Method for on-line monitoring of hydraulic parameters of pipe network of fire water system of ancient building and leakage positioning

Abstract

This application relates to the technical field of pipeline leakage monitoring, and in particular to a method for online monitoring and leakage location of hydraulic parameters in fire protection water systems of ancient buildings. The method includes: collecting sound waves at multiple sensing nodes in the pipeline network; extracting interference sources and their blind end lengths, and extracting reference sound velocities and impedance nodes based on material properties; superimposing signals and performing attenuation iterations until the envelope time width of each superimposed signal reaches a preset minimum convergence condition; combining the two superimposed signals corresponding to adjacent sensing nodes and the timestamps after synchronous compensation to extract the first arrival time difference; dividing the connected path into multiple material intervals using the coordinates of the impedance nodes as boundaries; calling the reference sound velocity corresponding to each material interval; constructing a time delay equation by combining the first arrival time difference; and solving the time delay equation to output the leakage coordinates. This application can solve the problem of inaccurate leakage location caused by multipath reflection and material abrupt changes in the pipeline network, achieving precise location of the leakage point.

Description

Technical Field

[0001] This application relates to the technical field of pipeline leakage monitoring, and in particular to a method for online monitoring of hydraulic parameters and leakage location of fire protection water system networks in ancient buildings. Background Technology

[0002] The fire-fighting water system network of ancient buildings is responsible for ensuring the building's fire safety. Due to the multi-courtyard layout, the pipeline routing of such networks is intricate, with a high density of tees and elbows. Furthermore, different construction stages often involve the use of dissimilar pipe materials such as cast iron and PE pipes. During long-term operation, minor leaks frequently occur due to aging or foundation settlement. Online monitoring and location of leaks in this special network are fundamental to maintaining its pressure-holding capacity and water resource management.

[0003] Chinese invention patent publication number CN116241807B discloses a pipeline leakage monitoring method and system based on negative pressure waves and ultrasonic guided waves. This scheme sets multiple sampling points at equal intervals at the beginning, end, and middle of the pipeline to acquire the time, propagation velocity, and stress value obtained by analyzing stress echoes using ultrasonic guided wave analysis of each sampling point. The system extracts the negative pressure wave time difference ratio, velocity difference ratio, and stress value difference ratio between adjacent sampling points, and uses trigonometric functions to extend these ratios into a propagation wave cluster. This method filters the difference between the maximum and minimum dimensions of each wave cluster segment, calculates the wave cluster segment differentiation data, and extracts the wave cluster segment with the largest differentiation value as the prominent wave cluster segment. Finally, the physical interval of adjacent sampling points corresponding to this wave cluster segment is divided into fixed shares, and a specific share interval is taken as the prediction interval for pipeline leakage.

[0004] The aforementioned scheme relies on equidistantly distributed sampling points at the data processing level and assumes the pipeline is a continuous physical medium with uniform acoustic properties. When this system is deployed in the fire water pipe network of ancient buildings, the high density of tees and elbows directly induces multipath reflections and phase superposition of transient sound waves. Simultaneously, an objective acoustic impedance step inevitably exists at the junction of cast iron and PE pipes in the network. Limited by the scheme's calculation of the transmission wave bundle, which only extracts the mathematical characteristics of the time-velocity difference ratio between adjacent sampling points and directly uses trigonometric functions for smoothing, it deviates from the actual spatial topology and physical material boundaries of the pipe network. Under the combined effect of multipath scattering interference and sudden changes in local wave velocity, the first arrival time and wave velocity collected by the sensor will undergo severe distortion. At this time, the calculated extreme values ​​of wave bundle segment differentiation cannot correspond to the actual physical leakage source, but instead randomly fall on data anomalies caused by multiple reflection superpositions, ultimately causing a severe shift in the predicted coordinates calculated based on equal division of a constant physical interval. Summary of the Invention

[0005] To address the inaccurate leak location caused by multi-path reflections and material abrupt changes in pipe networks, and to achieve precise leak point location, this application provides an online monitoring and leak location method for hydraulic parameters of fire protection water system pipe networks in ancient buildings.

