; calculate the Euclidean distance between the historical acoustic waveform image noise level mapping dataset and the historical received waveform image noise level dataset to obtain noise level mapping error data; when the noise level mapping error data is greater than or equal to the mapping error threshold, optimize the initial acoustic waveform noise level mapping equation until the noise level mapping error data is less than the mapping error threshold to obtain the final acoustic waveform noise level mapping equation; otherwise, there is no need to optimize the initial acoustic waveform noise level mapping equation, and the initial acoustic waveform noise level mapping equation is used as the final acoustic waveform noise level mapping equation.

[0029] By substituting the historical transducer parameter data matrix, historical acoustic amplification dataset, and historical concrete property parameter data matrix into the initial acoustic waveform noise level mapping equation for pre-mapping, it can be determined whether the current independent variable coefficients and biases ensure good mapping accuracy of the acoustic waveform noise level mapping equation. If the requirements are not met, the independent variable coefficients and biases are adjusted, thus ensuring the mapping accuracy of the mapping equation and providing a basis for determining whether adjustments to the independent variable coefficients and biases of the initial acoustic waveform noise level mapping equation are necessary.

[0030] Preferably, step S4 includes the following steps:

[0031] S41. Based on the transducer experimental parameter type set and the concrete property parameter type set, collect the experimental parameter data of the acoustic transducer, the property parameters of the concrete, and the acoustic wave receiving amplification factor data during the current wave velocity test to obtain the current transducer parameter dataset. Current concrete property parameter dataset and current sound wave receiving amplification data ; , These represent the test parameter data of the i-th type of acoustic wave test transducer and the property parameters of the concrete, respectively, during the current wave velocity test.

[0032] S42. Input the current transducer parameter dataset, the current concrete property parameter dataset, and the current acoustic wave receiving amplification factor data into the final acoustic wave waveform noise level mapping equation for mapping, and obtain the current acoustic wave waveform noise level data e.

[0033] S43. Adjust the current transducer parameter dataset and the current acoustic wave receiving amplification factor data according to the current acoustic wave waveform noise level data to obtain the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data.

[0034] S44. Set the test parameters of the acoustic wave test transducer for the current wave velocity test using the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data; after setting, the set acoustic wave test transducer is obtained; use the set acoustic wave test transducer to perform wave velocity test on concrete.

[0035] By adjusting the relevant parameters of the wave velocity measurement transducer and the amplification factor of the sound wave receiver in conjunction with the previously constructed final sound wave waveform noise level mapping equation, the noise of the received waveform image is reduced, thereby increasing the accuracy of the measurement.

[0036] Preferably, step S43 includes the following steps:

[0037] S431, Set the threshold for waveform image noise level;

[0038] S432. When the current acoustic waveform noise level data is greater than or equal to the waveform image noise level threshold, the current transducer parameter dataset and the current acoustic wave receiving amplification factor data are adjusted until the current acoustic waveform noise level data is less than the waveform image noise level threshold, thus obtaining the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data; otherwise, there is no need to adjust the current transducer parameter dataset and the current acoustic wave receiving amplification factor data, and the current transducer parameter dataset and the current acoustic wave receiving amplification factor data are respectively used as the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data.

[0039] By setting a threshold for the noise level of the waveform image, a quantitative basis is provided for determining whether the noise level data of the current sound waveform meets the test requirements.

[0040] Preferably, adjusting the current transducer parameter dataset and the current acoustic wave receiving amplification factor data in S432 includes the following steps:

[0041] S4321. Constructing wave speed test parameters to optimize the Mustang population , This indicates that the wave speed test parameters are optimized for the i-th horse in the wild horse population. The wave speed test parameters are used to optimize the size of the Przewalski's horse population. The maximum number of iterations for optimizing the Przewalski's horse population using the wave speed test parameters is set to [value missing]. And the current iteration number is Let be denoted as the maximum number of iterations for wave speed test parameter optimization and the current number of iterations for wave speed test parameter optimization, respectively; the search space dimension for wave speed test parameter optimization of the Przewalski's horse population is . ;

[0042] S4322. Set the range of test parameter data for each type of acoustic transducer and the range of acoustic wave receiving amplification factor data during the current wave velocity test to obtain the set of transducer parameter value ranges. and the range of magnification values ​​[d] 41 ,d 42 ]; , These represent the lower and upper limits of the experimental parameters for the i-th type of acoustic transducer during the current wave velocity test, respectively; d 41 d 42 These represent the lower and upper limits of the acoustic wave reception amplification factor data during the current wave velocity test, respectively.

[0043] Based on the set of transducer parameter values ​​and the range of amplification factor values, the wave velocity test parameters are set to optimize the initial position of each wild horse in the wild horse population, thus obtaining a second initial position matrix. ;as follows,

[0044] ;

[0045] in, The wave speed test parameters optimize the initial position of the kth wild horse in the wild horse population, which is the position component in the i-th search space dimension. , The calculation formulas are as follows:

[0046] ;

[0047] ;

[0048] In the formula, rand 2ki , They represent respectively targeting , Generate random numbers between 0 and 1;

[0049] S4323. Based on the current acoustic waveform noise level data e, construct wave speed test parameters to optimize the fitness function of the wild horse population. ;as follows,

[0050] ;

[0051] S4324. Begin iteration. Before iteration, set the current iteration count of the wave speed test parameter optimization to 1. During the first iteration, use the fitness function of the wave speed test parameter optimization for the wild horse population to calculate the fitness value of the initial position of each wild horse in the second initial position matrix, obtaining a third fitness value set. Take the largest fitness value in the third fitness value set and the corresponding initial position of the wild horse as the third global best fitness and the third global best position, respectively. Update the initial position of each wild horse in the second initial position matrix according to the third global best fitness and the third global best position. After the update is completed, increment the current iteration count of the wave speed test parameter optimization by 1 and enter the next iteration.

