Method for detecting quality of concrete-filled steel tube based on ultrasonic guided wave

Abstract

The present invention relates to the field of nondestructive testing technology, and proposes a method for detecting the quality of concrete-filled steel tubes based on ultrasonic guided waves, including: collecting the guided wave amplitudes at different initial selection frequencies at the quality detection positions of the concrete-filled steel tube components, and determining the guided wave sequence and spectrum sequence of the initial selection frequencies; defining the target initial selection frequency, determining the loss degree of the target initial selection frequency, and according to the guided wave sequence and spectrum sequence, determining the main frequency of the target initial selection frequency and the adjacent spectrum amplitudes of the main frequency, and further determining the frequency response matching degree of the target initial selection frequency; according to the numerical distribution of the frequency response matching degrees of all the initial selection frequencies, determining the matching degree threshold, and screening out all the final detection frequencies according to the values of the frequency response matching degree and the matching degree threshold; obtaining the quality detection results of the quality detection positions according to all the final detection frequencies screened out according to the quality detection positions and the frequency response matching degrees of the final detection frequencies. The purpose of the present invention is to improve the accuracy of the quality detection of concrete-filled steel tubes.

Description

Technical Field

[0001] The present invention relates to the technical field of non-destructive testing, and particularly relates to a method for detecting the quality of concrete-filled steel tubes based on ultrasonic guided waves. Background Art

[0002] Concrete-filled steel tube components are widely used in various buildings such as industrial factories, equipment framework columns, and high-rise buildings. However, under long-term compression, stress accumulation will inevitably occur in concrete-filled steel tube components. When the stress accumulation is too large, the bearing capacity and quality of concrete-filled steel tube components will be affected, which may cause cracks or other damages in concrete-filled steel tube components. Therefore, it is necessary to regularly detect concrete-filled steel tube components and repair and reinforce the damaged positions. The ultrasonic guided wave technology can be used to judge the internal damage condition of concrete-filled steel tube components according to the response characteristics of ultrasonic guided waves.

[0003] The waste incineration plant is a single-story or multi-story industrial factory building with a relatively high main structure height, a relatively large span of the plant building, and is composed of various concrete-filled steel tube components with different sizes. Moreover, the connection structure between concrete-filled steel tube components is relatively complex, making the stress changes at different positions of concrete-filled steel tubes relatively complex. The detection accuracy of ultrasonic guided waves with different frequencies is affected by the stress state of concrete-filled steel tubes to different degrees. Generally, the ultrasonic guided wave technology uses ultrasonic guided waves with a fixed frequency, ignoring the complex stress change characteristics of concrete-filled steel tube components in the waste incineration plant, and the accuracy of detecting the quality of concrete-filled steel tubes is often insufficient. Summary of the Invention

[0004] The present invention provides a method for detecting the quality of concrete-filled steel tubes based on ultrasonic guided waves to solve the problem of insufficient accuracy in detecting the quality of concrete-filled steel tubes caused by different degrees of frequency response of ultrasonic guided waves to the stress differences at different positions of concrete-filled steel tube components. The specific technical solutions adopted are as follows:

[0005] An embodiment of the present invention provides a method for detecting the quality of concrete-filled steel tubes based on ultrasonic guided waves. The method includes the following steps:

[0006] Collect the guided wave amplitudes at any quality detection position of the concrete-filled steel tube component at a preset number of different primary selection frequencies, and determine the guided wave sequence and frequency spectrum sequence of the primary selection frequency according to the guided wave amplitudes at the same primary selection frequency;

[0007] Denote any one of the primary selected frequencies as the target primary selected frequency. Determine the loss degree of the target primary selected frequency based on the differences between the guided wave amplitudes included in different subsequences divided according to the spectrum sequence of the target primary selected frequency. Determine the main frequency of the target primary selected frequency and the adjacent spectrum amplitudes of the main frequency according to the guided wave sequence and the spectrum sequence. Determine the frequency response matching degree of the target primary selected frequency based on the differences between the main frequency of the target primary selected frequency and all adjacent spectrum amplitudes of the main frequency, the differences between all values in the spectrum sequence of the target primary selected frequency that are not adjacent spectrum amplitudes, and the loss degree of the target primary selected frequency;

[0008] Determine the matching degree threshold according to the numerical distribution of the frequency response matching degrees of all primary selected frequencies, and screen out all final detection frequencies according to the values of the frequency response matching degree and the matching degree threshold;

[0009] Obtain the quality detection result of the quality detection position according to all the final detection frequencies screened according to the quality detection position and the frequency response matching degrees of the final detection frequencies.

