Vacuum pipe network passive end fault positioning method and system

Abstract

The application discloses a kind of vacuum pipe network passive end fault location method and system, the fault location method includes: in the main pipeline network node of vacuum drainage system deployment centralized sensing module;Through centralized sensing module, the aerodynamic acoustic signal and pressure transient signal generated under the action of gas-liquid two-phase fluid dynamics coupling inside vacuum pipe network are continuously and real-time collected;The time series waveform signal collected is extracted, and based on the preset fault classification model, the characteristic waveform signal caused by end equipment anomaly is identified and its fault category is judged;The first-order round-trip time delay and waveform distortion parameters of the characteristic waveform signal propagation in pipe network are extracted, the physical pipeline distance from the abnormal source point to the centralized sensing module is solved, and is associated with three-dimensional vacuum pipe network static topology database Calculation, the specific end equipment node where the abnormality occurs is located.The fault location method of the application makes full use of the inherent physical signal characteristics inside vacuum pipe network.

Description

Technical Field

[0001] This invention belongs to the field of terminal fault diagnosis technology for vacuum drainage systems, and more specifically, relates to a method and system for locating passive terminal faults in vacuum pipe networks. Background Technology

[0002] Vacuum drainage systems, as a novel wastewater collection and transportation technology, utilize negative pressure as a power source. By creating a vacuum environment within the pipeline system, wastewater is drawn in and transported to a collection station under atmospheric pressure. Developed since the 1970s, this technology has been widely applied globally, particularly in space-constrained, complex terrain, or environmentally demanding locations. Vacuum drainage systems offer significant advantages, including flexible layout, effective odor control, excellent water conservation, and strong adaptability to complex spaces, making them uniquely valuable in scenarios such as subway stations, underground utility tunnels, public buildings, and renovations of older residential areas. In subway station applications, vacuum drainage systems effectively solve the problem of difficult layout of traditional gravity drainage systems in underground spaces. In underground utility tunnels, their compact pipe layout and efficient transportation capacity significantly conserve valuable underground space resources. In public buildings, the odor-controlling properties of vacuum drainage systems provide a more hygienic and comfortable environment for densely populated areas. In renovation projects of older residential areas, their flexible installation methods avoid large-scale civil engineering projects, reducing renovation costs and construction time. However, during long-term operation, the terminal equipment of the vacuum drainage system (including vacuum valves, collection boxes, connecting pipes, etc.) is easily affected by factors such as impurities in sewage, humidity of the operating environment, and equipment aging, and frequently experiences typical faults such as air leakage, blockage, valve jamming, and abnormal vacuum. If such faults are not detected and dealt with in time, they will lead to system negative pressure imbalance, reduced drainage efficiency, and in severe cases, sewage overflow, equipment damage, etc., affecting the normal operation of the system and the surrounding environment.

[0003] Currently, the industry primarily employs two technical methods for fault diagnosis of terminal equipment in vacuum drainage systems. The first is the traditional manual inspection method, where professional technicians periodically inspect each terminal device on-site, judging their operating status through visual observation, manual operation, and instrument measurement. This method relies heavily on the experience and skills of the inspectors and typically requires detailed inspection plans and standard operating procedures, including a thorough check of key components such as vacuum valves, collection tanks, and sensors. The second method is sensor-based automated monitoring technology. This involves installing various sensors, such as pressure sensors, flow sensors, level sensors, and temperature sensors, on the terminal equipment of the vacuum drainage system. By collecting real-time equipment operating parameters and combining them with preset threshold judgment standards, the system can automatically monitor equipment status and provide fault warnings. Some advanced systems also incorporate data acquisition and monitoring systems, transmitting sensor data to a central control room for centralized management and analysis via host computer software. Furthermore, some research institutions and companies are exploring the application of Internet of Things (IoT) technology to vacuum drainage systems, using wireless sensor networks to achieve remote monitoring and data transmission, thereby improving the timeliness and accuracy of fault diagnosis.

[0004] Although existing technologies have made some progress in fault diagnosis of vacuum drainage systems, they still have many shortcomings and cannot effectively solve the practical problems faced in system operation. For example, manual inspection is costly, especially in large and complex systems where the workload is enormous. Moreover, the frequency of inspections is limited, making it impossible to achieve 24-hour uninterrupted monitoring and easily missing the best time to deal with sudden faults. Furthermore, the accuracy of fault diagnosis by manual inspection depends entirely on the experience of maintenance personnel and is easily affected by subjective factors. As for sensor monitoring, existing sensors are easily contaminated by oil and water vapor in sewage, leading to malfunctions, distorted monitoring data, and an inability to accurately distinguish fault types (such as abnormal vacuum levels which may be caused by various reasons such as air leakage, blockage, or valve jamming), thus failing to accurately determine the location of the fault. Summary of the Invention

[0005] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a passive end-point fault location method and system for vacuum pipe networks. By fully utilizing the inherent physical signal characteristics within the vacuum pipe network, it features passive operation, environmental friendliness, and reliability. It is particularly suitable for application scenarios with extremely high system reliability requirements, such as subway stations and underground integrated pipe corridors, providing an innovative solution for the intelligent operation and maintenance of vacuum drainage systems.

[0006] To achieve the above objectives, the present invention provides a method for locating faults at the passive end of a vacuum pipeline network, comprising the following steps: S1: Deploy centralized sensing modules at the main pipeline network nodes of the vacuum drainage system. The centralized sensing modules include at least an acoustic sensor array and a dynamic pressure sensor. S2: Through the centralized sensing module, continuously and in real time collect aerodynamic acoustic wave signals and pressure transient signals generated inside the vacuum pipeline under the coupling effect of gas-liquid two-phase fluid dynamics; S3: Extract features from the acquired time-series waveform signals, identify the characteristic waveform signals caused by the abnormality of the terminal equipment based on the preset fault classification model, and determine the fault category. S4: Extract the first-order round-trip time delay and waveform distortion parameters of the characteristic waveform signal propagating in the pipeline network, and use the aeroacoustic time-domain inversion algorithm to calculate the physical pipeline distance from the anomaly source point to the centralized sensing module; S5: The calculated physical pipeline distance is correlated with the three-dimensional vacuum pipeline static topology database to locate the specific end device node where the anomaly occurred.

