A Real-Time Monitoring Platform and Method for Geological Disasters Based on Intelligent AI

Abstract

This application relates to the field of geological disaster monitoring technology, and in particular to a real-time geological disaster monitoring platform and method based on intelligent AI. The method includes: acquiring multi-source monitoring data; analyzing the multi-source monitoring data to determine soil stability; determining soil characteristic parameters based on soil stability; acquiring environmental hydrological data and determining the dynamic response mode of the geological structure based on the environmental hydrological data; and determining the geological disaster risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure. Determining soil characteristic parameters provides building-specific quantitative indicators to support customized risk assessment. Acquiring environmental hydrological data and determining the dynamic response mode of the geological structure based on the environmental hydrological data captures the real-time behavioral characteristics of the geological structure under environmental changes and identifies vulnerable areas of the geological structure. Determining the geological disaster risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure enables building-level early warning and quantifies the risk evolution rate.

Description

Technical Field

[0001] This application relates to the field of geological disaster monitoring technology, and in particular to a real-time geological disaster monitoring platform and method based on intelligent AI. Background Technology

[0002] With the acceleration of urbanization, the excavation of underground infrastructure has become increasingly frequent, leading to a significant increase in the incidence of urban geological disasters. These geological disasters not only cause ground deformation but also lead to foundation subsidence, wall cracking, foundation instability, and even collapse of buildings.

[0003] In existing technologies, geological disaster monitoring mainly relies on a single data source, resulting in low monitoring accuracy and delayed early warning, which cannot meet the urgent needs of modern cities for "early identification, early warning, and early response" of geological disasters. Summary of the Invention

[0004] This application provides a real-time monitoring platform and method for geological disasters based on intelligent AI to solve the above-mentioned problems.

[0005] In a first aspect, this application provides a method for real-time monitoring of geological disasters based on intelligent AI, the method comprising:

[0006] Acquire multi-source monitoring data; analyze the multi-source monitoring data to determine soil stability;

[0007] Based on the soil stability, determine the soil characteristic parameters;

[0008] Acquire environmental hydrological data, and determine the dynamic response mode of the geological structure based on the environmental hydrological data;

[0009] The geological hazard risk level is determined based on the soil characteristic parameters and the dynamic response mode of the geological structure.

[0010] This solution acquires multi-source monitoring data, avoiding the limitations of a single data source and supporting a comprehensive assessment of soil stability. Analyzing multi-source monitoring data determines soil stability and quantifies the integrity of soil stress transmission paths, thereby improving monitoring accuracy. Based on soil stability, characteristic soil parameters are determined, providing building-specific quantitative indicators to support customized risk assessments. Environmental hydrological data is acquired, and based on this data, dynamic response patterns of geological structures are determined, capturing real-time behavioral characteristics of geological structures under environmental changes and identifying vulnerable areas. Based on soil characteristic parameters and dynamic response patterns of geological structures, geological hazard risk levels are determined, enabling building-level early warning and quantifying the rate of risk evolution.

[0011] Optionally, the multi-source monitoring data includes soil stress data, and the determination of soil characteristic parameters based on the soil stability includes:

[0012] By analyzing the multi-source monitoring data, the deformation data of the underground structure can be obtained;

[0013] Analyze the deformation data of the underground structure to determine the spatiotemporal distribution characteristics of the underground structure;

[0014] Based on the aforementioned spatiotemporal distribution characteristics, the soil stress data is analyzed to determine the soil stress transmission path;

[0015] Based on the stress transmission path of the soil, determine the characteristic parameters of the soil.

[0016] Optionally, the multi-source monitoring data also includes soil and rock layer distribution parameters. The step of analyzing the soil stress data and determining the soil stress transmission path based on the spatiotemporal distribution characteristics includes:

[0017] Based on the soil and rock layer distribution parameters and the soil stress data, three-dimensional geological data are constructed.

[0018] Based on the three-dimensional geological data, a stress transfer attenuation function is constructed.

[0019] Based on the spatiotemporal distribution characteristics and the stress transmission attenuation function, a soil stress transmission path is constructed.

[0020] Optionally, determining the dynamic response mode of the geological structure based on the environmental hydrological data includes:

[0021] Based on the aforementioned environmental hydrological data, determine the rate of change of seepage pressure;

[0022] Based on the soil stress transmission path and the seepage pressure change rate, the weak areas of the geological structure are identified;

[0023] Based on the rate of change of osmotic pressure and the weak areas of the geological structure, the dynamic response mode of the geological structure is determined.

[0024] Optionally, determining the geological hazard risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure includes:

[0025] Acquire historical disaster cases, analyze these cases, and determine disaster process data;

[0026] A spatiotemporal coupled prediction model is constructed using the disaster process data.

[0027] The soil characteristic parameters are input into the spatiotemporal coupling prediction model, and the displacement gradient change rate and energy accumulation coefficient output by the spatiotemporal coupling prediction model are obtained.

[0028] The risk evolution rate of each building to be monitored is calculated based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate.

[0029] The geological hazard risk level is determined based on the dynamic response mode of the geological structure and the risk evolution rate.

[0030] Optionally, constructing a spatiotemporally coupled prediction model using the disaster process data includes:

[0031] The disaster process data is classified to determine the disaster precursor parameters and disaster development data;

[0032] Based on time series analysis, the disaster precursor parameters and disaster development data are used to determine the correlation characteristics between the disaster precursor parameters and the disaster development time series.

[0033] Based on the aforementioned correlation features, a long short-term memory network is used to construct a time-related sub-model according to the disaster precursor parameters and the disaster development data, thereby obtaining a spatiotemporal coupled prediction model.

[0034] Optionally, calculating the risk evolution rate of each monitored building based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate includes:

[0035] Analyze the aforementioned environmental hydrological data to determine the historical rainfall intensity for the current season;

[0036] The rainfall infiltration coupling coefficient is determined based on the energy accumulation coefficient, the geologically weak area, and the historical rainfall intensity.

[0037] Obtain basic information for each building to be monitored;

[0038] The basic information is analyzed to determine the burial depth characteristics of each building to be monitored;

[0039] Based on the burial depth characteristics and the displacement gradient change rate, a burial depth influence weight matrix is ​​generated;

[0040] A three-dimensional risk evolution model is established based on the rainfall infiltration coupling coefficient and the burial depth influence weight matrix.