[0010] Furthermore, the specific method for determining the loss degree of the target primary selected frequency based on the differences between the guided wave amplitudes included in different subsequences divided according to the spectrum sequence of the target primary selected frequency is as follows:

[0011] Divide the spectrum sequence of the target primary selected frequency into the first preset threshold number of spectrum subsequences, and denote the root mean square of all guided wave amplitudes included in any one of the spectrum subsequences divided from the spectrum sequence of the target primary selected frequency as the dispersion degree of the any one spectrum subsequence;

[0012] Determine the loss degree of the target primary selected frequency based on the differences and numerical distribution between the dispersion degrees of all spectrum subsequences divided from the spectrum sequence of the target primary selected frequency.

[0013] Furthermore, the specific method for determining the loss degree of the target primary selected frequency based on the differences and numerical distribution between the dispersion degrees of all spectrum subsequences divided from the spectrum sequence of the target primary selected frequency is as follows:

[0014] Denote the maximum value of the dispersion degrees of all spectrum subsequences divided from the spectrum sequence of the target primary selected frequency as the maximum dispersion degree, and denote the mean value of the dispersion degrees of all spectrum subsequences divided from the spectrum sequence of the target primary selected frequency as the average dispersion degree of the target primary selected frequency;

[0015] Denote the difference between the maximum dispersion degree and the dispersion degree of the spectrum subsequence as the dispersion degree difference of the spectrum subsequence, and denote the cumulative sum of the dispersion degree differences of all spectrum subsequences divided from the spectrum sequence of the target primary selected frequency as the cumulative dispersion degree difference of the target primary selected frequency;

[0016] The ratio of the cumulative difference in the dispersion of the target primary selection frequency to the average dispersion is denoted as the loss degree of the target primary selection frequency.

[0017] Furthermore, the method for determining the main frequency of the target primary selection frequency and the spectral amplitudes of the neighboring frequencies of the main frequency is as follows:

[0018] The largest value in the spectral sequence of the target primary selection frequency is denoted as the main frequency of the target primary selection frequency;

[0019] Centering around the main frequency of the target primary selection frequency, a window with a length of the second preset threshold is established, and all the values within the window in the spectral sequence of the target primary selection frequency are denoted as the spectral amplitudes of the neighboring frequencies of the main frequency of the target primary selection frequency.

[0020] Furthermore, the method for determining the frequency response matching degree of the target primary selection frequency based on the differences between the main frequency of the target primary selection frequency and all the spectral amplitudes of the neighboring frequencies of the main frequency, the differences between all the values in the spectral sequence of the target primary selection frequency that are not the spectral amplitudes of the neighboring frequencies, and the loss degree of the target primary selection frequency includes the following specific method:

[0021] The sum of the accumulated differences between the main frequency of the target primary selection frequency and all the spectral amplitudes of the neighboring frequencies of the main frequency is denoted as the main frequency difference of the target primary selection frequency;

[0022] The standard deviation of all the values in the spectral sequence of the target primary selection frequency excluding the spectral amplitudes of the neighboring frequencies of the main frequency is denoted as the non-main frequency difference of the target primary selection frequency;

[0023] Based on the main frequency difference, non-main frequency difference, and loss degree of the target primary selection frequency, the frequency response matching degree of the target primary selection frequency is determined.

[0024] Furthermore, the method for determining the frequency response matching degree of the target primary selection frequency based on the main frequency difference, non-main frequency difference, and loss degree of the target primary selection frequency includes the following specific method:

[0025] The product of the main frequency difference and the non-main frequency difference of the target primary selection frequency is denoted as the first product of the target primary selection frequency;

[0026] The ratio of the first product of the target primary selection frequency to the loss degree is denoted as the frequency response matching degree of the target primary selection frequency.

[0027] Furthermore, the calculation formula for the matching degree threshold is:

[0028]

[0029] Wherein, represents the matching degree threshold; represents the minimum value of the frequency response matching degrees of all the primary selection frequencies; represents the range of the frequency response matching degrees of all the primary selection frequencies; represents the variance of the frequency response matching degrees of all initially selected frequencies; represents the mean value of the frequency response matching degrees of all initially selected frequencies; represents the first tuning parameter; represents the normalization function.

[0030] Furthermore, the method for screening the final detection frequencies is as follows:

[0031] All initially selected frequencies with a frequency response matching degree greater than or equal to the matching degree threshold are recorded as the final detection frequencies.

