A slope monitoring and early warning system for an open pit coal mine

Abstract

This invention discloses a monitoring and early warning system for open-pit coal mine slopes, relating to the field of coal mine slope monitoring technology. The system includes: a monitoring data acquisition module for acquiring slope surface deformation data, internal strain data, and environmental video streams; a hydrological model construction module for constructing a hydrological evolution model of the bottom sump of the spoil heap; a mechanical parameter correction module for deduce the evolution process of the slope sliding surface; a data alignment module for generating aligned monitoring data streams; a three-dimensional projection module for generating deformation cloud maps; an early warning signal generation module for dividing the slope deformation curve characteristics and sliding surface evolution trend into three sub-stages: initial acceleration, intermediate acceleration, and further acceleration, and generating corresponding graded early warning signals; a risk judgment module for judging the slope instability risk level; and an emergency response module for generating safety protection action commands when the risk level exceeds a preset threshold. This invention reduces the risk of landslides on mine slopes and improves the level of intelligent safety management of open-pit coal mine slopes.

Description

Technical Field

[0001] This invention relates to the field of coal mine slope monitoring technology, specifically to an open-pit coal mine slope monitoring and early warning system. Background Technology

[0002] The slopes of open-pit coal mine spoil heaps are artificially loosely deposited rock and soil structures, characterized by large deposit thickness, chaotic rock and soil structure, uneven compaction, and high permeability. Affected by multiple factors such as rainfall infiltration, base softening, and rock and soil creep, these slopes are highly susceptible to geological disasters such as deformation and slippage, local collapse, and even overall landslides, seriously threatening the safety of on-site workers, machinery, and the mine's production operations. In particular, the water collection pit area at the bottom of the spoil heap, which accumulates rainwater, surface runoff, and groundwater over a long period, is highly prone to softening of the rock and soil at the bottom of the slope and reduction of shear strength. This is the core weak point for slope instability and failure, placing extremely high demands on the precision, real-time nature, comprehensive coverage, and accuracy of early warning systems for slope monitoring. Currently… The monitoring technology for slopes in open-pit coal mines mostly adopts single fixed-point monitoring or single-equipment scanning monitoring modes, which have relatively limited monitoring dimensions and generally suffer from problems such as large monitoring blind spots, limited data dimensions, and weak anti-interference capabilities. Traditional monitoring methods cannot achieve multi-source collaborative monitoring of slope surface deformation, internal strain, and hydrological evolution, making it difficult to comprehensively capture the entire process evolution characteristics of slopes from micro-deformation to instability and landslides. At the same time, existing monitoring models mostly use fixed geotechnical mechanical parameters for stability simulation calculations, without combining the real-time deformation state of the slope to dynamically correct core parameters such as soil cohesion and internal friction angle. The simulation results deviate significantly from the actual working conditions on site and cannot accurately reflect the gradual instability law of slopes induced by seepage from sump pits.

[0003] Furthermore, existing monitoring systems lack a unified time-series calibration and error correction mechanism for multi-source heterogeneous monitoring data. Data from different sources suffers from time-series misalignment, noise interference, and atmospheric delay bias, resulting in low data fusion accuracy and delayed deformation feature identification. Traditional early warning methods rely solely on a single deformation threshold for risk assessment, failing to differentiate between the stages of initial, intermediate, and accelerated slope deformation, thus hindering tiered and precise early warning. Emergency response models are also singular and rigid, unable to dynamically adjust strategies based on real-time slope risk conditions. Simultaneously, the fixed deployment of traditional monitoring equipment suffers from blind spots, poor data transmission fault tolerance, and insufficient closed-loop emergency response, making it difficult to adapt to the complex hydrogeological, multi-interference, and high-risk operating conditions of open-pit coal mine spoil heaps. This results in numerous technical shortcomings, including low monitoring accuracy, delayed early warning, poor adaptability to operating conditions, and untimely emergency response. Summary of the Invention

[0004] To solve the above-mentioned technical problems, an open-pit coal mine slope monitoring and early warning system is provided. This technical solution solves the aforementioned problems.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: An open-pit coal mine slope monitoring and early warning system includes: The monitoring data acquisition module is used to conduct multi-source collaborative monitoring of open-pit coal mine slopes through slope deformation monitoring equipment, positioning monitoring equipment, internal strain monitoring equipment and on-site image acquisition equipment, and to acquire slope surface deformation data, internal strain data and environmental video streams. The hydrological model building module is used to construct a hydrological evolution model of the sump at the bottom of the spoil heap, and a non-steady flow calculation module is built based on Darcy's law to simulate the water level seepage rate. The mechanical parameter correction module is used to construct a subsystem for correcting the mechanical parameters of soil and rock. It uses a backpropagation neural network algorithm to update the cohesion and internal friction angle parameters in the model based on the internal strain data, and to deduce the evolution process of the slope sliding surface. The data alignment module is used to perform time-domain slicing and traffic correlation analysis on monitoring data from different sources based on spatiotemporal references, perform time-series alignment operations, and generate aligned monitoring data streams. The 3D projection module is used to project the aligned monitoring data and simulation results onto the 3D geographic information system model to generate deformable cloud maps. The early warning signal generation module is used to divide the slope deformation curve characteristics and the evolution trend of the sliding surface into three sub-stages: initial acceleration, intermediate acceleration and accelerated acceleration, and generate corresponding graded early warning signals. The risk assessment module is used to determine the slope instability risk level when an early warning signal is generated; The emergency response module is used to generate safety protection action commands when the risk level exceeds a preset threshold, and to link the on-site alarm equipment, on-site information display equipment and drainage execution equipment to perform emergency response.

