A Disaster Hazard Identification and Early Warning Method Based on Multi-Dimensional Data Fusion of Tailings Dams

Abstract

This invention discloses a disaster hazard identification and early warning method for tailings dams based on multidimensional data fusion, relating to the field of disaster hazard identification and early warning. The method includes: fusing multidimensional datasets to generate multidimensional data fusion results; rapidly extrapolating multiple environmental parameters and the multidimensional data fusion results; constructing a mathematical model for assessing the disaster impact of tailings dams to conduct risk assessments, generating multiple risk levels; analyzing tailings dam failure triggering criteria; generating multiple triggering indicators for tailings dams and conducting sensitivity analysis; determining disaster triggering sensitivity coefficients and the multiple risk levels to identify disaster hazards; generating disaster early warning signals and combining them with the multidimensional dataset to issue disaster alerts; and formulating emergency response measures to respond to disaster hazards in tailings dams. This method solves the technical problems of real-time analysis and comprehensive assessment of tailings dam data, achieving the technical effect of improving the accuracy of disaster hazard identification and effectively conducting early warning responses.

Description

Technical Field

[0001] This application relates to the field of disaster hazard identification and early warning, and in particular to a disaster hazard identification and early warning method based on multi-dimensional data fusion of tailings ponds. Background Technology

[0002] Tailings dams are common waste storage facilities in mining production, primarily used to store tailings generated during mining operations. Accidents involving tailings dams, such as dam failures, landslides, and leaks, can severely impact the ecological environment and surrounding communities. In recent years, due to the large size and complex structure of tailings dams, potential hazards have been difficult to detect in advance, and the consequences of such disasters can be disastrous. Therefore, effectively monitoring the stability of tailings dams, promptly identifying potential hazards, and providing effective early warnings are pressing technical challenges in tailings dam management.

[0003] Traditional tailings dam monitoring methods primarily rely on manual inspections or single data sources. These methods suffer from poor real-time performance, inaccurate data, and a lack of comprehensive assessment capabilities, making them ineffective in addressing complex disaster risks. Furthermore, due to the complex environmental factors affecting tailings dams, such as meteorological variations, hydrological changes, and soil properties, a single data source cannot comprehensively reflect the operational status and disaster risks of the tailings dam. Therefore, traditional tailings dam disaster monitoring methods cannot meet the needs of efficient early warning and emergency response. Summary of the Invention

[0004] This application provides a disaster hazard identification and early warning method based on multi-dimensional data fusion in tailings ponds. It solves the technical problem of real-time analysis and comprehensive evaluation of various monitoring data in tailings ponds, and achieves the technical effect of combining disaster early warning signals and emergency response measures to identify potential disaster risks in tailings ponds in advance, generate disaster early warning signals in a timely manner, and formulate effective emergency response strategies, thereby significantly improving the accuracy of tailings pond disaster hazard identification.

[0005] This application provides a disaster hazard identification and early warning method for tailings dams based on multi-dimensional data fusion, comprising: acquiring multi-dimensional initial datasets by using a group of sensor devices to collect data from the tailings dam in multiple dimensions; synchronizing the initial datasets to determine the multi-dimensional dataset; retrieving multiple environmental parameters of the tailings dam; fusing the multi-dimensional datasets to generate multi-dimensional data fusion results; using dynamic numerical simulation technology to rapidly extrapolate the multiple environmental parameters and the multi-dimensional data fusion results to construct a mathematical model for assessing the disaster impact of the tailings dam; conducting a risk assessment of the tailings dam using the mathematical model, generating multiple risk levels, and analyzing the tailings dam failure triggering criteria to generate multiple triggering indicators; performing sensitivity analysis according to the multiple triggering indicators to determine disaster triggering sensitivity coefficients; identifying disaster hazards based on the disaster triggering sensitivity coefficients and the multiple risk levels to generate disaster early warning signals; issuing disaster alerts based on the disaster early warning signals and the multi-dimensional datasets; formulating emergency response measures; and executing the emergency response measures to respond to the disaster hazards of the tailings dam.

[0006] In a possible implementation, the multidimensional initial dataset is synchronized to determine the multidimensional dataset, and the following processing is performed: wavelet decomposition is performed on the multidimensional data signal of the multidimensional initial dataset to obtain wavelet coefficients of the data signal; threshold quantization is performed on the wavelet coefficients of the data signal to determine the wavelet selection threshold of the data signal; the wavelet coefficients of the data signal are truncated according to the wavelet selection threshold of the data signal, and noise signals smaller than the wavelet selection threshold of the data signal are set to zero to obtain a valid signal greater than the wavelet selection threshold of the data signal; the valid signal is filtered and reconstructed to obtain the multidimensional dataset.

