A pipe lining stability analysis and health monitoring system for urban water supply networks

Abstract

The present invention discloses a pipe lining stability analysis and health monitoring system for urban water supply pipe networks, specifically relating to the field of water supply pipe network health monitoring, including: a data acquisition module acquiring triaxial vibration data and attaching timestamps and topological labels; a signal separation module extracting independent components of turbulence and structural vibration; a modal offset annotation module calculating the energy coupling ratio and identifying the offset mode; a stability attenuation assessment module quantifying the stiffness loss by comparing the benchmark data with the transfer function matrix and converting it into a pipe lining stability attenuation rate; a synchronous association module analyzing the timing characteristics and marking the synchronous association nodes; a pipe lining degradation tracing module constructing a tracing path tree and converting it into a decision result with geographic coordinates; the system can effectively distinguish between fluid disturbances and structural anomalies, perform pipe lining degradation monitoring and tracing, and realize visual early warning at the same time to ensure the long-term stable operation of the water supply pipe network.

Description

Technical Field

[0001] The present invention relates to the technical field of water supply network monitoring, and more particularly to a pipe lining stability analysis and health monitoring system for urban water supply networks. Background Art

[0002] Because the pipes of urban water supply networks operate deep underground for long periods of time, their structural health is significantly affected by changes in surrounding rock stress. The surrounding rock is affected by ground stress relaxation, changes in groundwater seepage pressure, or cyclical loads (such as earthquakes and construction disturbances), which can cause millimeter-level elastic deformation of the pipes. This deformation not only changes the geometric parameters of the pipe but also affects the stability attenuation of the pipe lining, thereby changing the internal natural vibration modes, especially the standing wave modes in the low-frequency band. However, existing pipe lining health monitoring technologies mainly rely on traditional methods such as laboratory testing or empirical judgment. These traditional methods require sampling and even shutting down the water supply network for testing, making it difficult to ensure timely and adaptable testing. This is especially true in the context of large-scale systems such as urban water supply networks, where traditional testing methods are extremely unsuitable.

[0003] In addition, existing water supply network health monitoring technologies often use distributed sensor arrangements and issue abnormal alarms based on single signal characteristics, but are unable to provide clear tracing of pipe lining degradation.

[0004] In order to solve the above problems, a technical solution is now provided. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a pipe lining stability analysis and health monitoring system for urban water supply pipe networks to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A pipe lining stability analysis and health monitoring system for urban water supply networks includes a data acquisition module, a signal separation module, a modal offset annotation module, a stability attenuation assessment module, a synchronization correlation module, and a pipe lining degradation tracing module.

[0008] The data acquisition module collects the original waveform data of the three-axis vibration including the time stamp, adds the time stamp and topology label to the data, and then outputs it to the signal separation module;

[0009] The signal separation module calculates the independent component projection directions of the turbulence component and the structural vibration component, and outputs the separated turbulence noise signal and structural vibration signal;

[0010] The modal offset annotation module extracts the amplitude spectrum energy of the vibration mode, calculates the percentage of energy in each direction to the total energy, constructs the energy coupling ratio matrix and marks the response offset mode;

[0011] The stability attenuation assessment module applies a variable-direction mechanical excitation wave to the modal offset marked pipeline node to calculate the transfer function matrix, compares it with the preset initial reference matrix frequency point by frequency point, and extracts phase difference data to quantify the stability attenuation of the pipe lining;

[0012] The synchronous correlation module maps the energy coupling ratio matrix and the lining stability attenuation into graph node features, analyzes the temporal evolution of modal parameters, captures synchronous events and marks synchronous correlation nodes;

[0013] The pipe lining degradation traceability module constructs a traceability tree structure consisting of a main path and alternative branches, and converts the verified traceability path tree into a decision result with geographic coordinate annotations.

[0014] In a preferred embodiment, collecting the original triaxial vibration waveform data including the timestamp, adding the timestamp and the topology label to the data and outputting it to the signal separation module specifically includes:

[0015] Three-axis low-frequency acceleration sensors are installed at preset intervals along the inner wall of the water supply network to collect three-axis vibration raw waveform data including time stamps;

[0016] At the same time, the surrounding rock strain data is collected at the same collection frequency as the original vibration waveform data;

[0017] According to the latitude and longitude coordinates of the sensor installation location and the pipeline node number, a spatial position-sensor identification mapping relationship table is established, and a standardized vibration data stream with topological labels is output;

[0018] The calibrated three-axis vibration data are encoded into double-precision floating-point format according to the axial, circumferential, and radial components, and then output to the signal separation module after adding timestamps and topology labels.

