Tailings pond groundwater real-time monitoring method and system based on internet of things

Abstract

The present application relates to the field of groundwater data monitoring, and more particularly to a tailing pond groundwater real-time monitoring method and system based on the Internet of Things, which comprises: real-time acquisition of groundwater parameters in each area; determination of abnormal node clusters; determination of pollution plume diffusion path; determination of seepage anomaly type; issuance of early warning for sustained enhanced anomaly; and adjustment of threshold value according to stable seepage type anomaly. The present application achieves accurate early warning of abnormal seepage by arranging multi-source monitoring nodes in the tailing pond area, the dam front area, the dam rear area and the seepage zone, comprehensively collecting key hydrogeological parameters, classifying and identifying different types of seepage anomalies, dynamically correcting the threshold value based on the spatial distribution of stable seepage type node clusters, and making the monitoring system self-adapt to the changing groundwater environment, effectively solving the problem of difficult timely discovery of complex abnormal seepage due to single monitoring parameter, data update lag and insufficient anomaly recognition capability.

Description

Technical Field

[0001] This invention relates to the field of groundwater data monitoring, and in particular to a method and system for real-time monitoring of groundwater in tailings ponds based on the Internet of Things. Background Technology

[0002] With the continuous expansion of mining scale and the continuous increase in tailings accumulation, the safety management of tailings dams faces increasingly complex hydrogeological conditions and environmental risks. Changes in parameters such as groundwater level, pore water pressure, tailings slurry temperature, and conductivity in the reservoir area, the area in front of the dam, the area behind the dam, and the seepage zone are closely related and directly affect the stability of the tailings dam and the safety of the surrounding ecology. However, traditional monitoring methods mostly rely on manual periodic observation, which makes it difficult to achieve real-time and continuous perception of abnormal seepage and pollution plumes. As a result, it is difficult to provide timely warnings in the event of sudden seepage or dam abnormalities, which increases the challenges of tailings dam management and emergency response.

[0003] Chinese Patent Application Publication No. CN111486926A discloses a dynamic monitoring system and method for determining the reverse osmosis water level in a deep-type wet tailings dam. The system includes multiple water level gauges evenly distributed around the tailings dam. The water level gauges sequentially transmit information to a remote data processing terminal via wireless output. The method involves using GPS positioning to evenly position the water level gauges around the tailings dam, generating a water level change map based on the gauge readings, and determining the lowest reverse osmosis point based on the range of water level fluctuations. The difference between the readings of the groundwater level gauges above the tailings dam and the liquid level reading in the tailings dam is then compared with the magnitude of the lowest reverse osmosis point to ensure that the tailings liquid does not exceed the lowest groundwater level.

[0004] Therefore, the existing dynamic monitoring system and method for determining the reverse seepage water level of deep-type wet tailings ponds has the following problems: the existing technology only monitors the single physical parameter of groundwater level, tailings pond leakage is often accompanied by the migration of pollutants, and water level changes cannot directly reflect changes in the chemical properties of water quality or the thermodynamic characteristics of pollution sources; the existing technology mainly relies on the periodic reading of water level gauges, and the information is transmitted wirelessly to a remote end for processing, lacking the ability to respond quickly to instantaneous or short-term anomalies; the existing technology is mainly based on the comparison of water level difference with the lowest reverse seepage point, and cannot identify complex seepage patterns. Summary of the Invention

[0005] To address this, the present invention provides a method and system for real-time monitoring of groundwater in tailings ponds based on the Internet of Things. This method uses multi-parameter sensors to collect groundwater information in the pond and dam areas in real time, and combines anomaly analysis and dynamic threshold adjustment to overcome the problems of existing technologies, such as difficulty in timely detection of complex abnormal seepage due to single monitoring parameters, delayed data updates, and insufficient anomaly identification capabilities.

[0006] To achieve the above objectives, on the one hand, the present invention provides a method for real-time monitoring of groundwater in tailings ponds based on the Internet of Things, comprising:

[0007] Real-time acquisition of tailings slurry temperature and pore water pressure at each monitoring node within the tailings dam area, groundwater level and temperature at each monitoring node in the front dam area, groundwater level and conductivity at each monitoring node in the rear dam area, and seepage rate at each monitoring node within the seepage zone.

[0008] Several abnormal node clusters are determined based on the groundwater level in front of the dam, the preset groundwater level threshold in front of the dam, the groundwater level behind the dam, the preset groundwater level threshold behind the dam, the pore water pressure, and the preset pressure threshold.

[0009] The pollution plume diffusion path is determined based on the variation characteristics of the conductivity, tailings slurry temperature, and water temperature in each of the aforementioned abnormal node clusters.

[0010] The seepage anomaly type of each abnormal node cluster is determined based on the variation characteristics of the seepage rate in the pollution plume diffusion path within a preset observation window.

[0011] An emergency early warning is generated for the cluster of abnormal nodes whose seepage anomalies are determined to be of the continuously enhancing type.

[0012] The preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold is adjusted based on the location distribution of the abnormal node clusters that are determined to be of the stable seepage type within the preset adjustment period.

[0013] Furthermore, the process of determining several abnormal node clusters based on the groundwater level in front of the dam, the preset groundwater level threshold in front of the dam, the groundwater level behind the dam, the preset groundwater level threshold behind the dam, the pore water pressure, and the preset pressure threshold includes:

[0014] The monitoring node whose groundwater level in front of the dam is greater than the preset groundwater level threshold in front of the dam is determined to be an abnormal groundwater level in front of the dam and is marked as an abnormal node in front of the dam.

[0015] The monitoring node whose groundwater level behind the dam is greater than the preset groundwater level threshold behind the dam is determined to be an abnormal groundwater level behind the dam and is marked as an abnormal node behind the dam.

[0016] The node type of the monitoring node whose pore water pressure is greater than the preset pressure threshold is determined to be pore water pressure abnormal, and it is marked as a pressure abnormal node.

[0017] The location coordinates of the abnormal nodes in front of the dam, the abnormal water level nodes, and the abnormal pressure nodes are obtained, and spatial density clustering is performed according to a preset clustering algorithm to obtain several spatial clusters.

[0018] Several abnormal node clusters are determined based on the node type of the monitoring nodes in each of the spatial clusters.

[0019] Furthermore, the process of determining several abnormal node clusters based on the node types of the monitoring nodes in each of the spatial clusters includes:

[0020] When there are more than or equal to the number of node types of a preset type in the spatial cluster, the spatial cluster is determined to be the abnormal node cluster, so as to identify several abnormal node clusters.

[0021] Furthermore, the process of determining the pollution plume diffusion path based on the variation characteristics of the conductivity, tailings slurry temperature, and water temperature in each of the aforementioned abnormal node clusters includes:

[0022] Obtain the location coordinates, conductivity, tailings slurry temperature, and water temperature of all monitoring nodes within each cluster of abnormal nodes and within the surrounding preset path determination range;

[0023] Based on the location coordinates of the monitoring nodes, spatial distribution fields of the electrical conductivity, tailings slurry temperature, and water temperature are constructed respectively to obtain the electrical conductivity distribution field, slurry temperature distribution field, and water temperature distribution field.

[0024] Calculate the spatial cross-correlation coefficient between the conductivity distribution field and the slurry temperature distribution field to obtain the first spatial correlation coefficient;

[0025] Calculate the spatial cross-correlation coefficient between the conductivity distribution field and the water temperature distribution field to obtain the second spatial correlation coefficient;

[0026] When the first spatial correlation coefficient and / or the second spatial correlation coefficient are greater than a preset correlation threshold, the region where the abnormal node cluster is located is determined to be a candidate abnormal region, and local spatial autocorrelation analysis is performed in the candidate abnormal region to determine the collaborative abnormal region.

[0027] The diffusion path of the pollution plume is determined based on the location coordinates of all the monitoring nodes in the coordinated anomaly zone and the conductivity.

[0028] Furthermore, the process of performing local spatial autocorrelation analysis within candidate anomaly regions to determine co-anomaly regions includes:

[0029] Based on the location coordinates of all monitoring nodes within the candidate anomaly region, a spatial weight matrix is ​​constructed;

[0030] The local spatial autocorrelation statistical value and the standardized Z-score corresponding to the local spatial autocorrelation statistical value of each monitoring node in the conductivity distribution field under the spatial weight matrix are calculated according to the preset statistical algorithm.

[0031] The monitoring nodes whose standardized Z-score is greater than a preset first significance threshold are determined as conductivity significance hotspots, so as to identify a number of conductivity significance hotspots;

[0032] The largest connected region formed in space by all the aforementioned conductivity hotspots is determined to be the core anomalous region.

[0033] A spatial buffer analysis is performed by shifting the outer boundary of the core anomaly region outward by a preset distance. All monitoring nodes covered by the buffer and located within the candidate anomaly region are included in the core anomaly region to determine the collaborative anomaly region.

[0034] Furthermore, the process of determining the pollution plume diffusion path based on the location coordinates of all the monitoring nodes in the coordinated anomaly zone and the conductivity includes:

[0035] The monitoring node corresponding to the maximum conductivity value of all the aforementioned components within the coordinated anomaly zone is identified as the primary pollution source.

