Pipe network water leakage point detection and distance positioning method based on flow analysis

Abstract

The invention discloses a pipe network water leakage point detection and distance positioning method based on flow analysis. The method comprises the following steps: S1, constructing a flow balance model of a pipeline network; s2, monitoring the flow state of each node of the pipeline network in real time based on the flow balance model constructed in the step S1, and performing preliminary positioning on a leakage point when pipeline leakage is detected; s3, flow gradient analysis and reverse hydraulic fine positioning: in the suspected leakage area locked in the step S2, meter-scale precision positioning of a water leakage point is realized through a flow gradient analysis and reverse hydraulic iterative model; and S4, a plurality of sensor nodes are arranged on the water leakage pipeline determined in the step S3, detection data are collected, the position of a leakage point is determined according to a related positioning fusion algorithm, comprehensive decision making is carried out on the leakage point and the leakage point determined through flow gradient analysis and reverse hydraulic power in the step S3, and finally the accurate position of the pipeline leakage point is judged. The problems of low precision, weak interference resistance and the like of a traditional method can be solved, and accurate detection of leakage points is realized.

Description

Technical Field

[0001] The present invention relates to the field of pipeline leakage detection, and in particular to a method for detecting and locating leakage points in a pipeline network based on flow analysis. Background Art

[0002] Currently, water supply in various regions is primarily delivered through water supply pipelines. However, water pipeline leakage and its effective management have long been a global concern. Pipeline leaks not only waste water resources but also pose a direct threat to regional water supply security and ecological balance. Once a localized leak occurs, quickly locating the leak point is a top priority. In recent years, with the advancement of the Internet of Things (IoT), pipeline leak detection and location technology has rapidly developed. From traditional listening and visual inspection methods to current data-driven machine learning techniques, a variety of methods have emerged.

[0003] Traditional leak detection and location techniques, such as visual inspection, locate pipeline leaks by observing water accumulation on the ground. The acoustic method uses listening rods or other acoustic sensors to capture acoustic signals generated by changes in water pressure in the pipeline. This requires specialized knowledge and experience to determine the leak location, and noise can affect location accuracy. The flow meter method determines whether a system has a leak by calculating changes in fluid flow in a pipeline, but its leak location accuracy is limited. Ground penetrating radar (GPR) technology utilizes the interaction between electromagnetic waves and the underground medium, transmitting high-frequency electromagnetic waves and receiving reflected signals from the ground to detect pipeline leaks. Tracer gas leak detection involves injecting an easily detectable and diffusible tracer gas into the pipeline. The gas diffuses to the leaking area, locating the leak, but this method lacks real-time performance. The pressure gradient method detects leak anomalies by measuring pressure changes along the pipeline, but it cannot pinpoint the leak. Infrared thermal imaging, which detects infrared radiation emitted by an object in a non-contact, real-time manner, measures surface temperature. However, it is susceptible to misidentification of leaks due to background temperature differences. Fiber optic technology uses optical fiber sensors cleverly placed around or inside pipelines for monitoring. However, high costs and sensitivity to environmental factors affect the accuracy of leak detection. However, these single detection technologies limit the accuracy and real-time performance of leak detection in complex environments.

[0004] In recent years, data-driven detection technologies such as support vector machines (SVMs), neural networks (ANNs), convolutional neural networks (CNNs), and recurrent neural networks (RNNs) have emerged. These technologies predict and identify water leaks by training on large amounts of historical data. However, SVMs require a large number of training samples to achieve good performance and have limited ability to handle data scarcity or imbalance. ANNs are unique in their strong nonlinear modeling capabilities, enabling them to capture and learn complex nonlinear relationships. They can automatically extract features from data and have strong generalization capabilities, making them suitable for processing a wide range of data types. However, ANNs' generalization capabilities are limited when dealing with data scarcity or imbalance. CNNs demonstrate strong application potential in complex water leak detection scenarios, especially when collecting information from multiple sensors. However, their training process requires high data volumes and computational resources. RNNs demonstrate unique advantages in the field of water leak detection. They excel at processing sequential data, extracting temporal features and contextual information, and can adapt their network structure (such as using LSTM or GRU units) to meet diverse processing requirements. However, these data-centric methods for detecting pipeline leakage have problems such as poor model interpretability and low trust by deeply mining and analyzing large amounts of data, which limits the accurate positioning of the leakage location.

[0005] In summary, the existing water leakage detection solutions for water supply networks still have defects such as poor positioning accuracy and multiple interferences, and there is room for improvement. A detection method that can accurately detect water leakage points in water supply pipelines is needed. Summary of the Invention

[0006] The present invention proposes a method for detecting and locating water leakage points in a pipe network based on flow analysis, which includes the following steps:

[0007] S1. Construct a flow balance model for the pipeline network;

[0008] S2. Based on the flow balance model constructed in step S1, real-time monitoring of the flow status of each node in the pipeline network, when a pipeline leak is detected, preliminary positioning of the leak area;

[0009] S3. Flow gradient analysis and reverse hydraulic refinement: Within the suspected leak area identified in step S2, the leak is located with meter-level accuracy using flow gradient analysis and a reverse hydraulic iteration model.

[0010] S4. Deploy multiple sensor nodes on the leaking pipe identified in step S3 to collect detection data. Determine the leak location using a relevant positioning fusion algorithm. Combined with the leak location determined by flow gradient analysis and reverse hydraulics in step S3, a comprehensive decision is made to ultimately determine the exact location of the pipeline leak.

[0011] Furthermore, the step S1 specifically includes:

[0012] S1.1 Data Collection: Install high-precision flow meters and pressure sensors at key points in the piped water supply network. Divide the large water supply area into several independent metering zones. Install flow meters at the entrance and exit of each zone, and deploy pressure sensors at the same locations as the flow meters.

[0013] S1.2 Data preprocessing: Apply wavelet denoising to the raw flow / pressure data, perform sliding average filtering on the pressure signal, and perform anomaly detection and cleaning on the collected data;

[0014] S1.3 Construct a hydraulic model of the water supply network: Based on the network topology modeling, establish a steady-state hydraulic model and a transient model.

[0015] Furthermore, in step S1.2, during data cleaning, gross error data is removed, and the isolation forest algorithm is used to identify complex abnormal patterns on the remaining data;

[0016] In step S1.3, when building the transient model, a dynamic pressure monitoring and compensation strategy and a pressure wave propagation frequency domain filtering strategy are adopted; in the dynamic pressure monitoring and compensation strategy, when the pressure change rate is detected to be greater than the threshold, the steady-state model is dynamically corrected, and the correction formula is: Q comp =Q base +α·ΔP· ;

[0017] Among them, Q comp is the actual flow rate after correction under the influence of pressure change, Q base is the theoretical flow value without pressure fluctuation and external interference, ΔP is the pressure change, Δt is the time step, and α is the pipe elasticity correction coefficient. In the frequency domain filtering strategy for pressure wave propagation, the original flow signal collected by the flow meter is decomposed by wavelet packets to separate the high-frequency fluctuation components caused by pressure wave propagation. After removing the high-frequency components, the reconstructed signal is used as the input data of the steady-state model.

