Operation monitoring method for steam pipe network

Abstract

The invention provides an operation monitoring method for a steam pipe network. The operation monitoring method comprises the following steps: acquiring pipeline burial depth information, a leakage signal, temperature, pressure and soil thermal conductivity change information of an underground steam pipe network; constructing a pipe network space model, marking pipe fitting positions, obtaining soil types, and calculating attenuation coefficients of leakage signals in different soil media; simulating a leakage signal propagation path based on a pipe network space model, obtaining a complex area of a pipe network structure according to the pipe burial depth information, and if the leakage signal is superposed or interfered in the complex area, separating out an independent signal component to obtain an initial propagation direction of the leakage signal; according to the position of the leakage point, in combination with the pipeline burial depth and the soil type attribute, a regression analysis method is adopted to calculate the quantitative relation between the intensity of the leakage signal and the leakage degree, and an evaluation result of the leakage degree is obtained.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an operation monitoring method for a steam pipe network. Background Art

[0002] Underground pipeline networks in large chemical parks serve as the lifeblood of energy and material transportation, and their operational stability directly impacts production efficiency and safety control. However, pipeline leaks can not only waste resources but also lead to environmental pollution and even major accidents. Therefore, accurately locating leaks and assessing their extent are of paramount importance. Currently, methods based on single sensors or traditional signal analysis are gaining increasing attention in complex pipeline network scenarios, but their importance and urgent technical needs continue to grow. Existing methods often exhibit significant limitations when dealing with complex pipeline structures. Traditional technologies often rely on a single data source, such as acoustic or temperature signals, making them difficult to adapt to the complex structures and coexistence of multiple media pipelines in densely populated areas. Furthermore, these methods lack the ability to integrate multi-source data when dealing with interference factors such as vibration and uneven heat dissipation, resulting in reduced positioning accuracy and a failure to meet practical requirements. The core challenges in this field are concentrated in several key technical factors. First, the propagation path and attenuation of leakage signals in complex pipeline networks are limited by structural features such as elbows, tees, and reducers, which can easily distort signal characteristics. Secondly, external environmental interference, such as vibrations from adjacent pipelines, changes in heat dissipation caused by insulation damage, and changes in soil thermal conductivity due to groundwater level fluctuations, further complicate the analysis. These factors have not been adequately addressed, leading to significant deviations in leak location in densely populated pipe networks, and even inability to accurately determine the extent of the leak, presenting unique technical challenges. Therefore, leveraging multi-source heterogeneous data to comprehensively analyze the propagation and attenuation characteristics of leak signals in complex pipe networks, while fully considering the relationship between pipe depth, soil type, and the relative position of sensors, to accurately locate leaks and quantitatively assess the extent of the leak, has become a critical issue that urgently needs to be addressed. Summary of the Invention

[0003] The present invention provides a method for monitoring the operation of a steam pipe network, which mainly includes: Obtain information on buried depth, leakage signals, temperature, pressure, and soil thermal conductivity changes of underground steam pipe networks; Build a spatial model of the pipe network, mark the location of pipe fittings, obtain soil types, and calculate the attenuation coefficient of leakage signals in different soil media; The propagation path of leakage signals is simulated based on the spatial model of the pipe network. Based on the pipeline burial depth information, the complex areas of the pipe network structure are obtained. If leakage signals overlap or interfere with each other in the complex areas, the independent signal components are separated to obtain the initial propagation direction of the leakage signal. According to the initial propagation direction of the leakage signal and the attenuation coefficient of the leakage signal in different soil media, combined with the change information of soil thermal conductivity, calculate the energy loss of the leakage signal in different soil media, and identify the relative position relationship between the leakage point and the sensor according to the signal energy attenuation characteristics, so as to obtain the preliminary positioning range of the leakage point; Extract the features of the obtained leakage signal, temperature and pressure data to obtain multi-dimensional signal feature parameters, and analyze whether there is a correlation between the leakage signal, temperature and pressure according to the multi-dimensional signal feature parameters. If so, combine the simulation results of the propagation path to correct the preliminary positioning range of the leakage point to obtain the leakage point position; According to the position of the leakage point, combined with the pipeline burial depth and soil type attributes, use the regression analysis method to calculate the quantitative relationship between the intensity of the leakage signal and the leakage degree, and obtain the evaluation result of the leakage degree; According to the evaluation result of the leakage degree, judge whether there is pipeline vibration interference or insulation layer damage. If so, combine the soil characteristic information to correct the signal propagation path and attenuation coefficient, update the leakage point positioning and evaluation results, and based on the updated results, continuously collect sensor data within the preset range and compare it with the normal operation baseline to monitor the pipeline health status.

[0004] Further, obtain the pipeline burial depth information, leakage signal, temperature, pressure and soil thermal conductivity change information of the underground steam pipe network, including: collect the temperature data around the pipeline through the intelligent sensor array of the underground pipe network for regional geothermal distribution scanning, and match the corresponding soil type parameters from the soil thermal conductivity database according to the multi-point temperature numerical change gradient to obtain the real-time thermal conductivity data around the pipeline. For the pipe network pressure data collected by the intelligent sensor array, if the pipe section pressure value is lower than the preset normal operation pressure threshold of the pipe network, calculate the leakage volume of the pipe section according to the cross-validation of the pressure fluctuation curve and the temperature anomaly data. Establish a temperature field distribution map based on the temperature gradient data collected by the intelligent sensor array, and perform spatial registration on the temperature field distribution map and the three-dimensional structure diagram of the underground pipe network to obtain the initial value of the pipeline burial depth. Use the intelligent sensor array to form a high-density monitoring network around the pipeline leakage point, and calculate the accurate spatial coordinates of the leakage point through the acoustic positioning algorithm according to the time difference between the peak moment of the pressure fluctuation curve and the arrival time of the pressure fluctuation at each monitoring point. According to the soil temperature data collected by the intelligent sensor array and the thermal conductivity data obtained in the first step, establish a soil thermodynamics characteristic model, and verify the accuracy of the initial value of the pipeline burial depth through the recursive least squares method. For the pipe network pressure data and temperature field data, combined with the obtained leakage point coordinate information, use the radial basis function to quantitatively estimate the leakage scale, and output the leakage volume value and the leakage duration. Based on the multi-dimensional monitoring data collected by the intelligent sensor array, construct a pipe network health status evaluation model through the support vector machine algorithm, and compare it with the preset pipe network parameter threshold to judge the pipeline operation status level.

[0005] Furthermore, a pipeline network space model is constructed, the positions of pipe fittings are marked, the soil type is obtained, and the attenuation coefficients of leakage signals in different soil media are calculated, including: collecting the spatial point cloud data of pipe segments through an underground pipeline network three-dimensional scanning device, performing spatial clustering on the point cloud data according to the point cloud density distribution characteristics, and generating an initial pipeline network space topology structure diagram. For the pipe segment connection nodes in the initial pipeline network space topology structure diagram, a density clustering method is used to extract the node feature vectors, and the spatial position coordinates of elbows, tees, and reduced-diameter pipe fittings are identified by matching with the pipe fitting shape feature library. According to the spatial position coordinates of the pipe fittings, the sampling area is calibrated, the soil conductivity values at different depths are measured by a soil resistivity sensor, and the soil layer boundary point set is obtained through the variation law of the conductivity values. Based on the soil layer boundary point set, a three-dimensional soil medium distribution map is constructed, and the acoustic parameters of the corresponding soil type are extracted from the soil parameter database to obtain the acoustic characteristic distribution map along the pipeline network. The acoustic response signal of the pipeline network is obtained through the pipeline network three-dimensional scanning device, and the signal propagation path is traced in combination with the pipeline network space topology structure diagram, and the sound intensity change curve during the signal propagation process is recorded. For the sound intensity change curve, different soil medium propagation sections are divided according to the acoustic characteristic distribution map, and the acoustic attenuation coefficients of each section are calculated by the least squares method. Based on the acoustic attenuation coefficients and the soil medium distribution map, a Markov chain algorithm is used to construct an acoustic signal propagation prediction model, and the propagation attenuation values of leakage signals in different soil media are calculated.

