A device and method for visualizing the mechanical state of a drainage pipe network under the influence of simulated rainfall

Abstract

The application discloses a kind of visual test device and method of drainage pipe network mechanical state under the influence of simulated rainfall, it is related to drainage pipe network physical model test technical field, the device includes test main body model box, model box bottom settlement simulation mechanism, rainfall simulation module, optical fiber monitoring unit, electromagnetic flowmeter, pressure type water level sensor and data acquisition and processing terminal.Set up the settlement actuating plate of independent lifting at bottom, actively simulate local void and uneven settlement working condition of subgrade.Set up OFDR high-precision distributed optical fiber to realize continuous collection of pipeline full-field strain, data terminal can complete multi-source data space-time registration, noise reduction processing, disease grading early warning, sensor self-calibration and rainfall seepage mechanics response prediction, and can realize closed-loop control under emergency state rainfall.The application can truly reproduce rainfall-seepage-settlement-pipeline stress multi-field coupling environment, with the advantages of high degree of visualization, high monitoring precision, high intelligent level.

Description

Technical Field

[0001] This invention relates to the field of physical model testing technology for drainage pipe networks, and in particular to a visualization test device and method for simulating the mechanical state of drainage pipe networks under the influence of rainfall. Background Technology

[0002] Urban drainage networks are core underground infrastructure for maintaining urban water circulation, flood control, drainage, and the stability of the municipal environment. They are widely buried in roadbeds and underground soil layers. Against the backdrop of frequent extreme rainfall events globally, heavy rainfall and continuous rainy weather significantly alter the balance of groundwater and soil. Rainwater continuously infiltrates through surface fissures and roadbed soil layers, constantly disturbing the original soil structure and seepage field distribution. Long-term seepage can easily cause softening of the roadbed soil, increased pore water pressure, and consequently, problems such as decreased soil compaction, localized voids, foundation delamination, and uneven settlement.

[0003] Underground drainage pipes are buried in a complex water-soil coupling environment for a long time. The combined effects of external loads such as roadbed settlement, soil compression, and seepage erosion cause continuous stress deformation in the pipes, gradually leading to problems such as pipe wall cracking, joint misalignment, and structural leakage. Once the drainage network suffers structural damage, it will not only significantly reduce the efficiency of rainwater collection and transportation, causing urban flooding and road waterlogging, but in severe cases, it can also induce road collapse, underground pipeline damage, and other safety accidents, seriously affecting urban traffic, resident safety, and municipal operation and maintenance.

[0004] Currently, the industry relies heavily on indoor model tests to study the disaster mechanism of drainage pipe networks, thereby guiding engineering design and pipe network operation and maintenance. However, existing traditional test devices have many technical defects, such as an enclosed overall structure and a lack of visualization design. They can only rely on post-testing to obtain test results and cannot observe the seepage direction inside the soil, pore water transport, and real-time deformation of the pipes in real time, making it difficult to intuitively reproduce the actual water and soil interaction process.

[0005] At the monitoring technology level, traditional tests mostly use single-point sensors for data acquisition, which can only obtain basic parameters such as local water level, pressure, and flow rate in a scattered manner. The monitoring range is limited and the data continuity is poor. It is impossible to achieve continuous monitoring of the strain and mechanical state of the entire drainage pipeline, and it is difficult to accurately quantify the mechanical response law of the pipeline under the coupled effects of multiple factors such as rainfall, seepage, and settlement.

[0006] In addition, the existing experimental systems operate with independent functional modules, lacking a coordinated mechanism between rainfall control, settlement simulation, and data acquisition equipment. Multi-source monitoring data cannot achieve unified spatiotemporal registration and fusion analysis, resulting in significant data noise interference and insufficient signal accuracy. Furthermore, conventional experimental equipment lacks intelligent disease identification, tiered early warning, and data prediction and analysis capabilities; it can only record basic data and cannot achieve early prediction of pipeline diseases or extrapolation of their evolution patterns, thus limiting the timeliness and practicality of experimental research. Therefore, developing an integrated experimental device and monitoring system that combines visual observation, high-precision fiber optic full-domain monitoring, multi-condition simulation, and intelligent data analysis is an urgent need to address current technical challenges and deepen research on drainage pipe network disasters.

[0007] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0008] The purpose of this invention is to address the shortcomings of existing technologies by proposing a visual test device and method for simulating the mechanical state of drainage pipe networks under the influence of rainfall. This invention incorporates complex engineering conditions such as rainfall infiltration, soil seepage evolution, and uneven settlement of roadbeds. It can solve the technical defects of existing drainage pipe network model test devices, such as low visualization level, single monitoring methods, discontinuous acquisition of mechanical parameters, poor fusion of multi-source data, and lack of early warning capabilities for defects.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: A visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall includes: The main test model box contains a roadbed soil test model and a drainage pipe model laid horizontally inside the roadbed soil test model. At least one side wall of the roadbed soil test model is set as a transparent baffle, and all pipe sections of the drainage pipe model are made of transparent pipe sections. The transparent baffle and the transparent pipe sections together constitute a visualization observation structure for real-time observation of water level changes, soil settlement and pipe structure deformation inside the roadbed soil. The settlement simulation mechanism at the bottom of the model box is set at the bottom of the test main model box, including at least one independently lifting settlement actuator plate. The settlement actuator plate is supported under the roadbed soil test model and is used to simulate the local void and uneven settlement conditions under the roadbed soil test model. A rainfall simulation module is installed above the main test model box and is used to apply controllable simulated rainfall to the subgrade soil test model; The soil internal seepage visualization and tracking unit includes a fluorescent tracer injection port set inside the roadbed soil test model, at least one transparent light guide tube pre-embedded in the roadbed soil test model along the depth direction, and an ultraviolet light source set outside; the side wall of the transparent light guide tube has multiple light-transmitting windows, and the ultraviolet light source can be introduced into the soil through the transparent light guide tube to realize the fluorescence observation of the three-dimensional seepage path and pore water transport trajectory inside the opaque soil under ultraviolet light irradiation; The fiber optic monitoring unit includes a distributed fiber optic strain sensor laid on the inner wall of the drainage pipe model, and an OFDR high-precision distributed fiber optic demodulator connected to the distributed fiber optic strain sensor, used to collect full-field strain data of the drainage pipe model under the influence of simulated rainfall and active settlement. An electromagnetic flow meter is installed at the bottom of the rainwater well of the drainage pipe model to monitor changes in flow rate within the pipe in real time. Pressure-type water level sensors are deployed in predetermined low-lying areas and well nodes within the roadbed soil test model to monitor water depth and water level changes in real time; and... The data acquisition and processing terminal is electrically and / or communicatively connected to the drive unit of the settlement actuation plate, the control unit of the rainfall simulation module, the OFDR high-precision distributed fiber optic demodulator, the electromagnetic flowmeter, and the pressure-type water level sensor, respectively. The data acquisition and processing terminal is configured to receive and process data from each sensing unit and drive unit, and to perform pipeline mechanical condition assessment and disease early warning.