[0007] In a possible implementation, the multidimensional dataset is fused to generate a multidimensional data fusion result, and the following processing is performed: the cooperative data array distribution probability of the effective signals in the multidimensional dataset is calculated to obtain a distribution probability array; a dot matrix similarity probability distribution function is constructed to reduce the dimensionality of the dot matrix distribution of the distribution probability array to generate a dot matrix dimensionality reduction result; and the multidimensional data fusion result is obtained by fusing the data based on the dot matrix dimensionality reduction result.

[0008] In one possible implementation, dynamic numerical simulation technology is used to rapidly extrapolate the fusion results of the multiple environmental parameters and the multidimensional data, constructing a mathematical model for assessing the catastrophic impact of the tailings dam. The following processes are performed: establishing a database of the physical and mechanical properties of tailings based on the three-dimensional structural evolution of the tailings dam; performing cluster analysis on the tailings dam according to the database to identify multiple mechanical distribution parameters, including similar and dissimilar distribution parameters; calculating the generalized layered structure data of the tailings dam based on the similar and dissimilar distribution parameters combined with pore water pressure monitoring data; using dynamic numerical simulation technology to dynamically simulate and extrapolate the fusion results of the multidimensional data of the tailings dam according to the generalized layered structure data, obtaining multiple catastrophic projection conditions; performing cross-validation based on the multiple catastrophic projection conditions, and constructing the mathematical model for assessing the catastrophic impact of the tailings dam based on the validation results.

[0009] In a possible implementation, the tailings dam is risk-assessed using the tailings dam disaster impact assessment mathematical model, generating multiple risk levels. The tailings dam failure triggering criteria are analyzed to generate multiple triggering indicators for the tailings dam. The following processes are then performed: The tailings dam disaster impact assessment mathematical model is activated to perform a disaster impact numerical simulation of the tailings dam, generating multiple disaster impact scores; a comprehensive risk indicator assessment is conducted based on the multiple disaster impact scores and the generalized hierarchical structure data of the tailings dam, generating a multi-layered risk assessment result; a risk analysis of the tailings dam is performed based on the multi-layered risk assessment result, dividing it into multiple risk levels; the stability of the tailings dam is calculated based on the multiple risk levels to obtain a stability coefficient; a dam failure trigger test is conducted based on the stability coefficient, and the multiple triggering indicators are set according to the test results.

[0010] In a possible implementation, sensitivity analysis is performed according to the multiple triggering indicators to determine the disaster triggering sensitivity coefficient, and the following processing is performed: the tailings dam is co-evolved according to the multiple triggering indicators to generate multi-indicator co-evolution results; sensitivity analysis is performed based on the multi-indicator co-evolution results to determine the disaster triggering sensitivity coefficient.

[0011] In a possible implementation, disaster hazard identification is performed based on the disaster trigger sensitivity coefficient and the multiple risk levels to generate a disaster early warning signal. The following processing is then performed: a weighted calculation is performed based on the disaster trigger sensitivity coefficient and the multiple risk levels to generate multiple weight coefficients; the multiple triggering indicators are matched and calculated according to the multiple weight coefficients to obtain a comprehensive risk score; the disaster trigger sensitivity coefficient and the multiple risk levels are correlated with disaster analysis to obtain a preset risk threshold, and it is determined whether the comprehensive risk score is greater than or equal to the preset risk threshold; when the comprehensive risk score is greater than or equal to the preset risk threshold, an early warning level is set according to the comprehensive risk score, and the early warning level is added to the disaster early warning signal.

[0012] In a possible implementation, a disaster warning is issued based on the disaster early warning signal and the multidimensional dataset. Emergency response measures are then formulated, and the following processes are performed: the warning level is mapped to the multidimensional dataset for disaster analysis, and a disaster rescue time limit is set; rescue feedback is provided to the multidimensional data of the tailings dam according to the disaster rescue time limit, and emergency feedback information is generated; the multidimensional dataset is monitored and adjusted in real time based on the emergency feedback information, and the emergency response measures are formulated.

