Optical fiber vibration and temperature measurement diagnosis method and system for drainage pipe leakage

Abstract

The invention discloses an optical fiber vibration and temperature measurement diagnosis method and system for drainage pipe leakage, and relates to the field of temperature measurement. Determining a temperature characteristic signal based on the vibration signal and a Raman scattering effect; performing wavelet packet transformation processing on the vibration signal to obtain a plurality of equal-broadband signals; eliminating noise in the plurality of equal-broadband signals, the similarity of which to a preset noise feature library is greater than a preset similarity threshold value, by adopting a minimum mean square algorithm to obtain a denoised vibration feature signal; an LSTM-GRU hybrid neural network model is constructed; splicing the output features of the LSTM layer and the GRU layer to obtain fusion features; and matching the fusion feature with a preset fault mode classification table to obtain a fault type of the drainage pipe, adding the fault type to a preset template, and outputting a diagnosis result of the drainage pipe. According to the method, the identification capability of tiny leakage in a complex environment is improved, the false alarm rate is reduced, and the accuracy of leakage detection of the drainage pipe network is improved.

Description

Technical Field

[0001] This application belongs to the field of temperature measurement, and in particular relates to a fiber optic vibration and temperature measurement diagnostic method and system for drainage pipe leakage. Background Technology

[0002] As a core pillar of modern urban infrastructure, the safe operation of drainage pipe networks has a significant impact on the sustainable development of cities. However, due to the failure of early pipe network designs to consider extreme weather conditions and structural defects in the pipes such as cracks and misaligned joints, infiltration of external water has frequently occurred, affecting the normal operation of the drainage pipe network.

[0003] In related technologies, fiber optic vibration sensing technology has been applied to monitor drainage pipe networks. This technology involves deploying fiber optic sensors along the pipe network, utilizing the Rayleigh scattering effect of optical fibers to detect pipe vibration signals, and employing traditional signal processing algorithms to identify leakage characteristics, thereby achieving real-time monitoring of pipe network leaks.

[0004] However, in complex underground environments, the vibration signals generated by pipeline leaks are relatively weak and easily masked by surrounding environmental noise. When a leak occurs, the vibration signal is superimposed on signals generated by external interference sources such as construction work and vehicle traffic, making it difficult to accurately extract the leak characteristics from the complex background signals, thereby reducing the accuracy of system detection. Summary of the Invention

[0005] This application provides a fiber optic vibration and temperature measurement diagnostic method and system for drainage pipe leaks, which improves the ability to identify minute leaks in complex environments, reduces false alarm rates, and improves the accuracy of drainage pipe network leak detection.

[0006] In the first aspect, this application provides a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage, which transmits pulsed lasers to distributed fiber optic sensors buried along the drainage pipe network to obtain pipe vibration signals. Temperature characteristic signals are determined based on vibration signals and Raman scattering effects; Wavelet packet transform is performed on the vibration signal to obtain several signals with equal bandwidth. The least mean square algorithm is used to eliminate noise from several equal-width frequency band signals whose similarity to a preset noise feature library is greater than a preset similarity threshold, thus obtaining the noise-reduced vibration feature signal; Construct an LSTM-GRU hybrid neural network model, where the LSTM layer is used to process temperature feature signals to capture slowly changing features, and the GRU layer is used to process vibration feature signals to capture abruptly changing features. The output features of the LSTM layer and the GRU layer are concatenated to obtain the fused features; The fused features are matched with a preset fault mode classification table to obtain the fault type of the drain pipe, and the fault type is added to a preset template to output the diagnostic results of the drain pipe.

[0007] By employing the above technical solution, vibration and temperature signals are acquired by deploying distributed fiber optic sensors along the drainage pipe network. Noise reduction is then performed using wavelet packet transform and the least mean square algorithm, allowing leakage characteristic signals to be extracted. An LSTM-GRU hybrid neural network model is used to process the temperature and vibration characteristic signals separately, enabling the model to simultaneously capture the slow temperature changes and abrupt vibration changes, thus improving the comprehensiveness and accuracy of feature extraction. The gating mechanism of the LSTM layer effectively preserves the long-term memory features of temperature changes, while the simplified structure of the GRU layer enables rapid response to abrupt vibration changes. The fusion of these two features enhances the model's ability to identify different types of leakage faults, improves the ability to identify minute leaks in complex environments, reduces the false alarm rate, and improves the accuracy of leakage detection in drainage pipe networks.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the temperature characteristic signal is determined based on the vibration signal and the Raman scattering effect, specifically including: Stokes intensity and anti-Stokes intensity are determined based on vibration signals and Raman scattering effects. Input the Stokes light intensity and anti-Stokes light intensity into the temperature calculation function to obtain the temperature value; The spatial diffusion characteristics of all temperature values ​​along the drainage network are analyzed using a sliding time window to obtain temperature characteristic signals.

[0009] By employing the above technical solution, the temperature value is determined by analyzing the Raman scattering effect in the optical fiber and calculating the ratio of Stokes light intensity to anti-Stokes light intensity, thus avoiding the limitations of traditional contact temperature measurement methods. Using a sliding time window to analyze the spatial diffusion characteristics of the temperature value allows the system to dynamically capture the changing patterns of the temperature field, improving the spatiotemporal resolution of the temperature characteristic signal. This non-contact, distributed temperature measurement method enables continuous monitoring of the entire drainage pipe network without damaging the pipe structure, and its measurement accuracy is unaffected by environmental factors.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the temperature calculation function is: ; In the above function, This is the temperature value. For Stokes light intensity, To counteract Stokes light intensity, is Planck's constant. At the speed of light, For Raman frequency shift, Boltzmann's constant, For anti-Stokes wavelength, This is the wavelength of Stokes light.