[0036] The monitoring node corresponding to the minimum conductivity of all the components at the downstream boundary of the coordinated anomaly zone is determined as the main pollution sink.

[0037] Based on the location coordinates of all the monitoring nodes in the collaborative anomaly zone and the conductivity, a kriging spatial interpolation method is used to generate a conductivity spatial distribution raster map.

[0038] Calculate the spatial gradient field of the conductivity spatial distribution grid, wherein the spatial gradient field characterizes the direction and rate of change of the conductivity at each point in space;

[0039] A particle tracking algorithm is used to release several virtual particles from the main pollution source.

[0040] The virtual particle is made to move along the negative gradient direction of the conductivity concentration in the spatial gradient field to obtain several movement trajectories.

[0041] All the movement trajectories converging towards the main pollution sink point are determined to be candidate diffusion paths;

[0042] The candidate diffusion paths that are spatially adjacent and oriented in the same direction are fused to determine the pollution plume diffusion path.

[0043] Furthermore, the process of determining the seepage anomaly type of each anomalous node cluster based on the variation characteristics of the seepage rate within a preset observation window in the pollution plume diffusion path includes:

[0044] Calculate the cumulative duration and maximum number of consecutive periods during which the seepage rate is greater than a preset seepage rate threshold for all monitoring nodes along the pollution plume diffusion path within the preset observation window;

[0045] Calculate the standard deviation of the rate of change of the seepage rate at all monitoring nodes along the pollution plume diffusion path within the preset observation window to obtain the change fluctuation value;

[0046] The long-term trend of seepage rate at each monitoring node along the pollution plume diffusion path was analyzed using the Mann-Kendall trend test method to obtain the trend test statistic.

[0047] The seepage anomaly type of each abnormal node cluster is determined based on the cumulative duration, the maximum number of consecutive time periods, the change fluctuation value, and the trend test statistic.

[0048] Furthermore, the process of determining the seepage anomaly type of each of the abnormal node clusters based on the cumulative duration, the maximum number of consecutive time periods, the change fluctuation value, and the trend test statistic includes:

[0049] When the cumulative duration is greater than a preset first persistence threshold, the maximum number of consecutive time periods is greater than a preset second persistence threshold, the change fluctuation value is less than a preset first fluctuation threshold, and the absolute value of the trend test statistic is less than a preset first trend threshold, the seepage anomaly type of the monitoring node is determined to be stable seepage type.

[0050] When the cumulative duration is greater than a preset third persistence threshold, the maximum number of consecutive time periods is greater than a preset fourth persistence threshold, and the trend test statistic is greater than a preset second trend threshold, the seepage anomaly type of the monitoring node is determined to be continuously enhanced.

[0051] When the cumulative duration is less than a preset fifth persistence threshold, the change fluctuation value is greater than a preset second fluctuation threshold, and the absolute value of the trend test statistic is less than the preset first trend threshold, the seepage anomaly type of the monitoring node is determined to be intermittent fluctuation type.

[0052] The seepage anomaly type of the abnormal node cluster is determined by the maximum number of monitoring nodes corresponding to the continuously enhanced type, the intermittent fluctuation type, and the stable seepage type in the abnormal node cluster.

[0053] Furthermore, the process of adjusting the preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold based on the location distribution of the abnormal node clusters determined to be of the stable seepage type within the preset adjustment period includes:

[0054] The abnormal node cluster whose seepage anomaly type is determined to be the stable seepage type is marked as a stable node cluster;

[0055] Calculate the Euclidean distance between the position coordinates of each stable node cluster and the coordinates of the preset reference center at each time point within the preset adjustment period to obtain several distribution distances;

[0056] Calculate the standard deviation of all the aforementioned distribution distances to obtain several stable concentrations;

[0057] Calculate the standard deviation of all the aforementioned stable concentrations to obtain the concentration fluctuation value;

[0058] When the concentration fluctuation value is less than the preset concentration fluctuation threshold, the number of monitoring nodes in all the stable node clusters at the end of the preset adjustment period in the reservoir area, the front area of ​​the dam, and the back area of ​​the dam are obtained respectively, so as to obtain the number of reservoir area, the number of front area of ​​the dam, and the number of back area of ​​the dam.

[0059] When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of reservoir areas, and the stability concentration at the end of the preset adjustment cycle is less than the preset concentration threshold, the preset pressure threshold is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold.

[0060] When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of areas in front of the dam, and the stability concentration at the end of the preset adjustment period is less than the preset concentration threshold, the preset water level threshold in front of the dam is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold.

[0061] When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of areas behind the dam, and the stability concentration at the end of the preset adjustment period is less than the preset concentration threshold, the preset water level threshold behind the dam is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold.

[0062] On the other hand, the present invention also provides an Internet of Things-based real-time monitoring system for groundwater in tailings ponds, comprising:

[0063] The acquisition module is used to acquire in real time the tailings slurry temperature and pore water pressure of each monitoring node in the tailings dam area, the groundwater level and water temperature of each monitoring node in the front area of ​​the dam, the groundwater level and conductivity of each monitoring node in the back area of ​​the dam, and the seepage rate of each monitoring node in the seepage zone.

[0064] A cluster determination module, which is connected to the acquisition module, is used to determine several abnormal node clusters based on the groundwater level in front of the dam, a preset groundwater level threshold in front of the dam, the groundwater level behind the dam, a preset groundwater level threshold behind the dam, the pore water pressure, and a preset pressure threshold.

[0065] A path determination module, which is connected to the cluster determination module and the acquisition module respectively, is used to determine the diffusion path of the pollution plume based on the change characteristics of the conductivity, tailings slurry temperature and water temperature in each of the abnormal node clusters.

[0066] A type determination module, which is connected to the path determination module and the acquisition module respectively, is used to determine the seepage anomaly type of each abnormal node cluster based on the change characteristics of the seepage rate in the pollution plume diffusion path within a preset observation window.

[0067] An early warning generation module, which is connected to the type determination module, is used to generate an emergency early warning prompt for the abnormal node cluster whose seepage anomaly type is determined to be continuously enhanced.

[0068] An adjustment module, which is connected to the type determination module and the cluster determination module respectively, is used to adjust the preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold according to the location distribution of the abnormal node clusters whose seepage anomaly type is determined to be stable seepage type within a preset adjustment period.

[0069] Compared with existing technologies, the advantages of this invention lie in its ability to comprehensively collect key hydrogeological parameters such as tailings slurry temperature, pore water pressure, groundwater level, water temperature, conductivity, and seepage rate by deploying multi-source monitoring nodes in the tailings dam area, front dam area, back dam area, and seepage zone. This allows for the construction of a dynamic coupling model that reflects the seepage state and pollution diffusion process in the dam area. On a time scale, threshold values ​​for water level and pore water pressure are used to identify anomalous node clusters. On a spatial scale, conductivity and temperature change characteristics are combined to identify the diffusion path of the pollution plume. Furthermore, the trend analysis of seepage rate changes reveals the response relationship between seepage field energy migration and groundwater dynamic changes. By classifying and identifying different types of seepage anomalies, accurate early warning of abnormal seepage is achieved. Simultaneously, the threshold is dynamically corrected based on the spatial distribution of stable seepage-type node clusters, enabling the monitoring system to adapt to changing groundwater environments. This achieves unified analysis of multi-parameter coupling, spatial correlation, and temporal trends, effectively solving the problem of difficulty in timely detection of complex abnormal seepage due to single monitoring parameters, delayed data updates, and insufficient anomaly identification capabilities.

[0070] Furthermore, by simultaneously considering three key hydraulic parameters—the groundwater level upstream of the dam, the groundwater level downstream of the dam, and pore water pressure—and comparing them with their respective preset thresholds, monitoring nodes exhibiting localized hydraulic response anomalies can be spatially identified. A rise in the water level upstream of the dam reflects increased seepage pressure in the upstream reservoir area, while a rise in the water level downstream indicates obstruction of downstream drainage or seepage channels. Anomalies in pore water pressure reveal the accumulation of seepage potential energy within the dam body. By performing spatial density clustering on these anomaly nodes, not only can the hydraulic coupling effects of the upstream, dam body, and downstream areas be comprehensively reflected, but the spatial continuity of abnormal seepage pressure gradients can also be revealed. This allows for the accurate identification of potential seepage concentration zones or structural hazard areas, enabling early spatial location and dynamic identification of tailings dam seepage instability risks.

[0071] Furthermore, by setting a preset number of types, a spatial cluster is only identified as an anomaly cluster when a sufficient number of different anomaly node types exist simultaneously within it. This effectively distinguishes between sporadic and systematic anomalies, reflecting the mutual influence between hydrological pressure, water level, and pore pressure in the tailings dam body and reservoir area. This ensures that the identified anomaly node clusters represent real potential risk areas, providing a reliable spatial basis for subsequent analysis of pollution plume diffusion paths and determination of seepage anomaly types, thereby improving the accuracy of monitoring and the scientific nature of early warning.

