A seepage monitoring system and method for reservoir safety monitoring

Abstract

This invention relates to the field of reservoir dam safety monitoring technology, specifically to a seepage monitoring system and method for reservoir safety monitoring, comprising: a distributed seepage pressure monitoring array, in which multiple high-precision vibrating wire piezometers are arranged on each monitoring cross section of the reservoir dam body to form a three-dimensional monitoring grid, wherein the piezometers are protected by a novel nanofiber filter layer. This invention solves the problems of traditional single-point piezometers being unable to reflect the overall seepage field and being prone to siltation by using the three-dimensional monitoring grid of the distributed seepage pressure monitoring array and the novel nanofiber filter layer. Combined with the double-weir structure of the intelligent water measuring weir group and the ultrasonic silt detection module, it avoids the accuracy deviation caused by siltation in traditional water measuring weirs. Edge computing nodes establish a three-dimensional correlation model of seepage pressure and flow rate with time using an embedded multi-parameter coupling analysis algorithm, which can automatically identify abnormal seepage patterns, solving the problem that traditional monitoring data is isolated and cannot automatically identify potential dam hazards.

Description

Technical Field

[0001] This invention relates to the field of reservoir dam safety monitoring technology, specifically to a seepage monitoring system and method for reservoir safety monitoring. Background Technology

[0002] Seepage monitoring is a core component of dam safety monitoring. By measuring the pressure difference between the inside and outside of the dam and the seepage flow rate, the dam's seepage prevention performance and stability can be assessed. Traditional methods involve manually observing water levels using piezometers or installing piezometers, combined with measuring weirs to measure seepage flow. These data are crucial for determining the location of the dam's phreatic line and the effectiveness of the seepage prevention system, directly impacting reservoir safety assessments. However, small reservoirs currently suffer from insufficient monitoring facilities and low levels of automation, making it difficult to detect seepage anomalies in a timely manner.

[0003] Existing seepage monitoring systems suffer from significant real-time and accuracy issues: manual observation of piezometers relies on on-site measurements by inspection personnel, making it impossible to capture sudden changes in water level; single-point piezometer deployment fails to reflect the overall seepage field distribution; measuring weirs are susceptible to siltation affecting accuracy, and conventional triangular weirs exhibit errors as high as 15%-20% at low flow rates (<1L / s). More critically, data from various sensors are collected in isolation, lacking a collaborative analysis mechanism. When the piezometer water level rises but the seepage flow does not increase synchronously, the system cannot automatically identify the underlying dam structure hazards behind this anomaly, resulting in severely delayed early warnings of seepage accidents. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a seepage monitoring system and method for reservoir safety monitoring, which solves the problems of difficulty in capturing sudden changes in water level through manual observation, difficulty in reflecting the overall seepage field by single-point piezometers, low accuracy due to easy siltation of measuring weirs, and delayed early warning caused by isolated data and lack of collaborative analysis.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a seepage monitoring system for reservoir safety monitoring, comprising: A distributed piezometric monitoring array is provided, in which multiple high-precision vibrating wire piezometers are arranged on each monitoring cross section of the reservoir dam to form a three-dimensional monitoring grid. The piezometers are protected by a novel nano-filter layer, which is used to improve the piezometers' resistance to siltation. The intelligent water measuring weir group includes a main weir and an auxiliary weir double weir structure. The main weir adopts a V-shaped and rectangular composite weir opening. The composite weir opening can automatically switch the measurement mode according to the magnitude of the seepage flow. The auxiliary weir is equipped with an ultrasonic silt detection module. The ultrasonic silt detection module is used to detect the silt thickness in real time and calibrate the zero point of the water measuring weir group. An edge computing node is embedded with a multi-parameter coupled analysis algorithm. The edge computing node is used to establish a three-dimensional correlation model of seepage pressure, seepage flow rate and time, and to identify preset abnormal seepage patterns based on the model. The abnormal seepage patterns include high pressure differential and low flow rate or low pressure differential and high flow rate, and automatically trigger graded early warning. An adaptive power supply system, integrating a thermoelectric generator module and a solar panel as dual power sources, is capable of maintaining uninterrupted monitoring for 30 days during continuous rainy weather; and The lightning protection and anti-interference transmission network adopts a hybrid networking of shielded twisted pair cable and LoRa wireless. The transmission network has a communication redundancy design, which ensures dual-path transmission of data packets.

[0006] Furthermore, each monitoring cross section of the distributed osmotic pressure monitoring array is equipped with 3 to 5 of the high-precision vibrating wire osmometers, the osmometers having a range of 0 to 700 kPa and a resolution less than or equal to 0.02%FS.

[0007] Furthermore, the main weir of the intelligent water measuring weir group uses a V-shaped groove for measurement under low flow conditions and automatically switches to a rectangular section for measurement under high flow conditions. The automatic switching of the measurement mode enables the measurement error of the water measuring weir group to be controlled within 5% of the full range.