[0019] In a preferred embodiment, it is characterized in that calculating the independent component projection directions of the turbulence component and the structural vibration component and outputting the separated turbulence noise signal and structural vibration signal specifically includes:

[0020] The input standardized vibration data stream is divided into time windows, and the Hamming window function is used to eliminate spectrum leakage, and a time domain signal frame sequence with overlapping rate is output;

[0021] The kurtosis statistic is introduced as a weight factor in the FastICA objective function, and the non-Gaussianity metric is used as the separation criterion to calculate the independent component projection directions of the turbulence component and the structural vibration component.

[0022] The unmixing matrix is updated iteratively through a fixed-point algorithm. The algorithm is terminated when the rate of change of the Frobenius norm of the mixing matrix of adjacent iterations is lower than the convergence threshold, and the separated turbulence noise signal and structural vibration signal are output.

[0023] Calculate the autocorrelation peak attenuation rate of the structural vibration component. If the attenuation rate of the main peak and the secondary peak amplitude reaches below the set minimum attenuation rate threshold, it is determined to be a valid modal signal.

[0024] The structural vibration components of the effective modal signal are reorganized according to the time domain frame sequence, attached with separation confidence labels and output to the modal offset annotation module, and invalid separation data that does not meet the standards is discarded.

[0025] In a preferred embodiment, extracting the amplitude spectrum energy of the vibration mode, calculating the percentage of energy in each direction to the total energy, constructing the energy coupling ratio matrix and marking the response offset mode specifically includes:

[0026] Perform frequency domain decomposition on the input structural vibration signal, use the frequency domain subspace method to extract the amplitude spectrum energy of the axial, circumferential and radial vibration modes, and output the modal energy density distribution curve in each direction;

[0027] Integrate the axial, circumferential, and radial modal energies within the set frequency band, calculate the percentage of energy in each direction to the total energy, and generate the three-degree-of-freedom energy proportion vector;

[0028] Based on the sensor topology index table of the acquisition module, the energy proportion vectors of adjacent nodes are feature-fused to construct a three-dimensional matrix containing axial-to-annular, annular-to-radial, and axial-to-radial energy coupling ratios.

[0029] The mean and variance of the historical energy ratio of each node are calculated based on a sliding time window. If the current coupling ratio deviates from the mean by more than the set dynamic threshold, the modal offset is marked and the offset direction is recorded.

[0030] The modal offset markers and the corresponding energy coupling ratio matrix are encoded into a structured data stream, which is then appended with timestamps and topology labels and output to the stability decay assessment module.

[0031] In a preferred embodiment, applying a variable-direction mechanical excitation wave to a modal offset marked pipeline node to calculate a transfer function matrix, comparing it with a preset initial reference matrix at each frequency point, and extracting phase difference data to quantify the attenuation of the pipe lining stability specifically includes:

[0032] Apply a variable-direction mechanical excitation wave at the pipeline node corresponding to the modal offset marker, synchronously record the vibration velocity and excitation force data corresponding to each frequency point, calculate the complex ratio of vibration velocity to excitation force through Fourier transform, and form a transfer function matrix;

[0033] Calculate the current transfer function matrix and compare it with the preset initial reference matrix at each frequency point, extract the phase lag of the transfer function, and generate a curve showing the phase difference changing with frequency;

[0034] Based on the elastic wave propagation theory, a stiffness-phase mapping model is established to show the physical relationship between the propagation phase difference and the axial stiffness. The stiffness loss of the current axial stiffness relative to the initial reference is quantified according to the phase difference.

[0035] The quantified value of stiffness loss is converted into the stability response degradation attenuation expression of the pipe lining structure under frequency domain excitation, and the equivalent pipe lining stability attenuation rate is generated;

[0036] The modal offset pipeline nodes whose equivalent pipe lining stability decay rate is greater than or equal to the set equivalent pipe lining stability decay rate threshold are sent to the synchronous association module.

[0037] In a preferred embodiment, mapping the energy coupling ratio matrix and the lining stability attenuation into graph node features, analyzing the temporal evolution of modal parameters, capturing synchronous events, and marking synchronously associated nodes specifically include:

[0038] The energy coupling ratio matrix and the transfer function matrix are mapped into graph node features, and the edge connection relationship is defined and the edge weights are initialized based on the sensor topology index table.

[0039] The edge weights are dynamically adjusted according to the temporal covariance of the energy coupling ratio between nodes, and the sliding window Pearson correlation coefficient is used to quantify the intensity of the vibration propagation path to generate a dynamic adjacency matrix.

