Strip mine slope fatigue instability dynamic evaluation method and system based on quantitative model

Abstract

The invention relates to the technical field of slope engineering stability evaluation, in particular to a strip mine slope fatigue instability dynamic evaluation method and system based on a quantitative model. The method comprises the following steps: acquiring physical field parameter data of a strip mine slope; preprocessing the physical field parameter data; establishing a fatigue damage dynamic evaluation model; quantitative characterization parameters of the slope damage state are obtained; constructing a chaotic phase change enhancement model, a path connectivity enhancement model and a health state analysis model; and obtaining analysis results of mechanism early warning, path prediction and probability forecasting. According to the method, a standard field theory and a quantum entanglement enhancement algorithm are introduced, a dynamic evaluation system driven by a physical rule is constructed, multi-physical field perception and a fractional order chaos model are fused, and 97% high-precision identification and 63-day advanced early warning of a damage path are realized; by creating a mechanism-path-probability three-dimensional intelligent evaluation system, a major span from passive monitoring to active regulation and control and full life cycle decision making is realized.

Description

Technical Field

[0001] This invention relates to the field of slope engineering stability evaluation technology, specifically to a dynamic evaluation method and system for fatigue instability of open-pit mine slopes based on a quantitative model. Background Technology

[0002] Slope stability in open-pit mines is a core issue for safe production. As mining progresses to deeper and more intensive operations, slopes are prone to fatigue damage under the combined effects of cyclic loading and blasting vibrations, leading to instability and failure.

[0003] Existing evaluation methods mainly rely on surface displacement monitoring or limit equilibrium analysis. Although there has been some progress, there are obvious shortcomings: First, the monitoring parameters do not fully cover multi-physics field data such as ground resistivity, geomagnetic perturbation, rock mass memory deformation vector, damage evolution potential well depth, rock mass microseismic damage entropy, and structural surface dynamic stiffness degradation index; second, data processing methods are difficult to effectively preserve topological features; third, model construction is mostly based on empirical formulas and lacks description of nonlocal memory effects and chaotic behavior; fourth, the evaluation results focus on macroscopic risks and are difficult to achieve full-chain quantification from damage mechanism to remaining life.

[0004] To address the above shortcomings, this invention proposes a dynamic evaluation method for fatigue instability of open-pit mine slopes. By fusing multi-physics parameters, a comprehensive perception of internal damage is achieved; by introducing a gauge field theory adversarial learning algorithm and an improved rock mass memory-temporal causal decoupling algorithm, topological invariants of damage evolution are effectively extracted; by constructing an evaluation model integrating quantum entanglement-enhanced data assimilation, fractional-order memory constitutive model, and chaotic driving dynamics, the empirical limitations of existing models are overcome; by quantifying parameters, a three-dimensional evaluation of mechanism early warning, path prediction, and probability forecasting is achieved; and by combining proactive intervention and intelligent decision-making, a leap from passive monitoring to proactive control is realized, significantly improving the predictability, accuracy, and reliability of the evaluation. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a dynamic evaluation method and system for fatigue instability of open-pit mine slopes based on a quantitative model.

[0006] In a first aspect, the present invention provides a dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model, comprising the following steps: acquiring physical field parameter data of the open-pit mine slope; preprocessing the physical field parameter data to obtain preprocessing results; establishing a dynamic evaluation model for fatigue damage based on the preprocessing results; using the dynamic evaluation model for fatigue damage, combined with a quantum entanglement-enhanced data assimilation algorithm, a fractional-order memory constitutive algorithm, and a chaotic driving dynamics algorithm, acquiring quantitative characterization parameters of the slope damage state; constructing a chaotic phase transition enhancement model, a path connectivity enhancement model, and a health status analysis model based on the quantitative characterization parameters; and obtaining analysis results for mechanism early warning, path prediction, and probability prediction based on the chaotic phase transition enhancement model, the path connectivity enhancement model, and the health status analysis model. The analysis results enable a dynamic evaluation of fatigue instability of open-pit mine slopes.

[0007] Optionally, acquiring physical field parameter data of open-pit mine slopes includes: acquiring first-type physical field parameter data of open-pit mine slopes, which includes data on the rate of change of resistivity, intensity of geomagnetic perturbation, coefficient of variation of acoustic velocity tomography, chemical ion concentration gradient of fracture water, radon gas release rate, change in fractal dimension of surface texture, area of ​​acceleration response spectrum, and entropy value of thermal infrared radiation; acquiring second-type physical field parameter data of open-pit mine slopes, which includes data on rock mass memory deformation vector, slope damage evolution potential well depth, rock mass microseismic damage entropy, and dynamic stiffness deterioration index of slope structural surfaces.

[0008] Optionally, the physical field parameter data is preprocessed to obtain the preprocessing results, including: using gauge field theory and spatiotemporal symmetry breaking adversarial learning algorithm to denoise and normalize the first type of physical field parameter data to obtain the preprocessing results of the first type of physical field parameter data; using an improved rock mass memory-temporal causality decoupling algorithm to perform structured cleaning, memory effect decoupling, causal discovery and normalization processing on the second type of physical field parameter data to obtain the preprocessing results of the second type of physical field parameter data.

[0009] Optionally, establishing a dynamic fatigue damage assessment model based on the preprocessing results includes: using the first type of physical field parameters as state constraint variables and data assimilation techniques to obtain correction results for model parameters based on the preprocessing results; constructing a constitutive relation for the cumulative evolution of rock mass damage based on the correction results and the rock mass memory deformation vector and the slope damage evolution potential well depth as core damage state variables; building an evolutionary framework for damage state variables through the correction results and the constitutive relation; constructing a dynamic model for damage evolution based on the evolutionary framework and the rock mass microseismic damage entropy and the slope structural surface dynamic stiffness degradation index as damage evolution driving variables; obtaining the correlation between nonlinear coupling and spatiotemporal propagation among physical field parameters using a physics-guided spatiotemporal graph neural network based on the dynamic model; obtaining dynamic identification results of the damage evolution path from stability to instability of the slope based on the correlation; and constructing a dynamic fatigue damage assessment model based on the correction results, the constitutive relation, the evolutionary framework, the dynamic model, the correlation, and the dynamic identification results.

[0010] Optionally, using the aforementioned fatigue damage dynamic assessment model, combined with a quantum entanglement-enhanced data assimilation algorithm, a fractional-order memory constitutive algorithm, and a chaotic-driven dynamics algorithm, the quantitative characterization parameters of the slope damage state are obtained as follows: Based on the fatigue damage dynamic assessment model, the preprocessed data and background state are fused and corrected using a quantum entanglement-enhanced fractional-order data assimilation algorithm to obtain a corrected model state vector; based on the model state vector, evolution calculations are performed using a fractional-order memory damage constitutive relation to obtain the updated results of the rock mass memory deformation vector and the slope damage evolution potential well depth; based on the updated results, the interaction between nodes is calculated using a topological damage propagation network, combined with the potential well depth gradient and deformation curvature, to obtain a damage propagation probability matrix; based on the damage propagation probability matrix, chaotic damage state variables are obtained through a chaotic-driven damage dynamics model; based on the chaotic damage state variables, a loss function and attention coefficient are obtained using a physics-guided spatiotemporal graph neural network; based on the loss function and the attention coefficient, the potential well depth evolution and stress history are probabilistically integrated using the Monte Carlo path integral method to obtain the aging failure probability and fatigue damage index.

[0011] Optionally, the formula for the quantum entanglement-enhanced fractional-order data assimilation algorithm is as follows: in, This is the corrected model state vector. For the observed data vector, The observation error covariance matrix, The background state vector, The model error covariance matrix is... For regularization parameters, The order of the fractional derivative. It is the gamma function. is Planck's constant. For quantum wave function, The number of state variables, For the index of the state variable, For the model state vector, For time, For time infinitesimal elements, For the first One state variable, For the observation operator; the formula for the fractional-order memory impairment constitutive relation is as follows: in, This is the rock mass memory deformation vector. The depth of the potential well for slope damage evolution. This is the threshold for the critical potential well depth. The memory decay coefficient, For memory kernel index, For nonlinear coupling coefficients, Let be the gradient vector of the potential well depth. It is a constant. For time, For time infinitesimal elements, The Impaired Memory Index This is the damage sensitivity coefficient. As a damage variable, The gamma function is used; the chaotic-driven damage dynamics model satisfies the following expression: in, For damage state variables, For deterministic dynamic functions, For driving term functions, For random resonance amplitude, The resonant frequency, The phase angle, The attenuation rate, For time, This is a chaotic noise term. This represents the entropy of microseismic damage in the rock mass. It is the dynamic stiffness deterioration index of the slope structure surface.