[0006] The online monitoring and leak location method for hydraulic parameters of fire protection water system networks in ancient buildings provided in this application adopts the following technical solution: The online monitoring and leak location method for hydraulic parameters of fire protection water system networks in ancient buildings includes: Sound waves are collected at multiple sensing nodes in the pipeline network. When the energy of the sound wave at an adjacent sensing node suddenly crosses a preset baseline, the corresponding initial sequence and timestamp are extracted, and the timestamp is synchronously compensated. Based on the adjacent sensing nodes corresponding to the compensated timestamps, the connected paths are extracted from the preset pipeline network map; The topology labels and material properties of the connected path are analyzed to extract the interference source and its blind end length, and the reference sound velocity and impedance node are extracted based on the material properties. For the interference source, the reflection delay is calculated by calling the corresponding reference sound speed and the blind end length. The reverse sequences corresponding to the adjacent sensing nodes are constructed respectively. Each reverse sequence is superimposed with the corresponding initial sequence to generate a corresponding superimposed signal and is attenuated iteratively until the envelope width of each superimposed signal reaches the preset minimum convergence condition. The first arrival time difference is extracted by combining the two superimposed signals corresponding to the adjacent sensing nodes and the timestamp after synchronization compensation. Using the coordinates of the impedance node as the boundary, the connected path is divided into multiple material intervals. The reference sound velocity corresponding to each material interval is called, and a time delay equation is constructed in conjunction with the first arrival time difference. The time delay equation is solved to output the leakage coordinates.

[0007] Optionally, when the energy change of the sound wave at an adjacent sensing node crosses a preset baseline, the corresponding initial sequence and timestamp are extracted, and the timestamp is synchronized and compensated, including: The energy derivative of the sound wave within the sliding time window is calculated as the energy mutation. When the energy derivative crosses the dynamic baseline, the dynamic baseline is set as the preset baseline, the initial sequence including the signal segment is extracted, and the timestamp including the hardware timestamp is obtained; The hardware timestamp is linearly compensated based on the temperature drift coefficient of the sensing node to complete the synchronization compensation.

[0008] Optionally, before setting the dynamic baseline to the preset baseline, the method further includes: Collect the integral difference of steady-state background noise; When the integral difference exceeds the preset noise drift threshold, the leakage compensation amount is calculated based on the integral difference, and the leakage compensation amount is superimposed on the dynamic baseline for updating.

[0009] Optionally, the step of parsing the topology labels and material properties of the connected path, extracting interference sources and their blind-end lengths, and extracting the reference sound velocity and impedance nodes based on the material properties includes: Traverse the connected paths; When the topology label is a branch node or a turning node, the corresponding physical coordinates are extracted as the interference source, and the associated physical length is read as the blind end length. When the connecting path has a material transition between adjacent pipe segments, the corresponding joint coordinates are extracted as the impedance node, and the physical sound velocity on both sides of the material transition is retrieved as the reference sound velocity.

[0010] Optionally, when traversing the connected path, if the connected path contains a detection blind zone lacking the topology label, the method further includes: A probe wave is emitted into the detection blind zone, and the echo with cross-sectional characteristics reflected from the detection blind zone is received; Based on the cross-sectional features, an equivalent pipe diameter step feature is extracted, and the equivalent pipe diameter step feature is converted into a supplementary label, so as to update the topology label corresponding to the detection blind zone using the supplementary label.