[0040] A time dimension sliding window is superimposed on the dynamic adjacency matrix, and the spectral graph convolution kernel is applied to extract spatial features. The gated recurrent unit is combined to capture the temporal evolution of modal parameters.

[0041] Calculate the cosine similarity of the modal energy ratio changes of adjacent nodes within a set time window. If the average similarity of a set number of consecutive windows exceeds the preset threshold, it is determined to be a synchronization event and the synchronization-related nodes are marked;

[0042] The synchronous associated node clusters and their topological connection relationships are encoded as a list of area numbers, attached with confidence scores, and then transmitted to the pipe lining degradation traceability module.

[0043] In a preferred embodiment, constructing a traceability tree structure including a main path and alternative branches, and converting the verified traceability path tree into a decision result with geographic coordinates specifically includes:

[0044] Align the synchronous associated node cluster with the surrounding rock strain history data according to the acquisition timestamp;

[0045] Based on the back-propagation confidence weighted algorithm, the posterior probability of each node being the source of pipe lining degradation is iteratively calculated on the pipeline topology map, and a heat map of the probability distribution of pipe lining degradation sources is output;

[0046] Based on the probability distribution heat map of the pipe lining degradation source, the maximum probability propagation path is searched and a tree structure consisting of the main path and alternative branches is constructed.

[0047] Calculate the tree path branch entropy value. If the main path entropy value is lower than the branch average entropy value and the confidence gradient descent rate meets the preset smoothness micro interval, the path convergence is determined to be valid.

[0048] The verified traceability path tree is converted into a decision result with geographic coordinate annotations, triggering an early warning signal and pushing it to the visual interface of the operation and maintenance platform.

[0049] In a preferred embodiment, triggering the warning signal and pushing it to the visual interface of the operation and maintenance platform specifically includes:

[0050] After receiving the early warning signal, the water supply network operation and maintenance platform marks the source node in red in the visual interface based on the decision results. The main path uses a solid line to continuously pass through the related nodes, and the alternative branch path uses a dotted line to mark the secondary impact area.

[0051] The technical effects and advantages of the pipe lining stability analysis and health monitoring system for urban water supply networks of the present invention are as follows:

[0052] The signal separation module optimizes the non-Gaussian metric based on the independent component analysis method, effectively separating turbulent noise from structural vibration signals. The modal offset annotation module calculates the vibration energy ratio, constructs an energy coupling ratio matrix, and identifies the response mode offset characteristics. The stability attenuation assessment module uses variable-direction mechanical excitation waves, combined with transfer function matrix analysis, to achieve a quantitative assessment of stiffness loss and generate an equivalent pipe lining stability attenuation rate. The synchronous correlation module constructs a synchronous event marking mechanism based on the temporal evolution of vibration modes, capturing spatial correlation features for traceability analysis. The pipe lining degradation traceability module constructs a traceability tree structure and, combined with geographic coordinate information, generates visual decision results, providing an accurate and efficient early warning method for pipe lining health monitoring of water supply pipelines.

[0053] The system can effectively distinguish high-frequency fluid disturbances, capture low-frequency abnormal disturbances in the pipe lining structure, monitor and trace the deterioration of the water supply pipeline lining, and realize visual early warning to ensure the long-term stable operation of the water supply network. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a structural schematic diagram of a pipe lining stability analysis and health monitoring system for urban water supply networks according to the present invention. DETAILED DESCRIPTION

[0055] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0056] Example 1

[0057] Figure 1 The present invention provides a pipe lining stability analysis and health monitoring system for urban water supply networks, including a data acquisition module, a signal separation module, a modal offset annotation module, a stability attenuation assessment module, a synchronization correlation module, and a pipe lining degradation tracing module.

[0058] The data acquisition module collects the original waveform data of the three-axis vibration including the time stamp, adds the time stamp and topology label to the data, and then outputs it to the signal separation module;

[0059] The signal separation module calculates the independent component projection directions of the turbulence component and the structural vibration component, and outputs the separated turbulence noise signal and structural vibration signal;

[0060] The modal offset annotation module extracts the amplitude spectrum energy of the vibration mode, calculates the percentage of energy in each direction to the total energy, constructs the energy coupling ratio matrix and marks the response offset mode;

[0061] The stability attenuation assessment module applies a variable-direction mechanical excitation wave to the modal offset marked pipeline node to calculate the transfer function matrix, compares it with the preset initial reference matrix frequency point by frequency point, and extracts phase difference data to quantify the stability attenuation of the pipe lining;

[0062] The synchronous correlation module maps the energy coupling ratio matrix and the lining stability attenuation into graph node features, analyzes the temporal evolution of modal parameters, captures synchronous events and marks synchronous correlation nodes;

[0063] The pipe lining degradation traceability module constructs a traceability tree structure consisting of a main path and alternative branches, and converts the verified traceability path tree into a decision result with geographic coordinate annotations.