[0018] Furthermore, in step S2, the difference between the inlet flow and the outlet flow of each partition is calculated based on the flow balance method, and the calculation formula is: , where Q in Indicates the flow rate value of the inlet flow meter, Q out Indicates the flow value of the outlet flow meter, For normal water consumption; When , it is determined that there is leakage in the partition, and the ε is a preset value.

[0019] Furthermore, the step S3 specifically includes: analyzing the flow gradient and calculating the flow gradient between adjacent flow meters, the formula is: , where ∇Q represents the flow gradient, Q i is the instantaneous flow rate of the i-th flow meter, Q i+1 is the instantaneous flow rate of the downstream adjacent flow meter, L is the length of the pipe section between the two flow meters, if the flow gradient value of a pipe section is significantly greater than that of its upstream pipe section, it is determined that the leak point is located in the pipe section near the upstream flow meter; when the dynamic pressure change rate is greater than the threshold, the formula Q is used to calculate the leak point. comp =Q base +α·ΔP·

[0020] Real-time calibration of flow value, based on the corrected flow Q comp , establish a reverse hydraulic inversion model to solve the leakage point location.

[0021] Furthermore, the reverse hydraulic inversion model for solving the leakage point location includes the following steps:

[0022] Assume the leak point location x and leakage flow Q leak As the parameter to be solved, the flow-leak point correlation equation is established. For the pipe section containing the leak point, the corrected flow monitored by the upstream and downstream flow meters satisfies the formula: Q comp , upstream =Q comp , downstream +Q leak ; wherein, the Q comp , upstream Corrected flow rate upstream of the leak point; Q comp , downstream Q is the corrected flow rate downstream of the leak point, leak is the leakage flow;

[0023] With the goal of "minimizing the residual between simulated flow and corrected flow", the objective function is constructed. , where x is the location of the leak point; Q sim,i (x) is the simulated flow rate of the ith flow meter when the leak point is assumed to be x; Q real,i is the corrected actual flow rate of the flow meter; n is the flow count of the target pipe section and the upstream and downstream associated pipe sections;

[0024] The nonlinear least squares optimization algorithm is used to iteratively solve the problem until the termination conditions of residual accuracy reaching the standard or position change convergence are met.

[0025] Furthermore, in step S4, the correlation positioning fusion algorithm includes multi-sensor node deployment and data collection, cross-correlation algorithm positioning, and multiple node correlation algorithm fusion positioning.

[0026] Furthermore, the cross-correlation algorithm positioning algorithm includes:

[0027] When sensor A and sensor B are located on both sides of the leak point, the sampled leak vibration signals are defined as 、 ,but 、 With similarity, if Hysteresis Time is , then approximately: , where A is the mutual correlation coefficient, then 、 The cross-correlation function is:

[0028] ,

[0029] When R ab When the maximum value is taken, the corresponding τ is , It is the time difference between the leakage vibration wave and the two probes. If the sound velocity of the pipeline medium is V, the pipeline length L between the leakage point and node B is (MV T) / 2, at this time the leak point location between the two points is completed.

[0030] Furthermore, the multiple node-related algorithm fusion positioning algorithm includes cluster head selection, clock synchronization algorithm and multi-sensor data fusion positioning algorithm; wherein,

[0031] The cluster head is the node at the center of the cluster with the most residual energy;

[0032] The clock synchronization algorithm includes the following steps: the cluster head node broadcasts a data packet carrying a reference timestamp beacon to each node in the cluster. Each node receives the data packet and records the reference timestamp and local time. The neighboring nodes of the cluster head sequentially transmit the integrated data packet carrying the reference time and their own local time until the last node. The node averages all local times and transmits them back to each node in reverse order. Each node updates its local time. The operation is repeated until the clock synchronization error is less than a threshold.

[0033] The multi-sensor data fusion positioning algorithm includes:

[0034] Under normal circumstances where there are no branches in the water supply pipeline within the cluster, the water leakage acoustic signal strength picked up by the nodes is grouped, paired, and sorted to ensure that the leakage point falls between the two nodes performing the cross-correlation algorithm. The size of the number reflects the strength of the signal received by the paired node. The leakage point location determined by all paired nodes is calculated according to the cross-correlation algorithm positioning method. The improved weighted average algorithm is used to obtain the final location of the leakage point for multiple groups of positioning results, where the weighting factor is inversely proportional to the distance between the two paired nodes. The sorting number is sorted each time the fusion is performed, and abnormal weighting factors are eliminated to exclude paired nodes affected by sudden noise interference.

[0035] In the special case of branching pipelines within a cluster, for multiple group pairing results formed by node signal strength, if the deviation of the positioning results of different groups is within the meter level range, the mean is taken as the leak point location. If the deviation between the two points is outside the meter level range, the leak point is determined through comprehensive decision-making.

[0036] Furthermore, the comprehensive decision-making in step S4 includes the following steps: S4.4.1, data preparation and structuring: collecting the positioning results obtained by the reverse hydraulic method and the multi-node data correlation positioning fusion algorithm, and establishing an information matrix containing the location information, source method and corresponding confidence / accuracy index of each candidate point;

[0037] S4.4.2. Preliminary Fusion and Overlap Analysis: Overlay all candidate points onto the pipe network GIS map to identify areas of high spatial concentration and select points that are consistent with both methods or have very high confidence or accuracy. S4.4.3. Introduce Field Constraints: Quantify field factors that influence the actual location of the leak, including pipe material and age, joint / valve location, burial depth and cover load, soil properties and corrosive environment, historical leak history, external disturbances, and pipe network topology.

[0038] S4.4.4. Establish a comprehensive decision-making model, design a scoring system, and calculate a comprehensive score for each candidate point;

[0039] S4.4.5. Ranking and preliminary decision: Determine the most likely leakage point or multiple high-risk areas by ranking based on the comprehensive scores.

