Portable pipeline blockage point quick positioning instrument and method based on pressure transient analysis

Abstract

The invention discloses a portable pipeline blockage point rapid positioning instrument and method based on pressure transient analysis, the positioning instrument comprises a positioning instrument host, a butt joint of a three-way structure, a high-speed electromagnetic valve and a high-frequency pressure sensor, the high-speed electromagnetic valve and the high-frequency pressure sensor are integrated in the butt joint; and signal and power supply connection with the host is realized through the quick connector. The positioning method comprises the following steps: connecting equipment and recovering a signal; the host controls the high-speed solenoid valve to instantly open and close to generate pressure pulse; the high-frequency pressure sensor collects incident and blocking point reflection pulse signals; the host preprocesses the data through wavelet transform, cross-correlation analysis and other algorithms and calculates a pulse time difference delta t; and calculating and displaying the distance of the blocking point according to a formula S = (wave velocity a * delta t) / 2, wherein the wave velocity can be calibrated through a self-adaptive algorithm. The positioning precision reaches the meter level, equipment is portable, operation is easy and convenient, point distribution and excavation in advance are not needed, the method is suitable for multi-material pipelines, and the detection cost and the technical threshold are reduced.

Description

Technical Field

[0001] This invention relates to the field of pipeline blockage location equipment technology, and in particular to a portable pipeline blockage rapid location instrument and method based on pressure transient analysis. Background Technology

[0002] Currently, the main technical solutions for locating blockages inside underground pipelines have significant drawbacks: 1. Manual segmented inspection method: This method requires technicians to rely on their experience to excavate and inspect sections of the pipeline or use endoscopes for exploration. This method is highly unreliable, inefficient, costly, and has a huge impact on traffic and the environment.

[0003] 2. Flow / Pressure Monitoring Method: This method involves installing flow meters or pressure gauges at multiple points along the pipeline and using data anomalies to roughly determine the blockage area. This method requires the pre-deployment of a large number of fixed monitoring devices, resulting in high initial investment, and it cannot achieve rapid, random single-point location, thus lacking flexibility.

[0004] 3. Traditional acoustic leak detection method: This method relies on the human ear or a listening rod to detect the sound produced by fluid passing through a leak or blockage point for localization. However, for blockage points, the acoustic signal is often very weak and easily drowned out by ambient noise, resulting in poor positioning accuracy, high dependence on operator experience, and is not suitable for non-metallic pipes.

[0005] The core flaw of existing technology can be summarized as follows: it is impossible to achieve rapid, accurate, and portable one-time location of blockage points without pre-installing equipment or excavation. Summary of the Invention

[0006] The technical problem to be solved by this invention is that existing pipe blockage detection methods cannot achieve rapid, accurate, and portable one-time location of blockage points without pre-installing equipment or excavation.

[0007] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a portable pipeline blockage rapid locator based on pressure transient analysis, including a locator host, a docking joint, a high-speed solenoid valve and a high-frequency pressure sensor; The docking connector is a tee connector. The positioning instrument main unit is connected to the pipeline to be tested through the docking connector. A high-speed solenoid valve and a high-frequency pressure sensor that are connected to the positioning instrument main unit are installed on the docking connector.

[0008] Furthermore, the total weight of the positioning device main unit does not exceed 5kg, its battery module supports hot-swapping, and its continuous working time under full charge is not less than 4 hours.

[0009] Furthermore, the high-speed solenoid valve and the high-frequency pressure sensor are fixedly arranged inside the mating joint. The end of the mating joint is sealed and equipped with a quick connector. The connection lines of the high-speed solenoid valve and the high-frequency pressure sensor are connected to the quick connector and connected to the positioning instrument host through the quick connector.

[0010] A portable method for rapid location of pipe blockages based on pressure transient analysis, using the aforementioned locator, includes the following steps: Step 1: Connection. Connect the docking connector to the pipe to be tested, connect the positioning instrument main unit and the docking connector, and restore the signal connection between the positioning instrument main unit and the docking connector. Step 2: Excitation. The positioning instrument host sends a signal and controls the high-speed solenoid valve to open and close instantaneously, generating a brief and steep pressure pulse signal. Step 3: Data Acquisition. The pressure pulse signal propagates downstream along the pipeline under test. The high-frequency pressure sensor synchronously and continuously records the pressure data at the moment of transmission and subsequent pressure changes, including the reflected pressure pulse signal after encountering a blockage point. Step 4: Analysis. The collected pressure-time data is analyzed by the signal processing algorithm unit built into the positioning device. Through cross-correlation analysis and wavelet transform algorithm, the feature points of the pressure pulse signal and the reflected pressure pulse signal are accurately identified, and the time difference Δt between the two is calculated. Step 5: Display. Based on the formula distance S=(wave velocity a×Δt) / 2, the distance to the blockage point is automatically calculated, and the result is clearly displayed on the main touch screen of the positioning device.

