A risk prediction method for valley tailings ponds threatened by floods in extremely small watersheds

Abstract

The present invention provides a risk prediction method for valley-type tailings ponds threatened by floods in small watersheds, including: S1: obtaining spatial distribution data of soil type, vegetation coverage, slope and gully in small watersheds of valley-type tailings ponds, and establishing an underlying surface classification system; S2: extracting nonlinear key inducements in the rainfall-runoff process in small watersheds of valley terrain, identifying the dominant factors affecting runoff formation, and constructing a nonlinear convergence calculation numerical model; S3: constructing a distributed sensor monitoring network through rainfall monitoring stations and runoff monitoring stations; S4: constructing a hierarchical three-dimensional hydrological and hydrodynamic model of coupled rainfall flow generation-slope convergence-shallow groundwater flow; S5: through the hierarchical three-dimensional hydrological and hydrodynamic model, combined with flood risk influencing factors, establishing a risk assessment index system and a risk prediction model. The present invention realizes the accurate prediction of valley-type tailings ponds under the threat of floods in small watersheds, and solves the problem of poor predictability of flood threats in small watersheds.

Description

Technical Field

[0001] The invention relates to the technical field of valley flood prediction, and in particular to a risk prediction method for a valley-type tailings pond threatened by floods in an extremely small watershed. Background Art

[0002] With the continuous expansion of mining activities, the safety of valley-type tailings ponds, as important facilities for storing mine waste, has received increasing attention. Especially when these tailings ponds are located in very small watersheds, they face safety threats caused by floods due to the characteristics of the terrain and the impact of climate change. Traditionally, the assessment of such risks mainly relies on empirical judgment and simple statistical methods, which is not only inefficient, but also difficult to accurately predict flood risks under extreme weather conditions. In addition, when dealing with flood risk prediction under complex terrain conditions, existing technical solutions often lack sufficient consideration of key factors such as soil type and vegetation coverage, resulting in greater uncertainty in the prediction results.

[0003] In the prior art, the risk prediction method for valley-type tailings ponds threatened by floods in very small watersheds usually relies on statistical analysis based on historical data and simple hydrological models. The advantage is that the implementation is relatively simple, limited data resources can be used for preliminary assessment, and in some cases reasonable risk estimates can be provided. However, its main disadvantage is the lack of comprehensive consideration of key factors such as soil type, vegetation coverage and other variables under complex terrain conditions, resulting in inaccurate prediction results; at the same time, the processing ability for nonlinear processes is weak, making it difficult to accurately simulate the flood response mechanism under extreme weather conditions.

[0004] Therefore, the risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds of the present invention adopts a comprehensive hydrological and hydrodynamic model, combines risk factors, and implements a risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds based on a risk assessment index system. Summary of the invention

[0005] The present invention provides a risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds, comprising:

[0006] S1: Obtain the spatial distribution data of soil type, vegetation coverage, slope and gully in the valley-type Tailings Kut small watershed, and establish an underlying surface classification system;

[0007] S2: Extract the nonlinear key factors in the rainfall-runoff process in the valley terrain and small watershed, use statistical learning methods to identify the dominant factors affecting runoff formation, and construct a nonlinear confluence calculation numerical model;

[0008] S3: Construct a distributed sensor monitoring network through rainfall monitoring stations and runoff monitoring stations;

[0009] S4: Obtain the controlling factors of rainfall intensity, vegetation interception, and permeability, and construct a layered three-dimensional hydrological and hydrodynamic model that couples rainfall runoff-slope runoff-shallow groundwater flow;

[0010] S5: Through the layered three-dimensional hydrological and hydrodynamic model, combined with the factors affecting flood risk, a risk assessment index system and risk prediction model are established to achieve accurate prediction of flood risks in valley-type tailings Kut small watersheds.

[0011] Preferably, the spatial distribution data of soil type, vegetation coverage, slope and gully in the valley-type tailings Kut small watershed are obtained by collecting high-resolution images of the valley-type tailings Kut small watershed through unmanned aerial vehicles, and obtaining three-dimensional surface information in combination with lidar multi-echo remote sensing technology; the spatial distribution data of slope and gully are extracted through a digital elevation model; the soil type and vegetation coverage of the valley-type tailings Kut small watershed are identified through remote sensing image classification technology.

[0012] Preferably, the establishment of the underlying surface classification system includes: based on the acquired soil type , vegetation coverage ,slope and valley distribution Data, hierarchical cluster analysis method was used to classify the underlying surface and calculate the Euclidean distance between each factor ,in Represents different sampling points, according to the set threshold The samples with high similarity are classified into the same category to form several underlying surface units with similar characteristics, and the formula Determine the comprehensive characteristic index of each unit, where is the characteristic function; after determining the comprehensive characteristic index of each unit, the construction of the underlying surface classification system is completed.

