Underground pipe gallery grouting material leakage monitoring method and system driven by Internet of Things

Abstract

The invention discloses an internet-of-things-driven underground pipe gallery grouting material leakage monitoring method and system, and relates to the technical field of data processing and analysis, and the method comprises the steps: obtaining a dynamic parameter data set; generating a grouting leakage feature vector; positioning a potential leakage coordinate according to the leakage risk grade; and triggering a grading early warning instruction through the Internet of Things platform. The technical problems of multi-source data isolation, low positioning precision and early warning lag in a traditional monitoring means in the underground pipe gallery grouting project are solved, and the technical effects that multi-source data fusion analysis is achieved, the positioning precision is improved, leakage early warning can be conducted in time, and then safe and stable operation of the underground pipe gallery grouting project is guaranteed are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing and analysis, and particularly to an Internet of Things-driven leakage monitoring method and system for grouting materials in an underground utility tunnel. Background Art

[0002] In the construction and operation and maintenance of underground utility tunnels, the problem of leakage of grouting materials is related to the structural safety and service life of the utility tunnels. At present, for the leakage monitoring of grouting materials in underground utility tunnels, traditional methods mainly rely on manual regular inspections or simple single-point sensor monitoring. Manual inspections not only have low efficiency and long time consumption, but are also greatly affected by subjective factors and are difficult to detect hidden leakage points; single-point sensor monitoring can only obtain local information and cannot comprehensively grasp the overall leakage situation of the utility tunnel. With the development of technologies such as the Internet of Things and big data, the requirements for the accuracy and real-time performance of underground utility tunnel leakage monitoring are increasing day by day, and the limitations of these traditional monitoring means are becoming more and more prominent, and they cannot meet the needs of accurately positioning, real-time monitoring and effective early warning of the leakage of grouting materials in underground utility tunnels. There is an urgent need for more advanced monitoring technologies to ensure the safe and stable operation of underground utility tunnels. Summary of the Invention

[0003] This application solves the technical problems of isolated multi-source data, low positioning accuracy and late warning in traditional monitoring means in the underground utility tunnel grouting project. This application deploys a multi-source sensor network in the grouting section of the underground utility tunnel to collect dynamic parameter data, generates a grouting leakage feature vector through multi-modal feature extraction, uses a leakage risk model to determine the leakage risk level and locate potential leakage coordinates, generates a leakage monitoring report by combining the risk level and coordinates, and triggers a hierarchical early warning, so as to accurately monitor the leakage of grouting materials in the underground utility tunnel and make the leakage monitoring of the underground utility tunnel grouting more efficient, accurate and reliable.

[0004] In view of the above technical problems, this application proposes a technical solution for an Internet of Things-driven leakage monitoring method and system for grouting materials in an underground utility tunnel.

[0005] In a first aspect, this application provides an Internet of Things-driven leakage monitoring method for grouting materials in an underground utility tunnel. The method includes: performing real-time collection on an underground utility tunnel through an Internet of Things sensor network to obtain a dynamic parameter data set; performing multi-modal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector; inputting the grouting leakage feature vector into a leakage risk model for real-time analysis to determine a leakage risk level, and locating potential leakage coordinates according to the leakage risk level; generating a leakage monitoring report according to the leakage risk level and the potential leakage coordinates, sending the leakage monitoring report to an Internet of Things platform, and triggering a hierarchical early warning instruction through the Internet of Things platform.

[0006] In a second aspect, the present application provides an Internet of Things (IoT)-driven leakage monitoring system for grouting materials in an underground utility tunnel. The system includes: a data set acquisition module for real-time collection of the underground utility tunnel through an IoT sensor network to obtain a dynamic parameter data set; a vector generation module for performing multi-modal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector; a coordinate positioning module for inputting the grouting leakage feature vector into a leakage risk model for real-time analysis to determine the leakage risk level, and positioning potential leakage coordinates based on the leakage risk level; and an instruction trigger module for generating a leakage monitoring report based on the leakage risk level and the potential leakage coordinates, sending the leakage monitoring report to the IoT platform, and triggering a hierarchical warning instruction through the IoT platform.

[0007] The present application proposes one or more technical solutions, having at least the following technical effects:

[0008] In the present application, an IoT sensor network is deployed in the grouting area of the underground utility tunnel to collect a dynamic parameter data set in real time. Multi-modal feature extraction is performed on the data to generate a grouting leakage feature vector. The vector is input into a leakage risk model to determine the leakage risk level and potential leakage coordinates. A leakage monitoring report is generated based on the risk level and coordinates, and a hierarchical warning is triggered. In this process, multi-source data fusion is achieved by collaborative collection of multi-source sensors, the positioning accuracy is improved by means of 3D modeling and sound source localization technologies, and the warning lag problem is solved through a real-time analysis and hierarchical warning mechanism, achieving the technical effects of realizing multi-source data fusion analysis, improving the positioning accuracy, and being able to perform leakage warning in a timely manner, thereby ensuring the safe and stable operation of the grouting project in the underground utility tunnel.

[0009] The above content outlines the present application for solving the method and system for monitoring the leakage of grouting materials in an IoT-driven underground utility tunnel. The technical solution steps of the present application will be described in detail in the following specific embodiments to facilitate a clear and complete understanding of the present application by those skilled in the art. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for description in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can obtain other drawings without creative efforts based on these drawings.

[0011] Figure 1 is a flowchart of the IoT-driven method for monitoring the leakage of grouting materials in an underground utility tunnel provided by an embodiment of the present application.

[0012] Figure 2 is a structural diagram of the IoT-driven leakage monitoring system for grouting materials in an underground utility tunnel provided by an embodiment of the present application.

[0013] Description of the reference numerals: data set acquisition module 1, vector generation module 2, coordinate positioning module 3, instruction trigger module 4. Specific implementation mode

[0014] In this application, a multi-source sensor network is deployed in the grouting section of the underground utility tunnel to collect data. After preprocessing, multi-modal feature extraction is performed to generate grouting leakage feature vectors. The vectors are input into the leakage risk model to determine the risk level and potential leakage coordinates, and accordingly, a leakage monitoring report is generated and a hierarchical early warning is triggered. If there are multiple abnormal sensor data areas, a comprehensive analysis is carried out to determine the key monitoring areas, and at the same time, the monitoring frequency is adjusted according to the risk level. Finally, the monitoring data is classified and stored to provide a basis for subsequent maintenance and management, achieving the technical effects of realizing multi-source data fusion analysis, improving the positioning accuracy, and being able to give timely leakage early warnings, thereby ensuring the safe and stable operation of the underground utility tunnel grouting project.