[0011] Optionally, for the interference source, the reflection delay is calculated by calling the corresponding reference sound speed and the blind end length, and reverse sequences corresponding to adjacent sensing nodes are constructed respectively. Each reverse sequence is superimposed with the corresponding initial sequence to generate a corresponding superimposed signal and is subjected to attenuation iteration until the envelope width of each superimposed signal reaches a preset minimum convergence condition. The first arrival time difference is extracted by combining the two superimposed signals corresponding to adjacent sensing nodes and the timestamp after synchronization compensation, including: The reflection delay is used to perform time-shift inversion on the initial sequence corresponding to the adjacent sensing nodes to generate the corresponding reverse sequence; The inverse sequences and the corresponding initial sequences are superimposed to generate the corresponding superimposed signals, and the envelope time width of each superimposed signal is extracted. The attenuation coefficient is adjusted according to a preset step size and the corresponding reverse sequence is updated to reconstruct the corresponding superimposed signal and the corresponding envelope time width in a loop until the envelope time width of each superimposed signal satisfies the preset minimum convergence condition. The two superimposed signals corresponding to adjacent sensing nodes are obtained, and the timestamps after synchronization compensation are combined to calculate the actual arrival time difference between adjacent sensing nodes, which is used as the first arrival time difference.

[0012] Optionally, the step of dividing the connected path into multiple material intervals using the coordinates of the impedance node as boundaries, calling the reference sound velocity corresponding to each material interval, constructing a time delay equation in conjunction with the first arrival time difference, and solving the time delay equation to output the leakage coordinates includes: The distance from the leak location to the corresponding impedance node is set as an unknown parameter, and a time consumption function with respect to the reference sound velocity is constructed in conjunction with the unknown parameter. Using the coordinates of the impedance node as the integration boundary, perform nonlinear integral accumulation on all the time consumption functions, and combine the accumulation result with the first arrival time difference to generate the time delay equation; Solve the time delay equation to extract the distance solution of the unknown parameter, and perform spatial mapping on the distance solution along the connected path to output the leakage coordinates.

[0013] Optionally, when constructing the time consumption function, if the material range includes a detection blind spot where the reference sound velocity is missing, the function further includes: Extract the dispersion characteristics generated by the sound wave passing through the detection blind segment; The dispersion characteristics are substituted into a preset transmission model to perform curve fitting and calculate the equivalent speed of sound. The equivalent sound velocity is used as the reference sound velocity corresponding to the detection blind segment to construct the time consumption function.

[0014] Optionally, after outputting the leakage coordinates, the method further includes: Extract the acoustic-structure coupling dispersion characteristics or high-frequency amplitude attenuation rate of the pipe wall within the frequency band corresponding to the initial sequence; When the acoustic-solid coupling dispersion characteristics of the pipe wall or the high-frequency amplitude attenuation rate meet the preset solid-propagation discrimination threshold, the acoustic wave corresponding to the leakage coordinate is determined to be a false signal, and a fuse command is triggered to discard the leakage coordinate.

[0015] Optionally, after the circuit breaker command is triggered and the leak coordinates are discarded, the method further includes: In response to the circuit breaker command, the amplitude characteristics of the pseudo signal are extracted, and the up-adjustment coefficient is calculated based on the amplitude characteristics; The adjustment coefficient is sent to the sensing node and then superimposed on the dynamic baseline for updating.

[0016] In summary, this application includes the following beneficial technical effects: 1. This application extracts interference sources and impedance nodes by parsing the topological labels and material properties of connected paths. It then calculates the reflection delay using the blind end length to construct a reverse sequence, which is superimposed on the initial sequence and iterated until the envelope width meets the minimum convergence condition. Finally, it constructs a time delay equation by combining the extracted first-arrival time difference and impedance node boundaries to solve for the coordinates. These features, through time shifting and attenuation reconstruction of the reverse sequence, offset the multipath reflection interference caused by blind ends in the pipeline network and match the reference sound velocity segmented according to the physical joints. This avoids the signal distortion error caused by directly using a globally uniform medium model, solves the problem of inaccurate leak location caused by high-density tee joints and heterogeneous pipe connections in ancient building fire water system networks, and achieves accurate leak point location under multipath reflection and material abrupt change environments.