[0064] Collect the original triaxial vibration waveform data including time stamps, add time stamps and topology labels to the data, and then output it to the signal separation module.

[0065] Three-axis low-frequency acceleration sensors are installed at preset intervals along the inner wall of the water supply network pipeline. The three-axis vibration raw waveform data including time stamps are collected in the axial (X), circumferential (Y) and radial (Z) directions of the pipeline respectively. The longitude and latitude coordinates of each sensor are recorded using a GNSS surveyor with an accuracy of better than ±0.5 meters to establish a high-precision position index.

[0066] At the same time, the surrounding rock strain data is collected at the same acquisition frequency as the original vibration waveform data. The vibration data and strain data use a unified clock synchronization protocol (IEEE 1588 PTP) to ensure that the timestamp error does not exceed 1ms to avoid phase offset caused by data alignment errors.

[0067] According to the latitude and longitude coordinates of the sensor installation location and the pipeline node number, a spatial location-sensor identification mapping table is constructed, which is stored using a hash index and outputs a standardized vibration data stream with topological labels;

[0068] The calibrated three-axis vibration data is encoded into double-precision floating-point format according to the axial, circumferential, and radial components. At the same time, the data is preprocessed and bandpass filtering is used to remove environmental noise. The vibration data is packaged according to information such as acquisition time, pipeline node number, and topology label to form a complete time series data stream and output it to the signal separation module.

[0069] The independent component projection directions of the turbulence component and the structural vibration component are calculated, and the separated turbulence noise signal and structural vibration signal are output.

[0070] The input standardized vibration data stream is segmented to ensure that each window contains sufficient time domain information for frequency domain analysis. A Hamming window function is applied to each time window to eliminate spectral leakage. A window overlap rate of 50% is used to output a time domain signal frame sequence with an overlap rate and smooth inter-frame transition.

[0071] In pipeline vibration analysis, the signal is usually composed of the superposition of water turbulence components and pipeline structure vibration components. The Kurtosis statistic is introduced as a non-Gaussian weighting factor in the FastICA objective function to enhance the signal separation effect.

[0072] Among them, kurtosis measures the non-Gaussianity of the signal. High kurtosis signals have sharp peaks, indicating fewer strong burst signals (mechanical resonance, impact vibration), and low kurtosis signals are close to Gaussian distribution, indicating more random noise (turbulent noise). Introducing kurtosis can make the algorithm more inclined to separate components with higher kurtosis and suppress components close to Gaussian noise.

[0073] The suppression of non-Gaussianity is used as the separation criterion, and the independent component projection directions are defined to calculate the turbulence component and the structural vibration component.

[0074] The fixed-point algorithm is used to iteratively update the unmixing matrix. When the rate of change of the Frobenius norm of the mixing matrix of adjacent iterations is lower than the convergence threshold, the algorithm is terminated and the separated turbulence noise signal and structural vibration signal are output. The specific process is as follows:

[0075] The formula for iteratively updating the unmixing matrix using the fixed point algorithm is:

[0076]

[0077] Where, is the separation vector at the tth iteration, is the objective function, set as , used to measure the non-Gaussianity of the signal, is the gradient of the objective function, a is the independent variable of the function, Set to , for The transposed vector of represents the mathematical expectation operation, is the normalized matrix of the input time domain signal frame sequence.

[0078] Construct an unmixing matrix based on the separation vectors in adjacent iterative steps and calculate the Frobenius norm rate of change of the unmixing matrix :

[0079]

[0080] Where, is the Frobenius norm of the unmixing matrix of the tth iteration, and when Less than When , the algorithm is considered to have converged, the iteration is terminated, and the separated turbulence noise signal and structural vibration signal are output.

[0081] Calculate the autocorrelation peak attenuation rate of the structural vibration component (calculated by the window length attenuation of the original peak after component extraction). If the main peak and secondary peak amplitude attenuation rates are below the set minimum attenuation rate threshold (specifically set according to the proportion of component peaks after decomposition, the default main peak attenuation rate threshold is 50% and the secondary peak attenuation rate threshold is 70%), it is determined to be a valid modal signal.