[0006] Optionally, the multi-source collaborative monitoring of open-pit coal mine slopes using slope deformation monitoring equipment, positioning monitoring equipment, internal strain monitoring equipment, and on-site image acquisition equipment includes: Slope deformation monitoring equipment was used to scan key areas of the slope, and location monitoring points were set up to assist in the monitoring of areas not covered by the monitoring equipment. Internal strain monitoring equipment was installed within the rock and soil of the spoil heap slope to collect strain data. A backup strain sensor is implanted in the borehole where each internal strain monitoring device is located, and the two form a parallel data acquisition channel. When the internal strain monitoring equipment breaks due to the tension of the soil and rock, the data channel of the backup strain sensor is switched.

[0007] Optionally, the subsystem for correcting the mechanical parameters of the soil and rock mass includes: Mark the transverse burial line along the slope height direction, and shorten the sensor placement spacing at the junction of soil and rock layers to increase the monitoring point density; The internal strain monitoring equipment is fitted with a protective sleeve, which is then filled with cement mortar and epoxy resin for protection. The collected strain data is input into the backpropagation neural network algorithm to correct the geomechanical parameters in the simulation model.

[0008] Optionally, the analysis of temporal slicing and traffic correlation of monitoring data from different sources based on a spatiotemporal reference includes: The core module is protected by encapsulation, and a signal isolation layer and grounding protection are added. The data processing algorithm based on Kalman filtering and wavelet analysis is executed. First, high-frequency vibration interference signals are removed by wavelet thresholding. Then, Kalman filtering is used to predict and correct atmospheric delay errors. By integrating data from positioning monitoring equipment, slope deformation monitoring equipment, and internal strain monitoring equipment, and calculating the correlation coefficient within each time-domain slice, time delay compensation and alignment are performed on asynchronous data streams.

[0009] Optionally, projecting the aligned monitoring data and simulation results onto the three-dimensional geographic information system model includes: In the data visualization platform, a unified reference point is determined, the coordinates of the reference point are measured and input into the platform, and correction parameters are calculated to adjust the spatial position of the 3D model. In the system trend analysis module, reduce the data sliding window size parameter to increase the weight of monitoring data in trend analysis; Based on the characteristics of the tangent angle change in the deformation curve, the slope deformation stage is identified and an early warning of the corresponding level is triggered.

[0010] Optionally, the emergency response performed by the alarm devices, on-site information display devices, and drainage execution devices linked to the site includes: Develop mobile applications and deploy on-site information display devices to push early warning information to technicians' terminals and on-site work areas via wireless network; When the system is determined to be in a state of imminent slippage, a start command is sent to the drainage execution equipment and drainage control system.

[0011] Optionally, before acquiring the slope surface deformation data, internal strain data, and environmental video stream, the method further includes: Curved steel rails were laid along the edge of the spoil heap, and slope deformation monitoring equipment was installed on a platform that could move along the rails. The slope deformation monitoring equipment is controlled by a drive device to move along the track and stop at the preset positioning sensor position for monitoring.

[0012] Optionally, the graded early warning signal includes: The initial acceleration stage warning corresponds to the state where the slope begins to show continuous deformation, and the patrol mechanism is activated. The warning for the accelerated deformation stage corresponds to a state where the deformation rate increases, and drainage and load reduction preparations are initiated. Acceleration phase warning corresponds to the state of an impending landslide, triggering audible and visual alarms and evacuating personnel and equipment.

[0013] Optionally, after performing time-domain slicing and traffic correlation analysis on monitoring data from different sources based on a spatiotemporal reference, the method further includes: When a data packet loss is detected for a specific monitoring device, the lost data packet is numerically filled using the average of the historical alignment data stored in the buffer medium. A data transmission system is constructed using a dual-link mode that combines wired transmission with long-distance wireless backup.

[0014] Optionally, after generating a safety protection action instruction when the risk level exceeds a preset threshold, the method further includes: After sending the safety protection action command, obtain on-site feedback data; Calculate the slope of change of the feedback data within a preset time window; When the magnitude of the change slope is less than the preset change rate threshold, the emergency response is deemed effective, and a downgrade command is sent. When the slope of change is not less than the preset rate of change threshold, the highest level alarm will be maintained and the drainage equipment will be increased.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention constructs a multi-source collaborative three-dimensional monitoring system, integrating multi-dimensional monitoring data such as slope surface deformation, internal strain, and environmental video. It is equipped with a dual-channel acquisition mechanism and a mobile full-domain scanning monitoring mode, which effectively eliminates the blind spots of fixed monitoring, realizes full-domain, uninterrupted, and highly reliable data acquisition of slopes, and solves the problems of single monitoring dimensions, missing data, and incomplete monitoring coverage, thereby improving the completeness and accuracy of basic slope monitoring. 2. This invention constructs a non-steady flow calculation model for the hydrological evolution of a sump based on Darcy's law, accurately simulating the dynamic changes in water infiltration and water level on slopes under rainfall conditions. At the same time, it relies on the backpropagation neural network algorithm combined with real-time strain data to dynamically correct the mechanical parameters of the soil and rock. Unlike the drawbacks of ordinary simulation models with fixed parameters and disconnected from the actual working conditions, this invention can accurately predict the entire process of the gradual evolution of the slope sliding surface, improving the fit and accuracy of slope stability simulation. 3. This invention adds a multi-source data time-series alignment and anti-interference correction mechanism, combining wavelet denoising and Kalman filtering algorithms to remove high-frequency interference and atmospheric delay errors in the field, and realizes time-series alignment of multi-source heterogeneous data through time-domain slice correlation analysis. At the same time, it configures a data missing fault-tolerant filling and dual-link backup transmission architecture to solve the problems of data time-series misalignment, high noise, packet loss and disconnection in the complex interference environment of the mining area, and ensure the stability, continuity and synchronization of monitoring data. 4. This invention uses a three-dimensional geographic information model to visualize and project monitoring data and simulation results, generating dynamic deformation cloud maps. Combined with an optimized sliding window weighting algorithm, it accurately captures the minute deformation characteristics of slopes. Based on the changing law of the tangent of the deformation curve, it divides the slope into three sub-stages of deformation and matches them with corresponding graded early warning mechanisms. This breaks through the traditional coarse mode of single threshold early warning and realizes the refined identification and graded early warning of the gradual evolution process of slope deformation, greatly improving the timeliness and accuracy of early warning. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the steps of the present invention. Detailed Implementation