[0011] By adopting the above technical solution, the temperature calculation method based on scattering spectroscopy analysis does not require external temperature calibration and can directly obtain the absolute temperature value, eliminating the cumulative error of the measurement system. By describing the relationship between temperature and light intensity through a precise mathematical model, the accuracy of temperature measurement is improved, enabling the system to detect minute temperature changes and enhancing the sensitivity of leak detection.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, the LSTM-GRU hybrid neural network model includes: Multiple LSTM units, each with an input gate, a forget gate, and an output gate, are connected in series along the time dimension to extract long-term variation features of temperature characteristic signals. Multiple GRU units, each with a reset gate and an update gate, are connected in series along the time dimension to extract abrupt changes in vibration characteristic signals. The input gate of the LSTM unit controls the proportion of new information input at the current moment, the forget gate controls the proportion of historical information forgotten, and the output gate controls the proportion of control unit state output. The reset gate of the GRU unit controls the proportion of historical information reset, and the update gate controls the proportion of historical information updated. The collected samples of normal pipeline operation, small hole leakage, and crack leakage are divided into training and test sets according to a preset ratio. The cross-entropy loss function is used to calculate the error between the prediction results and the true labels. Based on the error, the parameters of the LSTM-GRU hybrid neural network model are corrected by the backpropagation algorithm until the prediction accuracy on the test set is greater than the preset accuracy threshold.

[0013] By employing the aforementioned technical solution and designing a cascaded structure of multi-layer LSTM and GRU units, deep learning of temperature and vibration characteristics was achieved. The input, forget, and output gate mechanisms of the LSTM units selectively retain and update historical information, while the reset and update gate mechanisms of the GRU units simplify the parameter structure and improve computational efficiency. The model uses a cross-entropy loss function and backpropagation algorithm for parameter optimization, continuously improving prediction accuracy through iterative training. Normal operating samples and samples with different types of leaks are proportionally divided into training and test sets to ensure the model's generalization ability. This deep learning-based feature extraction method overcomes the limitations of traditional feature engineering, adaptively learning complex patterns in the data and improving the accuracy of leak diagnosis.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the fused features are matched with a preset fault mode classification table to obtain the fault type of the drain pipe, and the fault type is added to a preset template and output as the diagnostic result of the drain pipe, specifically including: The fused features are matched with a preset fault mode classification table to obtain the fault type of the drainage pipe. When the temperature change in the fused features is less than the first temperature value and the vibration energy is less than the first vibration energy value, the fault type of the drainage pipe is determined to be minor leakage; when the temperature change is not less than the first temperature value and is less than the second temperature value, and the vibration energy is not less than the first vibration energy value and is less than the second vibration energy value, the fault type of the drainage pipe is determined to be moderate leakage; when the temperature change is not less than the second temperature value and the vibration energy is not less than the second vibration energy value, the fault type of the drainage pipe is determined to be burst-level leakage. Determine the cause analysis and recommended handling solutions corresponding to the fault types in the preset fault database; Add the fault type, the corresponding cause analysis, and the recommended handling solution to the preset template and output the diagnostic results of the drain pipe.

[0015] By employing the above technical solution and matching the fused features with a preset fault mode classification table, based on dual threshold criteria of temperature change and vibration energy, three different fault types—minor leakage, moderate leakage, and pipe rupture-level leakage—can be accurately distinguished. Once the system detects a specific fault type, it automatically searches the preset fault database for corresponding cause analysis and handling solutions, integrates this information into a preset template, and outputs diagnostic results. This provides maintenance personnel with clear repair guidance, shortens fault response time, and reduces maintenance costs.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, after adding the fault type to a preset template and outputting the diagnostic results for the drain pipe, the method further includes: Obtain a pipe connection diagram of the drainage pipes, including the locations of the connection nodes between the main pipe and the branch pipes; Obtain the water flow temperature value at the upstream branch pipe of the connection node; Calculate the theoretical value of the mixing temperature at the connection node based on the thermodynamic law of conservation of energy and the water flow temperature; Calculate the difference between the theoretical value of the mixed temperature and the measured temperature value at the same connection node using the temperature characteristic signal; When the difference is greater than the first temperature threshold and less than the second temperature threshold, the temperature characteristic signal is marked as a normal mixing state. When the difference is greater than or equal to the second temperature threshold, the fault type is determined to be a leakage fault.

[0017] By adopting the above technical solution, a fault diagnosis mechanism based on temperature difference is established by acquiring the pipeline connection diagram and the water flow temperature value of the upstream branch pipe, calculating the theoretical value of the mixing temperature at the connection node using the thermodynamic law of conservation of energy, and comparing it with the measured temperature value. This mechanism utilizes the physical law of temperature change during fluid mixing, and distinguishes between normal mixing and leakage faults by setting two temperature thresholds. By comparing the difference between the theoretical and measured values, the system can filter out temperature fluctuations caused by normal mixing, accurately identify real leakage faults, and reduce the false alarm rate.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, the theoretical value of the mixing temperature at the connection node is calculated based on the thermodynamic law of conservation of energy and the water flow temperature value, specifically including: Obtain the cross-sectional area of ​​the upstream branch pipe; Calculate the flow ratio of each branch pipe based on its cross-sectional area; Substituting the flow rate ratio and water temperature into the thermodynamic energy conservation equation, the theoretical value of the mixing temperature is obtained.

[0019] By employing the above technical solution, the flow rate ratio is calculated by obtaining the cross-sectional area of ​​the upstream branch pipe, and then substituted into the thermodynamic energy conservation equation along with the water flow temperature, thus achieving an accurate calculation of the theoretical value of the mixing temperature. This calculation method considers the actual physical characteristics and fluid dynamic parameters of the pipeline, making the theoretical calculation results more consistent with reality and improving the scientific rigor and reliability of leak diagnosis.

[0020] Secondly, embodiments of this application provide a fiber optic vibration and temperature measurement diagnostic system for drainage pipe leaks. The fiber optic vibration and temperature measurement diagnostic system for drainage pipe leaks includes: one or more processors and a memory; the memory is coupled to one or more processors, and the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the system to perform the method described in the first aspect and any possible implementation of the first aspect.

[0021] Thirdly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a system, cause the system to perform the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer program product that, when run on a system, causes the system to execute the method described in any possible implementation of the first aspect.

[0023] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. This application provides a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leaks. By deploying distributed fiber optic sensors along the drainage pipe network to acquire vibration and temperature signals, and combining wavelet packet transform and least mean square algorithm for noise reduction, leakage characteristic signals can be extracted. An LSTM-GRU hybrid neural network model is used to process temperature and vibration characteristic signals separately, enabling the model to simultaneously capture the slow temperature changes and abrupt vibration changes, improving the comprehensiveness and accuracy of feature extraction. The gating mechanism of the LSTM layer effectively preserves the long-term memory features of temperature changes, while the simplified structure of the GRU layer can quickly respond to abrupt vibration changes. The fusion of these two features enhances the model's ability to identify different types of leakage faults, improves the ability to identify minor leaks in complex environments, reduces the false alarm rate, and improves the accuracy of drainage pipe network leak detection.