[0006] Furthermore, based on the pipeline network space model, the propagation path of the leakage signal is simulated, and at the same time, according to the pipeline burial depth information, the complex areas of the pipeline network structure are obtained. If the leakage signal appears superimposed or interfered in the complex area, the independent signal components are separated to obtain the initial propagation direction of the leakage signal, including: collecting the three-dimensional spatial data of the pipeline through a pipeline network detection device, calculating the pipeline node density value according to the pipe segment connection quantity and the burial depth distribution data, and marking the position coordinates of the complex areas of the pipeline network according to the change of the node density value and the burial depth gradient. The distributed sensor array in the pipeline network detection device is used to collect the acoustic response signal of the pipeline network, and the signal intensity attenuation curve at the sensor collection position is calculated according to the signal spectrum characteristics to judge the propagation loss value of the signal in the complex area. For the sensor nodes in the complex area of the pipeline network that exceed the preset attenuation threshold, a multi-channel signal collector is used to record the aliased waveform, and the waveform feature vector group is separated by an independent component extraction algorithm. According to the waveform feature vector group, the signal propagation delay between adjacent sensor nodes is calculated, and an adaptive Kalman filter is used to attenuate the waveform noise to obtain the reference signal waveform. Based on the reference signal waveform and the sensor node spatial distribution map, the signal amplitude change trend in the complex area is calculated, and a signal propagation path topology map is constructed by an acoustic wave positioning algorithm. Combining the signal propagation path topology map with the position coordinates of the complex areas of the pipeline network, a spatial mapping of the propagation path is performed, and the beam tracking method is used to determine the initial propagation direction of the leakage signal.

[0007] Furthermore, according to the initial propagation direction of the leakage signal and the attenuation coefficients of the leakage signal in different soil media, combined with the change information of soil thermal conductivity, calculate the energy loss of the leakage signal in different soil media, and identify the relative position relationship between the leakage point and the sensor according to the signal energy attenuation characteristics, so as to obtain the preliminary positioning range of the leakage point, including: collecting the amplitude data of the leakage signal through a distributed acoustic sensor array, constructing a sound wave propagation path vector diagram according to the spatial distribution coordinates of the sensor nodes, and obtaining the soil stratification structure diagram from the thermal conductivity monitoring points. For the soil stratification structure diagram, use a sound wave attenuation detector to measure the propagation loss values of sound waves in each soil layer, and generate a sound wave propagation loss distribution diagram through a sound wave intensity attenuation calculator. Based on the sound wave propagation loss distribution diagram, combined with the sound wave propagation path vector diagram, calculate the sound wave amplitude difference between adjacent sensor nodes, and obtain the sound wave attenuation curve per unit distance through distance normalization processing. According to the sound wave attenuation curve per unit distance, use a sound wave energy attenuation model to calculate the energy propagation loss of sound waves in soil media, and generate a sound wave energy spatial distribution diagram through an energy attenuation calibrator. Perform grid meshing on the sound wave energy spatial distribution diagram, use an energy gradient calculator to determine the direction with the fastest sound wave energy attenuation, and generate a leakage signal propagation trajectory diagram in combination with the sensor node coordinates. Based on the leakage signal propagation trajectory diagram, establish a distance attenuation prediction model through a support vector regression calculator, and determine the leakage point candidate area in combination with the sound wave energy threshold criterion. Use a sound source locator to perform spatial scanning on the leakage point candidate area, calculate the spatial position of the sound source according to the time difference of arrival of sound waves at multiple points, and obtain the accurate coordinates of the leakage point through trilateration.

[0008] Furthermore, feature extraction is performed on the acquired leakage signal, temperature, and pressure data to obtain multidimensional signal characteristic parameters. Correlations between the leakage signal, temperature, and pressure are analyzed based on the multidimensional signal characteristic parameters. If so, the initial location range of the leak point is revised based on the propagation path simulation results to determine the leak point location. This process involves collecting the leakage signal waveform, temperature distribution data, and pressure change data through an underground pipeline network monitoring array, extracting the waveform frequency eigenvalues using a wavelet decomposer, calculating the temperature change rate using a temperature gradient operator, and recording the pressure pulse values using a pressure collector. A spectrum distribution curve is generated based on the waveform frequency eigenvalues, and the coordinates of the spectrum peak points are extracted using a time-frequency analyzer. Characteristic frequency intervals are marked using an adaptive threshold splitter. A feature correlation calculator is used to generate a parameter correlation matrix for the temperature change rate and pressure pulse values within the characteristic frequency intervals, and the strength of the parameter correlation is determined using the Pearson correlation coefficient. Based on the parameter correlation matrix, a Kalman filter is used to fuse the leakage signal waveform, temperature change rate, and pressure pulse values to obtain a multidimensional feature vector set. A leakage feature space is constructed based on the multidimensional feature vector group. Cluster analysis of the feature space is performed using a support vector machine to generate a probability distribution map of the leakage point location. Spatial overlap is calculated based on the leakage point location probability distribution map, combined with the propagation path prediction results. Corrected leak point coordinates are determined using a three-dimensional coordinate mapper. A spatial mesher is used to finely partition the area surrounding the corrected leak point coordinates. The optimal leak point coordinates are then output using a multi-sensor data fusion algorithm.

[0009] Furthermore, according to the location of the leakage point, in combination with the pipeline burial depth and soil type attributes, a regression analysis method is used to calculate the quantitative relationship between the intensity of the leakage signal and the leakage degree, and the evaluation result of the leakage degree is obtained, including: collecting leakage acoustic signal data through the underground pipe network monitoring device, obtaining the depth value of the leakage point according to the pipeline burial depth distribution map, and extracting the acoustic wave propagation attenuation coefficient from the soil parameter database. For the acoustic wave propagation attenuation coefficient and the burial depth data, an acoustic wave attenuation compensator is used to calculate the distance attenuation value, and the medium absorption loss value is obtained in combination with the soil acoustic impedance parameter. Based on the distance attenuation value and the medium absorption loss value, the original sound pressure level value at the leakage point is obtained through an acoustic energy calculator, and the initial value of the leakage aperture is determined by using the sound pressure level calibration curve. According to the initial value of the leakage aperture, in combination with the pipe network pressure data, a Bernoulli flow calculator is used to generate an estimated leakage volume value, and the leakage volume value is adjusted by a hydrodynamic correction coefficient. For the leakage volume value and the sound pressure level data, a support vector regressor is used to establish the mapping relationship between the sound pressure level and the leakage volume, and the leakage volume classification threshold is obtained from the historical leakage database. Based on the leakage volume classification threshold, a Gaussian process regressor is used to construct a leakage degree prediction curve, and the leakage duration is calculated in combination with the pipe network pressure fluctuation characteristics. A multivariate regression analyzer is used to comprehensively calculate the leakage volume value, the duration, and the pressure fluctuation characteristics, and the leakage level determination result is output according to the preset leakage degree evaluation standard.