[0072] Furthermore, by acquiring conductivity, tailings slurry temperature, and water temperature information of anomalous node clusters and their surrounding monitoring nodes, and constructing a corresponding spatial distribution field, the spatial distribution differences and trends of each parameter can be revealed. Then, by calculating the spatial correlation between conductivity and slurry / water temperature, collaborative anomaly regions can be identified, reflecting the migration path of pollutants in the groundwater system under the combined influence of fluid and thermal conditions. This method comprehensively considers the interaction between multiple physical quantities such as conductivity, temperature, and water level, making the determined pollution plume diffusion path more accurate and reliable, and timely reflecting the potential movement direction and velocity of pollutants, providing a scientific basis for risk monitoring and emergency response in tailings dams.

[0073] Furthermore, by performing local spatial autocorrelation analysis on the location and conductivity information of each monitoring node within the candidate anomaly zone, hotspot areas with significantly elevated conductivity values ​​can be effectively identified. Core anomaly zones are defined based on the spatial connectivity of these hotspots, and surrounding affected monitoring nodes are incorporated through buffer zone expansion, thus forming collaborative anomaly zones. This approach reflects the coupled influence of upstream and downstream water level changes, pore pressure, and tailings slurry temperature on the migration of dissolved salts in the water body, fully reflecting the spatial concentration trend and local diffusion characteristics of seepage. This makes the subsequent determination of the pollution plume path more accurate and reliable. Simultaneously, the identification of anomaly zones maintains an inherent logical correlation with the type and intensity of seepage anomalies, improving the scientific rigor and practicality of overall early warning and risk assessment.

[0074] Furthermore, by coordinating the conductivity distribution and spatial location of monitoring nodes within the anomaly zone, the main pollution sources and sinks are accurately identified. A continuous conductivity distribution grid is constructed using Kriging spatial interpolation, and the spatial gradient field is calculated to reflect the direction and rate of change in pollution concentration. Then, by simulating the migration path of pollutants along the gradient direction through particle tracking, dynamic prediction of pollution plume diffusion is achieved. The spatial variation of conductivity, gradient direction, and particle trajectory are interconnected, which can reveal the migration law of pollutants in groundwater, accurately depict the possible diffusion path of pollution plumes, and provide a scientific basis for tailings dam environmental risk assessment and emergency response.

[0075] Furthermore, by comprehensively analyzing the cumulative duration, maximum number of consecutive periods, rate fluctuations, and long-term trends of seepage rates at each monitoring node along the pollution plume diffusion path within a preset observation window, the intensity, persistence, and fluctuation characteristics of seepage can be accurately reflected. This allows for the determination of the seepage anomaly type of the abnormal node cluster, revealing the intrinsic relationship between the cumulative effect, continuity, and fluctuation of seepage rates over time. By combining long-term trends with short-term fluctuations, the anomaly types can comprehensively reflect the dynamic characteristics of groundwater flow behavior and tailings dam seepage processes, providing a scientific basis for early warning and reasonable adjustment of monitoring parameters.

[0076] Furthermore, by comprehensively analyzing the cumulative duration, maximum number of consecutive periods, fluctuation values, and long-term trend statistics of seepage rate within a preset observation window, the seepage behavior of abnormal node clusters was classified. Stable seepage nodes exhibit long seepage duration, small fluctuations, and a stable trend, reflecting a balanced groundwater flow. Continuously increasing nodes show a continuous increase in seepage with a significant upward trend, suggesting potential tailings dam leakage or increased dam stress. Intermittently fluctuating nodes show large short-term changes in seepage rate with no clear trend, indicating that seepage is significantly affected by rainfall, drainage, or other external disturbances. By comparing the number of nodes of each type, the main seepage anomaly types of abnormal node clusters were determined, enabling accurate identification of the dynamic characteristics of tailings dam seepage, providing a reliable basis for early warning and control, and demonstrating the intrinsic correlation between seepage rate, duration, fluctuation, and trend.

[0077] Furthermore, by analyzing the location distribution of stable seepage anomaly node clusters within a preset adjustment period, the Euclidean distance and standard deviation of the node clusters relative to the reference center are calculated, thereby quantifying the spatial concentration and fluctuation of the node clusters. When the concentration fluctuation is small and the number of nodes in a certain area is dominant, the corresponding area's upstream water level threshold, downstream water level threshold, or pore water pressure threshold is dynamically adjusted based on the relative deviation between the concentration and the threshold, thereby achieving zoned optimization control of the seepage state in the reservoir area, upstream area, and downstream area. This method reflects the spatial diffusion characteristics and stable trend of seepage anomalies through the multidimensional correlation between location distribution, quantity proportion, and concentration, enabling threshold adjustment to conform to the seepage evolution law and improving the accuracy and response efficiency of tailings dam monitoring and early warning.

[0078] Furthermore, by acquiring key monitoring parameters in real time for the tailings dam area, the area in front of the dam, the area behind the dam, and the seepage zone, including tailings slurry temperature, pore water pressure, water level, conductivity, and seepage rate, and combining these with preset thresholds to determine abnormal node clusters, spatial clustering of abnormal areas and accurate identification of pollution plume paths are achieved. Further, by analyzing the cumulative duration, number of consecutive periods, fluctuation amplitude, and long-term trend of the seepage rate, stable seepage, intermittent fluctuations, and continuously enhancing anomalies are accurately distinguished, achieving dynamic anomaly type determination. Emergency early warning prompts are generated promptly for continuously enhancing node clusters, improving the tailings dam's safety response speed. Simultaneously, based on the location distribution of stable seepage node clusters within the adjustment cycle, concentration analysis is used to dynamically adjust the water level in front of the dam, the water level behind the dam, and the pressure threshold, achieving adaptive optimization of the thresholds. This ensures the matching of the reservoir area's hydrological conditions with the threshold settings, improving the monitoring system's sensitivity and response accuracy to abnormal seepage under different time and spatial conditions. Attached Figure Description

[0079] Figure 1 This is a flowchart of the IoT-based real-time monitoring method for groundwater in tailings ponds in this embodiment.

[0080] Figure 2 This is a flowchart for determining several abnormal node clusters in this embodiment;

[0081] Figure 3 This embodiment defines a logic diagram for determining several abnormal node clusters.

[0082] Figure 4 This is a schematic diagram of the IoT-based real-time groundwater monitoring system for tailings ponds in this embodiment. Detailed Implementation

[0083] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0084] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0085] Please see Figure 1 As shown, this is a flowchart of the real-time monitoring method for groundwater in tailings ponds based on the Internet of Things (IoT) in this embodiment. On one hand, this embodiment provides a real-time monitoring method for groundwater in tailings ponds based on the Internet of Things (IoT), including:

[0086] Real-time acquisition of tailings slurry temperature and pore water pressure at each monitoring node within the tailings dam area, groundwater level and temperature at each monitoring node in the front dam area, groundwater level and conductivity at each monitoring node in the rear dam area, and seepage rate at each monitoring node within the seepage zone.

[0087] Several abnormal node clusters are determined based on the groundwater level in front of the dam, the preset groundwater level threshold in front of the dam, the groundwater level behind the dam, the preset groundwater level threshold behind the dam, the pore water pressure, and the preset pressure threshold.

[0088] The pollution plume diffusion path is determined based on the variation characteristics of the conductivity, tailings slurry temperature, and water temperature in each of the aforementioned abnormal node clusters.

[0089] The seepage anomaly type of each abnormal node cluster is determined based on the variation characteristics of the seepage rate in the pollution plume diffusion path within a preset observation window.

[0090] An emergency early warning is generated for the cluster of abnormal nodes whose seepage anomalies are determined to be of the continuously enhancing type.

[0091] The preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold is adjusted based on the location distribution of the abnormal node clusters that are determined to be of the stable seepage type within the preset adjustment period.

[0092] In this embodiment, the tailings dam is an artificially constructed dammed reservoir used to store tailings slurry discharged during mining operations. Its overall structure typically consists of a reservoir area, a front dam area, a rear dam area, and a seepage zone. The reservoir area is the main storage area for the tailings slurry, responsible for solid-liquid separation and supernatant water recovery. The front dam area, located on the water-facing side of the tailings dam, primarily reflects the impact of water pressure within the reservoir on the dam body. The rear dam area, located on the back dam side, is a crucial area for monitoring seepage discharge and groundwater dynamics. The seepage zone is the area where water flow channels are concentrated in the tailings dam body and foundation, and is a key zone for assessing dam stability and leakage risk. The monitoring system deploys different types of IoT monitoring nodes in each area: real-time temperature data is collected via buried temperature sensors. The system collects tailings slurry temperature and upstream water temperature to reflect heat migration and water flow status; pore water pressure gauges monitor pore water pressure at the bottom of the dam and reservoir to assess saturation and potential seepage risk; water level gauges measure groundwater level changes upstream and downstream of the dam to reveal the hydraulic gradient trend in the reservoir area; conductivity sensors collect downstream groundwater conductivity to reflect pollutant concentration or ion migration; and seepage rate sensors are deployed in the seepage zone to measure the instantaneous flow velocity of water through the medium, characterizing the strength and direction of the seepage process. All of the above monitoring data are aggregated in real-time to a central monitoring platform via a wireless sensor network, enabling comprehensive perception and dynamic analysis of the tailings dam's groundwater dynamics and pollutant migration status.