[0052] In each iteration, the fitness function of the Przewalski's horse population optimized by the wave speed test parameters is used to calculate the fitness value of each horse's position in the wave speed test parameter optimized Przewalski's horse population updated in the previous iteration, resulting in a fourth fitness value set. The largest fitness value in the fourth fitness value set and the corresponding horse's position are taken as the fourth global best fitness and the fourth global best position, respectively. The position of each horse in the wave speed test parameter optimized Przewalski's horse population updated in the previous iteration is updated according to the fourth global best fitness and the fourth global best position. After the update is completed, the current iteration number of the wave speed test parameter optimization is incremented by 1 and the next iteration begins.

[0053] S4325, when If the first optimal position is reached, stop the iteration and obtain the second final global optimal position; otherwise, continue the iteration until the optimal position is reached. Up to this point; substitute each position component of the second final global optimal position and the current concrete property parameter dataset into the final acoustic waveform noise level mapping equation for mapping, and obtain the current optimized acoustic waveform noise level data;

[0054] When the current optimized acoustic waveform noise level data is less than the waveform image noise level threshold, the corresponding position components in the second final global optimal position are respectively used as the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data; otherwise, return to S4324 to continue iterating until the current optimized acoustic waveform noise level data is less than the waveform image noise level threshold.

[0055] In this scheme, the relevant parameters of the wave velocity measurement transducer and the acoustic wave receiving amplification factor data are simultaneously adjusted multiple times using the Mustang optimization algorithm. At the same time, the noise level of the waveform image corresponding to the relevant parameters of the wave velocity measurement transducer and the acoustic wave receiving amplification factor data obtained in each iteration is used as the fitness function. Therefore, as the iteration proceeds, the noise of the finally received waveform image data becomes lower and lower, which makes it easier to accurately measure the time data of ultrasonic waves passing through concrete.

[0056] The machine learning-based concrete wave velocity testing system under compressive conditions includes a wave velocity test parameter type setting module, a historical wave velocity test parameter data acquisition module, an existing wave velocity test parameter data acquisition module, a historical waveform image noise level data mapping module, an acoustic waveform noise level mapping equation construction module, a current wave velocity test parameter optimization module, and a wave velocity test module.

[0057] The present invention has the following beneficial effects:

[0058] 1. This invention determines whether the parameters of the acoustic wave measuring transducer need adjustment by examining the noise level data in the received acoustic waveform image to ensure that the noise in the received acoustic waveform image is minimal, thereby improving the accuracy of the measured time data of ultrasonic waves passing through concrete. Furthermore, by constructing a waveform image mapping model, a mapping tool is provided for acquiring the waveform image noise level data corresponding to the collected historical received acoustic waveform image data. Finally, by constructing a final acoustic waveform noise level mapping equation, a mapping model is provided for subsequent adjustments to the experimental parameters of the acoustic wave measuring transducer and the acoustic wave receiving amplification factor data during the current wave velocity test. This eliminates the need for wave velocity testing after each adjustment, improving operability.

[0059] 2. In this invention, by mapping the historical received waveform image dataset, the waveform image data is converted into digital data, which reduces the complexity of the data and makes it easier to construct the final acoustic waveform noise level mapping equation based on the data.

[0060] 3. In this invention, the relevant parameters of the wave velocity measurement transducer and the acoustic wave receiving amplification factor data are simultaneously adjusted iteratively multiple times using the Mustang optimization algorithm. At the same time, the noise level of the waveform image corresponding to the relevant parameters of the wave velocity measurement transducer and the acoustic wave receiving amplification factor data obtained in each iteration is used as the fitness function. Therefore, as the iteration proceeds, the noise of the finally received waveform image data becomes lower and lower, which facilitates the accurate measurement of the time data of ultrasonic waves passing through concrete.

[0061] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0062] To more clearly illustrate the technical solutions of the embodiments of the invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, the drawings can be obtained from these drawings without creative effort.

[0063] Figure 1 This is a flowchart illustrating the wave velocity testing method for concrete under compressive conditions based on machine learning, as described in this invention.

[0064] Figure 2 This is a schematic diagram of the wave velocity testing device for concrete under pressure in this invention, where F represents the pressure borne by the concrete. Detailed Implementation

[0065] The technical solutions of the embodiments of the invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the invention, and not all embodiments. Based on the embodiments of the invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the invention.