[0032] Furthermore, the specific method for obtaining the quality detection result of the quality detection position based on all the final detection frequencies screened according to the quality detection position and the frequency response matching degrees of the final detection frequencies includes:

[0033] The guided wave velocity, the first wave amplitude, the main frequency amplitude, the weighted spectral area, and the nonlinear coefficient are used as the guided wave detection parameters corresponding to the final detection frequencies;

[0034] All the final detection frequencies corresponding to the quality detection positions of 1000 concrete-filled steel tubular members with known quality detection results, the guided wave detection parameters of the final detection frequencies, the frequency response matching degrees of the final detection frequencies, and the quality detection results of the quality detection positions are used as the training set. The frequency response matching degree of the final detection frequencies is used as the guided wave eigenvalue of the decision tree of the random forest algorithm to train the random forest to obtain the quality detection model. Among them, all the final detection frequencies corresponding to the quality detection position and the guided wave detection parameters of the final detection frequencies are input into the quality detection model to obtain the decision value of the quality detection result, and the decision value of the quality detection result takes the value of -1 or 1;

[0035] All the final detection frequencies of the quality detection position and the guided wave detection parameters of the final detection frequencies are input into the quality detection model to obtain the decision values corresponding to all the final detection frequencies of the quality detection position, and 5 guided wave eigenvalues extracted by the decision tree of the random forest algorithm corresponding to each final detection frequency are obtained. The product of the sum of all the guided wave eigenvalues corresponding to the final detection frequency and the decision value of the final detection frequency is recorded as the frequency feature decision value of the final detection frequency;

[0036] The sum of the frequency feature decision values of all the final detection frequencies screened from the quality detection positions of the concrete-filled steel tubular members is recorded as the quality judgment value of the quality detection position;

[0037] Based on the quality judgment value of the quality detection position, the quality detection result of the quality detection position is obtained.

[0038] Further, obtaining the quality inspection result of the quality inspection position according to the quality judgment value of the quality inspection position includes the following specific methods:

[0039] When the quality judgment value of the quality inspection position is greater than or equal to 0, there is no quality problem at the quality inspection position;

[0040] When the quality judgment value of the quality inspection position is less than 0, there is a quality problem at the quality inspection position.

[0041] The beneficial effects of the present invention are:

[0042] This application takes into account the energy information of the guided wave exhibited by the guided wave amplitude contained in the guided wave sequence, directly determines the accuracy of the primary selected frequency corresponding to the guided wave sequence for steel tube concrete quality inspection, evaluates the strength of the energy information of the guided wave exhibited by the guided wave amplitude contained in the guided wave sequence, determines the loss degree of the target primary selected frequency. When the energy of the guided wave collected at the primary selected frequency is greater and the loss caused by passing through the quality inspection position of the steel tube concrete member is less, the loss degree of the primary selected frequency is smaller. Further, combined with the sensitivity characteristics of the frequency response of the steel tube concrete member to the guided wave of the primary selected frequency reflected by the frequency spectrum sequence of the primary selected frequency, the frequency response matching degree of the target primary selected frequency is determined. When the guided wave main frequency energy loss of the target primary selected frequency is smaller, the damage information contained in the frequency spectrum sequence obtained by selecting the target primary selected frequency is richer, the result obtained by performing steel tube concrete quality inspection is more accurate, and it is more sensitive to internal damage in the quality inspection position. At this time, the frequency response matching degree of the primary selected frequency is larger; according to the numerical distribution of the frequency response matching degrees of all primary selected frequencies, all final inspection frequencies are screened from both the time domain and the frequency domain, improving the accuracy and reliability of the frequency value for quality inspection at the current quality inspection position of the steel tube concrete member, and obtaining the quality inspection result of the quality inspection position according to all the final inspection frequencies screened by the quality inspection position and the frequency response matching degrees of the final inspection frequencies, solving the problem that the different stress differences at different positions of the steel tube concrete structure result in different frequency response degrees to ultrasonic guided waves, leading to insufficient accuracy of steel tube concrete quality inspection, and improving the accuracy of steel tube concrete quality inspection. Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0044] Figure 1 It is a schematic flow chart of a steel tube concrete quality inspection method based on ultrasonic guided waves provided by an embodiment of the present invention;

[0045] Figure 2 Flow chart for obtaining loss degree provided by an embodiment of the present invention. Detailed implementation manners

[0046] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0047] Please refer to Figure 1 , which shows a flow chart of a method for detecting the quality of concrete-filled steel tubes based on ultrasonic guided waves provided by an embodiment of the present invention. The method includes the following steps:

[0048] Step S001: Collect the guided wave amplitudes at an arbitrary quality detection position of a concrete-filled steel tube member at a preset number of different primary selection frequencies. According to the guided wave amplitudes at the same primary selection frequency, determine the guided wave sequence and the frequency spectrum sequence of the primary selection frequency.

[0049] The connection structure between concrete-filled steel tube members is relatively complex, making the stress changes at different positions of the concrete-filled steel tubes relatively complex. Therefore, a concrete-filled steel tube member includes multiple positions that need to be subjected to quality inspection, and the positions and specific quantities of the concrete-filled steel tube members that need to be subjected to quality inspection are determined by the engineer conducting the quality inspection.

