A detection and modeling method for urban pipe networks

Abstract

The present invention relates to a detection and modeling method for urban pipe networks, comprising: collecting pipe network data; taking on-site photos; decomposing pictures; deep learning; data analysis; establishing an industry algorithm model; and optimizing a statistical model. The present invention utilizes vector coding technology to generate 3D from 2D of the external surface, and at the same time, it can also conduct downward detection on invisible internal structures or shelters, and perform three-dimensional reconstruction on invisible internal parts. Generate 3D images from visible 2D images, especially for generating 3D vector models with physical attributes through vector coding downward detection technology without excavation, so as to clearly understand the details of underground pipes or pipes covered on the ground, providing an accurate and meticulous scientific basis for pipe repair and condition assessment.

Description

Technical Field

[0001] The present invention relates to a detection and modeling method for urban pipe networks, which is a method for modeling and digital processing of urban above-ground and underground pipe networks. Background Art

[0002] The urban underground pipeline system is an important part of urban infrastructure, covering various pipelines such as water supply, drainage, gas, electricity, and heat. Effective pipeline detection and management can not only reduce resource waste but also improve the safety of urban operation. However, traditional pipeline detection methods such as manual detection and ground radar often face problems such as limited detection range, complex data processing, and high costs. Therefore, there is an urgent need for a new detection technology to solve these problems. Summary of the Invention

[0003] In order to overcome the problems of the prior art, the present invention proposes a detection and modeling method for urban underground pipe networks. The method uses vector coding technology to generate 3D from the external surface of 2D, and at the same time can also detect the invisible internal structure or shelter, and establish accurate pipelines without excavation, laying a solid scientific foundation for pipeline emergency repair and maintenance.

[0004] The object of the present invention is achieved as follows: A detection and modeling method for urban pipe networks, and the steps of the method are as follows:

[0005] Step 1, collect pipe network data: Collect the pipe network data of the research area, including various data on the composition, orientation, connection status, pipe material, installation year, and corrosion situation of ground and underground pipe networks;

[0006] Step 2, on-site shooting: Use a camera and positioning equipment to take on-site photos of the pipe network. For the underground grid, locate and photograph the path of the pipe network, including the positions of underground pipe network manholes, and photograph the covered and uncovered pipe network orientations on the ground;

[0007] Step 3, image decomposition to form a 3D pipe network model: First, perform image preprocessing, denoising and normalization; then perform feature extraction, including edge detection and region segmentation, reduce the image resolution, and expand the image pixels into a vector matrix; then perform data integration to synthesize a 3D pipe network model;

[0008] Step 4, deep learning: Perform deep learning on the formed 3D model through a neural network, optimize and correct, and data analysis: A vector encoder with a variable coding algorithm for converting optical quantum signals is used for cyclic update; the update equation projects the current state variable as a prior estimate forward in time to the measurement update equation, and the measurement update equation corrects the prior estimate to obtain a posterior estimate of the state;

[0009] Model update equation:

[0010]

[0011]

[0012] State update equation:

[0013] K k = P - H T (HP - H T + R) -1 (2.3)

[0014]

[0015] P k+1 = (I - K k H)P k (2.5)

[0016] Where: A is the state transition matrix, which describes the relationship of the system state changing from one time period to the next. The current state is X K , then the next state X k+1 , then through X k+1 = A·X k + B·U k to represent, where B is the control input matrix; P is the state covariance matrix, which represents the uncertainty of the system state estimation. The diagonal elements of the covariance matrix represent the variances of each state variable, and the off-diagonal elements represent the covariances between state variables. The covariance matrix will be adjusted with the introduction of new observation data during the measurement update process; X is the state vector, which contains the estimated values of all state variables of the system at a certain moment. In an underground pipe network model, the state vector includes position and trend; U: control input vector, which represents the impact of the input of the extracted optical quantum signal converted into an electrical signal on the system state. In some cases, the state change of the system not only depends on the state of its own encoder, but also is affected by the control input; H: observation matrix, which describes how to generate the observation value Z from the state vector X. Specifically, the measurement equation can be expressed as Z = H·X + V, where V is the measurement noise. The measurement matrix maps the state space to the observation space and is usually used to convert the dimension of the state vector into the dimension of the observation vector; Q: process noise covariance matrix, which represents the uncertainty noise in the system process and is usually used to describe the inaccuracy of the model. Process noise refers to the random perturbation during the state transition process, and the covariance matrix Q describes the statistical characteristics of these perturbations. It is used to update the state covariance matrix P in the prediction step;

[0017] Step 5, Data Analysis: Organize, process, and interpret pipeline data to extract useful information. The data includes: pipe diameter size, burial depth, main pipeline alignment, location positioning, pipeline physical deformation analysis, pipeline corrosion condition detection, anti-corrosion layer detection, leak point screening, pipeline material analysis, and pipeline pressure analysis;

[0018] Step 6, Establish an Industry Algorithm Model: Input industry-specific data to transform the model into a model for a specific industry;

[0019] Step 7, Optimize the Statistical Model: Repeatedly output the data collected during operation, establish a professional database, and debug the model until a satisfactory state is achieved.

[0020] The advantages and beneficial effects of the present invention are as follows: The present invention utilizes vector encoding technology to generate 3D from 2D on the external surface, and at the same time, it can also detect and explore the invisible internal structure or obstacles, and perform three-dimensional reconstruction on the invisible part inside. Generate 3D images from visible 2D images, especially for generating a 3D vector model with physical attributes through the vector encoding exploration technology without excavation, so as to clearly understand the detailed conditions of underground pipelines or pipelines covered on the ground, providing an accurate and meticulous scientific basis for pipeline emergency repair and condition assessment.