[0041] Based on the three-dimensional risk evolution model, the risk evolution rate of each building to be monitored is output.

[0042] Optionally, the determination of the building to be monitored includes:

[0043] Analyze the weak areas of the geological structure to determine the extent of the structural weakness;

[0044] Obtain urban structure information, parse the urban structure information, and determine urban buildings and the coordinate information of each urban building;

[0045] The weak structural area and the coordinate information are coupled together, and the building to be monitored is determined based on the coupling result.

[0046] Optionally, the environmental hydrological data includes groundwater level monitoring data, and determining the rate of change of seepage pressure based on the environmental hydrological data includes:

[0047] Analyze the groundwater level monitoring data to determine the gradient of soil pore water pressure change;

[0048] Based on the soil and rock layer distribution parameters, determine the permeability coefficient and thickness of each soil layer;

[0049] Based on Darcy's law, establish the seepage velocity equation for each soil layer;

[0050] Based on the seepage velocity equation and the groundwater level monitoring data, the stratified seepage pressure attenuation rate is calculated.

[0051] The osmotic pressure transfer function is generated by weighting and superimposing the osmotic pressure decay rate of each layer.

[0052] The rate of change of osmotic pressure is generated based on the osmotic pressure transfer function and the gradient of pore water pressure change in the soil.

[0053] Secondly, this application provides a real-time geological disaster monitoring platform based on intelligent AI, the platform comprising:

[0054] The data analysis module is used to acquire multi-source monitoring data; analyze the multi-source monitoring data, and determine soil stability;

[0055] The parameter determination module is used to determine the characteristic parameters of the soil based on the soil stability.

[0056] The mode determination module is used to acquire environmental hydrological data and determine the dynamic response mode of the geological structure based on the environmental hydrological data.

[0057] The level determination module is used to determine the geological hazard risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure.

[0058] Optionally, the multi-source monitoring data includes soil stress data, and when the parameter determination module determines soil characteristic parameters based on the soil stability, it is used for:

[0059] By analyzing the multi-source monitoring data, the deformation data of the underground structure can be obtained;

[0060] Analyze the deformation data of the underground structure to determine the spatiotemporal distribution characteristics of the underground structure;

[0061] Based on the aforementioned spatiotemporal distribution characteristics, the soil stress data is analyzed to determine the soil stress transmission path;

[0062] Based on the stress transmission path of the soil, determine the characteristic parameters of the soil.

[0063] Optionally, the multi-source monitoring data also includes soil and rock layer distribution parameters. When the parameter determination module analyzes the soil stress data based on the spatiotemporal distribution characteristics to determine the soil stress transmission path, it is used for:

[0064] Based on the soil and rock layer distribution parameters and the soil stress data, three-dimensional geological data are constructed.

[0065] Based on the three-dimensional geological data, a stress transfer attenuation function is constructed.

[0066] Based on the spatiotemporal distribution characteristics and the stress transmission attenuation function, a soil stress transmission path is constructed.

[0067] Optionally, when the mode determination module determines the dynamic response mode of the geological structure based on the environmental hydrological data, it is used for:

[0068] Based on the aforementioned environmental hydrological data, determine the rate of change of seepage pressure;

[0069] Based on the soil stress transmission path and the seepage pressure change rate, the weak areas of the geological structure are identified;

[0070] Based on the rate of change of osmotic pressure and the weak areas of the geological structure, the dynamic response mode of the geological structure is determined.

[0071] Optionally, when determining the geological hazard risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure, the level determination module is used for:

[0072] Acquire historical disaster cases, analyze these cases, and determine disaster process data;

[0073] A spatiotemporal coupled prediction model is constructed using the disaster process data.

[0074] The soil characteristic parameters are input into the spatiotemporal coupling prediction model, and the displacement gradient change rate and energy accumulation coefficient output by the spatiotemporal coupling prediction model are obtained.

[0075] The risk evolution rate of each building to be monitored is calculated based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate.

[0076] The geological hazard risk level is determined based on the dynamic response mode of the geological structure and the risk evolution rate.

[0077] Optionally, when the level determination module constructs a spatiotemporally coupled prediction model using the disaster process data, it is used for:

[0078] The disaster process data is classified to determine the disaster precursor parameters and disaster development data;

[0079] Based on time series analysis, the disaster precursor parameters and disaster development data are used to determine the correlation characteristics between the disaster precursor parameters and the disaster development time series.

[0080] Based on the aforementioned correlation features, a long short-term memory network is used to construct a time-related sub-model according to the disaster precursor parameters and the disaster development data, thereby obtaining a spatiotemporal coupled prediction model.

[0081] Optionally, when the level determination module calculates the risk evolution rate of each building to be monitored based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate, it is used for:

[0082] Analyze the aforementioned environmental hydrological data to determine the historical rainfall intensity for the current season;

[0083] The rainfall infiltration coupling coefficient is determined based on the energy accumulation coefficient, the geologically weak area, and the historical rainfall intensity.

[0084] Obtain basic information for each building to be monitored;

[0085] The basic information is analyzed to determine the burial depth characteristics of each building to be monitored;

[0086] Based on the burial depth characteristics and the displacement gradient change rate, a burial depth influence weight matrix is ​​generated;

[0087] A three-dimensional risk evolution model is established based on the rainfall infiltration coupling coefficient and the burial depth influence weight matrix.

[0088] Based on the three-dimensional risk evolution model, the risk evolution rate of each building to be monitored is output.

[0089] Optionally, the AI-based real-time geological disaster monitoring platform further includes a building identification module, used for:

[0090] Analyze the weak areas of the geological structure to determine the extent of the structural weakness;

[0091] Obtain urban structure information, parse the urban structure information, and determine urban buildings and the coordinate information of each urban building;

[0092] The weak structural area and the coordinate information are coupled together, and the building to be monitored is determined based on the coupling result.

[0093] Optionally, the environmental hydrological data includes groundwater level monitoring data, and when the pattern determination module determines the rate of change of seepage pressure based on the environmental hydrological data, it is used for:

[0094] Analyze the groundwater level monitoring data to determine the gradient of soil pore water pressure change;

[0095] Based on the soil and rock layer distribution parameters, determine the permeability coefficient and thickness of each soil layer;

[0096] Based on Darcy's law, establish the seepage velocity equation for each soil layer;

[0097] Based on the seepage velocity equation and the groundwater level monitoring data, the stratified seepage pressure attenuation rate is calculated.

