An open-pit coal mine internal dump subsidence early warning method based on rainfall characteristics and radar subsidence data fusion

Abstract

The present application relates to the field of data fusion, in particular to a kind of open-pit coal mine internal dump subsidence early warning method based on rainfall feature and radar subsidence data fusion.First, based on the obtained subsidence displacement data, the subsidence rate is calculated;The subsidence rate and the obtained rainfall data are normalized to obtain the normalized subsidence rate and the normalized rainfall, and a historical rainfall sequence is constructed, the historical rainfall sequence is analyzed backtracking, and the equivalent response rainfall is obtained;Then, based on the equivalent response rainfall, the risk index is obtained;Based on the subsidence rate, screening is carried out to obtain the historical stable operation stage;Based on the risk index, the classification statistical threshold of the historical stable operation stage is obtained, and compared with the risk index, the subsidence early warning information is output.The technical problems that the actual response difference of rainfall to subsidence cannot be reflected, the real coupling relationship between rainfall and subsidence cannot be accurately described, and the cumulative evolution process of subsidence risk cannot be reflected are solved.

Description

Technical Field

[0001] This invention relates to the field of data fusion, and in particular to an early warning method for subsidence of spoil heaps in open-pit coal mines based on the fusion of rainfall characteristics and radar subsidence data. Background Technology

[0002] Open-pit coal mine spoil heaps, as large-scale artificial accumulations formed during mining, directly impact the safety of mining operations and the surrounding ecological environment. Because spoil heaps are typically formed from loose material deposits, their internal structure is highly heterogeneous, with well-developed pores and complex drainage conditions, making them prone to deformation and even instability under external disturbances. Rainfall infiltration is considered a significant external driving factor inducing spoil heap subsidence and slippage. Rainfall alters the moisture content, pore water pressure, and effective stress state of the heap, thereby affecting the mechanical structure between particles and causing accelerated subsidence or localized damage under certain conditions. Therefore, developing an early warning method that effectively reflects the coupling relationship between rainfall and subsidence is of significant engineering importance. In recent years, with the development of multi-source monitoring technology, the fusion analysis of radar subsidence monitoring data and rainfall information has gradually become a research hotspot. However, existing fusion methods mostly employ empirical or fixed-weight models, such as time-decay-based rainfall accumulation indices or simple linear superposition methods, failing to fully reflect the actual differences in the response of rainfall to subsidence. Meanwhile, the impact of rainfall on subsidence exhibits significant lag and phased characteristics across different time scales, and the simple assumption that "recent rainfall has a higher weight" does not always hold true in practical engineering. Furthermore, some existing methods repeatedly introduce similar variables when constructing coupled models, leading to information redundancy and increased statistical correlation, thereby reducing model stability and interpretability. Therefore, it is necessary to propose a subsidence early warning method based on the deep fusion of rainfall characteristics and radar subsidence data. By introducing a feature construction mechanism that reflects the true response relationship, this method can dynamically characterize and classify the subsidence risk of spoil heaps without relying on empirical parameters, thereby improving the accuracy, stability, and engineering applicability of the early warning results. Summary of the Invention

[0003] This invention provides an early warning method for subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data. This method addresses the problems of failing to reflect the actual differences in the response of rainfall to subsidence when using only cumulative rainfall or short-term rainfall intensity to characterize the effect of rainfall; failing to accurately depict the true coupling relationship between rainfall and subsidence when using fixed-time attenuation or empirical weights to process the impact of rainfall; and lacking a dynamic recursive mechanism based on historical states, which makes it difficult to reflect the cumulative evolution process of subsidence risk.

[0004] The present invention discloses an early warning method for subsidence of spoil heaps in open-pit coal mines based on the fusion of rainfall characteristics and radar subsidence data, comprising the following steps: S1. Obtain subsidence displacement data and rainfall data; calculate subsidence rate based on subsidence displacement data; normalize the subsidence rate and rainfall data to obtain normalized subsidence rate and normalized rainfall; obtain historical rainfall sequence based on normalized rainfall; perform retrospective analysis on historical rainfall sequence based on normalized subsidence rate, introduce rainfall response contribution weight, and obtain equivalent response rainfall. S2. Based on the equivalent response rainfall, the risk index is updated to obtain the risk index; based on the subsidence rate, the low fluctuation range is screened by the moving variance analysis method based on the subsidence rate sequence to obtain the historical stable operation stage; based on the risk index, the risk index sequence of the historical stable operation stage is extracted to obtain the graded statistical threshold; the risk index is compared with the graded statistical threshold to output the subsidence early warning information.

[0005] Preferably, S1 specifically includes: Based on the subsidence displacement data, the difference between the subsidence displacement data at two adjacent monitoring times is normalized over time to obtain the subsidence rate.