[0024] 2. This application provides a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leaks. By acquiring the pipe connection diagram and the water flow temperature value of the upstream branch pipe, and combining this with the thermodynamic law of conservation of energy, the theoretical value of the mixing temperature at the connection node is calculated and compared with the measured temperature value to establish a fault judgment mechanism based on temperature difference. This mechanism utilizes the physical laws of temperature change during fluid mixing, setting two temperature thresholds to distinguish between normal mixing and leakage faults. By comparing the difference between theoretical and measured values, the system can filter out temperature fluctuations caused by normal mixing, accurately identify true leakage faults, and reduce the false alarm rate. Attached Figure Description

[0025] Figure 1 This is an application scenario diagram of a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage, as described in this application embodiment.

[0026] Figure 2 This is another application scenario diagram of a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage, as described in the embodiments of this application.

[0027] Figure 3 This is a flowchart illustrating a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leaks, as described in an embodiment of this application.

[0028] Figure 4 This is another schematic flowchart of a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage in an embodiment of this application.

[0029] Figure 5 This is a schematic diagram of the physical device structure of a fiber optic vibration and temperature measurement diagnostic system for drainage pipe leakage provided in an embodiment of this application. Detailed Implementation

[0030] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0031] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0032] The fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage provided in this application is applied to a system consisting of three parts: a signal acquisition module, a signal recognition module, and an intelligent decision-making module.

[0033] For one application scenario of this application, please refer to [link / reference needed]. Figure 1 , Figure 1 The system includes a signal transmitter, an analyzer, drainage pipes, and distributed fiber optic sensors buried along the drainage network. The system control signal transmitter emits pulsed laser light to the distributed fiber optic sensors buried along the drainage network. The analyzer receives and analyzes the pipe vibration signals to obtain... Figure 1 The reverse discrete spectrum shown is from an optical analyzer.

[0034] Figure 2 Is Figure 1 The diagram shows another application scenario where a diagnostic decision system is added to the existing system. After obtaining the inverse discrete spectrum, the optical wave analyzer sends the inverse discrete spectrum to the photoelectric converter, which then converts it and sends it to the diagnostic decision system. The diagnostic decision system then determines the fault type of the drain pipe.

[0035] The following example is used in conjunction with Figure 1 This application describes a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leaks, as described in its embodiments: Please see Figure 1 This is a schematic flowchart of a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage in an embodiment of this application.

[0036] S101. Transmit pulsed laser to distributed optical fiber sensors buried along the drainage network to obtain pipeline vibration signals; In this step, the system utilizes distributed fiber optic sensing technology to acquire vibration signals from the drainage pipe network. First, distributed fiber optic sensors are buried along the drainage pipe network. These sensors can be coherent Rayleigh scattering fiber optic sensors, coherent Brillouin scattering fiber optic sensors, or distributed Raman scattering fiber optic sensors, etc. Then, the system emits pulsed laser light to the fiber optic sensors. As the laser propagates through the fiber, it scatters when it encounters vibration. By analyzing the scattered signal, vibration information at the fiber's location can be obtained. The system can optimize the signal-to-noise ratio and spatial resolution of the vibration signal by adjusting parameters such as the laser pulse width and peak power. Besides emitting pulsed laser light, the system can also use other suitable light sources such as continuous light or coded light; this is not limited here.

[0037] Specifically, the system can acquire pipeline vibration signals in the following two ways: The system emits narrow-pulse, high-peak-power pulsed laser light into a buried distributed optical fiber sensor. As the laser propagates through the fiber, it undergoes coherent Rayleigh scattering when it encounters vibration. The scattered light carries information such as the vibration location and amplitude. The system uses a photodetector to receive the scattered light and processes it through demodulation, amplification, and filtering to obtain the vibration signals at various points in the optical fiber.

[0038] The system emits wide-pulse, high-energy pulsed laser light into the buried distributed fiber optic sensors. As the laser propagates through the fiber, it undergoes nonlinear scattering, such as stimulated Brillouin scattering or stimulated Raman scattering, when it encounters vibration. The system uses an optical filter to separate the frequency-shifted light reflecting vibration information from the scattered spectrum, and then obtains the vibration signal at each point in the fiber through photoelectric conversion, amplification, and filtering.

[0039] S102. Determine the temperature characteristic signal based on vibration signal and Raman scattering effect; The system determines temperature characteristic signals based on vibration signals and Raman scattering effects. Specifically, it determines Stokes light intensity and anti-Stokes light intensity based on vibration signals and Raman scattering effects; inputs the Stokes light intensity and anti-Stokes light intensity into the temperature calculation function to obtain the temperature value; and uses a sliding time window to analyze the spatial diffusion characteristics of all temperature values ​​along the drainage network to obtain the temperature characteristic signals.

[0040] The temperature calculation function is as follows: ; In the above function, This is the temperature value. For Stokes light intensity, To counteract Stokes light intensity, is Planck's constant. At the speed of light, For Raman frequency shift, Boltzmann's constant, For anti-Stokes wavelength, This is the wavelength of Stokes light.

[0041] In this step, the system utilizes vibration signals and the Raman scattering effect of optical fibers to measure the temperature distribution along the drainage pipe network. When the optical fiber encounters vibration, in addition to coherent scattering, inelastic scattering such as Raman scattering also occurs. The Raman scattering spectrum includes two frequency-shifting peaks: Stokes light and anti-Stokes light, whose intensity ratio is exponentially related to the fiber temperature. Therefore, by measuring the intensity ratio of Stokes light and anti-Stokes light, combined with a known temperature sensitivity coefficient, the system can calculate the temperature value at various points in the optical fiber. After obtaining the temperature value, the system uses a sliding time window to analyze the spatiotemporal distribution characteristics of the temperature and extract characteristic signals reflecting the slow temperature change trend. The system can optimize parameters such as the time window size and sliding step size to balance the temporal resolution and noise level of the characteristic signals. In addition to sliding time window analysis, the system can also use other signal processing methods such as wavelet transform and empirical mode decomposition to extract multi-scale features of temperature, which are not limited here.

[0042] The system performs a short-time Fourier transform on the vibration signal to obtain its spectrum at different times. Then, it uses an optical bandpass filter to separate the frequency bands corresponding to Stokes light and anti-Stokes light, and calculates their energy ratio. According to the Boltzmann distribution law, this energy ratio is exponentially related to temperature. The system substitutes the energy ratio into the temperature inversion formula to calculate the temperature value at each point on the optical fiber. Finally, the system uses a sliding time window to extract the long-term trend of temperature change as the temperature characteristic signal.