[0013] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0014] The tailings dam multi-dimensional data fusion disaster hazard identification and early warning method provided in this application solves the technical problem of real-time analysis and comprehensive evaluation of various monitoring data in tailings dams. It achieves the technical effect of combining disaster early warning signals and emergency response measures to identify potential disaster risks in tailings dams in advance, generate disaster early warning signals in a timely manner, and formulate effective emergency response strategies, thereby significantly improving the accuracy of tailings dam disaster hazard identification. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments of the present invention will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the method according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0016] Figure 1 A flowchart illustrating the disaster hazard identification and early warning method based on multi-dimensional data fusion for tailings ponds provided in this application embodiment;

[0017] Figure 2This is a schematic diagram of the mathematical model for assessing the disaster impact of a tailings dam, illustrating the construction of a disaster hazard identification and early warning method based on multi-dimensional data fusion for tailings dams provided in this application embodiment. Detailed Implementation

[0018] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application.

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] In the following description, references to "some embodiments" describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. The terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules not explicitly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only.

[0021] This application provides a method for identifying and warning of potential disaster hazards based on multi-dimensional data fusion from tailings ponds, such as... Figure 1 As shown, the method includes:

[0022] Step A100: Multi-dimensional sensing data is collected from the tailings dam using a group of sensing devices to obtain a multi-dimensional initial dataset. Data synchronization is performed on the multi-dimensional initial dataset to determine the multi-dimensional dataset. In one possible implementation, step A100 further includes step A110: Wavelet decomposition is performed on the multi-dimensional data signal of the multi-dimensional initial dataset to obtain wavelet coefficients. Step A120: Threshold quantization is performed based on the wavelet coefficients to determine a wavelet selection threshold. Step A130: The wavelet coefficients are truncated according to the wavelet selection threshold, and noise signals smaller than the wavelet selection threshold are set to zero to obtain a valid signal larger than the wavelet selection threshold. Step A140: The valid signal is filtered and reconstructed to obtain the multi-dimensional dataset.

[0023] By using wavelet transform, the original signal is decomposed into multiple wavelet coefficients. These wavelet coefficients characterize the signal's features at different scales (or frequencies). The multidimensional data signal of the initial multidimensional dataset is decomposed into different frequency components, thus obtaining the wavelet coefficients of the data signal. These wavelet coefficients contain information about the original signal at different frequencies. Furthermore, based on the statistical characteristics of the wavelet coefficients (such as standard deviation and mean), a suitable threshold is selected to determine the wavelet selection threshold. This threshold is used to distinguish between valid and noise signals, allowing for better truncation of subsequent signals. Simultaneously, to remove noise and retain valid signals, each wavelet coefficient is compared with the threshold. If the wavelet coefficients are less than the threshold, they are considered noise signals and are set to zero; if the wavelet coefficients are greater than or equal to the threshold, they are considered valid signals and their values ​​are retained. Finally, a set of wavelet coefficients containing only valid signals is obtained, and noise signals are effectively removed, resulting in valid signals. In order to reconstruct the valid part of the original signal from the truncated wavelet coefficients, the truncated wavelet coefficients can be converted back to the spatial or time domain of the original signal through inverse wavelet transform, thereby obtaining the multidimensional dataset, which is the denoised initial multidimensional dataset.

[0024] Execute step A200, retrieve multiple environmental parameters of the tailings pond, perform data fusion on the multi-dimensional dataset to generate a multi-dimensional data fusion result, and use dynamic numerical simulation technology to quickly deduce the multiple environmental parameters and the multi-dimensional data fusion result, and construct a mathematical model for evaluating the impact of tailings pond disasters; In a possible implementation manner, step A200 further includes step A210, calculate and obtain the co-occurrence data array distribution probability of the effective signals in the multi-dimensional dataset to obtain a distribution probability array; Execute step A220, construct a dot matrix similarity probability distribution function, perform dimensionality reduction processing on the dot matrix distribution of the distribution probability array to generate a dot matrix dimensionality reduction result; Execute step A230, perform fusion based on the dot matrix dimensionality reduction result to obtain the multi-dimensional data fusion result.