[0015] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0016] It should be noted that any variations of the terms "including" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product or server including a series of steps or units is not necessarily limited to those clearly listed steps or units, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0017] Embodiment 1, as Figure 1 shown, an Internet of Things-driven method for monitoring the leakage of grouting materials in an underground utility tunnel, wherein the method includes:

[0018] Step A100: Real-time collection of the underground utility tunnel is carried out through the Internet of Things sensor network to obtain a dynamic parameter data set.

[0019] In the embodiments of the present application, the Internet of Things sensor network is a multi-source sensor network deployed in the grouting section of the underground utility tunnel. The underground utility tunnel is a public tunnel space in the city for centralized laying of municipal pipelines. The dynamic parameter data set is an original dynamic parameter data set generated by the multi-source sensor network for real-time collection of the grouting area of the underground utility tunnel.

[0020] Specifically, a multi-source sensor network is deployed in the underground pipe gallery grouting section based on the Internet of Things, such as distributed optical fiber pressure sensors, acoustic emission sensors, etc. They collect data in real time in the underground pipe gallery grouting area and generate a raw dynamic parameter data set containing a variety of information. Subsequently, these raw data are pre-processed at multiple levels to generate a dynamic parameter data set for subsequent analysis, laying a solid foundation for accurate leakage monitoring. The specific steps are described in detail in A110-A120.

[0021] Step A200: performing multimodal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector.

[0022] In the embodiment of the present application, multimodal feature extraction is a process of parsing multiple data from a dynamic parameter data set, extracting corresponding features using specific methods, and normalizing them. The grouting leakage feature vector is a vector generated by normalizing the pressure fluctuation feature matrix, flow velocity chaos feature vector, and frequency band energy distribution feature obtained in the multimodal feature extraction process.

[0023] Optionally, the dynamic parameter data set is parsed to obtain the grouting pressure time series data, slurry flow rate time series data, and soundprint signal waveform data, and then the time-frequency domain joint analysis, nonlinear dynamic analysis, and multi-scale decomposition are performed to extract the pressure fluctuation feature matrix, flow rate chaos feature vector, and frequency band energy distribution characteristics. These features are then normalized to generate the grouting leakage feature vector. The specific steps are described in detail in A210-A250.

[0024] By generating this characteristic vector, the leakage status of the grouting material can be reflected more accurately, providing key data support for subsequent leakage risk assessment.

[0025] Step A300: input the grouting leakage feature vector into the leakage risk model for real-time analysis, determine the leakage risk level, and locate the potential leakage coordinates according to the leakage risk level.

[0026] In the embodiment of the present application, the leakage risk model is a model constructed by integrating a convolutional neural network and a random forest algorithm. The leakage risk level is the result obtained by inputting the grouting leakage feature vector into the leakage risk model for real-time analysis. The potential leakage coordinates are the spatial position coordinates where leakage may occur determined according to the leakage risk level.

[0027] In one embodiment of the present application, the grouting leakage feature vector is synchronized to a pre-trained leakage risk model. The specific steps of the model generation process are described in detail in A351 - A353. The leakage event probability distribution map is generated by the multi-scale spatio-temporal feature extraction layer of the model, and the dominant modal data is extracted by performing dynamic modal decomposition on it. Based on this, the historical leakage similarity calculation is carried out to determine the similarity matrix, and then the leakage risk level is determined. The specific steps are described in detail in A310 - A340. Meanwhile, the strain data set is obtained by extracting the distributed fiber optic pressure sensor array through the multi-source sensor network for strain analysis, and based on this, a three-dimensional pressure distribution model of the underground utility tunnel is constructed. Then, the acoustic emission sensor is extracted to determine the sound source location data, and finally, the coordinate calculation is carried out by combining the three-dimensional pressure distribution model and the sound source location data to determine the potential leakage coordinates. The specific steps are described in detail in A360 - A390.

[0028] By determining the leakage risk level and locating the potential leakage coordinates, the effective assessment of the leakage risk and the accurate positioning of the potential leakage points are realized, ensuring the safe and stable operation of the underground utility tunnel.

[0029] Step A400: Generate a leakage monitoring report according to the leakage risk level and the potential leakage coordinates, send the leakage monitoring report to the Internet of Things platform, and trigger a hierarchical warning instruction through the Internet of Things platform.

[0030] In the embodiment of the present application, the leakage monitoring report is a comprehensive analysis report generated according to the leakage risk level and the potential leakage coordinates. The hierarchical warning instruction is a warning command with different risk levels issued by the Internet of Things platform according to the leakage monitoring report.

[0031] Specifically, according to the leakage risk level and the potential leakage coordinates, first, the grouting risk analysis of the underground utility tunnel is carried out to generate a risk level label, then the grouting leakage analysis is carried out to obtain the three-dimensional coordinates of the leakage point. Then, the grouting prediction is carried out by combining the risk level label and the three-dimensional coordinates of the leakage point to obtain the predicted range of leakage diffusion. Finally, according to the predicted range of leakage diffusion and the environmental parameter influence factors, the risk level label and the three-dimensional coordinates of the leakage point are associated and integrated to generate a leakage monitoring report. The specific steps are described in detail in A410 - A440.

[0032] The generated leakage monitoring report is transmitted to the cloud server through the communication interface of the Internet of Things platform. The platform system first analyzes and verifies the key data such as the risk level label, the three-dimensional coordinates of the leakage point, and the predicted range of leakage diffusion in the report. After passing the verification, the platform triggers the corresponding warning instruction according to the preset hierarchical warning rules (such as low, medium, and high risk thresholds). For example, when a high-risk leakage is detected, the platform will automatically send an instant notification containing the leakage coordinates and the diffusion range to the operation and maintenance personnel, and synchronously link the emergency equipment in the utility tunnel to start the local plugging procedure.

[0033] Through the above steps, the effective control of the grouting leakage risk of the underground utility tunnel is realized.

[0034] Furthermore, step A100 in the method provided by the embodiment of the present application includes:

[0035] A110: Deploy a multi-source sensor network in the grouting section of the underground utility tunnel, and through the multi-source sensor network, perform real-time acquisition on the grouting area of the underground utility tunnel to generate an original dynamic parameter data set.

[0036] A120: Perform multi-level preprocessing on the original dynamic parameter data set to generate a dynamic parameter data set.