[0010] Furthermore, the main experimental model box also includes an adjustable water level tank at the end. The adjustable water level tank is set at the outlet end of the drainage pipe model and is used to change the hydraulic boundary conditions at the outlet of the drainage pipe model by adjusting the water head height in the tank, so as to simulate the influence of the actual drainage network receiving water level change on the drainage capacity and mechanical state of the pipe.

[0011] Furthermore, the transparent tube segment is a high-transmittance acrylic tube; The distributed fiber optic strain sensor is wrapped with a transparent heat-shrink tubing encapsulation layer and is fixed by a transparent UV-curing adhesive along the bottom axis of the inner wall of the transparent tubing segment to avoid optical obstruction of the visualization observation structure by the fiber optic cable. The distributed fiber optic strain sensor used in the fiber optic monitoring unit is a single-mode sensing fiber based on optical frequency domain reflection technology, with a spatial resolution better than 1 mm, used to sense temperature changes inside and around the drainage pipe model; the distributed fiber optic strain sensor is continuously deployed along a preset path along the bottom and periphery of the entire pipe section of the drainage pipe model to realize three-dimensional spatial reconstruction of the strain of the entire pipe field; the data acquisition and processing terminal is configured to invert the acquired temperature field data into a visual cloud map indicating the water accumulation area around the pipe, the water flow path, and the flow characteristics.

[0012] Furthermore, the subgrade soil test model adopts a modular structure with layered filling, including at least a subgrade base layer for simulating the load-bearing and load-bearing structure of the subgrade, and a subgrade cushion layer located above the subgrade base layer for simulating the road drainage cushion layer; each filling layer is configured to be able to independently change the soil type, adjust the compaction degree and change the paving thickness according to the test conditions.

[0013] Furthermore, the data acquisition and processing terminal has a built-in multi-source heterogeneous data spatiotemporal registration and noise reduction algorithm module, which is configured to perform the following processing: Data from any of the optional pore water pressure and earth pressure sensors among the distributed fiber optic strain sensor, the electromagnetic flowmeter, the pressure water level sensor, and the soil internal seepage visualization tracking unit are uniformly timestamped and spatial coordinate mapped. An adaptive wavelet threshold denoising algorithm is used to remove environmental noise and abnormal jump signals. Weak strain signals are enhanced by a sliding window trend fitting algorithm, and high-fidelity multiphysics data are output.

[0014] Furthermore, the data acquisition and processing terminal has a built-in module for identifying and classifying pipeline defects as they gradually evolve. This module is configured as follows: Based on the pipeline strain time-series characteristics and strain rate data obtained by the distributed fiber optic strain sensor, combined with the water depth change rate obtained by the pressure water level sensor, the flow change data obtained by the electromagnetic flowmeter, and the settlement displacement data fed back by the settlement actuator plate, the early budding state of pipeline defects can be identified. The warning levels are divided into four levels: safe, alert, warning, and emergency, based on preset thresholds for settlement rate, stress increase, water accumulation depth, and seepage change. When the warning level reaches "emergency", the data acquisition and processing terminal generates a shutdown command and sends it to the control unit of the rainfall simulation module to cut off the simulated rainfall source and realize the warning and control closed loop.

[0015] Furthermore, the data acquisition and processing terminal has a built-in sensor data self-calibration and health diagnosis module, which is configured as follows: By utilizing the cross-verification relationship between the data collected by the distributed fiber optic strain sensor and the data collected by the piezometer, soil pressure gauge, and flow meter deployed at the same location, a cross-validation self-calibration model is constructed. Identify signal drift, disconnection, and abnormal output status of each sensor, and generate sensor health ratings and data reliability scores to support long-term stable monitoring without human intervention.

[0016] Furthermore, the data acquisition and processing terminal also has a built-in rainfall-seepage-mechanical response prediction algorithm module, which is configured as follows: Based on the current rainfall pattern, soil moisture content and seepage level time series data obtained by the pressure water level sensor and the soil internal seepage visualization tracking unit, the deformation trend of the drainage pipeline model and the probability of occurrence of various diseases in the future preset time period are predicted using a pre-constructed water-soil-pipeline coupled numerical model, and the early warning information is output in a visual form.

[0017] Furthermore, the experimental main model box, the rainfall simulation module, the fiber optic monitoring unit, the electromagnetic flowmeter, the pressure water level sensor, the soil internal seepage visualization tracking unit, and the data acquisition and processing terminal adopt a modular and detachable connection structure. The program modules stored in the memory of the data acquisition and processing terminal are encapsulated as independent microservice architecture containers, supporting independent upgrades and replacements of individual program modules. The data acquisition and processing terminal is also equipped with a cloud communication interface, which is used to upload the test measurement data that has been preprocessed locally and does not contain personal identification information to the cloud management platform. The cloud graphics processor is used to perform iterative optimization calculations of the three-dimensional reconstruction parameters of the pipeline stress field, and to receive the early warning threshold update data packets and remote control commands issued by the cloud management platform.