[0040] This method achieves precise positioning through flow analysis combined with multiple technologies. High-precision flow meters and pressure sensors are installed at key nodes in the pipeline network. Data is collected and preprocessed (synchronized, denoised, and cleaned). Steady-state and transient hydraulic models are constructed to eliminate interference such as water hammer. Flow balance is used to calculate the difference between inlet and outlet flow rates in each zone. Leaks are suspected when they exceed a threshold. Time series methods such as nighttime minimum flow and cluster analysis are then used to verify the true leak location. The leaking pipeline is then identified by calculating the flow gradient in the pipeline segment. A reverse hydraulic model is constructed based on the corrected flow rate. Through iteration, the residual difference between the simulated and actual flow rates is minimized, allowing the leak to be located to the meter level. Finally, multiple sensors are deployed along the target pipeline, and a cross-correlation algorithm is used for preliminary positioning. After optimization through cluster head selection and clock synchronization, the reverse hydraulic and multi-node positioning results are integrated, and comprehensive decisions are made based on site constraints (such as pipeline material and historical records). The final positioning accuracy reaches approximately 0.3 meters. This multi-stage progressive positioning solution, integrating multiple technologies and site factors, overcomes the low accuracy and poor interference resistance of traditional methods, achieving precise leak detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, Figure 1As shown, the following is a brief introduction to the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0042] Figure 1 Schematic diagram of the flow chart of the method for detecting and locating water leakage points in a pipe network based on flow analysis in the present invention;

[0043] Figure 2 This is a schematic diagram of the positioning principle of the cross-correlation algorithm in the present invention;

[0044] Figure 3 This is a schematic diagram of the principle of accurately locating leak points by fusing data from multiple sensor nodes in the present invention;

[0045] Figure 4 This is a schematic diagram of the pipe network structure with branches in the present invention. DETAILED DESCRIPTION

[0046] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0047] The present invention discloses a method for detecting and locating water leakage points in a pipe network based on flow analysis. Figure 1 As shown,

[0048] The key to leak location based on flow data lies in analyzing abnormal changes in flow distribution within the water supply network. When a pipeline leak occurs, flow upstream of the leak increases (due to the additional water flow caused by the leak) while flow downstream decreases (due to the loss of some water from the leak). Leak location is determined through a comprehensive approach using flow balance models, flow gradient analysis, reverse hydraulics, and related fusion algorithms. Key methods include:

[0049] Step S1, constructing a flow balance model of the pipeline network.

[0050] Specifically, step S1 includes:

[0051] S1.1 Data Collection

[0052] Install high-precision flow meters at key points of the pipe water supply network (such as pump stations, zone inlets / outlets, and main pipe intersections). Electromagnetic flow meters or ultrasonic flow meters are preferred. The selection principles are as follows:

[0053] Electromagnetic flowmeter: Suitable for stable measurement of full pipe flow and conductive liquids. The pipe diameter range is DN50~DN2000. During installation, ensure that the straight pipe section before and after is ≥5 times the pipe diameter.

[0054] Ultrasonic flowmeter: Suitable for large pipe diameters (DN300 and above) or partially full pipe flows. It supports clamp-on installation, but is sensitive to fluid turbidity and requires regular calibration.

[0055] Furthermore, for large water supply areas, the pipe network is divided into several independent metering areas (DMAs). Flow meters (same type to reduce system errors) are installed at the inlet and outlet of each area to monitor flow data in real time. Pressure sensors are deployed simultaneously with the flow meters, with a sampling frequency of ≥1Hz, for transient process analysis and leak verification.

[0056] S1.2 Data Preprocessing

[0057] Use PTP (Precision Time Protocol) or GPS clock synchronization to ensure that all flow meters and pressure sensors have a timestamp error of <0.1 second (which meets the requirements for subsequent flow gradient analysis). Apply wavelet denoising to the raw flow / pressure data to remove high-frequency noise (>1kHz) and low-frequency drift (<200Hz). Pressure signals are additionally filtered using a sliding average filter (with a 5-second window) to eliminate transient fluctuations.

[0058] Data cleaning, anomaly detection is performed on the collected data. In the first stage, the 3σ principle is used to eliminate data with gross errors (data within the range of ±3 times the standard deviation is retained). In the second stage, the isolation forest algorithm is used on the remaining data to identify complex abnormal patterns (such as sensor drift and intermittent interference).

[0059] S1.3 Constructing a hydraulic model for the water supply network

[0060] First, based on the pipe network topology model, a water supply network topology model is created using professional hydraulic modeling software such as CFD. Specific steps include: annotating parameters such as pipe length, diameter, material, and valve location. Node coordinates and connection relationships are input to generate a water supply network diagram (graph structure). Pipeline parameters such as length L, diameter D, material C, and valve location and status are input. Node parameters are then annotated.

[0061] Secondly, steady-state hydraulic model building includes: building a steady-state model based on the principles of conservation of mass and energy, and calculating the flow distribution of each node under normal operating conditions.

[0062] For each node i, it needs to satisfy:

[0063] (In case of leakage)

[0064] Among them, Q inrepresents the flow into node i, Q out represents the flow out of node i, Q leak Represents the amount of water leakage at node i.

[0065] Finally, transient modeling. Due to the water hammer effect and pressure wave propagation caused by transient processes in actual pipeline network operation (such as valve opening and closing, pump station startup and shutdown), the steady-state model calculation results are subject to uncertainty. Therefore, dynamic pressure monitoring and compensation strategies, as well as frequency-domain filtering strategies for pressure wave propagation, are required to mitigate these effects.

[0066] For dynamic pressure monitoring and compensation strategy, it is necessary to install pressure sensors at key nodes to monitor pressure fluctuations in real time. When the pressure change rate ∣∂P / ∂t∣>ε p When the pressure is εp = 0.1 MPa / s, it is considered that a significant transient process occurs. At this time, the steady-state model is dynamically modified:

[0067] Q comp =Q base +α·ΔP·

[0068] Among them, Q comp It represents the actual flow rate after correction under the influence of pressure change, Q base It represents the theoretical flow rate value without pressure fluctuation and external interference. ΔP is the pressure change, Δt is the time step, and α is the pipe elasticity correction coefficient.

[0069] Frequency-domain filtering of pressure wave propagation requires wavelet packet decomposition of the raw flow signal collected by the flowmeter to isolate the high-frequency fluctuation components caused by pressure wave propagation. After removing these high-frequency components, the reconstructed signal serves as input data for the steady-state model to eliminate interference from pressure wave propagation.

[0070] Step S2: Based on the flow balance model constructed in step S1, the flow status of each node in the pipeline network is detected in real time. When a pipeline leak is detected, the leak point is preliminarily located.

[0071] In step S2, the difference between the inlet flow and outlet flow of each partition is first calculated based on the flow balance method, and whether there is leakage in each pipe section is calculated according to the following formula:

[0072] Among them, Q in Indicates the flow rate value of the inlet flow meter, Q out Indicates the flow value of the outlet flow meter, is the normal water consumption (which can be estimated from historical data).