[0037] In the embodiment of the present application, the multi-source sensor network includes a distributed optical fiber pressure sensor array, a high-precision flow sensor, and an acoustic emission sensor network. The original dynamic parameter data set is a data set generated based on the real-time acquisition of the grouting area of the underground utility tunnel by the multi-source sensor network. The multi-level preprocessing includes noise filtering, abnormal data verification, and timestamp alignment. The dynamic parameter data set is a data set obtained after performing multi-level preprocessing on the original dynamic parameter data set.

[0038] Specifically, deploy a multi-source sensor network in the grouting section of the underground utility tunnel. Among them, the distributed optical fiber pressure sensor array can accurately collect grouting pressure data. A sensor is set every certain distance (such as 10 meters), and the pressure changes at different positions can be obtained in real time, with an accuracy of up to 0.1 MPa; the high-precision flow sensor is responsible for collecting slurry flow velocity data, and its measurement accuracy can reach ±1%, ensuring accurate flow velocity information; the acoustic emission sensor network is used to collect acoustic signal data, which can sensitively capture the minute sound changes in the utility tunnel, providing multi-dimensional data support for judging leakage. Then, these data are integrated and aggregated, that is, the data format is unified to make it consistent, and the data is arranged in chronological order, and the data collected by different sensors at the same moment is associated to form a complete data set, thus generating an original dynamic parameter data set containing grouting pressure data, slurry flow velocity data, and acoustic signal data.

[0039] Next, perform multi-level preprocessing on the original dynamic parameter data set. First, perform noise filtering to remove the noise generated by the sensor's own errors, environmental electromagnetic interference, etc., making the data more stable and reliable; then perform abnormal data verification. By setting reasonable data thresholds and logical rules, for the grouting pressure data, it can be set to 1 MPa - 8 MPa, and for the slurry flow velocity data, it can be set at 0.5 - 2 m 3 / h, the voiceprint signal data can be set between 50 - 500 Hz. Data outside this range can be determined as abnormal, and those abnormal data that clearly do not conform to the actual situation are identified and excluded. Finally, timestamp alignment is performed to ensure that the data collected by different types of sensors are synchronized in time, generating a dynamic parameter dataset for subsequent comprehensive analysis.

[0040] By obtaining the original dynamic parameter dataset and preprocessing it, a dynamic parameter dataset for subsequent analysis is generated, laying a solid data foundation for accurately monitoring the leakage of grouting materials in the underground pipe gallery.

[0041] Furthermore, step A200 in the method provided by the embodiments of the present application includes:

[0042] A210: Analyze the dynamic parameter dataset to obtain grouting pressure time - series data, slurry flow rate time - series data, and voiceprint signal waveform data.

[0043] A220: Perform a joint time - frequency domain analysis on the grouting pressure time - series data to extract a pressure fluctuation feature matrix.

[0044] A230: Perform non - linear dynamics analysis on the slurry flow rate time - series data to generate a flow rate chaos feature vector.

[0045] A240: Perform multi - scale decomposition on the voiceprint signal waveform data to extract frequency - band energy distribution features.

[0046] A250: Normalize the pressure fluctuation feature matrix, the flow rate chaos feature vector, and the frequency - band energy distribution features to generate the grouting leakage feature vector.

[0047] Optionally, first analyze the dynamic parameter dataset to separate the grouting pressure time - series data, the slurry flow rate time - series data, and the voiceprint signal waveform data from it.

[0048] Then, perform a joint time - frequency domain analysis on the grouting pressure time - series data. First, use the Fourier transform to decompose the grouting pressure signal that changes with time into a superposition of sine and cosine waves of different frequencies to obtain global frequency information. Then use the wavelet transform. By convolving the signal with a series of wavelet functions obtained by stretching and translating the mother wavelet function, the frequency changes of the signal at different time points are reflected. Combine the results of the Fourier transform and the wavelet transform, comprehensively consider the information of the pressure signal at different times and frequencies, and organize the pressure fluctuation features at different times and frequencies into a matrix form, thus generating the pressure fluctuation feature matrix.

[0049] Then, perform nonlinear dynamics analysis on the time series data of the slurry flow rate. Since the slurry leaks, the change in its flow rate often exhibits non-linear chaotic characteristics. By means of phase space reconstruction and calculating the Lyapunov exponent, first determine the two key parameters of the delay time and the embedding dimension. The delay time is generally determined by the autocorrelation function (calculate the correlation between the time series of the slurry flow rate data and its delayed sequence. As the delay time increases, the correlation will change. Generally, select the delay time corresponding to when the autocorrelation function value first drops to a certain proportion (such as 1 / e) of the initial value), and the embedding dimension is commonly calculated by the false nearest neighbor method (map the data points in the low-dimensional space to the high-dimensional space, calculate the nearest neighbor points of each data point, and judge whether these nearest neighbor points are still the nearest neighbors in the higher-dimensional space. As the embedding dimension increases, the proportion of false nearest neighbor points will change. When the proportion of false nearest neighbor points is lower than a certain threshold (such as 5%), the embedding dimension at this time is the appropriate embedding dimension). After determining the parameters, apply phase space reconstruction to the one-dimensional time series data of the slurry flow rate, expand it to the high-dimensional space, and display the non-linear characteristics contained in the data. In the phase space, select the phase point and its neighboring points, and calculate the change in the distance of the neighboring points during the evolution process. As time goes by, calculate the average exponential divergence rate of these distances, that is, the Lyapunov exponent. If the Lyapunov exponent is greater than zero, it indicates that the system has chaotic characteristics; if it is less than zero, it means that the system is stable, and finally generate the flow rate chaotic feature vector.

[0050] For the voiceprint signal waveform data, perform multi-scale decomposition. Use the empirical mode decomposition (EMD) algorithm to determine all the maximum and minimum points in the voiceprint signal, use cubic spline interpolation to fit the upper and lower envelopes respectively, calculate the mean of the upper and lower envelopes, and obtain an average envelope. Subtract the average envelope from the original voiceprint signal to get a preliminary intrinsic mode function (IMF). Then, judge whether the IMF meets the conditions of the intrinsic mode function. If it does not meet the conditions, repeat the above steps to screen it. The IMF that meets the conditions is separated, and the remaining signal continues to repeat the above decomposition process until the remaining signal is a monotonic function or a constant. After such processing, the complex voiceprint signal is decomposed into multiple intrinsic mode functions with different frequencies. By analyzing the energy distribution of these intrinsic mode functions in different frequency bands, the frequency band energy distribution characteristics are extracted.