[0011] Furthermore, in step four, the received data is preprocessed, including the following steps: S401. The original pressure signal is decomposed into multiple scales using the Daubechies wavelet basis, and the high-frequency coefficients are processed by soft thresholding to suppress environmental noise and high-frequency interference from the sensor. S402. Smooth the low-frequency signal to eliminate small fluctuations and retain the main characteristics of the pressure pulse; S403: Remove signal drift caused by temperature or pressure drift through polynomial fitting.

[0012] Furthermore, the specific decomposition steps of step S401 include: Let the original pressure signal be f(t), and use a low-pass filter h and a high-pass filter g to process the signal; In the j-th level decomposition, the signal f(t) is decomposed into approximate coefficients. and detail coefficient The approximation coefficient represents low-pass filtering, and the detail coefficient represents high-pass filtering. The specific calculation formula is as follows: Detail factor: ; Detail factor: ; in, It is the first Approximation coefficients of the layer and These are the coefficients of the low-pass filter and the high-pass filter, respectively. This represents the discrete-time index. By iterating through the above process, the signal is decomposed into multiple scales, yielding approximation coefficients and detail coefficients at different scales. These coefficients correspond to the low-frequency and high-frequency components of the signal, respectively. The specific decomposition steps for soft thresholding of high-frequency coefficients include: The high-frequency coefficients after wavelet transform are processed by setting a threshold. The coefficients with absolute values ​​less than the threshold are set to zero, and the coefficients with absolute values ​​greater than the threshold are shrunk to suppress noise and interference. The mathematical expression for the soft threshold function is: ; in, These are the high-frequency coefficients after wavelet transform. The threshold is a set value, determined based on an empirical range of noise levels. The determination of the threshold λ is based on the principle of balancing noise statistical characteristics and signal fidelity. Its theoretical basis is the wavelet threshold denoising theory proposed by Donoho and Johnstone. The core idea is that the threshold λ should be proportional to the noise standard deviation σ. For Gaussian white noise, the ideal universal threshold is λ=σ√(2ln(N)), where N is the signal length. For the non-stationary characteristics of pipeline pressure pulse signals, an improved empirical value determination method is adopted.

[0013] Furthermore, in step S402, the smoothing of the low-frequency signal uses an exponentially weighted moving average filtering method, and the specific calculation formula is as follows: ; in It is the filtered output at the current moment. It is the input signal at the current moment. It is the filtered output from the previous moment. It is a smoothing coefficient, with a typical value range of 0.1-0.3; The larger the value, the greater the weight given to the new data, and the faster the filter responds to changes in the signal. The smaller the value, the higher the dependence on historical data, and the better the smoothing effect.

[0014] Furthermore, the specific steps for removing signal offset using polynomial fitting in step S403 are as follows: Assume the collected pressure signal is The corresponding time point is ( );choose Fitting is performed using a polynomial of order 1, the polynomial of which is in the form of: ; in, These are the polynomial coefficients to be determined; These coefficients are then determined using the least squares method to minimize the sum of squared errors between the fitted polynomial and the original signal; Sum of squared errors The calculation formula is: ; In order to Minimum, for about ( Find the partial derivative and set it equal to Solving this system of linear equations will yield the polynomial coefficients. ; Obtain the fitting polynomial Then, remove the pressure signal after signal offset. for: .

[0015] Furthermore, in step four, when extracting pressure pulse features, the multi-resolution characteristics of wavelet transform in the time and frequency domain are used to detect abrupt changes in the pressure signal. Then, empirical mode decomposition (EMD) is performed on the signal to extract intrinsic mode functions. The instantaneous frequency is calculated through Hilbert transform to accurately locate the peak time. Finally, the standard incident wave template is compared with the measured signal to align the time series and eliminate the influence of propagation distortion.