[0066] Example 1

[0067] Please see Figure 1 This embodiment describes a machine learning-based method for testing wave velocity in concrete under compressive conditions, including the following steps:

[0068] (1) Adjust the parameters of the acoustic wave test transducer; including the following steps:

[0069] S1. Collect multiple sets of historical transducer experimental parameter data, acoustic wave amplification factor data, concrete property parameter data, and received acoustic wave waveform image data during historical wave velocity tests to obtain historical transducer parameter data matrix, historical acoustic wave amplification factor dataset, historical concrete property parameter data matrix, and historical received waveform image dataset; collect multiple sets of existing acoustic wave waveform image data and corresponding waveform noise level data to obtain existing acoustic wave waveform image dataset and existing waveform noise level dataset.

[0070] S1 includes the following steps:

[0071] S11. Set various test parameter types for the acoustic wave test transducer to obtain the transducer test parameter type set. a 1i This indicates the type of the i-th test parameter set for the acoustic wave test transducer. This indicates the total number of test parameter types set for the acoustic wave test transducer; then, various property parameters of concrete are set to obtain the concrete property parameter type set. a 2i This indicates the type of the i-th attribute parameter of the concrete. This indicates the total number of concrete property parameter types set.

[0072] S12. Based on the transducer experimental parameter type set and the concrete property parameter type set, when historically using acoustic wave testing transducers with different parameters to conduct wave velocity tests on concrete with different property parameters, the transducer experimental parameter data, acoustic wave receiving amplification factor data, concrete property parameter data, and received acoustic wave waveform image data of the corresponding acoustic wave testing transducer are collected to obtain a historical transducer parameter data matrix. Historical sound wave amplification dataset Historical concrete property parameter data matrix and historical received waveform image dataset , , These represent the acoustic amplification factor data and the corresponding received acoustic waveform image data collected during the i-th historical wave velocity test of concrete, respectively. This indicates the total number of historical wave velocity tests conducted on concrete. , They are as follows:

[0073] ; ;

[0074] in, , These represent the experimental parameter data of the j-th type of transducer and the concrete property parameter data collected during the i-th historical wave velocity test of concrete, respectively.

[0075] S13. Collect multiple sets of existing acoustic waveform image data and corresponding waveform noise level data to obtain an existing acoustic waveform image dataset. and existing waveform noise level datasets b 1i b 2i These represent the i-th existing acoustic waveform image data and the corresponding waveform noise level data, respectively. This indicates the total number of existing acoustic waveform image data collected. The total number of existing acoustic waveform image data collected is the same as the total number of corresponding waveform noise level data.

[0076] S2. Construct a waveform image mapping model using the existing acoustic waveform image dataset and the existing waveform noise level dataset, and map the noise level data corresponding to the historical received waveform image dataset to obtain the historical received waveform image noise level dataset.

[0077] S2 includes the following steps:

[0078] S21. Construct a waveform image mapping model using the existing acoustic waveform image dataset and the existing waveform noise level dataset;

[0079] S21 includes the following steps:

[0080] S211. Construct an initial CNN classification model and set a first training data ratio; divide the existing acoustic waveform image dataset and the existing waveform noise level dataset according to the first training data ratio to obtain an existing acoustic waveform image training dataset, an existing waveform noise level training dataset, an existing acoustic waveform image test dataset, and an existing waveform noise level test dataset.

[0081] S212. Set a first training error threshold; input the existing acoustic waveform image training dataset and the existing waveform noise level training dataset as training data and training labels respectively into the initial CNN classification model for training; during the training process, when the training error is less than the first training error threshold, stop training and obtain the trained CNN classification model; otherwise, continue training until the training error is less than the first training error threshold.

[0082] S213. Set a first test accuracy threshold; input the existing acoustic waveform image test dataset and the existing waveform noise level test dataset as test data and test labels respectively into the trained CNN classification model for testing; after the test is completed, obtain the first test accuracy; when the first test accuracy is greater than or equal to the first test accuracy threshold, use the trained CNN classification model as the waveform image mapping model; otherwise, return to S212 to continue training the trained CNN classification model until the first test accuracy is greater than or equal to the first test accuracy threshold, and obtain the waveform image mapping model;

[0083] S22. Input the historical received waveform image dataset into the waveform image mapping model for mapping to obtain the historical received waveform image noise level dataset. , This represents the noise level data of the received acoustic waveform image data collected during historical wave velocity tests on concrete.

[0084] S3. The final acoustic waveform noise level mapping equation is constructed using the historical transducer parameter data matrix, the historical acoustic amplification factor dataset, the historical concrete property parameter data matrix, and the historical received waveform image noise level dataset.

[0085] S3 includes the following steps:

[0086] S31. Construct the initial acoustic waveform noise level mapping equation; as follows.

[0087] ;

[0088] in, The dependent variable of the initial acoustic waveform noise level mapping equation represents the received waveform image noise level data; ceil represents the rounding function. Let be the i-th independent variable in the initial acoustic waveform noise level mapping equation, representing the transducer parameter data of the i-th type. for The coefficients of the independent variables; The first equation of the initial acoustic waveform noise level mapping equation There are *i* independent variables, representing the concrete property parameter data for the *i*th type. for The coefficients of the independent variables; The first equation of the initial acoustic waveform noise level mapping equation There are 1 independent variable, representing the sound wave amplification factor data. for The coefficients of the independent variables; This is the bias of the initial acoustic waveform noise level mapping equation;

[0089] S32. Substitute the historical transducer parameter data matrix, the historical acoustic amplification dataset, and the historical concrete property parameter data matrix into the initial acoustic waveform noise level mapping equation to perform mapping, thereby obtaining the historical acoustic waveform image noise level mapping dataset. c i This refers to the historical acoustic waveform image noise level mapping data obtained by inputting the data in the i-th row of the historical transducer parameter data matrix and the data in the historical concrete property parameter data matrix, as well as the i-th data in the historical acoustic amplification factor dataset, into the initial acoustic waveform noise level mapping equation.