[0013] Preferably, in S2, nonlinear key factors in the rainfall-runoff process in the valley terrain small watershed are extracted, and the historical rainfall and runoff data of the small watershed of the valley-type tailings pond are analyzed, and the nonlinear key factors of runoff generation and confluence in the valley terrain small watershed are extracted using a simulated annealing algorithm, and the factors that affect the formation of runoff, such as rainfall intensity, vegetation interception, soil permeability and underlying surface characteristics, are identified, and the nonlinear key factors that affect the formation of runoff under different rainfall conditions are determined;

[0014] The simulated annealing algorithm is used to extract the nonlinear key factors of runoff generation and confluence in small watersheds in valley terrain, and the objective function is defined according to the historical rainfall-runoff data set. , the formula is: ,in To observe the runoff, To simulate runoff for the model, is the total number of data sets; by setting the initial temperature and cooling coefficient , generates initial solutions randomly, and uses the Metropolis criterion to accept new solutions, gradually reducing the temperature until the preset termination condition or minimum temperature is reached , to optimize rainfall intensity , vegetation interception , soil permeability and underlying surface characteristics Parameters.

[0015] Preferably, the construction of the nonlinear confluence calculation numerical model is based on the identified nonlinear key inducements, and the finite volume method is used to discretize the hydrodynamic control equations, and the formula is: in Representative time The rainfall intensity, Representative time The amount of vegetation interception, Representative time The soil permeability, Representative time The comprehensive characteristic index of the underlying surface, is the time step; The parameters are obtained by calibration of historical data, and the runoff is affected by the changes in rainfall intensity, vegetation interception, soil permeability and underlying surface characteristics over time. Taking into account the nonlinear influence of the flow, a nonlinear confluence calculation numerical model was constructed, which can accurately simulate the rainfall-runoff response mechanism in the small watershed in the valley terrain.

[0016] Preferably, in S3, a distributed sensor monitoring network is constructed by using rainfall monitoring stations and runoff monitoring stations, and the rainfall monitoring stations are reasonably arranged in the valley-type tailings Kut small watershed. With runoff monitoring station , using the boundary-constrained gradient-free optimization algorithm to determine the optimal monitoring point location to maximize the monitoring coverage and ensure data accuracy; and runoff monitoring stations Collect data and calculate comprehensive monitoring indicators , the formula is: in and Rainfall monitoring stations and runoff monitoring stations The indicator weight coefficient is set according to the importance of the site and its influence on the runoff process; the comprehensive monitoring indicators A distributed sensor monitoring network that comprehensively reflects the dynamic changes of rainfall-runoff in the basin.

[0017] Preferably, the layout of the distributed sensor monitoring network is optimized by combining the boundary constraint gradient-free optimization algorithm, and the objective function is defined as , the formula is: ,in and are the actual locations of rainfall monitoring stations and runoff monitoring stations, and It is an ideal reference location for rainfall monitoring stations and runoff monitoring stations. By introducing boundary constraints to ensure that all monitoring points are located in the feasible area, a gradient-free optimization algorithm is used to iteratively search for the optimal solution to minimize the objective function value, thereby adjusting the location of each monitoring station, maximizing the coverage of the entire monitoring network and improving data collection accuracy.

[0018] Preferably, the rainfall intensity of the valley-type tailings Kut small watershed is obtained. , vegetation interception , soil permeability and underlying surface characteristics The controlling factors of the model are used to construct a layered three-dimensional hydrological-hydrodynamic model that couples rainfall runoff, slope runoff, and shallow groundwater flow. The controlling factors are determined through field observations and remote sensing data. The formula is: ,in is the comprehensive influence function of the underlying surface, are the parameters obtained by calibration with historical data, is the runoff; Describe the rainfall runoff process, combined with the slope runoff equation and the shallow groundwater flow equation A hierarchical three-dimensional hydrological-hydrodynamic model is formed that can reflect the interaction mechanism of rainfall-runoff-groundwater flow, in which is the depth of slope runoff, is the confluence area, is the friction coefficient, is the change of slope runoff depth, is the groundwater level change, is the permeability coefficient, For supply items, is the gradient operator.

[0019] Preferably, the risk assessment index system in S5 identifies the key disaster factors affecting the safety of the tailings pond through a comprehensive analysis of the tailings dam destruction mechanism, runoff characteristics and failure modes of flood control and drainage facilities, and determines the weight and threshold of each disaster factor using historical flood event data and layered three-dimensional hydrological-hydrodynamic model simulation results; uses data mining to clarify the impact of the interaction between disaster factors on the overall safety of the tailings pond, and forms a risk assessment index system for tailings pond threats by integrating dimensional indicators including flood level, burial depth of infiltration line, dry beach length, reliability of flood control and drainage structures and inherent risks of tailings ponds as risk factors.

[0020] Preferably, the flood risk prediction model for small watersheds of valley-type tailings ponds is constructed, the flood response process under different rainfall scenarios is simulated by a layered three-dimensional hydrological-hydrodynamic model, the parameters of the hydrological-hydrodynamic model are calibrated using historical flood data and real-time monitoring information, and the dynamic weighted Bayesian network is used in combination with risk factors to quantify the degree of influence of disaster-causing factors on flood risk, integrate the influence results, and form a comprehensive prediction model that accurately predicts the risk level of tailings ponds facing flood threats.

