Calculation method and system for multi-point detection of gas nonmetal pipeline based on active sound source

Abstract

The invention discloses a calculation method and system for multi-point detection of a gas nonmetal pipeline based on an active sound source, and relates to the technical field of pipeline monitoring and management, and the method comprises the following steps: 1, monitoring the change of a sound signal in the gas nonmetal pipeline according to the change of a transmitted laser signal through an optical fiber of a distributed acoustic assembly as a sensing medium, according to the invention, the optical fiber of the distributed acoustic assembly is used for monitoring the change of the acoustic signal in the gas non-metal pipeline, the frequency offset, intensity and delay change of the laser signal are converted into the frequency, amplitude intensity and sound wave source of the acoustic signal in a correlation manner, and the adaptive filter is used for dynamically updating the filter coefficient. A window function is adopted to carry out framing processing on sound signals, short-time Fourier transform is applied, real classification labels are distributed to the sound signals through a random forest algorithm, a triangulation positioning method is applied to determine space coordinates of the sound signals, and accurate monitoring and classification of the pipeline state are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of pipeline monitoring and management, and particularly to a calculation method and system for detecting a non-metallic gas pipeline based on active sound source multi-point detection. Background Art

[0002] With the acceleration of the urbanization process, the gas transmission system plays a crucial role in modern urban infrastructure. The safety and reliability of gas pipelines are directly related to the safety of the city and the lives and property of residents. Traditionally, metal pipelines have been widely used in gas transmission. However, in recent years, with the progress of non-metallic material technology, non-metallic pipelines have gradually become an important part of the gas transmission system.

[0003] The advantages of non-metallic pipelines are their light weight, corrosion resistance, and simple construction, which can not only effectively reduce the installation cost of pipelines but also extend the service life of pipelines. Especially in some harsh environments, their excellent durability can be better exerted. However, the wide application of non-metallic pipelines also brings some new challenges, especially the problem of leakage detection during pipeline operation. Due to the different structures of non-metallic pipelines and metal pipelines, the monitoring and detection of leakage become more complex. Non-metallic pipelines usually have low thermal conductivity and weak electrical conductivity, which makes the application of traditional leakage detection methods have certain limitations. In addition, non-metallic pipelines are generally buried or installed concealed, and it is very difficult to directly observe and discover when leakage occurs, especially when the gas leakage volume is small or occurs underground, the detection of leakage is more difficult.

[0004] The prior art has the following deficiencies: In the urban environment, there are rich background noise sources in the existing gas monitoring technology, which easily interfere with the reception of detection signals, making signal recognition difficult. The fluctuation of environmental noise leads to a decrease in the accuracy of the system in identifying the target sound source, thus affecting the judgment of the pipeline state. The propagation of sound waves in non-metallic pipelines is affected by various factors, such as soil type, pipeline material, and its surrounding environment. It is often impossible to comprehensively consider complex factors, resulting in difficult accurate prediction of propagation characteristics under different conditions. Sound waves will experience multiple reflections and scatterings during propagation, resulting in distortion of measurement signals, increasing the complexity of actual signal analysis, reducing the reliability of recognition, and improper selection of sensor positions leading to low signal capture efficiency. Especially in the monitoring of long pipelines, if the sensor span is too large or the layout is uneven, it is easy to miss key leakage signals.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The object of the present invention is to provide a calculation method and system for detecting gas non-metallic pipelines based on multi-point detection of active sound sources, so as to solve the problems in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions: A calculation method for detecting gas non-metallic pipelines based on multi-point detection of active sound sources, comprising the following steps:

[0008] Step 1: Use the optical fiber of the distributed acoustic component as a sensing medium to monitor the change of the sound signal in the gas non-metallic pipeline according to the change of the propagated laser signal.

[0009] Step 2: Establish a frequency shift mechanism, a reflection intensity change mechanism, and a delay change mechanism of the sound signal, and convert the frequency shift amount, intensity, and delay change of the laser signal into the frequency, amplitude intensity, and sound wave source of the sound signal through correlation conversion.

[0010] Step 3: Dynamically update the filter coefficient through an adaptive filter according to the feedback mechanism of the sound signal, and output the sound signal with noise removed.

[0011] Step 4: Perform frame segmentation on the sound signal using a window function and apply short-time Fourier transform. Visualize the time-frequency representation of the frequency components of the sound signal changing with the application time through a spectrogram. Extract the main frequency characteristics of the sound signal from the spectrogram, and assign the true classification label of the gas non-metallic pipeline to each sound signal through a random forest.