[0090] S33. Set a mapping error threshold; calculate the Euclidean distance between the historical acoustic waveform image noise level mapping dataset and the historical received waveform image noise level dataset to obtain noise level mapping error data; when the noise level mapping error data is greater than or equal to the mapping error threshold, optimize the initial acoustic waveform noise level mapping equation until the noise level mapping error data is less than the mapping error threshold to obtain the final acoustic waveform noise level mapping equation; otherwise, there is no need to optimize the initial acoustic waveform noise level mapping equation, and the initial acoustic waveform noise level mapping equation is used as the final acoustic waveform noise level mapping equation.

[0091] Optimizing the initial acoustic waveform noise level mapping equation described in S33 includes the following steps:

[0092] S331. Constructing a noise level mapping equation to optimize the Przewalski's horse population. , The noise level mapping equation is used to optimize the i-th horse in the wild horse population. The noise level mapping equation is used to optimize the size of the Przewalski's horse population. The maximum number of iterations for optimizing the Przewalski's horse population using the noise level mapping equation is set to [value missing]. And the current iteration number is Let be denoted as the maximum number of iterations for optimizing the noise level mapping equation and the current number of iterations for optimizing the noise level mapping equation, respectively; the search space dimension for optimizing the wild horse population using the noise level mapping equation is . ;

[0093] S332. Define the value range of each independent variable coefficient and the bias in the initial acoustic waveform noise level mapping equation to obtain the independent variable coefficient value range matrix d1 and the bias value range [d 21 ,d 22 ];d 21 d 22 These represent the lower and upper limits of the bias value in the initial acoustic waveform noise level mapping equation, respectively; d1 is as follows,

[0094] ;

[0095] in, , These represent the lower limit and upper limit of the coefficient of the i-th independent variable in the initial acoustic waveform noise level mapping equation, respectively. , The first and second parts of the initial acoustic waveform noise level mapping equation are respectively represented by the first part. The lower limit and upper limit of the values ​​of the coefficients of the independent variables corresponding to each independent variable;

[0096] Based on the independent variable coefficient value interval matrix and the bias value interval, a noise level mapping equation is set to optimize the initial position of each wild horse in the wild horse population, thus obtaining the first initial position matrix set. , The noise level mapping equation is used to optimize the initial position matrix of the k-th wild horse in the wild horse population; as follows:

[0097] ;

[0098] in, , The noise level mapping equation represents the initial position of the k-th wild horse in the wild horse population, as well as the i-th independent variable and the i-th independent variable in the initial sound waveform noise level mapping equation. The positional components of the coefficients of each independent variable on the dimension of the independent variable; The noise level mapping equation optimizes the initial position of the kth wild horse in the wild horse population in the position component of the bias dimension in the initial sound wave waveform noise level mapping equation. , , The calculation formulas are as follows:

[0099] ;

[0100] ;

[0101] ;

[0102] In the formula, rand 1k1i rand 1k2i rand 1k3 They represent respectively targeting , , Generate random numbers between 0 and 1;

[0103] S333. Based on the historical acoustic waveform image noise level mapping dataset and the historical received waveform image noise level dataset, construct a noise level mapping equation to optimize the fitness function of the Przewalski's horse population. ;as follows,

[0104] ;

[0105] S334. Begin iteration. Before iteration, set the current iteration count of the noise level mapping equation optimization to 1. During the first iteration, use the noise level mapping equation to optimize the fitness function of the wild horse population and calculate the fitness value of the initial position matrix of each wild horse in the first initial position matrix set to obtain the first fitness value set. Take the largest fitness value in the first fitness value set and the corresponding initial position matrix of the wild horse as the first global best fitness and the first global best position, respectively. Update the initial position matrix of each wild horse in the first initial position matrix set according to the first global best fitness and the first global best position. After the update is completed, increment the current iteration count of the noise level mapping equation optimization by 1 and enter the next iteration.

[0106] In each iteration, the fitness function of the Przewalski's horse population is optimized using the noise level mapping equation to calculate the fitness value of the position matrix of each Przewalski's horse in the population optimized by the noise level mapping equation obtained in the previous iteration, thus obtaining a second fitness value set. The maximum fitness value in the second fitness value set and the corresponding position matrix of the Przewalski's horse are taken as the second global best fitness and the second global best position, respectively. The position matrix of each Przewalski's horse in the population optimized by the noise level mapping equation obtained in the previous iteration is updated according to the second global best fitness and the second global best position. After the update is completed, the current iteration number of the noise level mapping equation optimization is incremented by 1 and the next iteration is started.