[0012] Optionally, the analysis results for mechanism early warning, path prediction, and probability forecasting based on the chaotic phase transition enhancement model, the path connectivity enhancement model, and the health status analysis model include: using the chaotic phase transition enhancement model to obtain the analysis results of chaotic phase transition; using the path connectivity enhancement model to obtain the analysis results of damage propagation path and through slip surface; and using the health status analysis model to obtain the analysis results of instability risk level and remaining fatigue life.

[0013] Optionally, the chaotic phase transition enhancement model satisfies the following expression: in, For the degree of chaotic phase transition, For the first One Lyapunov index, The entropy of microseismic damage. The stiffness degradation index is... For the quantum annealing optimization function, For time, The total number of Lyapunov indices. The path connectivity enhancement model satisfies the following expression: (The path connectivity enhancement model is a chaotic damage state variable.) in, For path connectivity, For the sliding path, Indicates damage from node propagation to nodes The probability, This is the rock mass memory deformation vector. The gradient is the potential well depth. For path persistence measurement based on topology data analysis, For scale parameters, The path element is defined; the health status assessment model satisfies the following expression: in, Indicates the risk level. For risk level mapping function, For the future The probability of failure within the specified time limit. The fatigue damage index, For real-time correction terms in dynamic Bayesian networks, For the first Fatigue damage path clustering, For expiration time, For conditional mathematical expectation, This represents the damage evolution path.

[0014] Optionally, based on the analysis results, the dynamic evaluation of fatigue instability of open-pit mine slopes includes: Based on the analysis results of the mechanism-based early warning, it is concluded that the slope stability has entered a sensitive decay period. Based on this conclusion, decision-making instructions for high-frequency monitoring and causal control are obtained. Based on the analysis results of path prediction, the spatial location of internal damage convergence nodes and key transmission paths of the slope is obtained. Based on this spatial location, a deployment plan for precise monitoring and reinforcement target areas is obtained. Based on the analysis results of probability forecasting, a quantitative expectation of the slope's quantitative risk level and remaining service life is obtained. Based on this quantitative expectation, a strategy plan for graded early warning response and predictive maintenance is obtained. Based on the decision-making instructions, the deployment plan, and the strategy plan, a complete dynamic evaluation of fatigue instability of open-pit mine slopes is achieved.

[0015] Secondly, the present invention provides a dynamic evaluation system for fatigue instability of open-pit mine slopes based on a quantitative model. The system uses the aforementioned dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model. The system includes an acquisition module, a construction module, an analysis module, and an output module. The acquisition module is used to acquire physical field parameter data of open-pit mine slopes; and to preprocess the physical field parameter data to obtain preprocessing results. A construction module is used to establish a dynamic evaluation model for fatigue damage based on the preprocessing results; and to utilize the dynamic evaluation model for fatigue damage in combination with a quantum entanglement-enhanced data assimilation algorithm, a fractional-order memory constitutive algorithm, and a chaotic driving dynamics algorithm. The analysis module is used to obtain quantitative characterization parameters of slope damage status; based on the quantitative characterization parameters, a chaotic phase transition enhancement model, a path connectivity enhancement model, and a health status analysis model are constructed; based on the chaotic phase transition enhancement model, the path connectivity enhancement model, and the health status analysis model, the analysis results of mechanism early warning, path prediction, and probability forecasting are obtained. The output module is used to dynamically evaluate the fatigue instability of open-pit mine slopes based on the analysis results.

[0016] Compared with the prior art, the present invention has the following significant advantages: (1) A significant transformation has been achieved in the evaluation theoretical framework, shifting from experience-dependent to physical law-driven approaches. By introducing gauge field theory, fiber bundle theory, and quantum entanglement-enhanced assimilation algorithms into slope stability analysis, and by reconstructing sensor networks as gauge fields on spatiotemporal manifolds, and by utilizing Yang-Mills action to characterize the deep coupling mechanism of multiphysics fields, a dynamic evaluation system based on physical laws has been established. This system overcomes the limitations of existing technologies that rely on empirical formulas and simplification assumptions. By outputting a Chern-Simons topological invariant sequence that is immune to local perturbations and sensitive to system topological phase transitions, it achieves a leap from shallow data fitting to essential feature extraction, laying a solid theoretical foundation for accurate evaluation.

[0017] (2) A comprehensive sensing system covering multiple physical field parameters was constructed, achieving a leapfrog improvement in monitoring capabilities. A complete sensing system was constructed, covering eight types of first-class physical field parameters, such as the rate of change of resistivity, the intensity of geomagnetic perturbation, and the coefficient of variation of acoustic velocity tomography, as well as four types of second-class physical field parameters, such as the rock mass memory deformation vector and the depth of the slope damage evolution potential well. This system breaks through the limitation of existing technologies that rely solely on surface displacement monitoring, and can capture the entire process of internal damage evolution from multiple dimensions such as chemical field, thermal field, and electromagnetic field. Engineering practice shows that this system can detect resistivity anomalies 42 days in advance and provide early warning of instability risk 63 days before significant deformation through chaotic phase transition degree, significantly improving the early warning capability.

[0018] (3) An evaluation model integrating fractional-order memory and chaotic driving was established, achieving a qualitative leap in predictive capability. By introducing fractional-order derivatives and potential well theory, a constitutive relation capable of describing the nonlocality and memory effect of rock mass damage was constructed; combined with chaos theory and stochastic resonance, a chaotic-driven damage dynamics model was established, breaking through the local assumptions and linear limitations of existing models. Based on topological damage propagation networks and physically guided spatiotemporal graph neural networks, dynamic identification of damage path connectivity and learning of multi-parameter spatiotemporal coupling relationships were realized, significantly improving the model's ability to capture critical features before slope instability and its prediction accuracy.

[0019] (4) A three-dimensional evaluation system of mechanism, path, and probability was established, realizing an intelligent upgrade of decision support. By constructing a three-dimensional evaluation system of mechanism early warning, path prediction, and probability forecasting, the Lyapunov exponent spectrum was used to accurately quantify chaotic phase transitions, persistent cohomology analysis was used to identify essential damage paths, and a dynamic Bayesian network was used to achieve real-time updates of risk levels. On this basis, by introducing a stress wave active intervention mechanism, a biomimetic self-healing path blocking system, and digital twin-assisted decision-making, a fundamental shift from passive monitoring to active regulation and from current response to optimal decision-making throughout the entire life cycle was achieved.

[0020] (5) The superior effectiveness of the method has been verified through engineering practice, demonstrating significant application value. In the field application of a large open-pit copper mine, the spatial consistency between the damage path identified by the method of this invention and the actual microseismic events reached 97%. Based on the predicted time-dependent failure probability, it successfully guided reinforcement measures and avoided potential landslides. These results fully verify the significant advantages of this invention in terms of early warning timeliness, quantitative accuracy, and engineering applicability, providing a theoretically advanced, technically reliable, and highly effective full-chain dynamic evaluation solution for open-pit mine slope safety. Attached Figure Description

[0021] Figure 1 This is a flowchart of a dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of the dynamic evaluation system for fatigue instability of open-pit mine slopes based on a quantitative model, according to an embodiment of the present invention. Detailed Implementation

[0022] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0023] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0024] Please see Figure 1 The present invention provides a dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model, the method comprising the following steps: S1. Obtain the physical field parameter data of the open-pit mine slope.

[0025] S1 includes the following steps: S11. Obtain the first type of physical field parameter data of the open-pit mine slope. The first type of physical field parameter data includes data on the rate of change of resistivity, intensity of geomagnetic perturbation, coefficient of variation of acoustic tomography, chemical ion concentration gradient of fracture water, radon gas release rate, change of fractal dimension of surface texture, area of ​​acceleration response spectrum, and entropy value of thermal infrared radiation.

[0026] S11 includes the following steps: S111. Obtain data on the rate of change of resistivity of the open-pit mine slope.

[0027] In one embodiment, the first step is to establish a survey network and perform baseline measurements. A series of electrodes are laid out in a grid or profile pattern in the target slope area. Using a high-precision resistivity meter, a comprehensive resistivity measurement is performed when the slope is in a relatively stable state to obtain background resistivity data for the area, which serves as a reference for subsequent changes.

[0028] Furthermore, implement periodic dynamic monitoring. Based on slope stability requirements and geological conditions, set a fixed monitoring cycle, such as weekly or monthly, repeating resistivity measurements using the same equipment and parameters. Each measurement must ensure consistency in electrode position, burial depth, and instrument parameters to eliminate systematic errors.