[0073] when (Threshold, usually ε = max(Q in, Q out )), if it is determined that there is a leak in the partition, then go to step S3 for verification; otherwise it is considered normal.

[0074] Furthermore, after the flow balance method initially identifies the suspected leak zone, it is necessary to verify the persistence of the leak through a time series analysis method to eliminate temporary interference (such as sudden water use and sensor fluctuations) and ensure the reliability of the detection results. For example, any of the following auxiliary methods should be used:

[0075] The nighttime minimum flow method uses low water usage to highlight leak characteristics. The analysis period is chosen between 2:00 AM and 4:00 AM. This is because this is the low water usage period, when residential and industrial water consumption is at its lowest and most stable (baseline flow is stable). If a leak persists, the leak will continuously generate additional flow (unrelated to normal water usage), causing the actual flow during this period to be significantly higher than the normal baseline. This method, by focusing on the "stable window of minimal interference," amplifies flow anomalies caused by leaks and reduces the impact of daily water usage fluctuations on detection.

[0076] DBSCAN (Density-Sensitivity Clustering Algorithm) analysis eliminates outliers from temporary, special events. Nighttime flow data may contain temporary anomalies that are not related to leaks (such as emergency water use for firefighting, pipe flushing, and transient sensor failures). These events manifest as short-term, large flow fluctuations (outliers) but do not represent sustained leaks. DBSCAN (Density-Sensitivity Clustering Algorithm) identifies "dense areas" (normal flow fluctuation ranges) and "sparse outliers" (special events) in the data. After eliminating these outliers, the remaining flow anomalies are more likely to be stable deviations caused by sustained leaks, thereby improving data purity.

[0077] Historical Baseline Comparison: Leak stability is verified by continuous deviation. The processed current nighttime flow rate is compared to a historical baseline (e.g., the average of normal nighttime flow rates for the same period over the past 30 days). The alarm threshold is set as "10% above the baseline for three consecutive days." The historical baseline represents the normal flow rate level in the absence of a leak. A leak could cause the current flow rate to consistently exceed the baseline. The "three consecutive days" setting avoids misjudgment of occasional daily fluctuations (such as short-term changes in water use due to unusual weather) and ensures that the anomaly is persistent. The "10% threshold" is based on the ratio of pipeline leakage to normal water use, balancing sensitivity and false alarm rate.

[0078] CUSUM and Bayesian change point detection: Accurately identify abnormal traffic trends. CUSUM (Cumulative Sum Algorithm): By accumulating the deviation of traffic flow from the baseline, a "change point" (a sudden change in traffic trend) is identified when the cumulative value exceeds a preset threshold. This method is suitable for detecting persistent, small increases in traffic flow caused by leaks (such as slow leaks). Compared to single-shot data comparisons, it is more capable of capturing long-term trend anomalies. Bayesian change point detection uses a probabilistic model to calculate the posterior probability of a "change point" (the point in time when a leak occurs or persists) in a traffic sequence. By quantifying uncertainty, it accurately identifies the turning point from normal to abnormal traffic conditions, verifying the persistence and onset of a leak.

[0079] These time series analysis methods verify the persistence of traffic anomalies through the logic of "focusing on low-interference periods → eliminating temporary anomalies → comparing historical baselines → detecting trend changes", and ultimately confirm whether there is a real leak in the suspicious partition, providing a reliable basis for subsequent refined positioning and reducing the misjudgment rate.

[0080] Step S3: Flow gradient analysis and reverse hydraulic fine-tuning. After the suspected leak area is identified in the coarse positioning phase, the refined positioning phase achieves meter-level positioning accuracy through flow gradient analysis and reverse hydraulic iteration. This step, following the initial identification of the leaking zone using the flow balance method in step S2, further narrows the leak point, improving positioning accuracy from the regional level to the meter level. This provides specific coordinates for subsequent precise repairs and addresses the rough positioning issues of the single flow balance method.

[0081] Specifically, step S3 includes:

[0082] S3.1. Based on the flow gradient analysis, dense flow meters (spacing ≤ 1 km) are arranged along the pipeline within the suspected leakage zone, and the time synchronization error of these flow meters is ≤ 0.1 second.

[0083] First, calculate the flow gradient between adjacent flow meters: ;

[0084] The flow gradient between adjacent flow meters is calculated by the formula, where represents the flow gradient, Q i is the instantaneous flow rate of the i-th flow meter, Q i+1 is the instantaneous flow rate of the adjacent downstream flow meter, and L is the length of the pipe section between the two flow meters. If the flow gradient value of a pipe section is significantly greater than that of its upstream pipe section, the leak is determined to be located in the pipe section near the upstream flow meter.

[0085] Then, for the real-time pressure data, when the dynamic pressure (the rate of change of pressure over time) |∂P|∂t|>0.1MPa / s, it indicates that the pipeline network transitions from a stable state to an unstable state. At this time, proceed to step S3.2;

[0086] S3.2, using the formula: Q comp =Q base +α·ΔP·

[0087] Real-time calibration of flow value, based on the corrected flow Q comp , establish a reverse hydraulic inversion model to solve the leakage point location.

[0088] S3.3 Reverse hydraulic calculation. The reverse hydraulic inversion model is based on the premise that "pipeline leakage will cause local flow anomalies" and establishes a mathematical relationship through the following assumptions and constraints:

[0089] Step 1: Assume that the leak is located in a certain pipe section, and calculate the leak point location x (the distance from the starting point of the pipe section, unit: meter) and the leakage flow Q leak As a parameter to be solved, the presence of a leak will change the hydraulic characteristics of the pipe section, which is manifested as an increase in flow upstream of the leak, a decrease in flow downstream, and an abnormal pressure attenuation pattern along the way.

[0090] The second step is to establish the flow-leak point correlation equation based on the corrected flow Q comp (Dynamic pressure interference has been eliminated) and the correlation equation between the leak point and the flow rate is constructed. For the pipe section containing the leak point, the corrected flow rate monitored by the upstream and downstream flow meters must meet the following requirements: Q comp , upstream =Q comp , downstream +Q leak Among them, Q comp , upstream Q is the corrected flow rate upstream of the leak point, comp , downstream Q is the corrected flow rate downstream of the leak point, leak is the leakage flow (related to the size of the leak point and pipeline pressure).