[0043] The system can also perform wavelet packet decomposition on the vibration signal, breaking it down into different frequency bands. Then, the system applies a modulus maxima algorithm to the wavelet coefficients corresponding to Stokes and anti-Stokes light to detect the amplitude of the scattered light. Based on Raman scattering intensity theory, the system constructs a functional relationship between Stokes and anti-Stokes light intensity and temperature, fits the function parameters using the least squares method, and then retrieves the temperature values ​​at various points in the optical fiber. Finally, the system performs empirical mode decomposition on the temperature sequence, extracting eigenmode functions reflecting slow temperature changes as temperature characteristic signals.

[0044] S103. Perform wavelet packet transform on the vibration signal to obtain several equal-width frequency band signals; In this step, the system employs wavelet packet transform to perform time-frequency analysis on the vibration signal, decomposing it into different frequency bands to obtain a series of equal-width sub-band signals. Wavelet packet transform iteratively divides the signal spectrum into equal parts, with each decomposition bisecting the current frequency band until a preset decomposition level is met. Compared to classical wavelet transform, wavelet packet transform can provide a more refined characterization of the high-frequency components and is more suitable for processing non-stationary signals such as vibration. The system can optimize the type of wavelet packet basis functions and the decomposition level based on factors such as the vibration frequency range and frequency resolution requirements of the drainage pipe network. Besides wavelet packet transform, the system can also use other time-frequency analysis methods such as Wigner-Ville distribution and Hilbert-Huang transform to achieve equal-width frequency band division; this is not limited here.

[0045] The system can select orthogonal wavelets such as Daubechies, Symlet, and Coiflet as wavelet packet basis functions to perform multi-level decomposition of the vibration signal. In each level of decomposition, the system uses low-pass and high-pass filters to divide the current frequency band into two sub-bands and downsamples the decomposition coefficients to obtain low-frequency and high-frequency sub-bands. Then, the system recursively performs the above decomposition process on the low-frequency and high-frequency sub-bands until the preset decomposition level J is reached. Finally, the system obtains 2^J sub-band signals with equal bandwidth, covering the frequency range of 0 to fs / 2, where fs is the sampling frequency.

[0046] The system can also use non-orthogonal wavelets such as Meyer, Morlet, and Mexican Hat as wavelet packet basis functions to perform multi-level decomposition of the vibration signal. In each level of decomposition, the system performs full-sample convolution on the wavelet coefficients of the current frequency band to obtain wavelet coefficients for two sub-bands: low-frequency and high-frequency. Then, the system recursively performs the above decomposition process on the low-frequency and high-frequency sub-bands until the preset time-frequency resolution requirement is met. Finally, the system performs inverse transform on the wavelet coefficients of each sub-band to reconstruct a series of equal-width frequency band signals. Compared with orthogonal wavelet packets, non-orthogonal wavelet packets are more flexible in frequency band division, but their computational complexity is also higher.

[0047] S104. The least mean square algorithm is used to eliminate noise in several equal-width frequency band signals whose similarity to the preset noise feature library is greater than the preset similarity threshold, so as to obtain the noise-reduced vibration feature signal. In this step, the system uses the least mean square algorithm to denoise the sub-band signals obtained from wavelet packet decomposition. The system pre-establishes a noise feature library containing frequency domain templates of common noise types in drainage pipe networks. For each sub-band signal, the system calculates its similarity to each template in the noise feature library, using metrics such as cross-correlation coefficient and Euclidean distance. When the similarity exceeds a preset threshold, the system considers that the sub-band mainly contains noise components and needs to be removed. The system uses the least mean square algorithm to find the optimal noise estimate, eliminating noise from the sub-band signal to obtain the denoised frequency band signal. After performing the denoising process on all sub-bands, the system concatenates the sub-bands to restore the denoised time-domain vibration signal and defines it as the vibration feature signal. The system can optimize the similarity threshold using methods such as cross-validation to achieve a balance between noise removal and signal feature preservation. Besides the least mean square algorithm, the system can also use other denoising methods such as wavelet thresholding and singular value decomposition, which are not limited here.

[0048] The system can calculate the envelope of each frequency domain template in the noise feature library, constructing upper and lower bounds for the noise sub-band energy. For each sub-band signal to be processed, the system calculates its envelope and compares it with the upper and lower bounds for the noise sub-band energy. If the envelope of the sub-band to be processed falls entirely between the upper and lower bounds, the system determines that the sub-band is noise-dominant and needs to be removed. The system uses least mean square estimation to obtain the noise component of the sub-band and subtracts it from the original signal to achieve noise reduction.

[0049] The system can also utilize the Dynamic Time Warp (DTW) algorithm to measure the similarity between the sub-band to be processed and the noise template. DTW allows for a certain degree of temporal scaling by finding the optimal matching path between two sequences and calculating their distance. The system calculates the DTW distance of the sub-band to be processed using each template in the noise feature library as a reference. If the minimum DTW distance is less than a preset threshold, the system determines that the sub-band is noise-dominated. Then, the system finds the noise signal corresponding to the optimal matching path and eliminates it using least mean square estimation.

[0050] S105. Construct an LSTM-GRU hybrid neural network model; The system constructs an LSTM-GRU hybrid neural network model, where LSTM layers are used to process temperature feature signals to capture slowly changing features, and GRU layers are used to process vibration feature signals to capture abrupt changes. The LSTM-GRU hybrid neural network model includes: Multiple LSTM units, each with an input gate, a forget gate, and an output gate, are connected in series along the time dimension to extract long-term variation features of temperature characteristic signals. Multiple GRU units, each with a reset gate and an update gate, are connected in series along the time dimension to extract abrupt changes in vibration characteristic signals. The input gate of the LSTM unit controls the proportion of new information input at the current moment, the forget gate controls the proportion of historical information forgotten, and the output gate controls the proportion of control unit state output. The reset gate of the GRU unit controls the proportion of historical information reset, and the update gate controls the proportion of historical information updated. The collected samples of normal pipeline operation, small hole leakage, and crack leakage are divided into training and test sets according to a preset ratio. The cross-entropy loss function is used to calculate the error between the prediction results and the true labels. Based on the error, the parameters of the LSTM-GRU hybrid neural network model are corrected by the backpropagation algorithm until the prediction accuracy on the test set is greater than the preset accuracy threshold.