[0021] Compared with the prior art, the technical solution of this application has the following technical effects:

[0022] The present invention obtains high-resolution image data of the valley-type tailings Kut small watershed through the use of unmanned aerial vehicles combined with lidar multi-echo remote sensing technology, and establishes an underlying surface classification system based on the hierarchical cluster analysis method, which solves the problems of insufficient data collection accuracy and difficulty in fully reflecting the actual ground conditions in traditional methods. It can efficiently and accurately collect key information such as soil type and vegetation coverage, and can also calculate the Euclidean distance between factors and classify similar samples according to the set threshold to form underlying surface units with similar characteristics, thereby providing a solid foundation for subsequent modeling and significantly improving the input data quality of the risk prediction model.

[0023] The present invention uses a simulated annealing algorithm to extract the nonlinear key factors in the rainfall-runoff process in the valley terrain and small watershed, and constructs a nonlinear confluence calculation numerical model, which solves the limitations of the existing technology in dealing with complex nonlinear hydrological processes. The objective function is defined using historical rainfall-runoff data sets, and the Metropolis criterion is used to accept new solutions and gradually optimize parameters, so that the model can more accurately simulate the mechanism of runoff formation under different conditions, enhance the model's ability to respond to extreme weather events, and improve the accuracy and reliability of flood risk prediction.

[0024] The present invention constructs a distributed sensor monitoring network by reasonably arranging rainfall monitoring stations and runoff monitoring stations and using a boundary-constrained gradient-free optimization algorithm to determine the optimal monitoring point locations. This solves the problem of incomplete or distorted data collection caused by unreasonable layout in traditional monitoring methods, ensures the optimal distribution of monitoring stations, maximizes the monitoring coverage while also ensuring the quality of the data, thereby improving the effectiveness and accuracy of the entire system in monitoring the dynamic changes of rainfall and runoff in the basin, and providing more reliable data support for risk assessment.

[0025] The present invention integrates multiple dimensional indicators including flood level, burial depth of infiltration line, dry beach length, etc., and uses a dynamic weighted Bayesian network to build a comprehensive prediction model, which solves the problem that previous risk assessment methods are too simplified and cannot fully consider the interaction of multiple factors. It can quantify the impact of different disaster-causing factors on flood risk, and give a more accurate risk level prediction by integrating the impact results. Such a prediction model not only improves the safety management level of tailings ponds, but also provides a scientific basis for decision makers, which helps to take preventive measures in advance to reduce potential disaster risks.

[0026] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application so that it can be implemented in accordance with the contents of the specification, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following is a detailed description of the preferred embodiments of the present application in conjunction with the accompanying drawings as follows.

[0027] Based on the detailed description of the specific embodiments of the present application in combination with the accompanying drawings below, those skilled in the art will become more aware of the above and other objects, advantages and features of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can also be obtained based on these drawings without creative work. In all drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, each element or part is not necessarily drawn according to the actual scale.

[0029] Figure 1 This is a flow chart of a risk prediction method for a valley-type tailings pond threatened by floods in an extremely small watershed according to the present invention;

[0030] Figure 2 A system diagram of the classification system of the underlying surface of a valley-type tailings pond threatened by floods in a very small watershed according to the present invention;

[0031] Figure 3A topographical marking map of a valley-type tailings pond threatened by floods in a particularly small watershed according to the present invention;

[0032] Figure 4 A marking diagram of a monitoring station for a valley-type tailings pond threatened by a flood in an extremely small watershed according to the present invention;

[0033] Figure 5 A runoff variation diagram of a site of a valley-type tailings pond threatened by a flood in an extremely small watershed according to the present invention;

[0034] Figure 6 This is a comparison chart of the actual measurement and forecast of a valley-type tailings pond threatened by floods in an extremely small watershed according to the present invention. DETAILED DESCRIPTION

[0035] To make the purpose, technical scheme and advantages of the embodiment of the present application clearer, the technical scheme in the embodiment of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiment of the present application. Obviously, the described embodiment is a part of the embodiment of the present application, rather than all of the embodiments. In the following description, specific details such as specific configuration and components are provided only to help fully understand the embodiments of the present application. Therefore, it should be clear to those skilled in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. In addition, for clarity and brevity, the description of known functions and structures is omitted in the embodiment.

[0036] It should be understood that the references to "one embodiment" or "this embodiment" throughout the specification mean that the specific features, structures, or characteristics associated with the embodiment are included in at least one embodiment of the present application. Therefore, the references to "one embodiment" or "this embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0037] In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplicity and clarity, and does not in itself indicate the relationship between the various embodiments and / or settings discussed.

[0038] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist at the same time. The term " / and" in this article describes another type of association object relationship, indicating that there can be two relationships. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " in this article generally indicates that the previous and next associated objects are in an "or" relationship.

[0039] The term "at least one" in this article is merely a description of the association relationship of associated objects, indicating that there may be three relationships. For example, at least one of A and B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0040] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusions.

[0041] Example 1

[0042] This embodiment mainly describes a risk prediction method for a valley-type tailings pond threatened by a flood in a small watershed. Figure 1 As shown, including:

[0043] S1: Obtain the spatial distribution data of soil type, vegetation coverage, slope and gully in the valley-type Tailings Kut small watershed, and establish an underlying surface classification system;

[0044] Further, if Figure 2 As shown, the spatial distribution data of soil type, vegetation coverage, slope and gully in the valley-type Tailings Kut small watershed are obtained by collecting high-resolution images of the valley-type Tailings Kut small watershed through drones, and combining lidar multi-echo remote sensing technology to obtain three-dimensional surface information; the spatial distribution data of slope and gully are extracted through digital elevation models; the soil type and vegetation coverage of the valley-type Tailings Kut small watershed are identified through remote sensing image classification technology.