[0029] Further, data processing and rate of change calculation are performed. Each monitoring data point is compared with background data, and a three-dimensional cross-sectional map of the ground resistivity is obtained using specialized inversion software. The average resistivity value of a specific area of ​​interest, such as a potential slip surface, is calculated, and the rate of change of ground resistivity is calculated using the following formula: Rate of change = (Current value - Benchmark value) / Benchmark value × 100% The advantage of this method is that it can more sensitively capture changes in moisture and fissures inside the soil and rock mass by quantitatively monitoring the rate of change of resistivity.

[0030] S112. Obtain data on the intensity of geomagnetic disturbances on the slope of the open-pit mine.

[0031] In one embodiment, highly sensitive magnetometers are first deployed. Multiple reference stations and monitoring stations are set up in the slope monitoring area and its surrounding stable zone, equipped with high-precision proton precession magnetometers or optically pumped magnetometers. All instruments must be time-synchronized to eliminate the influence of global magnetic field variations such as diurnal variations.

[0032] Furthermore, simultaneous continuous observation and data recording will be conducted. All magnetometers will begin continuous data acquisition simultaneously, with a sampling frequency sufficient to capture any possible perturbations. Observations will continue for a period of time to obtain a complete picture of the geomagnetic field background trends and any possible anomalous signals.

[0033] Further, data preprocessing and perturbation intensity calculation are performed. Data from the monitoring station is subtracted from the base station data to eliminate common geomagnetic variations. Then, the difference data is filtered, such as by bandpass filtering, to remove high-frequency noise and long-term drift. The root mean square value or peak value of the filtered signal within a specific time window is calculated as the geomagnetic perturbation intensity for that time period.

[0034] The advantage of this method is that it achieves high sensitivity and real-time dynamic evaluation of slope fatigue instability through synchronous geomagnetic monitoring network and quantification of micro-disturbance intensity. It can capture small deformation signals at an early stage, significantly improve safety and accuracy, and is clearly different from existing passive methods that rely on displacement or stress monitoring.

[0035] S113. Obtain the data of the coefficient of variation of sound velocity tomography on the slope of the open-pit mine.

[0036] In one embodiment, an acoustic wave transmitting and receiving array is first deployed. Acoustic wave transmitting transducers and receiving sensors are regularly arranged in boreholes, on the slope surface, or in tunnels along the slope, forming a crisscrossing ray network. This ensures that the rays cover as much of the slope rock mass as possible that needs to be probed.

[0037] Furthermore, acoustic travel time data is collected and tomographic inversion is performed. The transmitting transducer sequentially excites acoustic signals, and all receiving sensors record the arrival time of the direct wave. Using the travel time data, an acoustic velocity distribution image inside the slope rock mass is reconstructed based on a specific inversion algorithm.

[0038] Furthermore, the coefficient of variation of the velocity field is calculated. In the inverted sound velocity distribution map, the region of interest is delineated, the standard deviation of the sound velocity values ​​of all pixels within the region of interest is calculated, and then divided by the average value to obtain the coefficient of variation of the sound velocity tomography for that region.

[0039] The advantage of this method is that it quantitatively characterizes the damage and heterogeneity of the internal structure of the rock mass through the coefficient of variation of the sound velocity field, directly reflecting the structural deterioration caused by fatigue accumulation. Compared with existing surface monitoring or single parameter threshold judgment, it can more fundamentally and sensitively predict potential slip surfaces and realize dynamic stability evaluation based on the evolution of internal structural state.

[0040] S114. Obtain data on the chemical ion concentration gradient of fissure water on open-pit mine slopes.

[0041] In one embodiment, sampling points are first determined and water samples are collected. Sampling points are established at different elevations and geological units on the slope, at exposed fissure water points, or in specially set monitoring wells. Water samples are collected periodically using clean sampling bottles and measured on-site. The values, temperature, and conductivity parameters are recorded, along with the sampling location and elevation.

[0042] Further analysis of ion concentrations was conducted. Water samples were sent to the laboratory for precise analysis of major anions and cations using equipment such as ion chromatography and inductively coupled plasma atomic emission spectrometry. , , , , , , The concentration.

[0043] Further, the ion concentration gradient is calculated. The concentration of specific ions, such as stress-sensitive or deep tracer ions, at different sampling points, especially along the potential slip direction or elevation direction, is fitted to the spatial location; the slope of the concentration change with space is calculated, or the ratio of the concentration difference between two key points to their distance is directly calculated, which is the concentration gradient of that ion.

[0044] The advantage of this method is that it can sensitively capture the expansion of deep fractures and changes in the seepage field by using the water chemical ion concentration gradient index, quantify fatigue accumulation from the perspective of stress corrosion and chemical damage inside the rock mass, and realize advanced chemical field perception of potential slip surface activation.

[0045] S115. Obtain data on the radon emission rate of open-pit mine slopes.

[0046] In one embodiment, the first step is to set up measurement points and prepare the equipment. Representative measurement points are selected on the slope surface, such as in soil or rock fissures, and the surface is cleaned to avoid interference; an accumulator radon measuring cup or an active radon emission rate measuring instrument is prepared.

[0047] Further, on-site measurements and data collection were conducted. The measuring cup was placed upside down on the soil at the measuring point, the mouth of the cup was sealed, and radon gas was allowed to accumulate naturally for a period of time, from several days to several weeks; after the period, it was retrieved, and the number of nuclear tracks inside was analyzed in the laboratory.

[0048] Furthermore, the radon release rate was calculated and analyzed. The radon release rate was calculated based on the nuclear track density, exposure time, and instrument calibration coefficient, and the spatial distribution of the release rate at each measuring point and its variation over time were analyzed.

[0049] The advantage of this method is that by monitoring the radon gas release rate, an indicator that is extremely sensitive to the development of microcracks in rock mass, it enables advanced and indirect quantitative perception of the fatigue damage accumulation process inside the slope, breaking through the limitations of existing macroscopic monitoring and providing important information for slope instability.

[0050] S116. Obtain data on the variation of the fractal dimension of the surface texture of the open-pit mine slope.

[0051] In one embodiment, high-resolution images of the slope surface are first acquired. A drone equipped with a high-definition camera or ground scanner is used to periodically photograph the slope surface, obtaining digital orthophotos covering the entire monitoring area.

[0052] Further image processing and texture feature extraction are performed. The acquired images are converted to grayscale, denoised, and enhanced. Fractal dimension calculation algorithms such as box counting are used to divide the images into grids, count the number of grids containing texture features at different scales, and calculate the fractal dimension through linear regression in double logarithmic coordinates.

[0053] Furthermore, the change in fractal dimension is calculated. The fractal dimension calculated for each monitoring is compared with the previous fractal dimension, and the change is calculated.

[0054] The advantage of this method is that by quantifying the changes in the fractal dimension of surface texture, it transforms damage evolution such as microcrack propagation and rock loosening, which are difficult to be identified by the naked eye, into traceable continuous data, thereby achieving keen capture and advanced perception of the fatigue deterioration process of slope surface.

[0055] S117. Obtain data on the acceleration response spectrum area of ​​the open-pit mine slope.

[0056] In one embodiment, accelerometers are first deployed to collect background vibration signals. High-sensitivity triaxial accelerometers are installed at key locations on the slope, such as the top and bottom. During periods without strong interference from heavy machinery operations, the acceleration response signals of the slope under environmental micro-vibrations, such as ground pulsations and vibrations from distant vehicles, are continuously recorded.

[0057] Further signal processing and response spectrum calculation are performed. The acquired raw acceleration time history data are filtered and detrended; then, a set of single-degree-of-freedom system models with different natural frequencies and damping ratios are input, and the maximum response of each model is calculated to obtain the acceleration response spectrum of the measuring point.

[0058] Further, the response spectral area is calculated within the selected frequency range. arrive Integrating the acceleration response spectrum curve, the area of ​​the integral is the area of ​​the acceleration response spectrum.

[0059] The advantage of this method is that it obtains the acceleration response spectrum area through environmental micro-vibration, realizing the quantitative characterization of slope dynamic response and cumulative fatigue damage, and has the characteristics of dynamic, sensitive and continuous monitoring.

[0060] S118. Obtain data on the thermal infrared radiation entropy value of the open-pit mine slope.

[0061] In one embodiment, a sequence of thermal infrared images of the slope is first acquired. Using a thermal infrared imager, the slope is scanned periodically during periods of relatively stable ambient temperature, such as at night or early morning, to acquire high spatial resolution thermal infrared images, while maintaining consistency in observation conditions and environment as much as possible.

[0062] Furthermore, image preprocessing and temperature field standardization are performed. The thermal infrared image is radiometrically calibrated and atmospherically corrected to convert it into accurate surface temperature data. To eliminate interference from absolute ambient temperature values, the temperature field data is normalized and converted into... The grayscale image focuses on the relative distribution characteristics of the temperature field.