[0051] In this step, the system constructs an LSTM-GRU hybrid neural network model to fuse temperature and vibration features for pipeline fault diagnosis. LSTM (Long Short-Term Memory) and GRU (Gated Recurrent Unit) are both types of recurrent neural networks, effectively modeling long-short-term dependencies in time-series data. The system combines LSTM and GRU layers according to a specific topology to form a hybrid model. The LSTM layers are primarily used to extract the gradually varying features of the temperature signal, while the GRU layers are primarily used to extract the abrupt changes in the vibration signal. The system optimizes the number of layers, hidden units, and other hyperparameters of the hybrid model based on sample data and introduces attention mechanisms and residual connections to overcome problems such as vanishing gradients and overfitting. Besides LSTM and GRU, the system can also use other deep learning models such as convolutional neural networks and autoencoders to construct different hybrid structures; this is not limited here.

[0052] The system can combine LSTM and GRU layers in parallel. Temperature and vibration feature signals are input into the LSTM and GRU layers respectively to extract their respective deep features. Then, the system concatenates the outputs of the LSTM and GRU, performs feature fusion through a fully connected layer, and connects a softmax classifier at the output to achieve fault mode recognition. The parallel structure can fully leverage the respective advantages of LSTM and GRU while reducing information interference between the two paths.

[0053] The system can also combine LSTM and GRU layers in a cascaded manner. The temperature feature signal is first input into the LSTM layer to extract the slowly varying temperature characteristics. Then, the hidden state vector of the LSTM layer, along with the vibration feature signal, is input into the GRU layer, where feature sequence extraction and information fusion are performed. Finally, the system connects the GRU output to a fully connected layer and a classifier to output the fault diagnosis result. This cascaded structure enables deep interaction between the two feature sets, better uncovering the correlation between temperature and vibration signals.

[0054] S106. Concatenate the output features of the LSTM layer and the GRU layer to obtain the fused features; In this step, the system fuses the gradually varying temperature features extracted by the LSTM layer and the abrupt vibration features extracted by the GRU layer to obtain a fused feature vector that comprehensively reflects the pipeline state. The system uses feature concatenation to achieve this fusion, which involves concatenating the output vectors of the LSTM and GRU end-to-end to generate a longer feature vector. Concatenation is a simple and direct feature fusion method that can fully preserve the original information of different features. To improve the expressive power of the fused features, the system can stack several fully connected layers on the concatenated feature vector to achieve deep feature extraction and compression through nonlinear transformations. Besides feature concatenation, the system can also use other fusion methods such as feature weighted averaging and feature outer product to generate different types of fused features; this is not limited here.

[0055] The system can perform flattening operations at the output positions of the LSTM and GRU layers, respectively, converting the two-dimensional output matrices into one-dimensional vectors. Then, the system concatenates the two one-dimensional vectors sequentially to obtain the fused features. If the output dimension of the LSTM layer is [N, 64] and the output dimension of the GRU layer is [N, 128], the dimension of the flattened and concatenated fused features will be [N, 192], where N is the batch size. The concatenated fused features can be directly input into the subsequent fault mode classifier.

[0056] The system can also connect a fully connected layer at the output of the LSTM and GRU layers respectively, mapping the two-dimensional matrix into lower-dimensional feature vectors. The fully connected layer can learn a dimensionality reduction matrix during training, compressing the original features into a common low-dimensional space and extracting more abstract and expressive feature representations. Then, the system concatenates the two low-dimensional feature vectors to obtain the final fused features. Compared to method one, method two has lower dimensionality but higher feature quality in the fused features.

[0057] S107. Match the fused features with the preset fault mode classification table to obtain the fault type of the drain pipe, and add the fault type to the preset template to output the diagnostic result of the drain pipe.

[0058] The system matches the fused features with a preset fault mode classification table to obtain the fault type of the drain pipe. The fault type is then added to a preset template and output as the diagnostic result for the drain pipe. Specifically, this includes: matching the fused features with the preset fault mode classification table to obtain the fault type of the drain pipe; when the temperature change in the fused features is less than a first temperature value and the vibration energy is less than a first vibration energy value, the fault type of the drain pipe is determined to be minor leakage; when the temperature change is not less than a first temperature value and less than a second temperature value, and the vibration energy is not less than a first vibration energy value and less than a second vibration energy value, the fault type of the drain pipe is determined to be moderate leakage; when the temperature change is not less than a second temperature value and the vibration energy is not less than a second vibration energy value, the fault type of the drain pipe is determined to be burst-level leakage; the system determines the cause analysis and recommended treatment plan corresponding to the fault type in the preset fault database; and the fault type, its corresponding cause analysis, and recommended treatment plan are added to a preset template and output as the diagnostic result for the drain pipe.

[0059] In this step, the system uses fused features to identify and diagnose the fault types of drainage pipelines. The system pre-constructs a fault mode classification table, which contains typical characteristic patterns of various common faults (such as the range of temperature and vibration amplitude changes, spectral distribution, etc.) along with corresponding fault causes and handling suggestions. The system matches the fused features with each fault mode in the classification table, calculating the similarity between the fused features and each mode. Similarity can be measured using metrics such as Euclidean distance or Mahalanobis distance. The system selects the fault mode with the highest similarity and diagnoses it as the most likely fault type of the drainage pipeline. Then, the system extracts the cause analysis and handling suggestions for this fault type from the classification table, fills them into a preset report template, and automatically generates a fault diagnosis report. The system can provide a confidence level description of the fault type determination in the report based on the credibility of the diagnostic decision. Besides the similarity-based diagnostic method, the system can also use other classification algorithms such as support vector machines and decision trees to construct a fault diagnosis discrimination model; this is not limited here.

[0060] For example, when the fault type is minor leakage, the diagnostic report may include the following: Fault characteristics: Leakage rate ≤2 L / min, localized ground damp but no sewage overflow, manifested as temperature change ΔT<2℃ and vibration energy <0.5m / s²; Cause analysis: Micro-cracks may appear in plastic pipes due to ultraviolet radiation or chemical corrosion, and the rubber rings of socket joints may age, leading to sealing failure.

[0061] Recommended treatment: Use UV-CIPP (ultraviolet curing repair technology), repair time <4 hours, inject elastic sealant (such as polyurethane) at the interface.