[0012] Step 5: Select three points between the starting end and the ending end of the optical fiber as delay positioning points, and determine the sound signal coordinates through triangulation.

[0013] Preferably, introduce the optical fiber into the gas non-metallic pipeline to directly contact the area where the gas flows. Generate a continuous laser signal through the laser source connected to the starting end of the optical fiber and transmit it to different positions in the gas non-metallic pipeline. Receive the laser signal through the receiving device connected to the ending end of the optical fiber, and perform correlation conversion and preprocessing on the sound signal and the change of the laser signal through the signal processing center.

[0014] Preferably, a frequency offset mechanism of the acoustic signal is established. The acoustic signal affects the laser signal frequency through the scattering effect of the optical fiber. According to the velocity component existing in the relative motion between the acoustic signal and the laser signal, the frequency offset of the laser signal at the starting end and the ending end of the optical fiber is compared to be correlated and converted into the frequency of the acoustic signal. A reflection intensity change mechanism of the acoustic signal is established. The amplitude of the acoustic signal causes the vibration amplitude of the optical fiber. According to the proportional relationship between the amplitude of the acoustic signal and the deformation amplitude of the optical fiber, the intensities of the laser signals at the starting end and the ending end of the optical fiber are compared to be correlated and converted into the amplitude intensity of the acoustic signal. A delay change mechanism of the acoustic signal is established. According to the different transmission directions of the acoustic signals caused by the acoustic wave sources at different positions of the non-metallic gas pipeline, the signal delays of the laser signal reflections are different. By fixing the starting end of the optical fiber and comparing the delay changes of the laser signals at different ending ends, the acoustic wave sources of the acoustic signals are correlated and converted.

[0015] Preferably, set the initial weight of the adaptive filter, input the acoustic signal in real time, and output the acoustic signal passing through the adaptive filter. The specific formula is:

[0016]

[0017] Among them, y(t) represents the acoustic signal output passing through the adaptive filter, N represents the number of input acoustic signals, w(k) represents the adaptive filter parameter at the k-th moment, x(t - k) represents the acoustic signal input at the (t - k)-th moment. At each filtering time step, calculate the error between the reference acoustic signal and the acoustic signal output passing through the adaptive filter. The specific formula is:

[0018] e(t) = d(t) - y(t)

[0019] Among them, e(t) represents the error, d(t) represents the reference acoustic signal. Dynamically adjust the adaptive filter coefficients according to the error and update the weight. The specific formula is:

[0020] w(k)' = w(k) + μ * e(t) * x(t - k)

[0021] Among them, w(k)' represents the adaptive filter parameter dynamically adjusted at the k-th moment, and μ represents the filtering time step parameter controlling the update rate.

[0022] Preferably, select the window function width of the acoustic signal to be 512 points and the overlap ratio to be 50%. Frame the acoustic signal according to the window function width and multiply it by the window function. Apply the short-time Fourier transform to the windowed acoustic signal for each frame. The specific formula is:

[0023] S(T,f) = ∫s(τ)W(T - τ)e -j2πfτdτ 2πf

[0024] Among them, S(T, f) represents the time-frequency representation that describes the variation of the frequency components of the acoustic signal with the application time, s(τ) represents the amplitude of the acoustic signal at the time point τ, W(T - τ) represents the window function that limits the range of the acoustic signal analyzed by the short-time Fourier transform, 2πf represents the angular frequency corresponding to the frequency components of the acoustic signal. The time-frequency representation of the variation of the frequency components of the acoustic signal with the application time is visualized through a spectrogram, with the horizontal axis being time, the vertical axis being the frequency of the acoustic signal, and the depth of color representing the amplitude size of the acoustic signal. The main frequency characteristics of the acoustic signal are extracted according to the spectrogram.

[0025] Preferably, a tree structure of the gas non-metallic pipeline is constructed. The branch nodes inside the tree structure represent the main frequency characteristics, and the leaf nodes inside the tree structure represent the categories of the gas non-metallic pipeline. Multiple decision trees are constructed, and the gas non-metallic pipeline is classified according to the best splitting of the main frequency characteristics of randomly selected branch nodes by the leaf nodes. A true classification label of the gas non-metallic pipeline is assigned to each acoustic signal, including "pipeline normal", "pipeline slightly leaking", "pipeline moderately leaking", and "pipeline severely leaking".