[0021] Innovative points of the present invention:

[0022] 1. By collecting and extracting pipeline industry data, developing a dynamic statistical model, combining prior knowledge and real-time geographical data, dynamically updating the non-linear characteristics of pipeline status, adapting to changes in pipeline operating conditions, and providing a more flexible and adaptable detection solution, which is an effective supplement to existing detection technologies.

[0023] 2. Use the statistical model for anomaly detection, identify abnormal patterns during pipeline operation, and predict failures based on historical data, and take maintenance measures in advance to reduce risks.

[0024] 3. Innovatively research and develop the key technology for dynamic and rapid reconstruction of multi-dimensional and multi-scene models of integrated pipe corridors

[0025] In view of the problems such as the diverse types of equipment and facilities in the utility tunnel, the complexity of modeling each element, and the difficulty of monitoring the entire life cycle, this study focuses on the multi-dimensional dynamic connection and spatio-temporal correlation relationship of the "geometry-physics-behavior-rule" of the digital model of the utility tunnel. A multi-dimensional and multi-scenario high-precision digital model of the utility tunnel that includes all elements of pipeline facilities and multiple factors of people and the environment is constructed to achieve the dynamic control and synchronous update of the urban pipe network system and the state of the physical utility tunnel. For the dynamic reconstruction and assembly integration of the multi-level and multi-scale digital model of the utility tunnel, a fusion technology of the digital model of the utility tunnel is formed by means of unified interface call, feature correlation mapping, model dynamic coupling, etc., to realize the mutual feedback and reliable mapping of multi-modal data acquisition of the utility tunnel. Finally, a modular and general-purpose component library for constructing and managing the digital model of the utility tunnel is integrated and developed. The above results have important technical innovation value. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The present invention will be further described below in conjunction with the drawings and embodiments.

[0027] Figure 1 is a flowchart of the method described in the embodiment of the present invention;

[0028] Figure 2 is a cyclic update view of the vector encoder described in the embodiment of the present invention;

[0029] Figure 3 is a deformation diagram under the vertical load of the pipeline for calculating the standard value of the vertical overburden pressure on the top of the pipe described in the embodiment of the present invention;

[0030] Figure 4 is a mechanical model diagram of the pipeline for calculating the standard value of the vertical overburden pressure on the top of the pipe described in the embodiment of the present invention; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0031] Embodiment:

[0032] This embodiment is a method for detecting and modeling urban underground pipe networks. The research object, research process, and research key points of the method described in this embodiment are as follows:

[0033] I. Key scientific problems to be solved:

[0034] There are many types and diverse materials of underground pipelines. The laying time and methods are different. Pipelines often cross and overlap vertically and horizontally, increasing the difficulty of detection. In the case of non-excavation, using image reconstruction algorithms, in the image data, by measuring the propagation time of light particle signals and light field imaging data, a three-dimensional reconstruction of the pipeline scene is realized. Then, through computer machine learning technology, the accuracy and efficiency of imaging are improved, providing theoretical evidence for non-metallic pipelines, parallel-laid pipelines, long-distance and large-buried-depth pipelines, etc., improving the judgment accuracy rate and reducing blind construction.

[0035] (1) Detection of accident signs and quantitative characterization of integrated pipeline corridors using digital modeling;

[0036] Comprehensive utility corridors have many types of pipelines and coupled and intertwined safety risks, which makes it difficult to accurately identify signs of utility corridor accidents. Integrating the multimodal real-time data of the utility corridor digital twin in the form of digital-analog linkage and studying the characteristic indicators of typical accident scenarios of utility corridor pipelines and their explicit and implicit quantitative correlations is the basis for identifying signs of utility corridor accidents and quantifying risks.

[0037] (2) Holographic simulation reduction method for real-time modeling and risk evolution warning;

[0038] The application of multi-dimensional models has problems such as high computing load and low prediction efficiency. Innovative research and development of non-visual field exploration methods for multi-scenario models of pipeline corridors, further integration of real-time monitoring data, and quantitative prediction of the evolution path and key nodes of pipeline corridor accident risks are the key to achieving self-correcting pipeline corridor accident risk evolution preview and accurate early warning.

[0039] 2. Key technical issues to be solved:

[0040] Technical issues: research in application direction.

[0041] In the detection of long-distance and deeply buried pipelines, there are few mapping relationships of light particle signals on PE material pipelines, and the feature mapping of pipelines without data accumulation is missing, which leads to detection errors. The study uses computer simulation and optical simulation algorithms to analyze the impact of different materials on light quantum reflection and optimize the generation scheme of the prediction model.

[0042] (1) High-precision model construction technology for integrated utility corridors;

[0043] There are numerous pipelines and ancillary facilities inside the tunnel, so high-precision modeling is achieved. The multi-dimensional dynamic connection and spatiotemporal correlation of the "geometry-physics-behavior-rules" of the tunnel digital model are studied, and a high-precision digital model module with correlation and coupling of multiple factors such as the tunnel itself, pipelines inside the tunnel, ancillary facilities, personnel, and environment is constructed.

[0044] (2) Multi-dimensional and multi-scene dynamic three-dimensional reconstruction technology for integrated pipeline corridors;

[0045] The structural layout, node network, and time and space scales of the integrated pipeline corridor are complex and diverse, and it is difficult to assemble and reconstruct model scenes. We develop modular and universal pipeline corridor three-dimensional and even four-dimensional digital model scene dynamic reconstruction technology and tool components to achieve multi-modal data collection and feedback of all elements of the pipeline corridor's "man-machine-environment-pipeline" and credible mapping with the actual pipeline network.

[0046] 3. Research Methods:

[0047] 1. Data collection and preprocessing:

[0048] 1.1 Ensure that the pipeline-related data collected (such as flow rate, pressure, temperature, material properties, etc.) is representative and accurate.