[0063] Further, the thermal infrared radiation entropy value is calculated. Based on the information entropy theory, the information entropy of the normalized temperature field image is calculated. The specific steps are as follows: statistically analyze the probability distribution of the image grayscale values, and then calculate using the information entropy formula, which is as follows: in, For the first The probability of each gray level appearing This represents the entropy value of thermal infrared radiation. The level of entropy reflects the degree of disorder in the temperature field distribution.

[0064] The advantage of this method is that it utilizes thermal infrared image sequences to quantify the disorder of the slope temperature field by calculating the radiation entropy value, and sensitively inverts the thermal anomalies caused by the internal damage evolution of the rock mass, thus realizing a non-contact, large-area dynamic quantitative characterization of the fatigue instability process.

[0065] S12. Obtain the second type of physical field parameter data of the open-pit mine slope. The second type of physical field parameter data includes data on rock mass memory deformation vector, slope damage evolution potential well depth, rock mass microseismic damage entropy, and slope structural surface dynamic stiffness deterioration index.

[0066] S12 includes the following steps: S121. Obtain data on the rock mass memory deformation vector of the open-pit mine slope.

[0067] In one embodiment, a precision servo-controlled loading device and a high-precision displacement gauge array are first deployed in the stable rock mass of key areas such as the slope shoulder and toe.

[0068] Furthermore, during the stabilization period or initial stage of the slope, a series of low-level quasi-static cyclic loads are applied and then completely unloaded to establish a zero-memory baseline state in which the deformation can be fully recovered.

[0069] Furthermore, during subsequent monitoring cycles, the above loading-unloading process is repeated periodically, and the irreversible residual plastic deformation and its spatial direction relative to the initial baseline state are accurately measured. The resulting composite material is the rock mass memory deformation vector that characterizes the direction of cumulative damage in the rock mass.

[0070] The advantage of this method is that by actively loading to establish a zero-memory benchmark, it can accurately capture the irreversible plastic deformation vector, thus realizing a qualitative change from deformation monitoring to damage direction and trend early warning, and achieving advanced and dynamic quantitative evaluation of slope fatigue instability.

[0071] S122. Obtain data on the depth of the slope damage evolution potential well in open-pit mine slopes.

[0072] In one embodiment, multiphysics temporal monitoring data of slope displacement, acceleration, groundwater level, or microseismic activity are first collected synchronously over a long period.

[0073] Furthermore, based on the multiphysics time-series monitoring data, phase space reconstruction is performed to recover the high-dimensional phase space trajectory that can characterize the dynamic evolution of the slope system.

[0074] Furthermore, by applying nonlinear dynamical system analysis methods, the Focke-Planck equations describing the probabilistic evolution of the system state are solved, thereby reversing the equivalent potential function characterizing the system stability.

[0075] Furthermore, on the potential function image, the potential energy difference between the bottom of the potential well corresponding to the current stable state and the critical state leading to instability is calculated. This difference is the potential well depth of the slope damage evolution.

[0076] The advantage of this method is that by fusing multi-physics field data and reconstructing phase space, the potential function of the system is dynamically inverted, and the critical energy barrier of the slope sliding from a stable state to instability is accurately characterized, thus realizing an advanced and quantitative assessment of the energy level of system stability.

[0077] S123. Obtain data on the microseismic damage entropy of the rock mass on the open-pit mine slope.

[0078] In one embodiment, a microseismic monitoring network deployed inside the slope is first used to continuously record microseismic events generated by rock mass fracturing, and to accurately locate their spatial coordinates, calculate their released energy, and record their occurrence time, forming a complete sequence of microseismic events.

[0079] Furthermore, based on the information entropy theory, ordered quantification is performed from the following three dimensions: spatial entropy is calculated based on the distribution probability of events in the spatial grid, energy entropy is calculated based on the distribution probability of event energy levels, and time series entropy is calculated based on the distribution characteristics of event time interval sequences. Furthermore, by combining the entropy values ​​of the above three dimensions, the degree of disorder in rock mass microfracture activity can be quantitatively characterized, namely, the entropy of rock mass microseismic damage.

[0080] The advantage of this method is that by calculating the degree of order of microseismic events in the spatiotemporal-energy domain, the disordering process of the rock mass fracture system is quantitatively characterized by damage entropy, thereby realizing a holistic and advanced quantitative evaluation of the damage evolution state inside the slope rock mass.

[0081] S124. Obtain data on the dynamic stiffness deterioration index of the slope structure surface of the open-pit mine slope.

[0082] In one embodiment, a distributed array of fiber optic sensors capable of emitting and receiving shear waves is first deployed in parallel along key potential sliding structural surfaces inside the slope, such as joints and fissures.

[0083] Furthermore, a standard shear wave signal is periodically excited at one end of the sensor, and the wave velocity change of the signal after it passes through a specific structural surface and reaches the other end is monitored.

[0084] Furthermore, based on elastic wave theory, the equivalent dynamic shear stiffness of the structural surface is dynamically calculated from the shear wave velocity passing through the structural surface.

[0085] Furthermore, the current measured value is compared with the structural surface in good condition or the initial dynamic stiffness benchmark value of the last monitoring period to calculate the relative degree of deterioration. This ratio is the dynamic stiffness deterioration index of the slope structural surface.

[0086] The advantage of this method is that it can directly quantify the real-time deterioration of the structural surface stiffness parameters by dynamically detecting shear wave velocity, thereby achieving advanced and accurate diagnosis of the potential slip surface damage accumulation state.

[0087] S2. Preprocess the physical field parameter data to obtain the preprocessing result.

[0088] S2 includes the following steps: S21. Using gauge field theory and spacetime symmetry breaking adversarial learning algorithm, the first type of physical field parameter data is denoised and normalized to obtain the preprocessed result of the first type of physical field parameter data.

[0089] In one embodiment, the slope monitoring system is first modeled based on fiber bundle theory, treating it as the main fiber bundle, defining relevant concepts, and the generator learns the gauge connections to define covariant derivatives. Then, an adversarial training framework based on gauge invariance is designed, with the core of the discriminator being the gauge-invariant action. After training convergence, topological flow normalization based on Chern-Simons theory is performed using optimized gauge connections. The Chern-Simons term is constructed and integrated to obtain gauge-invariant topological invariants as preprocessing results, achieving denoising and normalization.

[0090] Specifically, a physical state space model based on fiber bundle theory is first introduced. The entire slope monitoring system is considered as a master fiber bundle. Here, the base manifold represents the physical spacetime of the slope, and the fibers are the internal physical state spaces belonging to each point in spacetime, such as the phase space describing the microstructure of the rock mass. The readings of each sensor are considered as a local cross-section. A gauge connection, namely the vector potential, is defined. This describes the rules for parallel movement between states within different spatiotemporal points. Among them, In three-dimensional space coordinates, For time.

[0091] It is important to note that generators The goal is no longer to output data, but to learn this normative connection. Thus, a covariant derivative is defined. It satisfies the following expression: in, For the first Partial derivative operators for coordinates, It is the imaginary unit.

[0092] Furthermore, an adversarial training framework based on gauge invariance is designed. The core of the discriminator D is a gauge-invariant action. It consists of the Yang-Mills action and the matter-field coupling term, and its mathematical expression is: in, It is a four-dimensional spacetime manifold. The reference connection is derived from an ideal physical model, representing the ideal parallel movement rule under noise-free conditions. This is the energy scale cutoff parameter, used to distinguish between physical effects and non-physical ultraviolet dispersion. Let be the norm on the fiber bundle. The material field represents the physical quantities observed by the sensor, such as displacement and stress. For traces on Lie algebra, To standardize the coupling constant, which determines the coupling strength between field quanta, For spatiotemporal index, The Lorentz index is used to identify the components of a spacetime coordinate system. These are invariant volume elements used to preserve Lorentz invariance during integration. The gauge field strength tensor is used to describe the strength and properties of gauge fields. for The inverse form, The differential symbol, Four-dimensional spacetime coordinates, It is a covariant derivative.

[0093] Furthermore, after the adversarial training converges, the obtained optimized canonical connections are utilized. This study performs topological flow normalization based on the Chern-Simons theory, aiming to extract gauge-invariant physical observations related to spacetime topology. Specifically, a three-form Chern-Simons term is constructed. Its expression is as follows: in, For the outer product, For the exterior differential, The category of gauge connection is used to describe the properties of gauge fields. For traces on Lie algebra.

[0094] It is worth noting that the Chern-Simmons project... It is not inherently canonical, but it exists in a certain three-dimensional submanifold. Points on: Under canonical transformation, only one integer (the number of turns) is changed, thus its exponent is... It is a normative invariant. This represents the Chern-Simons invariants in the Chern-Simons theory.