[0098] The osmotic pressure transfer function is generated by weighting and superimposing the osmotic pressure decay rate of each layer.

[0099] The rate of change of osmotic pressure is generated based on the osmotic pressure transfer function and the gradient of pore water pressure change in the soil. Attached Figure Description

[0100] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0101] Figure 1 This is a schematic diagram of an application scenario provided in an embodiment of this application;

[0102] Figure 2 A flowchart illustrating a real-time geological disaster monitoring method based on intelligent AI, provided as an embodiment of this application;

[0103] Figure 3 This is a schematic diagram of a real-time geological disaster monitoring platform based on intelligent AI, provided as an embodiment of this application. Detailed Implementation

[0104] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0105] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0106] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0107] In existing technologies, geological disaster monitoring mainly relies on a single data source, resulting in low monitoring accuracy and delayed early warning, which cannot meet the urgent needs of modern cities for "early identification, early warning, and early response" of geological disasters.

[0108] Based on this, this application provides a real-time geological disaster monitoring platform and method based on intelligent AI. It acquires multi-source monitoring data, avoiding the limitations of a single data source and thus supporting a comprehensive assessment of soil stability. By analyzing multi-source monitoring data, soil stability is determined, and the integrity of soil stress transmission paths is quantified, thereby improving monitoring accuracy. Based on soil stability, soil characteristic parameters are determined, providing building-specific quantitative indicators to support customized risk assessment. Environmental hydrological data is acquired, and based on this data, the dynamic response pattern of the geological structure is determined, capturing the real-time behavioral characteristics of the geological structure under environmental changes and identifying vulnerable areas. Based on soil characteristic parameters and the dynamic response pattern of the geological structure, the geological disaster risk level is determined, enabling building-level early warning and quantifying the risk evolution rate.

[0109] Figure 1 This is a schematic diagram of an application scenario provided by this application, illustrating the application of the method provided in this application during real-time monitoring of geological disasters.

[0110] Specifically, the method provided in this application can be applied to any server, where the server interacts with IoT sensors to acquire multi-source monitoring data. The multi-source monitoring data is analyzed to determine soil stability. Based on soil stability, soil characteristic parameters are determined, providing building-specific quantitative indicators to support customized risk assessment. Environmental hydrological data is acquired, and based on this data, the dynamic response pattern of the geological structure is determined, capturing the real-time behavioral characteristics of the geological structure under environmental changes and identifying vulnerable areas. Based on the soil characteristic parameters and the dynamic response pattern of the geological structure, the geological hazard risk level is determined, achieving building-level early warning and quantifying the risk evolution rate.

[0111] For specific implementation details, please refer to the following examples.

[0112] Figure 2 This is a flowchart illustrating a real-time geological disaster monitoring method based on intelligent AI, provided as an embodiment of this application. The method of this embodiment can be applied to the server in the above scenario. Figure 2 As shown, the method includes:

[0113] S201. Obtain multi-source monitoring data; analyze the multi-source monitoring data to determine soil stability;

[0114] Multi-source monitoring data can be a heterogeneous collection of data from multiple sensors.

[0115] Soil stability can be defined as the soil's resistance to deformation under external loads.

[0116] Specifically, multi-source monitoring data, including soil stress, displacement, and hydrological data, are collected in real time using IoT sensors. Then, the multi-source monitoring data is analyzed, and soil stress and displacement data are correlated to calculate a stress-displacement coupling index. Simultaneously, hydrological data is introduced to assess the impact of groundwater seepage on pore pressure. Subsequently, a spatiotemporal coupling prediction model is trained based on historical disaster case data, and soil stability is output through a machine learning model.

[0117] S202. Based on soil stability, determine soil characteristic parameters;

[0118] Soil characteristic parameters can be quantitative indicators derived from soil stability analysis.

[0119] Specifically, based on soil stability, the displacement gradient change rate is determined through time series analysis of displacement data; and the stress data is numerically integrated to calculate the cumulative value of stress in the time domain, i.e., the energy accumulation coefficient; thereby determining the soil characteristic parameters that include the displacement gradient change rate and the energy accumulation coefficient.

[0120] S203. Obtain environmental hydrological data and determine the dynamic response mode of geological structure based on the environmental hydrological data;

[0121] Environmental hydrological data can be real-time monitoring data related to the hydrological environment.

[0122] The dynamic response model of geological structure can be the real-time behavioral characteristics of geological structure under environmental changes.

[0123] Specifically, environmental hydrological data is acquired in real time through hydrological monitoring equipment; then, the rate of change of seepage pressure is calculated based on the environmental hydrological data; through the rate of change of seepage pressure, weak areas of geological structure are identified, and the dynamic response mode of geological structure is determined.

[0124] S204. Determine the geological hazard risk level based on soil characteristic parameters and dynamic response mode of geological structure.

[0125] Geological disaster risk level can be defined as the probability of disaster occurrence and the severity level.

[0126] Specifically, soil characteristic parameters and dynamic response patterns of geological structures are input into the risk assessment model; then, the soil characteristic parameters and dynamic response patterns of geological structures are correlated to quantify the dynamic changes in risk, thereby determining the geological hazard risk level.

[0127] This solution acquires multi-source monitoring data, avoiding the limitations of a single data source and supporting a comprehensive assessment of soil stability. Analyzing multi-source monitoring data determines soil stability and quantifies the integrity of soil stress transmission paths, thereby improving monitoring accuracy. Based on soil stability, characteristic soil parameters are determined, providing building-specific quantitative indicators to support customized risk assessments. Environmental hydrological data is acquired, and based on this data, dynamic response patterns of geological structures are determined, capturing real-time behavioral characteristics of geological structures under environmental changes and identifying vulnerable areas. Based on soil characteristic parameters and dynamic response patterns of geological structures, geological hazard risk levels are determined, enabling building-level early warning and quantifying the rate of risk evolution.

[0128] In some embodiments, multi-source monitoring data is parsed to obtain underground structure deformation data; the underground structure deformation data is analyzed to determine the spatiotemporal distribution characteristics of the underground structure; based on the spatiotemporal distribution characteristics, soil stress data is analyzed to determine the soil stress transmission path; and based on the soil stress transmission path, soil characteristic parameters are determined.

[0129] Subsurface structural deformation data can be time-series data of subsurface structural deformation information.

[0130] Underground structures can be underground infrastructure or geological formations.