[0007] Further, step S1 includes: S11: Before determining the final installation location of the centralized sensing module, perform a verification calculation of the ultimate acoustic signal-to-noise ratio of the pipeline network to ensure that it meets the preset signal-to-noise ratio safety threshold. S12: At the node that meets the requirements of the ultimate acoustic signal-to-noise ratio verification calculation, a linear array consisting of at least two broadband acoustic microphones is installed on the wall in a non-intrusive manner. S13: At the opening or reserved interface of the adjacent main pipe, a low-frequency static absolute pressure sensor is installed in an invasive manner to read the steady-state absolute value of the vacuum network in real time; then, high-frequency micro-pressure sensors are installed in parallel in an invasive manner to synchronously capture the transient pressure distortion wave propagating in the medium inside the pipe. S14: Deploy an edge computing and solution gateway in the electrical control cabinet closest to the centralized sensing module, and build a high-speed data acquisition link with hardwired connection between the two.

[0008] Further, step S2 includes: S21: Read the values ​​of the static absolute pressure sensor in real time. When the fluctuation amplitude of the pipeline vacuum degree within the continuous observation time window is less than the preset minimum value, and the monitoring network does not receive any pump station vacuum start / stop command, it is determined that the pipeline has entered a hydraulic idle state. Then, through the active audio frequency generator integrated in the centralized sensing module and acoustically coupled to the main pipeline, a linear sweep frequency detection sound wave with a known bandwidth and duration is emitted into the main pipeline to obtain the physical structure reflection response at different spatial scales in the pipeline. S22: While transmitting the detection sound wave, the acoustic sensor array simultaneously opens the acquisition time window to receive the transient sound pressure signal reflected back from the physical topology of the pipeline network; the digital signal processing module inside the edge computing and solution gateway performs bandpass filtering and noise reduction and analog-to-digital conversion on the acquired analog signal to generate a discrete time-series test sequence, and performs topological overlap detection with the pre-stored reference acoustic reflection feature sequence. S23: Logically compare the obtained topology overlap coefficient with the preset topology safety verification threshold; if the topology overlap coefficient is greater than or equal to the safety verification threshold, it is determined that the current physical pipeline structure has not been artificially modified, and the original three-dimensional vacuum pipeline static topology database mapping relationship remains unchanged; if the topology overlap coefficient is less than the safety verification threshold, it is determined that the pipeline has undergone physical line modification, the original spatial distance positioning calculation service is automatically intercepted and frozen, and a high-level alarm prompt for topology reconstruction and line change is output. At the same time, the abnormal test sequence collected this time is archived as a temporary benchmark, and after manual verification and entry of new pipeline building information model parameters, the benchmark mapping relationship in the local three-dimensional vacuum pipeline static topology database is overwritten and updated again.

[0009] Furthermore, the topological overlap detection method is as follows: ; In the formula, This is the topological overlap coefficient; This represents the total number of sampling points within the acoustic detection time window. The first time-aligned reference acoustic reflection feature sequence The amplitude of each discrete sampling point; The first in the latest test echo characteristic sequence The amplitude of each discrete sampling point; This is the arithmetic mean of the baseline sequence array; This is the arithmetic mean of the test sequence array.

[0010] Further, step S3 includes: S31: The discrete time-series data frame of the corresponding acoustic data is generated through edge computing and solution gateway, and a Hanning window is applied to it to suppress spectral leakage. Then, a fast Fourier transform is performed to convert the one-dimensional time domain signal into a frequency domain amplitude matrix and obtain the frequency domain feature quantity spectral centroid. S32: While extracting features in the frequency domain, envelope tracking is performed on the same data frames in the time domain to lock the initial peak moment of the abnormal event and extract the transient energy decay rate, a time domain feature. S33: Input the spectral centroid and transient energy decay rate into the preset fault classification matrix; if the spectral centroid is greater than the preset high-frequency threshold and the transient energy decay rate is lower than the preset decay safety threshold, it is a continuous leakage fault; if the spectral centroid is less than the preset low-frequency threshold and the transient energy decay rate is greater than the preset rapid decay threshold, it is a sudden blockage fault; if the frequency domain amplitude is discrete and multi-harmonic and the time domain decay rate shows non-convergent fluctuations within the observation window, it is an end mechanical jamming fault.

[0011] Further, step S4 includes: S41: The characteristic waveform signal is stripped to separate the direct sound wave sequence that is directly transmitted to the centralized sensing module along the main pipe, and the first-order endpoint reflected wave sequence that is reflected at the centralized sensing module and transmitted back via secondary reflection at the fault point; the discrete cross-correlation algorithm is used to calculate the extreme value of the similarity between the two sets of sequences on the time axis and obtain the corresponding first-order round-trip time delay. S42: Compensate for the nonlinear distortion of sound wave propagation velocity caused by sudden changes in local fluid density in the pipeline by synchronously extracting pressure characteristic parameters; the pressure characteristic parameters include the transient dynamic pressure difference extreme value and the absolute value of the steady-state vacuum degree of the pipeline network. S43: Based on the first-order round-trip time delay parameter and pressure characteristic parameter, the nonlinear sound velocity drift error caused by fluid compressibility and transient shock wave is compensated by the aeroacoustic time-domain inversion formula, and the absolute physical pipeline distance from the source of abnormal fluctuation to the centralized sensing module is obtained.

[0012] Furthermore, the first-order round-trip time delay is calculated as follows: ; ; In the formula, The magnitude of the discrete cross-correlation sequence; The number of sampling points within the time window length for cross-correlation calculation; For the first direct sound wave sequence The sound pressure amplitude at each sampling point; For sliding offset The sound pressure amplitude of the corresponding sampling point in the first-order endpoint reflected sound wave sequence after each step size; The optimal offset step size index value; This refers to the actual sampling frequency of the analog-to-digital converter. This is a first-order round-trip time delay; The absolute distance between the source of the abnormal fluctuation and the physical pipeline of the centralized sensing module is calculated as follows: ; In the formula, The absolute physical pipeline distance between the source of the abnormal fluctuation and the centralized sensing module; This is the theoretical phase velocity of sound wave propagation in a gas-liquid mixture; This is a first-order round-trip time delay; , These represent the extreme values ​​of transient dynamic pressure difference and the absolute value of steady-state vacuum, respectively. This is the Mach number distortion correction coefficient.

[0013] Further, step S5 includes: S51: Abstract the vacuum pipeline network into a directed weighted graph consisting of nodes and edges. Based on the absolute distance of the physical pipeline, traverse all path ends that start from the centralized sensing module and whose cumulative weight is within the error tolerance range in the topology database. S52: When the distance matching residuals of multiple paths are all less than a set threshold, based on the identified fault category characteristics, the end device branches with mismatched types are automatically eliminated, and the confidence score of each candidate node as a real fault point is obtained; wherein, the confidence score of each candidate node as a real fault point is calculated as follows: ; In the formula, For the first Each candidate node is a confidence score for a real fault point; For the first Device type matching factor for each candidate node; For the first Device type matching factor for each candidate node; For the first Distance matching residuals of candidate paths; For the first Distance matching residuals of candidate paths; The normalized scaling reference constant for the distance matching residual; , These are the discrete index variables for the local summation of the denominator and the discrete index variables for the numerator, respectively. This represents the total number of candidate nodes that fall within the distance error tolerance range. S53: Obtain the static attribute information corresponding to the candidate node with the highest confidence score. The static attribute information includes the device unique code and spatial location identifier.