[0082] The verified structural vibration signals are reorganized according to the time domain frame sequence to restore the continuous time series signal. After adding the separation confidence label, it is output to the modal energy analysis module, and the invalid separation data that does not meet the standard is discarded. The separation confidence label is calculated as the complement of the mean of the autocorrelation peak decay rate of the vibration component.

[0083] The amplitude spectrum energy of the vibration mode is extracted to calculate the percentage of energy in each direction to the total energy, the energy coupling ratio matrix is constructed and the response offset mode is marked.

[0084] The input structural vibration signal is decomposed in the frequency domain. By performing a Fast Fourier Transform (FFT) on the signal, the frequency domain components of the signal are extracted. A frequency domain subspace is then constructed. Within this subspace, the vibration is decomposed into three different directions: axial, circumferential, and radial. The amplitude spectrum of each mode is extracted. The amplitude spectrum represents the vibration intensity within different frequency ranges. By calculating the energy density of these modes, the energy distribution of the vibration signal in each direction is derived. The frequency domain energy spectrum of each modal direction is plotted as an energy density distribution curve, providing the energy distribution of each direction in different frequency bands.

[0085] The axial, circumferential and radial modal energies are integrated within the set frequency bands respectively. The frequency bands in each direction are dynamically set according to the pipeline material and geometric parameters. Specifically, the natural frequency in each direction measured by the tapping test is ±15% as the frequency band boundary to ensure that more than 95% of the effective modal energy is covered. The percentage of energy in each direction to the total energy is calculated to generate a three-degree-of-freedom energy share vector (that is, the energy share of the axial, circumferential and radial directions is combined into a three-dimensional vector).

[0086] Based on the pipeline layout and sensor locations, a topological index table is used to define the relative positional relationships between sensors. Each sensor is associated with its geographic location, node number, and other information through a number. The energy contribution vectors of adjacent nodes are feature-fused to construct a three-dimensional matrix R containing the axial-to-annular, annular-to-radial, and axial-to-radial energy coupling ratios. Specifically, it is: ,in, 、 、 They are the axial, circumferential and radial modal energies, and the matrix elements are 、 、 are the axial-circumferential, circumferential-radial, and axial-radial coupling ratios, respectively.

[0087] The mean and variance of the historical energy ratios of each node are statistically analyzed based on a sliding time window. If the current coupling ratio deviates from the mean by more than a set dynamic threshold (dynamically set based on pipeline material and geometric parameters, with a default setting of 30% to monitor sudden energy coupling ratio changes), and if the energy coupling ratio in a certain area suddenly changes by more than 30%, it may be a local bending of the pipeline or deterioration of the lining stability. A modal response shift is marked and the direction of the shift is recorded.

[0088] The modal response offset markers and the corresponding energy coupling ratio matrix are encoded into a structured data stream, which is then appended with a timestamp and topology label and output to the stability attenuation assessment module.

[0089] A variable-directional mechanical excitation wave is applied to the pipeline nodes marked by modal response offset to calculate the transfer function matrix, which is compared with the preset initial reference matrix at each frequency point, and the phase difference data is extracted to quantify the stiffness loss.

[0090] Mechanical excitation waves in the offset direction are applied to the pipeline nodes corresponding to the modal offset markers. These excitation waves cover different frequency points of the pipeline to simulate different loads of the external environment on the pipeline. The vibration velocity and excitation force data corresponding to each frequency point are recorded synchronously. The complex ratio of the vibration velocity and the excitation force is calculated through Fourier transform to form a transfer function matrix, where the rows of the transfer function matrix correspond to the frequency points, and the three columns store the axial, circumferential, and radial complex values respectively.

[0091] The current transfer function matrix is calculated and compared frequency-by-frequency with a preset initial reference matrix. The phase lag of the transfer function is extracted, and a phase difference vs. frequency curve is generated. During the initial operation of the pipeline (when the pipeline is undamaged), the aforementioned frequency sweep experiment is repeated to obtain benchmark data for the complex ratio of vibration velocity to excitation force. The initial transfer function matrix is constructed by averaging these repeated measurements, and the benchmark data is stored as an axial phase reference curve.

[0092] The preset stiffness-phase mapping model is used to calculate the percentage loss of the current axial stiffness relative to the initial baseline. Based on the theory of elastic wave propagation in pipes, the stiffness-phase mapping model states that the physical relationship between phase difference and axial stiffness is that the phase difference is inversely proportional to the square root of the frequency and directly proportional to the stiffness loss. The phase difference curve is integrated over a set frequency band (consistent with the frequency bands set in each direction in the modal offset annotation module) to derive the stiffness loss. The specific expression is:

[0093]

[0094] Where, is the stiffness loss, 、 are the measured actual phase and reference phase when the frequency is f, respectively. f1 and f2 are the upper and lower limits of the set frequency band.