[0017] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0018] Example 1: Refer to Figure 1 As shown, an open-pit coal mine slope monitoring and early warning system includes: The monitoring data acquisition module conducts multi-source collaborative monitoring of open-pit coal mine slopes using various monitoring methods. It simultaneously acquires slope surface deformation data, internal strain data, and on-site environmental video streams, constructing a comprehensive raw monitoring dataset covering the slope surface, interior, and environment. Specifically, it adopts a zoned collaborative monitoring mode, simultaneously collecting macroscopic surface deformation of the slope, deep strain of the soil and rock mass, and on-site real-scene video information. This enables synchronous perception of external deformation and internal structural changes of the slope, providing complete raw data support for subsequent simulation analysis and risk assessment. The multi-source collaborative monitoring mentioned here refers to an integrated three-dimensional data acquisition mode that combines spatial surface monitoring, deep structure monitoring, and on-site video monitoring. The hydrological model construction module constructs a hydrological evolution model of the sump at the bottom of the spoil heap. Based on Darcy's law, it establishes a non-steady flow calculation logic to simulate the water level infiltration rate and water transport patterns within the slope soil and rock mass under different rainfall and water accumulation conditions. Specifically, relying on the basic principle of seepage flow based on Darcy's law and combined with the permeability characteristics of the loose deposits in the spoil heap, a non-steady flow simulation calculation system adapted to the infiltration of water accumulation in the sump, surface runoff, and groundwater recharge is established to dynamically deduce the dynamic changes in water content and seepage pressure of the slope base soil and rock mass. The non-steady flow calculation mentioned here refers to a seepage simulation calculation mode adapted to real-time rainfall and water accumulation changes and dynamically iterating over time. The mechanical parameter correction module and the subsystem for correcting soil and rock mechanical parameters utilize a backpropagation neural network algorithm combined with real-time acquired internal slope strain data to dynamically update the core mechanical parameters of cohesion and internal friction angle of the soil and rock mass within the simulation model. This allows for real-time simulation of the entire evolution of the slope sliding surface, from initiation to expansion and connection. Specifically, it abandons the traditional fixed-parameter simulation mode, using the real-time deformation and strain characteristics of the slope as the driving factor. Through the nonlinear fitting capability of the neural network, it iteratively optimizes the soil and rock mechanical parameters, accurately matching the softening, creep, and damage evolution states of the soil and rock mass in the field, thereby improving the realism of slope instability simulation. The soil and rock mechanical parameters mentioned here refer to the key indicators of cohesion and internal friction angle that determine the slope's anti-sliding stability. The data alignment module performs time-domain slicing and data correlation analysis on monitoring data from different sources and dimensions based on a unified spatiotemporal benchmark. It performs latency compensation and time-series alignment operations on heterogeneous data streams with time misalignment and inconsistent sampling frequencies, generating aligned monitoring data streams with unified timing and matching dimensions. Specifically, using the system's globally unified time axis as a benchmark, it performs fixed-step time-domain slicing on multi-source monitoring data, calculates the correlation matching degree between different monitoring data dimensions segment by segment, and calculates the overall average time-series deviation by traversing the correlation fluctuation difference of all time-domain slices. Based on the average deviation, it performs frame-by-frame latency compensation on lagging or leading data streams, ultimately achieving complete synchronization and matching of multi-source data in the time dimension. The time-series alignment operation mentioned here refers to the process of eliminating the time difference in multi-source data acquisition and achieving unified calibration of the spatiotemporal dimensions of the data. The 3D projection module projects the time-aligned, precise monitoring data and hydrological and mechanical simulation results onto a 3D geographic information system model, constructing a global spatial data field for the slope and generating dynamic deformation cloud maps of slope displacement, strain, and seepage pressure. Specifically, relying on the 3D geographic information model to recreate the real topography of the slope, spatial interpolation algorithms are used for discrete monitoring point data to calculate the data change gradient of adjacent monitoring points grid by grid, completing the continuous data distribution of non-monitored areas, completing the gridded fitting of discrete data, and visually presenting the global deformation distribution of the slope, stress concentration areas, and seepage weak areas. The dynamic deformation cloud map mentioned here refers to a real-time updated, intuitively reflecting the spatial feature map of deformation differences in different areas of the slope. The early warning signal generation module, based on the long-term deformation curve characteristics of the slope and the simulated evolution trend of the sliding surface, divides the gradual instability process of the slope into three sub-stages: initial acceleration, intermediate acceleration, and accelerated acceleration. For each deformation stage, it generates a corresponding graded early warning signal based on the deformation rate and evolution characteristics. Specifically, it calculates the cumulative increase in slope deformation and the increase in deformation per unit time at each moment, statistically analyzes the year-on-year increase rate of deformation, and combines multi-dimensional quantitative indicators such as the curvature change of the deformation curve and the growth rate of the sliding surface expansion area to accurately distinguish between three states of slope: slow deformation, accelerated deformation, and sudden slippage, and matches differentiated early warning signals accordingly. The graded early warning signals mentioned here refer to standardized early warning information corresponding to different degrees of slope instability risk and increasing gradients. The risk assessment module, while generating early warning signals at all levels, dynamically quantifies and assesses the slope instability risk level by combining real-time monitoring data and simulation results, thus completing the accurate determination and classification of risk levels. Specifically, it normalizes and quantifies four core indicators: deformation rate, cumulative deformation, attenuation ratio of geotechnical parameters, and increase in seepage pressure, assigns corresponding weight coefficients to each, and obtains a comprehensive risk score by weighted summation. The risk level is then determined by comparing the score with a preset scoring range, quantifying the current probability and severity of slope instability. The emergency response module automatically generates standardized safety protection action instructions when the slope instability risk level is determined to exceed the system's preset safety threshold. It links multiple execution systems, such as on-site early warning prompts, information disclosure, and drainage control, to complete the fully automated emergency response process. Specifically, it matches corresponding disposal strategies according to different risk levels, realizing integrated linkage of early warning broadcasts, on-site disclosure, remote push, and water discharge, forming an automated emergency closed loop.