[0026] Preferably, three points between the starting end and the ending end of the optical fiber are selected as the delay positioning points, and the cross-correlation function between each delay positioning point is calculated. The specific formula is:

[0027]

[0028] Among them, R ij represents the cross-correlation function between each delay positioning point, p i (n) represents the acoustic signal received by the i-th delay positioning point, p j (n + m) represents the acoustic signal received after shifting the j-th delay positioning point forward by m unit times at time n. The maximum value point of the cross-correlation function is found to determine the delay change of different delay positioning points, and the acoustic signal coordinates are determined by the triangulation method. The specific formula is:

[0029] |r i -r source |-|r j -r source |=v*m ij

[0030] Among them, r i represents the coordinate of the i-th or j-th delay positioning point, r source represents the acoustic signal coordinate, v represents the propagation speed of the acoustic signal in the space of the gas non-metallic pipeline, m ij represents the unit time from the i-th delay positioning point to the j-th delay positioning point.

[0031] A calculation system for detecting gas non - metallic pipelines based on multi - point detection of active sound sources, including a laser monitoring module, a correlation conversion module, a dynamic filtering module, an acoustic signal analysis module, and a detection and positioning module;

[0032] Laser monitoring module: Using the optical fiber of the distributed acoustic component as the sensing medium, it monitors the change of the acoustic signal in the gas non - metallic pipeline according to the change of the propagated laser signal;

[0033] Correlation conversion module: Establish a frequency offset mechanism, a reflection intensity change mechanism, and a delay change mechanism of the acoustic signal, and correlate and convert the frequency offset, intensity, and delay change of the laser signal into the frequency, amplitude intensity, and acoustic wave source of the acoustic signal;

[0034] Dynamic filtering module: Through an adaptive filter, it dynamically updates the filtering coefficient according to the feedback mechanism of the acoustic signal and outputs the acoustic signal with noise removed;

[0035] Acoustic signal analysis module: Use a window function to frame the acoustic signal and apply the short - time Fourier transform. Visualize the time - frequency representation of the frequency components of the acoustic signal changing with the application time through a spectrogram, extract the main frequency characteristics of the acoustic signal according to the spectrogram, and assign the true classification label of the gas non - metallic pipeline to each acoustic signal through a random forest;

[0036] Detection and positioning module: Select three points between the starting end and the ending end of the optical fiber as the delay positioning points, and determine the acoustic signal coordinates through triangulation;

[0037] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0038] 1. By establishing a frequency offset mechanism, a reflection intensity change mechanism, and a delay change mechanism of the acoustic signal, the effective correlation conversion between the laser signal and the acoustic signal is realized. The frequency offset of the acoustic signal affects the propagation characteristics of the laser signal, the amplitude change of the acoustic signal causes the vibration of the optical fiber, thereby affecting the intensity of the laser signal, and the different positions of the acoustic wave source lead to the change of the signal delay. It can accurately extract the frequency, amplitude intensity, and acoustic wave source of the acoustic signal, improving the accuracy of monitoring. Using an adaptive filter to dynamically update the filtering coefficient can eliminate the interference of external noise on the acoustic signal in real - time. Through setting the initial weight and the feedback mechanism of the real - time input signal, the filtering effect can be continuously optimized to ensure that the output acoustic signal is purer. Framing the acoustic signal and applying the short - time Fourier transform can effectively extract the frequency components of the acoustic signal and visualize them through a spectrogram, making the dynamic characteristics of the acoustic signal clear at a glance, which is conducive to quickly identifying and classifying different types of pipeline states. By constructing a tree - like structure and a random forest classification model based on the main frequency characteristics, the acoustic signal can be intelligently classified, not only improving the accuracy of classification, but also effectively dealing with noise and outliers.

[0039] 2. By selecting three delay positioning points between the starting end and the ending end of the optical fiber, and analyzing the delay change of the acoustic signal using the cross-correlation function, the coordinates of the acoustic signal can be accurately determined. Combining the acoustic wave propagation speed and the time delay ensures the efficiency and accuracy of positioning, enables real-time monitoring of the state of the gas non-metallic pipeline, and can promptly detect potential leakage problems. By assigning classification labels, different degrees of faults in the gas non-metallic pipeline can be quickly identified, helping maintenance personnel take corresponding measures to reduce safety hazards, and can be adjusted and optimized according to the specific conditions of different gas non-metallic pipelines. Description of the Drawings

[0040] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0041] Figure 1 It is a method flow chart of the calculation method for detecting gas non-metallic pipelines based on active sound sources at multiple points of the present invention.