[0107] S335, when If the first optimal position is found, stop iterating until the first final global optimal position is obtained; otherwise, continue iterating until the optimal position is found. Up to the specified time; substitute each position component of the first final global optimal position into the initial acoustic waveform noise level mapping equation to obtain the optimized acoustic waveform noise level mapping equation; substitute the historical transducer parameter data matrix, the historical acoustic amplification factor dataset, and the historical concrete property parameter data matrix into the optimized acoustic waveform noise level mapping equation for mapping to obtain the optimized historical acoustic waveform image noise level mapping dataset; calculate the Euclidean distance between the optimized historical acoustic waveform image noise level mapping dataset and the historical received waveform image noise level dataset to obtain the optimized noise level mapping error data.

[0108] When the optimized noise level mapping error data is less than the mapping error threshold, the optimized sound wave waveform noise level mapping equation is used as the final sound wave waveform noise level mapping equation; otherwise, return to S324 to continue the iteration until the optimized noise level mapping error data is less than the mapping error threshold.

[0109] The Mustang optimization algorithm can effectively search for the global optimum of a problem, demonstrating strong optimization capabilities. It maintains high accuracy during the search process, avoiding getting trapped in local optima. It converges quickly to the optimum, improving algorithm efficiency. Furthermore, it is insensitive to the diversity of the initial population, exhibiting good robustness. Based on these advantages, this scheme employs the Mustang optimization algorithm to iteratively optimize each independent variable coefficient and bias of the initial acoustic waveform noise level mapping equation multiple times. The fitness function is the error between the mapped data and the actual data in the initial acoustic waveform noise level mapping equation. Therefore, as the iteration progresses, the mapping accuracy of the initial acoustic waveform noise level mapping equation increases, ultimately satisfying the mapping requirements.

[0110] S4. Adjust the test parameters of the acoustic transducer and the acoustic receiving amplification factor data during the current wave velocity test in accordance with the final acoustic waveform noise level mapping equation. After the adjustment is completed, conduct the wave velocity test of concrete under pressure.

[0111] S4 includes the following steps:

[0112] S41. Based on the transducer experimental parameter type set and the concrete property parameter type set, collect the experimental parameter data of the acoustic transducer, the property parameters of the concrete, and the acoustic wave receiving amplification factor data during the current wave velocity test to obtain the current transducer parameter dataset. Current concrete property parameter dataset and current sound wave receiving amplification data ; , These represent the test parameter data of the i-th type of acoustic wave test transducer and the property parameters of the concrete, respectively, during the current wave velocity test.

[0113] S42. Input the current transducer parameter dataset, the current concrete property parameter dataset, and the current acoustic wave receiving amplification factor data into the final acoustic wave waveform noise level mapping equation for mapping, and obtain the current acoustic wave waveform noise level data e.

[0114] S43. Adjust the current transducer parameter dataset and the current acoustic wave receiving amplification factor data according to the current acoustic wave waveform noise level data to obtain the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data.

[0115] S43 includes the following steps:

[0116] S431, Set the threshold for waveform image noise level;

[0117] S432. When the current acoustic waveform noise level data is greater than or equal to the waveform image noise level threshold, the current transducer parameter dataset and the current acoustic wave receiving amplification factor data are adjusted until the current acoustic waveform noise level data is less than the waveform image noise level threshold, thus obtaining the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data; otherwise, there is no need to adjust the current transducer parameter dataset and the current acoustic wave receiving amplification factor data, and the current transducer parameter dataset and the current acoustic wave receiving amplification factor data are respectively used as the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data.

[0118] The adjustment of the current transducer parameter dataset and the current acoustic wave receiving amplification factor data in S432 includes the following steps:

[0119] S4321. Constructing wave speed test parameters to optimize the Mustang population , This indicates that the wave speed test parameters are optimized for the i-th horse in the wild horse population. The wave speed test parameters are used to optimize the size of the Przewalski's horse population. The maximum number of iterations for optimizing the Przewalski's horse population using the wave speed test parameters is set to [value missing]. And the current iteration number is Let be denoted as the maximum number of iterations for wave speed test parameter optimization and the current number of iterations for wave speed test parameter optimization, respectively; the search space dimension for wave speed test parameter optimization of the Przewalski's horse population is . ;

[0120] S4322. Set the range of test parameter data for each type of acoustic transducer and the range of acoustic wave receiving amplification factor data during the current wave velocity test to obtain the set of transducer parameter value ranges. and the range of magnification values ​​[d] 41 ,d 42 ]; , These represent the lower and upper limits of the experimental parameters for the i-th type of acoustic transducer during the current wave velocity test, respectively; d 41 d 42 These represent the lower and upper limits of the acoustic wave reception amplification factor data during the current wave velocity test, respectively.

