Method for detecting leakage of barrier of refuse landfill

Abstract

The invention discloses a leakage detection method for a barrier barrier of a refuse landfill, and belongs to the technical field of leakage detection. According to the invention, a sound wave source and equidistant linear arrangement detection points are arranged in a refuse landfill, low-frequency, medium-frequency and high-frequency sound waves are emitted, and corresponding forced vibration signals are collected; calculating an energy value, an energy stray value and a frequency stray value of each signal, and constructing a signal energy ratio, an energy stray ratio and a frequency stray ratio enhancement vector of adjacent detection points under the same emission frequency; calculating change values of a signal energy ratio, an energy stray ratio and a frequency stray ratio according to the enhancement vectors under the three emission frequencies; a barrier leakage detection model is adopted to process an enhancement vector, a leakage degree value is obtained based on change value feature adjustment, and the problem of low leakage detection precision in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of leakage detection, in particular to a method for detecting leakage of an isolation barrier in a landfill. Background Art

[0002] Landfills are important infrastructure for the treatment of urban solid waste, and the integrity of their barriers is directly related to environmental safety and ecological balance. At present, traditional barrier leakage detection methods mainly rely on manual surveys, drilling sampling or simple permeability tests. However, for landfills with deep burial depths and large areas, these traditional detection methods are often unable to accurately locate and quantify the specific location and extent of leakage. Therefore, existing technologies have gradually adopted ultrasonic waves to detect leakage points in barrier barriers. Ultrasonic detection utilizes the propagation characteristics of sound waves in different media. By emitting multi-frequency sound waves and analyzing their reflection characteristics in the barrier, it can effectively identify potential leakage points. However, existing ultrasonic detection methods have difficulties in extracting the signal characteristics of leakage points from echoes, and it is difficult to accurately obtain the signal characteristics belonging to the leakage points, resulting in low leakage detection accuracy. Summary of the Invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for detecting leakage of a barrier barrier in a landfill, which solves the problem of low leakage detection accuracy in the prior art.

[0004] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: a method for detecting leakage of a barrier barrier in a landfill, comprising the following steps:

[0005] S1. Setting up an acoustic wave source at the landfill and setting up multiple detection points at equal intervals along the acoustic wave propagation path, wherein the multiple detection points form a straight line;

[0006] S2. Emit low-frequency, medium-frequency, and high-frequency sound waves at the sound wave source, respectively, and collect forced vibration signals of corresponding frequencies at each detection point;

[0007] S3, calculating the signal energy value, energy spurious value and frequency spurious value of each forced vibration signal;

[0008] S4. Under the same transmission frequency, construct a signal energy ratio enhancement vector, an energy spurious ratio enhancement vector, and a frequency spurious ratio enhancement vector according to the signal energy values, energy spurious values, and frequency spurious values ​​of adjacent detection points;

[0009] S5. Calculate the signal energy ratio change value, the energy spurious ratio change value, and the frequency spurious ratio change value according to the signal energy ratio enhancement vector, the energy spurious ratio enhancement vector, and the frequency spurious ratio enhancement vector at the three transmission frequencies;

[0010] S6. Use the barrier leakage detection model to process the signal energy ratio enhancement vector, energy spurious ratio enhancement vector and frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency, and adjust the characteristics of the signal energy ratio change value, energy spurious ratio change value and frequency spurious ratio change value to obtain the barrier leakage degree value.

[0011] Furthermore, the formula for calculating the signal energy value of each forced vibration signal in S3 is: Where E is the signal energy value of the forced vibration signal, x n is the nth signal value in the forced vibration signal, n is a positive integer, and N is the length of the forced vibration signal.

[0012] Furthermore, the process of calculating the energy spurious value and the frequency spurious value of each forced vibration signal in S3 includes the following steps:

[0013] A1. Perform spectrum analysis on each forced vibration signal to obtain the frequency value and amplitude;

[0014] A2. Extract each amplitude and calculate the energy spurious value;

[0015] A3. Extract each frequency value and calculate the frequency spurious value.

[0016] Furthermore, S4 includes the following sub-steps:

[0017] S41. Calculate a signal energy ratio based on signal energy values ​​of adjacent detection points at the same transmission frequency.