[0051] Finally, normalize the extracted pressure fluctuation feature matrix, flow velocity chaos feature vector, and band energy distribution feature through a weighted feature fusion algorithm. First, determine the respective weights of the pressure fluctuation feature matrix, flow velocity chaos feature vector, and band energy distribution feature, which are set by those skilled in the art according to the importance of each feature in reflecting the leakage situation of grouting materials. Then, perform normalization operations on each feature separately, mapping its value to a specific interval, such as [0, 1], to eliminate the differences in dimension and numerical range. For example, use the min-max normalization formula where x is the original feature value, x max and x min are the minimum and maximum values of this feature. After that, perform weighted calculation on the normalized features according to the determined weights, that is, multiply each normalized feature by the corresponding weight and then sum them up, finally obtaining a fused and normalized grouting leakage feature vector for subsequent leakage risk analysis.

[0052] Through the above steps, comprehensive and accurate data support is provided for the leakage risk model, improving the accuracy and reliability of leakage monitoring.

[0053] Furthermore, step A300 in the method provided by the embodiments of the present application includes:

[0054] A310: Input the grouting leakage feature vector into the leakage risk model for real-time analysis to determine the leakage risk level.

[0055] A320: Synchronize the grouting leakage feature vector to the pre-trained leakage risk model, and generate a leakage event probability distribution map through the multi-scale spatio-temporal feature extraction layer of the leakage risk model.

[0056] A330: Perform dynamic mode decomposition on the leakage event probability distribution map to extract the dominant mode data.

[0057] A340: Based on the dominant mode, perform historical leakage similarity calculation to determine the similarity matrix, and set the leakage risk level according to the similarity matrix.

[0058] In the embodiments of the present application, the multi-scale spatio-temporal feature extraction layer is a key component in the leakage risk model, and this layer processes the vector. The leakage event probability distribution map is a kind of map generated by the multi-scale spatio-temporal feature extraction layer of the model. The dominant mode data is the data extracted after performing dynamic mode decomposition on the leakage event probability distribution map.

[0059] Specifically, first, input the grouting leakage feature vector into the leakage risk model for real-time analysis. This vector comprehensively integrates feature information such as grouting pressure, slurry flow velocity, and acoustic signal, and can comprehensively reflect the state of the grouting process.

[0060] Next, synchronize the grouting leakage feature vector to the pre-trained leakage risk model. The multi-scale spatio-temporal feature extraction layer of this model deeply analyzes the vector through multiple convolutional layers and pooling layers. Convolution kernels of different sizes in the convolutional layer slide over the vector. Small convolution kernels capture local detail features such as short-term minute fluctuations in grouting pressure and small-area changes in slurry flow rate, while large convolution kernels obtain macroscopic features such as long-term trend changes in grouting pressure and distribution characteristics of slurry flow rate in different areas of the pipe gallery, and transform them into feature maps. The pooling layer then screens and reduces the dimensionality of the feature maps output by the convolutional layer. For example, max pooling selects the maximum value in a small area to highlight significant features and filter out unimportant details. After alternating processing by multiple convolutional layers and pooling layers, the vector is comprehensively and deeply analyzed from different scales and spatio-temporal dimensions. Finally, the processed features are integrated to generate a leakage event probability distribution map, which shows the probability of leakage events occurring in different areas.

[0061] Then, use a cascade classifier to classify the risk levels of the leakage event probability distribution map. The cascade classifier sequentially screens and classifies the data through multiple classifiers. The first classifier will, according to preset rules and features, preliminarily screen the data in the leakage event probability distribution map, identify some obvious features and trends, and classify the data into preliminary categories. Then, subsequent classifiers will further analyze and classify the data based on the results of the previous classifier. In this step-by-step screening and classification process, various factors are comprehensively considered (factors related to the grouting leakage feature vector, factors of the leakage event probability distribution map, factors of the dominant modal data, and factors of historical leakage similarity calculation), so as to accurately determine the low-risk, medium-risk, and high-risk levels. At the same time, the leakage event probability distribution map is processed by the dynamic modal decomposition technique to extract the dominant modal data. During the decomposition process, the amplitude and frequency parameters of the dominant mode are obtained. The amplitude reflects the intensity of the leakage-related features, and the frequency parameter reflects the speed of feature change. These parameters reflect the main features of the leakage event and are crucial for judging the risk level.

[0062] Finally, perform historical leakage similarity calculation based on the dominant mode. According to the extracted dominant modal data, which includes the amplitude and frequency parameters of the dominant mode. Then, collect historical known leakage data, which cover leakage feature information at different times and under different circumstances. Then, compare the current dominant modal data with the historical leakage data point by point, and use the cosine similarity algorithm to calculate the similarity between the two.

[0063] First, convert the current dominant mode and historical leakage data into vector form, where the elements in the vector are eigenvalue such as amplitude and frequency; calculate the result of the accumulation after multiplying the corresponding elements of the two vectors; calculate the square root of the sum of the squares of the elements of each of the two vectors respectively, which is similar to measuring the "length" of the vector; divide the previously calculated accumulation result by the product of the two "lengths" to obtain the cosine similarity value; the closer the value is to 1, the more similar they are; the closer it is to -1, the greater the difference; and 0 indicates that they are independent and irrelevant. By calculating in this way for each group of data, the similarity degree is determined. Set the leakage risk level according to the set three-level risk threshold, that is, similarity < 0.3 is low risk, 0.3 ≤ similarity < 0.7 is medium risk, and similarity ≥ 0.7 is high risk.

[0064] Through the above steps, the leakage risk can be evaluated more accurately, improving the accuracy and reliability of the leakage monitoring of the grouting material in the underground pipe gallery.

[0065] Furthermore, step A350 in the method provided by the embodiment of the present application includes:

[0066] A351: Use a convolutional neural network to extract spatio-temporal features from the grouting leakage feature vector to generate a multi-scale feature map, and the multi-scale feature map is in the multi-scale spatio-temporal feature extraction layer.

[0067] A352: Perform time series modeling on the multi-scale feature map to construct a leakage probability prediction function.

[0068] A353: Combine the random forest algorithm to optimize the decision boundary of the leakage probability prediction function to generate the leakage risk model.

[0069] In the embodiment of the present application, the multi-scale feature map is a map generated by using a convolutional neural network to extract spatio-temporal features from the grouting leakage feature vector. The leakage probability prediction function is a function constructed by performing time series modeling on the multi-scale feature map. The random forest algorithm is an algorithm used to optimize the decision boundary of the leakage probability prediction function.