[0008] Furthermore, the multi-parameter coupling analysis algorithm embedded in the edge computing node includes: The seepage field reconstruction module is used to convert the discrete seepage pressure data collected by the distributed seepage pressure monitoring array into a two-dimensional hydraulic head contour map of the dam body using the Kriging interpolation algorithm; and An anomaly pattern recognition module is used to establish a spatiotemporal prediction model for seepage parameters based on LSTM. When the measured value of the seepage pressure data or seepage flow rate deviates from the predicted value by 15%, root cause analysis is initiated.

[0009] Furthermore, the communication redundancy design of the lightning protection and anti-interference transmission network ensures that the packet loss rate is less than 0.1%.

[0010] The present invention also provides a seepage monitoring method for reservoir safety monitoring, applied to the seepage monitoring system for reservoir safety monitoring described in any one of the above claims, comprising: S1. Seepage pressure data of the reservoir dam body is obtained through a distributed seepage pressure monitoring array. The distributed seepage pressure monitoring array has multiple high-precision vibrating wire seepage gauges arranged in each monitoring cross section of the reservoir dam body to form a three-dimensional monitoring grid. The seepage gauges are protected by a new type of nano-filter layer. S2. Obtain seepage flow data of the reservoir dam body through intelligent water measuring weir group. The intelligent water measuring weir group includes a main weir and an auxiliary weir double weir structure. The main weir adopts a V-shaped and rectangular composite weir opening. The auxiliary weir is equipped with an ultrasonic silt detection module. S3. Based on edge computing nodes, perform multi-parameter coupling analysis on the seepage pressure data obtained in step S1 and the seepage flow data obtained in step S2 to establish a three-dimensional correlation model of seepage pressure, seepage flow and time, and identify preset abnormal seepage patterns, including high pressure difference and low flow or low pressure difference and high flow. S4. Automatically trigger graded early warnings based on the identified abnormal seepage patterns; S5. Execute a self-maintenance mechanism, which includes: the intelligent water measuring weir group automatically triggers the backwashing device to perform high-pressure water gun cleaning based on the sludge thickness detected by the ultrasonic sludge detection module; and the piezometer performs a self-check pulse once a day to remove the deposits on the nanofilter layer.

[0011] Furthermore, in step S1, the piezometer has a range of 0 to 700 kPa and a resolution of ≤0.02%FS.

[0012] Furthermore, in step S2, the main weir can automatically switch the measurement mode according to the magnitude of the seepage flow. Under low flow conditions, a V-shaped groove is used for measurement, and under high flow conditions, it automatically switches to a rectangular section for measurement.

[0013] Further, in step S3, the multi-parameter coupling analysis includes: The discrete seepage pressure data is converted into a two-dimensional hydraulic head contour map of the dam body using the Kriging interpolation algorithm, dynamically displaying changes in the dam body's phreatic line; and A spatiotemporal prediction model for seepage parameters is established based on LSTM, and the measured values ​​are monitored in real time to see if they deviate from the predicted values ​​by 15%. If they do, root cause analysis is initiated.

[0014] Furthermore, in step S4, the graded early warning strategy is set to multiple colors, wherein when a specific color warning is triggered, a preset emergency plan is simultaneously activated and the coordinates of the abnormal seepage area are automatically located.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention addresses the problems of traditional single-point piezometers failing to reflect the overall seepage field and being prone to siltation by using a three-dimensional monitoring grid of a distributed seepage monitoring array and a novel nanofiltration layer. Combined with the dual-weir structure of an intelligent water-measuring weir group and an ultrasonic silt detection module, it avoids the accuracy deviations caused by siltation in traditional water-measuring weirs. Edge computing nodes establish a three-dimensional correlation model of seepage pressure, flow rate, and time using an embedded multi-parameter coupled analysis algorithm, automatically identifying abnormal seepage patterns and solving the problem of isolated monitoring data failing to automatically identify dam hazards. The adaptive power supply system's dual-power design ensures uninterrupted monitoring even during continuous rainy weather, while the lightning-proof and interference-resistant transmission network's dual-path communication redundancy design reduces the risk of data loss. Overall, it achieves comprehensive automated monitoring of dam seepage, improving data acquisition accuracy and real-time performance, triggering timely warnings, and providing reliable support for the safe operation of the dam. Attached Figure Description