[0018] S42, arranging the signal energy ratios in sequence to construct a signal energy ratio vector, and performing enhancement processing on the signal energy ratio vector to obtain a signal energy ratio enhancement vector;

[0019] S43. Calculate the energy spurious ratio based on the energy spurious values ​​of adjacent detection points at the same transmission frequency;

[0020] S44, arranging the energy-stray ratio values ​​in sequence to construct an energy-stray ratio value vector, and performing enhancement processing on the energy-stray ratio vector to obtain an energy-stray ratio enhancement vector;

[0021] S45. Calculate the frequency spurious ratio based on the frequency spurious values ​​of adjacent detection points at the same transmission frequency.

[0022] S46. Arrange the frequency spurious ratios in sequence to construct a frequency spurious ratio vector, and perform enhancement processing on the frequency spurious ratio vector to obtain a frequency spurious ratio enhancement vector.

[0023] Furthermore, S5 includes the following sub-steps:

[0024] S51, calculating a signal energy ratio change value according to the signal energy ratio enhancement vector at low frequency, medium frequency, and high frequency;

[0025] S52, calculating the energy-spurious ratio change value according to the energy-spurious ratio enhancement vector at low frequency, medium frequency, and high frequency;

[0026] S53. Calculate a frequency spurious-to-interference ratio change value according to the frequency spurious-to-interference ratio enhancement vectors at low frequency, medium frequency, and high frequency.

[0027] Furthermore, the formula for calculating the signal energy ratio change value in S51 is: Among them, s E is the signal energy ratio change value, r H,E is the sum of the elements in the signal energy ratio enhancement vector at high frequency, r M,E is the sum of the elements in the signal energy ratio enhancement vector at the intermediate frequency, r L,E is the sum of the elements in the signal energy ratio enhancement vector at low frequency;

[0028] The formula for calculating the energy spurious ratio change value in S52 is: Among them, s A is the change value of energy spurious ratio, r H,A is the sum of the elements in the energy spurious ratio enhancement vector at high frequencies, r M,A is the sum of the elements in the energy spurious ratio enhancement vector at the intermediate frequency, r L,A is the sum of the elements in the energy spurious ratio enhancement vector at low frequency;

[0029] The formula for calculating the frequency spurious ratio change value in S53 is: Among them, s f is the frequency spurious ratio change value, r H,f is the sum of the elements in the frequency spurious ratio enhancement vector at high frequencies, r M,f is the sum of the elements in the frequency spurious ratio enhancement vector at the intermediate frequency, r L,f It is the sum of the elements in the frequency spurious ratio enhancement vector at low frequency.

[0030] Furthermore, the barrier leakage detection model in S6 includes: a first feature fusion unit, a second feature fusion unit, a third feature fusion unit, a first LSTM unit, a second LSTM unit, a third LSTM unit, a first tanh unit, a second tanh unit, a third tanh unit, a first feature adjustment unit, a second feature adjustment unit, a third feature adjustment unit and an output layer;

[0031] The input end of the first feature fusion unit is used to input the signal energy ratio enhancement vector at low frequency, medium frequency and high frequency; the input end of the second feature fusion unit is used to input the energy spurious ratio enhancement vector at low frequency, medium frequency and high frequency; the input end of the third feature fusion unit is used to input the frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency;

[0032] The first feature fusion unit, the first LSTM unit, the first tanh unit, and the first feature adjustment unit are connected in sequence;

[0033] The second feature fusion unit, the second LSTM unit, the second tanh unit and the second feature adjustment unit are connected in sequence;

[0034] The third feature fusion unit, the third LSTM unit, the third tanh unit, and the third feature adjustment unit are connected in sequence;

[0035] The output layer is connected to the first feature adjustment unit, the second feature adjustment unit and the third feature adjustment unit respectively.

[0036] Furthermore, the expression of the first feature fusion unit is: Among them, Y E is the signal energy ratio fusion feature, X H,E is the signal energy ratio enhancement vector at high frequency, X M,E is the signal energy ratio enhancement vector at the intermediate frequency, X L,E is the signal energy ratio enhancement vector at low frequency, For element-wise multiplication, F conv It is a point-by-point convolution operation;

[0037] The expression of the second feature fusion unit is: Among them, Y A is the energy-spurious ratio fusion feature, X H,A is the energy spurious ratio enhancement vector at high frequency, X M,A is the energy spurious ratio enhancement vector at the intermediate frequency, X L,A is the energy spurious ratio enhancement vector at low frequency;

[0038] The expression of the third feature fusion unit is: Among them, Y f is the frequency spurious ratio fusion feature, X H,f is the frequency spurious ratio enhancement vector at high frequency, X M,f is the frequency spurious ratio enhancement vector at the intermediate frequency, X L,f is the frequency spurious ratio enhancement vector at low frequency.