[0070] Specifically, first prepare convolutional kernels of different sizes. These convolutional kernels are like "feature detectors". When processing the grouting leakage feature vector, the convolutional kernels slide on the vector. When the small convolutional kernel slides, it focuses on the local area in the vector and can capture local detail features such as the minute fluctuations of the grouting pressure in a short period of time and the changes in the slurry flow rate in a small area; while when the large convolutional kernel slides, it pays attention to a more macroscopic range, such as the trend changes of the grouting pressure in a longer time period and the distribution characteristics of the slurry flow rate in different areas of the entire pipe gallery.

[0071] During the sliding process of the convolution kernel, each element in the convolution kernel is multiplied by the corresponding element of the grouting leakage feature vector area it covers, and then these products are added to obtain a convolution result value. As the convolution kernel slides point by point on the vector, a series of such convolution result values will be generated, which constitute a new feature map. Multiple convolution kernels of different sizes will generate multiple different feature maps, which reflect the characteristics of the grouting leakage feature vector from different scales. Combining these feature maps of different scales together is converted into multi-scale feature maps. These maps are in the multi-scale spatiotemporal feature extraction layer, which presents the grouting leakage characteristics from different scales, time and space dimensions.

[0072] Next, the multi-scale feature map is modeled as a time series, and the leakage probability prediction function is constructed with the help of long short-term memory network.

[0073] Step A: Arrange the multi-scale feature maps in chronological order as the input data of the long short-term memory network (LSTM). The LSTM network contains an input gate, a forget gate, and an output gate. The input gate controls the input of new information of the multi-scale feature map, the forget gate determines whether to retain or discard past information, and the output gate determines the output content, so as to process the multi-scale feature maps at different times and learn.

[0074] Step B: When processing input data, the LSTM network will learn multi-scale feature maps at different times based on its own structure. Through the gating mechanism, it can effectively capture long-term dependencies in time series without losing key information due to time span. For example, when faced with time-varying features such as grouting pressure and slurry flow rate, LSTM can remember the impact of early features on the current moment.

[0075] Step C: As time goes by, the LSTM network continuously updates its internal state, and gradually discovers the intrinsic connection between the multi-scale feature map and the leakage probability during the learning process (the changing trend and fluctuation of the characteristics such as grouting pressure, slurry flow rate, and soundprint signal in the multi-scale feature map over time, and the causal relationship and mutual influence relationship between them and the leakage probability of the underground pipeline corridor).

[0076] Step D: After multiple iterations of training, the back propagation algorithm first calculates the difference between the predicted leakage probability and the actual leakage probability to obtain the error, and then starts from the output layer, derives the activation function according to the error, and combines the network weights to gradually propagate the error forward to calculate the gradient of each layer, so as to determine how to adjust the weights and parameters to reduce the error and make the network output closer to the true value. Finally, the trained LSTM network is used as the core computing unit, and the multi-scale feature map data is input. After the internal calculation processing of the network, the predicted value of the corresponding leakage probability is output, thereby constructing a function model with multi-scale feature maps as input and leakage probability as output.

[0077] Finally, the decision boundary of the leakage probability prediction function is optimized by combining with the random forest algorithm to generate a leakage risk model.

[0078] The random forest algorithm first randomly extracts multiple sample subsets from the training dataset with replacement. For each sample subset, a decision tree is constructed. During the construction of the decision tree, for the splitting of each node, a part of the features is randomly selected from numerous features, and then the optimal splitting point is searched among these features. After numerous decision trees form a forest, during prediction, each decision tree predicts the input multi-scale feature map data. The prediction results of all decision trees are integrated. Generally, the majority voting method (for classification problems) is used. The input data is predicted by numerous decision trees respectively, and the prediction results of each decision tree are statistically analyzed. The prediction category with the most occurrences is used as the final prediction result. By continuously adjusting the parameters of the random forest, such as the number of decision trees, the number of features randomly selected for each node, etc., the prediction result is made closer to the true leakage probability situation, thereby optimizing the decision boundary of the leakage probability prediction function.

[0079] After the decision boundary of the leakage probability prediction function is optimized by the random forest algorithm, the optimized function is combined with the leakage probability prediction function previously constructed by extracting multi-scale feature maps through a convolutional neural network and using time series modeling. The grouting leakage feature vector obtained through multi-modal feature extraction is used as the input of the entire model. First, it passes through the convolutional neural network for spatio-temporal feature extraction to generate a multi-scale feature map. Then, the multi-scale feature map is input into the optimized leakage probability prediction function. According to the calculation of the function and the judgment of the decision boundary, the corresponding leakage risk level is output, thus generating a leakage risk model that can accurately evaluate the leakage risk of the grouting material in the underground utility tunnel.

[0080] Through the above steps, the generated leakage risk model can more accurately evaluate the leakage risk of the grouting material in the underground utility tunnel, effectively improving the accuracy and reliability of leakage monitoring, and providing strong support for ensuring the safe operation of the underground utility tunnel.

[0081] Furthermore, step A300 in the method provided by the embodiments of the present application includes:

[0082] A360: Extract the distributed fiber optic pressure sensor array through the multi-source sensor network, and perform strain analysis based on the distributed fiber optic pressure sensor array to obtain a strain dataset.

[0083] A370: Perform three-dimensional modeling of the underground utility tunnel according to the strain dataset to construct a three-dimensional pressure distribution model.

[0084] A380: Extract the acoustic emission sensors through the multi-source sensor network, perform sound source localization based on the acoustic emission sensors, and determine the sound source localization data.

[0085] A390: Perform coordinate calculation based on the three-dimensional pressure distribution model in combination with the sound source localization data to determine the potential leakage coordinates.

[0086] In the embodiments of the present application, the three-dimensional pressure distribution model is a model constructed by performing three-dimensional modeling on the underground pipe gallery according to the strain data set. The sound source localization data is data determined after performing sound source localization by utilizing the propagation characteristics of sound signals, such as the time difference of arrival at different sensors and other information.

[0087] Specifically, first, extract the distributed optical fiber pressure sensor array data with the help of the multi-source sensor network. Perform strain analysis on these pressure data. First, preprocess the collected pressure data to remove the noise and outliers therein. Then, according to the principles of material mechanics and the characteristics of optical fiber sensing, establish a mathematical relationship model between pressure and strain. Those skilled in the art first determine parameters such as the elastic modulus of the commonly used materials in the underground pipe gallery (such as a specific grade of concrete) based on theories such as Hooke's law in material mechanics, so as to clarify the basic ratio of stress to strain. In terms of optical fiber sensing, study the photoelastic effect of the optical fiber. For example, utilize the fiber Bragg grating sensing principle to obtain the drift data of the Bragg wavelength of a specific optical fiber under different pressures through experiments, and understand the relationship between pressure and the change in the refractive index of the optical fiber. Combining these two aspects, establish an equation relationship covering the material parameters of the pipe gallery, the optical parameters of the optical fiber, as well as pressure and strain. In addition, considering the influence of interference factors such as temperature on the optical fiber and the pipe gallery material, introduce correction coefficients through experiments or theoretical analysis, add temperature correction terms, etc., and finally form a mathematical model that can accurately describe the relationship between pressure and strain and is used to calculate strain data from pressure data.