[0049] 1.2 Data cleaning: Remove noise and outliers, and fill in missing values to improve data quality.

[0050] 2. Feature engineering: Transform the features of the image data to improve the regularity of the linear function.

[0051] 3. Hyperparameter tuning: Use a spatial grid to tune the hyperparameters of the existing model to find the best parameter combination.

[0052] 4. Linear regression analysis: Fit the pipeline data model using the pipeline industry algorithm model and analyze the linear pattern of the existing model.

[0053] 5. Repeat steps 2 - 4 to infinitely approach the prediction expectation, that is, the real pipeline data.

[0054] 6. Data acquisition and information mining. In view of the characteristics of multi-source, heterogeneous, massive, and real-time applications of urban underground pipe networks, construct a temporal data storage mode based on the base state correction model, and establish a unified data model of "static library + industry algorithm model + dynamic library". Achieve rapid acquisition of characteristic information of urban pipe network emergencies and in-depth data mining, and provide data support at three scales: local (laboratory), regional (urban area), and national (typical area) for preventing emergency assessment decisions and emergency response.

[0055] IV. Research objectives;

[0056] In view of the problems of existing detection technologies being difficult to detect and inaccurate in detecting magnetic field interference such as electricity and metal, large burial depths, PE materials, etc., through image reconstruction algorithms and by measuring the propagation time of light particle signals and light field imaging data in image data, realize the three-dimensional reconstruction of the pipeline scene, and cooperate with existing detection means to achieve accurate positioning and identification of underground pipe networks, including information such as pipe type, size, and material, for subsequent management and maintenance.

[0057] (1) Research objectives;

[0058] Study the multi-dimensional dynamic connection and spatio-temporal correlation relationship of the "geometry - physics - behavior - rule" of the integrated pipe gallery digital model, construct a multi-dimensional and multi-scene high-precision pipe gallery digital model including all elements of pipe gallery pipeline facilities and multiple factors such as human environment, form pipe gallery digital model construction and scene dynamic reconstruction technologies, and form modular and general-purpose pipe gallery digital model construction management tool components to provide a reliable model basis for subsequent research content.

[0059] (2) Main research contents;

[0060] Based on the experience of underground utility tunnel planning and design, safety risk prevention and control, and digital modeling technology achievements, study the multi-dimensional dynamic connection and spatio-temporal correlation relationship of "geometry-physics-behavior-rules" of the utility tunnel, and develop a high-precision non-line-of-sight digital model construction technology for the utility tunnel that couples multiple factors such as the tunnel body, in-tunnel pipelines, ancillary facilities, personnel, and environment. Study the dynamic reconstruction and data acquisition and feedback technology of digital models of various utility tunnels at multiple scales and levels (such as gas, heat, tap water, etc.), and realize the all-element multi-modal data acquisition, feedback, and credible mapping of "human-machine-environment-pipeline" in the utility tunnel. Develop a management tool component library for the construction of the digital model of the utility tunnel to provide new technical support for the establishment of the digital system platform for accident risk prevention and control of the utility tunnel.

[0061] (3) Major scientific problems and key technical problems to be solved. Identification and quantitative characterization of accident signs in urban utility tunnels. Utility tunnels have characteristics such as a large number of pipeline types and intertwined safety risks, which lead to utility tunnel accidents. It is difficult to accurately identify signs. Integrating multi-modal real-time data in the form of digital-analog linkage and studying the characteristic indicators of typical accident scenarios of utility tunnel pipelines and their explicit-implicit quantitative correlation relationship are the basis for identifying accident signs and quantitatively characterizing risks in utility tunnels.

[0062] The basic idea of the detection and modeling method for urban underground pipe networks is as follows: Customize a vector encoder architecture, and through end-to-end training, map instantaneous images to depth maps / models for representation. Three transient renderers are used during training to complete the reflective indirect light transmission, which can quickly generate a large amount of training data. Although the system structure design is complex, this software system can make meaningful predictions on experimental data. Even if the scene contains retroreflective objects that violate the assumptions of the (pure diffusion) forward model, the iterated prediction model is still effective - this is the successful application of deep learning in the vector encoder software module to the measurement of invisible underground pipe networks. Another advantage is that it is recursive and parameter-free, so the average time for predicting and generating a 3D model is only 60 minutes, and this method is also the fastest available technology currently. This embodiment is based on the underlying support graphics engine of the application software and the model platform to solve the core technology of detecting urban underground pipe networks under non-excavation conditions - enhancing the ability to construct graphic elements and realizing the deep integration of three types of algorithms: surface, mesh, and solid. The vector encoder incorporates capabilities such as parametric modeling, interactive modeling, and genetic algorithm parent-child set modeling, provides good human-computer interaction functions such as multi-screen multi-channel linkage, floating, and label marking menus, supports the UV mapping method to achieve more precise effect presentation, and fundamentally ensures the information security of digital applications. Support cross-industry applications and comprehensively contribute to the high-quality construction of the digital industry.

[0063] Digital construction solutions, digital collaborative management platform.

[0064] The digital collaborative management platform, based on a cloud architecture with multi-level application scenarios, provides a comprehensive solution for project management from construction to operation, covering the entire process of application from project planning, execution, monitoring, inspection to asset operation. Taking business collaboration, element collaboration, and data collaboration as the bus, it closely connects links such as project initiation, project management, quality inspection and assessment, and completion and delivery, realizes the interaction and analysis of multi-dimensional data, and provides strong support for digital design, intelligent construction, digital delivery, and asset management.