[0010] Further, based on the leakage degree assessment result, it is judged whether there is pipeline vibration interference or insulation layer damage. If so, in combination with the soil property information, the signal propagation path and attenuation coefficient are corrected, the leakage point positioning and assessment result are updated. Based on the updated result, the sensor data within a preset range is continuously collected and compared with the normal operation baseline to monitor the pipeline health status, including: collecting the pipeline vibration signal through a pipeline network vibration monitor, obtaining the surface temperature distribution data of the insulation layer in combination with a temperature monitor, and calculating the vibration characteristic frequency value according to a vibration spectrum analyzer. Based on the vibration characteristic frequency value, for the frequency interval exceeding the preset threshold, a vibration compensation calculator is used to generate an interference cancellation function, and a temperature field calculator is used to construct a temperature distribution map around the pipeline. According to the temperature distribution map and the vibration interference cancellation function, in combination with the soil medium acoustic parameter database, an acoustic path tracker is used to recalculate the leakage signal propagation path. For the leakage signal propagation path, an attenuation coefficient corrector is used to update the acoustic wave propagation loss value, and an acoustic wave propagation trajectory map is generated based on a medium boundary reflection calculator. An adaptive filter is used to suppress the noise of the acoustic wave propagation trajectory map, and the leakage point coordinate value is recalculated in combination with a spatial positioning algorithm. According to the leakage point coordinate value, a monitoring range is delimited, the pipeline operation state parameters are obtained through a multi-parameter data collector, and a parameter change trend map is constructed by using a recurrent neural network. Based on the parameter change trend map, in combination with the preset pipeline operation parameter baseline, a state evaluator is used to calculate the parameter deviation degree to judge the pipeline health state level.

[0011] The technical solution provided by the embodiment of the present invention may include the following beneficial effects: The present invention discloses an operation monitoring method for a steam pipeline network. By setting sensors to obtain information such as pipeline burial depth, leakage signal, temperature, pressure, etc., a pipeline network spatial model is constructed and the leakage signal propagation path is simulated. In combination with soil properties, the attenuation of the signal in different media is calculated to identify the relative position between the leakage point and the sensor, and the leakage range is initially located. Further, multi-dimensional signal features are extracted, the correlation is analyzed and the positioning result is corrected. Regression analysis is used to evaluate the leakage degree, and corrections are made considering factors such as pipeline vibration and insulation layer damage. Finally, the data is continuously monitored and compared with the normal operation baseline to realize the real-time monitoring of the pipeline health status. The present invention can accurately locate the leakage point of the underground steam pipeline network, evaluate the leakage degree, provide a basis for timely repair and preventive maintenance, and improve the operation safety and reliability of the pipeline network. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a flowchart of an operation monitoring method for a steam pipeline network of the present invention.

[0013] Figure 2 It is a schematic diagram of an operation monitoring method for a steam pipeline network of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0014] To make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0015] As Figure 1 , a method for monitoring the operation of a steam pipe network in this embodiment may specifically include: S101. Obtain the burial depth information, leakage signals, temperature field data, pressure changes, and soil thermal conductivity parameters of the underground steam pipe network by arranging a sensor network, and preliminarily calculate the spatial position and operating state of the leakage point based on the multi-dimensional data.

[0016] In the embodiment of the present invention, the sensor network is arranged in a distributed manner along the pipe network to form a monitoring area covering the pipeline, and key operating parameters are collected in real time.

[0017] S1011. Collect temperature data around the pipeline through the sensor network, match the soil thermal conductivity parameters using the temperature change gradient, and determine the burial depth information in combination with the three-dimensional model of the pipe network. At the same time, calculate the preliminary coordinates of the leakage point based on the time difference of pressure fluctuations.

[0018] In the embodiment of the present invention, a temperature sensor node is arranged every 50 meters to form a grid-shaped monitoring network. The sampling frequency is set to once every 10 minutes to ensure timely perception of temperature field changes. The collected temperature data is compared with a pre-set soil thermal conductivity database through gradient analysis. For example, the thermal conductivity of clay is about 0.15 watts per meter kelvin, and that of sandy soil is 0.58 watts per meter kelvin, to obtain the real-time thermal conductivity parameters. Subsequently, the temperature field distribution is spatially registered with the three-dimensional structure diagram of the pipe network to calculate the pipeline burial depth value, which is usually between 1.2 and 2 meters, and the accuracy can reach the centimeter level. At the same time, the pressure sensor monitors the operating pressure of the pipe network, and the normal range is 0.6 to 0.8 MPa. If it is lower than 0.5 MPa, an abnormal detection is triggered. Using the time difference when the pressure fluctuation peak reaches each monitoring point, combined with the acoustic wave propagation speed of about 1200 meters per second, calculate the preliminary spatial coordinates of the leakage point through the acoustic wave positioning algorithm.

[0019] S1012. Establish a distribution model for the collected temperature and pressure data, verify the accuracy of the burial depth information in combination with the soil thermal conductivity, quantify the leakage scale through radial basis functions, and construct a health status assessment model to analyze the operating level of the pipe network.

[0020] In the embodiments of the present invention, a temperature field distribution map in the form of isotherms is generated based on temperature data. Near the leakage point, it shows a significant temperature anomaly area, and the gradient is 3 to 5 times higher than that in the normal area. Combining with the thermal conductivity parameters obtained in the first step, a soil thermodynamics characteristic model is established, and the reliability of the burial depth data is verified by the recursive least squares method. If the pipe section pressure is lower than the threshold, the pressure fluctuation curve and temperature anomaly data are used for cross-verification to calculate the leakage amount of this section. For the assessment of the leakage scale, the radial basis function is used to analyze the pressure and temperature data, and the leakage amount and duration are output. In addition, the multi-dimensional monitoring data is integrated by the support vector machine algorithm to construct a pipeline network health status assessment model. By comparing with the preset parameter thresholds, the operating status is divided into three levels: normal, warning, and abnormal, and the assessment accuracy rate can reach more than 95%.

[0021] In the embodiments of the present invention, the sensor network forms a self-organizing structure through wireless communication. The temperature sampling is automatically increased to once per second when a leakage occurs to capture the dynamic changes. During normal operation, the temperature on the outer wall of the pipeline is about 60 degrees Celsius, and the soil temperature decreases with depth. At a depth of 1 meter, it is about 15 to 20 degrees Celsius. When a leakage occurs, the local soil temperature rises rapidly, accompanied by pressure fluctuation characteristics, providing a reliable data basis for subsequent analysis.

[0022] It can be understood that the embodiments of the present invention do not overly limit the specific settings of the sensor layout density and sampling frequency, which can be adjusted by technicians according to the actual scenario.

[0023] In the embodiments of the present invention, through the above steps, multi-source data can be effectively integrated, overcoming the limitations of traditional methods in complex pipeline networks, realizing high-precision positioning of leakage points and real-time monitoring of operating status, and providing technical guarantee for the safe operation of steam pipeline networks.

[0024] In the subsequent steps, based on further analysis of signal propagation and attenuation characteristics, the positioning results will be optimized and the evaluation process will be improved. The specific content will be detailed in the subsequent embodiments and will not be elaborated here.

[0025] S102. Construct a pipeline network space model and mark the positions of pipe fittings, and at the same time collect soil characteristic data to calculate the attenuation coefficient of leakage signals in different media.

[0026] In the embodiments of the present invention, through advanced space scanning technology and soil parameter measurement means, a three-dimensional structure of the pipeline network is generated and the signal propagation characteristics are analyzed.