[0088] Using this model, convert the pressure data into strain data. Considering the difference in pressure distribution at different positions of the pipe gallery, the data in different regions will be separately converted and calculated. During the calculation process, corrections will also be made in combination with factors such as the material characteristics and geometric structure of the pipe gallery to more accurately reflect the actual strain situation of the pipe gallery structure. Finally, organize and summarize these calculated and corrected strain data to form a strain data set reflecting the deformation situation of the pipe gallery structure.

[0089] Next, based on the obtained strain data set, a three-dimensional model of the underground utility tunnel is built to construct a three-dimensional pressure distribution model. The collected strain data is sorted and preprocessed (data cleaning, normalization, and format conversion). Then, using a professional three-dimensional modeling software (SolidWorks), the geometric structure of the underground utility tunnel is accurately constructed to clarify the shape, size, and relative positions of each part of the tunnel. Then, the strain data is mapped onto the three-dimensional model of the tunnel, associating the strain data at each position with the corresponding points in the model. According to the principles of material mechanics, the stress and pressure values at each point of the tunnel are calculated using the strain data. To more intuitively display the pressure distribution, visualization methods such as colors and contour lines are used to render the pressure values. Finally, the constructed model is verified and optimized. The calculation results of the model are compared with the actual monitoring data or theoretical analysis results, and the model parameters are adjusted according to the comparison results to improve the accuracy and reliability of the model, thus completing the construction of the three-dimensional pressure distribution model.

[0090] Then, acoustic emission sensor data is extracted through a multi-source sensor network. The acoustic emission sensors can capture the tiny acoustic signals generated inside the utility tunnel due to reasons such as leakage. Based on the propagation characteristics of these acoustic signals and information such as the time difference of arrival at the sensors, the sound source location is performed to determine the sound source location data.

[0091] Finally, coordinate calculations are performed by combining the three-dimensional pressure distribution model with the sound source location data. The three-dimensional pressure distribution model provides the spatial structure and pressure distribution background information of the utility tunnel, and the sound source location data determines the approximate area where leakage may occur. By integrating these two aspects of data, the coordinates of potential leakage can be accurately calculated.

[0092] Through the above steps, the accuracy and reliability of the leakage monitoring of the grouting material for the underground utility tunnel are improved, providing strong support for timely taking repair measures.

[0093] Furthermore, step A400 in the method provided by the embodiment of the present application includes:

[0094] A410: Perform a grouting risk analysis on the underground utility tunnel according to the leakage risk level and the potential leakage coordinates, and generate a risk level label.

[0095] A420: Perform a grouting leakage analysis on the underground utility tunnel according to the leakage risk level and the potential leakage coordinates, and generate the three-dimensional coordinates of the leakage point.

[0096] A430: Perform a grouting prediction on the underground utility tunnel according to the risk level label in combination with the three-dimensional coordinates of the leakage point to obtain the predicted range of leakage diffusion.

[0097] A440: Based on the predicted range of leakage diffusion according to the environmental parameter impact factors, associate and integrate the risk level label and the three-dimensional coordinates of the leakage point to generate the leakage monitoring report.

[0098] In the embodiment of the present application, the grouting risk analysis is carried out by constructing a grouting risk analysis checklist to analyze the leakage risk level and potential leakage coordinates. The grouting leakage analysis is to integrate the leakage risk level and potential leakage coordinates into the three-dimensional model of the underground utility tunnel by means of Geographic Information System (GIS) and Computer Aided Design (CAD) technologies. The grouting prediction is to simulate and predict the diffusion trend of the grouting material leakage and give the predicted range of leakage diffusion. The environmental parameter impact factors refer to the external environmental factors that affect the grouting leakage situation of the utility tunnel, including the humidity of the surrounding soil, the change of the groundwater level, the environmental temperature and air pressure, etc.

[0099] In one embodiment, when the system obtains the leakage risk level and potential leakage coordinates, the first step is to carry out the grouting risk analysis. Develop a detailed grouting risk analysis checklist, list various factors related to the leakage risk level and potential leakage coordinates, as well as the corresponding risk descriptions. When formulating the grouting risk analysis checklist, first widely collect the case materials of past underground utility tunnel grouting projects, including different types of leakage accidents and the corresponding grouting treatment situations, and at the same time refer to relevant industry standards and specifications. On this basis, sort out the factors closely related to the leakage risk level and potential leakage coordinates, such as the type of utility tunnel structure (rectangular, circular), the distribution of pipelines around the leakage point, the basis for dividing the leakage risk level, etc.

[0100] According to these factors, classify the risk levels, such as low risk, medium risk, and high risk. For each risk level, combined with the potential leakage coordinates, describe the corresponding risk situation in detail. For example, when the leakage risk level is low and the potential leakage coordinates are in the non-critical area of the utility tunnel and there are no important pipelines around, it is described as "Since the position of the leakage point is relatively independent and the surrounding environment is simple, the difficulty of grouting repair is relatively low, and it is determined to be a low risk". Arrange all the sorted out factors, risk levels and corresponding risk descriptions in a certain logical order to make a grouting risk analysis checklist, and continuously optimize and improve it according to new feedback in actual application.

[0101] Next, conduct grouting leakage analysis. With the help of the three-dimensional pressure distribution model of the underground utility tunnel constructed in step A370, integrate the leakage risk level and potential leakage coordinates. At the same time, the system synchronously collects multi-source sensor information such as strain data feedback by distributed optical fiber sensors in the utility tunnel and internal pressure data monitored by pressure sensors. Based on these data, simulate the leakage process of the grouting material starting from the potential leakage point in the three-dimensional model, and infer its possible flow direction according to the pressure difference, material properties, and the internal space structure of the utility tunnel. Use fluid mechanics algorithms to continuously correct and optimize the simulation path according to the flow law of fluids in complex spaces. Compare and verify the simulation path with actual data such as the approximate leakage area located by acoustic emission sensors. After multiple adjustments and calculations, accurately calculate the position of the leakage point in three-dimensional space.