[0050] For an arbitrary quality detection position of a concrete-filled steel tube member, on two opposite sides of the quality detection position of the concrete-filled steel tube member, respectively arrange two planar transducers of an ultrasonic detector, so that the propagation path of the ultrasonic guided wave of the ultrasonic detector is along the axial direction of the concrete-filled steel tube member, measure the width of the concrete-filled steel tube member between the two planar transducers, and input the width into the ultrasonic detector. The ultrasonic detector can automatically output the guided wave velocity according to the emitted ultrasonic guided wave.

[0051] The frequency change range of the guided wave emitted by the ultrasonic detector is , in the frequency change range, starting from 50 kHz, every 10 kHz, collect a primary selection frequency, and a total of 21 primary selection frequencies are collected.

[0052] Use the planar transducers of the ultrasonic detector to sequentially emit ultrasonic guided waves at each primary selection frequency, and use the planar transducers at the receiving end of the ultrasonic detector to collect the guided wave amplitudes corresponding to each primary selection frequency. Arrange all the guided wave amplitudes corresponding to the same primary selection frequency in the order of collection to obtain the guided wave sequence of the same primary selection frequency.

[0053] Among them, in this embodiment, the frequency of the ultrasonic guided wave collected by the planar transducer probe for the initial selected frequency is set to 1 MHz, and the number of guided wave amplitudes included in the guided wave sequence of the initial selected frequency is set to 2048.

[0054] The guided wave sequence of the initial selected frequency is processed using fast Fourier transform to obtain the spectral sequence of the initial selected frequency.

[0055] It can be understood that the values included in the spectral sequence of the initial selected frequency are the guided wave amplitudes.

[0056] So far, the guided wave sequence and the spectral sequence at the initial selected frequency for the quality inspection position of the concrete-filled steel tube member are obtained.

[0057] Step S002: Denote any one of the initial selected frequencies as the target initial selected frequency. According to the differences between the guided wave amplitudes included in different subsequences divided from the spectral sequence of the target initial selected frequency, determine the loss degree of the target initial selected frequency. According to the guided wave sequence and the spectral sequence, determine the main frequency of the target initial selected frequency and the adjacent spectral amplitudes of the main frequency. According to the differences between the main frequency of the target initial selected frequency and all adjacent spectral amplitudes of the main frequency, the differences between all values in the spectral sequence of the target initial selected frequency that are not adjacent spectral amplitudes, and the loss degree of the target initial selected frequency, determine the frequency response matching degree of the target initial selected frequency.

[0058] The energy information of the guided wave shown by the guided wave amplitudes included in the guided wave sequence directly determines the accuracy of the initial selected frequency corresponding to the guided wave sequence for the steel pipe quality inspection. When the energy information of the guided wave shown is weak, the energy loss of the guided wave inside the concrete-filled steel tube is more serious, the energy information of the guided wave shown by the guided wave amplitudes included in the guided wave sequence is more blurred, and the obtained results of the steel pipe quality inspection are less credible.

[0059] Denote any one of the initial selected frequencies as the target initial selected frequency, and divide the spectral sequence of the target initial selected frequency into R spectral subsequences. Different spectral subsequences correspond to the energy distribution information of the ultrasonic guided wave at different time periods.

[0060] Among them, R represents the first preset threshold. In this embodiment, the value of the first preset threshold is 16, that is, the first 128 guided wave amplitudes in the spectral sequence of the target initial selected frequency are arranged in the order of acquisition to obtain the first spectral subsequence, the 129th to 256th guided wave amplitudes in the spectral sequence of the target initial selected frequency are arranged in the order of acquisition to obtain the second spectral subsequence, the 257th to 384th guided wave amplitudes in the spectral sequence of the target initial selected frequency are arranged in the order of acquisition to obtain the third spectral subsequence, and so on. A total of 16 spectral subsequences are obtained.

[0061] Determine the loss degree of the target primary selection frequency based on the difference between the root mean squares of the guided wave amplitudes included in different spectral subsequences divided from the spectral sequence of the target primary selection frequency.

[0062] Preferably, as an embodiment of the present application, the root mean square of all the guided wave amplitudes included in any one spectral subsequence divided from the spectral sequence of the target primary selection frequency is denoted as the dispersion degree of the any one spectral subsequence, the maximum value of the dispersion degrees of all the spectral subsequences divided from the spectral sequence of the target primary selection frequency is denoted as the maximum dispersion degree, and the mean value of the dispersion degrees of all the spectral subsequences divided from the spectral sequence of the target primary selection frequency is denoted as the average dispersion degree of the target primary selection frequency; the difference between the maximum dispersion degree and the dispersion degree of the spectral subsequence is denoted as the dispersion degree difference of the spectral subsequence, and the cumulative sum of the dispersion degree differences of all the spectral subsequences divided from the spectral sequence of the target primary selection frequency is denoted as the cumulative dispersion degree difference of the target primary selection frequency; the ratio of the cumulative dispersion degree difference of the target primary selection frequency to the average dispersion degree is denoted as the loss degree of the target primary selection frequency.