[0062] When the fault type is moderate leakage, the diagnostic report may include the following: Fault characteristics: Leakage rate 2-10 L / min, continuous water accumulation on the ground, manifested as 2℃ ≤ temperature change ΔT < 5℃ and 0.5m / s² ≤ vibration energy < 1.5m / s²; Cause analysis: Acidic sewage (pH<5) caused pitting corrosion to penetrate the metal pipes; heavy vehicles running over the pipes caused a decrease in the compaction of the soil covering the pipes (compaction degree<90%); and rigid joints developed fatigue cracks due to temperature changes. Recommended solutions: Use HDPE spiral pipe segments for on-site repair (construction period < 6 hours), or in-situ curing (IPC) method, and install flexible joint compensators.

[0063] When the fault type is a pipe rupture-level leak, the diagnostic report may include the following: Fault characteristics: Leakage rate >10 L / min, sewage gushing to form a water pit with a diameter >3m, and severe liquefaction of the surrounding soil, manifested as temperature change ΔT >5℃ and vibration energy >1.5m / s²; Cause analysis: brittle fracture of cast iron pipelines due to hydrogen embrittlement (service life > 30 years), sudden settlement of soft soil foundation (single-day settlement > 20 mm), direct impact of third-party construction machinery on the pipeline (common in municipal road excavation), pump station failure causing instantaneous pressure to exceed the pipeline's pressure bearing limit, etc. Recommended solution: Activate the intelligent interception system to isolate the leak within 5 minutes, temporarily seal with quick-setting concrete, permanently repair the damaged section by blasting, replace the ductile iron pipe and install a reinforced inspection well.

[0064] The specific format and content of the diagnostic report are not limited here.

[0065] The system can construct a multi-classifier based on fused features, and use the softmax function to map the fused features to the probability distribution of each fault type. Building upon the fault mode classification table, the system further refines the feature representation of each fault category, extracts the fused features of fault samples as a training set, and trains the parameters of the softmax classifier. During diagnosis, the system inputs the fused features of the pipeline to be diagnosed into the trained classifier, selects the category with the highest output probability as the diagnostic result, and provides a confidence assessment based on the probability value.

[0066] The system can also construct a dictionary from the typical feature patterns in the fault mode classification table and use a sparse representation method to diagnose the fused features. By minimizing the reconstruction error, the system represents the fused features as a linear combination of several typical patterns in the dictionary, and the combination coefficient reflects the correlation between the fused features and each pattern. The system selects the fault mode with the largest combination coefficient as the diagnosis result. Compared with the softmax classifier, sparse representation diagnosis does not require a large number of training samples, but its generalization ability may be weaker.

[0067] In the above embodiments, vibration and temperature signals are acquired by deploying distributed fiber optic sensors along the drainage pipe network. Noise reduction is then performed using wavelet packet transform and the least mean square algorithm, allowing leakage characteristic signals to be extracted. An LSTM-GRU hybrid neural network model is used to process the temperature and vibration characteristic signals separately, enabling the model to simultaneously capture the slow temperature changes and abrupt vibration changes, improving the comprehensiveness and accuracy of feature extraction. The gating mechanism of the LSTM layer effectively preserves the long-term memory features of temperature changes, while the simplified structure of the GRU layer can quickly respond to abrupt vibration changes. The fusion of these two features enhances the model's ability to identify different types of leakage faults, improves the ability to identify minute leaks in complex environments, reduces the false alarm rate, and improves the accuracy of leakage detection in the drainage pipe network.

[0068] The above embodiments achieve intelligent diagnosis of leaks in drainage pipe networks through distributed fiber optic sensing measurement combined with deep learning. However, in practical applications, drainage pipe networks often have complex connection structures, and water mixing can occur at the connection nodes between main and branch pipes. This mixing may cause changes in temperature signals, thus affecting the accuracy of leak diagnosis. The following section will discuss... Figure 2 Another fiber optic vibration and temperature measurement diagnostic method for drainage pipe leakage is described in the embodiments of this application: Please see Figure 2 This is another flowchart illustrating a fiber optic vibration and temperature measurement diagnostic method for drainage pipe leaks in this application.

[0069] S201. Obtain a pipe connection diagram of the drainage pipe, including the location of the connection nodes between the main pipe and the branch pipe; In this step, the system acquires a pipeline connection diagram of the drainage network, which describes the topological connections between main pipes and branch pipes. The pipeline connection diagram can be in the form of electronic CAD drawings, vector layers in a GIS system, etc. The system uses image recognition and vectorization methods to extract the spatial location information of the main pipes, branch pipes, and their intersection nodes from the pipeline connection diagram. For complex pipeline structures, the system can also infer the upstream and downstream connections based on additional information such as pipe attribute labels and flow direction arrows. The acquisition of the pipeline connection diagram can be achieved through manual input, automatic scanning, etc., and is not limited here.

[0070] The system can access the GIS database of the drainage pipe network, retrieve the spatial coordinate information of main pipes, branch pipes, and their connecting nodes using SQL queries, and organize the query results into a topology diagram using spatial data formats such as shapefile and geodatabase. The topology diagram contains attribute information such as the spatial location, length, and diameter of the pipes, as well as the adjacency relationships between pipes. Based on the topology diagram, the system can generate pipe connection relationship diagrams required for subsequent analysis.

[0071] The system interface provides an upload interface for pipe connection diagrams, allowing users to upload pipe plan drawings created using vector design software such as AutoCAD. The system parses the uploaded vector graphics, extracting main pipes, branch pipes, and other elements corresponding to different layers. Then, the system traverses each pipe element, analyzes their adjacency and connectivity relationships in the topology, and automatically generates a node-pipe mapping table to form a complete pipe connection diagram.

[0072] S202. Obtain the water flow temperature value at the upstream branch pipe of the connection node; In this step, the system acquires real-time water flow temperature data by deploying temperature sensors at the upstream branch pipe location of the connection node between the main pipe and the branch pipe. Temperature sensors can be thermocouples, resistance temperature detectors (RTDs), infrared thermometers, etc., and the sensor model with a suitable measurement range and accuracy should be selected based on the branch pipe diameter, flow velocity, and other conditions. To ensure the representativeness of the temperature data, the sensors should be deployed at a certain distance from the connection node at the branch pipe inlet to avoid the influence of mixed flow at the connection node. The temperature sensors can be installed using methods such as pipe wall fixing or insertion. Data transmission between the sensors and the data acquisition terminal can be wired or wireless, without limitation here.