[0091] The third step is to construct the objective function with the goal of minimizing the residual between the simulated flow and the corrected flow:

[0092] The objective function is: ;

[0093] Where x is the location of the leak point, Q sim,i (x) is the simulated flow rate of the ith flow meter when the leak point is assumed to be x, Q real,i is the corrected actual flow rate of the flow meter (i.e. Q comp ); n is the number of flow meters in the target pipe section and its upstream and downstream associated pipe sections. The flow meters must cover the key nodes of the target pipe section and its upstream and downstream associated pipe sections, including at least 2 upstream and 2 downstream of the leak point to ensure that flow gradient changes can be captured.

[0094] The physical meaning of the objective function is the residual sum of squares between the simulated flow and the corrected flow. The smaller its value, the closer the assumed leak point location x is to the actual leak point. By minimizing it, the accurate inversion of the leak point location can be achieved. The initial position with the smallest residual is selected as the iterative starting point, and the nonlinear least squares optimization algorithm is used for iterative solution. The leak point inversion is aimed at "minimizing the residual of the simulated flow and the corrected flow". This process gradually reduces the residual of the simulated flow and the corrected flow by continuously adjusting the leak point location parameters, and finally converges to the optimal solution. The specific steps are as follows:

[0095] S3.3.1. Gradient calculation: For the current leak location, calculate the difference (residual) between the simulated flow and the corrected flow at each flow meter to determine the deviation of each sensor. Through hydraulic model parameter perturbation experiments, simulate the change in the simulated flow of each flow meter when the leak location changes slightly (e.g., ±0.1 meter). This determines the "change in simulated flow for every 1-meter change in the leak location," or the sensitivity coefficient. Multiply the residual by the sensitivity coefficient and add them together to obtain the gradient of the objective function with respect to the current leak location. The positive or negative direction of the gradient directly reflects "in which direction the position should be adjusted to reduce the residual." If the gradient is positive, the leak location needs to be moved to the left (upstream) to reduce the residual; if the gradient is negative, it needs to be moved to the right (downstream).

[0096] S3.3.2, Adaptive step size adjustment, the step size is the adjustment amplitude of the leak point position in each iteration, its size directly affects the iteration efficiency and convergence stability, and needs to be dynamically adapted according to the residual change: the initial step size setting is based on the pipeline length and the flow meter spacing to determine the initial adjustment amplitude (such as 5% of the target pipe length), to ensure that the initial iteration can quickly approach the optimal solution. After each iteration, the residual changes before and after the adjustment are compared. If the residual decreases significantly (for example, more than 10%), it means that the current step size direction is correct and the amplitude is reasonable. The step size is increased in the next iteration to accelerate convergence. If the residual increases or decreases slightly (for example, less than 1%), it means that the step size is too large and may cause oscillation. The step size is reduced in the next iteration to stably approach the optimal solution. If the residual fluctuates frequently, the minimum step size protection mechanism is triggered to forcibly limit the step size to within the threshold meter to avoid excessive adjustment.

[0097] S3.3.3. Position update: Iterate and adjust along the gradient descent direction, and update the leak point position according to the gradient direction and the current step length: if the gradient is positive, move the leak point position in the upstream direction of the pipeline by the current step length; if the gradient is negative, move the leak point position in the downstream direction of the pipeline by the current step length; after each position update, it is necessary to verify whether the new position exceeds the target pipe section range (for example, it must not exceed the monitoring range of the upstream and downstream flow meters). If it exceeds, it will be automatically truncated to the pipe section boundary to ensure physical rationality.

[0098] S3.3.4. The iteration terminates until either of the following two termination conditions is met to ensure the reliability of the inversion results: First, the residual accuracy meets the standard, the relative deviation between the simulated flow and the corrected flow of all flow meters is ≤0.5% (that is, the flow deviation of a single sensor does not exceed 0.5% of the actual flow), and the overall residual sum of squares is stable at the minimum value, and the change in three consecutive iterations is ≤0.001; Position change converges: Second, the leak point position adjustment between two adjacent iterations is ≤0.1 meter, indicating that the position is close to the actual leak point, and further iterations have no significant significance for improving accuracy.

[0099] In step S4, after the present invention locates the leaking pipe through flow gradient and hydraulic inversion in step S3, it further proposes step S4 to collect detection data at multiple sensor nodes arranged in the determined leaking pipe, determine the location of the leak point based on the relevant positioning fusion algorithm, and then make a comprehensive decision based on the flow gradient analysis and reverse hydraulic determination of the leak point, and finally determine the exact location of the pipeline leak point.

[0100] Specifically, step S4 includes the following key technologies:

[0101] (1) Multi-sensor node deployment and data collection

[0102] To enhance the reliability and accuracy of leak detection systems, this invention utilizes sensor nodes deployed across water supply pipelines. Data collected from multiple sensor nodes located at identified leaking water supply pipelines is calculated. This multi-sensor node data collection overcomes the limitations of a single sensor. Cross-validation of multi-node data reduces misjudgments caused by single sensor failures or localized noise. This provides "spatially distributed" physical signals (vibration and acoustics) for subsequent cross-correlation algorithms. The signal propagation characteristics are then used to infer the leak location, compensating for the inadequacy of flow / hydraulic models in capturing microscopic physical characteristics.

[0103] (2) Positioning principle of cross-correlation algorithm

[0104] like Figure 2 As shown in the figure, under normal circumstances: when sensor A and sensor B are located on both sides of the leak point, if nodes A and B are under synchronous clock control, the leakage vibration signals sampled are 、 ,but 、 With similarity, if Hysteresis Time is , then approximately: , where A is the mutual correlation coefficient. 、 The cross-correlation function is:

[0105]

[0106] When R ab When the maximum value is taken, the corresponding τ is ,Right now Figure 2 Where T is the time difference between the leakage vibration wave and the two probes. If the sound velocity of the pipeline medium is V, then the pipeline length L between the leakage point and node B is (MV T) / 2, the leak point is now located between the two points, essentially the same as in step S3. Leveraging the propagation patterns of physical signals (vibration waves), a direct correlation between signal time difference and spatial distance is established, enabling preliminary leak location. Combined with the flow-hydraulic inversion in step S3, this creates a dual-dimensional verification of "hydraulic data + physical signals," providing foundational positioning results for the subsequent multi-node fusion algorithm and serving as input for refined optimization.

[0107] (3) The above is the ideal location of water leakage, but the following problems still exist: 1. When the synchronous clock sampling of nodes A and B is out of sync, it will cause the time difference T to deviate, which directly affects the positioning accuracy; 2. If the water leakage point is located outside the two nodes or exactly at the location of sensor A or B, then the formula L=(MV T) / 2=(MM) / 2=0, which makes the positioning method invalid; 3. When the acoustic wave signal of the leak point is weak and there is strong noise in the surrounding environment (greater than 80db), the noise source may be mistakenly identified as the leak point.