[0017] 2. This application extracts the acoustic-structure coupling dispersion characteristics or high-frequency amplitude attenuation rate of the pipe wall within the corresponding frequency band of the initial sequence. When the solid-structure transmission discrimination threshold is met, a fuse command is triggered to discard the leakage coordinates. Furthermore, pseudo-signal amplitude characteristics are extracted to calculate an up-adjustment coefficient to update the dynamic baseline of the sensing node. These features utilize the underlying native dispersion and attenuation properties to independently identify the physical propagation mechanism, isolating high-speed acoustic interference waves propagating rapidly from solid structures. This effectively eliminates interference from fire hydrant metal wall vibrations caused by external activities such as tourists jumping in groups, reduces the probability of repeated triggering of similar acoustic-structure coupling pseudo-signals, avoids misjudgments caused by non-leakage abnormal noise, and improves the anti-interference capability of the positioning system.

[0018] 3. When encountering a detection blind zone with missing topology labels during the traversal of a connected path, this application emits a probe wave into the blind zone and receives reflected echoes containing cross-sectional features. This allows for the extraction of equivalent pipe diameter step features, which are then converted into supplementary labels to update the corresponding topology labels for the blind zone. These features utilize the acoustic reflection characteristics of the pipe diameter change interface to actively detect physical boundaries and obtain spatial geometric abrupt changes in unknown pipe segments to complete the map records in the mapping server. This compensates for the gaps in the digital twin map caused by missing pipeline construction data, prevents the failure of reverse sequence reconstruction due to unknown local topology structures, and ensures the consistency of the time delay calculation logic and the overall monitoring accuracy in pipelines with incomplete information. Attached Figure Description

[0019] Figure 1 A flowchart of a method for online monitoring and leak location of hydraulic parameters of fire-fighting water system network in ancient buildings, provided in an embodiment of this application; Figure 2 This application provides waveform diagrams illustrating the time-domain processing of the initial sequence, reverse sequence, and superimposed signal in embodiments of this application. Figure 3 Trend chart for identifying the acoustic-structure coupling dispersion characteristics or high-frequency amplitude attenuation rate of the pipe wall provided in the embodiments of this application. Detailed Implementation

[0020] The following combination Figures 1-3 This application will be described in further detail.

[0021] This application discloses an online monitoring and leak location method for hydraulic parameters of fire-fighting water system networks in ancient buildings, which is executed collaboratively by sensing nodes, mapping servers, and calculation terminals. Combined with... Figure 1 The method sequentially executes the following steps: energy mutation crossing the preset baseline and synchronous compensation, topology analysis to extract interference sources and impedance nodes, construction of reverse sequence and attenuation iteration to minimum convergence, joint superposition of signals to extract first arrival time difference, segmented matching of reference sound velocity to construct time delay equation, and solving the equation to obtain leakage coordinates. In the final positioning stage, based on whether the extracted pipe wall acoustic-structure coupling dispersion characteristics or high-frequency amplitude attenuation rate meets the preset solid-propagation discrimination threshold, the branch execution triggers the fuse command to discard the coordinates and update the dynamic baseline or outputs the leakage coordinates.

[0022] The sensing node is an edge acoustic vibration unit that is non-destructively adsorbed onto the metal wall of the ancient building's fire hydrant. It has a built-in piezoelectric acoustic emission sensor, FPGA timing control array and real-time clock module, and the sampling rate is configured to be no less than 100kHz. In some embodiments, the piezoelectric acoustic emission sensor is replaced by a fiber optic acoustic sensor or a microelectromechanical accelerometer.

[0023] The mapping server stores a digital twin map of the three-dimensional pipe network of the ancient building. The map records the physical length of reducers, tees and side blind ends with topological tags, and records the reference sound velocity and attenuation coefficient of each pipe with material properties. In some embodiments, the digital twin map uses a graph database to carry the association between topological tags and material properties.

[0024] The solution terminal acts as the master control device, sending trigger and attenuation adjustment commands to the sensing nodes, and undertaking the analysis of connectivity paths and the solution of time delay equations.