[0045] Furthermore, the establishment of the underlying surface classification system includes: based on the obtained soil types , vegetation coverage ,slope and valley distribution Data, hierarchical cluster analysis method was used to classify the underlying surface and calculate the Euclidean distance between each factor ,in Represents different sampling points, according to the set threshold The samples with high similarity are classified into the same category to form several underlying surface units with similar characteristics, and the formula Determine the comprehensive characteristic index of each unit, where is the characteristic function; after determining the comprehensive characteristic index of each unit, the construction of the underlying surface classification system is completed.

[0046] S2: Extract the nonlinear key factors in the rainfall-runoff process in the valley terrain and small watershed, use statistical learning methods to identify the dominant factors affecting runoff formation, and construct a nonlinear confluence calculation numerical model;

[0047] Furthermore, the nonlinear key factors in the rainfall-runoff process in the valley terrain small watershed are extracted. By analyzing the historical rainfall and runoff data of the small watershed of the valley-type tailings pond, the simulated annealing algorithm is used to extract the nonlinear key factors in the runoff generation and confluence of the valley terrain small watershed, identify the factors that affect the formation of runoff, such as rainfall intensity, vegetation interception, soil permeability and underlying surface characteristics, and determine the nonlinear key factors that affect the formation of runoff under different rainfall conditions.

[0048] The simulated annealing algorithm is used to extract the nonlinear key factors of runoff generation and confluence in small watersheds in valley terrain, and the objective function is defined according to the historical rainfall-runoff dataset. , the formula is: ,in To observe the runoff, To simulate runoff for the model, is the total number of data sets; by setting the initial temperature and cooling coefficient , generates initial solutions randomly, and uses the Metropolis criterion to accept new solutions, gradually reducing the temperature until the preset termination condition or minimum temperature is reached , to optimize rainfall intensity , vegetation interception , soil permeability and underlying surface characteristics Parameters.

[0049] Furthermore, the construction of the nonlinear confluence calculation numerical model is based on the identified nonlinear key factors, and the finite volume method is used to discretize the hydrodynamic control equations, which are as follows: in Representative time The rainfall intensity, Representative time The amount of vegetation interception, Representative time The soil permeability, Representative time The comprehensive characteristic index of the underlying surface, is the time step; The parameters are obtained by calibration of historical data, and the runoff is affected by the changes in rainfall intensity, vegetation interception, soil permeability and underlying surface characteristics over time. Taking into account the nonlinear influence of the flow, a nonlinear confluence calculation numerical model was constructed, which can accurately simulate the rainfall-runoff response mechanism in the small watershed in the valley terrain.

[0050] S3: Construct a distributed sensor monitoring network through rainfall monitoring stations and runoff monitoring stations;

[0051] Furthermore, a distributed sensor monitoring network is constructed through rainfall monitoring stations and runoff monitoring stations, and rainfall monitoring stations are reasonably arranged in the valley-type tailings Kut small watershed. With runoff monitoring station , using the boundary-constrained gradient-free optimization algorithm to determine the optimal monitoring point location to maximize monitoring coverage and ensure data accuracy; through the rainfall monitoring station and runoff monitoring stations Collect data and calculate comprehensive monitoring indicators , the formula is: in and Rainfall monitoring stations and runoff monitoring stations The indicator weight coefficient is set according to the importance of the site and its influence on the runoff process; through comprehensive monitoring indicators A distributed sensor monitoring network that fully reflects the dynamic changes of rainfall-runoff in the basin;

[0052] Furthermore, the layout of the distributed sensor monitoring network is optimized by combining the boundary constraint gradient-free optimization algorithm. By defining the objective function , the formula is: ,in and are the actual locations of rainfall monitoring stations and runoff monitoring stations, and It is an ideal reference location for rainfall monitoring stations and runoff monitoring stations. By introducing boundary constraints to ensure that all monitoring points are located in the feasible area, a gradient-free optimization algorithm is used to iteratively search for the optimal solution to minimize the objective function value, thereby adjusting the location of each monitoring station, maximizing the coverage of the entire monitoring network and improving data collection accuracy.

[0053] S4: Obtain the controlling factors of rainfall intensity, vegetation interception, and permeability, and construct a layered three-dimensional hydrological and hydrodynamic model that couples rainfall runoff-slope runoff-shallow groundwater flow;

[0054] Furthermore, the rainfall intensity of the valley-type Tailings Kut small watershed was obtained. , vegetation interception , soil permeability and underlying surface characteristics The controlling factors of the model are used to construct a layered three-dimensional hydrological-hydrodynamic model that couples rainfall runoff, slope runoff, and shallow groundwater flow. The controlling factors are determined through field observations and remote sensing data. The formula is: ,in is the comprehensive influence function of the underlying surface, are the parameters obtained by calibration with historical data, is the runoff; Describe the rainfall runoff process, combined with the slope runoff equation and the shallow groundwater flow equation A hierarchical three-dimensional hydrological-hydrodynamic model is formed that can reflect the interaction mechanism of rainfall-runoff-groundwater flow, in which is the depth of slope runoff, is the confluence area, is the friction coefficient, is the change of slope runoff depth, is the groundwater level change, is the permeability coefficient, For supply items, is the gradient operator.