[0014] A second aspect of the present invention provides a passive end-of-pipe fault location system for vacuum pipelines, which applies the fault location method described above, characterized in that it includes: The terminal equipment group, including multiple vacuum collection nodes, is configured as a passive silent node in the pipeline monitoring network; The centralized sensing module is deployed at the main pipeline convergence point of the vacuum pipeline network. It includes a wideband acoustic sensor array, a high-frequency micro-pressure sensor, a static absolute pressure sensor, and an active audio frequency generator. It is used to continuously and in parallel collect transient aerodynamic acoustic wave signals and dynamic pressure difference signals generated by fluid dynamics in the pipeline network, and to emit sweep frequency detection sound waves when the hydraulics are idle. The edge computing and solution gateway communicates with the centralized sensing module and contains a feature extraction module and a localization solver. It is used to identify abnormal waveforms based on the collected time-series signals and inversely calculate the physical distance to the fault source. The monitoring and topology mapping platform communicates with the edge computing and solution gateway. It internally stores a three-dimensional topology database of the vacuum pipeline network, which is used to map the physical distance of the inverted solution to the three-dimensional coordinates of the end device for positioning.

[0015] A third aspect of the present invention provides a computer-readable storage medium comprising a stored computer program, wherein the computer program, when executed by a processor, controls the device containing the storage medium to perform the fault location method as described above.

[0016] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: 1. The fault location method of the present invention, by making full use of the inherent physical signal characteristics inside the vacuum pipe network, has the characteristics of being passive, environmentally friendly and reliable. It is particularly suitable for application scenarios with extremely high system reliability requirements, such as subway stations and underground integrated pipe corridors, and provides an innovative solution for the intelligent operation and maintenance of vacuum drainage systems.

[0017] 2. The fault location method of the present invention deploys centralized sensing modules that have undergone extreme signal-to-noise ratio verification at key nodes of the main pipeline, and performs network silencing and electrical isolation configuration on the terminal equipment. This completely eliminates the potential faults and high maintenance costs of the terminal electronic nodes from a physical level. It also provides a standardized data environment with high signal-to-noise ratio and no electrical interference for subsequent fault type identification and accurate location using aeroacoustic time-domain inversion technology, thereby ensuring real-time response speed, positioning accuracy and operational stability under complex and variable vacuum pipeline network conditions.

[0018] 3. The fault location method of the present invention periodically verifies the acoustic impedance topology fingerprint of the pipeline network by actively transmitting audio frequencies and receiving characteristic echoes. This ensures that when the physical topology changes, it can identify, intercept false alarms, and trigger a baseline reconstruction alarm as soon as possible. This ensures that the spatial mapping relationship on which the inversion location algorithm depends always remains strictly consistent with the current real physical pipeline network. This fundamentally eliminates the risk of location failure due to sudden engineering situations such as line renovations, and greatly improves the adaptability and robustness of the passive monitoring network in complex scenarios.

[0019] 4. The fault location method of the present invention extracts the spectral centroid of the frequency centroid of quantized acoustic wave energy through fast Fourier transform, and combines it with time-domain envelope tracking to calculate the transient energy attenuation rate of the physical hysteresis characteristics of quantized energy dissipation, thereby constructing a fault judgment matrix with high robustness. This enables the effective shielding of background pipeline noise interference and accurate locking of effective abnormal characteristic waveforms, thereby achieving automatic and accurate classification of specific fault categories such as steady-state leakage, transient blockage, or mechanical jamming, providing a highly pure feature data source for subsequent inversion spatial location.

[0020] 5. The fault location method of the present invention accurately separates the true first-order round-trip time delay of the direct wave and the reflected wave under strong background noise interference through the cross-correlation function, and simultaneously extracts the transient dynamic pressure extreme value and steady-state vacuum degree to construct distortion correction weights. In this way, the nonlinear sound velocity drift error caused by fluid compressibility and transient shock wave is dynamically offset in the underlying mathematical model, so as to fundamentally ensure that the physical pipeline distance calculated by the inversion solution has extremely high absolute spatial accuracy, and provides the most core one-dimensional distance benchmark for finally locating the passive fault end in the three-dimensional static topology database with zero error.

[0021] 6. The fault location method of the present invention traverses the topology database through depth-first search, matches all candidate devices that meet the distance tolerance, and introduces the specific fault type identified in the previous steps as a logical filtering operator to eliminate interference nodes with inconsistent types from many ambiguous branches, and finally locks the fault device with unique spatial coordinates, thereby transforming the complex underlying data into a visualized fault location and dispatch guide that can be directly called by operation and maintenance personnel. Attached Figure Description

[0022] Figure 1 This is a schematic flowchart of the fault location method according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the logic of the fault location method executed by the fault location system in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0024] Example 1 Please refer to Figure 1 This invention provides a method for locating faults at the passive end of a vacuum pipeline network, comprising the following steps: S1: Deploy centralized sensing modules at the main pipeline network nodes of the vacuum drainage system. The centralized sensing modules include at least an acoustic sensor array and a dynamic pressure sensor. S2: Through the centralized sensing module, continuously and in real time collect aerodynamic acoustic wave signals and pressure transient signals generated inside the vacuum pipeline under the coupling effect of gas-liquid two-phase fluid dynamics; S3: Extract features from the acquired time-series waveform signals, identify the characteristic waveform signals caused by the abnormality of the terminal equipment based on the preset fault classification model, and determine the fault category. S4: Extract the first-order round-trip time delay and waveform distortion parameters of the characteristic waveform signal propagating in the pipeline network, and use the aeroacoustic time-domain inversion algorithm to calculate the physical pipeline distance from the anomaly source point to the centralized sensing module; S5: The calculated physical pipeline distance is correlated with the three-dimensional vacuum pipeline static topology database to locate the specific end device node where the anomaly occurred.

[0025] It should be noted that the passive end is a terminal device node in the vacuum drainage system that does not actively transmit detection signals but only passively receives monitoring.