[0095] The response degradation degrees corresponding to different stiffness loss amounts are uniformly normalized, and the standardized equivalent pipe lining stability attenuation rate between 0 and 1 is finally output. The degraded modal offset pipeline nodes whose pipe lining stability attenuation rate exceeds the set threshold are sent to the synchronous association module through the threshold comparison method.

[0096] The energy coupling ratio matrix and the lining stability attenuation are mapped into graph node features to analyze the temporal evolution of modal parameters, capture synchronous events and mark synchronous associated nodes.

[0097] A node feature vector is constructed based on the transfer function of the energy coupling ratio matrix and the liner stability decay rate. The feature dimensions are set to include the axial-to-circumferential coupling ratio, the circumferential-to-radial coupling ratio, the axial-to-radial coupling ratio, and the corresponding liner stability decay rate. Edge connections are defined based on the sensor topology index table of the acquisition module, and the edge weights are initialized as the inverse of the Euclidean distance between nodes.

[0098] The sliding window covariance of the node energy coupling ratio vector at consecutive moments is calculated, and the Pearson correlation coefficient is used to measure the vibration propagation path strength between nodes. A time dimension sliding window is superimposed on the constructed dynamic adjacency matrix, and the spectral convolution kernel is used to extract spatial features and calculate the node state change rate.

[0099] Combined with the gated recurrent unit to process time series data, the temporal evolution trend of modal energy is extracted.

[0100] The cosine similarity of the modal energy ratio changes of adjacent nodes within a set time window is calculated. If the average similarity of a set number of consecutive windows exceeds the preset threshold (specifically set according to the size of the modal energy, the default setting is 85%), it is determined to be a synchronization event and the synchronization-related nodes are marked.

[0101] The synchronous associated node clusters and their topological connection relationships are encoded as a list of area numbers, attached with a confidence score (the confidence score is obtained by combining the similarity of adjacent windows and the initialized edge weights), and then transmitted to the pipe lining degradation traceability module.

[0102] Construct a traceability tree structure containing the main path and alternative branches, and convert the verified traceability path tree into a decision result with geographic coordinate annotations.

[0103] For the synchronous associated node cluster data set and surrounding rock strain history data received by the pipe lining degradation traceability module, the linear interpolation method is used to time align the data with different sampling rates to ensure that all data points are arranged in a unified time series to form a complete time series data set.

[0104] Based on the adjacency matrix structure of the pipeline topology graph, all nodes are first assigned a uniformly distributed initial prior probability of the pipe lining degradation source. If there is a historical fault record, the initial weight is adjusted according to the node failure frequency.

[0105] An iterative optimization algorithm using a backpropagation confidence-weighted algorithm is used. Each iteration consists of two steps: forward propagation and backpropagation. The forward propagation phase evaluates the physical attenuation effect of deformation propagation along the flow direction, based on the pipe length, support point density, and material attenuation coefficient. The confidence level of downstream nodes is then probabilistically updated. The attenuation coefficient defaults to a range of 0.8 to 0.95, with the specific value dynamically set based on the pipe lining nondestructive testing report.

[0106] In the reverse correction stage, the current modal energy ratio outlier is combined, the measurement data is converted into likelihood probability through the Bayesian formula, the measurement error is modeled using the Gamma distribution, and the upstream node confidence is corrected in reverse probability.

[0107] During the iteration process, a time decay factor is introduced to assign exponential weight to historical confidence data, ensuring that newly measured data dominates the probability distribution. After three to five iterations, the posterior probability distribution of the source of pipe lining degradation at each node is output, forming a probability density field presented as a heat map.

[0108] In the probability distribution heat map, the node with the maximum a posteriori probability is selected as the starting point of the main traceability path. The dynamic programming method is used to search for the maximum probability propagation path, and a tree structure containing the main path and alternative branches is constructed.

[0109] Information entropy analysis is performed on the generated tree paths, and the Shannon entropy of the main path and each branch path is calculated to quantify path uncertainty. The main path entropy is obtained by accumulating the logarithmically weighted sum of the confidence levels of the path nodes, and the average branch entropy is the arithmetic mean of the entropy values of all branch paths.

[0110] At the same time, the confidence sequence of the main path nodes is extracted, the confidence gradient change rate between adjacent nodes is calculated, and the fluctuation range is analyzed to see if it meets the preset smoothness threshold. Convergence verification requires that the main path entropy value is less than 80% of the average branch entropy value, and the confidence gradient fluctuation range does not exceed plus or minus 15%.