[0027] Such as Figure 2 , S1021. Collect the point cloud data of the pipe section through the three-dimensional scanning equipment for underground pipeline networks, use the point cloud density distribution characteristics for spatial clustering to generate an initial pipeline network topology structure diagram, and use the density clustering method combined with the pipe fitting feature library to identify and mark the position coordinates of elbows, tees, and reduced-diameter pipe fittings.

[0028] In an embodiment of the present invention, an array ultrasonic probe is selected as the pipeline network three-dimensional scanning device. Detection points are set every 5 meters along the pipeline. The emission frequency of the probe is set to 200 kHz, and the receiving sensitivity reaches -80 dB. Each detection point stays for 10 seconds to complete a full-circle scan, with a sampling interval of 0.1 degree, and the point cloud data of the pipeline cross-section contour is collected. The point cloud data is spatially clustered through density distribution analysis. When there are obvious mutations in the cross-section contour, it indicates the existence of pipe fittings. For these mutation regions, a density clustering algorithm is used to extract feature vectors and match them with a pre-set pipe fitting shape feature library, which contains standard feature data of elbows, tees, and reducing pipe fittings. After the matching is completed, the type of pipe fitting is determined and its spatial coordinates are output, realizing the accurate marking of key nodes of the pipeline network and providing a spatial reference for subsequent signal analysis.

[0029] S1022. Divide the sampling area according to the position coordinates of the pipe fittings, use a soil resistivity sensor to measure the conductivity values at different depths, construct a three-dimensional soil medium distribution map, and calculate the attenuation coefficient of the acoustic signal in combination with the soil type parameters.

[0030] In an embodiment of the present invention, soil resistivity sensors are arranged near the calibrated positions of the pipe fittings. The Wenner four-electrode array method is used to measure the soil conductivity, and the measurement range is from the ground surface to the bottom of the pipeline, with samples taken every 20 cm. Different soil types show significant conductivity differences. For example, the conductivity of the clay layer is about 50 to 80 mS / m, that of the sand layer is 20 to 40 mS / m, and that of the pebble layer is less than 10 mS / m. By the vertical variation law of the conductivity, the soil layer boundaries are identified and a three-dimensional soil medium distribution map is generated. The acoustic properties of the corresponding soil types are extracted from the soil parameter database. For example, the sound wave velocity in clay is 1600 m / s and the attenuation coefficient is 0.8 dB / m, and the sound wave velocity in sand is 1800 m / s and the attenuation coefficient is 0.5 dB / m. Based on these parameters, combined with the propagation path of the pipeline network acoustic response signal, the attenuation coefficient in each soil medium is calculated to provide data support for signal attenuation analysis.

[0031] S1023. Trace the propagation path of the acoustic response signal through the pipeline network three-dimensional scanning device, record the sound intensity change curve, divide the propagation section in combination with the soil medium distribution map, and use the least square method and Markov chain algorithm to construct an acoustic signal propagation prediction model to deduce the attenuation characteristics of the leakage signal.

[0032] In the embodiments of the present invention, the acoustic response signal of the pipeline network is generated by an ultrasonic probe exciting through the pipeline wall. The signal frequency is consistent with the emission frequency and attenuates during propagation under the influence of the soil medium. The change of sound intensity with distance is recorded at multiple detection points to form a sound intensity change curve. In areas with significant soil stratification, the curve exhibits a stepped characteristic, and each step corresponds to the attenuation effect of different media. For this curve, the propagation sections are divided in combination with the soil medium distribution map, and the attenuation coefficients of each section are fitted using the least squares method to obtain a quantitative result. Subsequently, a Markov chain algorithm is used to construct an acoustic signal propagation prediction model. This algorithm describes the propagation law of the signal between different media through a state transition matrix. The model training is based on historical sound intensity data, and the prediction accuracy can reach more than 90%. It can effectively calculate the attenuation value of the leakage signal in complex media and provide a reliable basis for leakage point positioning.

[0033] In the embodiments of the present invention, the construction of the pipeline network spatial model makes full use of the high-precision characteristics of the point cloud data and realizes a comprehensive characterization of the underground environment in combination with soil conductivity measurement. The introduction of the acoustic signal propagation prediction model not only improves the calculation accuracy of the attenuation coefficient but also provides theoretical support for analyzing the propagation characteristics of the leakage signal.

[0034] It can be understood that the present invention does not impose excessive limitations on the specific frequency or sampling interval of the scanning device probe, which can be flexibly adjusted by technicians according to the pipeline network scale and environmental conditions.

[0035] In the embodiments of the present invention, the model and attenuation coefficient generated through the above steps lay a foundation for the preliminary positioning and further optimization of subsequent leakage points and can effectively cope with the challenges brought by the complexity of the pipeline network structure and the diversity of soil media.

[0036] Subsequent steps will further analyze the signal propagation direction and energy loss based on this model, gradually improving the leakage detection process, and the specific implementation will be elaborated in the subsequent description.

[0037] S103. Simulate the propagation path of the leakage signal based on the pipeline network spatial model, identify complex areas in combination with the pipeline burial depth information, and separate interference signals through a filtering algorithm to determine the initial propagation direction of the leakage signal.

[0038] In the embodiments of the present invention, the spatial characteristics of the pipeline network and the multi-dimensional data of the sensor array are used to analyze the signal propagation behavior and optimize the detection results.

[0039] S1031. Collect three-dimensional spatial data of the pipeline through a pipeline network detection device, calculate the node density and burial depth distribution characteristics to mark the position coordinates of complex areas, and use a distributed sensor array to obtain an acoustic response signal to generate an intensity attenuation curve.

[0040] In the embodiment of the present invention, the pipeline detection device adopts a subway-style layout method, with a row of sensors set every 5 meters along both sides of the pipeline, and the detection depth covers from the ground surface to 5 meters. The sensor array selects piezoelectric acoustic sensors with a sensitivity of -90 dB and a sampling frequency set at 1000 Hz, which can efficiently capture acoustic signals. By collecting the three-dimensional spatial data of the pipeline, the number of pipe fittings in the unit space is calculated to obtain the node density value. When the density exceeds 0.5 per cubic meter, this area is marked as a complex area. At the same time, according to the gradient change of the buried depth distribution data, the spatial range of the complex area is further confirmed. The distributed sensor array monitors the acoustic response signal in real time. Based on the signal spectrum characteristics, multiple peaks are extracted in the range of 200 to 800 Hz to generate a signal intensity attenuation curve. The results show that the signal loss in the complex area is 3 to 5 times higher than that in the normal area, and the typical attenuation value can reach 2 dB per meter, laying a foundation for analysis.

[0041] S1032. For the nodes in the complex area where the signal intensity attenuation curve exceeds the preset threshold, use a multi-channel signal collector to record the aliased waveform, separate the feature vector group through an independent component extraction algorithm, and calculate the propagation delay in combination with the sensor position to optimize the signal quality.

[0042] In the embodiment of the present invention, when the signal attenuation value exceeds the preset threshold of 6 dB per meter, the multi-channel signal collector is triggered to work, and at the same time, the waveform data of 8 adjacent sensor nodes are recorded. The collected aliased waveform shows obvious superposition characteristics. By analyzing the correlation between waveforms through an independent component extraction algorithm, it is decomposed into multiple independent waveform components, and each component corresponds to a single propagation path. This algorithm extracts the feature vector group through matrix decomposition technology to reflect the propagation characteristics of sound waves in different directions. Combining the spatial distribution of the sensor nodes, calculate the signal propagation delay between adjacent nodes. The speed of sound waves in normal soil is about 1500 m / s, while it can be increased to 2500 m / s in the pipeline acoustic waveguide. Using these delay data, verify the timing consistency of signal propagation, thereby providing reliable input for filtering processing.