[0042] Figure 2 It is a module schematic diagram of the calculation system for detecting gas non-metallic pipelines based on active sound sources at multiple points of the present invention. Detailed Embodiments

[0043] Now, the exemplary embodiments will be described more comprehensively with reference to the drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that the present disclosure will be more thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0044] Embodiment 1

[0045] The present invention provides a calculation method for detecting gas non-metallic pipelines based on active sound sources at multiple points as shown in Figure 1 and includes the following steps:

[0046] Step 1: Use the optical fiber of the distributed acoustic component as the sensing medium to monitor the change of the acoustic signal in the gas non-metallic pipeline according to the change of the propagated laser signal;

[0047] Introduce an optical fiber into a non-metallic gas pipeline to directly reach the area where the gas flows. A laser source connected to the starting end of the optical fiber generates a continuous laser signal and transmits it to different positions inside the non-metallic gas pipeline. Due to the high sensitivity of the optical fiber, when the laser signal passes through the optical fiber, the change in the acoustic signal along the non-metallic gas pipeline causes the optical fiber to undergo minute scattering or reflection. The laser signal is received by a receiving device connected to the end of the optical fiber, and the acoustic signal is correlated, converted, and preprocessed with the change in the laser signal through a signal processing center.

[0048] Assume that a vibration of a certain frequency occurs inside the pipeline. The optical signal is affected by the vibration during propagation, resulting in a frequency shift of the scattered light. By measuring this frequency change, the system can determine the frequency of the vibration source, thereby judging whether there is an abnormality in the non-metallic gas pipeline. If a large mechanical vibration occurs in the non-metallic gas pipeline (such as equipment startup or external impact), the optical fiber in this area will undergo a large deformation, thereby causing a change in the intensity of the reflected signal. By monitoring the fluctuations in these reflection intensities, the amplitude information of the sound wave can be analyzed.

[0049] Step 2: Establish a frequency shift mechanism, a reflection intensity change mechanism, and a delay change mechanism for the acoustic signal, and correlate and convert the frequency shift amount, intensity, and delay change of the laser signal into the frequency, amplitude intensity, and sound wave source of the acoustic signal;

[0050] Establish a frequency shift mechanism for the acoustic signal. The acoustic signal affects the frequency of the laser signal through the scattering effect of the optical fiber. According to the fact that there is a velocity component in the relative motion between the acoustic signal and the laser signal, that is, the frequency shift amount of the laser signal is proportional to the frequency of the acoustic signal. By comparing the frequency shift amounts of the laser signals at the starting end and the end of the optical fiber, it is correlated and converted into the frequency of the acoustic signal. Establish a reflection intensity change mechanism for the acoustic signal. The amplitude of the acoustic signal causes the vibration amplitude of the optical fiber. According to the proportional relationship between the amplitude of the acoustic signal and the deformation amplitude of the optical fiber, by comparing the intensities of the laser signals at the starting end and the end of the optical fiber, it is correlated and converted into the amplitude intensity of the acoustic signal. Establish a delay change mechanism for the acoustic signal. According to the fact that the transmission directions of the acoustic signals generated by sound wave sources at different positions in the non-metallic gas pipeline are different, resulting in different signal delays in the reflection of the laser signal, by fixing the starting end of the optical fiber and comparing the delay changes of the laser signals at different ends, it is correlated and converted into the sound wave source of the acoustic signal.

[0051] Step 3: Dynamically update the filter coefficient through an adaptive filter according to the feedback mechanism of the acoustic signal, and output the acoustic signal with noise removed;

[0052] Set the initial weight of the adaptive filter, input the acoustic signal in real time, and output the acoustic signal passing through the adaptive filter. The specific formula is:

[0053]

[0054] Among them, y(t) represents the acoustic signal output through the adaptive filter, N represents the number of input acoustic signals, w(k) represents the adaptive filter parameter at the k-th moment, x(t - k) represents the acoustic signal input at the (t - k)-th moment. At each filtering time step, the error between the reference acoustic signal and the acoustic signal output through the adaptive filter is calculated, and its specific formula is:

[0055] e(t) = d(t) - y(t)

[0056] Among them, e(t) represents the error, d(t) represents the reference acoustic signal. The adaptive filter coefficients are dynamically adjusted according to the error and the weights are updated, and its specific formula is:

[0057] w(k)' = w(k) + μ * e(t) * x(t - k)

[0058] Among them, w(k)' represents the adaptive filter parameter dynamically adjusted at the k-th moment, and μ represents the filtering time step parameter that controls the update rate.