[0055] S5: Through the layered three-dimensional hydrological and hydrodynamic model, combined with the factors affecting flood risk, a risk assessment index system and risk prediction model are established to achieve accurate prediction of flood risks in valley-type tailings Kut small watersheds.

[0056] Furthermore, the risk assessment index system in S5 identifies the key disaster factors that affect the safety of tailings dams through a comprehensive analysis of the tailings dam failure mechanism, runoff characteristics and failure modes of flood control and drainage facilities, and determines the weights and thresholds of each disaster factor using historical flood event data and layered three-dimensional hydrological-hydrodynamic model simulation results; uses data mining to clarify the impact of the interaction between disaster factors on the overall safety of the tailings dam, and forms a risk assessment index system for tailings dam threats by integrating dimensional indicators including flood level, burial depth of infiltration line, dry beach length, reliability of flood control and drainage structures and inherent risks of tailings dams as risk factors.

[0057] Furthermore, a flood risk prediction model for small watersheds of valley-type tailings ponds was constructed. The flood response process under different rainfall scenarios was simulated through a layered three-dimensional hydrological-hydrodynamic model. The parameters of the hydrological-hydrodynamic model were calibrated using historical flood data and real-time monitoring information. The dynamic weighted Bayesian network was used to combine risk factors. By quantifying the impact of disaster-causing factors on flood risk and integrating the impact results, a comprehensive prediction model was formed to accurately predict the risk level of tailings ponds facing flood threats.

[0058] This embodiment ensures the quality of basic data by acquiring high-resolution image data, establishes an underlying surface classification system using a hierarchical clustering analysis method, extracts nonlinear key factors through a simulated annealing algorithm and constructs a nonlinear confluence calculation numerical model, thereby achieving a more accurate description of the rainfall-runoff process; designs a distributed sensor monitoring network, optimizes the layout of monitoring points, and ensures the effectiveness and comprehensiveness of data collection; integrates historical data and real-time information, and uses a dynamic weighted Bayesian network to construct a risk prediction model, thereby greatly improving the prediction accuracy and reliability.

[0059] Example 2

[0060] This embodiment is based on Embodiment 1 and describes in detail an optimization scheme for constructing a distributed sensor monitoring network using rainfall monitoring stations and runoff monitoring stations, specifically including:

[0061] A distributed sensor monitoring network is built through rainfall monitoring stations and runoff monitoring stations. The observation data of the valley-type tailings pond small watershed is collected by using rainfall monitoring stations and runoff monitoring stations. The rain gauge, flow meter, level meter, and triangular weir equipment are combined to record and store the point rainfall, surface rainfall, rainfall type, rainfall distribution law of each duration, and time-varying characteristics of runoff generation in real time in the storage area.

[0062] Through edge computing embedded design, an adaptive monitoring data intelligent collection and transmission control method is developed with data interval collection and trigger transmission as the main method and timing device status collection and transmission as the auxiliary method.

[0063] The abnormal data in monitoring are located and eliminated through the anomaly detection algorithm, the missing monitoring data are supplemented by the correlation interpolation method, the Kalman filtering technology and wavelet packet algorithm are used to filter and smooth the noisy deformation data, and the control point and nearest point iterative adjustment algorithm is used to measure and adjust the sensor nodes of the rainfall and runoff monitoring modules to improve the data accuracy, spatiotemporal resolution and data integrity.

[0064] This embodiment describes in detail the data collection, transmission and filtering of the distributed sensor monitoring network constructed by rainfall monitoring stations and runoff monitoring stations. By filtering and smoothing the data, the accuracy and data integrity of the valley-type tailings pond small watershed data are improved, providing a better data flow for the subsequent evaluation model and prediction model.

[0065] Example 3

[0066] This embodiment, based on Embodiment 1, describes in detail the specific contents of the risk assessment index system and the prediction model, including:

[0067] When constructing a flood risk assessment index system for valley-type tailings ponds in small watersheds, it is necessary to identify key disaster factors that affect the safety of tailings ponds based on an in-depth analysis of the damage mechanism of tailings dams, runoff characteristics, and failure modes of flood control and drainage facilities, and through a comprehensive analysis of historical dam breach accident data and existing literature. Key disaster factors include but are not limited to rainfall intensity, vegetation coverage, soil permeability, terrain slope, natural factors of gully distribution, as well as the design standards of tailings ponds and the effectiveness of flood control and drainage structures;

[0068] After determining the key disaster factors, the impact of disaster factors on flood risk is quantified, and a layered three-dimensional hydrological-hydrodynamic model is used to simulate the flood response process under different rainfall scenarios. The model parameters are calibrated using historical flood data and real-time monitoring information to improve the prediction accuracy. The dynamic weighted Bayesian network method is used in combination with the flood level. , Depth of wetting line Length of dry beach , Reliability of flood control and drainage structures Key risk indicators, through the formula Calculate the flood risk, where They represent the weight coefficients of flood level, depth of infiltration line, length of dry beach and reliability of flood control and drainage structures respectively;

[0069] After quantifying the key risk indicators, they are integrated to form a risk assessment indicator system. The risk assessment indicator system includes indicators directly related to floods: flood level, depth of infiltration line, and length of dry beach. It also includes indirect influencing factors: the reliability of flood control and drainage structures, and considers the inherent risks of tailings ponds, such as the quality of dam construction materials and tailings particle size. All indicators together constitute a multi-dimensional risk assessment framework, which reflects the flood threat faced by valley-type tailings ponds. Through fuzzy theory and reliability analysis methods, the threshold and scoring criteria of each indicator are refined to form a risk assessment indicator system and prediction model.