[0102] After that, carry out grouting prediction. Collect a large number of historical cases with similar characteristics to the current underground utility tunnel, including the material of the utility tunnel, structural type, geological environment where it is located, leakage situation, etc. For example, find past cases with the same soil quality, similar utility tunnel structures, and similar leakage risk levels and leakage point positions. Analyze the leakage diffusion situation of the grouting material in these cases, such as diffusion speed, diffusion direction, and final diffusion range. According to the risk level label and three-dimensional coordinates of the leakage point of the current utility tunnel, refer to the leakage diffusion data in similar situations in historical cases, and analogously infer the predicted range of leakage diffusion of the grouting material in the current utility tunnel. For example, if in a historical case under similar conditions, the grouting material diffused 2 meters along the axial direction of the utility tunnel and 0.5 meters radially within 24 hours, this can be used as a reference for the leakage diffusion range of the current utility tunnel.

[0103] Finally, generate a leakage monitoring report. Take into account the influencing factors of environmental parameters, such as the humidity of the surrounding soil and the change of the underground water level. Integrate and correlate the risk level label, three-dimensional coordinates of the leakage point, and the predicted range of leakage diffusion, and use professional report generation software to automatically generate a leakage monitoring report with clear structure and detailed content. The report covers the basic information of the utility tunnel, the leakage risk situation, the accurate position of the leakage point, and the estimated range of future leakage diffusion, providing a comprehensive and reliable decision-making basis for the operation and maintenance management of the utility tunnel.

[0104] In summary, the method for monitoring the leakage of grouting materials in underground utility tunnels driven by the Internet of Things provided by the embodiments of this application has the following technical effects:

[0105] In this application, a data transmission link is constructed between the Internet of Things (IoT) sensor network and the leakage risk model. The data is processed using multi-modal feature extraction technology, and through operations such as parsing and analysis, the grouting leakage feature vector is obtained and analyzed in real time in the leakage risk model. Through historical leakage similarity calculation and dynamic mode decomposition, combined with mechanisms such as risk level division and coordinate positioning, the potential leakage coordinates are determined based on the three-dimensional pressure distribution model and the sound source positioning data, and a leakage monitoring report is generated, ensuring the effective monitoring of the grouting material leakage in the underground utility tunnel, achieving the technical effects of realizing multi-source data fusion analysis, improving the positioning accuracy, and being able to give leakage warnings in a timely manner, thereby ensuring the safe and stable operation of the grouting project in the underground utility tunnel.

[0106] Embodiment 2, as Figure 2 shown, based on the same inventive concept as the foregoing Embodiment 1, the embodiment of this application provides an IoT-driven underground utility tunnel grouting material leakage monitoring system, and the system includes:

[0107] A data set acquisition module 1, which is used to perform real-time acquisition of the underground utility tunnel through the IoT sensor network to obtain a dynamic parameter data set.

[0108] A vector generation module 2, which is used to perform multi-modal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector.

[0109] A coordinate positioning module 3, which is used to input the grouting leakage feature vector into the leakage risk model for real-time analysis, determine the leakage risk level, and locate the potential leakage coordinates according to the leakage risk level.

[0110] An instruction trigger module 4, which is used to generate a leakage monitoring report according to the leakage risk level and the potential leakage coordinates, send the leakage monitoring report to the IoT platform, and trigger a hierarchical warning instruction through the IoT platform.

[0111] Further, the data set acquisition module 1 is used to perform the following steps:

[0112] Deploy a multi-source sensor network in the grouting section of the underground utility tunnel based on the IoT, and perform real-time acquisition of the grouting area of the underground utility tunnel through the multi-source sensor network to generate an original dynamic parameter data set.

[0113] Perform multi-level preprocessing on the original dynamic parameter data set to generate a dynamic parameter data set.

[0114] Further, the vector generation module 2 is used to perform the following steps:

[0115] Parse the dynamic parameter data set to obtain the grouting pressure time series data, the slurry flow rate time series data, and the voiceprint signal waveform data.

[0116] Perform joint time-frequency domain analysis on the grouting pressure time series data to extract the pressure fluctuation feature matrix.

[0117] Perform non-linear dynamics analysis on the slurry flow rate time series data to generate the flow rate chaos feature vector.

[0118] Perform multi-scale decomposition on the voiceprint signal waveform data to extract the frequency band energy distribution feature.

[0119] Normalize the pressure fluctuation feature matrix, the flow rate chaos feature vector, and the frequency band energy distribution feature to generate the grouting leakage feature vector.

[0120] Furthermore, the coordinate positioning module 3 is used to perform the following steps:

[0121] Input the grouting leakage feature vector into the leakage risk model for real-time analysis to determine the leakage risk level.

[0122] Synchronize the grouting leakage feature vector to the pre-trained leakage risk model, and generate a leakage event probability distribution map through the multi-scale spatio-temporal feature extraction layer of the leakage risk model.

[0123] Perform dynamic mode decomposition on the leakage event probability distribution map to extract the dominant mode data.

[0124] Perform historical leakage similarity calculation based on the dominant mode to determine the similarity matrix, and set the leakage risk level according to the similarity matrix.

[0125] Furthermore, the coordinate positioning module 3 is used to perform the following steps:

[0126] Use a convolutional neural network to extract spatio-temporal features from the grouting leakage feature vector to generate a multi-scale feature map, and the multi-scale feature map is in the multi-scale spatio-temporal feature extraction layer.

[0127] Perform time series modeling on the multi-scale feature map to construct a leakage probability prediction function.

[0128] Optimize the decision boundary of the leakage probability prediction function by combining the random forest algorithm to generate the leakage risk model.

[0129] Furthermore, the coordinate positioning module 3 is used to perform the following steps:

[0130] Extract the distributed fiber optic pressure sensor array through the multi-source sensor network, perform strain analysis based on the distributed fiber optic pressure sensor array, and obtain a strain data set.

[0131] Perform three-dimensional modeling of the underground utility tunnel according to the strain data set, and construct a three-dimensional pressure distribution model.

[0132] Extract the acoustic emission sensor through the multi-source sensor network, perform sound source localization based on the acoustic emission sensor, and determine the sound source localization data.

[0133] Perform coordinate calculation according to the three-dimensional pressure distribution model in combination with the sound source localization data to determine the potential leakage coordinates.

[0134] Furthermore, the instruction trigger module 4 is used to execute the following steps:

[0135] Perform grouting risk analysis on the underground utility tunnel according to the leakage risk level and the potential leakage coordinates, and generate a risk level label.

[0136] Perform grouting leakage analysis on the underground utility tunnel according to the leakage risk level and the potential leakage coordinates, and generate the three-dimensional coordinates of the leakage point.