[0059] Step 4: Use a window function to frame the acoustic signal and apply the short-time Fourier transform. Visualize the time-frequency representation of the frequency components of the acoustic signal changing with the application time through a spectrogram. Extract the main frequency features of the acoustic signal according to the spectrogram, and assign the true classification label of the gas non-metallic pipeline to each acoustic signal through a random forest;

[0060] Select the window function width of the acoustic signal to be 512 points and the overlap ratio to be 50%. Frame the acoustic signal according to the window function width and multiply it by the window function. Apply the short-time Fourier transform to the windowed acoustic signal for each frame, and its specific formula is:

[0061] S(T,f) = ∫s(τ)W(T - τ)e -j2πfτdτ 2πf

[0062] Among them, S(T,f) represents the time-frequency representation that describes the change of the frequency components of the acoustic signal with the application time, s(τ) represents the amplitude of the acoustic signal at the time point τ, W(T - τ) represents the window function that limits the range of the acoustic signal analyzed by the short-time Fourier transform, 2πf represents the angular frequency corresponding to the frequency components of the acoustic signal. Visualize the time-frequency representation of the frequency components of the acoustic signal changing with the application time through a spectrogram, with the horizontal axis being the time, the vertical axis being the frequency of the acoustic signal, and the color depth representing the amplitude size of the acoustic signal. Extract the main frequency features of the acoustic signal according to the spectrogram.

[0063] Construct a tree structure for gas non-metallic pipelines. Let the branch nodes inside the tree structure represent the main frequency features, and the leaf nodes inside the tree structure represent the categories of gas non-metallic pipelines. Construct multiple decision trees, and classify gas non-metallic pipelines according to the best split of the main frequency features of randomly selected branch nodes by leaf nodes. Assign true classification labels of gas non-metallic pipelines to each acoustic signal, including "pipeline normal", "pipeline slightly leaking", "pipeline medium leaking", and "pipeline severely leaking".

[0064] Step Five: Select three points between the starting end and the ending end of the optical fiber as the delay positioning points, and determine the acoustic signal coordinates through triangulation positioning method;

[0065] Select three points between the starting end and the ending end of the optical fiber as the delay positioning points, and calculate the cross-correlation function between each pair of delay positioning points. The specific formula is:

[0066]

[0067] where, R ij represents the cross-correlation function between each pair of delay positioning points, p i (n) represents the acoustic signal received by the i-th delay positioning point, p j (n + m) represents the acoustic signal received by shifting the j-th delay positioning point forward by m unit time at time n. Find the maximum point of the cross-correlation function to determine the delay change of different delay positioning points, and determine the acoustic signal coordinates through triangulation positioning method. The specific formula is:

[0068] |r i - r source |-|r j - r source | = v * m ij

[0069] where, r i represents the coordinate of the i-th or j-th delay positioning point, r source represents the acoustic signal coordinate, v represents the propagation speed of the acoustic signal in the space of the gas non-metallic pipeline, m ij represents the unit time from the i-th delay positioning point to the j-th delay positioning point.

[0070] Embodiment 2

[0071] The present invention provides a calculation system for detecting gas non-metallic pipelines based on multi-point detection of active sound sources as shown in Figure 2 which includes a laser monitoring module, an association conversion module, a dynamic filtering module, an acoustic signal analysis module, and a detection and positioning module;

[0072] Laser monitoring module: Using the optical fiber of the distributed acoustic component as the sensing medium, it monitors the change of the acoustic signal in the non-metallic gas pipeline according to the change of the propagated laser signal;

[0073] In this embodiment, it is specifically necessary to explain the laser monitoring module. The optical fiber of the laser monitoring module uses low-loss optical fiber to ensure the effective transmission of the laser signal. The optical fiber is reasonably arranged in the pipeline to ensure that it is in direct contact with the flowing gas. This layout can optimize the interaction between the laser signal and the gas flow and improve the monitoring sensitivity. When the laser source is in the on state, it continuously generates laser signals that propagate along the optical fiber at a specific frequency and intensity, ensuring coverage of the key areas of the non-metallic gas pipeline. When the laser is transmitted through the optical fiber into the pipeline, these optical signals will be affected by the acoustic signal (such as frequency shift, change in reflection intensity, and time delay). These changes reflect the actual acoustic environment and flow state in the pipeline.