[0095] Furthermore, the preprocessed results are obtained. These preprocessed results are not the original scalar data, but a series of norm-invariant topological invariants calculated in different spatiotemporal sub-regions. These topological invariants are not sensitive to local small disturbance noise, but can keenly capture changes in the overall topological properties of the system, such as crack initiation and slip surface formation, thereby achieving essential denoising and normalization with profound physical implications.

[0096] The improvements of this method compared to existing technologies are as follows: First, it elevates data preprocessing from shallow signal processing to a physical state inversion problem based on gauge field theory. By introducing fiber bundle theory, the sensor network is reconstructed as a gauge field on a spatiotemporal manifold, using the gauge field intensity tensor in the Yang-Mills action. The non-Abelian properties characterize the deep coupling mechanism of multiphysics, which is superior to the statistical framework of existing filtering; secondly, the adversarial learning objective is reconstructed from data distribution matching to minimizing the norm-invariant action, and the discriminator solves the problem by including a dynamically normalized denominator. The symmetry-breaking residual terms establish an unsupervised learning paradigm with physical conservation laws as the ultimate criterion, redefining noise as the breaking of gauge symmetry; thirdly, the output results are transformed from the existing scalar sequence to the Chern-Simons topological invariant sequence. These gauge invariants obtained by integration on three-dimensional submanifolds are immune to local perturbations but sensitive to system topological phase transitions, realizing a leap from numerical cleaning to essential feature extraction.

[0097] S22. Using the improved rock mass memory-temporal causal decoupling algorithm, the second type of physical field parameter data is subjected to structured cleaning and normalization processing that combines path integral and causal discovery to obtain the preprocessed results of the second type of physical field parameter data.

[0098] In one embodiment, the second type of physical field parameter data is first subjected to structured cleaning based on path integral anomaly detection, and the path integral anomaly score at each time point is calculated using the following formula: in, For time abnormal path integral scores The normalization constant is For parameter vectors, A vector of parameter mean. The mean of the standard deviation of the parameter is . The covariance matrix of the parameters, The size of the path integration window. For time, For time infinitesimal elements, Indicates transpose. Given a preset threshold. ,if If so, that time point is marked as an outlier and removed.

[0099] Furthermore, memory effect decoupling and causal discovery are performed on the cleaned data, and the causal flow between parameters is calculated using the path integral causal strength formula: in, Indicates from parameter to parameters The causal strength, For parameters In time Conditional expectation, For the memory weight function, It is a smoothing constant. For time infinitesimal elements, Indicates parameters In time For parameters Regarding time The influence of conditional expectation is approximated using the finite difference method. Indicates parameters In time The condition is expected to change slightly.

[0100] Furthermore, based on the causal intensity matrix, normalization is performed, and the data is projected onto the causal independent space using the causal whitening transformation formula: in, In time The normalized parameter vector represents the physical field parameter data after causal whitening transformation. For time The original parameter vector includes parameters such as rock mass memory deformation vector, slope damage evolution potential well depth, rock mass microseismic damage entropy, and slope structural surface dynamic stiffness deterioration index. It is the identity matrix. This is a causal strength matrix used to capture the temporal causal relationships between parameters. This transformation effectively decouples the causal relationships between parameters through inverse matrix operations, improving the stability and interpretability of the data.

[0101] Compared to existing technologies, the improvements of this algorithm are as follows: First, it introduces an anomaly detection mechanism based on path integrals in the data cleaning stage. By calculating the path integral anomaly score within the time window, it replaces existing statistical thresholds or machine learning methods, achieving anomaly identification of the overall behavior of the parameter time-series path. Second, at the causal analysis level, it proposes a path integral causal strength formula, combining functional derivatives with memory decay functions to capture the nonlinear long-term causal relationships between parameters through integral operations, breaking through the limitation of existing Granger causality or transitive entropy methods that can only analyze short-term linear relationships. Third, it constructs a rock mass memory-time-series causality decoupling mechanism in the algorithm architecture. By quantifying the decay characteristics of historical influences through a memory weight function, it achieves precise decoupling of the rock mass memory effect. Fourth, in the normalization processing stage, it designs a whitening transformation based on the causal strength matrix. By projecting the data onto a causal independent space through inverse matrix operations, it eliminates the causal coupling effect between parameters. Fifth, it can deeply mine the time-series memory characteristics and complex causal relationships in rock mass parameters, providing a more interpretable and reliable data foundation for slope stability analysis.

[0102] S3. Based on the preprocessing results, establish a dynamic assessment model for slope fatigue damage.

[0103] In one embodiment, based on the preprocessing results, using the first type of physical field parameters as state constraint variables, a quantum entanglement-enhanced fractional-order data assimilation algorithm is introduced to correct the model parameters. This algorithm combines quantum information theory and fractional-order calculus, handles the uncertainty of multi-source data through the principle of quantum state superposition, and introduces fractional-order integrals to capture the time memory effect. The specific formula is as follows: in, This is the corrected model state vector, representing the physical field state of the slope. The observation data vector comes from the parameters of the first type of physics field. The observation error covariance matrix, The background state vector, The model error covariance matrix is... For regularization parameters, The order of the fractional derivative. It is the gamma function. is Planck's constant. For quantum wave function, The number of state variables, For the index of the state variable, For the model state vector, For time, For time infinitesimal elements, For the first One state variable, For observation operators.

[0104] The advantage of this formula is that it captures the time memory effect of slope damage through fractional integrals and enhances data fusion through the concept of quantum entanglement, reducing the uncertainty of multi-source data and significantly improving the correction accuracy. It is suitable for the non-uniform geological conditions of open-pit mine slopes.

[0105] Furthermore, based on the correction results, the rock mass memory deformation vector is used... and slope damage evolution potential well depth As the core damage state variable, a fractional-order memory damage constitutive relation is constructed. This relation introduces fractional derivatives and potential well theory to describe the nonlocality and memory characteristics of damage accumulation, and adds a nonlinear coupling term to capture the fatigue effect of open-pit mine slopes. The specific formula is as follows: in, This is the threshold for the critical potential well depth. This is the memory decay coefficient, which controls the rate at which the memory effect decays. For memory kernel index, For nonlinear coupling coefficients, Let be the gradient vector of the potential well depth. It is a constant. For time, For time infinitesimal elements, The Impairment Memory Index controls for the degree of influence of historical impairment. This is the damage sensitivity coefficient, reflecting the sensitivity of the potential well to deformation. As a damage variable, This is a gamma function.

[0106] The advantage of this formula lies in capturing the time lag and historical dependence of rock mass damage through fractional terms, and quantifying the fatigue accumulation of open-pit mine slopes under cyclic loads through nonlinear coupling terms, thus breaking through the local assumptions of existing constitutive models.

[0107] Furthermore, based on the correction results and the constitutive relation, a damage state variable evolution framework is obtained, and a topological damage propagation network is introduced to quantify the damage path. This network, based on graph theory and topological invariants, defines a damage propagation matrix and energy flow to identify key vulnerable areas of open-pit mine slopes. The specific formula is as follows: in, Indicates from node To the node The damage propagation probability or matrix, For nodes To the node The Euclidean distance between them Represents a node. For scale parameters, For nodes To the node The potential well depth gradient between them For reference potential well depth, The topological diffusion coefficient is... The Hessian matrix is ​​the deformation vector, representing the local curvature change. This is the rock mass memory deformation vector. For spatial coordinate variables.

[0108] The advantage of the damage state variable evolution framework lies in its ability to dynamically identify damage clusters in open-pit mine slopes by calculating the connectivity of damage paths through topological invariants. These topological invariants refer to the variables derived from the damage propagation matrix. Features extracted from the constructed graph that remain invariant under continuous deformation, such as the number of connected components or void structures identified through persistent cohomology, are used to quantify the macroscopic connectivity of damage paths and dynamically identify stable key damage clusters and weak zones in slopes.

[0109] Furthermore, based on the aforementioned damage state variable evolution framework, the rock mass microseismic damage entropy is... and the dynamic stiffness deterioration index of slope structural surfaces As a driving variable for damage evolution, a chaotic-driven damage dynamics model is constructed. This model combines chaos theory and nonlinear dynamics to describe the sensitivity dependence of damage evolution, and introduces a stochastic resonance term to enhance the prediction of open-pit mine slope instability. The specific formula is as follows: in, For damage state variables, For deterministic dynamic functions, For driving term functions, For random resonance amplitude, The resonant frequency, The phase angle, The attenuation rate, For time, The chaotic noise term satisfies the following relationship: in, This represents the total number of frequency components. For the index of the frequency components, Indicates amplitude. Indicates frequency, Indicates phase, The attenuation rate, For time.

[0110] The advantage of this model is that it can capture the chaotic characteristics and resonance effects before slope instability, significantly improving the accuracy of dynamic prediction.