[0025] To evaluate the协同性 or相关性 between effective signals in a multi-dimensional dataset, statistical methods or machine learning algorithms such as covariance matrix, correlation analysis, mutual information, etc. can be used to calculate the similarity or关联性 between data points, and construct a distribution probability array, where each element represents the probability of co-occurrence between data points, to obtain the distribution probability array, and the distribution probability array is used to characterize the协同分布特性 of effective signals in the multi-dimensional dataset. Further, construct a dot matrix similarity probability distribution function, and the dot matrix similarity probability distribution function is as follows:

[0026] L = ∑ i ∑ j KL[P(x i |x j )||Q(z i |z j )];

[0027] Among them, KL is a measure index characterizing the proximity between the dot matrix distribution of the distribution probability array and the dimensionality-reduced dot matrix distribution of the distribution probability array. Taking KL approaching 0 as a constraint condition for dimensionality reduction processing, L is a divergence characterizing the similarity between the dot matrix distribution of the distribution probability array and the dimensionality-reduced dot matrix distribution of the distribution probability array. P(x i |x j ) is the similarity probability regarding dot matrix x i and dot matrix x j in the dimensionality-reduced dot matrix distribution based on the distribution probability array. Q(z i |z j ) is the similarity probability regarding dot matrix z i and dot matrix z j in the dot matrix distribution based on the distribution probability array. (i, j) is any coordinate point in the co-occurrence data array of the effective signals. The value range of i is 0 < i < n, the value range of j is 0 < j < n, and n is the total number of coordinate points in the co-occurrence data array of the effective signals.

[0028] The dimensionality of the point distribution array is reduced using a point similarity probability distribution function. This generates a dimensionality-reduced point set, which represents the point set of the original data in a lower-dimensional space. Each point retains some key features or similarity information from the original data. For further analysis or visualization, the dimensionality-reduced data points need to be fused into a comprehensive representation. This can be achieved by performing clustering, classification, regression, or other forms of machine learning tasks on the dimensionality-reduced point set, resulting in a multi-dimensional data fusion result. This multi-dimensional data fusion result can be a single numerical value, a vector, a matrix, or a more complex structure, depending on the fusion method and purpose, providing a strong data foundation for subsequent disaster hazard identification.

[0029] In one possible implementation, such as Figure 2 As shown, step A200 further includes step A240, establishing a tailings physical and mechanical property database based on the three-dimensional structural evolution law of the tailings dam; executing step A250, performing cluster analysis on the tailings dam according to the tailings physical and mechanical property database to identify multiple mechanical distribution parameters, including similar distribution parameters and dissimilar distribution parameters; executing step A260, calculating and determining the generalized layered structure data of the tailings dam based on the similar distribution parameters, the dissimilar distribution parameters, and pore water pressure monitoring data; executing step A270, using dynamic numerical simulation technology to dynamically simulate and extrapolate the multi-dimensional data fusion results of the tailings dam according to the generalized layered structure data of the tailings dam to obtain multiple disaster extrapolation conditions; executing step A280, performing cross-validation based on the multiple disaster extrapolation conditions, and constructing the tailings dam disaster impact assessment mathematical model based on the validation results.

[0030] Data on the physical and mechanical properties of tailings were collected and organized. Through laboratory testing, field monitoring, and literature review, physical and mechanical parameters such as density, strength, deformation modulus, and permeability coefficient of the tailings were obtained, and a corresponding database was established. Then, clustering algorithms (such as K-means and hierarchical clustering) were used to analyze the database of tailings physical and mechanical properties, dividing the tailings pond into regions with similar or dissimilar mechanical properties. These regions could correspond to different tailings deposit layers, different mineral compositions, or different geological conditions. Regions with similar distribution parameters were merged into one layer, while regions with dissimilar distribution parameters were divided into different layers. Furthermore, considering pore water pressure monitoring data and the internal moisture distribution and permeability characteristics of the tailings pond, the layered structure was further refined to determine the generalized layered structure data of the tailings pond.

[0031] Furthermore, by employing dynamic numerical simulation technology, the evolution of tailings dams under different conditions is simulated to predict potential disaster risks. This involves using finite element analysis (FEA), discrete element analysis (DEM), or other numerical simulation methods, combined with generalized hierarchical structural data of the tailings dam, to conduct dynamic simulation and extrapolation of the tailings dam. During the extrapolation process, the influence of factors such as the physical and mechanical properties of the tailings, moisture distribution, and external loads is considered. The consistency and rationality of the simulation results are then verified by comparing them under different disaster extrapolation conditions. Simultaneously, the extrapolation results are further verified and adjusted by combining on-site monitoring data and historical disaster cases. Multiple disaster extrapolation conditions are set to ensure the accuracy and reliability of the disaster extrapolation results. Finally, by combining statistical, machine learning, or data mining techniques, multiple disaster extrapolation conditions can be cross-validated to generate cross-validation results. Based on the cross-validation results, a tailings dam disaster impact assessment model is constructed. The tailings dam disaster impact assessment model can output the disaster risk level or probability of the tailings dam under different conditions based on various factors such as the structural characteristics, physical and mechanical properties, moisture distribution, and external loads of the tailings dam, thereby improving the accuracy of the assessment of the disaster impact of tailings dams.