[0019] Through the above technical solutions, this system realizes intelligent monitoring and control of the entire process of multi-source data collaborative acquisition, hydrological-mechanical coupling simulation, precise time-series correction, three-dimensional visualization analysis, hierarchical early warning and intelligent emergency linkage for slopes of open-pit coal mine spoil heaps. It solves the problems of single monitoring data, fixed model parameters, crude early warning and delayed response in traditional monitoring systems, and improves the identification accuracy and timeliness of response to potential slope instability hazards induced by water collection pits.

[0020] In some embodiments, the data acquisition process for multi-source collaborative monitoring specifically includes the following steps: Step 201: Conduct continuous full-area scanning monitoring for key landslide risk areas of the slope to achieve rapid data coverage of large areas. For areas that are not effectively covered, such as scanning blind spots, obstructed areas, and edge dead corners, additional fixed monitoring points are set up to achieve full-area monitoring coverage without blind spots. Specifically, a mode combining large-area scanning and fixed-point supplementary measurement is adopted to balance monitoring efficiency and coverage integrity, eliminating the monitoring blind spot defects of the traditional single scanning mode. Step 202: Deploy deep monitoring points inside the loose rock and soil mass of the spoil heap slope. Continuously collect data on minute strain and deformation inside the rock and soil mass through pre-embedded deployment to capture early micro-damage and micro-deformation evolution characteristics inside the slope. Specifically, the deep monitoring points penetrate deep into the rock and soil mass inside the slope, which can capture internal sliding trends that cannot be identified by surface monitoring in advance, thus achieving early prediction of slope instability. Step 203: Install a backup monitoring sensor structure inside the borehole at each deep monitoring point to form a dual-channel parallel data acquisition system with the main monitoring structure and the backup monitoring structure, and simultaneously collect deep strain data at the same location; specifically, the main and backup dual channels work synchronously, and under normal conditions, the deviation values ​​of the two sets of collected data are compared frame by frame to calculate the data consistency and achieve mutual verification and calibration. Under abnormal conditions, they serve as backups for each other to ensure the reliability of monitoring data. Step 204: The system monitors the working status of the deep monitoring channel in real time and calculates the stability of the main channel data frame by frame. When the main monitoring structure data is detected to have fracture failure characteristics such as sudden changes, disconnection, or constant change, the system automatically identifies the channel abnormality and instantly switches to the backup sensor data channel to continuously acquire deep strain monitoring data. Specifically, through the dual-channel redundancy design, the monitoring interruption problem caused by deformation and failure of the soil and rock mass is effectively avoided, ensuring continuous and uninterrupted deep slope monitoring.

[0021] The above technical solution adopts a data acquisition mode of full-area scanning + fixed-point supplementary measurement + primary and backup redundant monitoring, which not only ensures full coverage of the slope surface, but also achieves accurate capture of deep hidden dangers. At the same time, the automatic channel switching mechanism solves the industry shortcomings of easy failure and data interruption in deep monitoring, and greatly improves the stability and reliability of the monitoring system under complex mining conditions.