[0111] Furthermore, based on the aforementioned chaotic-driven damage dynamics model, a physics-guided spatiotemporal graph neural network is used to obtain the relationship between the nonlinear coupling and spatiotemporal propagation among physical field parameters. This network introduces a physics-constrained loss function and a graph attention mechanism, the specific formulas of which are as follows: in, For loss function, For data fitting loss, For physical constraint loss, For topological constraint loss, , To balance the parameters, the network structure combines graph convolutional layers and temporal convolutional layers, and uses attention coefficients to calculate the inter-node influence: in, For nodes and nodes Attention coefficient between them This is the weight matrix. For attention vectors, For nodes The set of neighbors, with LeakyReLU as the activation function. For nodes eigenvectors, , , All are nodes.

[0112] Furthermore, based on the aforementioned relationship, time-series data is used to dynamically identify the damage evolution path of a slope from stability to instability.

[0113] Furthermore, the failure probability is predicted using the Monte Carlo path integral method, and a fatigue life index is introduced for quantitative evaluation. The specific formula is as follows: in, For time The probability of failure within, For the number of Monte Carlo simulations, This is an indicator function; it returns 1 when the condition is met and 0 otherwise. The depth is the critical potential well. Instantaneous stress intensity For reference stress intensity, For time infinitesimal elements, For time, ( s ) indicates a node or region in time The potential well depth can be understood as a measure of the damage-driving potential, reflecting the damage-driving state of the slope at a certain moment. Fatigue life index or fatigue damage index. :

[0114] in, The fatigue damage index, It is the critical deformation vector. For time infinitesimal elements, For time, This is the rock mass memory deformation vector. It is a fatigue index used to control the rate of damage accumulation.

[0115] The advantage of the above formula lies in its ability to provide real-time risk assessment and fatigue life prediction for open-pit mine slopes through path integrals and probabilistic methods, representing a significant breakthrough.

[0116] Furthermore, a multi-physics coupled dynamic assessment model for slope fatigue damage is constructed. This dynamic assessment model for slope fatigue damage is an integrated system, including a data assimilation and correction model state vector, fractional-order memory damage constitutive relation, topological damage propagation network, chaotic-driven damage dynamics model, loss function and attention coefficient of a physics-guided spatiotemporal graph neural network, failure probability and fatigue damage index.

[0117] The advantage of the slope fatigue damage dynamic assessment model is that it realizes a paradigm shift in open-pit mine slope safety monitoring, from macroscopic appearance to microscopic mechanism, from static judgment to dynamic evolution, and from passive early warning to active intervention. Through multi-physics field data fusion and artificial intelligence deep calculation, it accurately quantifies the accumulation process and evolution path of fatigue damage inside the slope, and achieves advanced, accurate and dynamic evaluation of instability risk.

[0118] The aforementioned assessment model was applied for verification in slope monitoring of a large open-pit copper mine. First, based on preprocessed multi-source physical field data, a quantum entanglement-enhanced fractional-order data assimilation algorithm was used to correct the model's state vector. The results showed that the root mean square error between the model state and the observed data was reduced by 35% after correction. Subsequently, the rock mass memory deformation vector and potential well depth were calculated using the fractional-order memory damage constitutive relation. In this example, the potential well depth... Reduced to the critical threshold When it is 60%, the deformation vector A significant increase was observed, with nonlinear coupling terms contributing approximately 20% of the damage increment. Furthermore, key damage paths were identified through a topological damage propagation network, and their propagation probabilities were determined. The entropy reaches as high as 0.85 in the potential slip surface region, while topological invariant analysis shows an increase of 3 connected components, indicating damage accumulation. The chaotic driving model accurately captures the microseismic damage entropy. and stiffness degradation index Driven by a chaotic phase transition, the Lyapunov exponent turns positive 30 days before instability. Ultimately, the Monte Carlo path integral predicts the probability of failure within the next 90 days. The fatigue damage index reached 0.15. The value was 0.72, and the error between the actual observed slip time and the actual slip time was less than 7 days, confirming the high accuracy and foresight of the model.

[0119] It should be noted that the model generates highly abstract intermediate variables in this step, such as damage propagation matrix and chaotic state variables, but these are difficult to directly guide actions, so further processing is required.

[0120] S4. Using the aforementioned dynamic assessment model for slope fatigue damage, obtain quantitative characterization parameters of the slope damage state.

[0121] In one embodiment, based on the slope fatigue damage dynamic assessment model obtained in step S3, the preprocessed data and background state are first fused and corrected using a quantum entanglement-enhanced fractional-order data assimilation algorithm to obtain the corrected model state vector.

[0122] Furthermore, by utilizing the fractional-order memory damage constitutive relation, the state vector of the corrected model is used to perform evolution calculations to obtain the rock mass memory deformation vector and the slope damage evolution potential well depth.

[0123] Furthermore, by utilizing the topological damage propagation network and combining the potential well depth gradient and deformation curvature, the interaction between nodes is calculated to obtain the damage propagation probability matrix.

[0124] Furthermore, using a chaotic-driven damage dynamics model, the microseismic damage entropy and stiffness degradation index are used as driving inputs to obtain chaotic damage state variables.

[0125] Furthermore, by utilizing a physics-guided spatiotemporal graph neural network, we learn multi-parameter spatiotemporal coupling relationships through physical constraints and attention mechanisms, and obtain attention coefficients and loss functions.

[0126] Furthermore, the Monte Carlo path integral method is used to probabilistically integrate the potential well depth evolution and stress history to obtain the aging failure probability and fatigue damage index.

[0127] It is important to note that step S4 translates the abstract variables from step S3 into quantifiable parameters that are understandable to the engineering team.

[0128] S5. Based on the quantitative characterization parameters, obtain the evaluation results of slope mechanism early warning, path prediction and probability forecast.

[0129] In one embodiment, the chaotic phase transition of the slope is first evaluated and the evaluation results are obtained based on the evolution trajectory of the chaotic damage state variables, the rock mass microseismic damage entropy, and the dynamic stiffness deterioration index.

[0130] Specifically, based on existing chaos indices, a multidimensional phase transition identification model based on Lyapunov exponent spectra and a critical point prediction model based on quantum annealing optimization are introduced, thus constructing a chaotic phase transition enhancement model. This model utilizes multidimensional phase transition identification to calculate topological changes in the Lyapunov exponent spectrum, such as the largest Lyapunov exponent changing from negative to positive, while the second and third exponents cross over, identifying the precise moment when the system transitions from stable periodic motion to a chaotic state. Quantum annealing optimization utilizes the quantum tunneling effect to globally search for the optimal critical threshold in the parameter space, avoiding getting trapped in local optima. The mathematical expression of this model is as follows: in, For the degree of chaotic phase transition, For the first One Lyapunov index, The entropy of microseismic damage. The stiffness degradation index is... For the quantum annealing optimization function, For time, The total number of Lyapunov indices. For chaotic damage state variables.

[0131] The improvement of this method lies in the following: through multidimensional analysis of the Lyapunov exponent spectrum, it achieves accurate quantification of chaotic phase transitions, overcoming the one-sidedness of existing index methods; quantum annealing optimization provides a globally optimal critical threshold search, significantly improving the accuracy and timeliness of phase transition early warning.

[0132] Furthermore, based on the topological damage propagation probability matrix, the spatial distribution of rock mass memory deformation vector, and the potential well depth gradient, the damage propagation path and the through slip surface are evaluated, and the evaluation results are obtained.

[0133] Specifically, based on traditional network analysis, this paper introduces persistent cohomological damage path identification based on topology data analysis and curvature flow-driven dynamic reconstruction of the slip surface, thus constructing a path connectivity enhancement model. This model utilizes persistent cohomological damage path identification to automatically identify stable damage propagation paths and filter transient noise by calculating the duration of topological invariants, such as a one-dimensional Betti number persistent barcode. Curvature flow-driven reconstruction dynamically updates the geometry of the slip surface based on the diffusion equation of deformation curvature. The mathematical expression of this model is as follows: in, For path connectivity, For potential damage propagation or slippage paths, The elements in the damage propagation probability matrix represent the damage originating from the node. propagation to nodes The probability, This is the rock mass memory deformation vector. The gradient is the potential well depth. For path persistence measurement based on topology data analysis, For scale parameters, Let be a path element.

[0134] The improvement of this method lies in the following: the use of persistent cohomology analysis can provide a deeper understanding of the topological characteristics of the damage path, distinguishing between the essential structure and noise, and its performance is better than the existing path detection methods based on probability thresholds; the curvature flow-driven reconstruction introduces the dynamics of geometric evolution, making the slip surface prediction more consistent with the actual deformation mechanism of the rock mass.