[0063] The loss degree of any one primary selection frequency can be obtained by the same method.

[0064] The calculation formula for the loss degree of any one primary selection frequency is:

[0065]

[0066] where represents the loss degree of the primary selection frequency ; represents the average dispersion degree of the primary selection frequency ; represents the number of spectral subsequences divided from the spectral sequence of the primary selection frequency , and in this embodiment, the value of is 16; represents the maximum dispersion degree of the primary selection frequency ; represents the dispersion degree of the th spectral subsequence divided from the spectral sequence of the primary selection frequency .

[0067] When the ultrasonic detector emits guided waves at different initial selected frequencies, the amplitudes of the guided waves are the same. After passing through the same quality inspection position of the concrete-filled steel tube member, due to the different degrees of energy dissipation of the guided waves with different initial selected frequencies by the concrete-filled steel tube member, the guided wave sequences collected at different initial selected frequencies are different. The spectral subsequences divided from the spectral sequence of the initial selected frequency correspond to the energy distribution information of the ultrasonic guided wave at different time periods. When the average dispersion degree of the initial selected frequency is larger, the energy of the guided wave collected at the initial selected frequency is larger, and the loss caused by passing through the quality inspection position of the concrete-filled steel tube member is less. At this time, the loss degree of the initial selected frequency is smaller. At the same time, when the difference in the dispersion degree of the spectral subsequences divided from the spectral sequence of the initial selected frequency is smaller, the loss caused by passing through the quality inspection position of the concrete-filled steel tube member is also less. At this time, the loss degree of the initial selected frequency is smaller.

[0068] The flowchart for obtaining the loss degree is as Figure 2 shown.

[0069] The waste incineration plant is a single-story or multi-story industrial plant with a relatively high main structure height, a relatively large plant span, and is composed of a variety of concrete-filled steel tube members with different sizes inside. Moreover, the connection structure between the concrete-filled steel tube members is relatively complex, making the stress changes at different positions of the concrete-filled steel tube more complex. The guided waves emitted by the ultrasonic detector at different initial selected frequencies will also be affected by the stress state of the concrete-filled steel tube. When the initial selected frequencies are different, the dispersion degrees of the guided waves with different initial selected frequencies by the concrete-filled steel tube member are different, and the sensitivity response degrees of different initial selected frequencies to different damages are different. The spectral sequence of the initial selected frequency reflects the frequency response of the concrete-filled steel tube member to the guided wave of the initial selected frequency.

[0070] The largest value in the spectral sequence of the target initial selected frequency is denoted as the main frequency of the target initial selected frequency. Centered on the main frequency of the target initial selected frequency, a window with a length of is established. All the values in the window of the spectral sequence of the target initial selected frequency are denoted as the adjacent spectral amplitudes of the main frequency of the target initial selected frequency. Among them, represents the second preset threshold, and the value of the second preset threshold in this embodiment is 21.

[0071] When the number of values included in the window established centered on the main frequency of the target initial selected frequency is less than the second preset threshold, only the values included in the window established centered on the main frequency of the target initial selected frequency are used for the following calculations. The number of values included in the window established centered on the main frequency of the target initial selected frequency is denoted as , then the value of is less than or equal to the second preset threshold.

[0072] Determine the frequency response matching degree of the target primary selection frequency based on the differences between the main frequency of the target primary selection frequency and the amplitudes of all adjacent frequency spectra, the differences between all values in the frequency spectrum sequence of the target primary selection frequency that are not adjacent frequency spectrum amplitudes, and the loss degree of the target primary selection frequency.

[0073] Preferably, as an embodiment of the present application, denote the cumulative sum of the differences between the main frequency of the target primary selection frequency and the amplitudes of all adjacent frequency spectra as the main frequency difference of the target primary selection frequency; denote the standard deviation of all values in the frequency spectrum sequence of the target primary selection frequency excluding the amplitudes of adjacent frequency spectra of the main frequency as the non-main frequency difference of the target primary selection frequency; denote the product of the main frequency difference and the non-main frequency difference of the target primary selection frequency as the first product of the target primary selection frequency, and denote the ratio of the first product of the target primary selection frequency to the loss degree as the frequency response matching degree of the target primary selection frequency.

[0074] The frequency response matching degree of any primary selection frequency can be obtained in the same way.