[0016] Figure 1 This is a system architecture diagram of the present invention; Figure 2 This is a schematic diagram of the deployment of the seepage pressure monitoring array of the present invention; Figure 3 This is a schematic diagram illustrating the working principle of the intelligent water measuring weir of the present invention. Figure 4 This is a flowchart of the edge computing analysis process of the present invention; Figure 5 This is a diagram of the graded early warning mechanism of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1-5 This invention provides a seepage monitoring system for reservoir safety monitoring, comprising: The distributed piezometric monitoring array has multiple high-precision vibrating wire piezometers arranged in each monitoring cross section of the reservoir dam to form a three-dimensional monitoring grid. The piezometers are protected by a new type of nano-filter layer, which is used to improve the piezometers' resistance to siltation. The intelligent water measuring weir group includes a double weir structure consisting of a main weir and an auxiliary weir. The main weir adopts a composite weir opening of V-shape and rectangle. The composite weir opening can automatically switch the measurement mode according to the magnitude of the seepage flow. The auxiliary weir is equipped with an ultrasonic silt detection module, which is used to detect the silt thickness in real time and calibrate the zero point of the water measuring weir group. Edge computing nodes, which are embedded with multi-parameter coupled analysis algorithms, are used to establish a three-dimensional correlation model of seepage pressure, seepage flow rate and time, and identify preset abnormal seepage patterns based on the model. Abnormal seepage patterns include high pressure differential and low flow rate or low pressure differential and high flow rate, and automatically trigger graded early warnings. The adaptive power supply system integrates a thermoelectric generator module and a solar panel as dual power sources, enabling it to maintain uninterrupted monitoring for 30 days during continuous rainy weather; and The lightning-proof and interference-resistant transmission network adopts a hybrid networking of shielded twisted-pair cables and LoRa wireless. The transmission network has a communication redundancy design, which ensures dual-path transmission of data packets.

[0019] Specifically, in reservoir dam seepage monitoring, distributed piezometric monitoring arrays are preferentially deployed at the maximum dam height of earth-rock dams, areas of abnormal seepage, and additional sections when the dam length exceeds 300m. Each monitoring cross section is equipped with multiple high-precision vibrating wire piezometers according to actual seepage monitoring needs. The piezometers are wrapped with a new type of nano-filter layer and are used in conjunction with piezometric tubes during installation. The piezometric tubes are made of galvanized steel pipes with an inner diameter of no more than 50mm. The tube openings are 10-20cm above the dam surface and equipped with protective devices. During the sealing stage, bentonite mud balls are used for layered backfilling to ensure that the piezometers stably collect seepage pressure data at different depths of the dam body, forming a three-dimensional monitoring grid and solving the problem that traditional single-point piezometers cannot reflect the overall seepage field distribution.

[0020] The intelligent water measuring weir group is set in the seepage collection ditch downstream of the dam's drainage body. The main weir adopts a composite weir structure of V-shape and rectangle, with a total weir channel length of not less than 2.0m. The weir plate is made of stainless steel and is kept vertical, perpendicular to the side wall of the weir channel. The secondary weir integrates an ultrasonic silt detection module to scan the silt accumulation in the weir channel in real time. Based on the detected silt thickness, it automatically calibrates the measurement zero point of the water measuring weir group, avoiding the accuracy deviation caused by silt accumulation in traditional water measuring weirs.

[0021] Edge computing nodes are deployed in observation boxes downstream of the dam, accessing seepage pressure data from the distributed seepage pressure monitoring array and seepage flow data from the intelligent water measurement weir group. A three-dimensional correlation model of seepage pressure, seepage flow, and time is established using an embedded multi-parameter coupled analysis algorithm. To simplify the correlation and ensure dimensional consistency, a simplified multivariate linear model is used to construct the correlation, as shown in the following formula:

[0022] In the formula: The seepage flow rate is expressed in cubic meters per second (m³ / s). This is osmotic pressure, with the dimension of kilopascal (kPa). The monitoring time is measured in hours (h) and is calculated cumulatively from the start time of monitoring as 0. The coefficient representing the influence of seepage pressure on seepage flow rate is expressed in cubic meters per second (kPa) [m³ / (s·kPa)], and is obtained by fitting historical monitoring data. The coefficient representing the influence of time on seepage flow is expressed in cubic meters per second per hour [m³ / (s·h)], and is obtained by fitting historical monitoring data. This is a constant term with dimensions in cubic meters per second (m³ / s), obtained by fitting historical monitoring data.

[0023] Edge computing nodes utilize the seepage field reconstruction function in the algorithm (based on the Kriging interpolation algorithm, a mature existing algorithm, which will not be elaborated here) to present the seepage distribution of the dam body. Combined with the abnormal pattern recognition function, it captures abnormal seepage patterns such as high pressure difference with small flow and low pressure difference with large flow, solving the problem of isolated traditional monitoring data and inability to automatically identify dam body hazards.

[0024] The adaptive power supply system installs solar panels on the dam crest or near the observation box, and deploys thermoelectric power generation modules in areas with significant temperature differences between the dam body and the foundation. The dual power sources are connected to the entire monitoring system through a switching device. In daily operation, solar power is used first to charge the battery. During continuous cloudy and rainy weather, the system automatically switches to thermoelectric power generation and battery power supply modes. With the energy storage capacity of the solar panels, the stable output of the thermoelectric power generation modules, and the backup capacity of the battery, at least 30 days of uninterrupted monitoring can be guaranteed. Even in the event of extreme continuous cloudy and rainy weather exceeding 30 days, the dual power sources can work together to maintain the operation of core monitoring functions, avoiding monitoring interruptions caused by weather conditions in traditional single power supply methods.