[0043] S1033. Use an adaptive Kalman filter to attenuate the noise of the aliased waveform, generate a reference signal waveform, construct a propagation path topology map based on the signal amplitude change trend and the acoustic wave positioning algorithm, and determine the initial direction of the leakage signal through the beam tracking method.

[0044] In the embodiment of the present invention, the adaptive Kalman filter performs noise reduction processing on the separated waveform according to the real-time updated noise covariance matrix, and the signal-to-noise ratio is increased by about 15 decibels to generate a clear reference signal waveform. Based on this waveform, combined with the spatial coordinates of the sensor nodes, a signal amplitude change trend graph is drawn, and it is found that the amplitude increases significantly near the leakage point, 6 to 8 decibels higher than the surrounding area, showing the characteristic of energy aggregation. The acoustic wave positioning algorithm is adopted to calculate the sound source position by using the time difference of arrival of signals from more than 4 nodes based on the principle of triangulation, and the positioning accuracy can reach the centimeter level. Subsequently, a propagation path topology graph is constructed to intuitively display the propagation trajectory of the acoustic wave in the pipe network. Further combined with the beam tracing method, by simulating the propagation path of the acoustic ray in the complex medium, the initial direction of the signal is traced, and the accuracy can reach 10 centimeters. This process makes full use of the pipe network space model and the buried depth information to effectively cope with the signal interference problem in complex areas.

[0045] In the embodiment of the present invention, the above steps solve the problem of superposition and attenuation of leakage signals in complex areas through multi-level signal processing, ensuring the accuracy of propagation direction identification.

[0046] It can be understood that the present invention does not strictly limit the layout spacing or sampling frequency of the sensor array, and can be flexibly adjusted according to the scale of the pipe network to meet the actual needs.

[0047] In the embodiment of the present invention, the combination of the propagation path simulation and the filtering technology provides key data support for the subsequent precise positioning of the leakage point, and the subsequent steps will further optimize the results.

[0048] S104. According to the initial propagation direction of the leakage signal and the attenuation coefficient in different soil media, calculate the energy loss in combination with the change of soil thermal conductivity, and identify the relative position between the leakage point and the sensor based on the signal attenuation characteristics to determine the preliminary positioning range.

[0049] In the embodiment of the present invention, the distributed acoustic sensor array is arranged around the pipeline in a grid pattern, with a node set every 10 meters, the sampling frequency is 1000 Hz, and the amplitude measurement range covers 0 to 120 dB, ensuring the comprehensiveness and high precision of signal acquisition. The acoustic wave propagation path vector graph is constructed by connecting adjacent sensor nodes. The direction of the node connection indicates the acoustic wave propagation trend, and the connection length reflects the propagation distance, providing a spatial basis for subsequent analysis. At the same time, thermal conductivity monitoring points are set near the sensors, and the monitoring depth ranges from the ground surface to the bottom of the pipeline. The thermal conductivity data is recorded every 0.5 meters to generate a soil stratification structure diagram, clearly showing the boundaries and characteristic differences of the surface cultivated soil, clay layer, sand layer, etc., providing environmental parameter support for acoustic wave attenuation calculation.

[0050] S1041. Measure the propagation loss values in each soil layer using a sound wave attenuation detector, generate a sound wave propagation loss distribution map, calculate the energy propagation loss through a sound wave energy attenuation model, and construct a sound wave energy spatial distribution map.

[0051] In an embodiment of the present invention, the sound wave attenuation detector measures the remaining intensity of a standard sound source signal in different soil layers and calculates the attenuation value per unit distance. For example, the attenuation of sound waves in the surface cultivated soil is about 0.8 decibels per meter, in the clay layer is 1.2 decibels per meter, and in the sandy soil layer is 0.6 decibels per meter. These data reflect the absorption effect of soil types on sound wave propagation. Based on the collected leakage signal amplitude data and vector diagram, calculate the sound wave amplitude difference between adjacent nodes, eliminate the influence of propagation distance through distance normalization processing, and generate a sound wave attenuation curve per unit distance. Combining this curve with the sound wave energy attenuation model, comprehensively considering geometric diffusion loss and medium absorption loss, simulate the logarithmic attenuation trend of sound wave energy with distance. For example, the sound pressure level is 90 decibels at a distance of 1 meter from the sound source and gradually weakens as the distance increases. Finally, generate a sound wave propagation loss distribution map through a sound wave intensity attenuation calculator, calculate the energy propagation loss in each soil layer based on this, and draw a sound wave energy spatial distribution map to visually present the distribution law of energy in space.

[0052] S1042. Mesh the sound wave energy spatial distribution map, establish a distance attenuation prediction model using support vector regression, and determine the precise coordinates of the leakage point by combining energy gradient and sound source localization technology.

[0053] In an embodiment of the present invention, the sound wave energy spatial distribution map is meshed into grids of 1 meter × 1 meter, and each grid records the corresponding sound wave energy value. Energy gradient calculation reveals the energy aggregation characteristics around the leakage point. The energy value is 15 to 20 decibels higher than the surrounding area, and the propagation trajectory is radial and refracts at the soil layer boundary. The refraction angle is closely related to the acoustic impedance difference. The support vector regression calculator is trained using historical leakage data to establish a non-linear mapping relationship between distance and energy attenuation, and the prediction accuracy can reach 90%, thus generating a distance attenuation prediction model. Combining with the sound wave energy threshold criterion, screen out the candidate areas of the leakage point. Subsequently, the sound source locator conducts a spatial scan of the candidate areas, uses the time difference of arrival of sound waves recorded by at least 3 sensor nodes, matches the signals through the cross-correlation algorithm, calculates the time difference accuracy up to the microsecond level, and uses the trilateration method to determine the coordinates of the leakage point, with a positioning accuracy of up to 0.5 meters. This method effectively captures the influence of soil moisture content changes on sound wave propagation. For example, when the moisture content increases by 10%, the attenuation value increases by about 0.3 decibels per meter, ensuring the adaptability and reliability of the positioning results.

[0054] In the embodiments of the present invention, through the above steps, by synthesizing the signal attenuation characteristics and the change of soil thermal conductivity, not only the preliminary positioning of the leakage point is achieved, but also a solid data foundation and analysis framework are provided for precise positioning and evaluation.

[0055] It can be understood that the present invention does not have fixed requirements for the sensor layout density or the grid division size, and can be flexibly adjusted according to the actual pipeline network environment to optimize the performance.

[0056] S105. Extract the feature values of the leakage signal, temperature and pressure data and analyze their correlations, and combine the propagation path simulation results to correct the preliminary positioning range of the leakage point and determine the position of the leakage point.

[0057] In the embodiments of the present invention, the underground pipeline network monitoring array adopts a distributed layout method. A group of sensor units are installed every 20 meters along the pipeline. Each group includes an acoustic sensor, a temperature sensor and a pressure sensor, which respectively collect the leakage signal waveform, temperature distribution and pressure change data to ensure the comprehensiveness of multi-source information. The waveform data captured by the acoustic sensor usually presents characteristic frequencies in the range of 100 to 500 Hz. The main frequency is related to the leakage aperture. For example, an aperture of 1 mm corresponds to about 400 Hz. The temperature sensor monitors the temperature field around the pipeline. The change rate is less than 0.1 degree Celsius per meter during normal operation, and can locally surge to 1 degree Celsius per meter during leakage. The pressure sensor records the pulse value at a sampling frequency of 100 Hz and automatically triggers data collection when the fluctuation amplitude exceeds 0.1 MPa, providing high-timeliness data support for analysis.