[0075] The calculation formula for the frequency response matching degree of any primary selection frequency is:

[0076]

[0077] Wherein, represents the frequency response matching degree of the primary selection frequency ; represents the non-main frequency difference of the primary selection frequency ; represents the loss degree of the primary selection frequency ; represents the number of values included in the window established with the main frequency of the primary selection frequency as the center; represents the main frequency of the primary selection frequency ; represents the th adjacent frequency spectrum amplitude of the main frequency of the primary selection frequency ;

[0078] When the natural frequency of the damaged structure or the natural frequency of the quality inspection position itself in the quality inspection position of the concrete-filled steel tube member is closer to the frequency of the guided wave, the main frequency energy loss of the guided wave emitted by the ultrasonic detector is smaller. At the same time, the energy loss difference of the remaining frequency clutter is larger, the damage information contained in the spectrum sequence obtained by the guided wave is richer, and the result obtained by selecting the initial frequency corresponding to the guided wave for the quality inspection of the concrete-filled steel tube is more accurate. Therefore, when the difference between the main frequency of the target initial frequency and the amplitudes of all adjacent spectra of the main frequency is larger, the difference between all values in the spectrum sequence of the target initial frequency that are not adjacent spectrum amplitudes is larger, and the loss degree of the initial frequency is smaller, the main frequency energy loss of the guided wave by selecting the target initial frequency is smaller, the damage information contained in the spectrum sequence obtained by selecting the target initial frequency is richer, the result obtained by performing the quality inspection of the concrete-filled steel tube is more accurate, and it is more sensitive to the internal damage in the quality inspection position. At this time, the frequency response matching degree of the initial frequency is larger.

[0079] Thus, the frequency response matching degrees of all the initial frequencies are obtained.

[0080] Step S003: Determine the matching degree threshold according to the numerical distribution of the frequency response matching degrees of all the initial frequencies, and screen out all the final detection frequencies according to the values of the frequency response matching degree and the matching degree threshold.

[0081] Determine the matching degree threshold according to the numerical distribution of the frequency response matching degrees of all the initial frequencies. The calculation formula of the matching degree threshold is:

[0082]

[0083] where represents the matching degree threshold; represents the minimum value of the frequency response matching degrees of all the initial frequencies; represents the range of the frequency response matching degrees of all the initial frequencies; represents the variance of the frequency response matching degrees of all the initial frequencies; represents the mean value of the frequency response matching degrees of all the initial frequencies; represents the first tuning parameter, whose function is to prevent the denominator from being 0. In this embodiment, the value of the first tuning parameter is 0.01; represents the normalization function, whose function is to take the normalized value of the value in the brackets.

[0084] It should be noted that in this embodiment, the Z-Score standard normalization method is used to calculate the normalized value. In the actual application process, the implementer can use other methods of existing technologies such as the maximum-minimum normalization method, sigmoid function, etc. to calculate the normalized value, which is not limited herein.

[0085] When the range and variance of the frequency response matching degrees of all the primary selection frequencies are larger, the differences in the sensitivity of the damage detection of concrete-filled steel tubes for different primary selection frequency values are greater. A larger matching degree threshold should be set to screen out the primary selection frequencies with poor detection performance and improve the accuracy of the quality detection of concrete-filled steel tubes. When the mean value of the frequency response matching degrees of all the primary selection frequencies is larger, the overall sensitivity of different primary selection frequencies to the damage detection of concrete-filled steel tubes is greater. A smaller matching degree threshold should be set to select more guided waves of different frequencies to monitor the quality of concrete-filled steel tubes and enhance the reliability of the quality detection.

[0086] All the primary selection frequencies with the frequency response matching degree greater than or equal to the matching degree threshold are recorded as the final detection frequencies. The final detection frequencies can be used as the frequency values for quality detection at the current quality detection positions of the concrete-filled steel tube members.

[0087] In summary, all the final detection frequencies are determined.

[0088] Step S004: Obtain the quality detection results of the quality detection positions according to all the final detection frequencies screened based on the quality detection positions and the frequency response matching degrees of the final detection frequencies.

[0089] The wave velocity of the guided wave, the amplitude of the first wave, the amplitude of the main frequency, the weighted spectral area, and the nonlinear coefficient are five commonly used guided wave detection parameters. These guided wave detection parameters are all well-known parameters. Specifically, the ultrasonic detector can automatically output the wave velocity of the ultrasonic guided wave according to the emitted ultrasonic guided wave. The amplitude of the first wave is the maximum value of the guided wave amplitudes in the first spectral subsequence divided by the guided wave sequence. The amplitude of the main frequency is the spectral amplitude of the main frequency of the spectral sequence. The weighted spectral area is the sum of all the spectral amplitudes in the spectral sequence. The nonlinear coefficient is the ratio of the spectral amplitude of twice the main frequency to the square of the spectral amplitude of the main frequency.