[0026] Specifically, traditional vacuum drainage system fault monitoring heavily relies on deploying expensive sensors and complex communication networks on each end device. This not only leads to extremely high hardware construction and maintenance costs, but also exposes the end devices to harsh environments characterized by high corrosion and humidity, which can easily cause electronic component failures. Furthermore, when attempting centralized monitoring, existing technologies lack methods to scientifically guide the selection of sensor installation nodes and cannot effectively eliminate electromagnetic or signal interference generated by the local electrical circuits of the end devices on the acoustic signals of the main pipeline fluid. This results in potential blind spots in the monitoring range, or the quality of the raw data collected may not meet the requirements of subsequent high-precision algorithms. Therefore, step S1 of this embodiment constructs a physical system architecture where the end device group is silent and the main pipeline is centrally sensed. Wherein: Step S1 includes: S11: Before determining the final installation location of the centralized sensing module, perform a verification calculation of the ultimate acoustic signal-to-noise ratio of the pipeline network to ensure that it meets the preset signal-to-noise ratio safety threshold. In an optional embodiment, the calculation must satisfy the condition that the calculated limit signal-to-noise ratio verification value is greater than a preset signal-to-noise ratio safety threshold, and the calculation method is as follows: ; In the formula, This is the limit signal-to-noise ratio verification value; The extreme values ​​of the initial aerodynamic sound pressure level induced by a typical terminal fault; This is the acoustic frequency attenuation coefficient of the gas-liquid two-phase medium in the current vacuum pipeline; The longest physical propagation path length from the selected deployment node to the farthest passive device in the pipeline network; The equivalent sound pressure level of the ambient background noise at the selected deployment node.

[0027] In an optional embodiment, to balance economy and accuracy of acoustic feature capture, the preset signal-to-noise ratio (SNR) safety threshold in the pipeline network limit acoustic SNR verification calculation ranges from 3dB to 10dB. Preferably, the preset SNR safety threshold is 6dB. It is understood that during use, if calculated using the SNR calculation formula, the target deployment node's... If the noise level is less than 3dB, it indicates that the node is too far from the passive end or the background noise is too high. It needs to be moved forward towards the passive end to find a secondary trunk node as a potential deployment point. If the target deployment node's... If the value is greater than 6dB, it confirms that the centralized sensing module can effectively cover and solve the characteristic waveform distortion caused by the abnormality of the farthest passive device.

[0028] S12: At the node that meets the requirements of the ultimate acoustic signal-to-noise ratio verification calculation, a linear array consisting of at least two broadband acoustic microphones is installed on the wall in a non-intrusive manner. S13: At the opening or reserved interface of the adjacent main pipe, a low-frequency static absolute pressure sensor is installed in an invasive manner to read the steady-state absolute value of the vacuum network in real time; then, high-frequency micro-pressure sensors are installed in parallel in an invasive manner to synchronously capture the transient pressure distortion wave propagating in the medium inside the pipe. In an optional embodiment, to ensure distortion-free capture of abnormal acoustic signals, the sampling frequency of the analog-to-digital conversion unit configured within the centralized sensing module satisfies the following formula: ; In the formula, This is the actual sampling frequency of the analog-to-digital conversion unit; This represents the highest characteristic frequency boundary of known end faults.

[0029] S14: Deploy an edge computing and solution gateway in the electrical control cabinet closest to the centralized sensing module, and build a high-speed data acquisition link with hardwired connection between the two.

[0030] It should be noted that the edge computing and resolution gateway is a common data processing device in the prior art. In this embodiment, a shielded coaxial cable or industrial-grade twisted pair is used to physically connect the analog signals output by the acoustic sensor array, low-frequency static absolute pressure sensor, and high-frequency micro-pressure sensor in the centralized sensing module to the analog input channel of the high-speed data acquisition card built into the edge computing and resolution gateway. At the same time, the driver interface of the active audio generator is connected to the digital / analog conversion output channel of the gateway. Thus, by using a local wired direct connection and electromagnetic interference-proof grounding configuration at the physical layer, it is ensured that the high-throughput acoustic timing signal can be continuously and losslessly imported into the memory buffer of the edge gateway with a physical delay of less than milliseconds, providing a foundation of underlying data streams with accurate timestamps for subsequent data analysis.

[0031] In an optional embodiment, all end devices within the vacuum drainage system are configured as telemetry nodes in a passive, silent state within the pipeline monitoring network. That is, the end devices do not deploy electrical sensors or network modules for collecting pipeline fluid dynamics parameters and performing uplink data communication with the centralized sensing module. Their own control loops, used only for local action triggering, are not connected to the fault location data link of the centralized sensing module. It is understood that during use, by physically blocking the fluid dynamics data acquisition and uplink communication links, and strictly isolating the electrical closed-loop control loops used only for maintaining basic local functions, the dependence on vulnerable electronic components in harsh end-user environments can be fundamentally eliminated. This eliminates data interference from the end-user electrical state on the centralized acoustic inversion algorithm, and ensures that accurate positioning is achieved entirely and exclusively based on the aeroacoustic time-domain inversion algorithm of the main pipeline.

[0032] Specifically, the inversion solution for spatial positioning based on one-dimensional acoustic wave propagation delay highly depends on the absolute accuracy of the underlying three-dimensional static pipeline topology database. However, in actual long-term engineering operations, vacuum pipelines are highly susceptible to changes in the actual acoustic propagation path due to artificial physical structural alterations such as local pipeline maintenance and renovation, branch additions and deletions, or pipe segment cuts. If the system lacks the dynamic perception capability for changes in the physical pipeline lines and passively relies only on the initially fixed static mapping topology, any unreported physical structural changes will inevitably lead to a severe mapping misalignment between the inverted fault distance and the actual spatial coordinates, causing a sharp collapse in the accuracy and reliability of the entire positioning system over the long term. Therefore, step S2 in this embodiment introduces an active frequency sweep detection and topology overlap calculation mechanism during pipeline idle periods, giving the positioning system the ability to actively perceive changes in the physical entity pipeline structure. Wherein: Step S2 includes: S21: Read the values ​​of the static absolute pressure sensor in real time. When the fluctuation amplitude of the pipeline vacuum degree within the continuous observation time window is less than the preset minimum value, and the monitoring network does not receive any pump station vacuum start / stop command, it is determined that the pipeline has entered a hydraulic idle state. Then, through the active audio frequency generator integrated in the centralized sensing module and acoustically coupled to the main pipeline, a linear sweep frequency detection sound wave with a known bandwidth and duration is emitted into the main pipeline to obtain the physical structure reflection response at different spatial scales in the pipeline. In an optional embodiment, the active audio generator is an externally clamped piezoelectric ceramic sound-generating unit, which is physically coupled to the outer wall of the vacuum main pipe in a non-invasive manner; the active audio generator excites the pipe wall through high-frequency mechanical vibration and couples and emits sweep frequency detection sound waves into the medium inside the pipe, thereby avoiding direct contact with corrosive fluids and solid contaminants inside the pipe.