[0073] A hole can be drilled in the wall of the upstream branch pipe at the connection node, and a resistance temperature sensor (RTS) can be inserted, allowing the sensor probe to directly contact the water flow. The RTS material can be platinum, copper, etc., with a temperature measurement range of 0~100℃ and a resolution better than 0.1℃. The temperature sensor is connected to a local data acquisition unit via a signal line. The data acquisition unit samples and quantizes the analog signal output by the sensor at regular intervals, and then uploads the digitized temperature value to the monitoring center.

[0074] An infrared thermometer can also be installed on the vertical section of the upstream branch pipe at the connection node to measure the surface temperature of the water flow in a non-contact manner. The infrared thermometer utilizes the principle of infrared radiation thermal imaging, receiving the infrared radiation energy emitted from the water surface and converting it into a corresponding temperature value. The optical system of the infrared thermometer can be selected according to the diameter of the branch pipe to ensure a suitable measurement field of view. The thermometer uploads the collected temperature data to the monitoring center via communication interfaces such as RS485.

[0075] S203. Calculate the theoretical value of the mixing temperature at the connection node based on the thermodynamic law of conservation of energy and the water flow temperature. The system calculates the theoretical mixing temperature at the connection node based on the thermodynamic law of conservation of energy and the water flow temperature. Specifically, this includes: obtaining the cross-sectional area of ​​the upstream branch pipe; calculating the flow rate ratio of each branch pipe based on the cross-sectional area; and substituting the flow rate ratio and the water flow temperature into the thermodynamic law of conservation of energy to obtain the theoretical mixing temperature.

[0076] In this step, the system utilizes the thermodynamic law of conservation of energy to establish an energy balance equation for the mixed water flow at the connection node, calculating the theoretical value of the mixed water flow temperature at the connection node. The system first acquires the cross-sectional area data of each upstream branch pipe and estimates the flow rate ratio of each branch pipe based on the continuity equation. Then, the system substitutes the flow rate ratio of the branch pipes and the measured water temperature into the energy conservation equation to solve for the mixed water flow temperature at the connection node. The energy conservation equation assumes that the mixing process is adiabatic, meaning there is no heat exchange with the outside environment during mixing. The system can assess the rationality of the adiabatic assumption based on factors such as the pipe's burial depth, material, and the thermal conductivity of the surrounding soil, and may introduce a heat loss correction factor if necessary; however, this is not limited here.

[0077] S204. Calculate the difference between the theoretical value of the mixed temperature and the measured temperature value of the temperature characteristic signal at the same connection node; In this step, the system compares the theoretical value of the mixed temperature at the connection node calculated in the previous step with the actual temperature value measured by the distributed fiber optic sensor at that node, and calculates the difference between the two. The system extracts the measured temperature value corresponding to the current connection node location from the temperature feature signal obtained in step S102. Then, the system subtracts the measured temperature value from the theoretical value of the mixed temperature to obtain the temperature difference value, which is used for subsequent leak fault diagnosis. The temperature difference value can be calculated using simple subtraction, or it can be expressed in the form of relative error, absolute error, etc., without limitation here.

[0078] S205. When the difference is greater than the first temperature threshold and less than the second temperature threshold, the temperature characteristic signal is marked as normal mixed flow state. In this step, the system compares the temperature difference with preset thresholds to determine whether the temperature anomaly at the current connection node is caused by normal flow mixing or by a pipe leak. The system first sets two temperature thresholds: the first threshold is the upper limit of normal fluctuation in the theoretical value of the mixing temperature, and the second threshold is the lower limit of abnormal temperature rise caused by a leak. When the temperature difference falls between the first and second thresholds, the system considers the temperature anomaly to be caused by normal branch pipe flow mixing, marks the temperature characteristic signal as normal flow mixing, and does not trigger a leak alarm. The setting of the two temperature thresholds needs to comprehensively consider factors such as daily and seasonal variations in the pipe network water temperature, as well as the influence of pipe material and burial depth on the leakage temperature rise amplitude; these are not limited here.

[0079] S206. When the difference is greater than or equal to the second temperature threshold, the fault type is determined to be a leakage fault.

[0080] In this step, the system determines whether a leak has occurred in the drainage pipe based on a comparison between the temperature difference and a preset threshold. The system first calculates the difference between the current pipe temperature and the normal operating temperature, then compares this difference with a second temperature threshold. This second temperature threshold is an empirical value, typically set based on factors such as pipe material, installation environment, and fluid temperature, and characterizes the severity of abnormal temperature changes. When the difference is greater than or equal to the second temperature threshold, it indicates a significant abnormal increase or decrease in pipe temperature, exceeding the normal fluctuation range, and the system determines that a leak has occurred. Leakage is a common form of damage to drainage pipes, and the temperature at the leak point is usually significantly different from the surrounding environment. The system can further assess the severity of the leak based on the magnitude of the temperature difference. In addition to a single temperature threshold judgment, the system can also integrate other monitoring indicators such as vibration and acoustic emission to form a multi-parameter joint diagnostic criterion, improving the reliability of leak diagnosis; however, this is not limited here.

[0081] In the above embodiments, by acquiring the pipeline connection diagram and the water flow temperature value of the upstream branch pipe, and combining this with the thermodynamic law of conservation of energy to calculate the theoretical value of the mixing temperature at the connection node, a fault judgment mechanism based on temperature difference is established by comparing and analyzing this theoretical value with the measured temperature value. This mechanism utilizes the physical law of temperature change during fluid mixing, and distinguishes between normal mixing and leakage faults by setting two temperature thresholds. By comparing the difference between the theoretical and measured values, the system can filter out temperature fluctuations caused by normal mixing, accurately identify real leakage faults, and reduce the false alarm rate.

[0082] The system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference needed]. Figure 3 This is a schematic diagram of the physical device structure of a fiber optic vibration and temperature measurement diagnostic system for drainage pipe leakage provided in an embodiment of this application.

[0083] It should be noted that, Figure 3 The structure of the system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0084] like Figure 3 As shown, the system includes a Central Processing Unit (CPU) 301, which can perform various appropriate actions and processes based on a program stored in Read-Only Memory (ROM) 302 or a program loaded from storage portion 308 into Random Access Memory (RAM) 303, such as executing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An Input / Output (I / O) interface 305 is also connected to the bus 304.