[0108] Based on the above situation, the present invention further proposes to use multiple node-related algorithm fusion to improve the accuracy of leak point judgment, and the specific implementation includes three key methods: cluster head selection, clock synchronization algorithm and multi-sensor data fusion positioning algorithm.

[0109] For cluster head selection, the present invention forms a cluster from the candidate area of ​​traffic positioning. The data perceived by each node in the cluster when detecting the same leak point is different due to their different locations. The cluster head is determined by considering factors such as the physical location, spatial distance and residual energy of ordinary nodes and cluster head nodes. In this algorithm, each node calculates its own position and residual energy, and passes the results to a designated node, which synthesizes and judges. The node at the center of the cluster with the most residual energy is the cluster head, and the cluster head information is broadcast to all nodes in the cluster. The selection of cluster heads can greatly reduce data transmission redundancy, centrally process data through cluster heads, improve algorithm efficiency, and prioritize nodes with sufficient energy and central location as the core to ensure data reliability and coverage, and avoid errors caused by weak signals at edge nodes.

[0110] In the clock synchronization algorithm, the cluster head node first broadcasts a data packet carrying a reference timestamp beacon to all nodes in the cluster. Upon receiving the data packet, each node immediately records the reference timestamp and local time. Next, node 1, a neighbor of the cluster head, sends a combined data packet containing the reference time and its own local time to its next neighbor, node 2. Upon receiving node 1's combined data packet, node 2 adds its own local time to the combined data packet and sends it to its next neighbor, node 3. This process continues until node n, at which point node n has obtained the local clock information of all nodes in the cluster (including the cluster head). Finally, node n averages all local times and transmits this average to all nodes in the cluster in the reverse order of the previous one. Each node then updates its local time to this average. This process repeats until the clock synchronization error falls below a threshold. This effectively addresses the problem of "time difference T calculation bias" in the cross-correlation algorithm, ensuring that the sampling times of multiple nodes are aligned and avoiding positioning errors caused by clock asynchrony.

[0111] Multi-sensor node data fusion algorithm, such as Figure 3 As shown, under normal circumstances, when the water supply pipelines within a cluster have no branches, the received water leakage vibration waves can be directly grouped, paired, and sorted based on their signal strength, and then cross-correlation and fusion localization can be performed. However, since the intensity of water leakage vibration waves from a pipeline leak decreases as the distance between the measurement node and the sound source increases during propagation, this characteristic can be exploited to group, pair, and sort the leak acoustic signals picked up by nodes within the cluster based on their strength. This ensures that the leak point falls between the two nodes performing the cross-correlation algorithm, with the number reflecting the strength of the signal received by the paired node. The multi-sensor node data fusion algorithm can address the issues of "leak points outside the range of two nodes" or "noise interference": by cross-validating multiple groups of nodes, anomalous results are eliminated, ensuring that the localization results fall within a reasonable range. A weighted average strategy emphasizes the contribution of highly reliable node pairs and reduces the impact of weak signals or noise, thereby improving the robustness of the algorithm.

[0112] According to the cross-correlation positioning method in step (2) in step S4, the leakage point locations determined by all paired nodes are calculated, and the improved weighted average algorithm is performed on multiple groups of positioning results to obtain the final location of the leakage point.

[0113] The principle of signal attenuation indicates that the farther a sensor node is from a pipe leak, the greater the attenuation of the received signal, the less clear the leak information carried in the signal, and the less reliable the positioning results calculated by the paired nodes. Therefore, a weighting factor (i.e., weight) inversely proportional to the distance between the two paired nodes is introduced as a coefficient for the positioning results of the paired nodes. This weighted average fusion algorithm can further improve the final positioning accuracy.

[0114] Various types of random burst noise within each segment can interfere with or even drown out the sound waves from leaks at pipeline leak points. These disrupted nodes can render the weighting factors of traditional weighted averaging ineffective, leading to greater errors. Therefore, an improved weighted averaging algorithm can mitigate the effects of these burst noises. This approach involves combing through the sorted numbers during each fusion to check for any abnormal weighting factors. Any such factors are removed, preventing paired nodes affected by burst noise from participating in data fusion and improving the robustness of the algorithm.

[0115] Special case: When there are branches in the pipeline of the cluster, this is the most difficult case to judge, such as Figure 4 As shown in the figure, 1 to 6 are nodes. Assuming that point A is the leaking point, there are three groups of pairs according to the node signal strength, namely Group 1: 2-5, 1-6; Group 2: 3-5, 4-6; Group 3: 2-3, 1-4. According to the cross-correlation algorithm, the leaking point determined by Groups 1 and 2 is A, while Group 3 is mistakenly identified as B. If the distance between point B and point A is within the meter range, we believe that the leaking point is at (A+B) / 2, and the leaking point can be accurately found within the specified construction range. If the distance between point B and point A is outside the meter range, then step (4) comprehensively determines the leaking point.

[0116] (4) Implementation plan for comprehensive decision-making loopholes

[0117] Based on the positioning results obtained through "flow gradient analysis and reverse hydraulic method" and "multi-node data correlation positioning fusion algorithm," combined with objective on-site conditions, the most reasonable and likely actual leak location is selected to guide precise excavation and repair. The core strategy utilizes data fusion, weighted analysis, and on-site verification.

[0118] Specific steps of the program:

[0119] S4.4.1. Data preparation and structuring

[0120] Collect the positioning results and clearly record the coordinates / section identifiers of the N-1 most likely leak points and their corresponding confidence levels (e.g., calculated error range, simulation fit) derived from the "flow gradient analysis and reverse hydraulic method." Also, record the coordinates / section identifiers of the M-1 most likely leak points and their corresponding positioning accuracy metrics (e.g., peak correlation coefficient, positioning error radius) derived from the "multi-node data correlation positioning fusion algorithm." Create a table using an information matrix to include each candidate point's location information (coordinates, section ID, distance from key nodes), the source method, and the confidence / accuracy metrics derived from that method. Standardize and structure the results of both positioning methods to provide a unified data foundation for subsequent fusion analysis, avoid analytical bias caused by confusing data formats, and clarify the reliability basis for each candidate point.

[0121] S4.4.2. Preliminary fusion and overlap analysis

[0122] All candidate points located by both methods (N+M in total) are superimposed on the pipe network GIS map. Spatial clustering is used to identify highly concentrated areas (i.e., multiple points located very close together). Prioritize points identified by both methods, or points where one method has extremely high confidence / accuracy (for example, points where the positioning error radius is much smaller than the pipe spacing), quickly narrowing the scope of potential leaks from a spatial perspective. Multi-method cross-validation is used to mitigate the limitations of a single method (e.g., the inverse hydraulic method may be affected by pipe topology, and the multi-node algorithm may be affected by noise). Screening for concentrated areas and high-confidence points can initially eliminate obviously unreasonable candidate points, reducing the complexity of subsequent analysis.