[0016] Furthermore, the screening termination condition for the intrinsic mode functions (IMFs) in the empirical mode decomposition (EMD) is that the standard deviation SD of two consecutive screening results is less than 0.001, SD=Σ[(IMF{k-1}(t)-IMF_k(t))²] / Σ[IMF{k-1}(t)²]); when the Hilbert transform calculates the instantaneous frequency, an analytical signal Z_i(t)=c_i(t)+j·H[c_i(t)] is constructed for each IMF component c_i(t), the instantaneous frequency f_i(t)=(1 / 2π)·d[argZ_i(t)] / dt, the effective frequency range is limited to 0.1-2kHz to suppress noise, and the peak time is determined by the maximum point of the instantaneous amplitude a_i(t)=|Z_i(t)|.

[0017] Furthermore, in step four, the signal processing algorithm unit employs a joint time delay estimation strategy combining cross-correlation analysis and generalized cross-correlation when calculating the time difference Δt between the pressure pulse signal and the reflected pressure pulse signal. Specifically, this includes: S404. Using the pressure pulse signal generated by the signal excitation sent by the positioning instrument host as the incident wave reference signal x(t), and the measured data including the reflected pressure pulse signal collected by the high-frequency pressure sensor as y(t), the initial time delay is obtained through cross-correlation calculation. The cross-correlation formula is: ; Where τ is the time delay, and the value of τ corresponding to the peak value of R(τ) is the initial time difference between the incident wave and the reflected wave; S405. The above cross-correlation results are optimized using PHAT weighting to enhance the robustness of time delay estimation in low signal-to-noise ratio environments. The generalized cross-correlation formula is as follows: ; Where X(f) and Y(f) are the Fourier transforms of x(t) and y(t) respectively, X(f) is the conjugate of X(f), and IFFT represents the inverse Fourier transform; S406. Combine the optimized generalized cross-correlation result with the matched filter output of the incident wave template on the measured signal, and accurately determine the final time difference Δt by finding the overlap position between the peak value of the matched filter output and the peak value of the generalized cross-correlation.

[0018] Furthermore, in step five, when calculating the distance to the blockage point and outputting the results, an enhancement strategy combining adaptive wave velocity calibration and result verification is adopted, specifically including: S501. If the pipe material or fluid properties are unknown, the positioning instrument host controls a high-speed solenoid valve to emit multiple pressure pulses. The pressure wave velocity 'a' is then calculated using a wave velocity adaptive calibration algorithm. The algorithm formula is as follows: ; Where L is the known characteristic length of the pipe input through on-site measurement, and Δt and Δt are the time differences between the two reflected pressure pulse signals and the incident wave signal, respectively; S502. The Levenberg-Marquardt algorithm is used to fit the multiple reflection time series, and the pressure wave velocity a and the distance S of the blockage point are solved simultaneously and optimized. The final distance of the blockage point is calculated by combining the formula S=(a×Δt) / 2. S503. The reflection signal features are identified by a machine learning classifier to distinguish between blockage reflections and non-blockage reflections caused by valves and elbows. The Bootstrap resampling method is used to generate the statistical distribution of Δt. The distance result containing a 95% confidence interval is output on the host touch screen in the format S±ΔS, where ΔS is the half-width of the confidence interval.

[0019] Furthermore, based on the amplitude ratio of the reflected pressure pulse signal (8) to the pressure pulse signal (7), it is determined whether the blockage point is completely blocked or partially blocked.

[0020] This invention provides a portable rapid pipe blockage location instrument and method based on pressure transient analysis, which has the following beneficial effects.

[0021] 1. The method of locating pipe blockages has been upgraded from traditional manual experience-based judgment to precise physical measurement. The positioning accuracy can reach the meter level or even higher, and the result can be obtained in one test, which greatly improves efficiency.

[0022] 2. The equipment is highly integrated, lightweight, and can be carried and operated by a single person. It only requires connecting to one point and can complete positioning within minutes, making it ideal for on-site emergency testing and routine inspections.

[0023] 3. No need to pre-bury sensors or conduct large-scale excavation, significantly reducing the economic and time costs of detection.

[0024] 4. The equipment is suitable for pipes of various materials and various fluids, and has a wide range of applications.

[0025] 5. Automated data processing and direct output of distance results greatly reduce the technical threshold and reliance on the operator's experience. Attached Figure Description

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a schematic diagram of the structure of an embodiment of the present invention.