[0070] This embodiment calculates the overall risk level based on the weight distribution in the risk assessment index system, and can provide quantitative risk assessment results. The prediction model can help decision makers identify potential risk points, so as to take effective preventive measures and reduce the losses caused by floods in valley-type tailings ponds.

[0071] Example 4

[0072] This embodiment is based on Embodiments 1-3 and illustrates a risk prediction method of a valley-type tailings pond threatened by floods in a small watershed in this application through detailed examples, specifically including:

[0073] like Figure 3As shown, a valley-type tailings pond area in Hubei Province was selected, which is located in a very small watershed with complex terrain, including different soil types, vegetation coverage, slope changes, and valley distribution;

[0074] The terrain of the valley tailings pond area was collected by drone with a resolution of 0.1 meters, and the collection coverage area was the entire tailings pond and its surrounding area of ​​1.2 kilometers;

[0075] The three-dimensional information of the valley-type tailings pond area was obtained through the laser radar multi-echo remote sensing technology, with an average error of no more than ±0.5 meters. The DEM data downloaded from the National Geographic Information System (NGIS) had a resolution of 1 meter. The images of the valley-type tailings pond area were obtained through remote sensing image classification technology. After analyzing the images, it was identified that the main soil types in the valley-type tailings pond area were red soil (accounting for 37.256%), yellow soil (accounting for 49.378%) and other types (13.366%). The vegetation coverage was about 72.652%, of which the forest coverage accounted for 42.775% and the grassland accounted for 26.524%.

[0076] The historical rainfall-runoff dataset in the valley tailings pond area was selected, the time period was selected from 2019 to 2023, the sample number was selected as 1826 daily record datasets, and the simulated annealing algorithm parameters were set, the initial temperature was 100°C, and the cooling coefficient was 0.95 for optimization. The optimized factor weights were: rainfall intensity (0.416), vegetation interception (0.215), and soil permeability (0.334);

[0077] The optimal solution was determined by performing 100 iterations of search to minimize the function value through the boundary constraint gradient-free optimization algorithm: the specific coordinates of 9 rainfall monitoring stations and 7 runoff monitoring stations were determined, so that the coverage rate of the entire monitoring network reached 98%, such as Figure 4 As shown, rainfall monitoring stations (A, B, C, D, E, F, G, H, I) and runoff monitoring stations (a, b, c, d, e, f, g) are set up in the area, and all stations are within the feasible area;

[0078] The rainfall-runoff data of the valley tailings pond area was collected through the rainfall monitoring station and the runoff monitoring station for 4 hours, and the water flow velocity at the station was obtained, such as Figure 5 As shown, through Figure 5The runoff changes at each station can be observed to obtain the controlling factors of rainfall intensity, vegetation interception, and permeability. According to the runoff data, the average value of vegetation interception is about 24.12 mm, with a standard deviation of 5 mm; the average value of soil permeability is about 0.001 cm / s, with a standard deviation of 0.0005. cm / s; the layered three-dimensional hydrological and hydrodynamic model of coupled rainfall runoff-slope runoff-shallow groundwater flow is used to determine the disaster factors in combination with flood risk influencing factors; through a comprehensive analysis of the tailings dam failure mechanism, runoff characteristics and failure mode of flood control and drainage facilities in valley-type tailings pond areas, the key disaster factors affecting the safety of tailings ponds are identified, and the weights and thresholds of each disaster factor are determined; data mining is used to clarify the impact of the interaction between disaster factors on the overall safety of tailings ponds, and a risk assessment index system for tailings pond threats is formed by integrating dimensional indicators including flood level (as shown in Table 2), burial depth of infiltration line, dry beach length, reliability of flood control and drainage structures (as shown in Table 3) and inherent risks of tailings ponds as risk factors; by quantifying the impact of disaster factors on flood risk and integrating the impact results, a comprehensive prediction model is formed to accurately predict the risk level of tailings ponds facing flood threats.