[0018] To achieve the above objectives, the present invention also employs the following technical solution: A visual experimental method for simulating the mechanical state of a drainage pipe network under the influence of rainfall, based on the device provided by this invention, includes the following steps: Step 1: Multi-source data acquisition: Simultaneously acquire multiple signals, perform timestamp alignment and data integrity verification, and construct the original dataset Z(t), Z(t) = {ε raw (x,t), Q raw (t), h raw (t), s raw (t), r raw (t), T raw (x,t)}, where ε raw (x,t), Q raw(t), h raw (t), s raw (t), r raw (t), T raw (x,t) represent the raw data of pipeline strain, flow rate, water level, settlement displacement, rainfall intensity, and temperature, respectively. Step 2, Spatiotemporal Registration: Perform spatiotemporal registration on the original dataset Z(t) to obtain the registered dataset X(t), where X(t) = {ε(x,t), Q(t), h(t), s(t), r(t), T(x,t)}; the spatiotemporal registration formula is as follows: Z(t-△t)]; Where Δt is the time delay and M is the spatial mapping operator; Step 3: Noise Reduction Processing: The registered dataset X(t) is denoised using an adaptive wavelet threshold function to obtain a denoised dataset. , ; Step 4, Feature Extraction: Extract the denoised dataset obtained in Step 3. Perform feature extraction to obtain feature vectors. The feature vector includes pipeline strain, strain rate of change, water level change, flow rate change, settlement displacement change, rainfall intensity, and temperature, specifically: ; in: ; ; ; ; in, For the current pipeline strain, For the rate of change of strain, This represents the change in water level. For changes in flow rate, This represents the change in settlement displacement. The rainfall intensity at the current moment. The temperature at the current moment; Step 5, Risk Assessment and Disease Identification: Based on the feature vector F t Calculate the risk index R t By combining historical test samples with predictive classification rules, a disease classification model was established to identify disease state C. t Risk index R t The calculation formula is as follows: ; in, is a weight coefficient; identifies the defect state C t The formula is as follows: ; where, C t is the defect state, g() is the defect classification model, and α is the classification model parameter; Step 6. Hierarchical early warning: Determine the early warning level L according to the risk index R t t , when R t <T1, L t is safe; when T1≤R t <T2, L t is caution; when T2≤R t <T3, L t is warning; when R t ≥T3, L t is emergency, is the early warning threshold; Step 7. Closed-loop control: When L t is emergency, a rain stop instruction is generated and sent to the rainfall module, otherwise the test operation is continued; Step 8. Trend prediction: According to the feature vector obtained in step 4 , the risk index obtained in step 5 and the defect state , construct a current state vector , which is used to characterize the comprehensive state of the drainage pipeline model at the current moment under the actions of rainfall, seepage, settlement and stress: ; where, is the pipeline strain at the current moment, is the strain change rate, is the water level variation, is the flow variation, is the settlement displacement variation, is the rainfall intensity at the current moment, is the temperature at the current moment, is the risk index, is the defect state; Input the current state vector into a pre-constructed water-soil-pipeline coupling prediction model to predict the pipeline deformation trend after a preset duration in the future: ; where, is the predicted value of pipeline deformation at the future moment , is the water-soil-pipeline coupling prediction model,​​ For model parameters, For the predicted duration; Meanwhile, according to the risk index and disease status Calculate the probability of disease occurrence : ; in, This represents the probability of disease occurrence. , , These are parameters for probability calculation; Step 9, Results Visualization and Report Output: Based on future trends and warning level L t It generates a visual cloud map, trend lines, and early warning reports. The visual cloud map is generated by spatial coordinate mapping and interpolation calculations based on pipeline strain, water level changes, settlement displacement, and predicted deformation results.

[0019] Compared with the prior art, the beneficial effects of this invention are as follows: 1. By using transparent baffles and transparent pipe sections to form a visual observation structure, and in conjunction with a soil internal seepage visualization tracking unit, it is possible to observe the changes in roadbed soil water level, soil settlement and pipe structure deformation in real time. It also enables fluorescent observation of the three-dimensional seepage path and pore water transport trajectory inside opaque soil. The experimental phenomena are intuitive and controllable, effectively improving the observability of the experimental process.

[0020] 2. The settlement actuator plate at the bottom of the model box can be raised and lowered independently, which can actively simulate the local voids and uneven settlement conditions under the subgrade soil. At the same time, with the adjustable water level tank at the end, the hydraulic boundary conditions at the outlet of the drainage pipe model can be changed, accurately restoring the influence of the receiving water level change on the drainage capacity and mechanical state of the pipe. The test conditions are close to the actual project and the degree of restoration is high.

[0021] 3. The OFDR type high-precision distributed fiber optic strain sensor is continuously deployed along the inner wall of the drainage pipe model, which can realize the three-dimensional spatial reconstruction of the strain of the entire pipe field; at the same time, it is combined with electromagnetic flowmeter and pressure water level sensor to synchronously collect flow and water level data, realize multi-physics field data collaborative monitoring, and significantly improve data integrity and detection accuracy.

[0022] 4. The data acquisition and processing terminal has a built-in multi-source heterogeneous data spatiotemporal registration and noise reduction algorithm module, which can perform unified timestamp calibration and spatial coordinate mapping on multi-source data such as optical fiber, electromagnetic flow meter, and water level sensor, and remove environmental noise through adaptive wavelet threshold denoising, effectively improving the consistency and reliability of multi-physics field data.

[0023] 5. Breaking through the limitations of traditional single-parameter threshold early warning systems, the system accurately identifies the early stages of pipeline defects through multi-source data coupling analysis of pipeline strain, seepage level, flow rate, and settlement displacement. The system can dynamically generate tiered early warnings based on operating conditions and link with a rainfall simulation module to execute closed-loop control, significantly improving the reliability of early warnings and the safety of testing.

[0024] 6. Its sensor data self-calibration and health diagnosis module can build a multi-sensor cross-validation self-calibration model, automatically identify signal drift, disconnection and abnormal output status, and generate sensor health rating and data reliability score, ensuring the stability and data reliability of long-term experimental monitoring.

[0025] 7. Through the rainfall-seepage-mechanical response prediction algorithm module, it is possible to predict the pipeline deformation trend and the probability of disease occurrence based on the current rainfall pattern, soil moisture content and seepage water level time series data, using a water-soil-pipeline coupled numerical model, and output advanced early warning information in a visualized form, providing advanced data support for the research on the disaster mechanism of drainage pipe network and risk prevention and control.

[0026] 8. The main experimental model box and various functional modules adopt a modular and detachable connection structure. The program module of the data acquisition and processing terminal adopts a microservice architecture containerization, which supports independent upgrades and cloud collaborative computing. It can flexibly adapt to different experimental conditions and has excellent versatility and scalability. Attached Figure Description

[0027] Figure 1 This is a schematic diagram of the overall structure assembly of the visualization test device and fiber optic monitoring system for simulating the mechanical state of drainage pipe networks under the influence of rainfall, as described in an embodiment of the present invention. Figure 2 This is a schematic diagram of the rainfall simulation module and water supply structure described in an embodiment of the present invention; Figure 3 This is a cross-sectional view of the experimental main model box and drainage pipe layout described in an embodiment of the present invention; Figure 4 This is a schematic diagram of the bottom settlement simulation mechanism of the model box according to an embodiment of the present invention; Figure 5 This is a block diagram illustrating the principle of the fiber optic monitoring and multi-source data acquisition and processing system described in an embodiment of the present invention.