[0025] The sensing node collects the sound waves from the fire hydrant wall and calculates the energy derivative dE / dt of the sound waves within the sliding time window as the energy mutation, where E is the energy integral value of the sound waves within that sliding time window. The window length of the sliding time window is between 2ms and 5ms. A window length shorter than 2ms makes it easy for microsecond-level transient noises such as bubble rupture to be falsely triggered, while a window length longer than 5ms causes the leakage pulse initiation edge and the steady segment to be intercepted together, weakening the time resolution. The preferred window length is 3ms to balance the suppression of transient false triggering and the preservation of the initiation edge.

[0026] Before setting the dynamic baseline to the preset baseline, the sensing node acquires the integral difference value of the steady-state background noise. When the integral difference value exceeds the preset noise drift threshold, the leakage compensation amount is calculated based on the integral difference value and superimposed on the dynamic baseline for updating. After the energy derivatives at two adjacent sensing nodes cross the dynamic baseline within the same sliding time window, the FPGA timing control array sets the dynamic baseline to the preset baseline, locks the initial sequence containing the sliding time window, and loads the hardware timestamp from the real-time clock module.

[0027] The sensing nodes perform linear compensation for the hardware timestamps based on their own temperature drift coefficients. The temperature drift coefficient, expressed in PPM, represents the shift in crystal oscillator frequency with temperature, thus completing the synchronization compensation of the timestamps. The synchronization accuracy of the real-time clock module is better than 10μs. The reference sound velocity in the water-filled pipe section within the pipeline network is in the range of 1000m / s to 1480m / s. A 10μs timing error corresponds to a spatial positioning deviation of approximately 1.0cm to 1.5cm. This synchronization accuracy is guaranteed by the IEEE 1588 Precision Time Protocol between the real-time clock module and the processing terminal. In courtyards without network timing capabilities, GPS second pulse timing is used as a substitute.

[0028] The adjacent sensing nodes corresponding to the compensated timestamp are uploaded to the solution terminal along with the initial sequence. If the data packet of either of the two adjacent sensing nodes does not reach the solution terminal within the preset waiting time window, the solution terminal discards the received data and re-enters listening. After the data is complete, the solution terminal requests the connection path between the two nodes from the mapping server.

[0029] After receiving the connected path, the solution terminal traverses the topology tags and material properties of the connected path. The physical coordinates of the topology tags marking the locations of branch nodes or turning nodes are extracted as interference sources, and the associated physical length is read as the blind end length. In some embodiments, reducers are also included as interference sources. The coordinates of joints where there is a material jump between adjacent pipe segments in the connected path are extracted as impedance nodes, and the physical sound velocities on both sides of the material jump are retrieved as the reference sound velocities.

[0030] When a connected path contains a detection blind zone with a missing topology label, the sensing node emits a probe wave into the blind zone and receives the echo with cross-sectional characteristics reflected from the blind zone. Based on the cross-sectional characteristics, it extracts the equivalent pipe diameter step feature and converts it into a supplementary label to update the topology label corresponding to the blind zone. The interference source, blind end length, impedance node, and reference sound velocity constitute the physical boundary for subsequent solutions.

[0031] The calculation terminal calculates the reflection delay for the interference source using the corresponding reference sound velocity and blind end length. The reflection delay is equal to the propagation time of the sound wave along the side blind end and the first-order reflection, and its value is twice the blind end length divided by the reference sound velocity. The calculation terminal uses the reflection delay to perform time-shifting and phase inversion on the initial sequences corresponding to adjacent sensing nodes to generate corresponding reverse sequences. In some embodiments, the reverse sequence is replaced by a time-reversed copy of the initial sequence to adapt to different multipath structures.