[0006] Preferably, S1 specifically includes: The contribution weight of rainfall response is calculated by multiplying the normalized rainfall amount and the absolute value of the normalized subsidence rate at each historical moment in the historical rainfall sequence as the numerator, and summing the products of the normalized rainfall amount and the absolute value of the normalized subsidence rate at each historical moment as the denominator.

[0007] Preferably, S1 specifically includes: The historical rainfall sequence is weighted and fused with the corresponding rainfall response contribution weight, and then combined with the normalized rainfall at the current moment to form the equivalent response rainfall.

[0008] Preferably, S2 specifically includes: The initial risk index is set to 0. The risk index is updated based on the equivalent response rainfall. The absolute value of the risk index at the previous moment is added by 1 and multiplied by the equivalent response rainfall at the current moment to obtain the coupling enhancement effect between the current rainfall effect and the historical risk state.

[0009] Preferably, S2 specifically includes: Based on the enhanced coupling effect between the current rainfall and the historical risk status, the newly added risk increment term is calculated; the newly added risk increment term is added to the risk index at the previous moment to calculate the risk index at the current moment.

[0010] Preferably, S2 specifically includes: In the sliding variance analysis method based on subsidence rate sequence, the subsidence displacement sequence is obtained according to a fixed monitoring time interval, and the subsidence rate sequence between adjacent time points is calculated.

[0011] Preferably, S2 specifically includes: The set continuous sliding window is moved step by step along the time axis. Each time it moves, the variance of the settlement rate within the sliding window is calculated. The lower quartile of the variance is used as the screening criterion to screen out the low fluctuation window and obtain the historical stable operation stage.

[0012] Preferably, S2 specifically includes: For each historical stable operation phase, the risk index sequence of the historical stable operation phase is extracted, the mean and standard deviation are calculated, and the three-σ discriminant method is used to construct the hierarchical statistical threshold corresponding to each historical stable operation phase.

[0013] Preferably, S2 specifically includes: The risk index at the current moment is compared with the hierarchical statistical threshold corresponding to the most recent and longest-continuous historical stable operating phase. If the risk index at the current moment does not exceed the mean plus two standard deviations, it is judged as low risk. If the risk index at the current moment is higher than the mean plus two standard deviations but not higher than the mean plus three standard deviations, it is judged as medium risk. If the risk index is higher than the mean plus three standard deviations, it is judged as high risk, and a subsidence warning is output.

[0014] The beneficial effects of the technical solution of the present invention are: 1. By fusing subsidence displacement data and rainfall data and constructing a subsidence rate characterization quantity, a dynamic characterization of the surface deformation evolution trend was achieved, thus providing a reliable data foundation for subsidence risk analysis; through the normalization process of the Z-score standardization method, the difference in dimensions of multi-source data was eliminated, and the consistency and stability of data fusion were improved.

[0015] 2. By introducing a rainfall response contribution weight allocation mechanism based on subsidence response information, the impact of rainfall is transformed from "time-driven" to "response-driven". A comprehensive equivalent rainfall index is constructed, which can truly reflect the actual intensity of rainfall's effect on subsidence and significantly improve the risk characterization capability.

[0016] 3. By constructing a progressive update mechanism for risk status, a unified expression of historical status inheritance and current disturbance response can be achieved, thereby enabling risk assessment to have continuity and sensitivity and improving the stability of assessment results. Attached Figure Description

[0017] Figure 1 This is a flowchart of an early warning method for subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, as described in this invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] 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 invention pertains.

[0020] The following describes in detail, with reference to the accompanying drawings, a specific scheme for an early warning method for subsidence of open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, provided by the present invention.

[0021] See attached document Figure 1 The diagram illustrates a flowchart of an early warning method for subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, according to an embodiment of the present invention. The method includes the following steps: S1. Obtain subsidence displacement data and rainfall data; calculate the subsidence rate based on the subsidence displacement data; normalize the subsidence rate and rainfall data to obtain the normalized subsidence rate and normalized rainfall; obtain the historical rainfall sequence based on the normalized rainfall; perform a retrospective analysis on the historical rainfall sequence based on the normalized subsidence rate, introduce rainfall response contribution weights, and obtain the equivalent response rainfall.

[0022] Two types of basic data are collected synchronously from the mining area monitoring system: one type is surface subsidence displacement data acquired by radar or InSAR systems. One type is used to characterize the surface deformation state at different time points; the other type is rainfall data acquired by meteorological monitoring systems. , is used to describe external hydrological input conditions.

[0023] By performing time normalization on the difference between surface subsidence displacement data at two adjacent monitoring times, the rate of change of surface subsidence displacement data per unit time can be obtained. Subsidence rate, or subsidence rate, is used to characterize the evolution trend of strata stability and can reflect the state of subsidence acceleration or deceleration.