[0090] Each final detection frequency corresponds to five guided wave detection parameters. The quality detection positions of each concrete-filled steel tube member may correspond to multiple final detection frequencies or may also correspond to one final detection frequency.

[0091] Taking all the final detection frequencies corresponding to the quality detection positions of 1000 concrete-filled steel tubular members with known quality detection results, the guided wave detection parameters of the final detection frequencies, the frequency response matching degrees of the final detection frequencies, and the quality detection results of the quality detection positions as the training set, and taking the frequency response matching degree of the final detection frequency as the guided wave eigenvalue of the decision tree of the random forest algorithm, training the random forest to obtain a quality detection model. Inputting all the final detection frequencies corresponding to the quality detection position and the guided wave detection parameters of the final detection frequency into the quality detection model, the decision value of the quality detection result can be obtained. The decision value of the quality detection result takes the value of -1 or 1. Among them, when the decision value of the quality detection result takes the value of -1, it means that there are quality problems at the quality detection position, and when the decision value of the quality detection result takes the value of 1, it means that there are no quality problems at the quality detection position.

[0092] Among them, the number of decision trees of the random forest algorithm is 100, and the number of guided wave features randomly selected for each decision tree is 5. Using the training set to train the random forest to obtain the quality detection model is a well-known technology and will not be elaborated here.

[0093] Inputting all the final detection frequencies corresponding to the quality detection position and the guided wave detection parameters of the final detection frequency into the quality detection model, obtaining the decision values corresponding to all the final detection frequencies of the quality detection position. At the same time, obtaining the 5 guided wave eigenvalues extracted by the decision tree of the random forest algorithm corresponding to each final detection frequency, and recording the product of the sum of all the guided wave eigenvalues corresponding to the final detection frequency and the decision value of the final detection frequency as the frequency feature decision value of the final detection frequency.

[0094] Recording the sum of the frequency feature decision values of all the final detection frequencies selected for the quality detection position of the concrete-filled steel tubular member as the quality judgment value of the quality detection position.

[0095] When the quality judgment value of the quality detection position is greater than or equal to 0, there are no quality problems at the quality detection position; when the quality judgment value of the quality detection position is less than 0, there are quality problems at the quality detection position, and an alarm signal is sent to the buzzer for alarm.

[0096] So far, the quality detection result of the concrete-filled steel tubular based on ultrasonic guided wave is obtained.

[0097] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A quality detection method for steel tube concrete based on ultrasonic guided waves, characterized in that: The method comprises the following steps: Collect the guided wave amplitudes at a preset number of different primary selected frequencies at any quality inspection position of the steel tube concrete component, and determine the guided wave sequence and spectrum sequence of the primary selected frequency according to the guided wave amplitudes at the same primary selected frequency; Any primary frequency is recorded as the target primary frequency, and the loss degree of the target primary frequency is determined according to the difference between the guided wave amplitudes contained in different subsequences divided from the spectrum sequence of the target primary frequency. The main frequency of the target primary frequency and the adjacent spectrum amplitude of the main frequency are determined according to the guided wave sequence and the spectrum sequence; the cumulative sum of the differences between the main frequency of the target primary frequency and all the adjacent spectrum amplitudes of the main frequency is recorded as the main frequency difference of the target primary frequency; the standard deviation of all values ​​of the adjacent spectrum amplitudes of the main frequency in the spectrum sequence of the target primary frequency except the main frequency is recorded as the non-main frequency difference of the target primary frequency; the product of the main frequency difference and the non-main frequency difference of the target primary frequency is recorded as the first product of the target primary frequency; the ratio of the first product of the target primary frequency to the loss degree is recorded as the frequency response matching degree of the target primary frequency; According to the numerical distribution of the frequency response matching degrees of all the preliminarily selected frequencies, a matching degree threshold is determined, and according to the values ​​of the frequency response matching degrees and the matching degree threshold, all the final detection frequencies are screened out; The quality detection result of the quality detection position is obtained according to the frequency response matching degree of all final detection frequencies screened out at the quality detection position and the final detection frequency.

2. The method for quality detection of steel tube concrete based on ultrasonic guided waves according to claim 1 is characterized in that: The loss degree of the target primary selected frequency is determined based on the difference between the guided wave amplitudes contained in different subsequences divided from the spectrum sequence of the target primary selected frequency, and the specific method includes: The spectrum sequence of the target preliminary frequency is equally divided into a first preset threshold number of spectrum subsequences, and the root mean square of all guided wave amplitudes contained in any spectrum subsequence divided from the spectrum sequence of the target preliminary frequency is recorded as the discreteness of any spectrum subsequence; The loss degree of the target primary frequency is determined according to the difference and value distribution between the discrete degrees of all spectrum subsequences divided from the spectrum sequence of the target primary frequency.