[0033] It should be noted that, in this embodiment, the preset minimum value ranges from 100 Pa to 1000 Pa.

[0034] S22: While transmitting the detection sound wave, the acoustic sensor array simultaneously opens the acquisition time window to receive the transient sound pressure signal reflected back from the physical topology of the pipeline network; the digital signal processing module inside the edge computing and solution gateway performs bandpass filtering and noise reduction and analog-to-digital conversion on the acquired analog signal to generate a discrete time-series test sequence, and performs topological overlap detection with the pre-stored reference acoustic reflection feature sequence. In an optional embodiment, the topological overlap detection method is as follows: ; In the formula, This is the topological overlap coefficient; This represents the total number of sampling points within the acoustic detection time window. The first time-aligned reference acoustic reflection feature sequence The amplitude of each discrete sampling point; The first in the latest test echo characteristic sequence The amplitude of each discrete sampling point; This is the arithmetic mean of the baseline sequence array; This is the arithmetic mean of the test sequence array.

[0035] It should be noted that the topological overlap coefficient ranges from -1 to 1, and the closer it is to 1, the better the current physical network matches the historical health model; This refers to the acoustic impedance topological fingerprint of the pipeline network in its known and manually verified initial state; the... After the active audio generator emits a swept-frequency probe wave, it is captured in real time by an acoustic sensor array and then filtered by a low-frequency ambient noise filter to generate one-dimensional sound pressure array data for the current test period.

[0036] S23: The edge computing and solution gateway will logically compare the obtained topology overlap coefficient with the preset topology security verification threshold. If the topology overlap coefficient is greater than or equal to the security verification threshold, it is determined that the current physical pipeline structure has not been artificially modified, and the original three-dimensional vacuum pipeline static topology database mapping relationship remains unchanged. If the topology overlap coefficient is less than the security verification threshold, it is determined that the pipeline has undergone physical line modification, automatically intercepts and freezes the original spatial distance positioning solution service, outputs a high-level alarm prompt for topology reconstruction and line change, and archives the abnormal test sequence collected this time as a temporary benchmark. After manual verification and entry of new pipeline building information model parameters, the benchmark mapping relationship in the local three-dimensional vacuum pipeline static topology database will be overwritten and updated again.

[0037] It should be noted that, in this embodiment, the preset topology security verification threshold is a constant generated by dynamic calibration based on the lowest correlation coefficient obtained from multiple health echo tests during the system commissioning and debugging phase, after deducting the security margin.

[0038] Specifically, the raw physical signals collected by sensing devices deployed within the vacuum pipeline network are often superimposed with a large amount of high-intensity broadband background noise generated by the operation of vacuum pumps and the friction between normal sewage discharge gas and liquid. In this complex fluid dynamic environment with strong noise and strong coupling, the initial waveform characteristics of different types of passive end-point faults (such as continuous micro-leakage and instantaneous hard object blockage) are easily confused with each other or submerged by background noise. Simply relying on traditional one-dimensional time-domain amplitude or simple physical quantity threshold judgment cannot accurately decouple signal characteristics, causing existing monitoring systems to not only easily generate false alarms and missed alarms, but also to be unable to accurately distinguish specific fault categories, resulting in the generated alarm information lacking specificity and failing to guide maintenance personnel to carry the correct tools for efficient emergency repairs. Therefore, step S3 of this embodiment introduces a time-frequency dual-domain feature decoupling algorithm to map the messy discrete one-dimensional time-series data into a high-dimensional mathematical feature classification space. Among them: Step S3 includes: S31: The discrete time-series data frame of the corresponding acoustic data is generated through edge computing and solution gateway, and a Hanning window is applied to it to suppress spectral leakage. Then, a fast Fourier transform is performed to convert the one-dimensional time domain signal into a frequency domain amplitude matrix and obtain the frequency domain feature quantity spectral centroid. In an optional embodiment, to distinguish between the high-frequency sound caused by air leakage and the low-frequency sound caused by blockage, the centroid of the spectrum is extracted by calculating using the following formula: ; In the formula, This represents the centroid frequency value of the spectrum. This represents the number of sampling points for the Fast Fourier Transform. For discrete frequency indexing; For the first The center frequency corresponding to each frequency point; For the first The absolute amplitude corresponding to each frequency point.

[0039] S32: While extracting features in the frequency domain, envelope tracking is performed on the same data frames in the time domain to lock the initial peak moment of the abnormal event and extract the transient energy decay rate, a time domain feature. In an optional embodiment, the transient energy decay rate is calculated using the following formula: ; In the formula, Transient energy decay rate; The length of the characteristic observation time window; This is the initial peak sound pressure level when the abnormal event is triggered. , These represent the acoustic energy at the initial peak time and the delayed observation time, respectively.

[0040] S33: Input the spectral centroid and transient energy decay rate into the preset fault classification matrix; if the spectral centroid is greater than the preset high-frequency threshold and the transient energy decay rate is lower than the preset decay safety threshold, it is a continuous leakage fault; if the spectral centroid is less than the preset low-frequency threshold and the transient energy decay rate is greater than the preset rapid decay threshold, it is a sudden blockage fault; if the frequency domain amplitude is discrete and multi-harmonic and the time domain decay rate shows non-convergent fluctuations within the observation window, it is an end mechanical jamming fault.

[0041] It should be noted that the preset high-frequency threshold is the frequency limit for distinguishing between low-frequency white noise from normal fluid flow and high-frequency components generated by abnormal airflow friction, and its value ranges from 2000Hz to 3000Hz, and is related to the pipe material; the preset attenuation safety threshold is the boundary of the energy change ratio per unit time for distinguishing between continuous signals and disappearing signals, and its value ranges from 0.1s. -1 Up to 0.5 -1 .