[0131] Spatiotemporal distribution characteristics can be the distribution patterns of data in time and space dimensions.

[0132] Soil stress data can be the quantitative data of the stress experienced by the soil.

[0133] The stress transmission path in soil can be the trajectory of stress propagation in the soil.

[0134] Specifically, multi-source monitoring data is analyzed using noise filtering and data alignment algorithms to separate and output underground structural deformation data. Then, time series analysis methods are applied to process the deformation data, calculating the deformation trend over time. Next, spatial interpolation algorithms are used to analyze the spatial distribution of the deformation data, generating spatiotemporal distribution characteristics. Based on these characteristics, a path tracing algorithm is used to analyze soil stress data and construct soil stress transmission paths. Finally, based on these stress transmission paths, the displacement gradient rate of change is determined using the displacement gradient rate of change; simultaneously, the energy accumulation coefficient is determined using the energy accumulation coefficient; and soil characteristic parameters incorporating both the displacement gradient rate of change and the energy accumulation coefficient are identified.

[0135] This scheme analyzes multi-source monitoring data to obtain underground structural deformation data, enabling the quantitative extraction of underground structural deformation information and enhancing the comprehensive assessment of soil stability. Analyzing the underground structural deformation data determines the spatiotemporal distribution characteristics of the underground structure, revealing the temporal and spatial variation patterns of underground structural deformation, aiding in the identification of weak areas in the geological structure, and providing support for constructing soil stress transmission paths. Based on the spatiotemporal distribution characteristics, soil stress data is analyzed to determine soil stress transmission paths, quantifying the dynamic trajectory of stress propagation, compensating for deficiencies in soil stress data processing, thereby improving the accuracy of dynamic monitoring of soil stability. Based on the soil stress transmission paths, soil characteristic parameters are determined, eliminating the problem of inaccurate risk assessment.

[0136] In some embodiments, three-dimensional geological data is constructed based on soil and rock layer distribution parameters and soil stress data; a stress transmission attenuation function is constructed based on the three-dimensional geological data; and a soil stress transmission path is constructed based on the spatiotemporal distribution characteristics and the stress transmission attenuation function.

[0137] The distribution parameters of soil and rock layers can be parameters that describe the characteristics of soil and rock layers.

[0138] Three-dimensional geological data can be three-dimensional model data representing geological structures.

[0139] The stress transfer attenuation function can be a function that describes the attenuation law of stress as it is transmitted in soil.

[0140] Specifically, spatial interpolation algorithms are used to fuse soil layer distribution parameters with soil stress data in three-dimensional space to generate three-dimensional geological data. Then, the continuity of soil layers in the three-dimensional geological data is analyzed, and combined with soil constitutive relations, a stress attenuation coefficient is derived, forming a stress transfer attenuation function. Furthermore, based on a path tracing algorithm, the stress propagation process in the soil is simulated to construct the soil stress transmission path: first, the spatiotemporal distribution characteristics and stress transfer attenuation function are used to determine the stress source point using temporal distribution characteristics; second, the stress transfer attenuation function is applied to calculate the direction of stress propagation and the degree of attenuation in three-dimensional space; finally, the soil stress transmission path is output.

[0141] This scheme constructs three-dimensional geological data based on soil and rock layer distribution parameters and soil stress data, eliminating the problem of insufficient integration of multi-source monitoring data and laying the foundation for a comprehensive assessment of soil stability. Based on the three-dimensional geological data, a stress transmission attenuation function is constructed to address the lack of dynamic response pattern recognition, identify weak areas in the geological structure, and capture the dynamic response patterns of the geological structure. Based on the spatiotemporal distribution characteristics and stress transmission attenuation function, soil stress transmission paths are constructed, providing input to address the issues of inaccurate and untargeted risk assessments and supporting the quantification of geological hazard risks.

[0142] In some embodiments, the seepage pressure change rate is determined based on environmental hydrological data; weak areas of the geological structure are determined based on the soil stress transmission path and the seepage pressure change rate; and the dynamic response mode of the geological structure is determined based on the seepage pressure change rate and the weak areas of the geological structure.

[0143] The rate of change of osmotic pressure can be the change in the pore water pressure gradient over time.

[0144] Areas with weak geological structures can be karst areas or the boundaries between soil layers.

[0145] Specifically, pore water pressure gradient data is extracted from environmental hydrological data, and the rate of change over time is calculated to obtain the seepage pressure change rate. Then, by combining the soil stress transmission path and the seepage pressure change rate, stress concentration areas and seepage anomaly areas are analyzed to identify areas with weak geological structures. Finally, based on the seepage pressure change rate and areas with weak geological structures, the real-time response mechanism of the geological body to hydrological changes is derived, thus forming a dynamic response model of the geological structure.

[0146] This scheme determines the rate of change of seepage pressure based on environmental hydrological data, thereby providing a measure of the impact of hydrological factors and eliminating the problem of lacking quantitative analysis of the rate of change of seepage pressure. Based on the soil stress transmission path and the rate of change of seepage pressure, it identifies weak areas in the geological structure, eliminating the problem of inaccurate identification of these areas and providing key input for determining the dynamic response model of the geological structure. Based on the rate of change of seepage pressure and weak areas in the geological structure, it determines the dynamic response model of the geological structure, thereby enabling a description of the dynamic response to geological hazards, eliminating the problem of the inability to determine the dynamic response model of the geological structure, and providing a reliable basis for risk early warning.

[0147] In some embodiments, historical disaster cases are acquired, analyzed, and disaster process data are determined; a spatiotemporal coupled prediction model is constructed using the disaster process data; soil characteristic parameters are input into the spatiotemporal coupled prediction model, and the displacement gradient change rate and energy accumulation coefficient output by the spatiotemporal coupled prediction model are obtained; the risk evolution rate of each building to be monitored is calculated based on the displacement gradient change rate, energy accumulation coefficient, and seepage pressure change rate; and the geological disaster risk level is determined based on the dynamic response mode of the geological structure and the risk evolution rate.

[0148] Historical disaster cases can be records of historical geological disaster events.

[0149] Disaster process data can be a complete record of a sequence of data from historical disaster events.

[0150] Spatiotemporal coupling prediction models can be mathematical models that combine time and spatial dimensions.

[0151] The rate of change of displacement gradient can be the soil displacement rate.

[0152] The energy accumulation coefficient can be a measure of the degree of stress accumulation.