3. The method for quality detection of steel tube concrete based on ultrasonic guided waves according to claim 2 is characterized in that: The method of determining the loss degree of the target primary frequency based on the difference and value distribution between the discrete degrees of all spectrum subsequences divided from the spectrum sequence of the target primary frequency includes the following specific methods: The maximum value of the discreteness of all the spectrum subsequences divided from the spectrum sequence of the target primary frequency is recorded as the maximum discreteness, and the mean value of the discreteness of all the spectrum subsequences divided from the spectrum sequence of the target primary frequency is recorded as the average discreteness of the target primary frequency; The difference between the maximum discreteness and the discreteness of the spectrum subsequence is recorded as the discreteness difference of the spectrum subsequence, and the cumulative sum of the discreteness differences of all spectrum subsequences divided from the spectrum sequence of the target primary frequency is recorded as the cumulative discreteness difference of the target primary frequency; The ratio of the cumulative difference in the dispersion of the target primary frequency to the average dispersion is recorded as the loss degree of the target primary frequency.

4. The method for quality detection of steel tube concrete based on ultrasonic guided waves according to claim 1 is characterized in that: The method for determining the amplitude of the main frequency and the adjacent spectrum of the main frequency of the target primary frequency is as follows: The largest value in the frequency spectrum sequence of the target primary selected frequency is recorded as the main frequency of the target primary selected frequency; A window with a length of the second preset threshold value is established with the main frequency of the target preliminary frequency as the center, and all values ​​in the spectrum sequence of the target preliminary frequency within the window are recorded as the adjacent spectrum amplitudes of the main frequency of the target preliminary frequency.

5. The method for quality inspection of concrete-filled steel tubes based on ultrasonic guided waves according to claim 1, characterized in that: The calculation formula of the matching degree threshold is: in, represents the matching threshold; It represents the minimum value of the frequency response matching of all the primary selected frequencies; Indicates the extreme difference in frequency response matching of all the primary frequencies; represents the variance of the frequency response matching of all the primary frequencies; represents the mean of the frequency response matching of all the primary selected frequencies; represents the first tuning parameter coefficient; Represents the normalization function.

6. The method for quality inspection of steel tube concrete based on ultrasonic guided waves according to claim 1, characterized in that: The screening method of the final detection frequency is: All the preliminarily selected frequencies whose frequency response matching degree is greater than or equal to the matching degree threshold are recorded as the final detection frequencies.

7. The method for quality inspection of steel tube concrete based on ultrasonic guided waves according to claim 1, characterized in that: The specific method of obtaining the quality detection result of the quality detection position according to all the final detection frequencies screened out at the quality detection position and the frequency response matching degree of the final detection frequency is as follows: The guided wave velocity, the first wave amplitude, the main frequency amplitude, the weighted spectrum area and the nonlinear coefficient are used as the guided wave detection parameters corresponding to the final detection frequency; All final detection frequencies corresponding to the quality detection positions of 1000 steel tube concrete components with known quality detection results, waveguide detection parameters of the final detection frequencies, frequency response matching of the final detection frequencies, and quality detection results of the quality detection positions are used as training sets, and the frequency response matching of the final detection frequencies is used as the waveguide feature value of the decision tree of the random forest algorithm. The random forest is trained to obtain a quality detection model, wherein all final detection frequencies corresponding to the quality detection positions and the waveguide detection parameters of the final detection frequencies are input into the quality detection model to obtain a decision value of the quality detection result, and the decision value of the quality detection result is -1 or 1; Input all final detection frequencies of the quality detection position and the waveguide detection parameters of the final detection frequencies into the quality detection model, obtain the decision values ​​corresponding to all the final detection frequencies of the quality detection position, and obtain 5 waveguide eigenvalues ​​extracted by the decision tree of the random forest algorithm corresponding to each final detection frequency, and record the product of the cumulative sum of all the waveguide eigenvalues ​​corresponding to the final detection frequency and the decision value of the final detection frequency as the frequency characteristic decision value of the final detection frequency; The cumulative sum of the frequency characteristic decision values ​​of all final detection frequencies screened out at the quality detection position of the steel tube concrete component is recorded as the quality judgment value of the quality detection position; According to the quality judgment value of the quality detection position, the quality detection result of the quality detection position is obtained.

8. The method for quality inspection of concrete-filled steel tubes based on ultrasonic guided waves according to claim 7, characterized in that: The specific method of obtaining the quality detection result of the quality detection position according to the quality judgment value of the quality detection position includes: When the quality judgment value of the quality detection position is greater than or equal to 0, there is no quality problem at the quality detection position; When the quality judgment value of the quality detection position is less than 0, there is a quality problem at the quality detection position.

Images