[0032] Step A300 involves conducting a risk assessment of the tailings dam using the tailings dam disaster impact assessment mathematical model, generating multiple risk levels, and analyzing the tailings dam failure triggering criteria to generate multiple triggering indicators for the tailings dam. In one possible implementation, step A300 further includes step A310, activating the tailings dam disaster impact assessment mathematical model to perform a disaster impact numerical simulation of the tailings dam, generating multiple disaster impact scores. Step A320 involves conducting a comprehensive risk index assessment based on the multiple disaster impact scores combined with the generalized hierarchical structure data of the tailings dam, generating a multi-layered risk assessment result. Step A330 involves conducting a risk analysis of the tailings dam based on the multi-layered risk assessment result, classifying it into multiple risk levels. Step A340 involves calculating the stability of the tailings dam based on the multiple risk levels, obtaining the stability coefficient of the tailings dam. Step A350 involves conducting a dam failure trigger test based on the stability coefficient, and setting the multiple triggering indicators based on the test results.

[0033] The tailings dam disaster impact assessment mathematical model is activated. Using this activated model, numerical simulations of the tailings dam's disaster impact under different conditions are performed. By inputting data such as the tailings dam's structural parameters, physical and mechanical properties, moisture distribution, and external loads, the mathematical model is run to generate multiple disaster impact scores. These scores are used to assess the potential disaster risks of the tailings dam under different conditions. Further, by combining the tailings dam's generalized hierarchical structure data, a comprehensive evaluation of these scores is conducted to generate a multi-layered risk assessment result. This involves combining the disaster impact scores with the tailings dam's hierarchical structure data, analyzing the mutual influence and constraints between different layers, and using methods such as weighted averaging and fuzzy comprehensive evaluation to calculate the risk assessment result for each layer. Finally, the risk assessment results from each layer are integrated to form a multi-layered risk assessment report. Based on the multi-layered risk assessment results, the tailings dam is divided into multiple risk levels for targeted risk management. This involves setting risk level classification standards, such as low risk, medium risk, high risk, and extremely high risk. Based on the results of multi-level risk assessment, each layer of the tailings dam or the entire dam is classified into the corresponding risk level.

[0034] Further quantitative calculations of tailings dam stability based on multiple risk levels involve using limit equilibrium methods, finite element methods, or other stability calculation methods. By inputting structural parameters and physical and mechanical properties of the tailings dam, a stability coefficient can be calculated. This stability coefficient reflects the stability state of the tailings dam under current conditions. Finally, a series of dam-break triggering conditions are set through dam-break trigger tests, such as rising water levels, increased rainfall, and earthquakes. Under each condition, numerical simulation models are run or physical experiments are conducted to observe the tailings dam's response. When a dam break occurs, the corresponding triggering conditions are recorded as dam-break triggering indicators. The dam-break triggering test results are then analyzed to determine the relationship between different triggering conditions and dam-break risk. Based on the analysis results, reasonable dam-break triggering indicators are set, such as water level thresholds, rainfall thresholds, and seismic intensity. These indicators will be used in the daily safety monitoring and early warning system of the tailings dam to promptly detect and respond to potential safety risks.

[0035] Step A400 involves performing sensitivity analysis based on the multiple triggering indicators to determine the disaster triggering sensitivity coefficient. Based on the disaster triggering sensitivity coefficient and the multiple risk levels, disaster hazard identification is performed, and a disaster early warning signal is generated. In one possible implementation, step A400 further includes step A410, performing co-evolution of the tailings dam according to the multiple triggering indicators to generate a multi-indicator co-evolution result. Step A420 involves performing sensitivity analysis based on the multi-indicator co-evolution result to determine the disaster triggering sensitivity coefficient.