[0153] The building to be monitored can be a high-risk target building.

[0154] The rate of risk evolution can be the speed at which a disaster develops.

[0155] Specifically, historical disaster cases are retrieved from disaster databases and monitoring records; multi-source data such as disaster occurrence time, spatial location, soil stress changes, and groundwater level fluctuations are extracted using data mining techniques; subsequently, multi-source data are integrated to determine the disaster process data. Based on the disaster process data, a machine learning algorithm is used to train the model according to a supervised learning framework; and through spatiotemporal correlation analysis, the disaster development pattern is learned to construct a spatiotemporally coupled prediction model that integrates spatiotemporal dimensions. Real-time collected soil characteristic parameters are input into the spatiotemporally coupled prediction model, and the displacement gradient change rate and energy accumulation coefficient are output through the spatiotemporal correlation analysis algorithm. For each building to be monitored, the risk evolution rate is calculated: first, the displacement gradient change rate, energy accumulation coefficient, and seepage pressure change rate are combined; then, a weighted or normalized algorithm is applied to calculate the risk evolution rate. The dynamic response pattern of the geological structure is matched with the risk evolution rate value; finally, based on historical disaster statistics and empirical knowledge, a preset threshold rule for the risk evolution rate is set to assign a risk level to each building to be monitored.

[0156] This solution acquires and analyzes historical disaster cases to determine disaster process data, ensuring that the prediction model is based on the actual disaster evolution process. Using this disaster process data, a spatiotemporal coupled prediction model is constructed, providing a computational framework for dynamically predicting soil behavior. Soil characteristic parameters are input into the spatiotemporal coupled prediction model, and the displacement gradient change rate and energy accumulation coefficient output by the model are obtained, thereby quantifying the dynamic changes of the soil and providing real-time parameters for risk calculation. Based on the displacement gradient change rate, energy accumulation coefficient, and seepage pressure change rate, the risk evolution rate of each monitored structure is calculated, enabling personalized risk assessment for different structures. Based on the dynamic response pattern of the geological structure and the risk evolution rate, the geological disaster risk level is determined to protect the safety of structures in a targeted manner.

[0157] In some embodiments, disaster process data are classified to determine disaster precursor parameters and disaster development data; based on time series, disaster precursor parameters and disaster development data are analyzed to determine the correlation characteristics between disaster precursor parameters and disaster development time series; based on the correlation characteristics, a long short-term memory network is used to construct a time correlation sub-model according to disaster precursor parameters and disaster development data to obtain a spatiotemporal coupled prediction model.

[0158] Precursor parameters for disasters can be sequences of monitoring data prior to a disaster.

[0159] Disaster development data can be a sequence of monitoring data during the disaster's expansion process.

[0160] A time series can be a sequence of data points arranged in chronological order.

[0161] Disaster development timeline can be a sequence of disaster development data arranged in chronological order.

[0162] Correlation features can be the correlation patterns or dependencies between disaster precursor parameters and the timing of disaster development in the time dimension.

[0163] Long Short-Term Memory (LSTM) networks can be a type of recurrent neural network used to process time-series data and capture long-term dependencies.

[0164] The temporal correlation sub-model can be a model component trained on a long short-term memory network for predicting the evolution of disasters over time.

[0165] Specifically, based on predefined classification rules established from physical attributes and historical data patterns in existing disaster monitoring theories, disaster process data is separated into disaster precursor parameters and disaster development data. Then, within a time-series framework, time-axis alignment analysis is performed on the time series of disaster precursor parameters and the time series of disaster development data. Furthermore, time-series analysis methods are used to compare the variation patterns of disaster precursor parameters with the evolution trends of disaster development data. Subsequently, pattern matching is used to identify the correlation features between disaster precursor parameters and disaster development time series. Next, the disaster precursor parameters and disaster development time series are input into a Long Short-Term Memory (LSTM) network, and network weights are initialized using correlation features. Then, the LTM network's gating mechanism is used to cyclically update the hidden states of the disaster precursor parameters and disaster development time series, capturing long- and short-term time dependencies. Finally, after the LTM network is trained, a time-related sub-model focusing on the time dimension is constructed. Finally, the set of time-related sub-models is embedded as a core module into a spatiotemporal coupled prediction model.

[0166] This scheme categorizes disaster process data, identifies precursor parameters and disaster development data, and provides structured input for time series analysis. Based on time series analysis, it determines the correlation characteristics between precursor parameters and disaster development time series, ensuring timestamp matching of data points and eliminating interference from time offsets. Based on these correlation characteristics, a Long Short-Term Memory (LSTM) network is used to construct a time-related sub-model based on precursor parameters and disaster development data, resulting in a spatiotemporal coupled prediction model. This model provides dynamic prediction capabilities in the time dimension, offering time-series evolution input for the spatial analysis module.

[0167] In some embodiments, environmental hydrological data is analyzed to determine the historical rainfall intensity of the current season; the rainfall infiltration coupling coefficient is determined based on the energy accumulation coefficient, areas with weak geological structures, and historical rainfall intensity; basic information of each building to be monitored is obtained; the basic information is analyzed to determine the burial depth characteristics of each building to be monitored; a burial depth influence weight matrix is ​​generated based on the burial depth characteristics and the displacement gradient change rate; a three-dimensional risk evolution model is established based on the rainfall infiltration coupling coefficient and the burial depth influence weight matrix; and the risk evolution rate of each building to be monitored is output based on the three-dimensional risk evolution model.

[0168] The current season can be determined by a clock or by user input.

[0169] Historical rainfall intensity can be a scalar value obtained by statistically analyzing historical rainfall data.

[0170] The rainfall-permeability coupling coefficient can be used to quantify the coupling effect of rainfall permeability on soil stability.

[0171] Basic information can include the unique identifier, location coordinates, and basic structural data of the building to be monitored.

[0172] The burial depth characteristic can be a normalized value of the foundation depth of a building.

[0173] The burial depth influence weight matrix can be a matrix form used to quantify the weighted influence of burial depth and displacement on risk.

[0174] A three-dimensional risk evolution model can be a three-dimensional space-time model used to predict risk distribution.