[0025] The lightning protection and anti-interference transmission network uses shielded twisted-pair cables buried along the dam body, and simultaneously establishes a LoRa wireless transmission channel to form a dual-path communication redundancy design. The shielded twisted-pair cables are protected by metal conduits, which are grounded at multiple points with a grounding resistance of less than 5Ω. The LoRa signal coverage covers the entire monitoring area, ensuring that monitoring data is transmitted to the monitoring platform through dual paths, reducing the risk of data loss and improving transmission stability.

[0026] This system enables comprehensive and automated monitoring of seepage in reservoir dams, improving the accuracy and real-time nature of seepage data acquisition, timely identifying seepage anomalies and triggering early warnings, and providing reliable support for the safe operation of dams.

[0027] In this embodiment, each monitoring cross section of the distributed osmotic pressure monitoring array is equipped with 3 to 5 high-precision vibrating wire osmometers, with a range of 0 to 700 kPa and a resolution of less than or equal to 0.02%FS.

[0028] Specifically, the selection of monitoring cross sections for distributed piezometric monitoring arrays should be based on the dam structure and seepage characteristics. Priority should be given to deploying them at the maximum dam height section of earth-rock dams, sections with complex dam foundation geology, and sections with historical seepage anomalies. Each cross section should be equipped with 3 to 5 high-precision vibrating wire piezometers. The piezometers should be steel wire type with a range of 0 to 700 kPa and a resolution of less than or equal to 0.02%FS. During installation, they should be distributed in the piezometer tubes at different depths. The borehole inclination of the piezometer tubes should not exceed 3°, and the hole depth should be about 50 cm below the designed measuring point elevation. After drilling, a coarse sand cushion layer should be filled in before lowering the piezometers to ensure that the piezometers can accurately collect seepage pressure data at different depths.

[0029] The deployment of multiple piezometers can form a denser network of monitoring points, more comprehensively reflecting the distribution of seepage pressure in the lateral and longitudinal directions of the dam body. This avoids the limitations of traditional single-point monitoring, making subsequent seepage field analysis more consistent with the actual seepage state of the dam body and improving the representativeness and reliability of seepage pressure data.

[0030] In this embodiment, the main weir of the intelligent water measuring weir group uses a V-shaped groove for measurement under low flow conditions and automatically switches to a rectangular section for measurement under high flow conditions. The automatic switching of the measurement mode ensures that the measurement error of the water measuring weir group is controlled within 5% across the entire range.

[0031] Specifically, during the installation of the main weir of the intelligent water measurement weir group, it is necessary to ensure a smooth connection between the V-shaped channel and the rectangular section, and that the elevation of the weir opening is not higher than the bottom elevation of the drainage ditch at the dam toe. When the seepage flow is small, the water mainly flows through the V-shaped channel and through the main weir, utilizing the high measurement accuracy of the V-shaped weir at small flow rates to achieve precise measurement; when the seepage flow increases to a set threshold, the main weir automatically switches to the rectangular section for measurement, leveraging the rectangular weir's ability to adapt to large flow rates to ensure measurement continuity.

[0032] The ultrasonic silt detection module of the auxiliary weir is installed on the upstream side of the main weir. It emits ultrasonic waves in real time to detect the thickness of silt at the bottom of the weir channel. The detection data is transmitted to the control unit of the measuring weir group. The control unit automatically adjusts the measurement zero point according to the silt thickness to offset the influence of silt accumulation on the seepage flow calculation. Through this automatic switching and zero-point calibration design, the measurement error of the measuring weir group can be controlled within 5% across the entire measurement range, ensuring the accuracy of seepage flow data under different flow conditions.

[0033] In this embodiment, the multi-parameter coupling analysis algorithm embedded in the edge computing node includes: The seepage field reconstruction module is used to convert discrete seepage pressure data collected by the distributed seepage pressure monitoring array into a two-dimensional hydraulic head contour map of the dam body using the Kriging interpolation algorithm; and The abnormal pattern recognition module is used to build a spatiotemporal prediction model for seepage parameters based on LSTM. When the measured value of seepage pressure or seepage flow deviates from the predicted value by 15%, root cause analysis is initiated.

[0034] Specifically, after acquiring discrete seepage pressure data from the distributed seepage pressure monitoring array, the seepage field reconstruction module of the edge computing node uses the Kriging interpolation algorithm (a mature existing algorithm, which will not be elaborated here) to process the data, transforming the discrete monitoring point data into a continuous two-dimensional head contour map of the dam body. This contour map can dynamically display the changing trend of the dam body's phreatic line, making it easier for monitoring personnel to intuitively grasp the seepage distribution inside the dam body.