[0093] The preset in-dam water level threshold is a benchmark value used to determine whether the groundwater level in front of the dam is abnormal. It depends on the design elevation of the tailings dam, the normal water storage depth in front of the dam, and the pressure-bearing capacity of the dam's impermeable layer. It is typically set within the effective water level range of 0.6 to 0.8 times the dam height. In this embodiment, it is set to 0.7 times the dam height, which can sensitively reflect abnormal rises in the water level in front of the dam while ensuring the safety of the dam. The preset post-dam water level threshold is a control benchmark used to identify abnormal changes in the groundwater level behind the dam. It depends on the design elevation of the post-dam seepage drainage system, the permeability coefficient of the dam foundation, and the groundwater recharge intensity. It is typically set between 2 and 5 meters above the dam foundation elevation. In this embodiment, it is set to 3 meters, which can effectively detect abnormal changes in the groundwater level behind the dam. The system effectively reflects early signs of seepage pressure accumulation or poor drainage behind the dam. The preset observation window is the time interval used to analyze the trend of seepage rate changes. It depends on the response cycle of the tailings dam seepage process and the sampling frequency of monitoring data. It is usually set between 24 hours and 7 days. In this embodiment, it is set to 72 hours, which can balance the observation effect of short-term fluctuations and medium-term trends. The preset adjustment cycle is the time interval used to re-evaluate the threshold parameters and the distribution of abnormal nodes. It depends on the tailings dam operation stage, seasonal hydrological changes and the update cycle of the monitoring system. It is usually set between 15 days and 30 days. In this embodiment, it is set to 20 days, which can correct the threshold deviation in a timely manner and maintain the dynamic stability of the monitoring model.

[0094] In this embodiment, generating an emergency early warning for abnormal node clusters whose seepage anomalies are determined to be of a continuously increasing type means that when a continuously increasing type of seepage anomaly is detected, the system automatically determines that there is a risk of seepage channel expansion or increased dam permeability in the area, immediately triggers an emergency early warning, sends alarm information to the monitoring center through the wireless communication module, and marks the abnormal area on the monitoring interface, so that operation and maintenance personnel can take timely intervention measures such as drainage, reinforcement or load reduction, thereby achieving early identification and proactive prevention and control of tailings dam seepage hazards.

[0095] By deploying multi-source monitoring nodes in the tailings dam area, upstream and downstream areas, and seepage zone, key hydrogeological parameters such as tailings slurry temperature, pore water pressure, groundwater level, water temperature, conductivity, and seepage rate were comprehensively collected. A dynamic coupling model reflecting the seepage state and pollution diffusion process in the dam area was constructed. At the time scale, threshold values ​​for water level and pore water pressure were used to identify anomalous node clusters. At the spatial scale, conductivity and temperature variation characteristics were combined to identify the diffusion path of the pollution plume. The response relationship between seepage field energy migration and groundwater dynamic changes was revealed through seepage rate trend analysis. By classifying and identifying different types of seepage anomalies, accurate early warning of abnormal seepage was achieved. Simultaneously, the threshold values ​​were dynamically corrected based on the spatial distribution of stable seepage-type node clusters, enabling the monitoring system to adapt to changing groundwater environments. This achieved unified analysis of multi-parameter coupling, spatial correlation, and temporal trends, effectively solving the problem of difficulty in timely detection of complex abnormal seepage due to single monitoring parameters, delayed data updates, and insufficient anomaly identification capabilities.

[0096] Please see Figure 2 The flowchart shown is for determining several abnormal node clusters in this embodiment. In this embodiment, the process of determining several abnormal node clusters based on the groundwater level in front of the dam, the preset groundwater level threshold in front of the dam, the groundwater level behind the dam, the preset groundwater level threshold behind the dam, the pore water pressure, and the preset pressure threshold includes:

[0097] The monitoring node whose groundwater level in front of the dam is greater than the preset groundwater level threshold in front of the dam is determined to be an abnormal groundwater level in front of the dam and is marked as an abnormal node in front of the dam.

[0098] The monitoring node whose groundwater level behind the dam is greater than the preset groundwater level threshold behind the dam is determined to be an abnormal groundwater level behind the dam and is marked as an abnormal node behind the dam.

[0099] The node type of the monitoring node whose pore water pressure is greater than the preset pressure threshold is determined to be pore water pressure abnormal, and it is marked as a pressure abnormal node.

[0100] The location coordinates of the abnormal nodes in front of the dam, the abnormal water level nodes, and the abnormal pressure nodes are obtained, and spatial density clustering is performed according to a preset clustering algorithm to obtain several spatial clusters.

[0101] Several abnormal node clusters are determined based on the node type of the monitoring nodes in each of the spatial clusters.

[0102] In this embodiment, the corresponding geographical coordinate information, including latitude and longitude or reservoir area plane coordinates, is obtained by the number or sensor ID of each monitoring node in the tailings dam. Combined with the real-time collected data, nodes that are identified as abnormal in front of the dam, abnormal water level, and abnormal pressure are screened out to form an abnormal node set. Then, the coordinates of these nodes are input into a preset clustering algorithm for calculation, which can form several spatial clusters according to the spatial distribution density of the nodes, thereby realizing the spatial aggregation analysis of abnormal nodes.

[0103] In this embodiment, the preset clustering algorithm adopts a clustering method based on node spatial density. Its main parameters include neighborhood radius and minimum node number threshold: the neighborhood radius is used to determine the vicinity of each node, and its value is set according to the spatial spacing of the tailings dam monitoring points and geological and hydrological characteristics, usually between 30 meters and 100 meters, and is set to 50 meters in this embodiment; the minimum node number threshold is used to determine whether the nodes in a certain area constitute an effective cluster, usually set to 3 to 10 nodes, and is set to 5 nodes in this embodiment. By adjusting these parameters, the algorithm can accurately identify spatially concentrated monitoring node clusters with similar anomaly characteristics, thereby facilitating subsequent pollution plume path analysis and seepage anomaly type determination.

[0104] By simultaneously considering three key hydraulic parameters—the groundwater level upstream of the dam, the groundwater level downstream of the dam, and pore water pressure—and comparing them with their respective preset thresholds, monitoring nodes exhibiting localized hydraulic response anomalies can be spatially identified. A rise in the water level upstream of the dam reflects increased seepage pressure in the upstream reservoir area, while a rise in the water level downstream indicates obstruction of downstream drainage or seepage channels. Anomalies in pore water pressure reveal the accumulation of seepage potential energy within the dam body. Spatial density clustering of these anomalous nodes not only comprehensively reflects the hydraulic coupling effects of the upstream, dam, and downstream areas but also reveals the spatial continuity of abnormal seepage pressure gradients. This allows for the accurate identification of potential seepage concentration zones or structural hazard areas, enabling early spatial location and dynamic identification of tailings dam seepage instability risks.

[0105] Please see Figure 3 As shown, this is a logic diagram for determining several abnormal node clusters in this embodiment. In this embodiment, the process of determining several abnormal node clusters based on the node type of the monitoring nodes in each spatial cluster includes:

[0106] When there are more than or equal to the number of node types of a preset type in the spatial cluster, the spatial cluster is determined to be the abnormal node cluster, so as to identify several abnormal node clusters.

[0107] In this embodiment, the preset number of types is a key threshold parameter used to determine whether a spatial cluster constitutes an anomalous node cluster. It represents the minimum number of different anomalous node types that need to appear simultaneously within the same spatial cluster. This parameter depends on the tailings dam body, hydrological structure, and monitoring accuracy, and is usually set between two and three types. In this embodiment, it is set to two types, which can effectively distinguish between occasional single anomalies and collaborative anomalies, ensuring that the spatial clustering results can reflect the interactive influence between parameters such as water level and pore pressure within the dam body and reservoir area, thereby accurately identifying potential risk areas.

[0108] By setting a preset number of types, a spatial anomaly cluster is only identified when a sufficient number of different anomaly node types exist simultaneously within the cluster. This effectively distinguishes between sporadic and systematic anomalies, reflecting the interaction between hydrological pressure, water level, and pore pressure in the tailings dam body and reservoir area. This ensures that the identified anomaly node clusters represent real potential risk areas, providing a reliable spatial basis for subsequent analysis of pollution plume diffusion paths and determination of seepage anomaly types, thereby improving the accuracy of monitoring and the scientific nature of early warning.

[0109] Specifically, the process of determining the diffusion path of the contamination plume based on the variation characteristics of the conductivity, tailings slurry temperature, and water temperature in each of the aforementioned abnormal node clusters includes:

[0110] Obtain the location coordinates, conductivity, tailings slurry temperature, and water temperature of all monitoring nodes within each cluster of abnormal nodes and within the surrounding preset path determination range;

[0111] Based on the location coordinates of the monitoring nodes, spatial distribution fields of the electrical conductivity, tailings slurry temperature, and water temperature are constructed respectively to obtain the electrical conductivity distribution field, slurry temperature distribution field, and water temperature distribution field.