[0137] Perform grouting prediction on the underground utility tunnel according to the risk level label in combination with the three-dimensional coordinates of the leakage point, and obtain the predicted range of leakage diffusion.

[0138] Based on the predicted range of leakage diffusion according to the environmental parameter influence factor, correlate and integrate the risk level label and the three-dimensional coordinates of the leakage point to generate the leakage monitoring report.

[0139] The Internet of Things-driven underground utility tunnel grouting material leakage monitoring system provided by the embodiments of the present invention can execute the Internet of Things-driven underground utility tunnel grouting material leakage monitoring method provided by any embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the method.

[0140] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or the server. The included individual units and modules are only divided according to the functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0141] The above specific embodiments do not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of the present application shall be included within the protection scope of the present application. In some cases, the actions or steps recited in the present application can be executed in a sequence different from that in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. Internet of Things-driven method for monitoring leakage of grouting materials in underground pipe corridors, characterized in that The method includes: Performing real-time acquisition on the underground utility tunnel through the Internet of Things sensor network to obtain a dynamic parameter dataset; Performing multi-modal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector; Inputting the grouting leakage feature vector into a leakage risk model for real-time analysis to determine the leakage risk level, and locating potential leakage coordinates according to the leakage risk level; Generating a leakage monitoring report according to the leakage risk level and the potential leakage coordinates, sending the leakage monitoring report to the Internet of Things platform, and triggering a hierarchical early warning instruction through the Internet of Things platform.

2. The method for monitoring the leakage of grouting materials in an underground pipe gallery driven by the Internet of Things according to claim 1, wherein, Performing real-time acquisition on the underground utility tunnel through the Internet of Things sensor network to obtain a dynamic parameter dataset, and the method includes: Deploying a multi-source sensor network in the grouting section of the underground utility tunnel based on the Internet of Things, and performing real-time acquisition on the grouting area of the underground utility tunnel through the multi-source sensor network to generate an original dynamic parameter dataset; Performing multi-level preprocessing on the original dynamic parameter dataset to generate a dynamic parameter dataset.

3. The method for monitoring the leakage of grouting materials in an underground pipe gallery driven by the Internet of Things according to claim 1, characterized in that, Performing multi-modal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector, and the method includes: Parsing the dynamic parameter dataset to obtain grouting pressure time series data, slurry flow rate time series data, and voiceprint signal waveform data; Performing joint time-frequency domain analysis on the grouting pressure time series data to extract a pressure fluctuation feature matrix; Performing non-linear dynamics analysis on the slurry flow rate time series data to generate a flow rate chaos feature vector; Performing multi-scale decomposition on the voiceprint signal waveform data to extract frequency band energy distribution features; Normalizing the pressure fluctuation feature matrix, the flow rate chaos feature vector, and the frequency band energy distribution features to generate the grouting leakage feature vector.

4. The IoT-driven underground pipe gallery grouting material leakage monitoring method according to claim 1, characterized in that, Inputting the grouting leakage feature vector into a leakage risk model for real-time analysis to determine the leakage risk level, and the method includes: Inputting the grouting leakage feature vector into a leakage risk model for real-time analysis to determine the leakage risk level; Synchronizing the grouting leakage feature vector to a pre-trained leakage risk model, and generating a leakage event probability distribution map through the multi-scale spatio-temporal feature extraction layer of the leakage risk model; Performing dynamic mode decomposition on the leakage event probability distribution map to extract dominant mode data; Performing historical leakage similarity calculation based on the dominant mode to determine a similarity matrix, and setting the leakage risk level according to the similarity matrix.

5. The IoT-driven leakage monitoring method for grouting materials in underground pipe corridors according to claim 4, wherein The generation process of the leakage risk model, and the method includes: Using a convolutional neural network to perform spatio-temporal feature extraction on the grouting leakage feature vector to generate a multi-scale feature map, and the multi-scale feature map is in the multi-scale spatio-temporal feature extraction layer; Performing time series modeling on the multi-scale feature map to construct a leakage probability prediction function; Combining a random forest algorithm to optimize the decision boundary of the leakage probability prediction function to generate the leakage risk model.

6. The Internet of Things-driven underground pipe gallery grouting material leakage monitoring method according to claim 2, characterized in that Locating potential leakage coordinates according to the leakage risk level, and the method includes: Extracting a distributed fiber optic pressure sensor array through the multi-source sensor network, and performing strain analysis according to the distributed fiber optic pressure sensor array to obtain a strain dataset; Perform three-dimensional modeling on the underground utility tunnel according to the strain data set to construct a three-dimensional pressure distribution model; Extract acoustic emission sensors through the multi-source sensor network, perform sound source localization based on the acoustic emission sensors, and determine sound source localization data; Perform coordinate calculation according to the three-dimensional pressure distribution model in combination with the sound source localization data to determine the potential leakage coordinates; 7. The method for monitoring the leakage of grouting materials in an underground pipe gallery driven by the Internet of Things according to claim 1, wherein, Generate a leakage monitoring report according to the leakage risk level and the potential leakage coordinates. The method includes: Perform grouting risk analysis on the underground utility tunnel according to the leakage risk level and the potential leakage coordinates to generate a risk level label; Perform grouting leakage analysis on the underground utility tunnel according to the leakage risk level and the potential leakage coordinates to generate three-dimensional coordinates of the leakage point; Perform grouting prediction on the underground utility tunnel according to the risk level label in combination with the three-dimensional coordinates of the leakage point to obtain the predicted range of leakage diffusion; Based on the predicted range of leakage diffusion according to the environmental parameter influence factor, associate and integrate the risk level label and the three-dimensional coordinates of the leakage point to generate the leakage monitoring report.

8. The Internet of Things-driven leakage monitoring system for grouting materials in underground pipe corridors, characterized in that, For implementing the Internet of Things-driven underground utility tunnel grouting material leakage monitoring method according to any one of claims 1-7, the system includes: A data set acquisition module for performing real-time acquisition on the underground utility tunnel through the Internet of Things sensor network to obtain a dynamic parameter data set; A vector generation module for performing multi-modal feature extraction on the dynamic parameter data to generate a grouting leakage feature vector; A coordinate positioning module for inputting the grouting leakage feature vector into a leakage risk model for real-time analysis to determine the leakage risk level, and positioning the potential leakage coordinates according to the leakage risk level; An instruction trigger module for generating a leakage monitoring report according to the leakage risk level and the potential leakage coordinates, sending the leakage monitoring report to the Internet of Things platform, and triggering a hierarchical early warning instruction through the Internet of Things platform.

Images