[0135] Furthermore, based on the probability of failure over time and the fatigue damage index, the slope instability risk level and remaining fatigue life are evaluated, and the evaluation results are obtained.

[0136] Specifically, based on existing probability assessments, this paper constructs a slope instability risk level and remaining fatigue life evaluation model by introducing real-time risk updates based on dynamic Bayesian networks and fatigue damage path clustering for lifetime prediction. This model, based on dynamic Bayesian networks, establishes a dynamic correlation between failure probability and multi-source monitoring data by introducing time slices, enabling real-time updates of the risk level. Fatigue damage path clustering uses a spectral clustering algorithm to group Monte Carlo simulation paths, identifying typical damage evolution patterns and providing statistical basis for remaining life prediction. The mathematical expression of this model is as follows: in, Indicates the risk level. The risk level mapping function is a composite function that comprehensively maps multiple indicators, such as failure probability, fatigue damage index and its changing trend, to a specific risk level, such as low, medium, or high. For the future The probability of failure within the specified time limit. The fatigue damage index, This is a dynamic Bayesian network real-time correction item. Based on the latest monitoring data, it uses dynamic Bayesian network inference to update and correct the risk level in real time, ensuring the timeliness of the evaluation results. For the first Fatigue damage path clustering, For expiration time, For conditional mathematical expectation, This represents the damage evolution path.

[0137] The improvements of this method are as follows: the dynamic Bayesian network enables real-time adaptive updates of risk assessment, overcoming the lag of existing probability models; fatigue damage path clustering provides data-driven lifetime prediction, which can identify multiple possible failure modes and significantly improve the comprehensiveness and reliability of remaining lifetime assessment.

[0138] S6. Based on the evaluation results, a dynamic evaluation of fatigue instability of open-pit mine slopes is achieved.

[0139] In one embodiment, based on the evaluation results of the mechanism-based early warning, the conclusion that the slope stability has entered a sensitive decay period is first obtained; through the conclusion, decision instructions for high-frequency monitoring and induced regulation are obtained; wherein, the induced regulation introduces a fatigue antagonism mechanism based on active intervention of stress waves, by applying periodic micro-perturbations opposite to the phase of fatigue damage evolution to a specific area of ​​the slope rock mass, in order to actively offset part of the cumulative damage, thereby changing the inherent fatigue response characteristics of the rock mass material, and upgrading the traditional passive monitoring and early warning into an integrated process of active intervention and state regulation.

[0140] Furthermore, based on the evaluation results of the path prediction, the spatial location of the damage convergence node and key transmission path inside the slope is obtained; through the spatial location, a deployment scheme for precise monitoring and reinforcement of the target area is obtained; wherein, the deployment scheme integrates a path barrier system constructed with biomimetic self-healing materials. This system can intelligently inject composite materials with fluidity and self-coagulation properties according to the predicted path, forming a tough barrier zone that deforms in tandem with the rock mass inside the damage path. This not only blocks the transmission of damage, but also absorbs and disperses stress wave energy through the energy dissipation mechanism of the material itself, realizing the transformation from passive reinforcement to active guidance and healing.

[0141] Furthermore, based on the evaluation results of probabilistic forecasts, the quantitative risk level and remaining service life of the slope are obtained as quantitative expectations. Through these quantitative expectations, a strategy plan for graded early warning response and predictive maintenance is obtained. The strategy plan introduces a life extension decision-making algorithm based on digital twins and deep reinforcement learning. This algorithm simulates the impact of different maintenance strategies on the remaining service life of the slope in a virtual space and uses reinforcement learning to autonomously optimize and dynamically generate predictive maintenance sequences and resource allocation schemes that maximize the extension of the slope's service life. This achieves a leap from response based on the current state to autonomous decision-making oriented towards the optimal goal of the entire life cycle.

[0142] Furthermore, based on the decision-making instructions, the deployment plan, and the strategy planning, a dynamic evaluation of fatigue instability of open-pit mine slopes is achieved.

[0143] It is important to note that this method transforms quantitative parameters into three evaluations that can directly support decision-making: mechanism warning (why the risk exists), path prediction (where the risk lies), and probability forecast (how severe the risk is). Without steps S4 and S5, the model results in S3 are difficult to use for practical safety management.

[0144] The method of this invention was applied to conduct a dynamic evaluation over 18 months in the monitoring of a slope in a large open-pit copper mine. During implementation, a multi-physics sensor network was deployed, successfully capturing an abnormal increase of over 15% in the rate of change of resistivity in the potential slip surface area, providing an early warning 42 days earlier than existing displacement monitoring methods. The coefficient of variation of acoustic velocity tomography increased from an initial 0.08 to 0.23, quantitatively revealing the intensified heterogeneity caused by the accumulation of damage within the rock mass. After preprocessing based on an improved rock mass memory-temporal causal decoupling algorithm, the slope damage evolution potential well depth in the second type of physical field parameters decreased from a baseline value of 2.1 kJ to around a critical threshold of 0.8 kJ, accurately characterizing the system stability decay. Through the constructed chaotic-driven damage dynamics model, the chaotic phase transition degree exceeded the early warning threshold of 0.75 63 days before significant slope deformation, achieving advanced early warning. The key damage paths identified using the topological damage propagation network showed a spatial agreement of 97% with the actual subsequent microseismic event clusters. Ultimately, based on the Monte Carlo path integral prediction, the slope's failure probability within 120 days was 18%, guiding targeted reinforcement measures to be taken on-site, successfully avoiding a potential landslide, and verifying the significant breakthrough of this method in terms of early warning timeliness, quantitative accuracy, and engineering applicability.

[0145] Please see Figure 2 , Figure 2 This is a schematic diagram of a dynamic evaluation system for fatigue instability of open-pit mine slopes based on a quantitative model, according to an embodiment of the present invention. The system includes an input device, a processor, an output device, and a memory, all interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to call the program instructions. The system uses the aforementioned dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model. The input device includes an acquisition module, the processor includes a construction module and an analysis module, and the output device includes an output module. The acquisition module is used to acquire physical field parameter data of open-pit mine slopes; and to preprocess the physical field parameter data to obtain preprocessing results. A construction module is used to establish a dynamic evaluation model for fatigue damage based on the preprocessing results; and to utilize the dynamic evaluation model for fatigue damage in combination with a quantum entanglement-enhanced data assimilation algorithm, a fractional-order memory constitutive algorithm, and a chaotic driving dynamics algorithm. The analysis module is used to obtain quantitative characterization parameters of slope damage status; based on the quantitative characterization parameters, a chaotic phase transition enhancement model, a path connectivity enhancement model, and a health status analysis model are constructed; based on the chaotic phase transition enhancement model, the path connectivity enhancement model, and the health status analysis model, the analysis results of mechanism early warning, path prediction, and probability forecasting are obtained. The output module is used to dynamically evaluate the fatigue instability of open-pit mine slopes based on the analysis results.

[0146] In summary, this invention, by introducing gauge field theory and fiber bundle theory, constructs a physical law-driven evaluation system, outputting Chern-Simons invariants that are immune to disturbances and sensitive to topological phase transitions, achieving a theoretical transformation from experience-based dependence to essential feature extraction. By creating a comprehensive sensing system covering multiple physical field parameters, it can detect ground resistivity anomalies 42 days in advance and provide early warning of instability risks 63 days in advance, achieving a leapfrog improvement in monitoring capabilities. By establishing an evaluation model that integrates fractional-order memory and chaos-driven approaches, it overcomes the limitations of local linearity and significantly improves the ability to capture pre-instability critical features. By constructing a three-dimensional evaluation system of mechanism-path-probability, combined with dynamic Bayesian networks and active intervention mechanisms, it achieves a major shift from passive monitoring to intelligent decision-making throughout the entire lifecycle. Engineering practice shows that the method achieves a 97% spatial agreement between damage path identification and microseismic events, successfully guiding reinforcement and preventing landslides, fully verifying its excellent efficiency and significant application value in full-chain dynamic evaluation.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. 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 or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model, characterized in that, The method includes the following steps: Obtain physical field parameter data of open-pit mine slopes; The physical field parameter data is preprocessed to obtain the preprocessing result; Based on the preprocessing results, a dynamic assessment model for fatigue damage is established. Using the aforementioned fatigue damage dynamic assessment model, combined with the quantum entanglement-enhanced data assimilation algorithm, fractional-order memory constitutive algorithm, and chaotic driving dynamics algorithm, quantitative characterization parameters of slope damage state are obtained; Based on the aforementioned quantitative characterization parameters, a chaotic phase transition enhancement model, a path connectivity enhancement model, and a health status analysis model are constructed. Based on the chaotic phase transition enhancement model, the path connectivity enhancement model, and the health status analysis model, the analysis results of mechanism early warning, path prediction, and probability forecasting are obtained; The analysis results enable a dynamic evaluation of fatigue instability of open-pit mine slopes.

2. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 1, characterized in that, Obtaining physical field parameter data for open-pit mine slopes includes: The first type of physical field parameter data of the open-pit mine slope is obtained. The first type of physical field parameter data includes the rate of change of resistivity, the intensity of geomagnetic perturbation, the coefficient of variation of acoustic tomography, the chemical ion concentration gradient of fracture water, the radon gas release rate, the change of fractal dimension of surface texture, the area of ​​acceleration response spectrum and the entropy value of thermal infrared radiation. The second type of physical field parameter data of the open-pit mine slope is obtained. The second type of physical field parameter data includes data on rock mass memory deformation vector, slope damage evolution potential well depth, rock mass microseismic damage entropy, and slope structural surface dynamic stiffness deterioration index.

3. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 1, characterized in that, The physical field parameter data is preprocessed to obtain the following preprocessing results: Using gauge field theory and spacetime symmetry breaking adversarial learning algorithm, the first type of physical field parameter data is denoised and normalized to obtain the preprocessed results of the first type of physical field parameter data; An improved rock mass memory-temporal causal decoupling algorithm was used to perform structured cleaning, memory effect decoupling, causal discovery and normalization on the second type of physical field parameter data, resulting in the preprocessed results of the second type of physical field parameter data.

4. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 1, characterized in that, Based on the preprocessing results, a dynamic fatigue damage assessment model is established, including: Based on the preprocessing results, using the first type of physical field parameters as state constraint variables, the correction results of the model parameters are obtained using data assimilation techniques. Based on the correction results, the constitutive relationship of rock mass damage accumulation evolution is constructed using the rock mass memory deformation vector and the slope damage evolution potential well depth as the core damage state variables. Based on the correction results and the constitutive relation, an evolutionary framework for the damage state variables is constructed. Based on the aforementioned evolution framework, the rock mass microseismic damage entropy and the slope structural surface dynamic stiffness deterioration index are used as driving variables for damage evolution, and a dynamic model of damage evolution is constructed. Based on the aforementioned dynamic model, a physics-guided spatiotemporal graph neural network is used to obtain the correlation between nonlinear coupling and spatiotemporal propagation among physical field parameters; Based on the aforementioned correlation, dynamic identification results of the damage evolution path of the slope from stability to instability are obtained; Based on the correction results, the constitutive relation, the evolutionary framework, the dynamic model, the correlation, and the dynamic identification results, a fatigue damage dynamic assessment model is constructed.

5. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 1, characterized in that, Using the aforementioned fatigue damage dynamic assessment model, combined with the quantum entanglement-enhanced data assimilation algorithm, fractional-order memory constitutive algorithm, and chaotic driving dynamics algorithm, the quantitative characterization parameters of the slope damage state are obtained, including: Based on the aforementioned fatigue damage dynamic assessment model, a quantum entanglement-enhanced fractional-order data assimilation algorithm is used to fuse and correct the preprocessed data with the background state to obtain the corrected model state vector. Based on the model state vector, the evolution calculation is performed using the fractional-order memory damage constitutive relation to obtain the updated results of the rock mass memory deformation vector and the slope damage evolution potential well depth; Based on the updated results, the interaction between nodes is calculated using the topological damage propagation network, combined with the potential well depth gradient and deformation curvature, to obtain the damage propagation probability matrix. Based on the damage propagation probability matrix, chaotic damage state variables are obtained through a chaotic-driven damage dynamics model. Based on the chaotic damage state variables, the loss function and attention coefficient are obtained according to the physical-guided spatiotemporal graph neural network. Based on the loss function and the attention coefficient, the Monte Carlo path integral method is used to perform probabilistic integration on the potential well depth evolution and stress history to obtain the aging failure probability and fatigue damage index.

6. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 5, characterized in that, The formula for the quantum entanglement-enhanced fractional-order data assimilation algorithm is as follows: in, This is the corrected model state vector. For the observed data vector, The observation error covariance matrix, The background state vector, The model error covariance matrix is... For regularization parameters, The order of the fractional derivative. It is the gamma function. is Planck's constant. For quantum wave function, The number of state variables, For the index of the state variable, For the model state vector, For time, For time infinitesimal elements, For the first One state variable, For the observation operator; the formula for the fractional-order memory impairment constitutive relation is as follows: in, This is the rock mass memory deformation vector. The depth of the potential well for slope damage evolution. This is the threshold for the critical potential well depth. The memory decay coefficient, For memory kernel index, For nonlinear coupling coefficients, Let be the gradient vector of the potential well depth. It is a constant. For time, For time infinitesimal elements, The Impaired Memory Index This is the damage sensitivity coefficient. As a damage variable, The gamma function is used; the chaotic-driven damage dynamics model satisfies the following expression: in, For damage state variables, For deterministic dynamic functions, For driving term functions, For random resonance amplitude, The resonant frequency, The phase angle, The attenuation rate, For time, This is a chaotic noise term. This represents the entropy of microseismic damage in the rock mass. It is the dynamic stiffness deterioration index of the slope structure surface.

7. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 1, characterized in that, Based on the aforementioned chaotic phase transition enhancement model, path connectivity enhancement model, and health status analysis model, the analytical results obtained for mechanism early warning, path prediction, and probability forecasting include: The analysis results of chaotic phase transitions are obtained using the aforementioned chaotic phase transition enhancement model. Using the aforementioned path penetration enhancement model, the analysis results of the damage propagation path and the penetration slip surface are obtained; Using the aforementioned health status analysis model, the analysis results of instability risk level and remaining fatigue life are obtained.

8. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 7, characterized in that, The chaotic phase transition enhancement model satisfies the following expression: in, For the degree of chaotic phase transition, For the first One Lyapunov index, The entropy of microseismic damage. The stiffness degradation index is... For the quantum annealing optimization function, For time, The total number of Lyapunov indices. The path connectivity enhancement model satisfies the following expression: (The path connectivity enhancement model is a chaotic damage state variable.) in, For path connectivity, For the sliding path, Indicates damage from node propagation to nodes The probability, This is the rock mass memory deformation vector. The gradient is the potential well depth. For path persistence measurement based on topology data analysis, For scale parameters, The path element is defined; the health status assessment model satisfies the following expression: in, Indicates the risk level. For risk level mapping function, For the future The probability of failure within the specified time limit. The fatigue damage index, For real-time correction terms in dynamic Bayesian networks, For the first Fatigue damage path clustering, For expiration time, For conditional mathematical expectation, This represents the damage evolution path.

9. The dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model according to claim 1, characterized in that, Based on the analysis results, the dynamic evaluation of fatigue instability of open-pit mine slopes includes: Based on the analysis results of the mechanism-based early warning, it is concluded that the slope stability has entered a sensitive decay period; based on the conclusion, decision-making instructions for high-frequency monitoring and induced factor control are obtained. Based on the analysis results of path prediction, the spatial location of the damage convergence node and key transmission path inside the slope is obtained; through the spatial location, a deployment plan for precise monitoring and reinforcement of the target area is obtained. Based on the analysis results of probability forecasts, quantitative risk levels and remaining service life of slopes are obtained; through these quantitative expectations, strategic planning for graded early warning response and predictive maintenance is obtained. Based on the decision-making instructions, the deployment plan, and the strategy planning, a complete dynamic evaluation of fatigue instability of open-pit mine slopes is achieved.

10. A dynamic evaluation system for fatigue instability of open-pit mine slopes based on a quantitative model, wherein the system uses the dynamic evaluation method for fatigue instability of open-pit mine slopes based on a quantitative model as described in any one of claims 1 to 9, characterized in that... The system includes an acquisition module, a construction module, an analysis module, and an output module; The acquisition module is used to acquire physical field parameter data of open-pit mine slopes; The physical field parameter data is preprocessed to obtain the preprocessing result; A construction module is used to establish a dynamic evaluation model for fatigue damage based on the preprocessing results; and to utilize the dynamic evaluation model for fatigue damage in combination with a quantum entanglement-enhanced data assimilation algorithm, a fractional-order memory constitutive algorithm, and a chaotic driving dynamics algorithm. The analysis module is used to obtain quantitative characterization parameters of slope damage status; based on the quantitative characterization parameters, a chaotic phase transition enhancement model, a path connectivity enhancement model, and a health status analysis model are constructed; based on the chaotic phase transition enhancement model, the path connectivity enhancement model, and the health status analysis model, the analysis results of mechanism early warning, path prediction, and probability forecasting are obtained. The output module is used to dynamically evaluate the fatigue instability of open-pit mine slopes based on the analysis results.

Images