[0175] Specifically, historical rainfall records are extracted from environmental hydrological data. Then, based on the current season, rainfall data for the same season in the historical database is queried. Next, a statistical method is used to average the rainfall data over time to determine the historical rainfall intensity for the current season. Attribute data of geologically weak areas are obtained from a geological exploration database. Then, combining the energy accumulation coefficient and historical rainfall intensity, a rainfall infiltration coupling coefficient is calculated using a coupling function. Furthermore, by accessing an urban building information database, basic information for each building to be monitored is retrieved in batches according to a predefined list of buildings. Subsequently, the basic information of each building is analyzed using an algorithm to extract the "foundation depth" field and normalize it, thus outputting the burial depth characteristics of each building to be monitored. Then, for each building to be monitored, a burial depth influence weight matrix is ​​calculated by combining the burial depth characteristics and the displacement gradient change rate. Then, the rainfall infiltration coupling coefficient is used as a global coupling factor. Finally, a three-dimensional risk evolution model is constructed using a numerical integration method, combined with the burial depth influence weight matrix. Finally, the time series of risk values ​​corresponding to the location of each building to be monitored is extracted from the three-dimensional risk evolution model to calculate the risk evolution rate per unit time.

[0176] This scheme analyzes environmental hydrological data to determine the historical rainfall intensity of the current season, avoiding early warning delays caused by ignoring seasonal rainfall. Based on the energy accumulation coefficient, areas with weak geological structures, and historical rainfall intensity, it determines the rainfall infiltration coupling coefficient, quantifying the coupling effect of rainfall infiltration on soil stability and correcting the bias of failing to construct soil stress transmission paths. It acquires basic information for each monitored building, avoiding overly broad or missed warnings due to a lack of urban structural information coupling. It analyzes the basic information to determine the burial depth characteristics of each monitored building, eliminating the problems of inaccurate and untargeted risk level predictions. Based on burial depth characteristics and displacement gradient change rate, it generates a burial depth influence weight matrix, quantifying the weight allocation of burial depth and displacement on risk, thereby eliminating the deficiency of inaccurate calculation of displacement gradient change rate and energy accumulation coefficient. Based on the rainfall infiltration coupling coefficient and burial depth influence weight matrix, it establishes a three-dimensional risk evolution model, achieving multi-source data fusion and correcting the inability to couple urban structural information and real-time parameters. Based on the three-dimensional risk evolution model, it outputs the risk evolution rate for each monitored building, enabling personalized early warnings for different monitored buildings.

[0177] In some embodiments, weak geological structural areas are analyzed to determine the scope of structural weakness; urban structural information is obtained, analyzed, and urban buildings and their coordinate information are determined; the scope of structural weakness and coordinate information are coupled, and the buildings to be monitored are determined based on the coupling results.

[0178] The area of ​​structural weakness can be generated by analyzing the geographic polygon data of areas with weak geological structures.

[0179] Urban structure information can be a raw dataset containing a list of buildings and their attributes.

[0180] Urban buildings can be a list of structured buildings.

[0181] Coordinate information can be the location data of buildings in each city.

[0182] The coupling result can be a list of Boolean values ​​generated from the structural weak areas and coordinate information.

[0183] Specifically, based on attribute data of geologically weak areas, a spatial analysis algorithm is used to read the coordinate point set of the attribute data, calculate the boundary to form a continuous geographical range, and thus determine the structurally weak area. Then, the urban building information database is accessed, and urban structural information is retrieved according to predefined query conditions. Next, based on the urban structural information, the algorithm traverses the building list, extracts the "coordinates" field, and verifies data integrity. Then, a unique identifier is assigned to each urban building to determine the building and its corresponding coordinate information. Finally, the structurally weak area and coordinate information are spatially coupled. The algorithm traverses the coordinate points of each urban building, checking whether it is located within the structurally weak area, thus determining the coupling result. Then, based on the coupling result, urban buildings whose coordinates are within the structurally weak area are selected. Finally, urban buildings within the structurally weak area are marked as buildings to be monitored.

[0184] This scheme analyzes areas with weak geological structures, determines the extent of these weaknesses, and quantifies the spatial distribution of high-risk geological disaster areas. It acquires and analyzes urban structural information to determine the coordinates of urban buildings and each building, ensuring the comprehensiveness and real-time nature of the data source and eliminating the drawback of overly broad warning ranges. The scheme couples the weak structural areas and coordinate information; based on the coupling results, it identifies buildings to be monitored, quantifies the correlation between buildings and geological risk areas, and eliminates the problem of missed warnings.

[0185] In some embodiments, groundwater level monitoring data is analyzed to determine the gradient of pore water pressure change in the soil; the permeability coefficient and thickness of each soil layer are determined based on the soil layer distribution parameters; a seepage velocity equation for each soil layer is established based on Darcy's law; the layered seepage pressure attenuation rate is calculated based on the seepage velocity equation and groundwater level monitoring data; a seepage pressure transfer function is generated by weighted superposition of the seepage pressure attenuation rates of each layer; and the seepage pressure change rate is generated based on the seepage pressure transfer function and the gradient of pore water pressure change in the soil.

[0186] Groundwater level monitoring data can be groundwater level time series data.

[0187] The gradient of soil pore water pressure change can be the rate of change of soil pore water pressure in the vertical direction per unit time.

[0188] Soil layers can be layered units with physical properties in a geological structure.

[0189] The permeability coefficient can be used as an indicator of the rate at which fluid passes through a soil layer.

[0190] Thickness can be the vertical height of the soil layer.

[0191] Darcy's law can be described as the principle that fluid flow rate is proportional to the hydraulic gradient.

[0192] The seepage velocity equation can be an equation that describes the relationship between fluid velocity and pressure within a soil layer.

[0193] The stratified permeability pressure decay rate can be the decrease in fluid pressure per unit path length in different soil layers.

[0194] The osmotic pressure transfer function can be a function that describes the relationship between total pressure decay and depth.

[0195] Specifically, the groundwater level monitoring data is processed differentially to calculate the water level change per unit time and determine the vertical gradient of pore water pressure in the soil. Then, the distribution parameters of the soil layers are analyzed to obtain the physical properties of several soil layers, determining the permeability coefficient and thickness of each layer. Next, based on Darcy's law and the permeability coefficient of each soil layer, a seepage velocity equation is constructed for each layer. Subsequently, the groundwater level monitoring data is input into the seepage velocity equation for each soil layer to calculate the pressure change as the fluid passes through different soil layers, thus determining the layered seepage pressure attenuation rate for each layer. Then, based on the soil layer depth order and with the thickness of each layer as a weight, the seepage pressure attenuation rates of each layer are linearly superimposed to generate a seepage pressure transfer function that comprehensively represents the overall geological profile. Finally, the pore water pressure change gradient is input into the seepage pressure transfer function to calculate the comprehensive change in seepage pressure per unit time under the current hydrological conditions, thus determining the seepage pressure change rate.