[0039] Furthermore, the expressions of the first feature adjustment unit, the second feature adjustment unit and the third feature adjustment unit are all: y = h1 [1 + Sigmoid (wh2 + b)], where y is the output of the feature adjustment unit, h1 is the input of the first input terminal of the feature adjustment unit, h2 is the input of the second input terminal of the feature adjustment unit, w is the weight of h2, b is the bias of h2, and Sigmoid is the S-type activation function.

[0040] The beneficial effects of the present invention are:

[0041] 1. The present invention places each detection point equidistantly, forming a straight line, to detect the propagation of sound waves in that direction. By constructing signal energy ratio enhancement vectors, energy spurious ratio enhancement vectors, and frequency spurious ratio enhancement vectors, the present invention can effectively amplify the signal characteristics of the leakage point and reduce background noise interference, thereby significantly improving the accuracy of leakage point identification.

[0042] 2. The leakage point of the blocking barrier has different reflection and attenuation characteristics for different frequencies. Therefore, the present invention calculates the signal energy ratio change value, energy spurious ratio change value and frequency spurious ratio change value to reflect the changes in the signal energy ratio value, energy spurious ratio value and frequency spurious ratio value at different frequencies, and further highlights the signal characteristics of the leakage point.

[0043] 3. The present invention adopts a barrier leakage detection model to process the signal energy ratio enhancement vector, energy spurious ratio enhancement vector and frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency, and performs feature adjustment based on the signal energy ratio change value, energy spurious ratio change value and frequency spurious ratio change value to improve the barrier leakage detection accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flow chart of a method for detecting leakage of a landfill barrier;

[0045] Figure 2 A schematic diagram of the detection point layout;

[0046] Figure 3 Schematic diagram of the structure of the barrier leakage detection model. DETAILED DESCRIPTION

[0047] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0048] like Figure 1 As shown, a method for detecting leakage of a barrier barrier in a landfill comprises the following steps:

[0049] S1. Arrange detection points: Set up a sound wave source on the landfill and set up multiple detection points at equal distances along the sound wave propagation path. The multiple detection points form a straight line, such as Figure 2 As shown;

[0050] S2. Collecting forced vibration signals: emitting low-frequency, medium-frequency, and high-frequency sound waves at the sound wave source, respectively, and collecting forced vibration signals of corresponding frequencies at each detection point;

[0051] S3. Extracting forced vibration signal features: calculating the signal energy value, energy spurious value, and frequency spurious value of each forced vibration signal;

[0052] S4. Extracting characteristic ratios of adjacent detection points: Under the same transmission frequency, constructing a signal energy ratio enhancement vector, an energy spurious ratio enhancement vector, and a frequency spurious ratio enhancement vector based on the signal energy values, energy spurious values, and frequency spurious values ​​of adjacent detection points;

[0053] S5. Extracting the change value of the characteristic ratio: Calculating the signal energy ratio change value, the energy spurious ratio change value, and the frequency spurious ratio change value according to the signal energy ratio enhancement vector, the energy spurious ratio enhancement vector, and the frequency spurious ratio enhancement vector at the three transmission frequencies;

[0054] S6. Predict the degree of leakage of the barrier: Use the barrier leakage detection model to process the signal energy ratio enhancement vector, energy spurious ratio enhancement vector and frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency, and adjust the characteristics of the signal energy ratio change value, energy spurious ratio change value and frequency spurious ratio change value to obtain the degree of leakage of the barrier.

[0055] In this embodiment, the frequency of the low-frequency sound wave is set to 5 kHz; the frequency of the medium-frequency sound wave is set to 50 kHz; and the frequency of the high-frequency sound wave is set to 500 kHz.

[0056] In this embodiment, the formula for calculating the signal energy value of each forced vibration signal in S3 is: Where E is the signal energy value of the forced vibration signal, x n is the nth signal value in the forced vibration signal, n is a positive integer, and N is the length of the forced vibration signal.