[0028] In the diagram: 1. Adjustable water level tank; 2. Main test model box; 3. Rainfall simulation module; 4. Rainfall nozzle; 5. Subgrade layer; 6. Subgrade base layer; 7. Drainage pipe model; 8. Transparent baffle; 9. Settlement actuation plate; 10. Settlement drive unit; 11. Support base; 12. Light-transmitting window; 13. Ultraviolet light source; 14. Pressure-type water level sensor; 15. Data acquisition and processing terminal; 16. Water supply tank; 17. Water supply pump; 18. Flow regulating valve. Detailed Implementation

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0030] Example 1: A visual experimental device for simulating the mechanical state of drainage pipe networks under the influence of rainfall, such as Figures 1-5 As shown, the system includes a test main model box 2, a settlement simulation mechanism at the bottom of the model box, a rainfall simulation module 3, a soil internal seepage visualization and tracking unit, a fiber optic monitoring unit, an electromagnetic flowmeter, a pressure water level sensor 14, an adjustable water level tank at the end, and a data acquisition and processing terminal 15. Each component is assembled using a modular, detachable structure, allowing for flexible combination, disassembly, and expansion according to test conditions. The structure is stable and rationally laid out, enabling it to realistically reproduce the multi-field coupled engineering environment of rainfall infiltration, soil seepage evolution, uneven roadbed settlement, and stress deformation of drainage pipes.

[0031] In this embodiment, the main test model box 2 adopts a rigid frame structure with sufficient structural strength and rigidity to withstand the lateral earth pressure generated during the filling and compaction of the roadbed soil and the test process, thus avoiding interference with the test results due to deformation of the model box. At least one side wall of the main test model box 2 is equipped with a transparent baffle 8. The transparent baffle 8 is made of high-transmittance, high-rigidity tempered glass or acrylic sheet, which has good light transmittance and impact resistance, and can realize direct external observation throughout the test process.

[0032] In this embodiment, the test model box 2 contains a subgrade soil test model. The subgrade soil test model adopts a modular structure with layered filling and compaction, with the subgrade base layer 6 and the subgrade cushion layer 5 arranged sequentially from bottom to top. The subgrade base layer 6, as the load-bearing structure, uses filler with high mechanical strength and stable compaction characteristics; the subgrade cushion layer 5, as the drainage and stress diffusion layer, uses materials with suitable permeability and reasonable gradation. The soil type, compaction degree, moisture content, and paving thickness of each filling layer can be independently adjusted according to the test conditions, which can simulate various typical engineering conditions such as weak subgrade, insufficiently compacted subgrade, and poorly graded subgrade, thus improving the versatility and applicability of the test device.

[0033] In this embodiment, a drainage pipe model 7 is horizontally laid inside the roadbed soil test model. The burial depth and laying slope are set according to current engineering specifications, which can realistically reflect the conventional burial state of underground drainage pipes. An adjustable water level tank 1 is set at the outlet end of the drainage pipe model 7. The height and volume of the tank can be adjusted according to the test requirements. By adjusting the water head height in the tank, the hydraulic boundary conditions at the outlet of the drainage pipe model 7 can be precisely controlled. This is used to simulate the impact of water level changes in receiving water bodies such as rivers, lakes, and municipal pipe networks on the drainage capacity of the pipe, the water flow state inside the pipe, and the mechanical response of the pipe, making the test conditions closer to the actual engineering scenario.

[0034] In this embodiment, the drainage pipe model 7 is made entirely of transparent pipe sections. The transparent pipe sections and the transparent baffle 8 together form a visual observation structure, which can observe the changes in water level inside the roadbed soil, soil settlement and deformation, crack development, and the structural deformation, interface displacement and pipe wall bending state of the drainage pipe model 7 in real time, thus solving the technical defects of traditional model tests that are closed, invisible and unable to be observed in real time.

[0035] In this embodiment, a settlement simulation mechanism is installed at the bottom of the test model box 2. The settlement simulation mechanism includes at least one independently movable settlement actuation plate 9, which is supported under the subgrade soil test model. Each settlement actuation plate 9 is equipped with an independent settlement drive unit 10, which can realize individual action or coordinated action of multiple plates. The settlement drive unit 10 adopts a servo electric cylinder or hydraulic drive device, which has high displacement control accuracy, stable operation, and fast response speed. The settlement displacement, settlement rate, and range of action can be digitally set and fed back in real time through the data acquisition and processing terminal 15. A support base 11 is provided at the bottom of the settlement drive unit 10 to support the settlement drive unit 10. During the test, by controlling the downward displacement of the settlement actuation plate 9, typical disease conditions such as local voids, stratum loss, uneven settlement, and foundation instability under the subgrade soil can be actively simulated, providing stable and controllable test conditions for studying the mechanical response of pipelines under settlement.

[0036] In this embodiment, a rainfall simulation module 3 is installed above the main experimental model box 2. The rainfall simulation module 3 consists of a support frame, a spray pipe, a flow regulating valve 18, a water supply tank 16, a water supply pump 17, and rainfall nozzles 4. It can realize continuous adjustment of rainfall intensity, rainfall duration, and rainfall spatial distribution. It can simulate various rainfall conditions such as light rain, moderate rain, heavy rain, rainstorm, and long-duration continuous rainfall. It has good rainfall uniformity and high stability, and can provide controllable, repeatable, and quantifiable rainfall infiltration conditions for the experiment. It can realistically reproduce the complete process of rainwater infiltrating the roadbed, disturbing the soil structure, and changing the stress state of the pipeline under heavy rainfall conditions.

[0037] In this embodiment, the soil internal seepage visualization and tracking unit includes a fluorescent tracer injection port, a transparent light guide tube, and an ultraviolet light source 13. The transparent light guide tube is pre-embedded along the depth direction of the roadbed soil test model, and multiple sets of light-transmitting windows 12 are opened along the axial direction on the side wall of the transparent light guide tube to ensure that ultraviolet light is uniformly injected into the soil. The ultraviolet light source 13 is located outside the test main model box 2. Ultraviolet light can be introduced into the soil through the transparent light guide tube. Under ultraviolet light excitation, the fluorescent tracer can form a clear and bright migration trajectory. During the test, the tracer is injected into the soil through the fluorescent tracer injection port. The tracer migrates synchronously with the pore water, which can intuitively display the three-dimensional seepage path, dominant seepage channels, pore water migration velocity, and stagnant area distribution inside the opaque soil, realizing the visualization observation and qualitative analysis of the seepage field inside the soil.

[0038] In this embodiment, the fiber optic monitoring unit is the core monitoring module of the device, including a distributed fiber optic strain sensor and an OFDR-type high-precision distributed fiber optic demodulator. The distributed fiber optic strain sensor uses a single-mode sensing fiber based on optical frequency domain reflection technology, with a spatial resolution better than 1 mm, enabling high-precision, high-density, and continuous acquisition of pipeline strain. The distributed fiber optic strain sensor is externally encapsulated in transparent heat-shrink tubing and fixed along the bottom axis of the inner wall of the drainage pipeline model 7 using transparent UV-curable adhesive, without causing optical obstruction and without affecting the visualization observation effect. The distributed fiber optic strain sensor is continuously deployed along the entire length of the pipeline and a preset path around the pipe, enabling the acquisition of full-field strain data of the drainage pipeline model 7 under the coupling effects of rainfall infiltration, soil settlement, and water flow impact. It also enables three-dimensional spatial reconstruction of the full-field strain of the pipeline and visualization analysis of the stress field, overcoming the shortcomings of traditional single-point sensor monitoring, such as discontinuous and incomplete data, and difficulty in reflecting full-domain deformation.