[0123] S4.4.3. Introducing site constraints (weighting factors)

[0124] The key constraint variables are to identify and quantify on-site factors that may affect the actual location of the leak, such as: 1. Pipe material and aging degree: cast iron pipes are prone to leak at the joints, and PVC pipes are prone to leak at damaged points; 2. Weak points in old pipelines (such as areas with severe corrosion) are weighted higher; 3. Joint / valve location: leaks are more common near joints or valves, and candidate points close to joints / valves should be given extra points; 4. Burial depth and cover load: areas that are too shallow or too deep, or areas with heavy loads (roads, buildings) above, have a higher risk of leakage; 5. Soil properties and corrosion environment: areas with loose soil or highly corrosive soil have a higher risk of leakage; 6. Historical leakage records: whether there has been a repair record nearby, and leak-prone pipe sections are weighted higher; 7. External disturbances: whether there has been recent construction (excavation, piling, etc.) nearby, and potential third-party damage points should be paid special attention to; 8. Pipeline network topology: stress concentration points such as elbows, tees, and reducers are more prone to damage. By combining the algorithm results with the actual conditions of the engineering site and assigning higher weights to high-risk scenarios (such as old pipeline interfaces), the evaluation of candidate points can be made closer to the actual leakage pattern, reducing the contradiction of "the algorithm has accurate positioning but there are no leaks on site".

[0125] S4.4.4. Establish a comprehensive decision-making model:

[0126] A scoring system is designed to quantitatively evaluate each candidate point, including the weight of positioning accuracy (Wt), the weight of site conditions (Wc), and the weight of expert experience (We).

[0127] Calculate the comprehensive score for each candidate point: Comprehensive score = Score t *Wt + Score c * Wc + Score e*We. By weightedly integrating "theoretical positioning accuracy," "actual on-site risk," and "expert experience," we achieve a quantitative balance in multi-dimensional decision-making. This weighting ensures the core accuracy of the algorithm (Wt accounts for the highest proportion) while also taking into account engineering practice (Wc) and difficult-to-quantify implicit knowledge (We), making the final results more practical.

[0128] It should be noted that if multiple candidate points (even from different methods) are clustered in a certain area, the comprehensive score of the area (or its representative points) should be improved.

[0129] S4.4.5. Ranking and preliminary decision-making:

[0130] All candidate points are sorted by their comprehensive scores, and the gap between the highest-scoring and second-highest-scoring points is determined. If the gap is significant (e.g., significantly greater than the model error or positioning accuracy itself), the highest-scoring point can be preliminarily identified as the "most likely leak point." If the scores of the top few candidates are very close, it is considered that there are several highly likely areas. When preparing the verification plan, these areas should be covered and prioritized to avoid decision-making confusion caused by an excessive number of candidate points. Determining a significant gap can quickly identify a single leak point, while close scores indicate the need for further verification. This balances decision-making efficiency and accuracy, providing clear target areas for subsequent on-site verification.

[0131] Furthermore, non-intrusive on-site verification can be adopted, and sound correlation method, gas tracer method, and infrared thermal imaging can also be used to assist decision-making under certain conditions.

[0132] S4.4.1 to S4.4.6 achieve a deep integration of theoretical algorithm positioning and actual engineering scenarios through the logic of data standardization → spatial screening → on-site weighting → quantitative decision-making → on-site verification, and ultimately improve the leakage point positioning accuracy to about 0.3 meters, providing a reliable basis for precise excavation and repair, and at the same time solving the problem that single-method positioning is susceptible to interference and disconnected from reality.

[0133] In summary, this invention achieves precise positioning through flow analysis combined with multiple technologies. High-precision flow meters and pressure sensors are installed at key nodes in the pipeline network. Data is collected and preprocessed (synchronized, denoised, and cleaned). Steady-state and transient hydraulic models are constructed to eliminate interference such as water hammer. Flow balance is used to calculate the difference between inlet and outlet flow rates in each zone. Leaks are suspected if they exceed a threshold. Time series methods such as nighttime minimum flow and cluster analysis are then used to verify the location of the actual leak. The flow gradient in the pipeline section is then calculated to identify the leaking pipeline. A reverse hydraulic model is constructed based on the corrected flow rate. Through iteration, the residual difference between the simulated and actual flow rates is minimized, allowing the leak to be located to the meter level. Finally, multiple sensors are deployed along the target pipeline, and a cross-correlation algorithm is used for preliminary positioning. After optimization through cluster head selection and clock synchronization, the reverse hydraulic and multi-node positioning results are integrated, and comprehensive decisions are made based on site constraints (such as pipeline material and historical records). The final positioning accuracy can reach approximately 0.3 meters. This multi-stage progressive positioning solution, integrating multiple technologies and site factors, overcomes the low accuracy and poor interference resistance of traditional methods, achieving precise leak detection.

Claims

1. A method for detecting and locating water leakage points in a pipe network based on flow analysis, characterized in that: The following steps are involved: S1. Construct a flow balance model for the pipeline network; S2. Based on the flow balance model constructed in step S1, real-time detection of the flow status of each node in the pipeline network, when a pipeline leak is detected, the leak point is initially located; S3. Flow gradient analysis and reverse hydraulic refinement: Within the suspected leak area identified in step S2, the leak is located with meter-level accuracy using flow gradient analysis and a reverse hydraulic iteration model. S4. Deploy multiple sensor nodes on the leaking pipe identified in step S3 to collect detection data. Determine the leak location using a relevant positioning fusion algorithm. Combined with the leak location determined by flow gradient analysis and reverse hydraulics in step S3, a comprehensive decision is made to ultimately determine the exact location of the pipeline leak.

2. The method according to claim 1, characterized in that The step S1 specifically includes: S1.1 Data Acquisition: Install high-precision flow meters and pressure sensors at key nodes in the piped water supply network, dividing the large water supply area into several independent metering zones. Flow meters are installed at the entrance and exit of each zone, and pressure sensors are deployed simultaneously with the flow meters. S1.2 Data Preprocessing: Apply wavelet denoising to the raw flow / pressure data, perform sliding average filtering on the pressure signal, and perform anomaly detection and cleaning on the collected data. S1.3 Construct a hydraulic model of the water supply network: Based on the network topology modeling, establish a steady-state hydraulic model and a transient model.