[0057] The present invention emits three sound waves of low frequency, medium frequency and high frequency at the sound wave source respectively, that is, three forced vibration signals will be detected at each detection point: the forced vibration signal corresponding to the low-frequency sound wave, the forced vibration signal corresponding to the medium-frequency sound wave and the forced vibration signal corresponding to the high-frequency sound wave.

[0058] In this embodiment, the process of calculating the energy spurious value and the frequency spurious value of each forced vibration signal in S3 includes the following steps:

[0059] A1. Perform spectrum analysis on each forced vibration signal to obtain the frequency value and amplitude;

[0060] A2. Extract each amplitude and calculate the energy spurious value: Among them, ε A is the energy spurious value, A main is the maximum value, A i Divide the maximum amplitude A by main The i-th amplitude except the maximum amplitude A main The number of amplitudes outside, i is a positive integer;

[0061] A3. Extract each frequency value and calculate the frequency spurious value: Among them, ε f is the frequency spurious value, f main is the frequency value corresponding to the maximum amplitude, f i is the i-th frequency value except the frequency value corresponding to the maximum amplitude, and L is the number of frequency values ​​except the frequency value corresponding to the maximum amplitude.

[0062] The present invention uses spectrum analysis to obtain frequency and amplitude values, based on which it calculates energy spurious values ​​and frequency spurious values. Energy spurious values ​​measure the energy contribution of amplitude components other than the main amplitude in the signal, while frequency spurious values ​​reflect the distribution of frequency components other than the main frequency. When a landfill barrier leaks, the propagation characteristics of the sound wave change, and the amplitude and frequency distribution of the forced vibration signal will also change accordingly. These spurious values ​​can accurately capture the changes in signal characteristics caused by the leak, providing a reliable basis for subsequent leak detection.

[0063] When there is a leakage point, the medium properties (such as density, elastic modulus, etc.) at the leakage point are different from those of the surrounding medium. When the sound wave propagates to the leakage point, it will interact with the fluid (such as leachate) and loose matter at the leakage point, causing part of the energy to be scattered, absorbed or reflected, so that in addition to the original main frequency components, some new frequency components will appear in the received sound wave signal. These new frequency components are usually caused by the nonlinear effect generated by the interaction between the sound wave and the complex medium at the leakage point. In the present invention, when emitting a single-frequency sound wave, in addition to the maximum amplitude and main frequency, the more amplitude and frequency values ​​there are, the more significant the leakage characteristics are.

[0064] In this embodiment, S4 includes the following sub-steps:

[0065] S41. Under the same transmission frequency, calculate the signal energy ratio based on the signal energy values ​​of adjacent detection points: Among them, μ E ,m is the energy ratio of the mth signal, E m is the signal energy value of the mth detection point, E m+1 is the signal energy value of the m+1th detection point, where m is a positive integer;

[0066] S42, arranging the signal energy ratios in sequence to construct a signal energy ratio vector, and performing enhancement processing on the signal energy ratio vector to obtain a signal energy ratio enhancement vector;

[0067] S43. Under the same transmission frequency, calculate the energy spurious ratio based on the energy spurious values ​​of adjacent detection points: Among them, μ A,m is the mth energy spurious ratio, ε A,m is the energy spurious value of the mth detection point, ε A,m+1 is the energy spurious value of the m+1th detection point;

[0068] S44, arranging the energy-stray ratio values ​​in sequence to construct an energy-stray ratio value vector, and performing enhancement processing on the energy-stray ratio vector to obtain an energy-stray ratio enhancement vector;

[0069] S45. Under the same transmission frequency, calculate the frequency spurious ratio based on the frequency spurious values ​​of adjacent detection points: Among them, μ f ,m is the mth frequency spurious ratio, ε f,m is the frequency spurious value of the mth detection point, ε f,m+1 is the frequency spurious value of the m+1th detection point;

[0070] S46. Arrange the frequency spurious ratios in sequence to construct a frequency spurious ratio vector, and perform enhancement processing on the frequency spurious ratio vector to obtain a frequency spurious ratio enhancement vector.

[0071] The present invention uses the signal energy ratio, energy spurious ratio, and frequency spurious ratio to clearly reflect the relative changes in signal energy, energy spurious, and frequency spurious that may be caused by leakage. If leakage occurs between a pair of detection points, the ratio between that pair of detection points will be significantly different from the ratios between other pairs of detection points, thus highlighting the leakage characteristics.