[0032] Combination Figure 2 The initial sequence contains multiple fluctuating waveforms, while the reverse sequence is time-displaced and has opposite waveform polarity. The processing terminal superimposes the corresponding reverse sequence and the corresponding initial sequence to obtain the corresponding superimposed signal and extracts the envelope time width of the superimposed signal. The amplitude attenuation coefficient is adjusted according to a preset step size, and the corresponding reverse sequence is updated. The superimposed signal and the corresponding envelope time width are reconstructed cyclically until the corresponding envelope time width meets the minimum convergence condition. At this point, the interference fluctuations in the superimposed signal are canceled out into a pure single-peak waveform. The processing terminal obtains two superimposed signals corresponding to adjacent sensing nodes and, together with the timestamps after synchronization compensation, extracts the true arrival time difference between adjacent sensing nodes at the corresponding single peak of the waveform as the first arrival time difference. The amplitude decay coefficient adopts an adaptive descent strategy with an initial step size of 0.01 and a halving when the envelope width increases. A step size greater than 0.05 causes the envelope width to oscillate around the minimum value and cross the convergence point. A step size less than 0.005 increases the number of iterations and lengthens the solution cycle. Two consecutive cycles of envelope width increase are identified as crossing the minimum value, triggering step size halving and reverse backtracking to avoid iteration divergence.

[0033] The solution terminal divides the connected path into multiple material intervals using the coordinates of the impedance nodes as boundaries. The distance from the leak location to the corresponding impedance node is set as an unknown parameter, and a time consumption function about the unknown parameter is constructed in conjunction with the reference sound velocity.

[0034] When the material range includes a detection blind segment with a missing reference sound velocity, the calculation terminal extracts the dispersion characteristics generated by the sound wave passing through the detection blind segment, substitutes them into the preset transmission model to perform curve fitting to calculate the equivalent sound velocity, and uses the equivalent sound velocity as the reference sound velocity corresponding to the detection blind segment to construct the time consumption function. In some embodiments, the transmission model adopts a frequency-varying transmission coefficient model with acoustic impedance mismatch.

[0035] The solution terminal performs nonlinear integral accumulation on all time-consuming functions using the coordinates of the impedance nodes as the integration boundary. The accumulated result is then combined with the first arrival time difference to generate a time delay equation. Solving the time delay equation yields the distance solution for the unknown parameters. After spatial mapping along the connected path, the leakage coordinates are output.

[0036] After calculating the leakage coordinates output by the terminal, the acoustic-structure interaction dispersion characteristics or high-frequency amplitude attenuation rate of the pipe wall within the corresponding frequency band of the initial sequence are extracted. Combined with... Figure 3 The data curve of the pseudo signal is generally distributed above the real signal curve. When the acoustic-solid coupling dispersion characteristics of the pipe wall or the high-frequency amplitude attenuation rate increases with the change of the signal sequence and meets the preset solid-solid transmission discrimination threshold to generate the boundary point judgment condition, the acoustic wave corresponding to the leakage coordinate is judged as a pseudo signal, and the calculation terminal triggers the fuse command to discard the leakage coordinate.

[0037] In the objective physical propagation mechanism, the upper limit of the preset range for the equivalent wave velocity of sound waves propagating in normal water is set at 1480 m / s, corresponding to the sound velocity of pure water at normal temperature and pressure, while the lower limit is set at 300 m / s. Below this value, it means that the signal is affected by attenuation interference from abnormally low-speed media such as air-filled cavities inside the pipe. Above this upper limit, it indicates that the signal propagates rapidly through the solid structure of the pipe wall with extremely high sound velocity.

[0038] Because calculating spatial coordinates based on the time delay equation and then dividing by the first arrival time difference inevitably leads to a mathematical deadlock due to algebraic degradation to the water body reference sound speed, this application uses the extraction of underlying native features independent of spatial secondary calculations for independent identification. Therefore, the acoustic-structure coupling pseudo-signal excited by tourists jumping on the fire hydrant wall is essentially propagated through a high-speed solid structure. Its acoustic-structure coupling dispersion characteristics or high-frequency amplitude attenuation rate can meet the solid-propagation identification threshold and be accurately intercepted.