[0196] This scheme analyzes groundwater level monitoring data to determine the gradient of pore water pressure changes in soil, quantifies the change of pore water pressure gradient over time, and eliminates the core deficiency of lacking analysis of seepage pressure change rate. Based on the distribution parameters of soil and rock layers, the permeability coefficient and thickness of each soil layer are determined, addressing the inability to construct three-dimensional geological data based on these parameters, ensuring the feasibility of layered analysis, and providing attribute data for Darcy's law. Based on Darcy's law, a seepage velocity equation is established for each soil layer, eliminating the inability to calculate the layered seepage pressure attenuation rate. Based on the seepage velocity equation and groundwater level monitoring data, the layered seepage pressure attenuation rate is calculated, quantifying the vertical attenuation of seepage pressure and capturing the cumulative effect of hydrological changes in different soil layers. By weighted superposition of the seepage pressure attenuation rates of each layer, a seepage pressure transfer function is generated, providing a unified expression for the cumulative effect of hydrological pressure in the vertical direction and eliminating blind spots in overall response analysis. Based on the seepage pressure transfer function and the gradient of pore water pressure changes in soil, a seepage pressure change rate is generated, enabling real-time assessment of the risk of soil instability caused by hydrological changes, thereby improving the accuracy and timeliness of geological disaster early warning.

[0197] Figure 3 A schematic diagram of the structure of a real-time geological disaster monitoring platform based on intelligent AI, as provided in one embodiment of this application, is shown below. Figure 3As shown, the AI-based real-time geological disaster monitoring platform 300 of this embodiment includes: a data analysis module 301, a parameter determination module 302, a mode determination module 303, and a level determination module 304.

[0198] Data analysis module 301 is used to acquire multi-source monitoring data; analyze the multi-source monitoring data, and determine soil stability;

[0199] The parameter determination module 302 is used to determine soil characteristic parameters based on the soil stability.

[0200] The mode determination module 303 is used to acquire environmental hydrological data and determine the dynamic response mode of the geological structure based on the environmental hydrological data.

[0201] The level determination module 304 is used to determine the geological hazard risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure.

[0202] Optionally, the multi-source monitoring data includes soil stress data, and when the parameter determination module 302 determines soil characteristic parameters based on the soil stability, it is used for:

[0203] By analyzing the multi-source monitoring data, the deformation data of the underground structure can be obtained;

[0204] Analyze the deformation data of the underground structure to determine the spatiotemporal distribution characteristics of the underground structure;

[0205] Based on the aforementioned spatiotemporal distribution characteristics, the soil stress data is analyzed to determine the soil stress transmission path;

[0206] Based on the stress transmission path of the soil, determine the characteristic parameters of the soil.

[0207] Optionally, the multi-source monitoring data also includes soil and rock layer distribution parameters. When the parameter determination module 302 analyzes the soil stress data based on the spatiotemporal distribution characteristics to determine the soil stress transmission path, it is used for:

[0208] Based on the soil and rock layer distribution parameters and the soil stress data, three-dimensional geological data are constructed.

[0209] Based on the three-dimensional geological data, a stress transfer attenuation function is constructed.

[0210] Based on the spatiotemporal distribution characteristics and the stress transmission attenuation function, a soil stress transmission path is constructed.

[0211] Optionally, when the mode determination module 303 determines the dynamic response mode of the geological structure based on the environmental hydrological data, it is used for:

[0212] Based on the aforementioned environmental hydrological data, determine the rate of change of seepage pressure;

[0213] Based on the soil stress transmission path and the seepage pressure change rate, the weak areas of the geological structure are identified;

[0214] Based on the rate of change of osmotic pressure and the weak areas of the geological structure, the dynamic response mode of the geological structure is determined.

[0215] Optionally, when the level determination module 304 determines the geological hazard risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure, it is used for:

[0216] Acquire historical disaster cases, analyze these cases, and determine disaster process data;

[0217] A spatiotemporal coupled prediction model is constructed using the disaster process data.

[0218] The soil characteristic parameters are input into the spatiotemporal coupling prediction model, and the displacement gradient change rate and energy accumulation coefficient output by the spatiotemporal coupling prediction model are obtained.

[0219] The risk evolution rate of each building to be monitored is calculated based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate.

[0220] The geological hazard risk level is determined based on the dynamic response mode of the geological structure and the risk evolution rate.

[0221] Optionally, when the level determination module 304 constructs a spatiotemporal coupled prediction model using the disaster process data, it is used for:

[0222] The disaster process data is classified to determine the disaster precursor parameters and disaster development data;

[0223] Based on time series analysis, the disaster precursor parameters and disaster development data are used to determine the correlation characteristics between the disaster precursor parameters and the disaster development time series.

[0224] Based on the aforementioned correlation features, a long short-term memory network is used to construct a time-related sub-model according to the disaster precursor parameters and the disaster development data, thereby obtaining a spatiotemporal coupled prediction model.

[0225] Optionally, when the level determination module 304 calculates the risk evolution rate of each building to be monitored based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate, it is used for:

[0226] Analyze the aforementioned environmental hydrological data to determine the historical rainfall intensity for the current season;

[0227] The rainfall infiltration coupling coefficient is determined based on the energy accumulation coefficient, the geologically weak area, and the historical rainfall intensity.

[0228] Obtain basic information for each building to be monitored;

[0229] The basic information is analyzed to determine the burial depth characteristics of each building to be monitored;

[0230] Based on the burial depth characteristics and the displacement gradient change rate, a burial depth influence weight matrix is ​​generated;

[0231] A three-dimensional risk evolution model is established based on the rainfall infiltration coupling coefficient and the burial depth influence weight matrix.

[0232] Based on the three-dimensional risk evolution model, the risk evolution rate of each building to be monitored is output.

[0233] Optionally, the AI-based real-time geological disaster monitoring platform further includes a building determination module 305, used for:

[0234] Analyze the weak areas of the geological structure to determine the extent of the structural weakness;

[0235] Obtain urban structure information, parse the urban structure information, and determine urban buildings and the coordinate information of each urban building;

[0236] The weak structural area and the coordinate information are coupled together, and the building to be monitored is determined based on the coupling result.