[0072] When leakage occurs, signal energy decays rapidly. Therefore, the ratio of the signal energy value at the mth detection point to the signal energy value at the m+1th detection point is used. Leakage also generates more frequency and amplitude components. Therefore, the ratio of the energy spurious value at the m+1th detection point to the energy spurious value at the mth detection point, and the ratio of the frequency spurious value at the m+1th detection point to the frequency spurious value at the mth detection point, is used.

[0073] In this embodiment, the formula for the enhancement process in S42, S44 and S46 is: m =s m · m , where s′ m is the mth enhancement value, s m is the mth element in the vector.

[0074] The present invention multiplies the elements in a vector by themselves, so as to enhance the significant features in the vector to be more significant and weaken the non-significant features.

[0075] In this embodiment, S5 includes the following sub-steps:

[0076] S51, calculating a signal energy ratio change value according to the signal energy ratio enhancement vector at low frequency, medium frequency, and high frequency;

[0077] S52, calculating the energy-spurious ratio change value according to the energy-spurious ratio enhancement vector at low frequency, medium frequency, and high frequency;

[0078] S53. Calculate a frequency spurious-to-interference ratio change value according to the frequency spurious-to-interference ratio enhancement vectors at low frequency, medium frequency, and high frequency.

[0079] In this embodiment, the formula for calculating the signal energy ratio change value in S51 is: Among them, s E is the signal energy ratio change value, r H,E is the sum of the elements in the signal energy ratio enhancement vector at high frequency, r M,E is the sum of the elements in the signal energy ratio enhancement vector at the intermediate frequency, r L,E is the sum of the elements in the signal energy ratio enhancement vector at low frequency;

[0080] The formula for calculating the energy spurious ratio change value in S52 is: Among them, s A is the energy spurious ratio change value, r H,A is the sum of the elements in the energy spurious ratio enhancement vector at high frequencies, r M,A is the sum of the elements in the energy spurious ratio enhancement vector at the intermediate frequency, r L,A is the sum of the elements in the energy spurious ratio enhancement vector at low frequency;

[0081] The formula for calculating the frequency spurious ratio change value in S53 is: Among them, s f is the frequency spurious ratio change value, r H,f is the sum of the elements in the frequency spurious ratio enhancement vector at high frequencies, r M,f is the sum of the elements in the frequency spurious ratio enhancement vector at the intermediate frequency, r L,fIt is the sum of the elements in the frequency spurious ratio enhancement vector at low frequency.

[0082] By calculating the signal energy ratio, energy-to-spurious ratio, and frequency-to-spurious ratio variation at low, medium, and high frequencies, this method can synthesize information about the impact of leakage on sound wave propagation at different frequencies. Sound waves of different frequencies have different propagation characteristics within a medium. Low-frequency sound waves have strong penetration, while high-frequency sound waves are sensitive to local variations. If the medium at the leak site is uneven, high-frequency sound waves are significantly affected, resulting in significant changes in the frequency-to-spurious ratio. Low-frequency sound waves, on the other hand, reflect overall structural changes. Combining this multi-frequency information provides a comprehensive understanding of the leakage situation.

[0083] like Figure 3 As shown, the barrier leakage detection model in S6 includes: a first feature fusion unit, a second feature fusion unit, a third feature fusion unit, a first LSTM unit, a second LSTM unit, a third LSTM unit, a first tanh unit, a second tanh unit, a third tanh unit, a first feature adjustment unit, a second feature adjustment unit, a third feature adjustment unit and an output layer;

[0084] The input end of the first feature fusion unit is used to input the signal energy ratio enhancement vector at low frequency, medium frequency and high frequency; the input end of the second feature fusion unit is used to input the energy spurious ratio enhancement vector at low frequency, medium frequency and high frequency; the input end of the third feature fusion unit is used to input the frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency;

[0085] The input end of the first LSTM unit is connected to the output end of the first feature fusion unit, and the output end thereof is connected to the input end of the first tanh unit; the first input end of the first feature adjustment unit is connected to the output end of the first tanh unit, and the second input end thereof is used to input the signal energy ratio change value;