[0035] The anomaly pattern recognition module uses LSTM to construct a spatiotemporal prediction model for seepage parameters. The model training data covers historical data on seepage pressure and flow rate during different reservoir operating cycles, such as flood season and non-flood season. In actual monitoring, the model outputs predicted values ​​of seepage pressure and flow rate in real time and compares them with measured values. The deviation rate is used to determine whether root cause analysis is triggered. The deviation rate is calculated using the following formula:

[0036] In the formula: The deviation rate is dimensionless (expressed as a percentage). Reality The measured values ​​are the seepage pressure or seepage flow rate (the dimension of seepage pressure is kPa, and the dimension of seepage flow rate is m³ / s). Pre This is a predicted value for seepage pressure or seepage flow (with dimensions consistent with measured values ​​to ensure uniformity of calculation dimensions).

[0037] when When the rate reaches 15%, the edge computing node automatically initiates the root cause analysis process to investigate the causes of anomalies, such as piezometer failure, local leakage in the dam body, and siltation in the measuring weir, providing direction for subsequent treatment measures.

[0038] By applying this algorithm, in-depth analysis of seepage data can be achieved. It can not only dynamically present the seepage field status, but also quickly identify anomalies and trace their causes, thereby improving the efficiency of handling dam seepage anomalies.

[0039] In this embodiment, the communication redundancy design of the lightning protection and anti-interference transmission network includes: shielded twisted pair cable and LoRa wireless hybrid networking to form a dual-path transmission channel. The dual-path transmission channel achieves a packet loss rate of <0.1% through a consistency verification mechanism for simultaneously transmitting data packets and monitoring and control centers.

[0040] Specifically, the shielded twisted-pair cable of the lightning protection and anti-interference transmission network is selected from models that meet the standards for water conservancy monitoring. The cable is laid in a cable trench excavated along the dam body. Metal protective pipes are laid in the cable trench. Flat iron is welded between the metal pipes for conduction and multiple grounding points are used, and the grounding resistance is less than 5Ω. The LoRa wireless transmission module is installed on the signal tower on the top of the observation box. The signal coverage covers all monitoring equipment, forming a hybrid network that combines wired and wireless technologies.

[0041] The communication redundancy design is achieved through dual-path transmission. Data from each monitoring device is simultaneously transmitted to the monitoring and control center via shielded twisted-pair cable and LoRa wireless channel. The control center has a built-in data verification module that pairs and verifies the data packets received from both paths according to "device ID + acquisition timestamp": if both data packets pass verification and the parameter deviation is ≤0.5%, one path is randomly selected for storage; if only a single path receives a valid data packet, the data from that path is stored directly and the other path is marked as abnormal; if the deviation between the two data paths is >0.5%, a retransmission command is sent to the corresponding monitoring device to ensure valid data reception.

[0042] This invention also provides a seepage monitoring method for reservoir safety monitoring, comprising: S1. Seepage pressure data of the reservoir dam body is obtained through a distributed seepage pressure monitoring array. The distributed seepage pressure monitoring array has multiple high-precision vibrating wire piezometers arranged in each monitoring cross section of the reservoir dam body to form a three-dimensional monitoring grid. The piezometers are protected by a new type of nano-filter layer. S2. The seepage flow data of the reservoir dam body is obtained through the intelligent water measuring weir group. The intelligent water measuring weir group includes a main weir and an auxiliary weir double weir structure. The main weir adopts a V-shaped and rectangular composite weir mouth, and the auxiliary weir is equipped with an ultrasonic silt detection module. S3. Based on edge computing nodes, perform multi-parameter coupling analysis on the seepage pressure data obtained in step S1 and the seepage flow data obtained in step S2 to establish a three-dimensional correlation model of seepage pressure, seepage flow and time, and identify preset abnormal seepage patterns, including high pressure differential and low flow or low pressure differential and high flow. S4. Automatically trigger graded early warnings based on the identified abnormal seepage patterns; S5. Implement a self-maintenance mechanism, which includes: the intelligent water measuring weir group automatically triggers the backwashing device to perform high-pressure water gun cleaning based on the sludge thickness detected by the ultrasonic sludge detection module; and the piezometer performs a self-check pulse once a day to remove the deposits on the nanofilter layer.

[0043] Specifically, when conducting seepage monitoring of reservoir dams, seepage pressure data is first obtained through a distributed seepage pressure monitoring array. This array is set up with monitoring cross sections at the maximum dam height section and the seepage anomaly section of the reservoir dam. Multiple high-precision vibrating wire piezometers with novel nano-filter layers are installed in each cross section. The piezometers penetrate into different depths of the dam body through the piezometer tube to continuously collect seepage pressure data from various parts of the dam body.

[0044] Subsequently, seepage flow data is obtained through an intelligent water-measuring weir group. The water-measuring weir group is set up in the seepage collection ditch behind the dam. The main weir automatically switches between the measurement modes of the V-shaped groove and the rectangular section according to the size of the seepage flow. The ultrasonic silt detection module of the secondary weir monitors the silt thickness in real time, providing data support for the zero point calibration of the water-measuring weir group.