[0022] In some embodiments, the layout and parameter correction process of the soil and rock mechanical parameter correction subsystem specifically includes the following steps: Step 301: Divide the slope vertically into multiple layers of transverse monitoring layout lines and evenly distribute conventional monitoring points; for key risk locations such as the interface between soil and rock layers, the transition zone between soft and hard soil and rock, and weak interlayers in the accumulation body, shorten the layout spacing and increase the density of monitoring points; specifically, the interface between soil and rock is a high-incidence area for slope slippage and seepage softening, and the monitoring sensitivity and data density of key layers are improved by increasing the number of points. Step 302: Install protective sleeves on the outside of all monitoring structures buried in the rock and soil to form a rigid protective structure; fill the space between the sleeve and the borehole wall with a protective material made of cement mortar and epoxy resin to achieve sealing, waterproofing, seepage prevention, and compression protection; specifically, the protective sleeve resists the compression and shearing damage of the rock and soil, the filling material isolates groundwater corrosion, fixes the monitoring position, and prevents the deep monitoring structure from shifting, failing, or being damaged by water ingress; Step 303: The real-time slope strain time-series data collected by the deployed deep monitoring structure is continuously input into the backpropagation neural network algorithm model to complete the dynamic correction of mechanical parameters. The specific calculation process is as follows: First, the real-time strain monitoring data is normalized, and the original strain values ​​are uniformly mapped to the standard range of zero to one to eliminate dimensional differences. Second, the normalized data is input into the input layer of the neural network. Through multi-layer nonlinear weighted calculation in the hidden layer, the predicted values ​​of the current soil cohesion and internal friction angle are output. Then, the error difference between the predicted parameters and the historical optimal parameters is calculated, and the error is propagated back layer by layer to correct the network weight parameters. The process is iterated repeatedly until the error difference is less than the preset convergence threshold. Finally, the optimal mechanical parameters that adapt to the current slope deformation state are output, and the simulation model parameters are updated adaptively to accurately match the mechanical deterioration law of the slope soil and rock mass with seepage, creep, and load changes.

[0023] The above technical solutions ensure the accuracy and effectiveness of deep monitoring data through layered and encrypted deployment and multiple protective encapsulation. By relying on neural network iterative calculation and error convergence optimization, dynamic adaptive correction of geotechnical parameters is achieved, which completely solves the defects of fixed parameters and disconnection from the harsh on-site conditions of traditional simulation models, and significantly improves the fit of slope stability simulation and deduction.

[0024] In some embodiments, the multi-source monitoring data time-series alignment process specifically includes the following steps: Step 401: The core data processing module of the system is encapsulated with mining-grade protective packaging. An external signal isolation shielding layer and a dedicated grounding protection structure are added to isolate electromagnetic interference, noise signals and electrostatic interference in the mining area. Specifically, large equipment in the mining area is prone to generating strong electromagnetic interference. Through shielding and grounding design, the stability of signal transmission is improved from the hardware level. Step 402: Execute the wavelet analysis and Kalman filter fusion data processing algorithm to complete data denoising and error correction. The specific calculation process is as follows: First, the wavelet multi-level decomposition algorithm is used to decompose the original monitoring data into low-frequency effective components and high-frequency noise components. High-frequency components are filtered by a preset wavelet threshold, and high-frequency interference components exceeding the threshold are set to zero, and the denoised clean monitoring data is reconstructed. Then, Kalman filter iterative calculation is used. First, the data prediction value at the current time is calculated based on the optimal data estimate value at the previous time. Then, the observation residual is calculated by combining the current measured denoised data. The Kalman gain coefficient is updated iteratively through the residual. Finally, high-precision monitoring data that eliminates atmospheric delay and environmental drift errors is obtained, which greatly improves the data purity and accuracy. Step 403: Integrate three types of heterogeneous monitoring data: positioning monitoring, surface deformation monitoring, and deep strain monitoring. Divide the multi-source data into time-domain slices according to a unified time step, and calculate the correlation coefficient of different data dimensions for each slice. The specific calculation method is as follows: extract the time-series change sequences of the three types of data within the same time-domain slice, calculate the mean and variance of each group of data, solve the correlation matching degree between different data through covariance operation, and obtain the correlation coefficient in the range of zero to one after normalization. The closer the coefficient is to one, the higher the consistency of the time-series changes of the data, providing a quantitative basis for time-series calibration. Step 404: Determine the data timing synchronization status based on the correlation coefficient of each slice, count the difference in the number of leading and lagging frames of different data streams, calculate the delay compensation amount frame by frame, perform frame-by-frame interpolation compensation and timing shift calibration for data streams with timing deviations, unify the sampling timing nodes of all monitoring data, complete the timing alignment of multi-source heterogeneous data, and generate a spatiotemporally synchronized fused monitoring dataset.

[0025] Through the above technical solution, by combining hardware anti-interference design with wavelet denoising and Kalman error correction dual-layer fusion calculation, complex noise interference in the mining area is eliminated. Combined with time-domain slice correlation quantification calculation and precise time delay compensation, accurate alignment of multi-source data is achieved, which solves the problems of disordered time sequence, high noise and low fusion accuracy of multi-source data, and provides a high-precision data foundation for subsequent simulation analysis and risk assessment.