[0074] Correlation conversion module: Establish a frequency shift mechanism, reflection intensity change mechanism, and time delay change mechanism of the acoustic signal, and correlate and convert the frequency shift amount, intensity, and time delay change of the laser signal into the frequency, amplitude intensity, and sound wave source of the acoustic signal;

[0075] In this embodiment, it is specifically necessary to explain the correlation conversion module. The correlation conversion module can determine the frequency of the acoustic signal by comparing the frequency shift amounts of the laser signals received at the starting end and the ending end of the optical fiber. For example, when the sound wave source moves relative to the optical fiber, the movement speed of the sound wave will cause a change in the frequency of the laser signal transmitted to the receiving end. By analyzing the frequency change, it can be successfully converted into the frequency information of the acoustic signal, thereby realizing the effective monitoring of the acoustic signal. By monitoring the intensities of the laser signals at the starting end and the ending end of the optical fiber, the change in the amplitude intensity of the acoustic signal can be extracted. This process involves real-time signal analysis and comparison, and can accurately convert the intensity change of the optical signal into the amplitude characteristics of the acoustic signal. Fixing the starting end of the optical fiber and analyzing the time delay change of the laser signals received at different ending ends, the time delay can be calculated with a certain algorithm, and these delays can be converted into the position of the sound wave source, greatly improving the accuracy and real-time performance of fault location.

[0076] Dynamic filtering module: Dynamically updates the filtering coefficient through an adaptive filter according to the feedback mechanism of the acoustic signal, and outputs the acoustic signal with noise removed;

[0077] In this embodiment, it is specifically necessary to explain the dynamic filtering module. The dynamic filtering module sets the initial weights of the adaptive filter, inputs the acoustic signal in real time, and realizes the efficient processing of the acoustic signal by dynamically adjusting the filter parameters. It not only improves the quality of the acoustic signal, but also provides reliable data support for subsequent analysis and decision-making. Especially in applications such as gas pipeline monitoring, it can significantly improve the accuracy and timeliness of fault detection.

[0078] Acoustic signal analysis module: The acoustic signal is framed using a window function and the short-time Fourier transform is applied. The time-frequency representation of the frequency components of the acoustic signal changing with the applied time is visualized through a spectrogram. The main frequency features of the acoustic signal are extracted from the spectrogram, and the true classification labels of the gas non-metal pipelines are assigned to each acoustic signal through a random forest;

[0079] In this embodiment, specifically, for the acoustic signal analysis module, the window function width of the acoustic signal selected by the acoustic signal analysis module is 512 points, and the overlapping ratio is 50%. Then the signal length of each frame is 512 samples, and the overlapping part between adjacent frames is 256 samples. The acoustic signal is framed according to the window function width to obtain multiple frame signals. Each frame signal is multiplied by the window function to reduce the edge effect and smooth the signal. The short-time Fourier transform is applied to the windowed acoustic signal of each frame to obtain the time-frequency representation, which can accurately extract the frequency components of the acoustic signal, provide rich time-frequency information, and help identify the pipeline state. The result of the short-time Fourier transform is visualized as a spectrogram, where the horizontal axis is time, the vertical axis is the frequency of the acoustic signal, and the color depth represents the amplitude size of the acoustic signal. The change of the frequency components of the acoustic signal with time is intuitively observed through the spectrogram to help extract the main frequency features of the acoustic signal. Multiple decision trees are constructed, and the main frequency features of the branch nodes are randomly selected according to the leaf nodes for the best split, aiming to maximize the information gain or minimize the impurity to effectively distinguish different categories, and can flexibly handle different acoustic signal features, improving the accuracy and robustness of classification.

[0080] Detection and positioning module: Three points between the starting end and the ending end of the optical fiber are selected as the delay positioning points, and the acoustic signal coordinates are determined by the triangulation method;

[0081] In this embodiment, specifically, for the detection and positioning module, between the starting end and the ending end of the optical fiber, three delay positioning points are selected to receive the acoustic signal for delay analysis and positioning. For each pair of delay positioning points, their cross-correlation function is calculated to determine the similarity of the acoustic signal between different positioning points. By calculating the value of the cross-correlation function, the maximum value point is found to determine the delay change between different delay positioning points, which is used to estimate the propagation time of the acoustic signal between different positioning points. Through the delay information, the triangulation method can be used to determine the coordinates of the acoustic signal. Using the triangulation method, through the known coordinates of the delay positioning points and the calculated delay information, the position of the coordinates of the acoustic signal in space can be solved.