[0036] Multiple triggering indicators (such as water level, rainfall, seismic intensity, and changes in the physical and mechanical properties of tailings) are used as input variables. Numerical simulation models or system dynamics models are employed to simulate the evolution of the tailings dam. By adjusting the values ​​of the input variables, changes in the tailings dam's response are observed, simulating the evolution of the tailings dam under the combined influence of different triggering indicators. This generates multi-indicator co-evolution results, which can be visualized using time series plots, scatter plots, and heat maps. Comparative analysis identifies key triggering indicators and their interactions, providing a foundation for subsequent sensitivity analysis.

[0037] Further sensitivity analysis methods, such as local sensitivity analysis (e.g., differential method, finite difference method) or global sensitivity analysis (e.g., analysis of variance, sensitivity index method), can be used to further process the results of the co-evolution of multiple indicators. By calculating the influence of each triggering indicator on the probability or severity of tailings dam disasters, a disaster triggering sensitivity coefficient is determined. This coefficient may include a water level sensitivity coefficient and a settlement sensitivity coefficient. The water level sensitivity coefficient characterizes the risk of dam failure when water level changes reach a certain critical point, while the settlement sensitivity coefficient characterizes the possibility that rapid settlement indicates structural problems in the tailings dam, increasing the probability of disaster. Quantifying the sensitivity of each triggering indicator to tailings dam disasters provides a scientific basis for the safety management and early warning of tailings dams.

[0038] In one possible implementation, step A400 further includes step A430, which involves performing a weighted calculation based on the disaster trigger sensitivity coefficient and the multiple risk levels to generate multiple weight coefficients; step A440, which involves matching and calculating the multiple trigger indicators according to the multiple weight coefficients to obtain a comprehensive risk score; step A450, which involves performing a correlation disaster analysis between the disaster trigger sensitivity coefficient and the multiple risk levels to obtain a preset risk threshold, and determining whether the comprehensive risk score is greater than or equal to the preset risk threshold; and step A460, which involves setting an early warning level based on the comprehensive risk score when the comprehensive risk score is greater than or equal to the preset risk threshold, and adding the early warning level to the disaster early warning signal.

[0039] The disaster trigger sensitivity coefficient is weighted and calculated with various risk levels to identify the differences in the sensitivity of trigger indicators under different risk levels. A reasonable weight is assigned to each trigger indicator, generating multiple weight coefficients. The calculation process of these weight coefficients can include methods such as expert scoring, statistical analysis, or machine learning to reflect their relative importance in the tailings dam disaster risk. Furthermore, the actual value of each trigger indicator is multiplied by its corresponding weight coefficient, and all products are summed to obtain a comprehensive risk score. This comprehensive risk score is used to characterize the contribution of different trigger indicators to the tailings dam disaster risk.

[0040] By associating disaster trigger sensitivity coefficients and risk levels, a preset risk threshold is set. The preset risk threshold can be determined through methods such as historical disaster data, expert experience, and statistical analysis. The preset risk threshold is used to determine whether the disaster risk of the tailings dam has reached the level that requires early warning. It can characterize the disaster risk level of the tailings dam under different conditions and is used to trigger corresponding early warning signals.

[0041] Further, the comprehensive risk score is compared with a preset risk threshold to determine whether the tailings dam's disaster risk reaches the warning level. If the comprehensive risk score is greater than or equal to the preset risk threshold, the tailings dam is considered to have a high disaster risk, requiring the triggering of a warning signal. If the comprehensive risk score is less than the preset risk threshold, the tailings dam's disaster risk is considered to be within an acceptable range, and a warning signal is not required for the time being. When the comprehensive risk score reaches or exceeds the preset risk threshold, a corresponding warning level is set based on the comprehensive risk score and added to the disaster warning signal. This means determining the corresponding warning level (such as Level 1 warning, Level 2 warning, etc.) based on the magnitude of the comprehensive risk score and the preset warning level classification standards. Then, the warning level is combined with the disaster warning signal, and relevant personnel are promptly notified via SMS, email, automatic alarm systems, etc., so that they can take timely countermeasures.

[0042] Finally, step A500 is executed, in which a disaster warning is issued based on the disaster early warning signal and the multidimensional dataset, emergency response measures are formulated, and the emergency response measures are executed to respond to the disaster risks of the tailings dam.