[0065] Overview of Vector Encoder Technology:

[0066] Briefly speaking, when light irradiates an object, it will cause the object to emit electrons. At the same time, light is not a continuous wave but a stream of moving particles, so light has wave-particle duality. The vector encoder precisely utilizes this characteristic of light having wave-particle duality. When the photoelectric effect occurs in the object to be measured, the transition electron data generated is combined with the algorithm model of the pipeline industry for deep learning, and after N iterations, a statistical analysis data model that meets industry requirements is obtained. The vector encoder is an efficient multi-modal analysis technology that can be used for urban pipe network signal processing and data analysis. Its core is to decompose, couple, and reconstruct the modes of optoelectronic signals to extract useful factors. In pipeline downhole detection, the vector encoder technology can calculate and analyze the field energy signals of underground pipelines, obtain the spectral frequency and vector intensity, and based on parameter information such as kinetic energy, potential energy, and force, identify and calculate the chemical and physical states of the pipeline to be measured, such as the position, material density, pressure, depth (elevation), corrosion, stress, etc.

[0067] Principle of Vector Encoder:

[0068] The basic principle of the vector encoder is to utilize the collection of invisible light signals and field energy density information at different positions around the underground pipeline, and process the signals through a modal analysis algorithm. The vector encoder can effectively distinguish different types of signals, thereby improving the detection dimension and accuracy.

[0069] At the same time, 3D vision algorithms and AI machine learning algorithms quickly construct a three-dimensional space, supporting the export of rich data formats such as VR links, mesh models, and engineering files in CAD and BIM. At the same time, it can achieve the global coupling of underground - ground - outdoor - indoor, and construct a global visual 3D real scene space. This embodiment can also be compatible with third-party laser scanners, oblique photography, panoramic photos, GIS maps, 3Dmax model files, etc., to achieve data sharing and truly realize the use of visible and invisible (underground, concealed spaces, etc.) real scene three-dimensional models from acquisition to application.

[0070] 1. Use the virtual wall model built in the main program to establish a data analysis library, conduct data mining and algorithm model training on historical data information such as steel pipes and PE pipes, and establish map management specifically for urban pipelines. In the future, predict the potential location and trend of new data pipelines and evaluate the health of pipelines.

[0071] 2. At the same time, combined with the characteristics of the virtual wall that can give multiple data model results each time, fusion algorithm analysis is carried out simultaneously, which can also improve the accuracy of underground detection.

[0072] 3. The main component of the experimental results on metal pipeline corrosion is iron oxide, and the main causes are two types of corrosion:

[0073] (1) Electrochemical corrosion: The current state (trend + position) of the pipeline surface can be reversed by the different degrees of electrochemical corrosion.

[0074] (2) Stress corrosion: Stress corrosion accelerates with the increase of longitudinal strain and can reverse the state of welds (interfaces) and valves.

[0075] 4. Determine the location and trend of underground pipe network by the density of light particles:

[0076] Use vector encoders to monitor the location and health of underground gas, heat and other pipelines:

[0077] (1) Connecting pipe network sensors: Sensors can be used to monitor parameters such as temperature, pressure, and strain in underground pipes. Vector encoders can detect small changes in pipes by analyzing the propagation characteristics of light and the above parameters, thereby providing early warning of corrosion and potential leaks or failures.

[0078] (2) Measuring photon density: Photon density is related to the intensity and propagation characteristics of light. By adjusting the intensity and wavelength of the light source, the vector encoder can optimize the sensitivity of the sensor, thereby improving the accuracy of 3D modeling and the reliability of pipeline health assessment.

[0079] 5. Multi-physics coupling matrix:

[0080] Multi-scale: The urban underground pipeline network is a combination of macroscopic, mesoscopic and microscopic phenomena.

[0081] Multi-physical and chemical mechanisms: coupling of mathematical algorithm models of multiple physical and chemical processes.

[0082] From nonlinear discretization to linear regression: from complex underground pipe networks to visualized statistical data models.

[0083] High precision: From two-dimensional to three-dimensional and then to four-dimensional, higher space-time dimensions and model accuracy.

[0084] Application Advantages of Vector Encoder:

[0085] High efficiency: The vector encoder can process a large amount of data quickly, significantly improving the detection efficiency.

[0086] Accuracy: Through modal analysis, it can effectively reduce noise interference and improve the accuracy of detection results.

[0087] Cost-effectiveness: Compared with traditional methods, the vector encoder has obvious advantages in terms of equipment and labor costs.

[0088] 6. Vector Encoding Method

[0089] 1) Data acquisition

[0090] In urban pipeline detection, it is first necessary to conduct a field survey on the surface of the target area to collect invisible light signals and field energy data information around the underground pipelines. Screen out the signal information suitable for the underground environment to ensure the accuracy of data acquisition.

[0091] 2) Data processing

[0092] The collected signal data will be processed through vector encoding technology. The specific steps include ---

[0093] Signal preprocessing: Denoise and normalize the original signal.

[0094] Modal analysis: Use the vector encoding algorithm to decompose the signal into modes and extract the features related to the pipeline.

[0095] Position recognition: According to the extracted features, use machine learning algorithms to identify the position, depth, and trend of the pipeline.

[0096] Pipeline health status: Establish a pipeline health assessment model and conduct a comprehensive assessment by combining detection data and environmental factors.

[0097] The urban pipeline downhole measurement method based on the vector encoder demonstrates its advantages in improving detection efficiency and accuracy. Future research can further explore the application potential of vector encoders in detection in different industries (military, intelligence, aerospace) to promote visualization and intelligent management.

[0098] The algorithm fusion of the vector encoder performs deep learning on the data model and iteratively fits the health status of the pipe network. The detection of the health status of urban underground pipe networks is an important link to ensure the safe and efficient operation of urban infrastructure. With the acceleration of urbanization, the underground pipeline system faces problems such as aging, corrosion, and settlement. Therefore, it is particularly important to detect and evaluate the health status of pipelines in a timely and accurate manner. The following are some key points and method principles for the detection of the health status of urban underground pipelines.