[0112] Calculate the spatial cross-correlation coefficient between the conductivity distribution field and the slurry temperature distribution field to obtain the first spatial correlation coefficient;

[0113] Calculate the spatial cross-correlation coefficient between the conductivity distribution field and the water temperature distribution field to obtain the second spatial correlation coefficient;

[0114] When the first spatial correlation coefficient and / or the second spatial correlation coefficient are greater than a preset correlation threshold, the region where the abnormal node cluster is located is determined to be a candidate abnormal region, and local spatial autocorrelation analysis is performed in the candidate abnormal region to determine the collaborative abnormal region.

[0115] The diffusion path of the pollution plume is determined based on the location coordinates of all the monitoring nodes in the coordinated anomaly zone and the conductivity.

[0116] In this embodiment, the collected monitoring point data is first subjected to quality control and time window aggregation, namely noise reduction, missing value imputation, and taking the mean or median according to the observation window. After converting the latitude and longitude of the points into a unified planar coordinate system, a regular grid is generated within the study area. The grid spacing is usually several meters to tens of meters depending on the monitoring point density and engineering needs. Then, based on the processed point data, spatial interpolation methods are used to construct the distribution fields for conductivity, tailings slurry temperature, and water temperature, with Kriging interpolation being preferred. This involves calculating the empirical variability function and fitting the sump and sump height. Variation model parameters such as degree and oscillation are used to reflect spatial autocorrelation characteristics. Ordinary kriging is used to calculate the interpolation value of each grid point on the grid and output the estimated variance simultaneously. When the monitoring point density is low or there is a clear trend, inverse distance weighted interpolation or spline interpolation can be combined with trend surface decomposition to improve stability. Finally, the accuracy of the interpolation results is evaluated by leave-one-out method. As needed, the grid field is smoothed or the holes are filled, and continuous conductivity field, slurry temperature field and water temperature field with uncertainty information are output for subsequent correlation analysis and trajectory tracking.

[0117] The preset path determination range is used to determine the spatial range of monitoring nodes participating in the pollution plume analysis around the abnormal node cluster. It depends on the geological structure and hydrological conditions of the tailings dam and is usually set between 50 meters and 200 meters. In this embodiment, it is set to 100 meters, which can cover key areas that may be affected by pollution. The preset correlation threshold is used to determine whether there is a significant correlation between conductivity and the spatial distribution field of temperature and water temperature. It depends on the physical properties of tailings slurry and groundwater and the amplitude of environmental fluctuations. It is usually set between 0.6 and 0.9. In this embodiment, it is set to 0.75, which can accurately identify the cooperative anomaly area and exclude the influence of random fluctuations.

[0118] By acquiring conductivity, tailings slurry temperature, and water temperature information of anomalous node clusters and their surrounding monitoring nodes, and constructing corresponding spatial distribution fields, this method can reveal the spatial distribution differences and trends of each parameter. Furthermore, by calculating the spatial correlation between conductivity and slurry / water temperature, collaborative anomaly regions can be identified, reflecting the migration path of pollutants in the groundwater system under the combined influence of fluid and thermal conditions. This method comprehensively considers the interplay between multiple physical quantities such as conductivity, temperature, and water level, making the determined pollution plume diffusion path more accurate and reliable, and timely reflecting the potential movement direction and velocity of pollutants, thus providing a scientific basis for risk monitoring and emergency response in tailings dams.

[0119] Specifically, the process of performing local spatial autocorrelation analysis within candidate anomaly regions to determine co-anomaly regions includes:

[0120] Based on the location coordinates of all monitoring nodes within the candidate anomaly region, a spatial weight matrix is ​​constructed;

[0121] The local spatial autocorrelation statistical value and the standardized Z-score corresponding to the local spatial autocorrelation statistical value of each monitoring node in the conductivity distribution field under the spatial weight matrix are calculated according to the preset statistical algorithm.

[0122] The monitoring nodes whose standardized Z-score is greater than a preset first significance threshold are determined as conductivity significance hotspots, so as to identify a number of conductivity significance hotspots;

[0123] The largest connected region formed in space by all the aforementioned conductivity hotspots is determined to be the core anomalous region.

[0124] A spatial buffer analysis is performed by shifting the outer boundary of the core anomaly region outward by a preset distance. All monitoring nodes covered by the buffer and located within the candidate anomaly region are included in the core anomaly region to determine the collaborative anomaly region.

[0125] In this embodiment, the process of constructing a spatial weight matrix based on the location coordinates of all monitoring nodes in the candidate anomaly area includes: firstly, calculating the Euclidean distance or proximity relationship between each monitoring node, and then determining the weight value between nodes according to the distance or proximity. The closer the distance or the greater the weight of adjacent nodes, the greater the weight of distant or non-adjacent nodes, thereby forming a matrix that reflects the degree of spatial mutual influence between nodes, which is used for subsequent local spatial autocorrelation analysis.

[0126] In this embodiment, the standardized Z-score is used to measure the degree of deviation of the local spatial autocorrelation statistic of a monitoring node from the overall distribution. The calculation result reflects whether the node exhibits significant high-value or low-value clustering characteristics in the conductivity distribution field. The larger the Z-score obtained by subtracting the average value of all nodes from the local spatial autocorrelation statistic and dividing by the standard deviation, the stronger the clustering of conductivity values ​​around the node; the smaller the Z-score, the more significant the low-value clustering in its neighborhood. This index can quantitatively identify local abnormal regions in the conductivity field, providing an accurate basis for subsequent determination of pollution plume direction and identification of seepage anomalies.

[0127] The preset statistical algorithm is a hotspot analysis method based on spatial autocorrelation. It depends on the conductivity distribution characteristics and spatial distribution density of the monitoring nodes and is usually selected from various spatial statistical methods. In this embodiment, it is set to Getis-OrdGi*, which can identify local significant abnormal regions and determine significant conductivity hotspots. The preset first significance threshold is the standard used to determine whether a monitoring node is a significant conductivity hotspot. It depends on the statistical distribution characteristics of the conductivity of the monitoring nodes and is usually set between 90% and 99%. In this embodiment, it is set to 95%, which can effectively distinguish abnormal hotspot nodes from normal nodes. The preset push distance is the spatial buffer distance used to extend the boundary of the core abnormal region. It depends on the spatial distribution density of the monitoring nodes in the abnormal region and is usually set between 10 meters and 20 meters. In this embodiment, it is set to 20 meters, which can cover the monitoring nodes that may be affected around the core abnormal region.

[0128] By performing local spatial autocorrelation analysis on the location and conductivity information of each monitoring node within the candidate anomaly zone, hotspots with significantly elevated conductivity values ​​can be effectively identified. Core anomaly zones are defined based on the spatial connectivity of these hotspots, and surrounding affected monitoring nodes are incorporated through buffer zone expansion, thus forming collaborative anomaly zones. This approach reflects the coupled influence of upstream and downstream water level changes, pore pressure, and tailings slurry temperature on the migration of dissolved salts in the water, fully demonstrating the spatial concentration trend and local diffusion characteristics of seepage. This makes the subsequent determination of pollution plume paths more accurate and reliable. Furthermore, the identification of anomaly zones maintains an inherent logical connection with seepage anomaly types and intensities, improving the scientific rigor and practicality of overall early warning and risk assessment.

[0129] Specifically, the process of determining the pollution plume diffusion path based on the location coordinates of all monitoring nodes in the coordinated anomaly zone and the conductivity includes:

[0130] The monitoring node corresponding to the maximum conductivity value of all the aforementioned components within the coordinated anomaly zone is identified as the primary pollution source.

[0131] The monitoring node corresponding to the minimum conductivity of all the components at the downstream boundary of the coordinated anomaly zone is determined as the main pollution sink.

[0132] Based on the location coordinates of all the monitoring nodes in the collaborative anomaly zone and the conductivity, a kriging spatial interpolation method is used to generate a conductivity spatial distribution raster map.

[0133] Calculate the spatial gradient field of the conductivity spatial distribution grid, wherein the spatial gradient field characterizes the direction and rate of change of the conductivity at each point in space;

[0134] A particle tracking algorithm is used to release several virtual particles from the main pollution source.

[0135] The virtual particle is made to move along the negative gradient direction of the conductivity concentration in the spatial gradient field to obtain several movement trajectories.

[0136] All the movement trajectories converging towards the main pollution sink point are determined to be candidate diffusion paths;

[0137] The candidate diffusion paths that are spatially adjacent and oriented in the same direction are fused to determine the pollution plume diffusion path.

[0138] In this embodiment, based on the location coordinates and conductivity of all monitoring nodes within the cooperative anomaly zone, the node with the maximum conductivity is first identified as the main pollution source, and the downstream node with the minimum conductivity is identified as the main pollution sink. Subsequently, a spatial distribution raster map of conductivity is generated using the Kriging space interpolation method, and its spatial gradient field is calculated to characterize the direction and rate of conductivity change at each location. Virtual particles are released from the main pollution source using a particle tracking algorithm, causing them to move along the negative conductivity gradient direction to form a trajectory. The trajectory that converges to the main pollution sink is used as a candidate diffusion path. Then, adjacent paths with the same direction are fused to determine the pollution plume diffusion path.