[0111] If verification fails, a new round of iterative calculations is automatically triggered until the convergence condition is met or the maximum number of iterations is reached. The dual constraint mechanism of entropy and gradient is used to ensure the physical rationality and statistical significance of the traceability path.

[0112] The verified traceability path tree is converted into a decision result with geographic coordinate annotations, triggering an early warning signal and pushing it to the visual interface of the operation and maintenance platform.

[0113] After receiving the early warning signal, the water supply network operation and maintenance platform marks the source node in red in the visual interface based on the decision results. The main path uses a solid line to continuously pass through the related nodes, and the alternative branch path uses a dotted line to mark the secondary impact area.

[0114] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0115] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0116] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0117] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0118] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0119] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0120] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0121] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0122] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0123] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A pipe lining stability analysis and health monitoring system for urban water supply networks, characterized by: It includes data acquisition module, signal separation module, modal offset annotation module, stability attenuation assessment module, synchronization correlation module and pipe lining degradation traceability module: The data acquisition module collects the original waveform data of the three-axis vibration including the time stamp, adds the time stamp and topology label to the data, and then outputs it to the signal separation module; The signal separation module calculates the independent component projection directions of the turbulence component and the structural vibration component, and outputs the separated turbulence noise signal and structural vibration signal; The modal offset annotation module extracts the amplitude spectrum energy of the vibration mode, calculates the percentage of energy in each direction to the total energy, constructs the energy coupling ratio matrix and marks the response offset mode; The stability attenuation assessment module applies a variable-direction mechanical excitation wave to the modal offset marked pipeline node to calculate the transfer function matrix, compares it with the preset initial reference matrix frequency point by frequency point, and extracts phase difference data to quantify the stability attenuation of the pipe lining; The synchronization correlation module maps the energy coupling ratio matrix and the lining stability attenuation into graph node features, analyzes the temporal evolution of modal parameters, captures synchronization events, and marks synchronization correlation nodes. Specifically, it includes: The energy coupling ratio matrix and the transfer function matrix are mapped into graph node features, and the edge connection relationship is defined based on the sensor topology label and the edge weight is initialized; The edge weights are dynamically adjusted according to the temporal covariance of the energy coupling ratio between nodes, and the sliding window Pearson correlation coefficient is used to quantify the intensity of the vibration propagation path to generate a dynamic adjacency matrix. A time dimension sliding window is superimposed on the dynamic adjacency matrix, and the spectral graph convolution kernel is applied to extract spatial features. The gated recurrent unit is combined to capture the temporal evolution of modal parameters. Calculate the cosine similarity of the modal energy ratio changes of adjacent nodes within a set time window. If the average similarity of a set number of consecutive windows exceeds the preset threshold, it is determined to be a synchronization event and the synchronization-related nodes are marked; The synchronous associated node clusters and their topological connection relationships are encoded into a list of area numbers, attached with confidence scores, and then transmitted to the pipe lining degradation traceability module; The pipe lining degradation traceability module constructs a traceability tree structure consisting of a main path and alternative branches. It converts the verified traceability path tree into a decision result with geographic coordinate annotations, specifically including: Align the synchronous associated node cluster with the surrounding rock strain history data according to the acquisition timestamp; Based on the back-propagation confidence weighted algorithm, the posterior probability of each node being the source of pipe lining degradation is iteratively calculated on the pipeline topology map, and a heat map of the probability distribution of pipe lining degradation sources is output; Based on the probability distribution heat map of the pipe lining degradation source, the maximum probability propagation path is searched and a tree structure consisting of the main path and alternative branches is constructed. Calculate the tree path branch entropy value. If the main path entropy value is lower than the branch average entropy value and the confidence gradient descent rate meets the preset smoothness micro interval, the path convergence is determined to be valid. Convert the verified traceability path tree into a decision result with geographic coordinates, trigger an early warning signal and push it to the visual interface of the operation and maintenance platform; The triggering of the warning signal and pushing it to the visual interface of the operation and maintenance platform specifically includes: After receiving the early warning signal, the water supply network operation and maintenance platform marks the source node in red in the visual interface based on the decision results. The main path uses a solid line to continuously pass through the related nodes, and the alternative branch path uses a dotted line to mark the secondary impact area.