[0039] In this embodiment, to achieve multi-physics collaborative monitoring, an electromagnetic flowmeter is installed at the bottom of the rainwater well in the drainage pipeline model 7 to collect real-time changes in flow rate and velocity within the pipe, reflecting the pipeline's drainage efficiency. Pressure-type water level sensors 14 are deployed in low-lying areas, well nodes, and typical monitoring sections within the roadbed soil test model to monitor real-time changes in soil water depth and pore water level, reflecting rainwater infiltration and soil saturation. All sensor signals are synchronously transmitted to the data acquisition and processing terminal 15, achieving integrated acquisition of multi-source data such as strain, flow rate, water level, and settlement displacement.

[0040] In this embodiment, the data acquisition and processing terminal 15 serves as the core of the device's control and analysis. It establishes electrical and communication connections with the settlement actuator plate 9 drive unit, rainfall simulation module 3, OFDR distributed fiber optic demodulator, electromagnetic flowmeter, and pressure-type water level sensor 14, respectively, to achieve centralized control, data acquisition, processing and analysis, and early warning output. The terminal has built-in modules for multi-source heterogeneous data spatiotemporal registration and denoising, pipeline defect progressive evolution identification and graded early warning, sensor data self-calibration and health diagnosis, and rainfall seepage mechanical response prediction. It can perform functions such as unified timestamp calibration of multi-source data, spatial coordinate mapping, adaptive wavelet threshold denoising, weak signal enhancement, pipeline defect identification, graded early warning, sensor self-diagnosis, and deformation trend prediction.

[0041] In this embodiment, when the monitoring data reaches a preset threshold, the system automatically executes a warning according to four levels: safety, attention, early warning, and emergency. When the warning level reaches the emergency level, the terminal automatically generates a control command to shut down the rainfall simulation module 3 and cut off the simulated rainfall source, forming a closed loop of "monitoring-analysis-early warning-control" to improve the safety and intelligence level of the experiment.

[0042] In this embodiment, the device adopts a modular and detachable structure, allowing for flexible hardware assembly. The software employs a microservice containerized architecture, with each algorithm module operating independently, supporting individual upgrades, replacements, and maintenance. The terminal is equipped with a cloud communication interface, enabling the uploading of locally pre-processed test data to the cloud management platform. Leveraging cloud computing power, it performs three-dimensional reconstruction of the stress field and iterative optimization of parameters. It also receives early warning threshold update data packets and remote control commands from the cloud, achieving remote monitoring, data sharing, and intelligent operation and maintenance.

[0043] This embodiment presents a visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall. The test operation steps are as follows: 1. Model preparation. According to the test plan, the subgrade base layer 6 and subgrade cushion layer 5 were filled in layers in the test main model box 2, and the compaction degree and thickness of each layer were strictly controlled; the drainage pipe model 7, distributed fiber optic strain sensor, pressure water level sensor 14 and transparent light guide tube were pre-embedded in the preset position to complete the preparation of the subgrade soil and pipe model.

[0044] 2. System Connection and Debugging. Connect each hardware module to the data acquisition and processing terminal 15, check the working status of the drive unit, sensor, demodulation equipment and rainfall module to ensure that the signal is normal and the operation is reliable; set the outlet hydraulic boundary conditions through the adjustable water level tank 1 at the end, and preset the settlement parameters through the settlement simulation mechanism.

[0045] 3. Visual observation and data acquisition. Turn on the ultraviolet light source 13 and inject the tracer through the fluorescent tracer injection port; start the rainfall simulation module 3 and data acquisition system, and carry out the experiment according to the set rainfall pattern; observe the soil seepage trajectory, settlement development and pipeline deformation status in real time, and simultaneously collect multi-physical field data such as strain, flow rate, water level and settlement.

[0046] 4. Intelligent Analysis and Early Warning Control. The data acquisition and processing terminal 15 performs real-time data processing, fusion analysis, and disease identification, and issues graded early warnings based on preset thresholds. If necessary, it automatically shuts down the rainfall simulation module 3 to ensure experimental safety.

[0047] 5. Test Completion and Data Processing. After the test is completed, data acquisition, rainfall simulation, and ultraviolet light source 13 are stopped sequentially; the test data are exported and post-processed for analysis to generate a pipeline mechanical condition assessment report, disease evolution curves, and visualization cloud maps.

[0048] This embodiment presents a visualization test device for simulating the mechanical state of drainage pipe networks under the influence of rainfall. It realistically reproduces the multi-field coupled environment of rainfall, seepage, soil settlement, and pipe stress, enabling visualized tracking of seepage paths within the soil and high-precision continuous monitoring of mechanical strain across the entire pipe network. Simultaneously, it collects multi-dimensional parameters such as flow rate, water level, and settlement displacement. Through multi-source data fusion algorithms and intelligent analysis modules, it achieves early identification of pipe defects, graded early warning, and prediction of mechanical response trends. This solves the technical challenges of traditional tests, such as difficulty in quantifying pipe network disaster mechanisms, poor observability, insufficient monitoring accuracy, and low test condition reproduction. It provides reliable experimental support and data for urban drainage pipe network structural safety assessment, defect evolution mechanism research, and disaster prevention and maintenance technology optimization.

[0049] The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall, as described in this embodiment, has the following beneficial effects: 1. By using transparent baffles and transparent pipe sections to form a visual observation structure, and in conjunction with a soil internal seepage visualization tracking unit, it is possible to observe the changes in roadbed soil water level, soil settlement and pipe structure deformation in real time. It also enables fluorescent observation of the three-dimensional seepage path and pore water transport trajectory inside opaque soil. The experimental phenomena are intuitive and controllable, effectively improving the observability of the experimental process.

[0050] 2. The settlement actuator plate at the bottom of the model box can be raised and lowered independently, which can actively simulate the local voids and uneven settlement conditions under the subgrade soil. At the same time, with the adjustable water level tank at the end, the hydraulic boundary conditions at the outlet of the drainage pipe model can be changed, accurately restoring the influence of the receiving water level change on the drainage capacity and mechanical state of the pipe. The test conditions are close to the actual project and the degree of restoration is high.