[0082] By establishing the frequency offset mechanism, reflection intensity change mechanism, and delay change mechanism of the acoustic signal, the effective correlation conversion between the laser signal and the acoustic signal is achieved. The frequency offset of the acoustic signal affects the propagation characteristics of the laser signal, the amplitude change of the acoustic signal causes the vibration of the optical fiber, thereby affecting the intensity of the laser signal, and the different positions of the acoustic wave source result in the change of signal delay, enabling the accurate extraction of the frequency, amplitude intensity, and acoustic wave source of the acoustic signal, improving the monitoring accuracy. By using an adaptive filter to dynamically update the filter coefficients, the interference of external noise on the acoustic signal can be eliminated in real time. Through setting the initial weight and the feedback mechanism of the real-time input signal, the filtering effect can be continuously optimized to ensure that the output acoustic signal is purer. The acoustic signal is frame-processed and the short-time Fourier transform is applied to effectively extract the frequency components of the acoustic signal and visualize them through a spectrogram, making the dynamic characteristics of the acoustic signal clear at a glance, which is conducive to quickly identifying and classifying different types of pipeline states. By constructing a tree-like structure and a random forest classification model based on the main frequency characteristics, the acoustic signal can be intelligently classified, not only improving the classification accuracy but also effectively dealing with noise and outliers.

[0083] By selecting three delay positioning points between the starting end and the ending end of the optical fiber and using the cross-correlation function to analyze the delay change of the acoustic signal, the coordinates of the acoustic signal can be accurately determined. Combining the acoustic wave propagation speed and the time delay ensures the efficiency and accuracy of positioning, enables real-time monitoring of the state of the gas non-metallic pipeline, and timely discovers potential leakage problems. By assigning classification labels, different degrees of faults of the gas non-metallic pipeline can be quickly identified, helping maintenance personnel take corresponding measures to reduce potential safety hazards, and can be adjusted and optimized according to the specific conditions of different gas non-metallic pipelines.

[0084] Only some exemplary embodiments of the present invention have been described by way of illustration. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A calculation method based on active sound source multi-point detection of non-metallic gas pipelines, characterized in that: The following steps are involved: Step 1: Using the optical fiber of the distributed acoustic component as a sensing medium to monitor the acoustic signal changes in the gas non-metallic pipeline according to the propagated laser signal changes; Step 2: Establish the frequency offset mechanism, reflection intensity change mechanism and delay change mechanism of the acoustic signal, and convert the frequency offset, intensity and delay change of the laser signal into the frequency, amplitude intensity and sound wave source of the acoustic signal; Step 3: Dynamically update the filter coefficients through the adaptive filter according to the feedback mechanism of the sound signal, and output the sound signal with noise removed; Step 4: Use the window function to frame the acoustic signal and apply short-time Fourier transform. Visualize the time-frequency representation of the frequency components of the acoustic signal changing with the application time through the spectrum diagram. Extract the main frequency characteristics of the acoustic signal according to the spectrum diagram, and assign the true classification label of the gas non-metallic pipeline to each acoustic signal through random forest. Step 5: Select three points between the starting end and the end of the optical fiber as delay positioning points, and determine the coordinates of the acoustic signal by triangulation.

2. The calculation method based on active sound source multi-point detection of gas non-metallic pipelines according to claim 1 is characterized by: In step one, the distributed acoustic component specifically includes a laser source that generates a continuous laser signal, an optical fiber that independently senses the sound wave changes in the gas non-metallic pipeline, a receiving device that receives and analyzes the reflected signal returned from the optical fiber, and a signal processing center that processes and analyzes the reflected signal.

3. The calculation method based on active sound source multi-point detection of non-metallic gas pipelines according to claim 1 is characterized by: In the step 2, the frequency offset mechanism specifically includes: according to the velocity component of the relative movement between the acoustic signal and the laser signal, the frequency offset of the laser signal at the starting end and the end of the optical fiber is compared to correlate the frequency of the acoustic signal; the reflection intensity change mechanism specifically includes: according to the proportional relationship between the amplitude of the acoustic signal and the deformation amplitude of the optical fiber, the intensity of the laser signal at the starting end and the end of the optical fiber is compared to correlate the amplitude intensity of the acoustic signal; the delay change mechanism specifically includes: according to the different transmission directions of the acoustic signal caused by the acoustic wave sources located at different positions of the gas non-metallic pipeline, resulting in different signal delays of the laser signal reflection, the delay changes of the laser signals at different ends are compared by fixing the starting end of the optical fiber to correlate the acoustic wave source converted into the acoustic signal.