[0027] In the diagram: 1. Pipe to be tested; 2. Blockage point; 3. Positioning instrument main unit; 4. Quick connector; 5. High-speed solenoid valve; 6. High-frequency pressure sensor; 7. Pressure pulse signal; 8. Reflected pressure pulse signal; 9. Main unit touch screen display; 10. Connecting joint. Detailed Implementation

[0028] like Figure 1 As shown, the present invention provides a portable pipeline blockage rapid location instrument based on pressure transient analysis, including a locator host 3, a docking joint 10, a high-speed solenoid valve 5 and a high-frequency pressure sensor 6; The docking connector 10 is a tee connector. The positioning instrument host 3 is connected to the pipeline 1 to be tested through the docking connector 3. The docking connector 10 is equipped with a high-speed solenoid valve 5 and a high-frequency pressure sensor 6 that are connected to the positioning instrument host 3 via signal.

[0029] The instantaneous opening and closing of the high-speed solenoid valve 5 generates a pressure pulse signal 7 within the test pipe 1. This pressure pulse signal 7 is detected by the high-frequency pressure sensor 6. As the pressure pulse signal 7 propagates downstream along the test pipe 1, it passes through the blockage point 2, generating a reverse reflected pressure pulse signal 8, which is also detected by the high-frequency pressure sensor 6 and transmitted to the positioning device host 3. The positioning device is installed by connecting the tee connector 10 in series to the test pipe 1, ensuring a tight connection and improving installation efficiency. The positioning device host 3 weighs no more than 5 kg, its battery module supports hot-swapping, and its continuous operating time on a full charge is no less than 4 hours. like Figure 1 As shown. The high-speed solenoid valve 5 and the high-frequency pressure sensor 6 are fixedly arranged inside the docking joint 10. The end of the docking joint 10 is sealed and equipped with a quick connector 4. The connection lines of the high-speed solenoid valve 5 and the high-frequency pressure sensor 6 are connected to the quick connector 4 and connected to the positioning instrument host 3 through the quick connector 4. A stud is provided on the connection end of the docking joint 10 and the positioning instrument host 3. The base plate of the positioning instrument host 3 is locked to the docking joint 10 with a nut. After the positioning instrument host 3 is fixed, the positioning instrument host 3 is connected to the high-speed solenoid valve 5 and the high-frequency pressure sensor 6 through the quick connector 4 to realize control connection, power supply and signal connection. The high-speed solenoid valve 5 and the high-frequency pressure sensor 6 are integrated and installed inside the docking joint 10. During installation, the docking joint 10 is directly installed to improve the overall installation efficiency of the positioning instrument.

[0030] A portable method for rapid location of pipe blockages based on pressure transient analysis, using the aforementioned locator, includes the following steps: S1. Connect the docking connector 10 to the pipe to be tested 1, connect the positioning instrument host 3 and the docking connector 10, and restore the signal connection between the positioning instrument host 3 and the docking connector 10. S2, Excitation, through the positioner host 3, the positioner host 3 sends a signal and controls the high-speed solenoid valve 5 to open and close instantaneously, generating a short and steep pressure pulse signal 7; S3. Acquisition: The pressure pulse signal 7 propagates downstream along the pipeline 1 under test. The high-frequency pressure sensor 6 synchronously and continuously records the pressure data at the moment of transmission and subsequent pressure changes, including the reflected pressure pulse signal 8 reflected back after encountering the blockage point 2. S4. Analysis: The collected pressure-time data is analyzed by the signal processing algorithm unit built into the positioning instrument host 3. Through cross-correlation analysis and wavelet transform algorithm, the feature points of the pressure pulse signal 7 and the reflected pressure pulse signal 8 are accurately identified, and the time difference Δt between them is calculated. In the process of extracting pressure pulse features, the multi-resolution characteristics of wavelet transform in the time and frequency domain are used to detect abrupt changes in the pressure signal. Then, empirical mode decomposition (EMD) is performed on the signal to extract the intrinsic mode functions. The instantaneous frequency is calculated by Hilbert transform to accurately locate the peak time. Finally, the standard incident wave template is compared with the measured signal to align the time series and eliminate the influence of propagation distortion. Empirical mode decomposition (EMD) refinement: The termination condition of IMF decomposition is that the standard deviation SD of two consecutive screening results is <0.001 (SD=Σ[(IMF{k-1}(t)-IMF_k(t))²] / Σ[IMF{k-1}(t)²]).