[0237] Optionally, the environmental hydrological data includes groundwater level monitoring data, and when the pattern determination module 303 determines the rate of change of seepage pressure based on the environmental hydrological data, it is used for:

[0238] Analyze the groundwater level monitoring data to determine the gradient of soil pore water pressure change;

[0239] Based on the soil and rock layer distribution parameters, determine the permeability coefficient and thickness of each soil layer;

[0240] Based on Darcy's law, establish the seepage velocity equation for each soil layer;

[0241] Based on the seepage velocity equation and the groundwater level monitoring data, the stratified seepage pressure attenuation rate is calculated.

[0242] The osmotic pressure transfer function is generated by weighting and superimposing the osmotic pressure decay rate of each layer.

[0243] The rate of change of osmotic pressure is generated based on the osmotic pressure transfer function and the gradient of pore water pressure change in the soil.

[0244] The platform in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A real-time monitoring method for geological disasters based on intelligent AI, characterized in that, include: Acquire multi-source monitoring data; Analyze the multi-source monitoring data to determine soil stability; Based on the soil stability, determine the soil characteristic parameters; Acquire environmental hydrological data, and determine the dynamic response mode of the geological structure based on the environmental hydrological data; Based on the soil characteristic parameters and the dynamic response mode of the geological structure, the geological hazard risk level is determined, including: Acquire historical disaster cases, analyze these cases, and determine disaster process data; A spatiotemporal coupled prediction model is constructed using the disaster process data. The soil characteristic parameters are input into the spatiotemporal coupling prediction model, and the displacement gradient change rate and energy accumulation coefficient output by the spatiotemporal coupling prediction model are obtained. The risk evolution rate of each building to be monitored is calculated based on the displacement gradient change rate, the energy accumulation coefficient, and the seepage pressure change rate. The geological hazard risk level is determined based on the dynamic response mode of the geological structure and the risk evolution rate. The step of calculating the risk evolution rate of each monitored building based on the displacement gradient change rate, the energy accumulation coefficient, and the osmotic pressure change rate includes: Analyze the aforementioned environmental hydrological data to determine the historical rainfall intensity for the current season; The rainfall infiltration coupling coefficient is determined based on the energy accumulation coefficient, the geologically weak area, and the historical rainfall intensity. Obtain basic information for each building to be monitored; The basic information is analyzed to determine the burial depth characteristics of each building to be monitored; Based on the burial depth characteristics and the displacement gradient change rate, a burial depth influence weight matrix is ​​generated; A three-dimensional risk evolution model is established based on the rainfall infiltration coupling coefficient and the burial depth influence weight matrix. Based on the three-dimensional risk evolution model, the risk evolution rate of each building to be monitored is output.

2. The method of claim 1, wherein, The multi-source monitoring data includes soil stress data, and the determination of soil characteristic parameters based on the soil stability includes: By analyzing the multi-source monitoring data, the deformation data of the underground structure can be obtained; Analyze the deformation data of the underground structure to determine the spatiotemporal distribution characteristics of the underground structure; Based on the aforementioned spatiotemporal distribution characteristics, the soil stress data is analyzed to determine the soil stress transmission path; Based on the stress transmission path of the soil, determine the characteristic parameters of the soil.

3. The method of claim 2, wherein, The multi-source monitoring data also includes soil and rock layer distribution parameters. The analysis of soil stress data based on the spatiotemporal distribution characteristics to determine the soil stress transmission path includes: Based on the soil and rock layer distribution parameters and the soil stress data, three-dimensional geological data are constructed. Based on the three-dimensional geological data, a stress transfer attenuation function is constructed. Based on the spatiotemporal distribution characteristics and the stress transmission attenuation function, a soil stress transmission path is constructed.

4. The method of claim 3, wherein, The step of determining the dynamic response mode of the geological structure based on the environmental hydrological data includes: Based on the aforementioned environmental hydrological data, determine the rate of change of seepage pressure; Based on the soil stress transmission path and the seepage pressure change rate, the weak areas of the geological structure are identified; Based on the rate of change of osmotic pressure and the weak areas of the geological structure, the dynamic response mode of the geological structure is determined.

5. The method of claim 4, wherein, The construction of a spatiotemporal coupled prediction model using the disaster process data includes: The disaster process data is classified to determine the disaster precursor parameters and disaster development data; Based on time series analysis, the disaster precursor parameters and disaster development data are used to determine the correlation characteristics between the disaster precursor parameters and the disaster development time series. Based on the aforementioned correlation features, a long short-term memory network is used to construct a time-related sub-model according to the disaster precursor parameters and the disaster development data, thereby obtaining a spatiotemporal coupled prediction model.

6. The method of claim 5, wherein, The determination of the building to be monitored includes: Analyze the weak areas of the geological structure to determine the extent of the structural weakness; Obtain urban structure information, parse the urban structure information, and determine urban buildings and the coordinate information of each urban building; The weak structural area and the coordinate information are coupled together, and the building to be monitored is determined based on the coupling result.

7. The method of claim 4, wherein, The environmental hydrological data includes groundwater level monitoring data, and the determination of the seepage pressure change rate based on the environmental hydrological data includes: Analyze the groundwater level monitoring data to determine the gradient of soil pore water pressure change; Based on the soil and rock layer distribution parameters, determine the permeability coefficient and thickness of each soil layer; Based on Darcy's law, establish the seepage velocity equation for each soil layer; Based on the seepage velocity equation and the groundwater level monitoring data, the stratified seepage pressure attenuation rate is calculated. The osmotic pressure transfer function is generated by weighting and superimposing the osmotic pressure decay rate of each layer. The rate of change of osmotic pressure is generated based on the osmotic pressure transfer function and the gradient of pore water pressure change in the soil.

8. A real-time monitoring platform for geological disasters based on intelligent AI, applied to the method of any one of claims 1-7, characterized in that, include: The data analysis module is used to acquire multi-source monitoring data; Analyze the multi-source monitoring data to determine soil stability; The parameter determination module is used to determine the characteristic parameters of the soil based on the soil stability. The mode determination module is used to acquire environmental hydrological data and determine the dynamic response mode of the geological structure based on the environmental hydrological data. The level determination module is used to determine the geological hazard risk level based on the soil characteristic parameters and the dynamic response mode of the geological structure.

Images