[0039] After the circuit breaker command is triggered, the calculation terminal extracts the amplitude characteristics of the pseudo signal and calculates the adjustment coefficient accordingly. The adjustment coefficient is then sent to the sensing node and superimposed on the dynamic baseline to complete the update. This effectively reduces the probability of repeated triggering of similar acoustic-structure coupling pseudo signals and avoids the misjudgment of solid structure vibrations excited by tourist activities as pipe leaks.

[0040] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for online monitoring of hydraulic parameters of a pipe network of a fire water system of an ancient building and leakage positioning, characterized in that, include: Sound waves are collected at multiple sensing nodes in the pipeline network. When the energy of the sound wave at an adjacent sensing node suddenly crosses a preset baseline, the corresponding initial sequence and timestamp are extracted, and the timestamp is synchronously compensated. Based on the adjacent sensing nodes corresponding to the compensated timestamps, the connected paths are extracted from the preset pipeline network map; The topology labels and material properties of the connected path are analyzed to extract the interference source and its blind end length, and the reference sound velocity and impedance node are extracted based on the material properties. For the interference source, the reflection delay is calculated by calling the corresponding reference sound speed and the blind end length. The reverse sequences corresponding to the adjacent sensing nodes are constructed respectively. Each reverse sequence is superimposed with the corresponding initial sequence to generate a corresponding superimposed signal and is attenuated iteratively until the envelope width of each superimposed signal reaches the preset minimum convergence condition. The first arrival time difference is extracted by combining the two superimposed signals corresponding to the adjacent sensing nodes and the timestamp after synchronization compensation. Using the coordinates of the impedance node as the boundary, the connected path is divided into multiple material intervals. The reference sound velocity corresponding to each material interval is called, and a time delay equation is constructed in conjunction with the first arrival time difference. The time delay equation is solved to output the leakage coordinates.

2. The method according to claim 1, characterized in that, When the energy change of the sound wave at an adjacent sensing node crosses a preset baseline, the corresponding initial sequence and timestamp are extracted, and the timestamp is synchronized and compensated, including: The energy derivative of the sound wave within the sliding time window is calculated as the energy mutation. When the energy derivative crosses the dynamic baseline, the dynamic baseline is set as the preset baseline, the initial sequence including the signal segment is extracted, and the timestamp including the hardware timestamp is obtained; The hardware timestamp is linearly compensated based on the temperature drift coefficient of the sensing node to complete the synchronization compensation.

3. The method according to claim 2, characterized in that, Before setting the dynamic baseline to the preset baseline, the method further includes: Collect the integral difference of steady-state background noise; When the integral difference exceeds the preset noise drift threshold, the leakage compensation amount is calculated based on the integral difference, and the leakage compensation amount is superimposed on the dynamic baseline for updating.

4. The method for online monitoring and leak location of hydraulic parameters of ancient building fire protection water system network according to claim 1, characterized in that, The process of parsing the topology labels and material properties of the connected path, extracting interference sources and their blind-end lengths, and extracting the reference sound velocity and impedance nodes based on the material properties includes: Traverse the connected paths; When the topology label is a branch node or a turning node, the corresponding physical coordinates are extracted as the interference source, and the associated physical length is read as the blind end length. When the connecting path has a material transition between adjacent pipe segments, the corresponding joint coordinates are extracted as the impedance node, and the physical sound velocity on both sides of the material transition is retrieved as the reference sound velocity.

5. The method for online monitoring and leak location of hydraulic parameters of ancient building fire protection water system network according to claim 4, characterized in that, When traversing the connected path, if the connected path contains a detection blind zone with a missing topology label, the method further includes: A probe wave is emitted into the detection blind zone, and the echo with cross-sectional characteristics reflected from the detection blind zone is received; Based on the cross-sectional features, an equivalent pipe diameter step feature is extracted, and the equivalent pipe diameter step feature is converted into a supplementary label, so as to update the topology label corresponding to the detection blind zone using the supplementary label.