[0058] S1051. Extract the frequency feature values of the leakage signal through wavelet decomposition and generate a spectrum distribution curve, use adaptive threshold segmentation to mark the characteristic frequency interval, and calculate the temperature change rate and pressure pulse value at the same time to construct a parameter correlation matrix.

[0059] In the embodiments of the present invention, the leakage signal waveform is processed by a wavelet decomposer, and an 8-layer decomposition structure is adopted to gradually separate the energy distribution of different frequency bands and extract the waveform frequency feature values. Based on these feature values, a spectrum distribution curve is generated using the fast Fourier transform. When a leakage occurs, a significant peak appears in the curve, and its position is related to the aperture size. To accurately mark the characteristic frequency interval, an adaptive threshold segmenter is used to divide the spectrum curve into a background area and a characteristic area through the maximum inter-class variance method, highlighting the main frequency band of the leakage signal. At the same time, the temperature gradient operator calculates the temperature change rate based on the central difference method, reflecting the dynamic trend of local thermal anomalies. The pressure pulse value is recorded in real time by the pressure collector, capturing the instantaneous fluctuation characteristics of the pressure in the pipeline. For the temperature change rate and pressure pulse value within the characteristic frequency interval, the feature correlation calculator calculates the correlation between parameters with a 60-second sliding window and generates a parameter correlation matrix through the Pearson correlation coefficient. When the absolute value of the coefficient is greater than 0.8, it is regarded as a strong correlation, indicating that there is a significant co-variation between the parameters, providing a basis for fusion.

[0060] S1052. The Kalman filter is used to fuse the leakage signal waveform, the temperature change rate, and the pressure pulse value to generate a multi-dimensional feature vector group. Through support vector machine clustering analysis and combined with the propagation path prediction result, the corrected coordinates of the leakage point are determined.

[0061] In the embodiment of the present invention, the Kalman filter fuses data with a 6-dimensional state vector, including signal amplitude, frequency, temperature, temperature gradient, pressure, and pressure change rate. The measurement noise variance is set to 0.01 to ensure the stability of the fusion result. Through filtering processing, a multi-dimensional feature vector group is generated, and each group of samples contains 6 components to construct a leakage feature space. The support vector machine uses a radial basis kernel function to cluster the feature space. The clustering center represents the potential leakage position, generating an elliptical probability distribution map with its major axis consistent with the pipeline direction, and the probability value exponentially decays from the center to the outside. Combining the previous propagation path simulation results, the spatial overlap area between the probability distribution map and the path prediction is calculated, and the point with the maximum probability in the overlap area is taken as the preliminary corrected coordinate. For further refined positioning, the spatial grid divider divides the area around the corrected coordinate. The grid resolution near the leakage point is encrypted to 0.1 m, and gradually thinned to 1 m at the periphery, and the exact position is output through a three-dimensional coordinate mapper.

[0062] In the embodiment of the present invention, the multi-sensor data fusion adopts the Bayesian estimation method, assigns weights according to the data reliability, the weight of the acoustic data is 0.5, the weight of the temperature data is 0.3, and the weight of the pressure data is 0.2. The optimal leakage point coordinates are calculated through weighted average to improve the positioning robustness and accuracy.

[0063] It can be understood that the present invention does not make rigid regulations on the specific spacing or decomposition layers of the sensor unit, and can be flexibly adjusted according to the complexity of the pipe network to optimize the effect.

[0064] In the embodiment of the present invention, through the multi-step collaboration of feature extraction and data fusion, the accuracy of leakage point positioning is significantly improved, and its applicability in complex environments can be further verified subsequently.

[0065] S106. According to the leakage point position, combined with the pipe burial depth and soil type, the regression analysis method is used to calculate the relationship between the leakage signal intensity and the leakage degree, and the leakage degree evaluation result is generated.

[0066] In the embodiment of the present invention, an acoustic sensor array is arranged around the leakage point by the underground pipeline network monitoring device. The sensitivity of the sensor reaches -90 dB, and the sampling frequency is 1000 Hz. The collected acoustic signals contain frequency and amplitude information, providing high-precision data for subsequent analysis. The depth of the leakage point is determined by the pipeline burial depth distribution map, usually in the range of 1.2 to 2.5 meters. Combining with the soil parameter database, the acoustic wave propagation attenuation coefficient corresponding to the depth is extracted. For example, the attenuation coefficient of sandy soil is 0.5 dB per meter, that of clay is 0.8 dB per meter, and that of gravel layer is 0.3 dB per meter. These parameters reflect the propagation characteristics of acoustic waves in different media and lay the foundation for attenuation calculation.

[0067] S1061. Calculate the distance attenuation value through the acoustic wave attenuation compensator and obtain the medium absorption loss value by combining with the soil acoustic impedance parameters. Use the acoustic energy calculator to deduce the original sound pressure level at the leakage point and determine the initial value of the leakage aperture.

[0068] In the embodiment of the present invention, the acoustic wave propagation attenuation is divided into distance attenuation caused by geometric diffusion and material attenuation caused by soil medium absorption. The acoustic wave attenuation compensator calculates the distance attenuation according to the spherical wave propagation law. The sound pressure level decreases logarithmically with the increase of distance. For example, 90 dB is measured at a distance of 1 meter from the sound source and then gradually weakens. At the same time, the medium absorption loss is calculated by combining with the soil acoustic impedance parameters. The acoustic impedance is determined by the soil density and sound velocity. For example, the acoustic impedance of sandy soil is about 3.2 MPa·s / m, and that of clay is 2.8 MPa·s / m. Different acoustic impedances result in differences in absorption loss. Based on the distance attenuation value and the medium absorption loss value, the acoustic energy calculator restores the original sound pressure level value at the leakage point. The initial value of the leakage aperture is further determined by using the sound pressure level calibration curve. The curve shows that for every 6 dB increase in the sound pressure level, the aperture approximately doubles. This relationship provides an intuitive basis for aperture estimation. Through this step, the acoustic characteristics of the leakage point can be deduced from the signal attenuation, providing a key input for subsequent flow calculation.

[0069] S1062. Estimate the leakage volume according to the initial value of the leakage aperture and the pipeline network pressure data. Construct the mapping relationship between the sound pressure level and the leakage volume by support vector regression and Gaussian process regression and generate the leakage degree prediction curve.

[0070] In the embodiments of the present invention, the normal operating pressure of the pipe network is maintained at 0.6 MPa. When there is a leak, the pressure fluctuation shows periodic characteristics, and the fluctuation period is related to the aperture size. The Bernoulli flow calculator estimates the leakage volume based on the pressure difference and the aperture area. The calculation result needs to be adjusted by a hydrodynamics correction coefficient, which is related to the Reynolds number, 0.6 in the laminar flow state and 0.8 in the turbulent flow state, to ensure the accuracy of the estimated value. The support vector regressor uses a radial basis kernel function and is trained with historical leakage data to establish a non-linear mapping relationship between the original sound pressure level and the leakage volume. The Gaussian process regressor further constructs a leakage degree prediction curve, calculates the leakage duration in combination with the pressure fluctuation characteristics of the pipe network, and determines it as abnormal when the fluctuation amplitude exceeds 0.1 MPa or the frequency exceeds 10 Hz. The classification threshold is extracted from the historical leakage database, and the leakage volume is divided into four levels: trace, less than 0.1 cubic meters per hour; slight, 0.1 to 1 cubic meters per hour; moderate, 1 to 5 cubic meters per hour; severe, greater than 5 cubic meters per hour, providing a quantitative standard for evaluation.