[0031] Hilbert transform refinement: For each IMF component c_i(t), construct an analytic signal Z_i(t) = c_i(t) + j·H[c_i(t)], with an instantaneous frequency f_i(t) = (1 / 2π)·d[argZ_i(t)] / dt. The effective frequency range is limited to 0.1-2kHz to suppress noise, and the peak time is determined by the maximum point of the instantaneous amplitude a_i(t) = |Z_i(t)|.

[0032] First, the data needs to be denoised; S401, the original pressure signal is decomposed into multiple scales using the Daubechies wavelet basis, and the high-frequency coefficients are processed by soft thresholding to suppress environmental noise and high-frequency interference from the sensor; Let the original pressure signal be f(t), and use a low-pass filter h and a high-pass filter g to process the signal; In the j-th level decomposition, the signal f(t) is decomposed into approximate coefficients. and detail coefficient The approximation coefficient represents low-pass filtering, and the detail coefficient represents high-pass filtering. The specific calculation formula is as follows: Detail factor: ; Detail factor: ; in, It is the first Approximation coefficients of the layer and These are the coefficients of the low-pass filter and the high-pass filter, respectively. This represents the discrete-time index. By iterating through the above process, the signal is decomposed into multiple scales, yielding approximation coefficients and detail coefficients at different scales. These coefficients correspond to the low-frequency and high-frequency components of the signal, respectively. The specific decomposition steps for soft thresholding of high-frequency coefficients include: The high-frequency coefficients after wavelet transform are processed by setting a threshold. The coefficients with absolute values ​​less than the threshold are set to zero, and the coefficients with absolute values ​​greater than the threshold are shrunk to suppress noise and interference. The mathematical expression for the soft threshold function is: ; in, These are the high-frequency coefficients after wavelet transform. It is a set threshold, set according to the range of empirical values. Considering the characteristics of pipeline pressure pulse signals, λ can be 0.02-0.08 times the peak value of the original signal. When the signal-to-noise ratio is below 10dB, the upper limit is taken, and when it is above 20dB, the lower limit is taken, taking into account both noise suppression and signal feature preservation.

[0033] The screening termination condition for intrinsic mode functions (IMFs) in the empirical mode decomposition (EMD) is that the standard deviation (SD) of two consecutive screening results is less than 0.001; when the Hilbert transform is used to calculate the instantaneous frequency, its effective frequency analysis range is limited to 0.1 kHz to 2 kHz.

[0034] The threshold λ is determined based on the principle of balancing noise statistical characteristics and signal fidelity. Its theoretical basis is the wavelet threshold denoising theory proposed by Donoho and Johnstone. The core idea is that the threshold λ should be proportional to the noise standard deviation σ. For Gaussian white noise, the ideal universal threshold is λ=σ√(2ln(N)), where N is the signal length. For the non-stationary characteristics of pipeline pressure pulse signals, an improved empirical value determination method is adopted.

[0035] S402. Smooth the low-frequency signal to eliminate small fluctuations and retain the main characteristics of the pressure pulse; The smoothing of low-frequency signals uses an exponentially weighted moving average filtering method, and the specific calculation formula is as follows: ; in It is the filtered output at the current moment. It is the input signal at the current moment. It is the filtered output from the previous moment. It is a smoothing coefficient, and its value ranges from... between; The larger the value, the greater the weight given to the new data, and the faster the filter responds to changes in the signal. The smaller the value, the higher the dependence on historical data, and the better the smoothing effect.

[0036] S403. Remove signal drift caused by temperature or pressure drift by polynomial fitting; The specific steps for removing signal offset using polynomial fitting are as follows: Assume the collected pressure signal is The corresponding time point is ( );choose Fitting is performed using a polynomial of order 1, the polynomial of which is in the form of: ; in, These are the polynomial coefficients to be determined; These coefficients are then determined using the least squares method to minimize the sum of squared errors between the fitted polynomial and the original signal; Sum of squared errors The calculation formula is: ; In order to Minimum, for about ( Find the partial derivative and set it equal to Solving this system of linear equations will yield the polynomial coefficients. ; Obtain the fitting polynomial Then, remove the pressure signal after signal offset. for: .

[0037] In step four, when extracting pressure pulse features, the multi-resolution characteristics of wavelet transform in the time and frequency domain are used to detect abrupt changes in the pressure signal. Then, empirical mode decomposition (EMD) is performed on the signal to extract intrinsic mode functions. The instantaneous frequency is calculated through Hilbert transform to accurately locate the peak time. Finally, the standard incident wave template is compared with the measured signal to align the time series and eliminate the influence of propagation distortion.