[0042] Specifically, traditional positioning algorithms based on fixed sound velocity and simple time delay calculations suffer from serious fundamental ranging defects when dealing with complex vacuum drainage pipe networks. Vacuum pipe networks are filled with low-pressure gas-liquid two-phase mixtures. When passive equipment at the end experiences blockages, impacts, or sudden leaks, it generates strong transient pressure shock waves, causing drastic changes in local fluid density within the pipe. This leads to severe nonlinear distortion of the sound wave propagation velocity, i.e., Mach number distortion. Furthermore, the strong background noise superimposed within the main pipe easily obscures the true arrival boundaries of the characteristic waves. If uncompensated, directly applying a constant sound velocity under ideal conditions and a simple peak time difference for spatial ranging will result in unacceptably large errors in physical distance calculation, causing subsequent spatial node mapping to completely fail. Therefore, step S4 of this embodiment introduces a discrete cross-correlation algorithm and a multi-dimensional parametric distortion compensation mechanism to construct a high-precision aeroacoustic time-domain inversion positioning model that can adapt to complex fluid dynamic environments. Wherein: Step S4 includes: S41: The characteristic waveform signal is stripped to separate the direct sound wave sequence that is directly transmitted to the centralized sensing module along the main pipe, and the first-order endpoint reflected wave sequence that is reflected at the centralized sensing module and transmitted back via secondary reflection at the fault point; the discrete cross-correlation algorithm is used to calculate the extreme value of the similarity between the two sets of sequences on the time axis and obtain the corresponding first-order round-trip time delay. In an optional embodiment, the first-order round-trip time delay is calculated as follows: ; ; In the formula, The magnitude of the discrete cross-correlation sequence; The number of sampling points within the time window length for cross-correlation calculation; For the first direct sound wave sequence The sound pressure amplitude at each sampling point; For sliding offset The sound pressure amplitude of the corresponding sampling point in the first-order endpoint reflected sound wave sequence after each step size; The optimal offset step size index value; This refers to the actual sampling frequency of the analog-to-digital converter. This is the first-order round-trip time delay.

[0043] S42: Compensate for the nonlinear distortion of sound wave propagation velocity caused by sudden changes in local fluid density in the pipeline by synchronously extracting pressure characteristic parameters; the pressure characteristic parameters include the transient dynamic pressure difference extreme value and the absolute value of the steady-state vacuum degree of the pipeline network. It should be noted that, in this embodiment, the extreme value of transient dynamic pressure difference is extracted by calculating the absolute difference between the amplitude of the highest point of the transient pressure wave peak and the amplitude of the lowest point of the transient pressure wave captured by the high-frequency micro-pressure sensor within the trigger time window; the absolute value of the steady-state vacuum degree of the pipeline network is obtained by reading in real time from a conventional static absolute pressure sensor installed at the centralized sensing module and calculating the sliding arithmetic mean within the idle time window before the fault is triggered.

[0044] S43: Based on the first-order round-trip time delay parameter and pressure characteristic parameter, the nonlinear sound velocity drift error caused by fluid compressibility and transient shock wave is compensated by the aeroacoustic time-domain inversion formula, and the absolute physical pipeline distance from the source of abnormal fluctuation to the centralized sensing module is obtained.

[0045] In an optional embodiment, the absolute distance between the source of the abnormal fluctuation and the physical pipeline of the centralized sensing module is calculated as follows: ; In the formula, The absolute physical pipeline distance between the source of the abnormal fluctuation and the centralized sensing module; This is the theoretical phase velocity of sound wave propagation in a gas-liquid mixture; This is a first-order round-trip time delay; , These are the extreme values ​​of transient dynamic pressure difference and the absolute value of steady-state vacuum, respectively. ; This is the Mach number distortion correction coefficient.

[0046] It should be noted that, The active audio generator in the centralized sensing module emits high-frequency calibration sound waves into the tube, which are received by a wideband acoustic sensor array. The edge computing and resolution gateway dynamically obtains the calibration sound waves by performing a real-time division operation based on the flight time of the calibration sound waves within the known physical installation distance between the two. During the initial commissioning and joint debugging phase, technicians artificially generated standard fault pulses at the end nodes with known three-dimensional coordinates, recorded the deviation data, and after calibration by univariate linear regression fitting, permanently stored the calibration constants in the local non-volatile database of the edge computing gateway.

[0047] Specifically, in complex vacuum pipe networks containing numerous branches, tees, and overlapping pipelines, the single one-dimensional distance parameter obtained through aeroacoustic inversion often faces severe spatial ambiguity due to the one-distance-multiple-point distance. Starting from the centralized sensing module, the same pipeline physical length may simultaneously point to multiple end devices located on different floors or in different service zones within the network. Without establishing a rigorous topological association and feature constraint resolution mechanism, the monitoring system will be unable to accurately pinpoint the true fault origin from multiple ambiguous branches with similar distances, easily leading to fatal deviations in the location results. This not only renders intelligent monitoring alarms ineffective but also results in ineffective on-site troubleshooting and wasted time for maintenance personnel. Therefore, step S5 of this embodiment constructs a graph-based path search algorithm and a fault category feature constraint model to achieve accurate cross-dimensional mapping from abstract one-dimensional pipeline distances to specific three-dimensional physical device nodes. Wherein: Step S5 includes: S51: Abstract the vacuum pipeline network into a directed weighted graph consisting of nodes and edges. Based on the absolute distance of the physical pipeline, traverse all path ends that start from the centralized sensing module and whose cumulative weight is within the error tolerance range in the topology database. In an optional embodiment, the distance matching residual of each candidate path is calculated as follows: ; In the formula, For the first Distance matching residuals of candidate paths; The index number of the candidate path; The absolute physical pipeline distance between the source of the abnormal fluctuation and the centralized sensing module; This represents the total number of pipe segments traversed from the centralized sensing module to the candidate end device. This is the index number of the pipe segment in a single candidate path; For the first The first candidate path The physical spatial length of the pipe segment.

[0048] S52: When the distance matching residuals of multiple paths are all less than the set threshold, based on the identified fault category features, the end device branches that do not match the type are automatically removed, and the confidence score of each candidate node as a real fault point is obtained. In an optional embodiment, the confidence score of each candidate node as a real fault point is calculated as follows: ; In the formula, For the first Each candidate node is a confidence score for a real fault point; For the first Device type matching factor for each candidate node; For the first Device type matching factor for each candidate node; For the first Distance matching residuals of candidate paths; For the first Distance matching residuals of candidate paths; The normalized scaling reference constant for the distance matching residual; , These are the discrete index variables for the local summation of the denominator and the discrete index variables for the numerator, respectively. This represents the total number of candidate nodes that fall within the distance error tolerance range.

[0049] It should be noted that the device type matching factor is generated by logically comparing the fault category features identified in the previous steps with the device attribute tags of the target node in the topology database. If the attributes match, the value is one; otherwise, the value is zero.

[0050] S53: Obtain the static attribute information corresponding to the candidate node with the highest confidence score. The static attribute information includes the device unique code and spatial location identifier.