[0043] In one possible implementation, step A500 further includes step A510, mapping the early warning level to the multidimensional dataset for disaster analysis and setting a disaster rescue time limit; executing step A520, providing rescue feedback on the multidimensional data of the tailings dam according to the disaster rescue time limit and generating emergency feedback information; executing step A530, performing real-time monitoring and adjustment of the multidimensional dataset based on the emergency feedback information and formulating the emergency response measures.

[0044] Utilizing data analysis tools (such as data mining and machine learning algorithms) to process multidimensional datasets involves combining early warning levels with multidimensional datasets (which may include data on the tailings dam's geography, geology, structure, environment, and operations) for in-depth disaster analysis. This analysis, based on early warning levels, identifies key disaster risk factors, potential disaster paths, and the scope of disaster impact, leading to a more comprehensive understanding of the disaster risks and possible consequences faced by the tailings dam. Simultaneously, based on the tailings dam's disaster characteristics, the availability of rescue resources, and the urgency of the disaster's impact, a disaster rescue timeframe is set. This timeframe should be sufficiently urgent to prompt rapid action from relevant personnel, but not so urgent as to compromise rescue effectiveness, ensuring that effective rescue measures can be implemented quickly before or after the disaster to minimize losses.

[0045] By utilizing real-time monitoring systems and data analysis tools, multi-dimensional data of tailings ponds are monitored and fed back in real time within the set disaster rescue time limit. Emergency feedback information is generated in a timely manner, including disaster development trends, rescue resource needs, and personnel safety status, so as to obtain the latest disaster information and rescue progress and provide a basis for formulating and adjusting emergency response measures.

[0046] Finally, the multidimensional dataset is updated and optimized based on emergency feedback information to ensure data accuracy and timeliness. Simultaneously, emergency response measures are formulated or adjusted according to the latest disaster situation and rescue needs, including evacuation plans, allocation of relief supplies, and personnel deployment, to reflect the latest disaster situation and rescue progress, and emergency response measures are formulated and adjusted accordingly.

[0047] The embodiments of this application solve the technical problem of real-time analysis and comprehensive evaluation of various monitoring data in tailings dams, and achieve the technical effect of combining disaster early warning signals and emergency response measures to identify potential disaster risks in tailings dams in advance, generate disaster early warning signals in a timely manner, and formulate effective emergency response strategies, thereby significantly improving the accuracy of identifying potential disaster hazards in tailings dams.

[0048] The specific embodiments described above do not constitute a limitation on the scope of protection of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application. In some cases, the actions or steps described in this application can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A method for identifying and warning of potential disaster hazards based on multi-dimensional data fusion from tailings dams, characterized in that, The method includes: The tailings dam is subjected to multi-dimensional sensing data acquisition by a group of sensing devices to obtain a multi-dimensional initial dataset. The multi-dimensional initial dataset is then synchronized to determine the multi-dimensional dataset. Multiple environmental parameters of the tailings dam are retrieved, the multidimensional dataset is fused to generate a multidimensional data fusion result, and dynamic numerical simulation technology is used to quickly extrapolate the multiple environmental parameters and the multidimensional data fusion result to construct a mathematical model for assessing the disaster impact of the tailings dam. The tailings dam disaster impact assessment mathematical model is used to conduct a risk assessment of the tailings dam, generate multiple risk levels, and analyze the tailings dam failure trigger criteria to generate multiple trigger indicators for the tailings dam. Sensitivity analysis is performed on the multiple triggering indicators to determine the disaster triggering sensitivity coefficient. Based on the disaster triggering sensitivity coefficient and the multiple risk levels, disaster risks are identified and disaster early warning signals are generated. Based on the disaster early warning signal and the multidimensional dataset, a disaster warning is issued, emergency response measures are formulated, and the emergency response measures are implemented to respond to the disaster risks of the tailings dam. A database of the physical and mechanical properties of tailings was established based on the three-dimensional structural evolution law of tailings ponds; Cluster analysis was performed on the tailings dam according to the tailings physical and mechanical properties database to identify multiple mechanical distribution parameters, including similar distribution parameters and dissimilar distribution parameters. Based on the similarity distribution parameters, the dissimilar distribution parameters, and pore water pressure monitoring data, the generalized stratified structure data of the tailings dam is determined. Based on the generalized hierarchical structure data of the tailings dam, dynamic numerical simulation technology is used to dynamically simulate and extrapolate the multidimensional data fusion results of the tailings dam to obtain multiple disaster extrapolation conditions. Based on the cross-validation of the multiple disaster simulation conditions, a mathematical model for assessing the disaster impact of the tailings dam is constructed according to the validation results.