[0045] After receiving the seepage pressure and seepage flow data, the edge computing node starts multi-parameter coupling analysis, generates a two-dimensional head contour map of the dam body through the Kriging interpolation algorithm, and at the same time uses the LSTM model combined with formula (1) to establish a three-dimensional correlation model of seepage pressure, seepage flow and time, and identifies abnormal seepage patterns such as high pressure difference and small flow, low pressure difference and large flow; calculates the deviation rate through formula (2) to determine the degree of deviation between the measured value and the predicted value.

[0046] When an abnormal seepage pattern is detected, the system automatically triggers a tiered early warning system, pushing corresponding warning information to the monitoring platform and management personnel terminals according to the severity of the anomaly. At the same time, a self-maintenance mechanism is implemented. The intelligent water measuring weir group starts the backwashing device based on the silt thickness detected by ultrasonic waves, and cleans the silt in the weir channel with a high-pressure water gun. The piezometer performs a self-check pulse once a day in the early morning (at this time the seepage state of the dam body is relatively stable, so as to avoid interference with the monitoring data by the self-check), and removes the attachments on the surface of the nanofilter layer to ensure the continuous and stable operation of the monitoring equipment.

[0047] This method automates the entire seepage monitoring process, taking into account data acquisition, analysis, early warning, and equipment maintenance, thereby improving the efficiency and reliability of dam seepage monitoring.

[0048] In this embodiment, in step S1, the piezometer has a range of 0 to 700 kPa and a resolution of ≤0.02%FS.

[0049] Specifically, the high-precision vibrating wire piezometer selected in step S1 must be adapted to the actual seepage pressure monitoring requirements of the reservoir dam: considering that the seepage pressure values ​​from the dam crest to the dam foundation at different depths of the earth-rock dam body usually fluctuate within the range of 0 to 600 kPa, in order to reserve a safety monitoring margin of 100 kPa, a vibrating wire piezometer with a range of 0 to 700 kPa is selected to avoid the seepage pressure exceeding the range due to insufficient range, or the measurement error of small seepage pressure values ​​being increased due to excessive range; at the same time, since abnormal seepage in the dam body is often accompanied by minute seepage pressure changes at the level of 0.1 kPa, conventional piezometers are difficult to capture accurately, so a model with a resolution of ≤0.02%FS is selected to ensure that early seepage hazards in the dam body can be identified.

[0050] Before installation, the piezometer should be soaked in clean water for more than 2 hours to remove air bubbles from the permeable stone. After the sensitivity test of the pressure measuring tube is qualified, the piezometer is suspended to the design elevation, and the elevation deviation is controlled within ±5cm. The cable is led along the dam slope to the observation box and covered with a PVC protective pipe to avoid aging of the cable due to sun and rain.

[0051] The selection of this measurement range and resolution is based on conventional adaptation and optimization of the seepage pressure distribution law of the reservoir dam body. It can ensure that the piezometer can stably collect data under different dam depths and different operating conditions, providing accurate basic data for subsequent seepage field analysis and anomaly identification. At the same time, the new nano-filter layer on the outside of the piezometer reduces the interference of siltation on data acquisition, and the adapted measurement range and resolution ensure the accuracy of the data itself, jointly solving the problem of "inaccurate data and inability to reflect subtle anomalies" in traditional seepage pressure monitoring.

[0052] In this embodiment, in step S2, the main weir can automatically switch the measurement mode according to the magnitude of the seepage flow. Under low flow conditions, a V-shaped groove is used for measurement, and under high flow conditions, it automatically switches to a rectangular section for measurement.

[0053] Specifically, in step S2, the main weir of the intelligent water measuring weir group must be installed with a smooth transition at the junction of the V-shaped channel and the rectangular section. A flow threshold sensor is set up to monitor the seepage flow through the main weir in real time. When the seepage flow is less than the threshold, the main weir only uses the V-shaped channel for measurement, utilizing the high sensitivity of the V-shaped weir to small flow rates for accurate measurement. When the seepage flow exceeds the threshold, the main weir control mechanism automatically opens the rectangular section channel, switches to the rectangular section measurement mode, and simultaneously closes the main water flow channel of the V-shaped channel to ensure the stability of seepage flow measurement under high flow rates.

[0054] Through this automatic switching design, the main weir can adapt to the changes in seepage flow during different operating stages of the reservoir. Whether it is the large seepage flow during the flood season or the small seepage flow during the non-flood season, it can meet the seepage flow parameters in formula (1). The accurate input and the measured value of seepage flow in formula (2) Reliable calculations provide data support, ensuring the continuity and accuracy of seepage flow data.

[0055] In this embodiment, step S3, the multi-parameter coupling analysis includes: Discrete seepage pressure data are transformed into two-dimensional hydraulic head contour maps of the dam body using the Kriging interpolation algorithm, dynamically displaying changes in the dam's phreatic lines; and A spatiotemporal prediction model for seepage parameters is established based on LSTM, and the measured values ​​are monitored in real time to see if they deviate from the predicted values ​​by 15%. If they do, root cause analysis is initiated.