[0051] In an optional embodiment, the calculation results are mapped to three-dimensional physical coordinates in the building information model in real time, driving the corresponding device icon on the monitoring interface to enter a high-brightness alarm state, and calculating the list of affected downstream branches based on the fault location and topological connectivity, and simultaneously generating and pushing a fault report containing standardized handling suggestions.

[0052] It should be noted that in this embodiment, all the devices used are common acoustic monitoring devices in the prior art. In other embodiments, other types of monitoring devices may also be used, which are not specifically limited here.

[0053] Example 2 Please refer to Figure 2 This invention provides a passive end-of-pipe fault location system for vacuum pipelines. The terminal equipment group, including multiple vacuum collection nodes, is configured as a passive silent node in the pipeline monitoring network; The centralized sensing module is deployed at the main pipeline convergence point of the vacuum pipeline network. It includes a wideband acoustic sensor array, a high-frequency micro-pressure sensor, a static absolute pressure sensor, and an active audio frequency generator. It is used to continuously and in parallel collect transient aerodynamic acoustic wave signals and dynamic pressure difference signals generated by fluid dynamics in the pipeline network, and to emit sweep frequency detection sound waves when the hydraulics are idle. The edge computing and solution gateway communicates with the centralized sensing module and contains a feature extraction module and a localization solver. It is used to identify abnormal waveforms based on the collected time-series signals and inversely calculate the physical distance to the fault source. The monitoring and topology mapping platform communicates with the edge computing and solution gateway. It internally stores a three-dimensional topology database of the vacuum pipeline network, which is used to map the physical distance of the inverted solution to the three-dimensional coordinates of the end device for positioning.

[0054] In an optional embodiment, the passive silent node does not deploy electrical sensors and network modules for collecting fluid parameters and uplink communication, and its local action triggering control loop is not connected to the fault location data link.

[0055] Other technical features are the same as in Embodiment 1 and can achieve the same technical effects, so they will not be described in detail here.

[0056] It should be noted that the passive end-of-pipe fault location system provided in this embodiment can be a computer program (including program code) running on a computer device. For example, a passive end-of-pipe fault location system for a vacuum pipeline is an application program that can be used to execute the corresponding steps in the method provided in the embodiment of the present invention.

[0057] In some feasible implementations, the vacuum pipeline passive end-of-line fault location system provided in this embodiment can be implemented using a combination of hardware and software. As an example, the vacuum pipeline passive end-of-line fault location system of this embodiment can be a processor in the form of a hardware decoding processor, which is programmed to execute the vacuum pipeline passive end-of-line fault location method provided in this embodiment. For example, the processor in the form of a hardware decoding processor can be one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0058] In some feasible implementations, the passive end-of-pipe fault location system for vacuum pipelines provided in this embodiment can be implemented in software. It can be software in the form of programs and plug-ins, and includes a series of modules to implement the passive end-of-pipe fault location method for vacuum pipelines provided in this embodiment of the invention.

[0059] Example 3 This invention also provides a computer-readable storage medium storing a computer program that is executed by a processor to implement the various steps in the fault location method described above. For details, please refer to the implementation methods provided for each of the above steps, which will not be repeated here.

[0060] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0061] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0062] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0063] In this invention, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes said element.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit it; those skilled in the art will readily understand that the above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for locating faults at the passive end of a vacuum pipeline network, characterized in that, Includes the following steps: S1: Deploy centralized sensing modules at the main pipeline network nodes of the vacuum drainage system. The centralized sensing modules include at least an acoustic sensor array and a dynamic pressure sensor. S2: Through the centralized sensing module, continuously and in real time collect aerodynamic acoustic wave signals and pressure transient signals generated inside the vacuum pipeline under the coupling effect of gas-liquid two-phase fluid dynamics; S3: Extract features from the acquired time-series waveform signals, identify the characteristic waveform signals caused by the abnormality of the terminal equipment based on the preset fault classification model, and determine the fault category. S4: Extract the first-order round-trip time delay and waveform distortion parameters of the characteristic waveform signal propagating in the pipeline network, and use the aeroacoustic time-domain inversion algorithm to calculate the physical pipeline distance from the anomaly source point to the centralized sensing module; S5: The calculated physical pipeline distance is correlated with the three-dimensional vacuum pipeline static topology database to locate the specific end device node where the anomaly occurred.

2. The fault location method according to claim 1, characterized in that, Step S1 includes: S11: Before determining the final installation location of the centralized sensing module, perform a verification calculation of the ultimate acoustic signal-to-noise ratio of the pipeline network to ensure that it meets the preset signal-to-noise ratio safety threshold. S12: At the node that meets the requirements of the ultimate acoustic signal-to-noise ratio verification calculation, a linear array consisting of at least two broadband acoustic microphones is installed on the wall in a non-intrusive manner. S13: At the opening or reserved interface of the adjacent main pipe, a low-frequency static absolute pressure sensor is installed in an invasive manner to read the steady-state absolute value of the vacuum network in real time; then, high-frequency micro-pressure sensors are installed in parallel in an invasive manner to synchronously capture the transient pressure distortion wave propagating in the medium inside the pipe. S14: Deploy an edge computing and solution gateway in the electrical control cabinet closest to the centralized sensing module, and build a high-speed data acquisition link with hardwired connection between the two.

3. The fault location method according to claim 2, characterized in that, Step S2 includes: S21: Read the values ​​of the static absolute pressure sensor in real time. When the fluctuation amplitude of the pipeline vacuum degree within the continuous observation time window is less than the preset minimum value, and the monitoring network does not receive any pump station vacuum start / stop command, it is determined that the pipeline has entered a hydraulic idle state. Then, through the active audio frequency generator integrated in the centralized sensing module and acoustically coupled to the main pipeline, a linear sweep frequency detection sound wave with a known bandwidth and duration is emitted into the main pipeline to obtain the physical structure reflection response at different spatial scales in the pipeline. S22: While transmitting the detection sound wave, the acoustic sensor array simultaneously opens the acquisition time window to receive the transient sound pressure signal reflected back from the physical topology of the pipeline network; the digital signal processing module inside the edge computing and solution gateway performs bandpass filtering and noise reduction and analog-to-digital conversion on the acquired analog signal to generate a discrete time-series test sequence, and performs topological overlap detection with the pre-stored reference acoustic reflection feature sequence. S23: Logically compare the obtained topology overlap coefficient with the preset topology safety verification threshold; if the topology overlap coefficient is greater than or equal to the safety verification threshold, it is determined that the current physical pipeline structure has not been artificially modified, and the original three-dimensional vacuum pipeline static topology database mapping relationship remains unchanged; if the topology overlap coefficient is less than the safety verification threshold, it is determined that the pipeline has undergone physical line modification, the original spatial distance positioning calculation service is automatically intercepted and frozen, and a high-level alarm prompt for topology reconstruction and line change is output. At the same time, the abnormal test sequence collected this time is archived as a temporary benchmark, and after manual verification and entry of new pipeline building information model parameters, the benchmark mapping relationship in the local three-dimensional vacuum pipeline static topology database is overwritten and updated again.