[0024] The subsidence rate and rainfall data were normalized using the Z-score normalization method to obtain the normalized subsidence rate. and normalized rainfall .

[0025] A retrospective analysis was performed on historical rainfall sequences composed of normalized historical rainfall amounts, and subsidence response information at corresponding moments was incorporated as the basis for allocating rainfall response contribution weights. These weights do not represent traditional time decay, but rather indicate the intensity of the contribution of a rainfall event at a given historical moment to the actual surface subsidence process. Specifically, if the rainfall at a particular moment is stronger and subsequently triggers a significant subsidence response, the corresponding rainfall response contribution weight is higher; conversely, even if the rainfall intensity is high but does not trigger significant subsidence changes, the rainfall response contribution weight is low, thus achieving a response-driven expression of rainfall impact.

[0026] The historical rainfall sequence is weighted and fused with its corresponding rainfall response contribution weights, and then combined with the normalized rainfall at the current moment to form a comprehensive equivalent rainfall index. This comprehensive equivalent rainfall index no longer simply reflects the total rainfall or average intensity, but emphasizes the effective contribution of rainfall to the actual subsidence. The formula is as follows: in, Indicates time The equivalent response rainfall, i.e., the comprehensive equivalent rainfall index; Indicates time Normalized rainfall; The historical retrospective window length is determined based on the geological response cycle of the mining area. The geological response cycle of the mining area refers to the effective response time range in which the impact of rainfall is continuously reflected in the change of the surface subsidence rate after a rainfall event occurs. Indicates the weight of the contribution of rainfall response; Representing historical moments The normalized rainfall data is derived from historical rainfall sequences; Representing historical moments The normalized subsidence rate, i.e. the normalized subsidence rate corresponding to one time step after rainfall, is used to reflect the subsidence response information at the corresponding time. This represents the time step offset index in the historical backtracking sequence, i.e., the number of backtracking steps in the historical time series; Represents a minimal constant, with a range of values ​​of 1 to 2. This is used to prevent the denominator from being zero.

[0027] S2. Based on the equivalent response rainfall, the risk index is updated to obtain the risk index; based on the subsidence rate, the low fluctuation range is screened by the moving variance analysis method based on the subsidence rate sequence to obtain the historical stable operation stage; based on the risk index, the risk index sequence of the historical stable operation stage is extracted to obtain the graded statistical threshold; the risk index is compared with the graded statistical threshold to output the subsidence early warning information.

[0028] The initial risk index is set to zero to ensure that subsequent risks are generated entirely by actual monitoring data and to avoid human interference from prior knowledge. The current risk index is updated at each moment. The update process consists of two parts: one part is the risk index of the previous moment to maintain historical continuity, and the other part is the newly added risk increment to characterize the immediate driving effect of the current rainfall on the risk.

[0029] In the calculation of the new risk increment, the equivalent response rainfall at the current moment is combined with the risk index at the previous moment to characterize the "actual impact capacity of rainfall under the current geological conditions". That is, when the risk index at the previous moment is higher, the regional structure tends to be unstable, and the same rainfall will trigger a stronger risk response. Conversely, in the low-risk stage, the response to rainfall is relatively weak, thus realizing the adaptive adjustment of risk sensitivity to rainfall.

[0030] The formula for calculating the risk index is: in, Indicates time The risk index; Indicates the previous moment The risk index is used to reflect the inheritance of historical status and maintain historical continuity; Indicates time The equivalent response rainfall; This indicates a newly added risk increment, used to characterize the current coupling increment and the immediate driving effect of current rainfall on risk; Used to represent the enhanced coupling effect between current rainfall and historical risk status.

[0031] By using the approach of "current coupling increment + historical state inheritance", the risk state is gradually updated, enabling the risk evolution process to have both immediate response capability and long-term memory constraint capability, thus forming a stable dynamic causal risk generation mechanism.

[0032] The method of using moving variance analysis based on subsidence rate sequences to screen for low-fluctuation intervals involves obtaining subsidence displacement sequences at fixed monitoring time intervals, calculating the subsidence rate sequences between adjacent time points, and setting a continuous sliding window, such as continuous... sky, The sliding window length can be determined based on the regular monitoring frequency of the mining area, either daily or over a complete monitoring cycle. The sliding window is moved gradually along the time axis. Each time it moves, the variance of the subsidence rate within the sliding window is calculated to characterize the fluctuation of subsidence changes within the window. To avoid setting arbitrary thresholds, the lower quartile of the variance of the subsidence rate within all historical sliding windows can be used as the screening criterion, i.e., those with lower variances... The sliding window within the range is used as a low fluctuation window to obtain several historical stable operation stages. If multiple low fluctuation windows are consecutive or overlap in time, they are merged into a single historical stable operation stage.