[0071] In the embodiments of the present invention, the multivariate regression analyzer comprehensively evaluates the leakage volume, duration, and pressure fluctuation characteristics by using the weighted least squares method, assigning weights of 0.5, 0.3, and 0.2 respectively, and outputs the leakage level determination result, comprehensively reflecting the severity and potential impact of the leakage.

[0072] It can be understood that the present invention does not specifically limit the sensor sampling frequency or the specific value of the attenuation coefficient, and can be adjusted according to the actual soil conditions and the state of the pipe network to improve the applicability.

[0073] S107. Judge the pipeline vibration interference or the damage of the insulation layer according to the leakage degree evaluation result, correct the signal propagation path and the attenuation coefficient in combination with the soil characteristics, update the leakage point positioning result, continuously collect data and compare it with the baseline, and complete the pipeline health status monitoring.

[0074] In the embodiments of the present invention, the pipeline vibration monitor uses a piezoelectric acceleration sensor, and a monitoring point is arranged every 50 meters along the pipeline. The sampling frequency is 1000 Hz, covering the vibration signal range of 10 to 500 Hz. During normal operation, the vibration spectrum shows broadband characteristics, the main frequency is concentrated in 50 to 200 Hz, and the amplitude is lower than 0.1 g. If it exceeds this range, it may indicate abnormal interference. The temperature monitor uses an infrared array sensor, with a temperature measurement accuracy of 0.1 °C, scans once every 0.5 meters, and generates a surface temperature distribution map of the pipeline. When the insulation layer is intact, the surface temperature is stable at 30 to 35 °C, and a hot spot is formed at the damaged part, with the temperature rising by 5 to 10 °C, providing an intuitive basis for analysis.

[0075] S1071. Calculate the characteristic frequency value through the vibration signal and the temperature data and generate an interference cancellation function, and recalculate the leakage signal propagation path in combination with the temperature distribution map and the soil acoustic parameters.

[0076] In an embodiment of the present invention, the vibration spectrum analyzer performs time-frequency analysis on the collected vibration signals, extracts the characteristic frequency values. If the frequency exceeds a preset threshold, such as the typical interference range of 80 to 120 Hz, the vibration compensation calculator is used to generate an interference cancellation function. This function is based on the adaptive filtering method, identifies and suppresses the interference components through the least mean square error criterion. The length of the filtering window is set to 256 points, and the convergence step size is 0.01 to ensure the real-time and accuracy of signal processing. At the same time, the temperature field calculator constructs a temperature distribution map according to the heat conduction equation, considering the change of the thermal conductivity of the thermal insulation layer. The normal value is 0.035 watts per meter Kelvin, and it rises to 0.1 watts per meter at the damaged part. Combining with the acoustic parameter database of the soil medium, the sound speed and attenuation coefficient of different soils are provided. The sound speed range is between 1200 and 2000 m / s, the attenuation coefficient is between 0.3 and 1.0 dB / m, and the attenuation coefficient increases by 0.2 dB / m for every 10% increase in the water content. The acoustic path tracer calculates the propagation path of the leakage signal using the ray tracing method, considering the refraction of the sound ray caused by soil stratification. The refraction angle is determined by the sound speed ratio, so as to optimize the path accuracy.

[0077] S1072. Update the acoustic wave propagation loss value according to the corrected propagation path and recalculate the leakage point coordinates, and evaluate the pipeline health status based on the comparison of multi-parameter data with the baseline.

[0078] In an embodiment of the present invention, the attenuation coefficient corrector updates the acoustic wave propagation loss value in combination with the soil characteristics, and uses the medium boundary reflection calculator to generate an acoustic wave propagation trajectory map, showing the equal sound path lines and the boundary refraction phenomenon. The adaptive filter further suppresses the noise of the trajectory map, recombines with the spatial positioning algorithm to recalculate the leakage point coordinates, and improves the positioning reliability. Subsequently, the pipeline operation state parameters, including pressure, temperature and flow rate, are obtained through the multi-parameter data collector. The sampling period is 1 minute, and continuous sampling is carried out for more than 24 hours. The recurrent neural network adopts a long short-term memory structure to analyze the temporal variation of the parameters and predict the trend. The pipeline operation parameter baseline is set based on historical data, the pressure is 0.6 MPa, the temperature is 60 °C, and the flow rate depends on the pipe diameter. The state evaluator calculates the parameter deviation degree using the weighted Euclidean distance, and the weights are 0.4 for pressure, 0.3 for temperature, and 0.3 for flow rate. The health status is divided into four levels: normal, attention, warning, and alarm, reflecting the real-time condition of the pipeline.

[0079] It can be understood that the present invention does not strictly limit the sensor layout spacing or the sampling duration, and can be adjusted according to actual needs to adapt to different scenarios.

[0080] The above-disclosed is only a preferred embodiment of the present invention. Of course, the scope of the rights of the present invention cannot be limited thereby. Those of ordinary skill in the art can understand all or part of the processes of implementing the above embodiments, and the equivalent changes made according to the claims of the present invention still fall within the scope covered by the invention.

Claims

1. A method for monitoring the operation of a steam pipe network, characterized in that, The method includes: Obtaining the pipe burial depth information, leakage signal, temperature, pressure, and soil thermal conductivity change information of the underground steam pipe network; Constructing a pipe network spatial model, marking the positions of pipe fittings, obtaining the soil type, and calculating the attenuation coefficients of leakage signals in different soil media; Simulating the propagation path of the leakage signal based on the pipe network spatial model. At the same time, according to the pipe burial depth information, obtain the complex areas of the pipe network structure. If the leakage signal shows superposition or interference in the complex area, separate the independent signal components to obtain the initial propagation direction of the leakage signal; According to the initial propagation direction of the leakage signal and the attenuation coefficients of the leakage signal in different soil media, combined with the soil thermal conductivity change information, calculate the energy loss of the leakage signal in different soil media. According to the signal energy attenuation characteristics, identify the relative position relationship between the leakage point and the sensor to obtain the preliminary positioning range of the leakage point; Extract the feature parameters of the multi-dimensional signal from the obtained leakage signal, temperature, and pressure data. Analyze whether there is a correlation between the leakage signal, temperature, and pressure according to the multi-dimensional signal feature parameters. If so, combine the simulation results of the propagation path to correct the preliminary positioning range of the leakage point to obtain the leakage point position; According to the position of the leakage point, combined with the pipe burial depth and soil type attributes, use the regression analysis method to calculate the quantitative relationship between the intensity of the leakage signal and the leakage degree to obtain the evaluation result of the leakage degree; According to the evaluation result of the leakage degree, judge whether there is pipeline vibration interference or insulation layer damage. If so, combine the soil characteristic information to correct the signal propagation path and attenuation coefficient, update the leakage point positioning and evaluation results. Based on the updated results, continuously collect sensor data within the preset range and compare it with the normal operation baseline to monitor the pipeline health status.