[0099] (1) Importance of Detection:

[0100] Safety: Timely detection of pipeline damage or leakage can prevent accidents and ensure public safety.

[0101] Economy: Regular detection can identify problems in advance, reducing maintenance costs and economic losses caused by emergencies.

[0102] Environmental protection: Prevent pipeline leakage from polluting soil and water sources and protect the ecological environment.

[0103] (2) Detection methods:

[0104] Detecting the position and status of underground pipelines by using the distribution of invisible light can provide high-resolution images, 2D drawings and 3D models.

[0105] The specific steps of the detection and modeling method for urban underground pipe networks are as follows, and the process is as Figure 1 shown:

[0106] Step 1, collect pipe network data: Collect the pipe network data of the area under study, including the composition, orientation, connection status, pipe material, installation date, corrosion situation of various surface and underground pipe networks;

[0107] The area under study refers to the area where pipe network renovation or repair is planned, such as a building or structure of one or several houses, or a compound of a certain unit, or a residential community, etc. The pipe network data includes: the drawings and written materials used during the construction of the pipe network facilities and corresponding buildings or structures under study, including the pipe diameter, burial depth, pipeline trend, position location, etc. The literature drawing materials should also cover the changes that have occurred during the use of the pipe network itself or the surrounding environment. These changes include: corrosion of the pipe network itself or damage caused by external forces, whether other pipe networks have been built around the pipe network, whether the underground or surface part of the pipe network has been covered or occupied by other foreign objects (for example, a house has been built on the path where the underground pipeline is buried, or the riser has been covered by decorative boards or mortar during decoration and is thus concealed), resulting in various obstacle factors such as difficulties in maintenance or excavation.

[0108] Step 2, on-site photography: Use a camera and positioning equipment to take on-site photos of the pipe network. For the underground grid, locate and photograph the path of the pipe network, including the positions of underground pipe network manholes. For the above-ground pipe network, photograph the covered and uncovered pipe network orientations;

[0109] The pipe network is usually invisible underground, but the ground conditions corresponding to the pipe network orientation can be photographed, such as the overlying objects on the ground above the pipe network and the manholes that appear on the ground. Some of the above-ground pipes are visible, while most of those installed in buildings are not. The photos are mainly taken according to the pipe network orientation.

[0110] Step 3, Image Decomposition to Form a 3D Pipeline Network Model: First, perform image preprocessing, including noise removal and normalization. Then, perform feature extraction, including edge detection and region segmentation, reduce the image resolution, and expand the image pixels into a vector matrix. For example, reduce a photo to a resolution of 28×28 pixels, with 784 pixel points, and expand it into a vector matrix of size 1×784. Then, perform data integration to synthesize the 3D pipeline network model;

[0111] 2D images usually contain complex visual information. Decomposition can simplify this information into more manageable parts. Through decomposition, image data can be processed more efficiently, the computational burden can be reduced, and the generation speed can be accelerated. Decomposition helps to better understand the image content, thus generating a more realistic 3D model. The decomposed information can be used as training data to help the data statistical model learn how to generate 3D structures from 2D images. The purpose of decomposing the image: By decomposing the 2D image, extract important information in the image, such as the shape, grayscale, texture, color, and spatial relationship of the object, filter out visible light, and retain invisible light after inversion, etc.

[0112] The decomposed information can be used to construct a 3D primary model to help the data model understand the three-dimensional structure and features of the object. Image decomposition can improve the accuracy of generating a 3D model, making the final result more consistent with real-world objects. The data after image decomposition can be used in the algorithm model to regularize the data statistical model.

[0113] Through photo decomposition, the following effects can be achieved ultimately: The 3D model generated from the 2D image has a high degree of restoration and detailed texture, and can accurately reflect the shape and features of the object. The generated 3D model can be widely applied in the field of urban underground pipeline networks to enhance the user's health detection of pipelines. The decomposed data can be used for further analysis and processing, such as object recognition, scene understanding, physical mechanics analysis, electrochemical corrosion, and stress corrosion. Use the generated 3D model for predictive assessment to transform the ex post emergency repair work of the underground pipeline network into ex ante prevention.

[0114] Photo decomposition is an important step in realizing the three-dimensional reconstruction of the underground pipeline network. By extracting and processing image information, high-quality three-dimensional models can be generated, promoting the development of various application scenarios including urban underground pipeline networks.

[0115] Step 4, Deep Learning: Perform deep learning on the formed 3D model through a neural network for optimization, correction, and data analysis: Use a vector encoder with an algorithm for converting optical quantum signals into variable codes for cyclic update; The update equation projects the current state variable as a prior estimate forward in time to the measurement update equation, and the measurement update equation corrects the prior estimate to obtain a posterior estimate of the state;

[0116] Model Update Equation:

[0117]

[0118] State update equation:

[0119] K k = P - H T (HP - H T + R) -1 (2.3)

[0120]

[0121] p k+1 = (I - K k H)P k (2.5)

[0122] Where: A is the state transition matrix, which describes the relationship between the system states from one time period to the next. If the current state is X K , then the next state X k+1 , then it can be obtained through X k+1 = A·X k + B·U k to represent, where B is the control input matrix; P is the state covariance matrix, which represents the uncertainty of the system state estimation. The diagonal elements of the covariance matrix represent the variances of the respective state variables, and the off-diagonal elements represent the covariances between the state variables. The covariance matrix is adjusted with the introduction of new observation data during the measurement update process; X is the state vector, which contains the estimated values of all state variables of the system at a certain moment. In an underground pipeline network model, the state vector includes position and trend; U: control input vector, which represents the influence of the input of the extracted optical quantum signal converted into an electrical signal on the system state. In some cases, the state change of the system depends not only on the state of its own encoder but also on the control input; H: observation matrix, which describes how to generate the observation value Z from the state vector X. Specifically, the measurement equation can be expressed as Z = H·X + V, where V is the measurement noise. The measurement matrix maps the state space to the observation space and is usually used to convert the dimension of the state vector to the dimension of the observation vector; Q: process noise covariance matrix, which represents the uncertainty noise in the system process and is usually used to describe the inaccuracy of the model. Process noise refers to the random perturbation during the state transition process, and the covariance matrix Q describes the statistical characteristics of these perturbations. It is used to update the state covariance matrix P in the prediction step.