4. The fault location method according to claim 3, characterized in that, The topological overlap detection method is as follows: ； In the formula, This is the topological overlap coefficient; This represents the total number of sampling points within the acoustic detection time window. The first time-aligned reference acoustic reflection feature sequence The amplitude of each discrete sampling point; The first in the latest test echo characteristic sequence The amplitude of each discrete sampling point; This is the arithmetic mean of the baseline sequence array; This is the arithmetic mean of the test sequence array.

5. The fault location method according to claim 3, characterized in that, Step S3 includes: S31: The discrete time-series data frame of the corresponding acoustic data is generated through edge computing and solution gateway, and a Hanning window is applied to it to suppress spectral leakage. Then, a fast Fourier transform is performed to convert the one-dimensional time domain signal into a frequency domain amplitude matrix and obtain the frequency domain feature quantity spectral centroid. S32: While extracting features in the frequency domain, envelope tracking is performed on the same data frames in the time domain to lock the initial peak moment of the abnormal event and extract the transient energy decay rate, a time domain feature. S33: Input the spectral centroid and transient energy decay rate into the preset fault classification matrix; if the spectral centroid is greater than the preset high-frequency threshold and the transient energy decay rate is lower than the preset decay safety threshold, it is a continuous leakage fault; if the spectral centroid is less than the preset low-frequency threshold and the transient energy decay rate is greater than the preset rapid decay threshold, it is a sudden blockage fault; if the frequency domain amplitude is discrete and multi-harmonic and the time domain decay rate shows non-convergent fluctuations within the observation window, it is an end mechanical jamming fault.

6. The fault location method according to claim 5, characterized in that, Step S4 includes: S41: The characteristic waveform signal is stripped to separate the direct sound wave sequence that is directly transmitted to the centralized sensing module along the main pipe, and the first-order endpoint reflected wave sequence that is reflected at the centralized sensing module and transmitted back via secondary reflection at the fault point; the discrete cross-correlation algorithm is used to calculate the extreme value of the similarity between the two sets of sequences on the time axis and obtain the corresponding first-order round-trip time delay. S42: Compensate for the nonlinear distortion of sound wave propagation velocity caused by sudden changes in local fluid density in the pipeline by synchronously extracting pressure characteristic parameters; the pressure characteristic parameters include the transient dynamic pressure difference extreme value and the absolute value of the steady-state vacuum degree of the pipeline network. S43: Based on the first-order round-trip time delay parameter and pressure characteristic parameter, the nonlinear sound velocity drift error caused by fluid compressibility and transient shock wave is compensated by the aeroacoustic time-domain inversion formula, and the absolute physical pipeline distance from the source of abnormal fluctuation to the centralized sensing module is obtained.

7. The fault location method according to claim 6, characterized in that, The first-order round-trip time delay is calculated as follows: ； ； In the formula, The magnitude of the discrete cross-correlation sequence; The number of sampling points within the time window length for cross-correlation calculation; For the first direct sound wave sequence The sound pressure amplitude at each sampling point; For sliding offset The sound pressure amplitude of the corresponding sampling point in the first-order endpoint reflected sound wave sequence after each step size; The optimal offset step size index value; This refers to the actual sampling frequency of the analog-to-digital converter. This is a first-order round-trip time delay; The absolute distance between the source of the abnormal fluctuation and the physical pipeline of the centralized sensing module is calculated as follows: ； In the formula, The absolute physical pipeline distance between the source of the abnormal fluctuation and the centralized sensing module; This is the theoretical phase velocity of sound wave propagation in a gas-liquid mixture; This is a first-order round-trip time delay; , These represent the extreme values ​​of transient dynamic pressure difference and the absolute value of steady-state vacuum, respectively. This is the Mach number distortion correction coefficient.

8. The fault location method according to claim 6, characterized in that, Step S5 includes: S51: Abstract the vacuum pipeline network into a directed weighted graph consisting of nodes and edges. Based on the absolute distance of the physical pipeline, traverse all path ends that start from the centralized sensing module and whose cumulative weight is within the error tolerance range in the topology database. S52: When the distance matching residuals of multiple paths are all less than a set threshold, based on the identified fault category characteristics, the end device branches with mismatched types are automatically eliminated, and the confidence score of each candidate node as a real fault point is obtained; wherein, the confidence score of each candidate node as a real fault point is calculated as follows: ； In the formula, For the first Each candidate node is a confidence score for a real fault point; For the first Device type matching factor for each candidate node; For the first Device type matching factor for each candidate node; For the first Distance matching residuals of candidate paths; For the first Distance matching residuals of candidate paths; The normalized scaling reference constant for the distance matching residual; , These are the discrete index variables for the local summation of the denominator and the discrete index variables for the numerator, respectively. This represents the total number of candidate nodes that fall within the distance error tolerance range. S53: Obtain the static attribute information corresponding to the candidate node with the highest confidence score. The static attribute information includes the device unique code and spatial location identifier.

9. A passive end-of-pipe fault location system for a vacuum pipeline network, employing the fault location method as described in any one of claims 1-8, characterized in that, include: The terminal equipment group, including multiple vacuum collection nodes, is configured as a passive silent node in the pipeline monitoring network; The centralized sensing module is deployed at the main pipeline convergence point of the vacuum pipeline network. It includes a wideband acoustic sensor array, a high-frequency micro-pressure sensor, a static absolute pressure sensor, and an active audio frequency generator. It is used to continuously and in parallel collect transient aerodynamic acoustic wave signals and dynamic pressure difference signals generated by fluid dynamics in the pipeline network, and to emit sweep frequency detection sound waves when the hydraulics are idle. The edge computing and solution gateway communicates with the centralized sensing module and contains a feature extraction module and a localization solver. It is used to identify abnormal waveforms based on the collected time-series signals and inversely calculate the physical distance to the fault source. The monitoring and topology mapping platform communicates with the edge computing and solution gateway. It internally stores a three-dimensional topology database of the vacuum pipeline network, which is used to map the physical distance of the inverted solution to the three-dimensional coordinates of the end device for positioning.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program that, when executed by a processor, controls the device containing the storage medium to perform the fault location method as described in any one of claims 1-8.

Images