3. The method according to claim 2, wherein: In step S1.2, during data cleaning, gross error data are removed, and the isolation forest algorithm is used to identify complex abnormal patterns on the remaining data; In step S1.3, when building the transient model, a dynamic pressure monitoring and compensation strategy and a pressure wave propagation frequency domain filtering strategy are adopted; in the dynamic pressure monitoring and compensation strategy, when the pressure change rate is monitored to be greater than the threshold, the steady-state model is dynamically corrected, and the correction formula is: ; Among them, Q comp is the actual flow rate after correction under the influence of pressure change, Q base is the theoretical flow value without pressure fluctuation and external interference, ΔP is the pressure change, Δt is the time step, and α is the pipe elasticity correction coefficient. In the frequency domain filtering strategy for pressure wave propagation, the original flow signal collected by the flow meter is decomposed by wavelet packets to separate the high-frequency fluctuation components caused by pressure wave propagation. After removing the high-frequency components, the reconstructed signal is used as the input data of the steady-state model.

4. The method according to claim 1, wherein In step S2, the difference between the inlet flow and the outlet flow of each partition is calculated based on the flow balance method, and the calculation formula is: , where Q in Indicates the flow rate value of the inlet flow meter, Q out Indicates the flow value of the outlet flow meter, For normal water consumption; When , it is determined that there is leakage in the partition, and the ε is a preset value.

5. The method according to claim 1, wherein The step S3 specifically includes: analyzing the flow gradient and calculating the flow gradient between adjacent flow meters. The formula is: , where ∇Q represents the flow gradient, Q i is the instantaneous flow rate of the i-th flow meter, Q i+1 is the instantaneous flow rate of the downstream adjacent flow meter, L is the length of the pipe section between the two flow meters, if the flow gradient value of a pipe section is significantly greater than that of its upstream pipe section, it is determined that the leak point is located in the pipe section near the upstream flow meter; when the dynamic pressure change rate is greater than the threshold, the formula is used Real-time calibration of flow value, based on the corrected flow Q comp , establish a reverse hydraulic inversion model to solve the leakage point location.

6. The method according to claim 5, characterized in that The reverse hydraulic inversion model for solving the leakage point location includes the following steps: Assume the leak point location x and leakage flow Q leak As the parameter to be solved, the flow-leak point correlation equation is established. For the pipe section containing the leak point, the corrected flow monitored by the upstream and downstream flow meters satisfies the formula: Q comp , upstream =Q comp , downstream +Q leak ; wherein, the Q comp , upstream Corrected flow rate upstream of the leak point; Q comp , downstream Q is the corrected flow rate downstream of the leak point, leak is the leakage flow; With the goal of "minimizing the residual between simulated flow and corrected flow", the objective function is constructed. , where x is the location of the leak point; Q sim,i (x) is the simulated flow rate of the ith flow meter when the leak point is assumed to be x; Q real,i is the corrected actual flow rate of the flow meter; n is the flow count of the target pipe section and the upstream and downstream associated pipe sections; The nonlinear least squares optimization algorithm is used to iteratively solve the problem until the termination conditions of residual accuracy reaching the standard or position change convergence are met.

7. The method according to claim 1, characterized in that In step S4, the related positioning fusion algorithm includes multi-sensor node deployment and data collection, cross-correlation algorithm positioning, and multiple node related algorithm fusion positioning.

8. The method according to claim 7, characterized in that The cross-correlation algorithm positioning algorithm includes: When sensor A and sensor B are located on both sides of the leak point, the sampled leak vibration signals are defined as a(t) and b(t), respectively. Then a(t) and b(t) have similarity. If a(t) lags behind b(t) by a time of , then approximately: , where A is the mutual correlation coefficient. At this time, the cross-correlation function of a(t) and b(t) is: ， When R ab When the maximum value is taken, the corresponding τ is , It is the time difference between the leakage vibration wave and the two probes. If the sound velocity of the pipeline medium is V, the pipeline length L between the leakage point and node B is (MV T) / 2, at this time the leak point location between the two points is completed.

9. The method according to claim 7, characterized in that The multiple node-related algorithm fusion positioning algorithm includes cluster head selection, clock synchronization algorithm and multi-sensor data fusion positioning algorithm; wherein, The cluster head is the node at the center of the cluster with the most residual energy; The clock synchronization algorithm includes the following steps: the cluster head node broadcasts a data packet carrying a reference timestamp beacon to each node in the cluster. After receiving the data packet, each node records the reference timestamp and local time. The neighboring nodes of the cluster head sequentially transmit the integrated data packet carrying the reference time and their own local time until the last node. The node averages all local times and transmits them back to each node in reverse order. Each node updates its local time. The operation is repeated until the clock synchronization error is less than a threshold. The multi-sensor data fusion positioning algorithm includes: under normal circumstances where the water supply pipeline in the cluster has no branches, grouping, pairing and sorting the water leakage acoustic signal strengths picked up by the nodes, ensuring that the location of the water leakage point falls between the two nodes performing the cross-correlation algorithm, and the size of the number reflects the strength of the signal received by the paired node. The leakage point location determined by all paired nodes is calculated according to the cross-correlation algorithm positioning method, and the improved weighted average algorithm is used to obtain the final location of the leakage point for multiple groups of positioning results, wherein the weighting factor is inversely proportional to the distance between the two paired nodes, and the sorting number is sorted each time the fusion is performed, and abnormal weighting factors are eliminated to exclude paired nodes interfered by sudden noise; in the special case where the pipeline in the cluster has branches, for the multiple groups of grouping and pairing results formed by the node signal strength, if the deviation of the positioning results of different groups is within the meter range, the average is taken as the leakage point location; if the deviation of the two points is outside the meter range, the leakage point is determined through comprehensive decision-making.

10. The method according to claim 9, characterized in that The comprehensive decision-making in step S4 includes the following steps: S4.4.

1. Data preparation and structuring: collecting the positioning results obtained by flow gradient analysis, reverse hydraulic method and multi-node data correlation positioning fusion algorithm, and establishing an information matrix containing the location information, source method and corresponding confidence / accuracy index of each candidate point; S4.4.

2. Preliminary Fusion and Overlap Analysis: Overlay all candidate points onto the pipe network GIS map to identify areas of high spatial concentration and select points that are consistent with both methods or have very high confidence or accuracy. S4.4.

3. Introduce Field Constraints: Quantify field factors that influence the actual location of the leak, including pipe material and age, joint / valve location, burial depth and cover load, soil properties and corrosive environment, historical leak history, external disturbances, and pipe network topology. S4.4.

4. Establish a comprehensive decision-making model, design a scoring system, and calculate a comprehensive score for each candidate point; S4.4.

5. Ranking and preliminary decision: Determine the most likely leakage point or multiple high-risk areas by ranking based on the comprehensive scores.

Images