[0121] Based on the set of transducer parameter values ​​and the range of amplification factor values, the wave velocity test parameters are set to optimize the initial position of each wild horse in the wild horse population, thus obtaining a second initial position matrix. ;as follows,

[0122] ;

[0123] in, The wave speed test parameters optimize the initial position of the kth wild horse in the wild horse population, which is the position component in the i-th search space dimension. , The calculation formulas are as follows:

[0124] ;

[0125] ;

[0126] In the formula, rand 2ki , They represent respectively targeting , Generate random numbers between 0 and 1;

[0127] S4323. Based on the current acoustic waveform noise level data e, construct wave speed test parameters to optimize the fitness function of the wild horse population. ;as follows,

[0128] ;

[0129] S4324. Begin iteration. Before iteration, set the current iteration count of the wave speed test parameter optimization to 1. During the first iteration, use the fitness function of the wave speed test parameter optimization for the wild horse population to calculate the fitness value of the initial position of each wild horse in the second initial position matrix, obtaining a third fitness value set. Take the largest fitness value in the third fitness value set and the corresponding initial position of the wild horse as the third global best fitness and the third global best position, respectively. Update the initial position of each wild horse in the second initial position matrix according to the third global best fitness and the third global best position. After the update is completed, increment the current iteration count of the wave speed test parameter optimization by 1 and enter the next iteration.

[0130] In each iteration, the fitness function of the Przewalski's horse population optimized by the wave speed test parameters is used to calculate the fitness value of each horse's position in the wave speed test parameter optimized Przewalski's horse population updated in the previous iteration, resulting in a fourth fitness value set. The largest fitness value in the fourth fitness value set and the corresponding horse's position are taken as the fourth global best fitness and the fourth global best position, respectively. The position of each horse in the wave speed test parameter optimized Przewalski's horse population updated in the previous iteration is updated according to the fourth global best fitness and the fourth global best position. After the update is completed, the current iteration number of the wave speed test parameter optimization is incremented by 1 and the next iteration begins.

[0131] S4325, when If the first position is found, stop iterating to obtain the second final global optimum; otherwise, continue iterating until... Up to this point; substitute each position component of the second final global optimal position and the current concrete property parameter dataset into the final acoustic waveform noise level mapping equation for mapping, and obtain the current optimized acoustic waveform noise level data;

[0132] When the current optimized acoustic waveform noise level data is less than the waveform image noise level threshold, the corresponding position components in the second final global optimal position are respectively used as the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data; otherwise, return to S4324 to continue iterating until the current optimized acoustic waveform noise level data is less than the waveform image noise level threshold.

[0133] S44. Set the test parameters of the acoustic wave test transducer for the current wave velocity test using the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data; after setting, the set acoustic wave test transducer is obtained; use the set acoustic wave test transducer to perform wave velocity test on concrete.

[0134] (2) After the parameters of the acoustic wave test transducer are adjusted, apply coupling agent grease to the test surface of the acoustic wave test transducer, place it in the test box 2, rotate the transducer to ensure that the test surface of the transducer and the test box 2 are filled with grease; cover the test box 1, the spring is in close contact with the back of the transducer, and the transducer is held in place to ensure that the test surface of the transducer and the test box 2 are always in close contact.

[0135] (3) Apply butter to the outer surface of test box 2 and install the first test box;

[0136] (4) Repeat (2) and (3) to make the second test box. Make close contact between the first test box and the second test box with the butter-coated sides, start the wave speed test, and obtain the time t0.

[0137] (5) Separate the test box, test the dimensions of the upper and lower surfaces of the concrete test block, input the data into the control computer of the pressure testing machine, apply grease to the upper and lower surfaces of the concrete test block, place it between the test boxes, install the resistance displacement gauge, place the test box and the concrete test block as a whole in the pressure testing machine, and make contact with the upper and lower pressure plates of the testing machine.

[0138] (6) Power-on test; First, obtain the wave velocity test time t1 and the distance between the test boxes when F=0; As the test begins, t i At any given moment, the press records the pressure F. i Wave speed test time t i Distance l i And obtain t according to the following formula. i At any moment, pressure F i Lower concrete wave velocity value:

[0139] ;

[0140] In the formula: Indicates t i Under constant pressure i The concrete wave velocity value.

[0141] Example 2

[0142] This embodiment discloses a machine learning-based concrete wave velocity testing system under compressive conditions. The system can implement the method of the above embodiment and includes a wave velocity test parameter type setting module, a historical wave velocity test parameter data acquisition module, an existing wave velocity test parameter data acquisition module, a historical waveform image noise level data mapping module, a sound wave waveform noise level mapping equation construction module, a current wave velocity test parameter optimization module, and a wave velocity test module.

[0143] The wave velocity test parameter type setting module is used to set various test parameter types for the acoustic wave test transducer and various property parameters of concrete, so as to obtain a set of transducer test parameter types and a set of concrete property parameter types.

[0144] The historical wave velocity test parameter data acquisition module is used to collect multiple sets of transducer experimental parameter data, acoustic wave receiving amplification factor data, concrete property parameter data and received acoustic wave waveform image data during historical wave velocity tests according to the transducer experimental parameter type set and the concrete property parameter type set, so as to obtain historical transducer parameter data matrix, historical acoustic wave amplification factor dataset, historical concrete property parameter data matrix and historical received waveform image dataset.

[0145] The existing wave velocity test parameter data acquisition module is used to acquire multiple sets of existing acoustic waveform image data and corresponding waveform noise level data to obtain existing acoustic waveform image dataset and existing waveform noise level dataset.

[0146] The historical waveform image noise level data mapping module is used to construct a waveform image mapping model using the existing acoustic waveform image dataset and the existing waveform noise level dataset, and to map the noise level data corresponding to the historical received waveform image dataset to obtain the historical received waveform image noise level dataset.