[0026] In some embodiments, the three-dimensional projection visualization process of the monitoring data and simulation results specifically includes the following steps: Step 501: Select the benchmark point for the overall stability of the slope in the system data visualization platform, obtain the accurate coordinate parameters of the benchmark point by actual measurement and enter them into the platform system. By calculating the three-dimensional difference between the measured coordinates and the original coordinates of the model in the horizontal, vertical and elevation directions, solve the overall correction offset parameters of the three-dimensional model, correct the spatial coordinates of the three-dimensional geological model grid by grid, complete the spatial position calibration of the model, ensure that the model is accurately matched with the actual terrain on site, and eliminate modeling deviations. Step 502: Optimize the parameter configuration of the system trend analysis module, reduce the data sliding window size, shorten the time series statistical interval, and redistribute the time series weight ratio. The specific calculation method is as follows: reduce the number of historical data samples within the sliding window, reduce the weight ratio of early historical data, and simultaneously and proportionally increase the weight coefficient of the latest frame to the latest multi-frame real-time monitoring data, so that the trend fitting calculation results respond first to the current small deformation changes of the slope, strengthen the system's ability to capture the small and new deformation trends of the slope, and improve the sensitivity of early hidden danger identification. Step 503: Real-time plotting of the slope deformation time-series curve, calculating the tangent angle of the deformation curve at each time step, and completing the deformation stage identification; the specific calculation process is as follows: select multiple consecutive adjacent sets of time-series deformation data, fit to obtain the local deformation curve, solve the tangent slope of the curve at the current time, and obtain the tangent angle through the arctangent operation of the slope; compare the angle change amplitude and angle growth rate frame by frame, the slow and stable angle change is determined to be the initial acceleration stage, the continuous uniform increase of the angle is determined to be the medium acceleration stage, and the sudden and sharp increase of the angle is determined to be the accelerated and imminent slip stage, and the corresponding level of graded early warning mechanism is triggered according to the change characteristics.

[0027] Through the above technical solutions, a triple mechanism of three-dimensional coordinate difference correction, temporal weight quantification optimization, and iterative calculation of curve tangent angle is used to achieve refined, visualized, and intelligent analysis of the slope deformation evolution process, effectively improving the ability to identify early micro-deformation hazards of slopes and solving the problems of lagging and low sensitivity of traditional trend analysis.

[0028] In some embodiments, the tiered emergency response process specifically includes the following steps: Step 601: Establish a mobile remote early warning push system and a fixed on-site information display system. After the system generates early warning information at all levels, it pushes it to the mobile terminals of management personnel in real time via wireless network with encryption. At the same time, it is simultaneously displayed on the information display equipment in the on-site work area, realizing two-way early warning display from both remote and on-site perspectives. Specifically, it ensures that management personnel can remotely grasp the risk status in real time, and on-site workers can intuitively receive early warning prompts. Step 602: The system combines multi-dimensional quantitative indicators to comprehensively determine the slope status. The specific calculation method is as follows: integrate three core indicators: the year-on-year growth rate of slope deformation rate, the growth rate of sliding surface expansion area, and the percentage increase in seepage pressure. Calculate the weighted comprehensive risk mutation coefficient. When the coefficient reaches the preset threshold for imminent sliding, the slope is determined to be in a dangerous state of imminent sliding. The system automatically sends a one-click start command to the on-site drainage execution equipment and drainage control system to quickly start the drainage of water accumulation in the sump and the seepage dredging operation on the slope. This reduces the water content of the soil and rock, weakens the instability risk caused by water softening, and achieves proactive risk management. Through the above technical solutions, an integrated emergency response system combining remote push, on-site public display, and automatic drainage is constructed. Relying on multi-indicator weighted quantitative judgment of the landslide state, the system can achieve all-round delivery of early warning information and precise and automated handling of risks and hazards, effectively shortening the emergency response time and improving the ability to handle slope hazards in advance. In some embodiments, the process of deploying supplementary monitoring in the monitoring blind zone specifically includes the following steps: Step 701: Lay an arc-shaped steel rail moving track along the edge of the slope of the open-pit coal mine spoil heap, and use a concrete base to fix the track as a whole so that the track trajectory conforms to the outer contour of the slope and is adapted to the monitoring needs of the curved slope surface; specifically, the arc-shaped track can adapt to the arc contour of the slope and solve the problem that fixed monitoring cannot conform to the curved slope surface. Step 702: Fix the slope surface deformation scanning device on a mobile support platform with its own drive device, so that the scanning device can move adaptively along the track with the platform; specifically, the mobile mounting mode breaks through the field of view limitation of traditional fixed point monitoring and can realize dynamic scanning of the whole area. Step 703: Deploy positioning sensor structures at preset key monitoring points on the arc-shaped track. The drive device controls the mobile platform to move at a constant speed along the track. After reaching the preset positioning point, it automatically brakes and stops to complete the precise scanning and monitoring of the fixed point. Specifically, the system pre-enters the coordinate parameters of each point and compares the coordinate difference between the real-time position of the platform and the preset position in real time. When the difference is less than the preset positioning accuracy threshold, it determines that the monitoring position has been reached and brakes and stops. After traversing all blind spot points, it completes the full-area data collection without blind spots. The above technical solution adopts an arc-shaped mobile dynamic monitoring mode, which relies on coordinate difference comparison and positioning calculation to achieve precise fixed-point monitoring, replacing the traditional fixed monitoring method. This effectively eliminates monitoring blind spots in slope edges, corners, and areas with elevation differences, achieving full coverage and no dead angles in fine-grained scanning monitoring of slopes, and greatly improving the integrity of monitoring.