[0085] The following components are connected to I / O interface 305: input section 306 including a camera, infrared sensor, etc.; output section 307 including a liquid crystal display (LCD) and speakers, etc.; storage section 308 including a hard disk, etc.; and communication section 309 including a network interface card such as a LAN (Local Area Network) card and a modem, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0086] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs the various functions defined in the present invention.

[0087] It should be noted that the computer-readable medium shown in the embodiments of the present invention can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, wherein a computer-readable computer program is carried. The transmitted data signal can take many forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof.

[0088] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0089] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the system described in the above embodiments; or it may exist independently and not assembled into the system. The storage medium carries one or more computer programs that, when executed by a processor of a system, cause the system to implement the methods provided in the above embodiments.

[0090] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0091] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0092] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. 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 (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0093] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A fiber optic vibration and temperature measurement diagnostic method for drainage pipe leaks, characterized in that, include: Pulsed lasers are emitted to distributed fiber optic sensors buried along the drainage network to obtain pipeline vibration signals. The temperature characteristic signal is determined based on the vibration signal and the Raman scattering effect. The vibration signal is processed by wavelet packet transform to obtain several equal-width frequency band signals; The least mean square algorithm is used to eliminate noise in the several equal-width frequency band signals whose similarity to a preset noise feature library is greater than a preset similarity threshold, so as to obtain the noise-reduced vibration feature signal; Construct an LSTM-GRU hybrid neural network model, wherein the LSTM layer is used to process the temperature feature signal to capture slowly changing features, and the GRU layer is used to process the vibration feature signal to capture abruptly changing features; The output features of the LSTM layer and the GRU layer are concatenated to obtain the fused features; The fused features are matched with a preset fault mode classification table to obtain the fault type of the drain pipe, and the fault type is added to a preset template and output as the diagnostic result of the drain pipe.

2. The method according to claim 1, characterized in that, The determination of the temperature characteristic signal based on the vibration signal and Raman scattering effect specifically includes: The Stokes intensity and anti-Stokes intensity are determined based on the vibration signal and the Raman scattering effect. The Stokes light intensity and the anti-Stokes light intensity are input into the temperature calculation function to obtain the temperature value; The spatial diffusion characteristics of all temperature values ​​along the drainage network are analyzed using a sliding time window to obtain temperature characteristic signals.

3. The method according to claim 2, characterized in that, The temperature calculation function is: ； In the above function, the The temperature value, the The Stokes light intensity, the For the anti-Stokes light intensity, the Let be Planck's constant, the For the speed of light, the For Raman frequency shift, the The Boltzmann constant is stated as follows. For the anti-Stokes wavelength, the This is the wavelength of Stokes light.

4. The method according to claim 1, characterized in that, The LSTM-GRU hybrid neural network model includes: Multiple LSTM units connected in series along the time dimension, each LSTM unit having an input gate, a forget gate, and an output gate, are used to extract the long-term variation features of the temperature feature signal; multiple GRU units connected in series along the time dimension, each GRU unit having a reset gate and an update gate, are used to extract the abrupt change features of the vibration feature signal; the input gate of the LSTM unit is used to control the input ratio of new information at the current moment, the forget gate is used to control the forgetting ratio of historical information, and the output gate is used to control the output ratio of the control unit state; the reset gate of the GRU unit is used to control the reset ratio of the historical information, and the update gate is used to control the update ratio of the historical information. The collected samples of normal pipeline operation, small hole leakage, and crack leakage are divided into training and test sets according to a preset ratio; the error between the prediction result and the true label is calculated using the cross-entropy loss function; based on the error, the parameters of the LSTM-GRU hybrid neural network model are corrected using the backpropagation algorithm until the prediction accuracy on the test set is greater than a preset accuracy threshold.

5. The method according to claim 1, characterized in that, The step of matching the fused features with a preset fault mode classification table to obtain the fault type of the drain pipe, and adding the fault type to a preset template to output the diagnostic result of the drain pipe, specifically includes: The fused features are matched with a preset fault mode classification table to obtain the fault type of the drain pipe. When the temperature change in the fused features is less than a first temperature value and the vibration energy is less than a first vibration energy value, the fault type of the drain pipe is determined to be minor leakage; when the temperature change is not less than the first temperature value and is less than a second temperature value, and the vibration energy is not less than the first vibration energy value and is less than the second vibration energy value, the fault type of the drain pipe is determined to be moderate leakage; when the temperature change is not less than the second temperature value and the vibration energy is not less than the second vibration energy value, the fault type of the drain pipe is determined to be burst-level leakage. Determine the cause analysis and recommended handling solution corresponding to the fault type in the preset fault database; The fault type, the corresponding cause analysis, and the recommended handling solution are added to the preset template and output as the diagnostic results of the drain pipe.

6. The method according to claim 1, characterized in that, After adding the fault type to the preset template and outputting it as the diagnostic result of the drain pipe, the method further includes: Obtain a pipe connection diagram showing the connection points of the main pipe and branch pipes in the drainage pipe; The water flow temperature value is obtained at the upstream branch pipe of the connection node; Calculate the theoretical value of the mixing temperature at the connection node based on the thermodynamic law of conservation of energy and the water flow temperature value; Calculate the difference between the theoretical value of the mixed temperature and the measured temperature value of the temperature characteristic signal at the same connection node; When the difference is greater than the first temperature threshold and less than the second temperature threshold, the temperature characteristic signal is marked as a normal mixing state; When the difference is greater than or equal to the second temperature threshold, the fault type is determined to be a leakage fault.

7. The method according to claim 6, characterized in that, The calculation of the theoretical mixing temperature at the connection node based on the thermodynamic law of conservation of energy and the water flow temperature specifically includes: Obtain the cross-sectional area of ​​the upstream branch pipe; Calculate the flow ratio of each branch pipe based on the cross-sectional area; Substituting the flow rate ratio and the water temperature into the thermodynamic energy conservation equation, the theoretical value of the mixing temperature is obtained.

8. A fiber optic vibration and temperature measurement diagnostic system for drainage pipe leaks, characterized in that, The system includes: One or more processors and a memory; the memory is coupled to the one or more processors, the memory being used to store computer program code, the computer program code including computer instructions, the one or more processors invoking the computer instructions to cause the system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instructions are executed on the system, the system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the system, the system performs the method as described in any one of claims 1-7.

Images