2. The pipe lining stability analysis and health monitoring system for urban water supply networks according to claim 1 is characterized in that: Collecting the original triaxial vibration waveform data with timestamps, adding timestamps and topology labels to the data and outputting it to the signal separation module specifically includes: Three-axis low-frequency acceleration sensors are installed at preset intervals along the inner wall of the water supply network to collect three-axis vibration raw waveform data including time stamps; At the same time, the surrounding rock strain data is collected at the same collection frequency as the original vibration waveform data; According to the latitude and longitude coordinates of the sensor installation location and the pipeline node number, a spatial position-sensor identification mapping relationship table is established, and a standardized vibration data stream with topological labels is output; The calibrated three-axis vibration data are encoded into double-precision floating-point format according to the axial, circumferential, and radial components, and then output to the signal separation module after adding timestamps and topology labels.

3. The pipe lining stability analysis and health monitoring system for urban water supply networks according to claim 2 is characterized in that: Calculate the independent component projection direction of the turbulence component and the structural vibration component, and output the separated turbulence noise signal and structural vibration signal. Specifically include: The input standardized vibration data stream is divided into time windows, and the Hamming window function is used to eliminate spectrum leakage, and a time domain signal frame sequence with overlapping rate is output; The kurtosis statistic is introduced as a weight factor in the FastICA objective function, and the non-Gaussianity metric is used as the separation criterion to calculate the independent component projection directions of the turbulence component and the structural vibration component. The unmixing matrix is updated iteratively through a fixed-point algorithm. The algorithm is terminated when the rate of change of the Frobenius norm of the mixing matrix of adjacent iterations is lower than the convergence threshold, and the separated turbulence noise signal and structural vibration signal are output. Calculate the autocorrelation peak attenuation rate of the structural vibration component. If the attenuation rate of the main peak and the secondary peak amplitude reaches below the set minimum attenuation rate threshold, it is determined to be a valid modal signal. The structural vibration components of the effective modal signal are reorganized according to the time domain frame sequence, attached with separation confidence labels and output to the modal offset annotation module, and invalid separation data that does not meet the standards is discarded.

4. The pipe lining stability analysis and health monitoring system for urban water supply networks according to claim 3 is characterized in that: Extract the amplitude spectrum energy of the vibration mode and calculate the percentage of energy in each direction to the total energy. Construct the energy coupling ratio matrix and mark the response offset mode. Specifically include: Perform frequency domain decomposition on the input structural vibration signal, use the frequency domain subspace method to extract the amplitude spectrum energy of the axial, circumferential and radial vibration modes, and output the modal energy density distribution curve in each direction; Integrate the axial, circumferential, and radial modal energies within the set frequency band, calculate the percentage of energy in each direction to the total energy, and generate the three-degree-of-freedom energy proportion vector; Based on the sensor topology labels of the acquisition module, the energy proportion vectors of adjacent nodes are feature fused to construct a three-dimensional matrix containing axial-to-annular, annular-to-radial, and axial-to-radial energy coupling ratios. The mean and variance of the historical energy ratio of each node are calculated based on a sliding time window. If the current coupling ratio deviates from the mean by more than the set dynamic threshold, the modal offset is marked and the offset direction is recorded. The modal offset markers and the corresponding energy coupling ratio matrix are encoded into a structured data stream, which is then appended with timestamps and topology labels and output to the stability decay assessment module.

5. The pipe lining stability analysis and health monitoring system for urban water supply networks according to claim 4 is characterized in that: Apply a variable-directional mechanical excitation wave to the modal offset marked pipeline node to calculate the transfer function matrix, compare it with the preset initial reference matrix at each frequency point, and extract phase difference data to quantify the attenuation of the pipe lining stability. The specific steps include: Apply a variable-direction mechanical excitation wave at the pipeline node corresponding to the modal offset marker, synchronously record the vibration velocity and excitation force data corresponding to each frequency point, calculate the complex ratio of vibration velocity to excitation force through Fourier transform, and form a transfer function matrix; Calculate the current transfer function matrix and compare it with the preset initial reference matrix at each frequency point, extract the phase lag of the transfer function, and generate a curve showing the phase difference changing with frequency; Based on the elastic wave propagation theory, a stiffness-phase mapping model is established to show the physical relationship between the propagation phase difference and the axial stiffness. The stiffness loss of the current axial stiffness relative to the initial reference is quantified according to the phase difference. The quantified value of stiffness loss is converted into the stability response degradation attenuation expression of the pipe lining structure under frequency domain excitation, and the equivalent pipe lining stability attenuation rate is generated; The modal offset pipeline nodes whose equivalent pipe lining stability decay rate is greater than or equal to the set equivalent pipe lining stability decay rate threshold are sent to the synchronous association module.

Images