[0147] The acoustic waveform noise level mapping equation construction module is used to construct the final acoustic waveform noise level mapping equation using historical transducer parameter data matrix, historical acoustic amplification factor dataset, historical concrete property parameter data matrix, and historical received waveform image noise level dataset.

[0148] The current wave velocity test parameter optimization module is used to adjust the test parameter data of the acoustic transducer and the acoustic wave receiving amplification factor data during the current wave velocity test in conjunction with the final acoustic waveform noise level mapping equation, so as to obtain the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data.

[0149] The wave velocity testing module is used to perform wave velocity testing of concrete under pressure conditions in conjunction with the current final transducer parameter dataset and the current final acoustic wave receiving amplification factor data.

[0150] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0151] The preferred embodiments of the invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A method for testing wave velocity of concrete under pressure based on machine learning, characterized in that: The following steps are involved: S1, collecting multiple sets of transducer experimental parameter data, sound wave receiving magnification data, concrete property parameter data and received sound wave waveform image data during historical wave velocity tests, and obtaining a historical transducer parameter data matrix, a historical sound wave magnification data set, a historical concrete property parameter data matrix and a historical received waveform image data set; collecting multiple sets of existing sound wave waveform image data and corresponding waveform noise level data, and obtaining an existing sound wave waveform image data set and an existing waveform noise level data set; S2, using the existing sound wave waveform image data set and the existing waveform noise level data set to construct a waveform image mapping model and map the noise level data corresponding to the historical received waveform image data set to obtain the historical received waveform image noise level data set; S3, using the historical transducer parameter data matrix, the historical sound wave amplification data set, the historical concrete property parameter data matrix and the historical received waveform image noise level data set to construct the final sound wave waveform noise level mapping equation; S4. Adjust the test parameter data of the acoustic wave test transducer and the acoustic wave receiving amplification factor data during the current wave velocity test in accordance with the final acoustic wave waveform noise level mapping equation. After the adjustment is completed, perform a wave velocity test of the concrete under pressure.

2. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 1 is characterized in that: The S1 comprises the following steps: S11, setting a plurality of test parameter types of the acoustic wave test transducer to obtain a transducer test parameter type set; and then setting a plurality of property parameters of concrete to obtain a concrete property parameter type set; S12, according to the transducer experimental parameter type set and the concrete property parameter type set, when the wave velocity test was performed on concrete with different property parameters using an acoustic wave test transducer with different parameters set in history, the transducer experimental parameter data of the transmitted acoustic wave, the acoustic wave receiving amplification factor data, the concrete property parameter data and the received acoustic wave waveform image data of the corresponding acoustic wave test transducer were collected to obtain a historical transducer parameter data matrix, a historical acoustic wave amplification factor data set, a historical concrete property parameter data matrix and a historical received waveform image data set; S13, collecting multiple groups of existing sound wave waveform image data and corresponding waveform noise level data to obtain an existing sound wave waveform image data set and an existing waveform noise level data set.

3. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 2 is characterized in that: The S2 comprises the following steps: S21, constructing a waveform image mapping model using the existing sound wave waveform image dataset and the existing waveform noise level dataset; S22: Input the historical received waveform image data set into a waveform image mapping model for mapping, to obtain a historical received waveform image noise level data set.

4. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 3 is characterized in that: The waveform image mapping model adopts a CNN neural network model.

5. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 4 is characterized in that: The S3 comprises the following steps: S31, constructing an initial sound wave waveform noise level mapping equation; S32, substituting the historical transducer parameter data matrix, the historical sound wave amplification data set, and the historical concrete property parameter data matrix into the initial sound wave waveform noise level mapping equation for mapping, to obtain a historical sound wave waveform image noise level mapping data set; S33, setting a mapping error threshold; calculating the Euclidean distance between the historical sound wave waveform image noise level mapping data set and the historical received waveform image noise level data set to obtain noise level mapping error data; when the noise level mapping error data is greater than or equal to the mapping error threshold, optimizing the initial sound wave waveform noise level mapping equation until the noise level mapping error data is less than the mapping error threshold, to obtain the final sound wave waveform noise level mapping equation; otherwise, there is no need to optimize the initial sound wave waveform noise level mapping equation, and the initial sound wave waveform noise level mapping equation is used as the final sound wave waveform noise level mapping equation.

6. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 5 is characterized in that: In S33, the initial sound wave waveform noise level mapping equation is optimized using the Mustang optimization algorithm.

7. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 6 is characterized in that: The S4 comprises the following steps: S41, collecting test parameter data of the acoustic wave test transducer during the current wave velocity test, property parameters of concrete, and acoustic wave receiving amplification factor data according to the transducer experimental parameter type set and the concrete property parameter type set, to obtain a current transducer parameter data set, a current concrete property parameter data set, and a current acoustic wave receiving amplification factor data; S42, inputting the current transducer parameter data set, the current concrete property parameter data set and the current sound wave receiving amplification factor data into the final sound wave waveform noise level mapping equation for mapping, to obtain the current sound wave waveform noise level data; S43, adjusting the current transducer parameter data set and the current sound wave receiving amplification factor data according to the current sound wave waveform noise level data to obtain the current final transducer parameter data set and the current final sound wave receiving amplification factor data; S44, using the current final transducer parameter data set and the current final sound wave receiving amplification factor data to set the test parameter data of the sound wave test transducer during the current wave velocity test; after the setting is completed, a set sound wave test transducer is obtained; and the set sound wave test transducer is used to perform a wave velocity test on concrete.

8. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 7 is characterized in that: The S43 comprises the following steps: S431, setting a waveform image noise level threshold; S432. When the current sound wave waveform noise level data is greater than or equal to the waveform image noise level threshold, the current transducer parameter data set and the current sound wave reception amplification data are adjusted until the current sound wave waveform noise level data is less than the waveform image noise level threshold, and the current final transducer parameter data set and the current final sound wave reception amplification data are obtained; otherwise, there is no need to adjust the current transducer parameter data set and the current sound wave reception amplification data, and the current transducer parameter data set and the current sound wave reception amplification data are used as the current final transducer parameter data set and the current final sound wave reception amplification data, respectively.

9. The method for testing wave velocity of concrete under pressure based on machine learning according to claim 8 is characterized in that: The step of adjusting the current transducer parameter data set and the current sound wave receiving amplification factor data in S432 includes the following steps: S4321, constructing a wild horse population for wave speed test parameter optimization: setting the maximum number of iterations for the wave speed test parameter optimization wild horse population to And the current iteration number is , respectively recorded as the maximum number of iterations for wave speed test parameter optimization and the current number of iterations for wave speed test parameter optimization; S4322, setting the value intervals of each type of test parameter data and the sound wave receiving amplification factor data of the sound wave test transducer when performing the current wave velocity test, and obtaining a transducer parameter value interval set and amplification factor value interval; The wave velocity test parameters are set according to the transducer parameter value interval set and the amplification factor value interval to optimize the initial position of each wild horse in the wild horse population, and obtain a second initial position matrix; S4323, constructing a wave speed test parameter to optimize the fitness function of the wild horse population according to the current sound wave waveform noise level data; S4324, start iteration; in the first round of iteration, use the wave speed test parameter to optimize the fitness function of the wild horse population to calculate the fitness value of the initial position of each wild horse in the second initial position matrix, and update the initial position of each wild horse in the second initial position matrix; in each other round of iteration, use the wave speed test parameter to optimize the fitness function of the wild horse population to calculate the fitness value of the position of each wild horse in the wild horse population optimized by the wave speed test parameter updated in the previous round of iteration, and update the position of each wild horse in the wild horse population optimized by the wave speed test parameter updated in the previous round of iteration; S4325, when When , stop the iteration and get the second final global optimal position, otherwise, continue to iterate until until the time; each position component of the second final global optimal position and the current concrete attribute parameter data set are substituted into the final sound wave waveform noise level mapping equation for mapping to obtain the current optimized sound wave waveform noise level data; when the current optimized sound wave waveform noise level data is less than the waveform image noise level threshold, each corresponding position component in the second final global optimal position is used as the current final transducer parameter data set and the current final sound wave receiving amplification data; otherwise, return to S4324 to continue iterating until the current optimized sound wave waveform noise level data is less than the waveform image noise level threshold.

10. A system for implementing the method for testing wave velocity of concrete under pressure based on machine learning as claimed in any one of claims 1 to 9, characterized in that: It includes a wave speed test parameter type setting module, a historical wave speed test parameter data acquisition module, an existing wave speed test parameter data acquisition module, a historical waveform image noise level data mapping module, an acoustic waveform noise level mapping equation construction module, a current wave speed test parameter optimization module and a wave speed test module; The wave velocity test parameter type setting module is used to set various test parameter types of the acoustic wave test transducer and various property parameters of concrete, and obtain a transducer experimental parameter type set and a concrete property parameter type set; The historical wave velocity test parameter data acquisition module is used to collect multiple groups of transducer experimental parameter data, sound wave receiving amplification data, concrete property parameter data and received sound wave waveform image data during historical wave velocity tests according to the transducer experimental parameter type set and the concrete property parameter type set, and obtain a historical transducer parameter data matrix, a historical sound wave amplification data set, a historical concrete property parameter data matrix and a historical received waveform image data set; The existing wave velocity test parameter data acquisition module is used to collect multiple groups of existing sound wave waveform image data and corresponding waveform noise level data to obtain an existing sound wave waveform image data set and an existing waveform noise level data set; The historical waveform image noise level data mapping module is used to construct a waveform image mapping model using the existing sound wave waveform image data set and the existing waveform noise level data set and map the noise level data corresponding to the historical received waveform image data set to obtain the historical received waveform image noise level data set; The acoustic waveform noise level mapping equation construction module is used to construct a final acoustic waveform noise level mapping equation using a historical transducer parameter data matrix, a historical acoustic amplification factor data set, a historical concrete property parameter data matrix, and a historical received waveform image noise level data set; The current wave velocity test parameter optimization module is used to adjust the test parameter data of the sound wave test transducer and the sound wave receiving amplification factor data during the current wave velocity test in coordination with the final sound wave waveform noise level mapping equation to obtain the current final transducer parameter data set and the current final sound wave receiving amplification factor data; The wave velocity test module is used to perform a wave velocity test of concrete under pressure in cooperation with the current final transducer parameter data set and the current final sound wave receiving amplification factor data.

Images