[0079] Table 2 Correspondence between flood level and inundation depth

[0080] Flood Level describe Submergence depth range (m) Risk level 1 Small flood <0.5 Lower 2 Moderate flood 0.5-1.0 medium 3 Major floods 1.0-2.0 Higher 4 Severe floods 2.0-3.0 high 5 Extreme flooding >3.0 Very high

[0081] Table 2 classifies flood levels according to the inundation depth within the tailings pond and assesses the corresponding risk level accordingly. A lower flood level means less risk, while a higher flood level indicates a more serious potential threat;

[0082] Table 3 Reliability evaluation table of flood control and drainage structures

[0083] Reliability Level describe Defining Standards Maintenance Recommendations A Very reliable All facilities meet the design standards without any defects Check regularly to keep in good condition B reliable The facilities are basically intact, but there are some minor aging or problems Monitor and fix minor problems in time, preventive maintenance C generally Some visible signs of ageing, requires regular maintenance Increase monitoring frequency and plan necessary fixes and improvements D Not very reliable Significantly aged or damaged, performance degraded Take immediate action to repair or replace E Unreliable Severely damaged, unable to work properly, causing safety issues Stop using immediately and perform a complete overhaul or rebuild

[0084] The reliability of tailings dam flood control and drainage structures is evaluated according to Table 3, ranging from very reliable to unreliable. Specific definition standards and maintenance recommendations are provided for each level to help managers understand the current status of the facilities and guide them on how to ensure the safety and effectiveness of these critical structures;

[0085] We selected 20% of the daily rainfall-runoff data from 2019 to 2023 as an independent test set, which covers rainfall events of different intensities and durations. We used the constructed hierarchical three-dimensional hydrological and hydrodynamic model to simulate and predict the test data of the test set. Figure 5 The predicted runoff volume for each station at each time point is shown.

[0086] like Figure 6As shown in the figure, the model-predicted runoff was compared with the actual observed runoff. The Pearson correlation coefficient was calculated to quantify the strength of the linear relationship between the two. The results showed that in the prediction test, the correlation coefficient between the model prediction value and the actual occurrence reached an average of more than 0.85, indicating that the model has high prediction accuracy and reliability.

[0087] This embodiment describes in detail the specific effects of a risk prediction method for valley-type tailings ponds threatened by floods in ultra-small watersheds. It uses high-resolution drone images, lidar data and digital elevation models, combined with historical rainfall-runoff data and real-time monitoring information, to construct a layered three-dimensional hydrological and hydrodynamic model, and quantifies the impact of key disaster-causing factors through a dynamic weighted Bayesian network. The results show that this method not only improves the accuracy of flood prediction, but also effectively supports decision makers to take preventive measures in advance, significantly reduces potential disaster risks, and improves the safety management level of tailings ponds. Through simulation tests, the correlation coefficient between the model prediction value and the actual occurrence reached 0.85, which can warn of flood events with major losses, proving the effectiveness and reliability of this application.

[0088] The above are only preferred embodiments of the present invention, which do not limit the scope of protection of the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any changes, modifications, replacements, integrations and parameter changes to these embodiments within the spirit and principles of the present invention through conventional substitutions or without departing from the principles and spirit of the present invention fall within the scope of protection of the present invention.

Claims

1. A risk prediction method for valley-type tailings ponds threatened by floods in small watersheds, characterized in that: include: S1: Obtain the spatial distribution data of soil type, vegetation coverage, slope and gully in the valley-type Tailings Kut small watershed, and establish an underlying surface classification system; S2: Extract the nonlinear key factors in the rainfall-runoff process in the valley terrain and small watershed, use statistical learning methods to identify the dominant factors affecting runoff formation, and construct a nonlinear confluence calculation numerical model; S3: Construct a distributed sensor monitoring network through rainfall monitoring stations and runoff monitoring stations; S4: Obtain the controlling factors of rainfall intensity, vegetation interception, and permeability, and construct a layered three-dimensional hydrological and hydrodynamic model that couples rainfall runoff-slope runoff-shallow groundwater flow; S5: Through the layered three-dimensional hydrological and hydrodynamic model, combined with the factors affecting flood risk, a risk assessment index system and risk prediction model are established to achieve accurate prediction of flood risk in the valley-type tailings Kut small watershed; The construction of the nonlinear confluence calculation numerical model in S2 is based on the identified nonlinear key causes and the finite volume method is used to discretize the hydrodynamic control equations. The formula is: ,in Representative time The rainfall intensity, Representative time The amount of vegetation interception, Representative time The soil permeability, Representative time The comprehensive characteristic index of the underlying surface, is the time step; The parameters are obtained by calibration of historical data, and the runoff is affected by the changes in rainfall intensity, vegetation interception, soil permeability and underlying surface characteristics over time. Taking into account the nonlinear influence of the flow, a nonlinear confluence calculation numerical model was constructed, which can accurately simulate the rainfall-runoff response mechanism in the small watershed in the valley terrain.

2. The risk prediction method for a valley-type tailings pond threatened by floods in a small watershed according to claim 1, characterized in that: The spatial distribution data of soil type, vegetation coverage, slope and gully in the valley-type tailings Kut small watershed are obtained by collecting high-resolution images of the valley-type tailings Kut small watershed through drones, and obtaining three-dimensional surface information in combination with lidar multi-echo remote sensing technology; the spatial distribution data of slope and gully are extracted through digital elevation models; the soil type and vegetation coverage of the valley-type tailings Kut small watershed are identified through remote sensing image classification technology.