[0051] 3. The OFDR type high-precision distributed fiber optic strain sensor is continuously deployed along the inner wall of the drainage pipe model, which can realize the three-dimensional spatial reconstruction of the strain of the entire pipe field; at the same time, it is combined with electromagnetic flowmeter and pressure water level sensor to synchronously collect flow and water level data, realize multi-physics field data collaborative monitoring, and significantly improve data integrity and detection accuracy.

[0052] 4. The data acquisition and processing terminal has a built-in multi-source heterogeneous data spatiotemporal registration and noise reduction algorithm module, which can perform unified timestamp calibration and spatial coordinate mapping on multi-source data such as optical fiber, electromagnetic flow meter, and water level sensor, and remove environmental noise through adaptive wavelet threshold denoising, effectively improving the consistency and reliability of multi-physics field data.

[0053] 5. Breaking through the limitations of traditional single-parameter threshold early warning systems, the system accurately identifies the early stages of pipeline defects through multi-source data coupling analysis of pipeline strain, seepage level, flow rate, and settlement displacement. The system can dynamically generate tiered early warnings based on operating conditions and link with a rainfall simulation module to execute closed-loop control, significantly improving the reliability of early warnings and the safety of testing.

[0054] 6. Its sensor data self-calibration and health diagnosis module can build a multi-sensor cross-validation self-calibration model, automatically identify signal drift, disconnection and abnormal output status, and generate sensor health rating and data reliability score, ensuring the stability and data reliability of long-term experimental monitoring.

[0055] 7. Through the rainfall-seepage-mechanical response prediction algorithm module, it is possible to predict the pipeline deformation trend and the probability of disease occurrence based on the current rainfall pattern, soil moisture content and seepage water level time series data, using a water-soil-pipeline coupled numerical model, and output advanced early warning information in a visualized form, providing advanced data support for the research on the disaster mechanism of drainage pipe network and risk prevention and control.

[0056] 8. The main experimental model box and various functional modules adopt a modular and detachable connection structure. The program module of the data acquisition and processing terminal adopts a microservice architecture containerization, which supports independent upgrades and cloud collaborative computing. It can flexibly adapt to different experimental conditions and has excellent versatility and scalability.

[0057] Example 2: This embodiment is basically the same as Embodiment 1 in terms of structure and working principle, and the same parts will not be repeated. The difference is that: while collecting the strain of the pipeline, the distributed fiber optic strain sensor can simultaneously sense the temperature changes of the surrounding soil and water flow, forming temperature field monitoring data. The data acquisition and processing terminal 15 inverts and calculates the temperature field data, transforming it into a visualized cloud map of the water accumulation area around the pipe, the water flow path, and the flow characteristics, further enriching the monitoring dimensions and realizing a more intuitive and comprehensive expression of the water-soil coupling state.

[0058] Example 3: This embodiment has the same overall structure and test process as Embodiment 1, and the same parts will not be repeated. The difference is that: in the process of multi-source data processing, the data acquisition and processing terminal 15 adopts an adaptive wavelet threshold denoising algorithm to remove noise signals such as environmental vibration, electromagnetic interference, and soil fluctuation, and enhances weak strain signals through a sliding window trend fitting algorithm to improve the data recognition capability under weak deformation conditions, output high-fidelity, high signal-to-noise ratio multi-physics field data, and improve the accuracy and reliability of test results.

[0059] Example 4: This embodiment has the same main structure and functional modules as Embodiment 1, and the same parts will not be repeated. The difference is that the data acquisition and processing terminal 15 has a built-in sensor data self-calibration and health diagnosis module. It uses the physical correlation and mutual verification relationship of multi-source data such as distributed fiber optic strain sensors, piezometers, soil pressure gauges, and electromagnetic flowmeters to build a cross-validation self-calibration model, automatically identify sensor signal drift, disconnection, abnormal jump and other faults, and generate sensor health rating and data credibility score, supporting stable and reliable monitoring under long-term unattended conditions.

[0060] Example 5: This embodiment has the same overall structure and working mode as Embodiment 1, and the same parts will not be repeated. The difference is that the data acquisition and processing terminal 15 has a built-in rainfall seepage mechanical response prediction algorithm module. Based on real-time rainfall pattern, soil moisture content and seepage water level time series data, combined with the pre-constructed water-soil pipeline coupled numerical model, it can predict the deformation trend, stress development and probability of disease occurrence of drainage pipeline model 7 in the future preset time period, and output it in the form of visualization such as cloud map, curve and early warning text, providing advanced data support for experimental process control, disaster mechanism analysis and engineering risk prevention and control.

[0061] Example 6: A visual experimental method for simulating the mechanical state of a drainage pipe network under the influence of rainfall, based on any one of the devices in Examples 1 to 5, includes the following steps: Step 1: Multi-source data acquisition: Simultaneously acquire multiple signals, perform timestamp alignment and data integrity verification, and construct the original dataset Z(t), Z(t) = {ε raw (x,t), Q raw (t), h raw (t), s raw (t), r raw (t), T raw (x,t)}, where ε raw (x,t), Q raw (t), h raw (t), s raw (t), rraw (t), T raw (x,t) represent the raw data for pipeline strain, flow rate, water level, settlement displacement, rainfall intensity, and temperature, respectively.

[0062] In this embodiment, in step 1, ε raw (x,t), Q raw (t), h raw (t), s raw (t), r raw (t), T raw (x,t) are obtained through an OFDR-type high-precision distributed optical fiber demodulator, an electromagnetic flowmeter, a water level sensor, a sedimentation feedback signal, a rainfall control status, and an optical fiber monitoring unit, respectively.

[0063] Step 2, Spatiotemporal Registration: Perform spatiotemporal registration on the original dataset Z(t) to obtain the registered dataset X(t), where X(t) = {ε(x,t), Q(t), h(t), s(t), r(t), T(x,t)}; the spatiotemporal registration formula is as follows: Z(t-△t)]; Where Δt is the time delay and M is the spatial mapping operator.

[0064] Step 3: Noise Reduction Processing: The registered dataset X(t) is denoised using an adaptive wavelet threshold function to obtain a denoised dataset. , .

[0065] In this embodiment, the adaptive wavelet threshold function formula in step 3 is as follows: ; ; in, The wavelet coefficients of the input signal at scale j and position k are... For noise estimation, N is the sample length.

[0066] Step 4, Feature Extraction: Extract the denoised dataset obtained in Step 3. Perform feature extraction to obtain feature vectors. The feature vector includes pipeline strain, strain rate of change, water level change, flow rate change, settlement displacement change, rainfall intensity, and temperature, specifically: ; in: ; ; ; ; wherein, is the pipeline strain at the current moment, is the strain change rate, is the water level variation, is the flow rate variation, is the settlement displacement variation, is the rainfall intensity at the current moment, is the temperature at the current moment; This step outputs a feature vector , which is used as the input for subsequent risk assessment, disease identification and trend prediction.