4. The calculation method based on active sound source multi-point detection of non-metallic gas pipelines according to claim 1 is characterized by: In step 3, the specific formula for outputting the noise-removed acoustic signal is: Wherein, y(t) represents the acoustic signal output through the adaptive filter, N represents the number of input acoustic signals, w(k) represents the adaptive filter parameter at the kth moment, and x(tk) represents the laser signal input at the tkth moment.

5. The calculation method based on active sound source multi-point detection of gas non-metallic pipelines according to claim 1 is characterized by: In step 3, the specific formula for dynamically updating the filter coefficient is: e(t)=d(t)-y(t) w(k)'=w(k)+μ*e(t)*x(tk) Wherein, e(t) represents the error, d(t) represents the reference acoustic signal, the adaptive filter coefficients are dynamically adjusted and the weights are updated according to the error, w(k)' represents the adaptive filter parameters dynamically adjusted at the kth moment, and μ represents the filter time step parameter that controls the update rate.

6. The calculation method based on active sound source multi-point detection of non-metallic gas pipelines according to claim 1 is characterized by: In step 4, the specific formula of the short-time Fourier transform is: S(T,f)=∫s(τ)W(T-τ)e -j2πfτdτ 2πf Among them, S(T,f) represents the time-frequency representation that describes the change of the frequency components of the acoustic signal with the application time, s(τ) represents the amplitude of the acoustic signal at the time point τ, W(T-τ) represents the window function that limits the range of the acoustic signal analyzed by the short-time Fourier transform, and 2πf represents the angular frequency corresponding to the frequency components of the acoustic signal.

7. The calculation method based on active sound source multi-point detection of gas non-metallic pipelines according to claim 1 is characterized by: In the step 4, the specific steps of random forest are: constructing a tree structure of gas non-metallic pipelines, the branch nodes inside the tree structure represent the main frequency features, the leaf nodes inside the tree structure represent the categories of gas non-metallic pipelines, constructing multiple decision trees, and classifying gas non-metallic pipelines by randomly selecting the best splitting of the main frequency features of branch nodes according to the leaf nodes.

8. The calculation method based on active sound source multi-point detection of non-metallic gas pipelines according to claim 1 is characterized by: In step 5, the cross-correlation function between each delayed positioning point has the following specific formula: Among them, R ij represents the cross-correlation function between each delayed positioning point, p i (n) represents the acoustic signal received by the ith delayed positioning point, p j (n+m) represents the sound signal received after the j-th delayed positioning point is shifted forward by m units of time at time n.

9. The calculation method based on active sound source multi-point detection of gas non-metallic pipelines according to claim 1 is characterized by: In step 5, the coordinates of the acoustic signal are determined by triangulation, and the specific formula is: |r i -r source |-|r j -r source |=v*m ij Among them, r i represents the coordinates of the i-th or j-th delayed positioning point, r source represents the coordinates of the acoustic signal, v represents the speed at which the acoustic signal propagates in the space of the gas non-metallic pipeline, m ij Represents the unit time from the i-th delay positioning point to the j-th delay positioning point.

10. A computing system for detecting non-metallic gas pipelines based on active sound sources at multiple points, used to implement the computing method for detecting non-metallic gas pipelines based on active sound sources at multiple points as described in any one of claims 1 to 9, characterized in that: It includes laser monitoring module, correlation conversion module, dynamic filtering module, acoustic signal analysis module and detection and positioning module; Laser monitoring module: uses the optical fiber of the distributed acoustic component as the sensing medium to monitor the acoustic signal changes in the gas non-metallic pipeline according to the propagating laser signal changes; Correlation conversion module: establish the frequency offset mechanism, reflection intensity change mechanism and delay change mechanism of the acoustic signal, and convert the frequency offset, intensity and delay change of the laser signal into the frequency, amplitude intensity and sound wave source of the acoustic signal; Dynamic filtering module: dynamically updates the filter coefficients based on the feedback mechanism of the sound signal through an adaptive filter, and outputs the sound signal with noise removed; Acoustic signal analysis module: Frame the acoustic signal using the window function and apply short-time Fourier transform. Visualize the time-frequency representation of the frequency components of the acoustic signal changing with the application time through the spectrum diagram. Extract the main frequency characteristics of the acoustic signal based on the spectrum diagram. Use random forest to assign the true classification label of the gas non-metallic pipeline to each acoustic signal. Detection and positioning module: select three points between the starting end and the end of the optical fiber as delay positioning points, and determine the coordinates of the acoustic signal through the triangulation positioning method.

Images