[0038] In step four, the signal processing algorithm unit employs a joint time delay estimation strategy combining cross-correlation analysis and generalized cross-correlation when calculating the time difference Δt between the pressure pulse signal and the reflected pressure pulse signal. Specifically, this includes: S404. Using the pressure pulse signal 7 generated by the signal excitation sent by the positioning instrument host as the incident wave reference signal x(t), and the measured data including the reflected pressure pulse signal 8 collected by the high-frequency pressure sensor 6 as y(t), the initial time delay is obtained through cross-correlation calculation. The cross-correlation formula is: ; Where τ is the time delay, and the value of τ corresponding to the peak value of R(τ) is the initial time difference between the incident wave and the reflected wave; S405. The above cross-correlation results are optimized using PHAT weighting to enhance the robustness of time delay estimation in low signal-to-noise ratio environments. The generalized cross-correlation formula is as follows: ; Where X(f) and Y(f) are the Fourier transforms of x(t) and y(t) respectively, X(f) is the conjugate of X(f), and IFFT represents the inverse Fourier transform; S406. Combine the optimized generalized cross-correlation result with the matched filter output of the incident wave template on the measured signal, and accurately determine the final time difference Δt by finding the overlap position between the peak value of the matched filter output and the peak value of the generalized cross-correlation.

[0039] S5. Display: Based on the formula S = wave velocity a × Δt / 2, the distance to blockage point 2 is automatically calculated, and the result is clearly displayed on the host touch screen 9 on the host unit 3 of the positioning device. In step five, when calculating the distance to the blockage point and outputting the results, an enhancement strategy combining adaptive wave velocity calibration and result verification is adopted, specifically including: S501. If the pipe material or fluid properties are unknown, the high-speed solenoid valve 5 is controlled by the positioning instrument host to emit multiple pressure pulses. The pressure wave velocity 'a' is then inverted using the wave velocity adaptive calibration algorithm. The algorithm formula is as follows: ; Where L is the known characteristic length of the pipe input through on-site measurement, and Δt and Δt are the time differences between the two reflected pressure pulse signals and the incident wave signal, respectively; S502. The Levenberg-Marquardt algorithm is used to fit the multiple reflection time series, and the pressure wave velocity 'a' and the distance S from the blockage point are simultaneously optimized and solved. The final distance from the blockage point is calculated using the formula S=(a×Δt) / 2. The rise time of the pressure pulse signal is controlled within 2ms, the pulse width is 1-10ms, and the peak pressure is 1.2-1.5 times the working pressure of the pipeline. In the wave velocity adaptive calibration, if the pipeline material is unknown, the initial wave velocity is preset to 800-1200m / s for metal pipelines or 300-500m / s for non-metal pipelines. S503. The characteristics of the reflected signal are identified by a machine learning classifier to distinguish between blockage reflections and non-blockage reflections caused by valves and elbows. The statistical distribution of Δt is generated by the Bootstrap resampling method and the distance result containing a 95% confidence interval is output on the host touch screen (9). The format is S±ΔS, where ΔS is the half-width of the confidence interval. In the test of a DN150 ductile iron pipe with a length of 255 meters, the wave velocity calibration value is 1062 m / s. The location result of the blockage point at 150.2 meters is 149.7 meters, with an error of 0.5 meters. The total time of a single test is less than 5 minutes.

[0040] Based on the amplitude ratio of the reflected pressure pulse signal 8 to the pressure pulse signal 7, the blockage point is determined to be either completely blocked or partially blocked.

Claims

1. A portable rapid pipe blockage location instrument based on pressure transient analysis, characterized in that: It includes the positioning device main unit (3), the docking connector (10), the high-speed solenoid valve (5) and the high-frequency pressure sensor (6); The docking connector (10) is a three-way connector. The positioning instrument host (3) is connected to the pipeline (1) to be tested through the docking connector (10). The docking connector (10) is equipped with a high-speed solenoid valve (5) and a high-frequency pressure sensor (6) that are connected to the positioning instrument host (3) via signal.