3. The risk prediction method for a valley-type tailings pond threatened by floods in a small watershed according to claim 1 or 2, characterized in that: The establishment of the underlying surface classification system includes: based on the obtained soil types , vegetation coverage ,slope and valley distribution Data, hierarchical cluster analysis method was used to classify the underlying surface and calculate the Euclidean distance between each factor ,in Represents different sampling points, according to the set threshold The samples with high similarity are classified into the same category to form several underlying surface units with similar characteristics, and the formula Determine the comprehensive characteristic index of each unit, where is the characteristic function; after determining the comprehensive characteristic index of each unit, the construction of the underlying surface classification system is completed.

4. The risk prediction method for a valley-type tailings pond threatened by floods in a small watershed according to claim 1, characterized in that: In the S2, nonlinear key factors in the rainfall-runoff process in the valley terrain small watershed are extracted, and the historical rainfall and runoff data of the small watershed of the valley tailings pond are analyzed, and the simulated annealing algorithm is used to extract the nonlinear key factors of runoff generation and confluence in the valley terrain small watershed, identify the factors that affect the formation of runoff, such as rainfall intensity, vegetation interception, soil permeability and underlying surface characteristics, and determine the nonlinear key factors that affect the formation of runoff under different rainfall conditions; The simulated annealing algorithm is used to extract the nonlinear key factors of runoff generation and confluence in small watersheds with valley terrain, and the objective function is defined based on the historical rainfall-runoff dataset. , the formula is: ,in To observe the runoff, To simulate runoff for the model, is the total number of data sets; by setting the initial temperature and cooling coefficient , generates the initial solution randomly, and uses the Metropolis criterion to gradually reduce the temperature until the preset termination condition or minimum temperature is reached. , to optimize rainfall intensity , vegetation interception , soil permeability and underlying surface characteristics Parameters.

5. The risk prediction method for a valley-type tailings pond threatened by floods in a small watershed according to claim 1, characterized in that: In S3, a distributed sensor monitoring network is constructed by means of rainfall monitoring stations and runoff monitoring stations. With runoff monitoring station , using the boundary-constrained gradient-free optimization algorithm to determine the optimal monitoring point location to maximize the monitoring coverage and ensure data accuracy; Runoff monitoring station Collect data and calculate comprehensive monitoring indicators , the formula is: in and Rainfall monitoring stations and runoff monitoring stations The indicator weight coefficient is set according to the importance of the site and its influence on the runoff process; through comprehensive monitoring indicators A distributed sensor monitoring network that comprehensively reflects the dynamic changes of rainfall-runoff in the basin.

6. The risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds according to claim 5 is characterized in that: The distributed sensor monitoring network is optimized by using a boundary-constrained gradient-free optimization algorithm. , the formula is: ,in and are the actual locations of rainfall monitoring stations and runoff monitoring stations, and It is an ideal reference location for rainfall and runoff monitoring stations; By introducing boundary constraints to ensure that all monitoring points are located in the feasible area, the gradient-free optimization algorithm is used to iteratively search for the optimal solution to minimize the objective function value, thereby adjusting the position of each monitoring station to maximize the coverage of the entire monitoring network and improve the data collection accuracy.

7. The risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds according to claim 3 is characterized in that: The rainfall intensity of the valley-type Tailings Kut small watershed is obtained , vegetation interception , soil permeability and underlying surface characteristics The controlling factors of the model are used to construct a layered three-dimensional hydrological-hydrodynamic model that couples rainfall runoff, slope runoff, and shallow groundwater flow. The controlling factors are determined through field observations and remote sensing data. The formula is: ,in is the comprehensive influence function of the underlying surface, are the parameters obtained by calibration with historical data, is the runoff; Describe the rainfall runoff process, combined with the slope runoff equation and the shallow groundwater flow equation A hierarchical three-dimensional hydrological-hydrodynamic model is formed that can reflect the interaction mechanism of rainfall-runoff-groundwater flow, in which is the depth of slope runoff, is the confluence area, is the friction coefficient, is the change of slope runoff depth, is the groundwater level change, is the permeability coefficient, For supply items, is the gradient operator.

8. The risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds according to claim 1, characterized in that: The risk assessment index system in S5 identifies the key disaster factors that affect the safety of tailings dams through a comprehensive analysis of the tailings dam destruction mechanism, runoff characteristics and failure modes of flood control and drainage facilities, and determines the weights and thresholds of each disaster factor using historical flood event data and layered three-dimensional hydrological-hydrodynamic model simulation results; uses data mining to clarify the impact of the interaction between disaster factors on the overall safety of the tailings dam, and forms a risk assessment index system for tailings dam threats by integrating dimensional indicators including flood level, burial depth of infiltration line, dry beach length, reliability of flood control and drainage structures and inherent risks of tailings dams as risk factors.

9. The risk prediction method for valley-type tailings ponds threatened by floods in extremely small watersheds according to claim 1, characterized in that: In the S5, a risk prediction model for valley-type tailings pond small watershed is constructed. The flood response process under different rainfall scenarios is simulated through a layered three-dimensional hydrological-hydrodynamic model. The parameters of the hydrological-hydrodynamic model are calibrated using historical flood data and real-time monitoring information. The dynamic weighted Bayesian network is used to combine risk factors. By quantifying the impact of disaster-causing factors on flood risk, the impact results are integrated to form a comprehensive prediction model that accurately predicts the risk level of tailings ponds facing flood threats.

Images