[0067] Step 5. Risk Assessment and Disease Identification: Calculate the risk index R based on the feature vector F t t , and establish a disease classification model in combination with historical test samples and prediction classification rules to identify the disease state C t , wherein the calculation formula of the risk index R t is as follows: ; wherein, is a weight coefficient; The formula for identifying the disease state C t is as follows: ; wherein, C t is the disease state, g() is the disease classification model, and α is the classification model parameter; Step 6. Graded Early Warning: Determine the early warning level L based on the risk index R t t , when R t <T1, L t is safe; when T1≤R t <T2, L t is attention; when T2≤R t <T3, L t is early warning; when R t ≥T3, L t is emergency, is the early warning threshold; Step 7. Closed-Loop Control: When L t is emergency, a rain pause instruction is generated and sent to the rainfall module, otherwise the test continues to run; Step 8. Trend Prediction: Construct the current state vector based on the feature vector obtained in Step 4 , the risk index obtained in Step 5 and the disease state ​​​It is used to characterize the combined state of the drainage pipeline model at the current moment under the combined effects of rainfall, seepage, settlement, and stress: ; in, For the current pipeline strain, For the rate of change of strain, This represents the change in water level. For changes in flow rate, This represents the change in settlement displacement. The rainfall intensity at the current moment. The temperature at the current moment. As a risk index, The condition is in a diseased state; The current state vector Input a pre-built water-soil-pipeline coupled prediction model to predict the future preset duration. Subsequent pipeline deformation trends: ; in, For the future Predicted values ​​of pipeline deformation For water-soil-pipeline coupled prediction model, For model parameters, For the predicted duration; Meanwhile, according to the risk index and disease status Calculate the probability of disease occurrence : ; in, This represents the probability of disease occurrence. , , These are parameters for probability calculation; This step outputs the future pipeline deformation trend. and the probability of disease occurrence This is then passed to step 9 to generate trend curves, visualization cloud maps, and early warning reports.

[0068] Step 9, Results Visualization and Report Output: Based on future trends and warning level L t It generates a visual cloud map, trend lines, and early warning reports. The visual cloud map is generated by spatial coordinate mapping and interpolation calculations based on pipeline strain, water level changes, settlement displacement, and predicted deformation results.

[0069] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A visual experimental device for simulating the mechanical state of a drainage pipe network under the influence of rainfall, characterized in that, include: The main test model box contains a roadbed soil test model and a drainage pipe model laid horizontally inside the roadbed soil test model. At least one side wall of the roadbed soil test model is set as a transparent baffle, and all pipe sections of the drainage pipe model are made of transparent pipe sections. The transparent baffle and the transparent pipe sections together constitute a visualization observation structure for real-time observation of water level changes, soil settlement and pipe structure deformation inside the roadbed soil. The settlement simulation mechanism at the bottom of the model box is set at the bottom of the test main model box, including at least one independently lifting settlement actuator plate. The settlement actuator plate is supported under the roadbed soil test model and is used to simulate the local void and uneven settlement conditions under the roadbed soil test model. A rainfall simulation module is installed above the main test model box and is used to apply controllable simulated rainfall to the subgrade soil test model; The soil internal seepage visualization and tracking unit includes a fluorescent tracer injection port set inside the roadbed soil test model, at least one transparent light guide tube pre-embedded in the roadbed soil test model along the depth direction, and an ultraviolet light source set outside; the side wall of the transparent light guide tube has multiple light-transmitting windows, and the ultraviolet light source can be introduced into the soil through the transparent light guide tube to realize the fluorescence observation of the three-dimensional seepage path and pore water transport trajectory inside the opaque soil under ultraviolet light irradiation; The fiber optic monitoring unit includes a distributed fiber optic strain sensor laid on the inner wall of the drainage pipe model, and an OFDR high-precision distributed fiber optic demodulator connected to the distributed fiber optic strain sensor, used to collect full-field strain data of the drainage pipe model under the influence of simulated rainfall and active settlement. An electromagnetic flow meter is installed at the bottom of the rainwater well of the drainage pipe model to monitor changes in flow rate within the pipe in real time. Pressure-type water level sensors are deployed in predetermined low-lying areas and well nodes within the roadbed soil test model to monitor water depth and water level changes in real time; and... The data acquisition and processing terminal is electrically and / or communicatively connected to the drive unit of the settlement actuation plate, the control unit of the rainfall simulation module, the OFDR high-precision distributed fiber optic demodulator, the electromagnetic flowmeter, and the pressure-type water level sensor, respectively. The data acquisition and processing terminal is configured to receive and process data from each sensing unit and drive unit, and to perform pipeline mechanical condition assessment and disease early warning.

2. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The main experimental model box also includes an adjustable water level tank at the end. The adjustable water level tank is set at the outlet end of the drainage pipe model and is used to change the hydraulic boundary conditions at the outlet of the drainage pipe model by adjusting the water head height in the tank, so as to simulate the influence of the actual drainage network receiving water level changes on the drainage capacity and mechanical state of the pipe.

3. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The transparent tube segment is a high-transmittance acrylic tube; The distributed fiber optic strain sensor is wrapped with a transparent heat-shrink tubing encapsulation layer and is fixed by a transparent UV-curing adhesive along the bottom axis of the inner wall of the transparent tubing segment to avoid optical obstruction of the visualization observation structure by the fiber optic cable. The distributed fiber optic strain sensor used in the fiber optic monitoring unit is a single-mode sensing fiber based on optical frequency domain reflection technology, with a spatial resolution better than 1 mm, used to sense temperature changes inside and around the drainage pipe model; the distributed fiber optic strain sensor is continuously deployed along a preset path along the bottom and periphery of the entire pipe section of the drainage pipe model to realize three-dimensional spatial reconstruction of the strain of the entire pipe field; the data acquisition and processing terminal is configured to invert the acquired temperature field data into a visual cloud map indicating the water accumulation area around the pipe, the water flow path, and the flow characteristics.

4. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The subgrade soil test model adopts a modular structure with layered filling. From bottom to top, it includes at least a subgrade base layer for simulating the load-bearing and load-bearing structure of the subgrade, and a subgrade cushion layer located above the subgrade base layer for simulating the road drainage cushion layer. Each filling layer is configured to be able to independently change the soil type, adjust the compaction degree and change the paving thickness according to the test conditions.

5. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The data acquisition and processing terminal has a built-in multi-source heterogeneous data spatiotemporal registration and noise reduction algorithm module, which is configured to perform the following processing: Data from any of the optional pore water pressure and earth pressure sensors among the distributed fiber optic strain sensor, the electromagnetic flowmeter, the pressure water level sensor, and the soil internal seepage visualization tracking unit are uniformly timestamped and spatial coordinate mapped. An adaptive wavelet threshold denoising algorithm is used to remove environmental noise and abnormal jump signals. Weak strain signals are enhanced by a sliding window trend fitting algorithm, and high-fidelity multiphysics data are output.

6. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The data acquisition and processing terminal has a built-in pipeline defect progressive evolution identification and hierarchical early warning module, which is configured as follows: Based on the pipeline strain time-series characteristics and strain rate data obtained by the distributed fiber optic strain sensor, combined with the water depth change rate obtained by the pressure water level sensor, the flow change data obtained by the electromagnetic flowmeter, and the settlement displacement data fed back by the settlement actuator plate, the early budding state of pipeline defects can be identified. The warning levels are divided into four levels: safe, alert, warning, and emergency, based on preset thresholds for settlement rate, stress increase, water accumulation depth, and seepage change. When the warning level reaches "emergency", the data acquisition and processing terminal generates a shutdown command and sends it to the control unit of the rainfall simulation module to cut off the simulated rainfall source and realize the warning and control closed loop.

7. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The data acquisition and processing terminal has a built-in sensor data self-calibration and health diagnosis module, which is configured as follows: By utilizing the cross-verification relationship between the data collected by the distributed fiber optic strain sensor and the data collected by the piezometer, soil pressure gauge, and flow meter deployed at the same location, a cross-validation self-calibration model is constructed. Identify signal drift, disconnection, and abnormal output status of each sensor, and generate sensor health ratings and data reliability scores to support long-term stable monitoring without human intervention.

8. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall according to claim 1, characterized in that, The data acquisition and processing terminal also has a built-in rainfall-seepage-mechanical response prediction algorithm module, which is configured as follows: Based on the current rainfall pattern, soil moisture content and seepage level time series data obtained by the pressure water level sensor and the soil internal seepage visualization tracking unit, the deformation trend of the drainage pipeline model and the probability of occurrence of various diseases in the future preset time period are predicted using a pre-constructed water-soil-pipeline coupled numerical model, and the early warning information is output in a visual form.

9. The visualization test device for simulating the mechanical state of a drainage pipe network under the influence of rainfall as described in claim 1, characterized in that, The experimental main model box, the rainfall simulation module, the fiber optic monitoring unit, the electromagnetic flowmeter, the pressure water level sensor, the soil internal seepage visualization tracking unit, and the data acquisition and processing terminal adopt a modular and detachable connection structure. The program modules stored in the memory of the data acquisition and processing terminal are encapsulated as independent microservice architecture containers, supporting independent upgrades and replacements of individual program modules. The data acquisition and processing terminal is also equipped with a cloud communication interface, which is used to upload the test measurement data that has been preprocessed locally and does not contain personal identification information to the cloud management platform. The cloud graphics processor is used to perform iterative optimization calculations of the three-dimensional reconstruction parameters of the pipeline stress field, and to receive the early warning threshold update data packets and remote control commands issued by the cloud management platform.

10. A visual experimental method for simulating the mechanical state of a drainage pipe network under the influence of rainfall, characterized in that, The apparatus according to any one of claims 1 to 9 includes the following steps: Step 1: Multi-source data acquisition: Simultaneously acquire multiple signals, perform timestamp alignment and data integrity verification, and construct the original dataset Z(t), Z(t) = {ε raw (x,t), Q raw (t), h raw (t), s raw (t), r raw (t), T raw (x,t)}, where ε raw (x,t), Q raw (t), h raw (t), s raw (t), r raw (t), T raw (x,t) represent the raw data of pipeline strain, flow rate, water level, settlement displacement, rainfall intensity, and temperature, respectively. Step 2, Spatiotemporal Registration: Perform spatiotemporal registration on the original dataset Z(t) to obtain the registered dataset X(t), where X(t) = {ε(x,t), Q(t), h(t), s(t), r(t), T(x,t)}; the spatiotemporal registration formula is as follows: Z(t-△t)]; Where Δt is the time delay and M is the spatial mapping operator; Step 3: Noise Reduction Processing: The registered dataset X(t) is denoised using an adaptive wavelet threshold function to obtain a denoised dataset. , ; Step 4, Feature Extraction: Extract the denoised dataset obtained in Step 3. Perform feature extraction to obtain feature vectors. The feature vector includes pipeline strain, strain rate of change, water level change, flow rate change, settlement displacement change, rainfall intensity, and temperature, specifically: ； in: ； ； ； ； in, For the current pipeline strain, For the rate of change of strain, This represents the change in water level. For changes in flow rate, This represents the change in settlement displacement. The rainfall intensity at the current moment. The temperature at the current moment; Step 5, Risk Assessment and Disease Identification: Based on the feature vector F t Calculate the risk index R t By combining historical test samples with predictive classification rules, a disease classification model was established to identify disease state C. t Risk index R t The calculation formula is as follows: ； in, These are the weighting coefficients; Identify disease status C t The formula is as follows: ； Among them, C t Let g represent the disease state, g() represent the disease classification model, and α represent the classification model parameters. Step 6: Hierarchical warning: According to the risk index R t determine the warning level L t , R t is less than T1, L t is safe; when T1 ≤ R t is less than T2, L t is caution; when T2 ≤ R t is less than T3, L t is warning; when R t ≥ T3, L t is emergency, which are warning thresholds; Step 7, Closed-loop control: When L t In case of emergency, a rain listening command is generated and sent to the rainfall module; otherwise, the test operation continues. Step 8, Trend Prediction: Based on the feature vector obtained in Step 4 The risk index obtained in step 5 and disease status Construct the current state vector It is used to characterize the combined state of the drainage pipeline model at the current moment under the combined effects of rainfall, seepage, settlement, and stress: ； in, For the current pipeline strain, For the rate of change of strain, This represents the change in water level. For changes in flow rate, This represents the change in settlement displacement. The rainfall intensity at the current moment. The temperature at the current moment. As a risk index, The condition is in a diseased state; The current state vector Input a pre-built water-soil-pipeline coupled prediction model to predict the future preset duration. Subsequent pipe deformation trends: ； in, For the future Predicted values ​​of pipeline deformation For water-soil-pipeline coupled prediction model, For model parameters, For the predicted duration; Meanwhile, according to the risk index and disease status Calculate the probability of disease occurrence : ； in, This represents the probability of disease occurrence. , , These are parameters for probability calculation; Step 9, Results Visualization and Report Output: Based on future trends and warning level L t It generates a visual cloud map, trend lines, and early warning reports. The visual cloud map is generated by spatial coordinate mapping and interpolation calculations based on pipeline strain, water level changes, settlement displacement, and predicted deformation results.

Images