[0056] Specifically, in the multi-parameter coupling analysis of step S3, the discrete seepage pressure data obtained in step S1 is first preprocessed. After removing outliers, the data is converted into a two-dimensional head contour map of the dam body using the Kriging interpolation algorithm (a mature existing algorithm, which will not be elaborated here). This map can dynamically display the position changes of the dam body seepage line and intuitively determine whether there is a concentrated seepage area in the dam body.

[0057] Simultaneously, a spatiotemporal prediction model for seepage parameters is constructed based on LSTM. The model input parameters include historical seepage pressure data, seepage flow data, and reservoir operation parameters such as reservoir water level and rainfall. The parameters in formula (1) are determined through model training. , , The system establishes the correlation between seepage pressure, seepage flow rate and time, and outputs predicted values ​​in real time. Substitute the measured values ​​and predicted values ​​obtained in steps S1 and S2 into formula (2) (which is consistent with the deviation rate formula and will not be repeated here) to calculate the deviation rate. When the deviation rate reaches 15%, the system automatically starts root cause analysis to investigate whether the abnormality is caused by changes in the dam structure, monitoring equipment failure or external environmental influences, providing a basis for subsequent handling.

[0058] This analysis process enables in-depth mining of seepage data, not only presenting the dynamic changes in the seepage field, but also timely locating the causes of anomalies, thus improving the level of intelligence in seepage monitoring.

[0059] In this embodiment, in step S4, the graded early warning strategy is set to multiple colors. When a specific color warning is triggered, a preset emergency plan is simultaneously activated and the coordinates of the abnormal seepage area are automatically located.

[0060] Specifically, the graded early warning strategy in step S4 uses multiple colors, specifically four levels of warning: blue, yellow, orange, and red. Each level corresponds to a different severity of seepage anomalies. When the system calculates a small deviation rate using formula (2) and identifies a minor anomaly such as a slight fluctuation in seepage pressure, a blue warning is triggered to remind monitoring personnel to pay closer attention. When the deviation rate increases and the anomaly worsens, such as a slow increase in seepage flow, a yellow warning is triggered, and the frequency of monitoring is increased. When the deviation rate approaches 15% and the anomaly approaches a dangerous threshold, such as a rapid rise in seepage pressure, an orange warning is triggered, and personnel are organized to conduct on-site investigations. When the deviation rate reaches or exceeds 15% and a serious anomaly occurs, such as a high pressure differential and low flow rate that continues to deteriorate, a red warning is triggered. The system simultaneously activates a preset emergency plan, such as shutting down relevant drainage facilities and notifying downstream personnel to prepare for prevention. At the same time, the coordinates of the seepage anomaly area are automatically located using the measurement point location information of the distributed seepage pressure monitoring array, providing accurate guidance for on-site handling.

[0061] By implementing tiered early warning and emergency response coordination, targeted measures can be taken based on the severity of the anomaly, effectively reducing the risk of dam seepage accidents and ensuring the safe operation of the dam.

[0062] In summary, this invention addresses the problems of traditional single-point piezometers failing to reflect the overall seepage field and being prone to siltation by utilizing a three-dimensional monitoring grid of a distributed seepage monitoring array and a novel nanofiltration layer. Combined with the dual-weir structure of the intelligent water-measuring weir group and the ultrasonic silt detection module, it avoids the accuracy deviations caused by siltation in traditional water-measuring weirs. Edge computing nodes establish a three-dimensional correlation model between seepage pressure, flow rate, and time using an embedded multi-parameter coupled analysis algorithm, automatically identifying abnormal seepage patterns and solving the problem of isolated monitoring data failing to automatically identify dam hazards. The adaptive power supply system's dual-power design ensures uninterrupted monitoring even during continuous rainy weather, while the lightning protection and anti-interference transmission network's dual-path communication redundancy design reduces the risk of data loss. Overall, this invention achieves comprehensive automated monitoring of dam seepage, improving data acquisition accuracy and real-time performance, triggering timely warnings, and providing reliable support for the safe operation of the dam.

[0063] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0064] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A seepage monitoring system for reservoir safety monitoring, characterized in that, include: A distributed piezometric monitoring array is provided, in which multiple high-precision vibrating wire piezometers are arranged on each monitoring cross section of the reservoir dam to form a three-dimensional monitoring grid. The piezometers are protected by a novel nano-filter layer, which is used to improve the piezometers' resistance to siltation. The intelligent water measuring weir group includes a main weir and an auxiliary weir double weir structure. The main weir adopts a V-shaped and rectangular composite weir opening. The composite weir opening can automatically switch the measurement mode according to the magnitude of the seepage flow. The auxiliary weir is equipped with an ultrasonic silt detection module. The ultrasonic silt detection module is used to detect the silt thickness in real time and calibrate the zero point of the water measuring weir group. An edge computing node is embedded with a multi-parameter coupled analysis algorithm. The edge computing node is used to establish a three-dimensional correlation model of seepage pressure, seepage flow rate and time, and to identify preset abnormal seepage patterns based on the model. The abnormal seepage patterns include high pressure differential and low flow rate or low pressure differential and high flow rate, and automatically trigger graded early warning. An adaptive power supply system, which integrates a thermoelectric power generation module and a solar panel dual power supply, can maintain uninterrupted monitoring for at least 30 days under continuous rainy weather. as well as The lightning protection and anti-interference transmission network adopts a hybrid networking of shielded twisted pair cable and LoRa wireless. The transmission network has a communication redundancy design, which ensures dual-path transmission of data packets.