2. The portable pipeline blockage rapid location instrument based on pressure transient analysis as described in claim 1, characterized in that: The high-speed solenoid valve (5) and the high-frequency pressure sensor (6) are fixedly arranged in the docking joint (10). The end of the docking joint (10) is sealed and equipped with a quick connector (4). The connection lines of the high-speed solenoid valve (5) and the high-frequency pressure sensor (6) are connected to the quick connector (4) and connected to the positioning instrument host (3) through the quick connector (4).

3. The portable pipeline blockage rapid location instrument based on pressure transient analysis as described in claim 2, characterized in that: The total weight of the positioning device host (3) does not exceed 5kg, its battery module supports hot-swapping, and its continuous working time under full charge is not less than 4 hours.

4. A portable method for rapid location of pipeline blockages based on pressure transient analysis as described in any one of claims 1-3, characterized in that, Using the positioning device includes the following steps: Step 1: Connection. Connect the docking connector (10) to the pipe to be tested (1), connect the locator host (3) and the docking connector (10), and restore the signal connection between the locator host (3) and the docking connector (10). Step 2: Excitation. The positioning host (3) sends a signal and controls the high-speed solenoid valve (5) to open and close instantaneously, generating a short and steep pressure pulse signal (7). Step 3: Acquisition. The pressure pulse signal (7) propagates downstream along the pipeline to be tested (1). The high-frequency pressure sensor (6) synchronously and continuously records the pressure data at the moment of emission and the subsequent pressure changes, including the reflected pressure pulse signal (8) after encountering the blockage point (2). Step 4: Analysis. The collected pressure-time data is analyzed by the signal processing algorithm unit built into the positioning instrument host (3). Through cross-correlation analysis and wavelet transform algorithm, the feature points of the pressure pulse signal (7) and the reflected pressure pulse signal (8) are accurately identified, and the time difference Δt between the two is calculated. Step 5: Display. Based on the formula distance S=(wave velocity a×Δt) / 2, the distance to the blockage point (2) is automatically calculated, and the result is clearly displayed on the host touch screen (9) on the locator host (3).

5. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 4, characterized in that, In step four, the received data is preprocessed, including the following steps: S401. The original pressure signal is decomposed into multiple scales using the Daubechies wavelet basis, and the high-frequency coefficients are processed by soft thresholding to suppress environmental noise and high-frequency interference from the sensor. S402. Smooth the low-frequency signal to eliminate small fluctuations and retain the main characteristics of the pressure pulse; S403: Remove signal drift caused by temperature or pressure drift through polynomial fitting.

6. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 5, characterized in that, The specific decomposition steps of step S401 include: Let the original pressure signal be f(t), and use a low-pass filter h and a high-pass filter g to process the signal; In the j-th level decomposition, the signal f(t) is decomposed into approximate coefficients. and detail coefficient The approximation coefficient represents low-pass filtering, and the detail coefficient represents high-pass filtering. The specific calculation formula is as follows: Approximation coefficient: ; Detail factor: ; in, It is the first Approximation coefficients of the layer and These are the coefficients of the low-pass filter and the high-pass filter, respectively. This represents the discrete-time index. By iterating through the above process, the signal is decomposed into multiple scales, yielding approximation coefficients and detail coefficients at different scales. These coefficients correspond to the low-frequency and high-frequency components of the signal, respectively. The specific decomposition steps for soft thresholding of high-frequency coefficients include: The high-frequency coefficients after wavelet transform are processed by setting a threshold. The coefficients with absolute values ​​less than the threshold are set to zero, and the coefficients with absolute values ​​greater than the threshold are shrunk to suppress noise and interference. The mathematical expression for the soft threshold function is: ; in, These are the high-frequency coefficients after wavelet transform. It is the set threshold.

7. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 5, characterized in that, In step S402, the smoothing of the low-frequency signal uses an exponentially weighted moving average filtering method. The specific calculation formula is as follows: ； in It is the filtered output at the current moment. It is the input signal at the current moment. It is the filtered output from the previous moment. It is a smoothing coefficient, with a value ranging from 0.1 to 0.3; The larger the value, the greater the weight given to the new data, and the faster the filter responds to changes in the signal. The smaller the value, the higher the dependence on historical data, and the better the smoothing effect.

8. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 5, characterized in that, The specific steps for removing signal offset using polynomial fitting in step S403 are as follows: Assume the collected pressure signal is The corresponding time point is ( );choose Fitting is performed using a polynomial of order 1, the polynomial of which is in the form of: ； in, These are the polynomial coefficients to be determined; These coefficients are then determined using the least squares method to minimize the sum of squared errors between the fitted polynomial and the original signal; Sum of squared errors The calculation formula is: ； In order to Minimum, for about ( Find the partial derivative and set it equal to 0. Solving this system of linear equations will yield the polynomial coefficients. ; Obtain the fitting polynomial Then, remove the pressure signal after signal offset. for: .

9. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 4, characterized in that, In step four, when extracting pressure pulse features, the multi-resolution characteristics of wavelet transform in the time and frequency domain are used to detect abrupt changes in the pressure signal. Then, empirical mode decomposition (EMD) is performed on the signal to extract intrinsic mode functions. The instantaneous frequency is calculated through Hilbert transform to accurately locate the peak time. Finally, the standard incident wave template is compared with the measured signal to align the time series and eliminate the influence of propagation distortion.

10. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 9, characterized in that, The screening termination condition for intrinsic mode functions (IMFs) in the empirical mode decomposition (EMD) is that the standard deviation (SD) of two consecutive screening results is less than 0.001, where SD = Σ[(IMF{k-1}(t)-IMF_k(t))²] / Σ[IMF{k-1}(t)²]). When calculating the instantaneous frequency using the Hilbert transform, an analytic signal Z_i(t) = c_i(t) + j·H[c_i(t)] is constructed for each IMF component c_i(t), and the instantaneous frequency f_i(t) = (1 / 2π)·d[argZ_i(t)] / dt is given. The effective frequency range is limited to 0.1-2kHz to suppress noise, and the peak time is determined by the maximum point of the instantaneous amplitude a_i(t) = |Z_i(t)|.

11. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 4, characterized in that, In step four, the signal processing algorithm unit employs a joint time delay estimation strategy combining cross-correlation analysis and generalized cross-correlation when calculating the time difference Δt between the pressure pulse signal and the reflected pressure pulse signal. Specifically, this includes: S404. The pressure pulse signal (7) generated by the signal excitation sent by the positioning instrument host is used as the incident wave reference signal x(t), and the measured data including the reflected pressure pulse signal (8) collected by the high-frequency pressure sensor (6) is used as y(t). The initial time delay is obtained by cross-correlation calculation. The cross-correlation formula is: ； Where τ is the time delay, The τ value corresponding to the peak value is the initial time difference between the incident wave and the reflected wave; S405. The above cross-correlation results are optimized using PHAT weighting to enhance the robustness of time delay estimation in low signal-to-noise ratio environments. The generalized cross-correlation formula is as follows: ； Where X(f) and Y(f) are the Fourier transforms of x(t) and y(t) respectively, X(f) is the conjugate of X(f), and IFFT represents the inverse Fourier transform; S406. Combine the optimized generalized cross-correlation result with the matched filter output of the incident wave template on the measured signal, and accurately determine the final time difference Δt by finding the overlap position between the peak value of the matched filter output and the peak value of the generalized cross-correlation.

12. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 4, characterized in that, In step five, when calculating the distance to the blockage point and outputting the results, an enhancement strategy combining adaptive wave velocity calibration and result verification is adopted, specifically including: S501. If the pipe material or fluid properties are unknown, the high-speed solenoid valve (5) is controlled by the positioning instrument host to emit multiple pressure pulses, and the pressure wave velocity a is inverted using the wave velocity adaptive calibration algorithm. The algorithm formula is: ； Where L is the known characteristic length of the pipe input through on-site measurement, Δ and Δ These represent the time differences between the two reflected pressure pulse signals and the incident wave signal, respectively. S502. The Levenberg-Marquardt algorithm is used to fit the multiple reflection time series, and the pressure wave velocity a and the distance S of the blockage point are solved simultaneously and optimized. The final distance of the blockage point is calculated by combining the formula S=(a×Δt) / 2. S503. The characteristics of the reflected signal are identified by a machine learning classifier to distinguish between blockage reflection and non-blockage reflection caused by valves and elbows. The Bootstrap resampling method is used to generate the statistical distribution of Δt. The distance result containing a 95% confidence interval is output on the host touch screen (9) in the format S±ΔS, where ΔS is the half width of the confidence interval.

13. The portable method for rapid location of pipeline blockages based on pressure transient analysis as described in claim 4, characterized in that: It also includes determining whether the blockage point is completely blocked or partially blocked based on the amplitude ratio of the reflected pressure pulse signal (8) to the pressure pulse signal (7).

Images