[0123] What the state update equation does first is to calculate the gain K k , and secondly is to measure the output to obtain Z k, then generate the posterior estimate of the state according to Equation 2.4, and finally estimate the posterior covariance of the state according to Equation 2.5.

[0124] Calculating the model update equation requires repeating the entire process based on the state update equation. The posterior estimate obtained from the previous calculation is used as the prior estimate for the next calculation. This recursive calculation is one of the characteristics of measuring pipeline probing, and it is easier to implement than other algorithms. Because the current state estimate is recursively calculated only based on the previous measurement variables each time. Figure 2 Shows the entire operation process.

[0125] Regardless of whether there is a reasonable standard model to select the final model, better performance parameters can be obtained by continuously fitting through a statistical mathematical model and adjusting the coefficients of the self-developed encoder. When adjusting to track the position of the pipeline network in a three-dimensional environment, the degree of model overfitting will gradually decrease, and finally the optimal solution is selected according to the user's usage intention and the uncertain multiple solutions of the model. The distribution of discrete random variables (the density of continuous random variables) is asymptotically linearly normal after being transformed by a nonlinear system. The vector encoder is a special state estimator that achieves an asymptotically optimal Bayesian decision through nonlinear to linear transformation. This algorithm variant of the vector encoder enables the random variables after nonlinear transformation to still have the characteristics of a normal distribution.

[0126] Calculation of the vector encoder:

[0127] To estimate a process with nonlinear difference and measurement relationships, a new linearized representation is first given:

[0128]

[0129] Where: X k and Z k are the true values of the state vector and the observation vector;

[0130] and are the approximate values of the state vector and the observation vector;

[0131] is the posterior estimate of the k state vector;

[0132] The random variable w k and V k represent the process excitation noise and the observation noise;

[0133] A is the partial derivative matrix of f with respect to x:

[0134]

[0135] W is the partial derivative matrix of f with respect to w:

[0136]

[0137] H is the partial derivative matrix of h with respect to x:

[0138]

[0139] V is the partial derivative matrix of h with respect to v:

[0140]

[0141] For convenience, the subscript k is not added to A, W, H, V in the expressions, but they actually vary with time. Define a new expression for the prediction error:

[0142]

[0143] and the residue of the observed variable:

[0144]

[0145] In practice, X in Equation 2.7 cannot be directly obtained k , which is the true value of the state vector, that is, the object to be estimated. Similarly, Z in Equation 2.8 cannot be obtained k , which is the true value of the observation vector used to estimate X k . The expressions for the error process can be written from Equations 2.7 and 2.8:

[0146]

[0147] E k and η k represent independent random variables with zero mean and covariance matrices WQW T and V RV T . Equations 2.9 and 2.10 are linear, which are the state difference equation and measurement equation in the encoder. Using the true value of the observed residue in Equation 2.8 and the encoder to estimate the prediction error in Equation 2.9, the estimated result is denoted as Combined with Equation 2.7, the posterior state estimate of the initial nonlinear process can be obtained:

[0148]

[0149] The random variables in Equations 2.9 and 2.10 have the following probability distributions:

[0150]

[0151] P(E k ) ~ N(0, WQk W T )

[0152] P(ηk)~N(0,VR k V T )

[0153] Let the estimated value of be zero. From the above approximation, the encoder expression for estimating can be written as:

[0154]

[0155] Substituting Equation (2.8) back into Equation (2.11), the actual true encoder expression is:

[0156]

[0157] Therefore, the encoder equation and parameter expressions are:

[0158] Model update equation:

[0159]

[0160] P - = P k k -1 + Q (2.14)

[0161] Measurement update equation:

[0162]

[0163] P k+1 = (I - K k )P - k(2.16)

[0164] The above is the update of the observation variable of the vector encoder. By appropriately substituting the observation error covariance, the encoder gain K k。 can be obtained. After matrix transposition, the useful part of the observed quantum optical information can be correctly transmitted or weighted for downlink. When all the expressions of the encoder are given, use to replace to express the prior probability. At the same time, the matrices A, W, H, and V will make the encoder converge finally in any case. This indicates that they have changing values at different positions and time nodes and need to be recalculated each time. This ensures the effectiveness of the final number of iterations and the best authenticity of the fitting model.

[0165] Obtain the three-dimensional model of the pipeline through the physical field energy density ranging technology, and analyze the geometric shape and deformation of the pipeline. Figure 3 、Figure 4 The calculation method of the standard value of the vertical earth pressure on the pipe top is given. Among them, Figure 3 is the deformation diagram under the vertical load of the pipeline, Figure 4 is the mechanical model diagram of the pipeline:

[0166] q ck = y s H (2.17)

[0167] Among them: q ck is the standard value of the vertical earth pressure on the pipe top (KN / m 2 ); r s is the gravity density of the backfill soil (KN / m 2 ); H is the depth from the ground to the pipe top.