[0086] The input end of the second LSTM unit is connected to the output end of the second feature fusion unit, and the output end thereof is connected to the input end of the second tanh unit; the first input end of the second feature adjustment unit is connected to the output end of the second tanh unit, and the second input end thereof is used to input the energy-spurious ratio change value;

[0087] The input end of the third LSTM unit is connected to the output end of the third feature fusion unit, and the output end thereof is connected to the input end of the third tanh unit; the first input end of the third feature adjustment unit is connected to the output end of the third tanh unit, and the second input end thereof is used to input the frequency spurious ratio change value;

[0088] The input end of the output layer is connected to the output end of the first feature adjustment unit, the output end of the second feature adjustment unit and the output end of the third feature adjustment unit respectively, and its output end serves as the output end of the isolation barrier leakage detection model.

[0089] In this embodiment, the output layer may be a fully connected layer.

[0090] This paper uses long-short-term memory (LSTM) units, which have powerful processing capabilities for sequential data and can learn long-term temporal and spatial dependencies between signal energy ratios, energy-to-spurious ratios, and frequency-to-spurious ratios. In landfill leak detection, data collected at different times or locations exhibit correlations. LSTM can capture these correlations, unlocking deeper data features and providing richer information for leak detection.

[0091] The feature fusion unit fuses multi-dimensional features, avoiding the limitations of a single feature and comprehensively reflecting the barrier status. After LSTM processing, the tanh unit performs a nonlinear transformation on the features, enhancing their expressive power and highlighting leakage-related features, enabling the model to better distinguish between normal and leakage states.

[0092] In this embodiment, the expression of the first feature fusion unit is: Among them, Y E is the signal energy ratio fusion feature, X H,E is the signal energy ratio enhancement vector at high frequency, X M,E is the signal energy ratio enhancement vector at the intermediate frequency, X L,E is the signal energy ratio enhancement vector at low frequency, For element-wise multiplication, F conv It is a point-by-point convolution operation;

[0093] The expression of the second feature fusion unit is: Among them, Y A is the energy-spurious ratio fusion feature, X H,A is the energy spurious ratio enhancement vector at high frequency, X M,A is the energy spurious ratio enhancement vector at the intermediate frequency, X L,A is the energy spurious ratio enhancement vector at low frequency;

[0094] The expression of the third feature fusion unit is: Among them, Y f is the frequency spurious ratio fusion feature, X H,f is the frequency spurious ratio enhancement vector at high frequency, X M,f is the frequency spurious ratio enhancement vector at the intermediate frequency, X L,f is the frequency spurious ratio enhancement vector at low frequency.

[0095] In this embodiment, the point-by-point convolution operation is a convolution layer of size 1*1.

[0096] The present invention fuses the signal energy ratio enhancement vectors corresponding to low frequency, medium frequency and high frequency through a first feature fusion unit, fuses the energy spurious ratio enhancement vectors corresponding to low frequency, medium frequency and high frequency through a second feature fusion unit, and fuses the frequency spurious ratio enhancement vectors corresponding to low frequency, medium frequency and high frequency through a third feature fusion unit. Sound waves of different frequencies have different propagation characteristics in landfill media and are affected differently by leakage. This fusion method can fully integrate multi-frequency information, comprehensively reflect the impact of leakage on sound wave propagation, and avoid the one-sidedness of single frequency information.

[0097] In this embodiment, the expressions of the first tanh unit, the second tanh unit, and the third tanh unit are all: Among them, g is the output of the tanh unit, c j is the jth eigenvalue of the tanh unit input, w j c j The weight of b j c j , R is the number of eigenvalues ​​of the tanh unit input, tanh is the hyperbolic tangent activation function, and j is a positive integer.

[0098] In this embodiment, the expressions of the first feature adjustment unit, the second feature adjustment unit and the third feature adjustment unit are all: y = h1 [1 + Sigmoid (wh2 + b)], where y is the output of the feature adjustment unit, h1 is the input of the first input terminal of the feature adjustment unit, h2 is the input of the second input terminal of the feature adjustment unit, w is the weight of h2, b is the bias of h2, and Sigmoid is the S-type activation function.

[0099] In this embodiment, the weights and biases in the barrier leakage detection model can be obtained by training using the existing gradient descent method.