2. The disaster hazard identification and early warning method based on multi-dimensional data fusion of tailings dams as described in claim 1, characterized in that, The method for synchronizing the initial multidimensional dataset and determining the multidimensional dataset includes: Wavelet decomposition is performed on the multidimensional data signal of the multidimensional initial dataset to obtain the wavelet coefficients of the data signal. Threshold quantization is performed based on the wavelet coefficients of the data signal to determine the wavelet selection threshold of the data signal; The wavelet coefficients of the data signal are truncated according to the wavelet selection threshold of the data signal, and the noise signals smaller than the wavelet selection threshold of the data signal are set to zero to obtain the effective signal greater than the wavelet selection threshold of the data signal. The effective signal is filtered and reconstructed to obtain the multidimensional dataset.

3. The disaster hazard identification and early warning method based on multi-dimensional data fusion of tailings ponds as described in claim 2, characterized in that, The method for fusing the multidimensional dataset to generate a multidimensional data fusion result includes: Calculate the cooperative data array distribution probability of the effective signals within the multidimensional dataset to obtain the distribution probability array; Construct a dot matrix similarity probability distribution function, perform dimensionality reduction processing on the dot matrix distribution of the probability distribution array, and generate dot matrix dimensionality reduction results; The multidimensional data fusion result is obtained by fusing the dimensionality reduction results of the dot matrix.

4. The disaster hazard identification and early warning method based on multi-dimensional data fusion of tailings dams as described in claim 1, characterized in that, The tailings dam disaster impact assessment mathematical model is used to conduct a risk assessment of the tailings dam, generating multiple risk levels. The tailings dam failure triggering criteria are analyzed, generating multiple triggering indicators for the tailings dam. The method includes: The tailings dam disaster impact assessment mathematical model is activated to perform a numerical simulation of the tailings dam's disaster impact, generating multiple disaster impact scores. A comprehensive risk assessment is conducted based on the multiple disaster impact scores combined with the generalized hierarchical structure data of the tailings dam, generating a multi-layered risk assessment result. Based on the results of the multi-level risk assessment, a risk analysis of the tailings dam was conducted, and multiple risk levels were classified. The stability of the tailings dam is calculated based on the multiple risk levels to obtain the stability coefficient of the tailings dam. Based on the stability coefficient, a dam failure trigger test is conducted, and the multiple trigger indicators are set according to the test results.

5. The disaster hazard identification and early warning method based on multi-dimensional data fusion of tailings dams as described in claim 1, characterized in that, Sensitivity analysis is performed on the aforementioned multiple triggering indicators to determine the disaster triggering sensitivity coefficient. The methods include: The tailings dam is co-evolved according to multiple trigger indicators to generate multi-indicator co-evolution results; Sensitivity analysis was conducted based on the results of multi-indicator co-evolution to determine the disaster trigger sensitivity coefficient.

6. The disaster hazard identification and early warning method based on multi-dimensional data fusion of tailings dams as described in claim 1, characterized in that, Based on the disaster triggering sensitivity coefficient and the multiple risk levels, a disaster hazard identification method is used to generate a disaster early warning signal, including: Multiple weighting coefficients are generated by weighting the disaster triggering sensitivity coefficient with the multiple risk levels; A comprehensive risk score is obtained by matching and calculating the multiple triggering indicators according to the multiple weighting coefficients. The disaster trigger sensitivity coefficient is correlated with the multiple risk levels to perform disaster analysis, a preset risk threshold is obtained, and it is determined whether the comprehensive risk score is greater than or equal to the preset risk threshold. When the comprehensive risk score is greater than or equal to the preset risk threshold, an early warning level is set according to the comprehensive risk score, and the early warning level is added to the disaster early warning signal.

7. The method for identifying and warning of potential hazards in tailings ponds based on multi-dimensional data fusion as described in claim 6, characterized in that, Based on the disaster early warning signal and the multidimensional dataset, disaster alerts are issued, and emergency response measures are formulated. The methods include: The warning levels are mapped to the multidimensional dataset for disaster analysis, and disaster relief time limits are set. According to the disaster rescue time limit, the multidimensional data of the tailings dam are used for rescue feedback to generate emergency feedback information; Based on the emergency feedback information, the multidimensional dataset is monitored and adjusted in real time, and the emergency response measures are formulated.

Images