[0168] Under this experimental condition, the influence of vehicle load is not considered, and only the deformation calculation is considered according to the stacking load. The compaction degree of the backfill soil is the leading factor related to the safe operation of buried pipelines. Therefore, in the selection of pipeline series, pipelines with a small diameter-thickness ratio should be used to enhance the anti-vertical deformation ability and annular stability resistance of the pipelines. At the same time, the buried pipelines should select an appropriate soil covering depth (burial depth). When the local burial depth is too large or too small, the modeling and checking calculations of the circumferential bending stress, vertical deformation and position stability should be carried out in these three aspects. If necessary, the pipeline diameter-thickness ratio can also be locally increased to ensure the safe operation of the underground pipe network. Therefore, through spectral analysis and the mirror image change of kinetic and potential energy, a mathematical model can be established in real time and parameters such as pipe network pressure, temperature and humidity, electrochemical corrosion, and stress corrosion can be monitored. The statistical model and data analysis technology are used to evaluate the position trend state of the pipe network and the health status of the pipelines.

[0169] Deep learning is a branch of machine learning, mainly through neural network models to learn and process data. The main body of deep learning is the urban underground pipe network system, especially the model based on deep neural network. These models can be pre-trained or trained immediately according to specific tasks. Deep learning relies on the computing power of the computer (CPU), especially the graphics processing unit (GPU) and tensor processing unit (TPU), to process a large amount of data collected by on-site surveys and complex calculations. In a statistical sense, data sets such as photos and videos can also be regarded as part of "learning" because they provide the input information required for model training.

[0170] The content of deep learning includes: Feature extraction: Deep learning models automatically extract features from image and video data through multi-layer neural networks. These features can be edges and textures in images, or the content of each frame in videos, etc. Pattern recognition: The model learns how to recognize patterns in data, such as specific patterns in tasks like image classification, dimensional tolerances, shape tolerances, and position tolerances. Mapping relationship: Deep learning models learn the mapping relationship between input data and output results. For example, in an image classification task, the model learns to map an image to the corresponding class label. Optimization strategy: Through the backpropagation algorithm, the algorithm model learns how to adjust its parameters to minimize the error between the predicted result and the true result. Regarding the digital model iteration optimization and error loss reduction of deep neural networks.

[0171] Neural network is one of the important algorithms in deep inversion and has many applications in graphic classification, detection, and secondary modeling. It will be explained from two aspects: optimization and error in urban underground pipeline network detection.

[0172] The learning objectives achieved:

[0173] Automated decision-making: The deep learning image factor feature set can automatically make primary predictions and decisions of the model based on input data, and it is widely used in the field of underground pipeline network diagnosis.

[0174] Improve accuracy: By learning pipeline theory specifications and combining with actual work surveying and mapping, deep learning models can achieve high accuracy in specific tasks, exceeding traditional machine learning methods in terms of implementation means and effects.

[0175] Process complex data: Deep learning can process high-dimensional and complex data of known pipeline models, such as the trend depth of pipe jacking and the damage of PE pipes, and output useful information from it to form a decision tree.

[0176] Generate new content: Some deep learning models (such as generative adversarial networks for underground pipeline networks) can simulate and analyze data content such as electrostatic corrosion and safety distances between different types of pipelines to promote the improvement of construction technology and digital early warning management during the laying of underground pipeline networks.

[0177] Enhance human-computer interaction: The application of deep learning in urban underground pipeline networks in the field of computer vision has improved the naturalness and intelligence level of human-computer interaction.

[0178] Deep learning is a process in which an artificial intelligence system learns by processing a large amount of urban underground pipeline network data. The content of learning includes feature extraction, pattern recognition and mapping relationship, constructing models of the pipeline network's location trend and health status, etc. The ultimate goal is to achieve various applications such as automated decision-making of the prediction model, improving accuracy, early warning analysis of pipeline network health detection, and processing difficult pipeline data.

[0179] Step 5, Data Analysis: Organize, process, and interpret pipeline data to extract useful information. The data includes: pipe diameter size, burial depth, main pipeline orientation, location positioning, pipeline physical deformation analysis, pipeline corrosion detection, anti-corrosion layer detection, leak point screening, pipeline material analysis, and pipeline pressure analysis;

[0180] Provide digital drive for urban underground pipe networks through data analysis, help management make informed decisions, reduce the operation risks of urban pipe networks. At the same time, use industry algorithm models to continuously identify potential pipeline health risks (such as leak points), help predict the future operation status of pipe networks and maintenance arrangements. Identify bottlenecks and improvement points by analyzing the current work processes and pipeline operation data to improve efficiency and reduce costs. Identify risks and problems in potential pipe networks through data analysis, and develop corresponding strategy models through artificial intelligence and data analysis to reduce the operation risks of pipeline companies, changing from emergency repair after the event to preventive measures in advance.

[0181] Data analysis includes:

[0182] Structured data:

[0183] Tabular data: Databases of urban pipe network management departments such as pipe model tables, operation spreadsheets, etc., which contain row and column formats and are convenient for establishing statistical data models.

[0184] Time series data: Pipeline data arranged in chronological order, establish a trend analysis algorithm model and a prediction decision model with the time T axis as the inspection point.

[0185] Image and video data: Such as product pictures, surveillance videos, etc., which require computer vision technology to establish BIM for analysis to form a prediction model.

[0186] Semi-structured data:

[0187] Formats such as JSON and XML: These data formats contain tags and structures and do not fully conform to the traditional tabular format, which is suitable for data exchange and storage when establishing statistical models.

[0188] User behavior data:

[0189] Application usage data: Such as operation records in use by pipe network companies, which are used to immediately analyze the existing data statistical model and judge the usage of the improved algorithm model.

[0190] Data fusion: Perform data model fusion on the data obtained by the vector encoder detection method to provide a more comprehensive pipeline health assessment.