[0139] In this embodiment, the Kriging spatial interpolation method establishes a spatial statistical model by utilizing the spatial location of monitoring points and their corresponding conductivity values. This transforms discrete monitoring data into a continuous spatial distribution grid, considering spatial autocorrelation—that is, conductivity values ​​at similar locations are more likely to be similar—thus accurately reflecting the spatial distribution and changing trends of pollutant concentrations in groundwater. The particle tracking algorithm, within the constructed conductivity spatial gradient field, uses pollution source points as particle release points and simulates pollutant migration trajectories along the negative direction of the conductivity gradient. Each virtual particle represents a moving unit of pollutants in groundwater. By tracking the particle trajectories, the migration path, convergence location, and potential diffusion range of pollutants can be revealed, providing an intuitive and quantitative basis for judging the diffusion path of pollution plumes.

[0140] By coordinating the conductivity distribution and spatial location of monitoring nodes within the anomaly zone, the main pollution sources and sinks are accurately identified. A continuous conductivity distribution grid is constructed using Kriging spatial interpolation, and the spatial gradient field is calculated to reflect the direction and rate of change in pollution concentration. Then, particle tracking is used to simulate the migration path of pollutants along the gradient direction, enabling dynamic prediction of pollution plume diffusion. The spatial variation of conductivity, gradient direction, and particle trajectory are interconnected, revealing the migration patterns of pollutants in groundwater and accurately depicting the possible diffusion paths of pollution plumes, providing a scientific basis for tailings dam environmental risk assessment and emergency response.

[0141] Specifically, the process of determining the seepage anomaly type of each anomalous node cluster based on the variation characteristics of the seepage rate within a preset observation window in the pollution plume diffusion path includes:

[0142] Calculate the cumulative duration and maximum number of consecutive periods during which the seepage rate is greater than a preset seepage rate threshold for all monitoring nodes along the pollution plume diffusion path within the preset observation window;

[0143] Calculate the standard deviation of the rate of change of the seepage rate at all monitoring nodes along the pollution plume diffusion path within the preset observation window to obtain the change fluctuation value;

[0144] The long-term trend of seepage rate at each monitoring node along the pollution plume diffusion path was analyzed using the Mann-Kendall trend test method to obtain the trend test statistic.

[0145] The seepage anomaly type of each abnormal node cluster is determined based on the cumulative duration, the maximum number of consecutive time periods, the change fluctuation value, and the trend test statistic.

[0146] In this embodiment, the Mann-Kendall trend test is a non-parametric statistical method used to determine whether time series data exhibits a monotonically increasing or decreasing trend. By comparing the magnitude of data at any two time points in the series, the cumulative positive and negative differences are calculated to obtain the trend test statistic. The significance level can be further obtained through standardization to determine the strength and direction of the trend. It does not require the assumption that the data follows a specific distribution and can effectively analyze the long-term changing trends of continuous monitoring data such as seepage rate, meteorology, and water level, providing a reliable basis for the determination and early warning of seepage anomalies.

[0147] By comprehensively analyzing the cumulative duration, maximum number of consecutive periods, rate fluctuations, and long-term trends of seepage rates at each monitoring node along the pollution plume diffusion path within a preset observation window, the intensity, persistence, and fluctuation characteristics of seepage can be accurately reflected. This allows for the determination of the seepage anomaly type of the abnormal node cluster, revealing the intrinsic relationship between the cumulative effect, continuity, and fluctuation of seepage rates over time. Furthermore, by combining long-term trends with short-term fluctuations, the anomaly types can comprehensively reflect the dynamic characteristics of groundwater flow behavior and tailings dam seepage processes, providing a scientific basis for early warning and reasonable adjustment of monitoring parameters.

[0148] Specifically, the process of determining the seepage anomaly type of each of the abnormal node clusters based on the cumulative duration, the maximum number of consecutive time periods, the change fluctuation value, and the trend test statistic includes:

[0149] When the cumulative duration is greater than a preset first persistence threshold, the maximum number of consecutive time periods is greater than a preset second persistence threshold, the change fluctuation value is less than a preset first fluctuation threshold, and the absolute value of the trend test statistic is less than a preset first trend threshold, the seepage anomaly type of the monitoring node is determined to be stable seepage type.

[0150] When the cumulative duration is greater than a preset third persistence threshold, the maximum number of consecutive time periods is greater than a preset fourth persistence threshold, and the trend test statistic is greater than a preset second trend threshold, the seepage anomaly type of the monitoring node is determined to be continuously enhanced.

[0151] When the cumulative duration is less than a preset fifth persistence threshold, the change fluctuation value is greater than a preset second fluctuation threshold, and the absolute value of the trend test statistic is less than the preset first trend threshold, the seepage anomaly type of the monitoring node is determined to be intermittent fluctuation type.

[0152] The seepage anomaly type of the abnormal node cluster is determined by the maximum number of monitoring nodes corresponding to the continuously enhanced type, the intermittent fluctuation type, and the stable seepage type in the abnormal node cluster.

[0153] A first persistence threshold is preset as the judgment threshold for the cumulative duration, which depends on the length of the preset observation window and the statistical distribution of historical normal seepage states. It is typically set between 60% and 80% of the observation window duration; in this embodiment, it is set to 70% to ensure sufficient persistence of the abnormal state and eliminate short-term fluctuation interference. A second persistence threshold is preset as the judgment threshold for the maximum number of consecutive time periods, which depends on the noise level of the monitoring data and the minimum identification standard for consecutive abnormal events. It is typically set between 1.5 and 2.5 times the minimum number of time periods that can constitute a consecutive abnormal event; in this embodiment, it is set to 2 times to effectively identify... Continuous, non-intermittent abnormal events; a preset first volatility threshold is the maximum allowable volatility value for identifying a steady state, which depends on the standard deviation of the rate of change of seepage rate during historical steady seepage stages, and is usually set between 1.0 and 1.5 times the historical mean of this standard deviation. In this embodiment, it is set to 1.2 times, which can effectively define the upper limit of the allowable volatility range for steady seepage; a preset first trend threshold is the absolute value of the maximum statistic for determining whether a trend is insignificant, which depends on the critical value of the Mann-Kendall trend test at the selected significance level, and is usually set between 1.96 and 2.58. In this embodiment, it is set to 2.0, which can effectively define the upper limit of the allowable volatility range for steady seepage; High statistical power determines that the seepage rate does not show a significant upward or downward trend; the preset third persistence threshold is a more stringent cumulative duration threshold required to identify continuously enhancing anomalies, and its value is usually higher than the first persistence threshold, set between 75% and 90% of the observation window duration. In this embodiment, it is set to 85%, which can ensure that the anomaly has fully manifested before it is determined to be continuously enhancing; the preset fourth persistence threshold is a more stringent maximum number of consecutive time periods required to identify continuously enhancing anomalies, and its value is usually higher than the second persistence threshold. In this embodiment, it is set to 3 times the minimum number of time periods that can constitute a continuous anomalous event, which can further confirm the continuity of the anomaly. The second trend threshold is a preset second statistical measure to determine the existence of a significant upward trend. It is based on the same statistical test principle as the first trend threshold and is usually set between 1.96 and 2.58. In this embodiment, it is set to 2.0, which can confirm the existence of a significant monotonically increasing trend in the seepage rate with high statistical power. The second volatility threshold is a preset second volatility value to determine the minimum volatility anomaly. Its value is usually higher than the first volatility threshold and is set between 1.8 and 2.5 times the standard deviation of the seepage rate change rate in the historical stable seepage stage. In this embodiment, it is set to 2.0 times, which can effectively capture intermittent anomalies with high volatility characteristics.

[0154] By comprehensively analyzing the cumulative duration, maximum number of consecutive periods, fluctuation values, and long-term trend statistics of seepage rate within a preset observation window, the seepage behavior of abnormal node clusters is classified. Stable seepage nodes are characterized by long seepage duration, small fluctuations, and a stable trend, reflecting a balanced groundwater flow. Continuously increasing nodes show a continuous increase in seepage with a significant upward trend, suggesting potential tailings dam leakage or increased dam stress. Intermittently fluctuating nodes exhibit large short-term changes in seepage rate with no clear trend, indicating that seepage is significantly affected by rainfall, drainage, or other external disturbances. By comparing the number of nodes of each type, the main seepage anomaly types of abnormal node clusters are determined, enabling accurate identification of the dynamic characteristics of tailings dam seepage. This provides a reliable basis for early warning and control, and also demonstrates the intrinsic correlation between seepage rate, duration, fluctuation, and trend.

[0155] Specifically, the process of adjusting the preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold based on the location distribution of the abnormal node clusters determined to be of the stable seepage type within a preset adjustment period includes:

[0156] The abnormal node cluster whose seepage anomaly type is determined to be the stable seepage type is marked as a stable node cluster;

[0157] Calculate the Euclidean distance between the position coordinates of each stable node cluster and the coordinates of the preset reference center at each time point within the preset adjustment period to obtain several distribution distances;

[0158] Calculate the standard deviation of all the aforementioned distribution distances to obtain several stable concentrations;

[0159] Calculate the standard deviation of all the aforementioned stable concentrations to obtain the concentration fluctuation value;

[0160] When the concentration fluctuation value is less than the preset concentration fluctuation threshold, the number of monitoring nodes in all the stable node clusters at the end of the preset adjustment period in the reservoir area, the front area of ​​the dam, and the back area of ​​the dam are obtained respectively, so as to obtain the number of reservoir area, the number of front area of ​​the dam, and the number of back area of ​​the dam.