[0029] In some embodiments, the matching and processing logic for the graded early warning signal specifically includes the following steps: Step 801: The system calculates the incremental deformation of the slope per unit time every hour and statistically analyzes the changes in the deformation rate. When the deformation rate is detected to be continuously greater than the foundation micro-deformation threshold and the growth rate is slow without sudden changes, the slope is determined to have entered the initial acceleration deformation stage, triggering an early warning signal for the initial acceleration stage. At this time, the slope is generally stable, with only continuous micro-deformation and no immediate risk of instability. The system automatically starts a routine manual inspection mechanism, increases the frequency of on-site inspections, and tracks the deformation development trend. Step 802: The system continuously calculates the month-on-month increase rate of deformation rate. When the month-on-month increase rate of deformation rate continues to rise and the cumulative deformation amount steadily increases, exceeding the upper limit threshold of the initial acceleration stage, the slope is determined to have entered the medium-acceleration deformation stage, triggering the medium-acceleration stage warning signal. At this time, the slope deformation rate continues to accelerate, the soil and rock mass shows obvious deterioration trend, and there is a potential instability risk. The system automatically starts on-site drainage pretreatment and slope load reduction preparation work to eliminate risk factors in advance. Step 803: The system calculates the rate of increase of the sliding surface penetration ratio and the tangent angle of the deformation curve in real time. When the rate of increase of the sliding surface penetration ratio and the tangent angle reaches the threshold of imminent sliding, it determines that the slope has entered the accelerated deformation stage and triggers the early warning signal of the accelerated stage. At this time, the slope deformation increases sharply, the sliding surface is basically connected, and landslide instability is about to occur. The system immediately activates the full-area audible and visual alarm, pushes the evacuation command at the same time, and organizes the emergency evacuation of on-site personnel and machinery. Through the above technical solutions, relying on multi-dimensional quantitative calculations of deformation rate, year-on-year growth rate, and sliding surface penetration ratio, the three core stages of progressive slope instability can be accurately distinguished. Differentiated early warning and response strategies can be matched to achieve full-gradient hierarchical management from early inspection and prevention, mid-term risk control to late-stage emergency evacuation, avoiding the problems of over-handling or delayed handling by a single early warning mode.

[0030] In some embodiments, the fault-tolerant transmission process after the multi-source data timing alignment is completed specifically includes the following steps: Step 901: The system monitors the data packet reception status of each monitoring channel in real time frame by frame, and counts the number of data packets received and the time interval within a unit of time. When a missing data packet or an abnormally large time interval is detected, it is determined that a data packet loss anomaly has occurred. The system automatically retrieves the recent historical normal alignment time sequence data stored in the system cache medium and selects multiple consecutive frames of valid sample data before and after the packet loss time. Step 902: Calculate the mean of the retrieved historical normal sample data. The specific calculation method is as follows: sum up the values ​​of all selected valid sample data, divide by the total number of samples, and obtain the statistical mean. Use this mean as the fitting and filling value for the missing data packet to complete the repair and completion of the missing data, ensuring the continuity and integrity of the monitoring data time series. This mean filling method can restore the real working condition data to the greatest extent and avoid analysis interruption and judgment distortion caused by data loss. Step 903: Establish a dual-link data transmission system that combines wired fiber optic transmission with long-distance wireless backup transmission. Under normal circumstances, wired transmission is given priority to ensure high-precision transmission. The data transmission delay and bit error rate of the wired link are compared in real time. When the delay or bit error rate exceeds the normal threshold and the link is judged to be faulty, the system automatically switches to the wireless backup link to continue transmission, ensuring stable and reliable data transmission around the clock. Through the above technical solutions, a highly fault-tolerant and highly reliable data transmission system is constructed by intelligently identifying data packet loss, calculating the mean value and filling the data, and using a dual-link redundant transmission mechanism. This effectively solves the problems of data packet loss, transmission interruption, and sequence breakage under the interference environment of the mining area, and ensures the continuous and stable operation of monitoring, analysis, and risk assessment.

[0031] In some embodiments, the closed-loop verification and handling process after the emergency command is issued specifically includes the following steps: Step 1001: After the system issues the emergency action command for safety protection, it continuously collects on-site monitoring feedback data in real time, obtains the slope deformation, seepage, and environmental time-series feedback data after the emergency response, and constructs a fixed-duration sliding window emergency effect evaluation dataset. Step 1002: Calculate the dynamic change slope of the field feedback monitoring data within the sliding time window. The specific calculation process is as follows: extract the values ​​of the first and last sets of core monitoring data within the window and their corresponding time nodes, calculate the difference between the data value changes and the time change, divide the difference between the value changes and the time change to obtain the dynamic change slope of the data within the window, and quantitatively evaluate the trend of slope deformation convergence or continuous deterioration by the magnitude of the slope. Step 1003: Set a fixed rate of change threshold, calculate the absolute value of the slope. When the absolute value of the slope is less than the set rate of change threshold, it is determined that the slope deformation tends to converge, the risk change is gradual, the risk is effectively controlled, the emergency response effect meets the standard, and the system automatically issues an emergency flow restriction and downgrade command to gradually lift the high-intensity emergency control status. Step 1004: When the absolute value of the slope is greater than or equal to the preset rate of change threshold, it is determined that the slope deformation continues to develop, the risk changes drastically, the risk is not under control and continues to rise, the system continues to maintain the highest level alarm, and at the same time automatically expands the capacity to increase the investment in drainage equipment, strengthens the drainage and dredging efforts, and continuously manages the risk. Through the above technical solutions, the effectiveness of emergency response is quantified by standardized slope calculation, so that the effectiveness of emergency response can be quantified, verified and dynamically adjusted, avoiding the problems of blind early warning or inadequate response, and greatly improving the scientific and intelligent level of slope emergency management. The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A slope monitoring and early warning system for open-pit coal mines, characterized in that, include: The monitoring data acquisition module is used to conduct multi-source collaborative monitoring of open-pit coal mine slopes through slope deformation monitoring equipment, positioning monitoring equipment, internal strain monitoring equipment and on-site image acquisition equipment, and to acquire slope surface deformation data, internal strain data and environmental video streams. The hydrological model building module is used to construct a hydrological evolution model of the sump at the bottom of the spoil heap, and a non-steady flow calculation module is built based on Darcy's law to simulate the water level seepage rate. The mechanical parameter correction module is used to construct a subsystem for correcting the mechanical parameters of soil and rock. It uses a backpropagation neural network algorithm to update the cohesion and internal friction angle parameters in the model based on the internal strain data, and to deduce the evolution process of the slope sliding surface. The data alignment module is used to perform time-domain slicing and traffic correlation analysis on monitoring data from different sources based on spatiotemporal references, perform time-series alignment operations, and generate aligned monitoring data streams. The 3D projection module is used to project the aligned monitoring data and simulation results onto the 3D geographic information system model to generate deformable cloud maps. The early warning signal generation module is used to divide the slope deformation curve characteristics and the evolution trend of the sliding surface into three sub-stages: initial acceleration, intermediate acceleration and accelerated acceleration, and generate corresponding graded early warning signals. The risk assessment module is used to determine the slope instability risk level when an early warning signal is generated; The emergency response module is used to generate safety protection action commands when the risk level exceeds a preset threshold, and to link the on-site alarm equipment, on-site information display equipment and drainage execution equipment to perform emergency response.

2. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, The multi-source collaborative monitoring of open-pit coal mine slopes using slope deformation monitoring equipment, positioning monitoring equipment, internal strain monitoring equipment, and on-site image acquisition equipment includes: Slope deformation monitoring equipment was used to scan key areas of the slope, and location monitoring points were set up to assist in the monitoring of areas not covered by the monitoring equipment. Internal strain monitoring equipment was installed within the rock and soil of the spoil heap slope to collect strain data. A backup strain sensor is implanted in the borehole where each internal strain monitoring device is located, and the two form a parallel data acquisition channel. When the internal strain monitoring equipment breaks due to the tension of the soil and rock, the data channel of the backup strain sensor is switched.

3. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, The constructed subsystem for correcting the mechanical parameters of soil and rock mass includes: Mark the transverse burial line along the slope height direction, and shorten the sensor placement spacing at the junction of soil and rock layers to increase the monitoring point density; The internal strain monitoring equipment is fitted with a protective sleeve, which is then filled with cement mortar and epoxy resin for protection. The collected strain data is input into the backpropagation neural network algorithm to correct the geomechanical parameters in the simulation model.

4. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, The analysis of temporal slicing and traffic correlation of monitoring data from different sources based on spatiotemporal reference includes: The core module is protected by encapsulation, and a signal isolation layer and grounding protection are added. The data processing algorithm based on Kalman filtering and wavelet analysis is executed. First, high-frequency vibration interference signals are removed by wavelet thresholding. Then, Kalman filtering is used to predict and correct atmospheric delay errors. By integrating data from positioning monitoring equipment, slope deformation monitoring equipment, and internal strain monitoring equipment, and calculating the correlation coefficient within each time-domain slice, time delay compensation and alignment are performed on asynchronous data streams.

5. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, The process of projecting the aligned monitoring data and simulation results onto the 3D geographic information system model includes: In the data visualization platform, a unified reference point is determined, the coordinates of the reference point are measured and input into the platform, and correction parameters are calculated to adjust the spatial position of the 3D model. In the system trend analysis module, reduce the data sliding window size parameter to increase the weight of monitoring data in trend analysis; Based on the characteristics of the tangent angle change in the deformation curve, the slope deformation stage is identified and an early warning of the corresponding level is triggered.

6. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, The emergency response functions of the alarm devices, on-site information display devices, and drainage execution devices at the linked site include: Develop mobile applications and deploy on-site information display devices to push early warning information to technicians' terminals and on-site work areas via wireless network; When the system is determined to be in a state of imminent slippage, a start command is sent to the drainage execution equipment and drainage control system.

7. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, Before acquiring the slope surface deformation data, internal strain data, and environmental video stream, the following steps are also included: Curved steel rails were laid along the edge of the spoil heap, and slope deformation monitoring equipment was installed on a platform that could move along the rails. The slope deformation monitoring equipment is controlled by a drive device to move along the track and stop at the preset positioning sensor position for monitoring.

8. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, The tiered early warning signals include: The initial acceleration stage warning corresponds to the state where the slope begins to show continuous deformation, and the patrol mechanism is activated. The warning for the accelerated deformation stage corresponds to a state where the deformation rate increases, and drainage and load reduction preparations are initiated. Acceleration phase warning corresponds to the state of an impending landslide, triggering audible and visual alarms and evacuating personnel and equipment.

9. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, After performing time-domain slicing and traffic correlation analysis on monitoring data from different sources based on a spatiotemporal reference, the method further includes: When a data packet loss is detected for a specific monitoring device, the lost data packet is numerically filled using the average of the historical alignment data stored in the buffer medium. A data transmission system is constructed using a dual-link mode that combines wired transmission with long-distance wireless backup.

10. The open-pit coal mine slope monitoring and early warning system according to claim 1, characterized in that, After generating a safety protection action instruction when the risk level exceeds a preset threshold, the method further includes: After sending the safety protection action command, obtain on-site feedback data; Calculate the slope of change of the feedback data within a preset time window; When the magnitude of the change slope is less than the preset change rate threshold, the emergency response is deemed effective, and a downgrade command is sent. When the slope of change is not less than the preset rate of change threshold, the highest level alarm will be maintained and the amount of drainage equipment put into operation will be increased.

Images