6. The method for online monitoring and leak location of hydraulic parameters of ancient building fire protection water system network according to claim 4, characterized in that, For the interference source, the reflection delay is calculated by calling the corresponding reference sound speed and the blind end length. Reverse sequences corresponding to adjacent sensing nodes are constructed respectively. Each reverse sequence is superimposed with the corresponding initial sequence to generate a superimposed signal, which is then attenuated iteratively until the envelope width of each superimposed signal reaches a preset minimum convergence condition. The first arrival time difference is extracted by combining the two superimposed signals corresponding to adjacent sensing nodes and the timestamp after synchronization compensation, including: The reflection delay is used to perform time-shift inversion on the initial sequence corresponding to the adjacent sensing nodes to generate the corresponding reverse sequence; The inverse sequences and the corresponding initial sequences are superimposed to generate the corresponding superimposed signals, and the envelope time width of each superimposed signal is extracted. The attenuation coefficient is adjusted according to a preset step size and the corresponding reverse sequence is updated to reconstruct the corresponding superimposed signal and the corresponding envelope time width in a loop until the envelope time width of each superimposed signal satisfies the preset minimum convergence condition. The two superimposed signals corresponding to adjacent sensing nodes are obtained, and the timestamps after synchronization compensation are combined to calculate the actual arrival time difference between adjacent sensing nodes, which is used as the first arrival time difference.

7. The method for online monitoring and leak location of hydraulic parameters of ancient building fire protection water system network according to claim 6, characterized in that, The process involves dividing the connected path into multiple material intervals using the coordinates of the impedance nodes as boundaries, calling the reference sound velocity corresponding to each material interval, constructing a time delay equation in conjunction with the first arrival time difference, and solving the time delay equation to output the leakage coordinates. This includes: The distance from the leak location to the corresponding impedance node is set as an unknown parameter, and a time consumption function with respect to the reference sound velocity is constructed in conjunction with the unknown parameter. Using the coordinates of the impedance node as the integration boundary, perform nonlinear integral accumulation on all the time consumption functions, and combine the accumulation result with the first arrival time difference to generate the time delay equation; Solve the time delay equation to extract the distance solution of the unknown parameter, and perform spatial mapping on the distance solution along the connected path to output the leakage coordinates.

8. The method for online monitoring and leak location of hydraulic parameters of fire protection water system network in ancient buildings according to claim 7, characterized in that, When constructing the time consumption function, if the material range includes a detection blind spot where the reference sound velocity is missing, the function further includes: Extract the dispersion characteristics generated by the sound wave passing through the detection blind segment; The dispersion characteristics are substituted into a preset transmission model to perform curve fitting and calculate the equivalent speed of sound. The equivalent sound velocity is used as the reference sound velocity corresponding to the detection blind segment to construct the time consumption function.

9. The method for online monitoring and leak location of hydraulic parameters of fire protection water system network in ancient buildings according to claim 7, characterized in that, Following the output of the leakage coordinates, the following is also included: Extract the acoustic-structure coupling dispersion characteristics or high-frequency amplitude attenuation rate of the pipe wall within the frequency band corresponding to the initial sequence; When the acoustic-solid coupling dispersion characteristics of the pipe wall or the high-frequency amplitude attenuation rate meet the preset solid-propagation discrimination threshold, the acoustic wave corresponding to the leakage coordinate is determined to be a false signal, and a fuse command is triggered to discard the leakage coordinate.

10. The method for online monitoring and leak location of hydraulic parameters of fire protection water system network in ancient buildings according to claim 9, characterized in that, After the circuit breaker command is triggered and the leak coordinates are discarded, the following steps are also included: In response to the circuit breaker command, the amplitude characteristics of the pseudo signal are extracted, and the up-adjustment coefficient is calculated based on the amplitude characteristics; The adjustment coefficient is sent to the sensing node and then superimposed on the dynamic baseline for updating.

Images