[0100] The present invention sets an S-type activation function to measure the influence of the signal energy ratio change value, the energy spurious ratio change value and the frequency spurious ratio change value on the output of the tanh unit, thereby achieving an accurate assessment of the leakage situation.

[0101] The present invention places detection points equidistantly, forming a straight line, to detect the propagation of sound waves in that direction. By constructing signal energy ratio enhancement vectors, energy spurious ratio enhancement vectors, and frequency spurious ratio enhancement vectors, the present invention can effectively amplify the signal characteristics of leak points and reduce background noise interference, thereby significantly improving the accuracy of leak point identification.

[0102] The leakage point of the blocking barrier has different reflection and attenuation characteristics for different frequencies. Therefore, the present invention calculates the signal energy ratio change value, the energy spurious ratio change value and the frequency spurious ratio change value to reflect the changes in the signal energy ratio value, the energy spurious ratio value and the frequency spurious ratio value at different frequencies, and further highlights the signal characteristics of the leakage point.

[0103] The present invention adopts a barrier leakage detection model to process the signal energy ratio enhancement vector, energy spurious ratio enhancement vector and frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency, and performs feature adjustment based on the signal energy ratio change value, energy spurious ratio change value and frequency spurious ratio change value to improve the barrier leakage detection accuracy.

[0104] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for detecting leakage of a barrier barrier in a landfill, characterized in that: The following steps are involved: S1. Setting up an acoustic wave source at the landfill and setting up multiple detection points at equal intervals along the acoustic wave propagation path, wherein the multiple detection points form a straight line; S2. Emit low-frequency, medium-frequency, and high-frequency sound waves at the sound wave source, respectively, and collect forced vibration signals of corresponding frequencies at each detection point; S3, calculating the signal energy value, energy spurious value and frequency spurious value of each forced vibration signal; S4. Under the same transmission frequency, construct a signal energy ratio enhancement vector, an energy spurious ratio enhancement vector, and a frequency spurious ratio enhancement vector according to the signal energy values, energy spurious values, and frequency spurious values ​​of adjacent detection points; S5. Calculate the signal energy ratio change value, the energy spurious ratio change value, and the frequency spurious ratio change value according to the signal energy ratio enhancement vector, the energy spurious ratio enhancement vector, and the frequency spurious ratio enhancement vector at the three transmission frequencies; S6. Use the barrier leakage detection model to process the signal energy ratio enhancement vector, energy spurious ratio enhancement vector and frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency, and adjust the characteristics of the signal energy ratio change value, energy spurious ratio change value and frequency spurious ratio change value to obtain the barrier leakage degree value.

2. The landfill barrier leakage detection method according to claim 1, characterized in that: The formula for calculating the signal energy value of each forced vibration signal in S3 is: Where E is the signal energy value of the forced vibration signal, x n is the nth signal value in the forced vibration signal, n is a positive integer, and N is the length of the forced vibration signal.

3. The landfill barrier leakage detection method according to claim 1, characterized in that: The process of calculating the energy spurious value and the frequency spurious value of each forced vibration signal in S3 includes the following steps: A1. Perform spectrum analysis on each forced vibration signal to obtain the frequency value and amplitude; A2. Extract each amplitude and calculate the energy spurious value; A3. Extract each frequency value and calculate the frequency spurious value.

4. The landfill barrier leakage detection method according to claim 1, characterized in that: The S4 comprises the following sub-steps: S41. Calculate a signal energy ratio based on signal energy values ​​of adjacent detection points at the same transmission frequency. S42, arranging the signal energy ratios in sequence to construct a signal energy ratio vector, and performing enhancement processing on the signal energy ratio vector to obtain a signal energy ratio enhancement vector; S43. Calculate the energy spurious ratio based on the energy spurious values ​​of adjacent detection points at the same transmission frequency; S44, arranging the energy-stray ratio values ​​in sequence to construct an energy-stray ratio value vector, and performing enhancement processing on the energy-stray ratio vector to obtain an energy-stray ratio enhancement vector; S45. Calculate the frequency spurious ratio based on the frequency spurious values ​​of adjacent detection points at the same transmission frequency. S46. Arrange the frequency spurious ratios in sequence to construct a frequency spurious ratio vector, and perform enhancement processing on the frequency spurious ratio vector to obtain a frequency spurious ratio enhancement vector.