[0191] Machine learning: Use machine learning algorithms to analyze historical data, couple with algorithm models adapted to the pipeline industry to predict the health status and potential risks of pipelines.

[0192] Health assessment model: Establish a pipeline health assessment model, combine detection data and environmental factors for comprehensive assessment.

[0193] Data analysis and assessment

[0194] Data fusion: Perform data model fusion on the data obtained by the vector encoder detection method to provide a more comprehensive pipeline health assessment.

[0195] Machine learning: Use machine learning algorithms to analyze historical data, couple algorithm models adapted to the pipeline industry to predict the health status and potential risks of pipelines.

[0196] Health assessment model: Establish a pipeline health assessment model, combine detection data and environmental factors for comprehensive assessment.

[0197] To ensure the safe operation of various buried pipelines in cities and towns, it is necessary to focus on evaluating their circumferential bending stress, deformation, and corrosion stability. First, give the mathematical and geometric methods of the force condition and load of the 3D model of the buried pipeline, and then analyze the health assessment of the stress, deformation, and corrosion stability of the buried pipeline. Considering various factors such as different pipe material series, burial depth, and compactness of backfill soil for comparative analysis, the many factors affecting the safe operation of buried pipelines can provide technical support for the safe operation of various buried pipe networks.

[0198] The position measurement and health condition detection of urban underground pipelines are complex and important scientific research tasks. Giving full play to the downward exploration function advantages of vector encoder (artificial intelligence and data engineering) technology can effectively improve the accuracy and efficiency of detection. In the future, with the continuous progress of technology, intelligent and automated detection means will play an increasingly important role in urban pipeline management.

[0199] Step 6, establish an industry algorithm model: Input industry-specific data and transform the model into a model for a specific industry.

[0200] For example: If the research object is a gas pipeline, input the professional data of the gas pipeline into the model, thus becoming a professional model for gas pipelines.

[0201] Step 7, optimize the statistical model: Repeatedly output the data collected during operation, establish a professional database, and debug the model until a satisfactory state is reached.

[0202] Under normal circumstances, it is necessary to repeatedly verify the model and require a large amount of data for a large number of trainings and learning by the neural network. The learning process is a process of data processing and large-scale accumulation.

[0203] Finally, it should be noted that the above is only used to illustrate the technical solution of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred arrangement, those of ordinary skill in the art should understand that the technical solution of the present invention (such as changes or refinements to the digital arithmetic system used, the application of various formulas, the sequence of steps, etc.) can be modified or equivalently replaced without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A detection and modeling method for urban pipe networks, characterized in that The steps of the method are as follows: Step 1, collect pipeline network data: Collect the pipeline network data of the area under study, including the composition, orientation, connection status, pipeline material, installation year, and corrosion conditions of various surface and underground pipeline networks; Step 2, on-site photography: Use a camera and positioning equipment to take on-site photos of the pipeline network. For underground pipeline networks, locate and photograph the paths passed by the pipeline network, including the positions of underground pipeline network manholes, and photograph the covered and uncovered pipeline network orientations on the ground; Step 3, image decomposition to form a 3D pipeline network model: First, perform image preprocessing, including noise removal and normalization; then perform feature extraction, including edge detection and region segmentation, reduce the image resolution, and expand the image pixels into a vector matrix; then perform data integration to synthesize the 3D pipeline network model; Step 4, deep learning: Perform deep learning on the formed 3D model through a neural network, optimize and correct, and data analysis: A vector encoder with a light quantum signal conversion into variable coding algorithm performs cyclic updates; the update equation projects the current state variable as a prior estimate forward in time to the measurement update equation, and the measurement update equation corrects the prior estimate to obtain a posterior estimate of the state; Model update equation: State update equation: K k = P - H T (HP - H T + R) -1 (2.3) p k+1 = (I - K k H)P k (2.5) Where: A is the state transition matrix, which describes the relationship between the system states from one time period to the next; the current state is X K , then the next state X k+1 , then through X k+1 = A·X k + B·U k is represented, where B is the control input matrix; P is the state covariance matrix, which represents the uncertainty of the system state estimation; the diagonal elements of the covariance matrix represent the variances of the respective state variables, and the off-diagonal elements represent the covariances between the state variables; during the measurement update process, the covariance matrix is adjusted with the introduction of new observation data; X is the state vector, which contains the estimated values of all state variables of the system at a certain moment; in an underground pipe network model, the state vector includes position and trend; U: the control input vector, which represents the influence of the input of the extracted optical quantum signal converted into an electrical signal on the system state; in some cases, the state change of the system depends not only on the state of its own encoder but also on the control input; H: the observation matrix, which describes how to generate the observation value Z from the state vector X; specifically, the measurement equation can be expressed as Z = H·X + V, where V is the measurement noise; the measurement matrix maps the state space to the observation space and is usually used to convert the dimension of the state vector into the dimension of the observation vector; Q: the process noise covariance matrix, which represents the uncertainty noise in the system process and is usually used to describe the inaccuracy of the model; the process noise refers to the random perturbation during the state transition, and the covariance matrix Q describes the statistical characteristics of these perturbations; it is used to update the state covariance matrix P in the prediction step; Step 5, data analysis: Organize, process, and interpret the pipeline data to extract useful information; the data includes: pipe diameter size, burial depth, main pipeline orientation, position location, pipeline physical deformation analysis, pipeline corrosion condition detection, anti-corrosion layer detection, leak point screening, pipeline material analysis, pipeline pressure analysis; Step 6, establish an industry algorithm model: Input industry professional data to transform the model into a model for a specific industry; Step 7, optimize the statistical model: Repeatedly output the data collected during operation, establish a professional database, and debug the model until a satisfactory state is reached.

Images