[0033] For each historical stable operation phase, the risk index sequence of the historical stable operation phase is extracted, and the mean and standard deviation are calculated. Using the existing three-σ discriminant method, a hierarchical statistical threshold corresponding to each historical stable operation phase is constructed: "mean plus two standard deviations" is used as the medium risk trigger boundary, and "mean plus three standard deviations" is used as the high risk trigger boundary, avoiding the artificial setting of fixed thresholds.

[0034] The risk index at the current moment is compared with the hierarchical statistical threshold corresponding to the most recent and longest-continuous historical stable operating phase. If the risk index at the current moment does not exceed "mean plus two standard deviations", it is judged as low risk; if the risk index at the current moment is higher than "mean plus two standard deviations" but not higher than "mean plus three standard deviations", it is judged as medium risk; if the risk index is higher than "mean plus three standard deviations", it is judged as high risk, and a subsidence warning is output.

[0035] In summary, a method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data has been developed.

[0036] The order of the embodiments is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0037] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0038] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, characterized in that, Includes the following steps: S1. Obtain settlement displacement data and rainfall data; Based on the subsidence displacement data, the subsidence rate is calculated; the subsidence rate and rainfall data are normalized to obtain the normalized subsidence rate and normalized rainfall. Based on the normalized rainfall, historical rainfall sequences are obtained; based on the normalized subsidence rate, a retrospective analysis of the historical rainfall sequences is performed, and rainfall response contribution weights are introduced to obtain the equivalent response rainfall. S2. Based on the equivalent response rainfall, the risk index is updated to obtain the risk index; Based on the subsidence rate, the low fluctuation range is screened by the sliding variance analysis method based on the subsidence rate sequence to obtain the historical stable operation stage; Based on the risk index, the risk index sequence of historical stable operation phases is extracted to obtain the hierarchical statistical threshold; the risk index is compared with the hierarchical statistical threshold to output subsidence early warning information.

2. The method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data as described in claim 1, characterized in that, S1 specifically includes: Based on the subsidence displacement data, the difference between the subsidence displacement data at two adjacent monitoring times is normalized over time to obtain the subsidence rate.

3. The method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data as described in claim 1, characterized in that, S1 specifically includes: The contribution weight of rainfall response is calculated by multiplying the normalized rainfall amount and the absolute value of the normalized subsidence rate at each historical moment in the historical rainfall sequence as the numerator, and summing the products of the normalized rainfall amount and the absolute value of the normalized subsidence rate at each historical moment as the denominator.

4. The method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data as described in claim 3, characterized in that, S1 specifically includes: The historical rainfall sequence is weighted and fused with the corresponding rainfall response contribution weight, and then combined with the normalized rainfall at the current moment to form the equivalent response rainfall.

5. The method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data as described in claim 1, characterized in that, S2 specifically includes: The initial risk index is set to 0. The risk index is updated based on the equivalent response rainfall. The absolute value of the risk index at the previous moment is added by 1 and multiplied by the equivalent response rainfall at the current moment to obtain the coupling enhancement effect between the current rainfall effect and the historical risk state.

6. A method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, as described in claim 5, is characterized in that... S2 specifically includes: Based on the enhanced coupling effect between the current rainfall and the historical risk status, the newly added risk increment term is calculated; the newly added risk increment term is added to the risk index at the previous moment to calculate the risk index at the current moment.

7. The method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data as described in claim 1, characterized in that, S2 specifically includes: In the sliding variance analysis method based on subsidence rate sequence, the subsidence displacement sequence is obtained according to a fixed monitoring time interval, and the subsidence rate sequence between adjacent time points is calculated.

8. A method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, as described in claim 7, is characterized in that... S2 specifically includes: The set continuous sliding window is moved step by step along the time axis. Each time it moves, the variance of the settlement rate within the sliding window is calculated. The lower quartile of the variance is used as the screening criterion to screen out the low fluctuation window and obtain the historical stable operation stage.

9. A method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, as described in claim 8, is characterized in that... S2 specifically includes: For each historical stable operation phase, the risk index sequence of the historical stable operation phase is extracted, the mean and standard deviation are calculated, and the three-σ discriminant method is used to construct the hierarchical statistical threshold corresponding to each historical stable operation phase.

10. A method for early warning of subsidence in open-pit coal mine spoil heaps based on the fusion of rainfall characteristics and radar subsidence data, as described in claim 9, is characterized in that... S2 specifically includes: The risk index at the current moment is compared with the hierarchical statistical threshold corresponding to the most recent and longest-continuous historical stable operating phase. If the risk index at the current moment does not exceed the mean plus two standard deviations, it is judged as low risk. If the risk index at the current moment is higher than the mean plus two standard deviations but not higher than the mean plus three standard deviations, it is judged as medium risk. If the risk index is higher than the mean plus three standard deviations, it is judged as high risk, and a subsidence warning is output.

Images