5. The landfill barrier leakage detection method according to claim 1, characterized in that: The S5 comprises the following sub-steps: S51, calculating a signal energy ratio change value according to the signal energy ratio enhancement vector at low frequency, medium frequency, and high frequency; S52, calculating the energy-spurious ratio change value according to the energy-spurious ratio enhancement vector at low frequency, medium frequency, and high frequency; S53. Calculate a frequency spurious-to-interference ratio change value according to the frequency spurious-to-interference ratio enhancement vectors at low frequency, medium frequency, and high frequency.

6. The landfill barrier leakage detection method according to claim 5, characterized in that: The formula for calculating the signal energy ratio change value is: Among them, s E is the signal energy ratio change value, r H,E is the sum of the elements in the signal energy ratio enhancement vector at high frequency, r M,E is the sum of the elements in the signal energy ratio enhancement vector at the intermediate frequency, r L,E is the sum of the elements in the signal energy ratio enhancement vector at low frequency; The formula for calculating the change in energy spurious ratio is: Among them, s A is the energy spurious ratio change value, r H,A is the sum of the elements in the energy spurious ratio enhancement vector at high frequencies, r M,A is the sum of the elements in the energy spurious ratio enhancement vector at the intermediate frequency, r L,A is the sum of the elements in the energy spurious ratio enhancement vector at low frequency; The formula for calculating the frequency spurious ratio change is: Among them, s f is the frequency spurious ratio change value, r H,f is the sum of the elements in the frequency spurious ratio enhancement vector at high frequencies, r M,f is the sum of the elements in the frequency spurious ratio enhancement vector at the intermediate frequency, r L,f It is the sum of the elements in the frequency spurious ratio enhancement vector at low frequency.

7. The landfill barrier leakage detection method according to claim 1, characterized in that: The barrier leakage detection model in S6 includes: a first feature fusion unit, a second feature fusion unit, a third feature fusion unit, a first LSTM unit, a second LSTM unit, a third LSTM unit, a first tanh unit, a second tanh unit, a third tanh unit, a first feature adjustment unit, a second feature adjustment unit, a third feature adjustment unit and an output layer; The input end of the first feature fusion unit is used to input the signal energy ratio enhancement vector at low frequency, medium frequency and high frequency; the input end of the second feature fusion unit is used to input the energy spurious ratio enhancement vector at low frequency, medium frequency and high frequency; the input end of the third feature fusion unit is used to input the frequency spurious ratio enhancement vector at low frequency, medium frequency and high frequency; The first feature fusion unit, the first LSTM unit, the first tanh unit and the first feature adjustment unit are connected in sequence; the second feature fusion unit, the second LSTM unit, the second tanh unit and the second feature adjustment unit are connected in sequence; the third feature fusion unit, the third LSTM unit, the third tanh unit and the third feature adjustment unit are connected in sequence; the output layer is respectively connected to the first feature adjustment unit, the second feature adjustment unit and the third feature adjustment unit.

8. The landfill barrier leakage detection method according to claim 7, characterized in that: The specific process of the first feature fusion unit includes: performing point-by-point convolution operations on the signal energy ratio enhancement vectors at high frequency, medium frequency, and low frequency, and then multiplying the three results after the point-by-point convolution operations element-by-element to obtain a signal energy ratio fusion feature; The specific process of the second feature fusion unit includes: performing point-by-point convolution operations on the energy-spurious ratio enhancement vectors at high frequency, medium frequency, and low frequency, and then multiplying the three results after the point-by-point convolution operations element-by-element to obtain the energy-spurious ratio fusion feature; The specific process of the third feature fusion unit includes: performing point-by-point convolution operations on the frequency spurious ratio enhancement vectors at high frequency, medium frequency, and low frequency respectively, and then multiplying the three results after the point-by-point convolution operations element-by-element to obtain the frequency spurious ratio fusion feature.

9. The landfill barrier leakage detection method according to claim 7, characterized in that: The expressions of the first feature adjustment unit, the second feature adjustment unit and the third feature adjustment unit are all: y=h1·[1+Sigmoid(wh2+b)], where y is the output of the feature adjustment unit, h1 is the input of the first input terminal of the feature adjustment unit, h2 is the input of the second input terminal of the feature adjustment unit, w is the weight of h2, b is the bias of h2, and Sigmoid is the S-type activation function.

Images