[0161] When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of reservoir areas, and the stability concentration at the end of the preset adjustment cycle is less than the preset concentration threshold, the preset pressure threshold is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold, where Q1'=Q1×(1-k1×︱J-J0︱ / J0), Q1' is the preset pressure threshold after reduction, Q1 is the preset pressure threshold before reduction, k1 is the preset first adjustment coefficient, J is the stability concentration, and J0 is the preset concentration threshold;

[0162] When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of areas in front of the dam, and the stability concentration at the end of the preset adjustment period is less than the preset concentration threshold, the preset water level threshold in front of the dam is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold, where Q2'=Q2×(1-k2×︱J-J0︱ / J0), Q2' is the reduced preset water level threshold in front of the dam, Q2 is the original preset water level threshold in front of the dam, and k2 is the preset second adjustment coefficient;

[0163] When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of areas behind the dam, and the stability concentration at the end of the preset adjustment period is less than the preset concentration threshold, the preset water level threshold behind the dam is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold, where Q3'=Q3×(1-k3×︱J-J0︱ / J0), Q3' is the reduced preset water level threshold behind the dam, Q3 is the original preset water level threshold behind the dam, and k3 is the preset third adjustment coefficient.

[0164] The preset concentration fluctuation threshold is a threshold for judging whether the change in the distribution of stable node clusters is significant. It depends on the fluctuation range of node cluster concentration under historical stable seepage conditions and is usually set between 0.5 and 1.0 times the standard deviation of historical concentration. In this embodiment, it is set to 0.7 times, which can effectively distinguish between stable and fluctuating node distributions. The preset concentration threshold is a threshold for judging whether the spatial concentration of node clusters has reached the adjustment condition. It depends on the node distribution characteristics and spatial density of each area of ​​the tailings dam and is usually set between 0.6 and 0.9 times the standard deviation of concentration. In this embodiment, it is set to 0.8 times, which can reasonably trigger threshold adjustment to optimize seepage control. The preset first adjustment coefficient is a proportional coefficient for reducing the pressure threshold of the dam area, which depends on the seepage. The sensitivity of the stability concentration to pressure threshold adjustment is typically set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which allows for smooth adjustment of the pressure threshold and avoids over-correction. The preset second adjustment coefficient is a proportional coefficient for reducing the upstream water level threshold. It depends on the relationship between the stability concentration in the upstream area and the water level threshold, and is typically set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which can effectively adjust the upstream water level threshold to adapt to local seepage changes. The preset third adjustment coefficient is a proportional coefficient for reducing the downstream water level threshold. It depends on the responsiveness of the downstream stability concentration to threshold adjustment, and is typically set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which can precisely adjust the downstream water level threshold and achieve regional seepage management.

[0165] By analyzing the location distribution of stable seepage anomaly node clusters within a preset adjustment period, the Euclidean distance and standard deviation of the node clusters relative to the reference center are calculated, thereby quantifying the spatial concentration and fluctuation of the node clusters. When the concentration fluctuation is small and the number of nodes in a certain area is dominant, the threshold values ​​for the upstream and downstream water levels or pore water pressure in the corresponding area are dynamically adjusted according to the relative deviation between the concentration and the threshold, thereby achieving zoned optimization control of the seepage state in the reservoir area, the upstream area, and the downstream area. This method reflects the spatial diffusion characteristics and stable trend of seepage anomalies through the multidimensional correlation between location distribution, quantity ratio, and concentration, enabling the threshold adjustment to conform to the seepage evolution law and improving the accuracy and response efficiency of tailings dam monitoring and early warning.

[0166] Please see Figure 4 As shown, this is a schematic diagram of the IoT-based real-time monitoring system for groundwater in tailings ponds according to this embodiment. Furthermore, this embodiment also provides an IoT-based real-time monitoring system for groundwater in tailings ponds, including:

[0167] The acquisition module is used to acquire in real time the tailings slurry temperature and pore water pressure of each monitoring node in the tailings dam area, the groundwater level and water temperature of each monitoring node in the front area of ​​the dam, the groundwater level and conductivity of each monitoring node in the back area of ​​the dam, and the seepage rate of each monitoring node in the seepage zone.

[0168] A cluster determination module, which is connected to the acquisition module, is used to determine several abnormal node clusters based on the groundwater level in front of the dam, a preset groundwater level threshold in front of the dam, the groundwater level behind the dam, a preset groundwater level threshold behind the dam, the pore water pressure, and a preset pressure threshold.

[0169] A path determination module, which is connected to the cluster determination module and the acquisition module respectively, is used to determine the diffusion path of the pollution plume based on the change characteristics of the conductivity, tailings slurry temperature and water temperature in each of the abnormal node clusters.

[0170] A type determination module, which is connected to the path determination module and the acquisition module respectively, is used to determine the seepage anomaly type of each abnormal node cluster based on the change characteristics of the seepage rate in the pollution plume diffusion path within a preset observation window.

[0171] An early warning generation module, which is connected to the type determination module, is used to generate an emergency early warning prompt for the abnormal node cluster whose seepage anomaly type is determined to be continuously enhanced.

[0172] An adjustment module, which is connected to the type determination module and the cluster determination module respectively, is used to adjust the preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold according to the location distribution of the abnormal node clusters whose seepage anomaly type is determined to be stable seepage type within a preset adjustment period.

[0173] By acquiring key monitoring parameters in real time from the tailings dam area, the area in front of the dam, the area behind the dam, and the seepage zone, including tailings slurry temperature, pore water pressure, water level, conductivity, and seepage rate, and combining these with preset thresholds to determine abnormal node clusters, the system achieves spatial clustering of abnormal areas and accurate identification of pollution plume paths. Furthermore, by analyzing the cumulative duration, number of consecutive periods, fluctuation amplitude, and long-term trend of the seepage rate, the system accurately distinguishes between stable seepage, intermittent fluctuations, and continuously intensifying anomalies, achieving dynamic anomaly type determination. For continuously intensifying node clusters, timely emergency warnings are generated, improving the tailings dam's safety response speed. Simultaneously, based on the location distribution of stable seepage node clusters within the adjustment cycle, concentration analysis is used to dynamically adjust the water level in front of the dam, the water level behind the dam, and the pressure thresholds, achieving adaptive optimization of the thresholds. This ensures the matching of the reservoir area's hydrological conditions with the threshold settings, improving the monitoring system's sensitivity and response accuracy to abnormal seepage under different time and spatial conditions.

[0174] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for real-time monitoring of groundwater in tailings dams based on the Internet of Things, characterized in that, include: Real-time acquisition of tailings slurry temperature and pore water pressure at each monitoring node within the tailings dam area, groundwater level and temperature at each monitoring node in the front dam area, groundwater level and conductivity at each monitoring node in the rear dam area, and seepage rate at each monitoring node within the seepage zone. Several abnormal node clusters are determined based on the groundwater level in front of the dam, the preset groundwater level threshold in front of the dam, the groundwater level behind the dam, the preset groundwater level threshold behind the dam, the pore water pressure, and the preset pressure threshold. The pollution plume diffusion path is determined based on the variation characteristics of the conductivity, tailings slurry temperature, and water temperature in each of the aforementioned abnormal node clusters. The seepage anomaly type of each abnormal node cluster is determined based on the variation characteristics of the seepage rate in the pollution plume diffusion path within a preset observation window. An emergency early warning is generated for the cluster of abnormal nodes whose seepage anomalies are determined to be of the continuously enhancing type. The preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold is adjusted based on the location distribution of the abnormal node clusters that are determined to be of the stable seepage type within the preset adjustment period.

2. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 1, characterized in that, The process of determining several abnormal node clusters based on the groundwater level in front of the dam, a preset threshold for the groundwater level in front of the dam, the groundwater level behind the dam, a preset threshold for the groundwater level behind the dam, the pore water pressure, and a preset pressure threshold includes: The monitoring node whose groundwater level in front of the dam is greater than the preset groundwater level threshold in front of the dam is determined to be an abnormal groundwater level in front of the dam and is marked as an abnormal node in front of the dam. The monitoring node whose groundwater level behind the dam is greater than the preset groundwater level threshold behind the dam is determined to be an abnormal groundwater level behind the dam and is marked as an abnormal node behind the dam. The node type of the monitoring node whose pore water pressure is greater than the preset pressure threshold is determined to be pore water pressure abnormal, and it is marked as a pressure abnormal node. The location coordinates of the abnormal nodes in front of the dam, the abnormal water level nodes, and the abnormal pressure nodes are obtained, and spatial density clustering is performed according to a preset clustering algorithm to obtain several spatial clusters. Several abnormal node clusters are determined based on the node type of the monitoring nodes in each of the spatial clusters.

3. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 2, characterized in that, The process of determining several abnormal node clusters based on the node type of the monitoring nodes in each of the spatial clusters includes: When there are more than or equal to the number of node types of a preset type in the spatial cluster, the spatial cluster is determined to be the abnormal node cluster, so as to identify several abnormal node clusters.

4. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 3, characterized in that, The process of determining the pollution plume diffusion path based on the variation characteristics of the conductivity, tailings slurry temperature, and water temperature in each of the aforementioned abnormal node clusters includes: Obtain the location coordinates, conductivity, tailings slurry temperature, and water temperature of all monitoring nodes within each cluster of abnormal nodes and within the surrounding preset path determination range; Based on the location coordinates of the monitoring nodes, spatial distribution fields of the electrical conductivity, tailings slurry temperature, and water temperature are constructed respectively to obtain the electrical conductivity distribution field, slurry temperature distribution field, and water temperature distribution field. Calculate the spatial cross-correlation coefficient between the conductivity distribution field and the slurry temperature distribution field to obtain the first spatial correlation coefficient; Calculate the spatial cross-correlation coefficient between the conductivity distribution field and the water temperature distribution field to obtain the second spatial correlation coefficient; When the first spatial correlation coefficient and / or the second spatial correlation coefficient are greater than a preset correlation threshold, the region where the abnormal node cluster is located is determined to be a candidate abnormal region, and local spatial autocorrelation analysis is performed in the candidate abnormal region to determine the collaborative abnormal region. The diffusion path of the pollution plume is determined based on the location coordinates of all the monitoring nodes in the coordinated anomaly zone and the conductivity.

5. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 4, characterized in that, The process of performing local spatial autocorrelation analysis within candidate anomaly regions to determine co-anomaly regions includes: Based on the location coordinates of all monitoring nodes within the candidate anomaly region, a spatial weight matrix is ​​constructed; The local spatial autocorrelation statistical value and the standardized Z-score corresponding to the local spatial autocorrelation statistical value of each monitoring node in the conductivity distribution field under the spatial weight matrix are calculated according to the preset statistical algorithm. The monitoring nodes whose standardized Z-score is greater than a preset first significance threshold are determined as conductivity significance hotspots, so as to identify a number of conductivity significance hotspots; The largest connected region formed in space by all the aforementioned conductivity hotspots is determined to be the core anomalous region. A spatial buffer analysis is performed by shifting the outer boundary of the core anomaly region outward by a preset distance. All monitoring nodes covered by the buffer and located within the candidate anomaly region are included in the core anomaly region to determine the collaborative anomaly region.

6. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 5, characterized in that, The process of determining the pollution plume diffusion path based on the location coordinates of all monitoring nodes in the coordinated anomaly zone and the conductivity includes: The monitoring node corresponding to the maximum conductivity value of all the aforementioned components within the coordinated anomaly zone is identified as the primary pollution source. The monitoring node corresponding to the minimum conductivity of all the components at the downstream boundary of the coordinated anomaly zone is determined as the main pollution sink. Based on the location coordinates of all the monitoring nodes in the collaborative anomaly zone and the conductivity, a kriging spatial interpolation method is used to generate a conductivity spatial distribution raster map. Calculate the spatial gradient field of the conductivity spatial distribution grid, wherein the spatial gradient field characterizes the direction and rate of change of the conductivity at each point in space; A particle tracking algorithm is used to release several virtual particles from the main pollution source. The virtual particle is made to move along the negative gradient direction of the conductivity concentration in the spatial gradient field to obtain several movement trajectories. All the movement trajectories converging towards the main pollution sink point are determined to be candidate diffusion paths; The candidate diffusion paths that are spatially adjacent and oriented in the same direction are fused to determine the pollution plume diffusion path.

7. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 6, characterized in that, The process of determining the seepage anomaly type of each anomalous node cluster based on the variation characteristics of the seepage rate within a preset observation window in the pollution plume diffusion path includes: Calculate the cumulative duration and maximum number of consecutive periods during which the seepage rate is greater than a preset seepage rate threshold for all monitoring nodes along the pollution plume diffusion path within the preset observation window; Calculate the standard deviation of the rate of change of the seepage rate at all monitoring nodes along the pollution plume diffusion path within the preset observation window to obtain the change fluctuation value; The long-term trend of seepage rate at each monitoring node along the pollution plume diffusion path was analyzed using the Mann-Kendall trend test method to obtain the trend test statistic. The seepage anomaly type of each abnormal node cluster is determined based on the cumulative duration, the maximum number of consecutive time periods, the change fluctuation value, and the trend test statistic.

8. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things according to claim 7, characterized in that, The process of determining the seepage anomaly type of each of the abnormal node clusters based on the cumulative duration, the maximum number of consecutive time periods, the change fluctuation value, and the trend test statistic includes: When the cumulative duration is greater than a preset first persistence threshold, the maximum number of consecutive time periods is greater than a preset second persistence threshold, the change fluctuation value is less than a preset first fluctuation threshold, and the absolute value of the trend test statistic is less than a preset first trend threshold, the seepage anomaly type of the monitoring node is determined to be stable seepage type. When the cumulative duration is greater than a preset third persistence threshold, the maximum number of consecutive time periods is greater than a preset fourth persistence threshold, and the trend test statistic is greater than a preset second trend threshold, the seepage anomaly type of the monitoring node is determined to be continuously enhanced. When the cumulative duration is less than a preset fifth persistence threshold, the change fluctuation value is greater than a preset second fluctuation threshold, and the absolute value of the trend test statistic is less than the preset first trend threshold, the seepage anomaly type of the monitoring node is determined to be intermittent fluctuation type. The seepage anomaly type of the abnormal node cluster is determined by the maximum number of monitoring nodes corresponding to the continuously enhanced type, the intermittent fluctuation type, and the stable seepage type in the abnormal node cluster.

9. The method for real-time monitoring of groundwater in tailings dams based on the Internet of Things as described in claim 8, characterized in that, The process of adjusting the preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold based on the location distribution of the abnormal node clusters determined to be of the stable seepage type within the preset adjustment period includes: The abnormal node cluster whose seepage anomaly type is determined to be the stable seepage type is marked as a stable node cluster; Calculate the Euclidean distance between the position coordinates of each stable node cluster and the coordinates of the preset reference center at each time point within the preset adjustment period to obtain several distribution distances; Calculate the standard deviation of all the aforementioned distribution distances to obtain several stable concentrations; Calculate the standard deviation of all the aforementioned stable concentrations to obtain the concentration fluctuation value; When the concentration fluctuation value is less than the preset concentration fluctuation threshold, the number of monitoring nodes in all the stable node clusters at the end of the preset adjustment period in the reservoir area, the front area of ​​the dam, and the back area of ​​the dam are obtained respectively, so as to obtain the number of reservoir area, the number of front area of ​​the dam, and the number of back area of ​​the dam. When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of reservoir areas, and the stability concentration at the end of the preset adjustment cycle is less than the preset concentration threshold, the preset pressure threshold is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold. When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of areas in front of the dam, and the stability concentration at the end of the preset adjustment period is less than the preset concentration threshold, the preset water level threshold in front of the dam is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold. When the maximum value of the number of reservoir areas, the number of areas in front of the dam, and the number of areas behind the dam is equal to the number of areas behind the dam, and the stability concentration at the end of the preset adjustment period is less than the preset concentration threshold, the preset water level threshold behind the dam is reduced according to the relative deviation between the preset stability concentration and the preset concentration threshold.

10. A real-time monitoring system for groundwater in tailings ponds based on the Internet of Things (IoT), constructed based on the real-time monitoring method for groundwater in tailings ponds based on the IoT as described in any one of claims 1-9, characterized in that, include: The acquisition module is used to acquire in real time the tailings slurry temperature and pore water pressure of each monitoring node in the tailings dam area, the groundwater level and water temperature of each monitoring node in the front area of ​​the dam, the groundwater level and conductivity of each monitoring node in the back area of ​​the dam, and the seepage rate of each monitoring node in the seepage zone. A cluster determination module, which is connected to the acquisition module, is used to determine several abnormal node clusters based on the groundwater level in front of the dam, a preset groundwater level threshold in front of the dam, the groundwater level behind the dam, a preset groundwater level threshold behind the dam, the pore water pressure, and a preset pressure threshold. A path determination module, which is connected to the cluster determination module and the acquisition module respectively, is used to determine the diffusion path of the pollution plume based on the change characteristics of the conductivity, tailings slurry temperature and water temperature in each of the abnormal node clusters. A type determination module, which is connected to the path determination module and the acquisition module respectively, is used to determine the seepage anomaly type of each abnormal node cluster based on the change characteristics of the seepage rate in the pollution plume diffusion path within a preset observation window. An early warning generation module, which is connected to the type determination module, is used to generate an emergency early warning prompt for the abnormal node cluster whose seepage anomaly type is determined to be continuously enhanced. An adjustment module, which is connected to the type determination module and the cluster determination module respectively, is used to adjust the preset upstream water level threshold, the preset downstream water level threshold, or the preset pressure threshold according to the location distribution of the abnormal node clusters whose seepage anomaly type is determined to be stable seepage type within a preset adjustment period.

Images