2. The seepage monitoring system for reservoir safety monitoring as claimed in claim 1 wherein, Each monitoring cross section of the distributed osmotic pressure monitoring array is equipped with 3 to 5 of the high-precision vibrating wire osmometers, the osmometers having a range of 0 to 700 kPa and a resolution less than or equal to 0.02%FS.

3. The seepage monitoring system for reservoir safety monitoring as claimed in claim 1 wherein, The main weir of the intelligent water measuring weir group uses a V-shaped groove for measurement under low flow conditions and automatically switches to a rectangular section for measurement under high flow conditions. The automatic switching of the measurement mode ensures that the measurement error of the water measuring weir group is controlled within 5% across the entire range.

4. The seepage monitoring system for reservoir safety monitoring as claimed in claim 1 wherein, The multi-parameter coupling analysis algorithm embedded in the edge computing node includes: The seepage field reconstruction module is used to convert the discrete seepage pressure data collected by the distributed seepage pressure monitoring array into a two-dimensional head contour map of the dam body using the Kriging interpolation algorithm; and the anomaly pattern recognition module is used to establish a spatiotemporal prediction model of seepage parameters based on LSTM, and to initiate root cause analysis when the measured value of the seepage pressure data or seepage flow deviates from the predicted value by 15%.

5. The seepage monitoring system for reservoir safety monitoring as claimed in claim 1 wherein, The communication redundancy design of the lightning protection and anti-interference transmission network includes: shielded twisted pair cable and LoRa wireless hybrid networking to form a dual-path transmission channel. The dual-path transmission channel achieves a packet loss rate of <0.1% through a consistency verification mechanism for simultaneously transmitting data packets and monitoring and control centers.

6. The seepage monitoring method for reservoir safety monitoring, applied to the seepage monitoring system for reservoir safety monitoring in any one of claims 1-5, characterized in that, include: S1. Seepage pressure data of the reservoir dam body is obtained through a distributed seepage pressure monitoring array. The distributed seepage pressure monitoring array has multiple high-precision vibrating wire seepage gauges arranged in each monitoring cross section of the reservoir dam body to form a three-dimensional monitoring grid. The seepage gauges are protected by a new type of nano-filter layer. S2. Obtain seepage flow data of the reservoir dam body through intelligent water measuring weir group. The intelligent water measuring weir group includes a main weir and an auxiliary weir double weir structure. The main weir adopts a V-shaped and rectangular composite weir opening. The auxiliary weir is equipped with an ultrasonic silt detection module. S3. Based on edge computing nodes, perform multi-parameter coupling analysis on the seepage pressure data obtained in step S1 and the seepage flow data obtained in step S2 to establish a three-dimensional correlation model of seepage pressure, seepage flow and time, and identify preset abnormal seepage patterns, including high pressure difference and low flow or low pressure difference and high flow. S4. Automatically trigger graded early warnings based on the identified abnormal seepage patterns; S5. Execute the self-maintenance mechanism, which includes: the intelligent water measuring weir group automatically triggers the backwashing device to perform high-pressure water gun cleaning based on the silt thickness detected by the ultrasonic silt detection module; The osmometer performs a self-test pulse once a day to remove deposits on the nanofilter layer.

7. A seepage monitoring method for reservoir safety monitoring according to claim 6, characterized in that, In step S1, the osmotic pressure gauge has a range of 0 to 700 kPa and a resolution of ≤0.02%FS.

8. A seepage monitoring method for reservoir safety monitoring according to claim 6, characterized in that, In step S2, the main weir can automatically switch the measurement mode according to the magnitude of the seepage flow. Under low flow conditions, a V-shaped groove is used for measurement, and under high flow conditions, it automatically switches to a rectangular section for measurement.

9. A seepage monitoring method for reservoir safety monitoring according to claim 6, characterized in that, In step S3, the multi-parameter coupling analysis includes: The discrete seepage pressure data is converted into a two-dimensional hydraulic head contour map of the dam body using the Kriging interpolation algorithm, dynamically displaying changes in the dam body's phreatic line; and A spatiotemporal prediction model for seepage parameters is established based on LSTM, and the measured values ​​are monitored in real time to see if they deviate from the predicted values ​​by 15%. If they do, root cause analysis is initiated.

10. A seepage monitoring method for reservoir safety monitoring according to claim 6, characterized in that, In step S4, the graded early warning strategy is set to multiple colors. When a specific color warning is triggered, a preset emergency plan is simultaneously activated and the coordinates of the abnormal seepage area are automatically located.

Images