2. The method according to claim 1, characterized in that, The obtaining of the pipe burial depth information, leakage signal, temperature, pressure, and soil thermal conductivity change information of the underground steam pipe network includes: Collecting the temperature data around the pipe by the underground pipe network sensor array, and matching the thermal conductivity parameters around the pipe from the soil thermal conductivity database through the temperature numerical change gradient; Performing a spatial registration operation on the thermal conductivity parameter and the temperature field distribution map using the three-dimensional pipe network structure diagram to obtain the pipe burial depth value; Forming a monitoring network around the pipe through the sensor array, and obtaining the spatial coordinates of the leakage point by using the acoustic wave positioning algorithm according to the peak time of the pressure fluctuation curve and the time difference of the pressure fluctuation arrival at the monitoring point; Calculating the leakage amount value and leakage duration using the radial basis function for the spatial coordinates of the leakage point and the pressure fluctuation data; 3. The method according to claim 1, wherein The constructing of the pipe network spatial model, marking the positions of pipe fittings, obtaining the soil type, and calculating the attenuation coefficients of leakage signals in different soil media includes: Collecting the spatial point cloud data of the pipe section by the pipe network three-dimensional scanning device, and generating the initial pipe network spatial topology structure diagram according to the density distribution characteristics of the point cloud data; For the pipe section connection nodes in the initial pipe network spatial topology structure diagram, using the density clustering method to extract the node feature vectors, and obtaining the spatial position coordinates of the pipe fittings by matching with the pipe fitting shape feature library; Calibrate the sampling area according to the spatial position coordinates of the pipe fittings, measure the soil conductivity values at different depths using a soil resistivity sensor, and construct a three-dimensional soil medium distribution map through the conductivity values; For the three-dimensional soil medium distribution map, obtain the acoustic response signal of the pipe network, and calculate the acoustic attenuation coefficient according to the propagation path of the acoustic response signal.

4. The method according to claim 1, characterized in that Simulate the propagation path of the leakage signal based on the pipe network spatial model. At the same time, according to the pipeline burial depth information, obtain the complex area of the pipe network structure. If the leakage signal appears superimposed or interfered in the complex area, separate the independent signal components to obtain the initial propagation direction of the leakage signal, including: Collect the three-dimensional spatial data of the pipeline by the pipe network detection device, and obtain the position coordinates of the complex area of the pipe network by calculating the pipeline node density value and the burial depth distribution data; Use a distributed sensor array to obtain the acoustic response signal in the complex area of the pipe network, and calculate the signal intensity attenuation curve according to the signal frequency spectrum characteristics and the sensor acquisition position; For the sensor nodes whose signal intensity attenuation curve exceeds the preset attenuation threshold, use a multi-channel signal collector to record the aliased waveform, and obtain the waveform feature vector group through the independent component extraction algorithm; Calculate the signal propagation delay according to the waveform feature vector group and the sensor node spatial distribution map, use an adaptive Kalman filter to attenuate the waveform noise, and obtain the initial propagation direction of the leakage signal through the acoustic wave positioning algorithm.

5. The method according to claim 1, characterized in that Calculate the energy loss of the leakage signal in different soil media according to the initial propagation direction of the leakage signal and the attenuation coefficient of the leakage signal in different soil media, and combine the change information of the soil thermal conductivity. Identify the relative position relationship between the leakage point and the sensor according to the signal energy attenuation characteristics, and obtain the preliminary positioning range of the leakage point, including: Collect the leakage signal amplitude data by the distributed acoustic sensor array, and obtain the acoustic wave propagation path vector map through the spatial distribution coordinates of the sensor nodes; For the acoustic wave propagation path vector map, use an acoustic wave attenuation detector to measure the propagation loss values in each soil layer, and generate an acoustic wave propagation loss distribution map through an acoustic wave intensity attenuation calculator; Based on the acoustic wave propagation loss distribution map, if the energy propagation loss value in the soil medium is calculated by the acoustic wave energy attenuation model, then generate the acoustic wave energy spatial distribution map; Use the acoustic wave energy spatial distribution map for grid meshing, establish a distance attenuation prediction model through a support vector regression calculator, and determine the preliminary positioning range of the leakage point in combination with the acoustic wave energy threshold criterion.

6. The method according to claim 1, wherein Extract the features of the obtained leakage signal, temperature and pressure data to obtain multi-dimensional signal feature parameters, analyze whether there is a correlation between the leakage signal, temperature and pressure according to the multi-dimensional signal feature parameters. If so, combine the simulation results of the propagation path to correct the preliminary positioning range of the leakage point to obtain the leakage point position, including: Use the underground pipe network monitoring array to obtain the leakage signal waveform data, temperature distribution data and pressure change data, and obtain the waveform frequency characteristic value through the wavelet decomposer according to the leakage signal waveform data; Generate a spectral distribution curve for the waveform frequency eigenvalue, and analyze the spectral distribution curve through an adaptive threshold segmenter to obtain a characteristic frequency interval, which is used to mark the leakage characteristics; According to the temperature distribution data and pressure change data within the characteristic frequency interval, use a characteristic correlation calculator to calculate the temperature change rate and pressure pulse value, and obtain a parameter correlation matrix through Pearson correlation coefficient operation; For the parameter correlation matrix, use a Kalman filter to perform data fusion on the leakage signal waveform, temperature change rate, and pressure pulse value, and perform clustering analysis on the data fusion result through a support vector machine to obtain the leakage point coordinate value.

7. The method according to claim 1, wherein According to the position of the leakage point, combined with the pipeline burial depth and soil type attributes, use a regression analysis method to calculate the quantitative relationship between the intensity of the leakage signal and the leakage degree, and obtain the evaluation result of the leakage degree, including: Collect leakage acoustic signal data through an underground pipe network monitoring device, obtain the leakage point depth value according to the leakage acoustic signal data, and extract the acoustic wave propagation attenuation coefficient corresponding to the leakage point depth value from the soil parameter database; For the acoustic wave propagation attenuation coefficient, use an acoustic wave attenuation compensator to calculate the distance attenuation value, and combine the soil acoustic impedance parameter to obtain the medium absorption loss value at the leakage point; According to the distance attenuation value and the medium absorption loss value, use an acoustic energy calculator to obtain the original sound pressure level value at the leakage point, and use a sound pressure level calibration curve to determine the initial leakage aperture value at the leakage point; For the initial leakage aperture value and the pipe network pressure data, use a Bernoulli flow calculator to generate an estimated leakage volume value, establish a mapping relationship between the original sound pressure level value and the estimated leakage volume value through a support vector regressor, and construct a leakage degree prediction curve based on a Gaussian process regressor to obtain the evaluation result of the leakage degree.

8. The method according to claim 1, wherein According to the leakage degree evaluation result, determine whether there is pipeline vibration interference or insulation layer damage. If so, combine the soil characteristic information to correct the signal propagation path and attenuation coefficient, update the leakage point positioning and evaluation result, and based on the updated result, continuously collect sensor data within a preset range and compare it with the normal operation baseline to monitor the pipeline health status, including: Use a pipe network vibration monitor to obtain the pipeline vibration signal, collect the surface temperature data of the insulation layer through a temperature monitor, and calculate the vibration characteristic frequency value according to the vibration signal and temperature data; For the frequency interval where the vibration characteristic frequency value exceeds the preset threshold, use a vibration compensation calculator to generate an interference cancellation function, and construct a temperature distribution map according to the interference cancellation function; Based on the temperature distribution map and the interference cancellation function, combined with the parameter values in the soil medium acoustic parameter database, use an acoustic path tracker to calculate the leakage signal propagation path; According to the leakage signal propagation path, update the acoustic wave propagation loss value through an attenuation coefficient corrector, calculate the leakage point coordinate value using a spatial positioning algorithm, and judge the pipeline health status based on the leakage point coordinate value.

Images