Flexible slope support optimal design system

Abstract

The invention discloses a flexible slope support optimization design system, and relates to the technical field of flexible slope support, a data acquisition module is constructed, dynamic data and image data of a slope surface are acquired based on a multi-array sensor deployed on the slope surface, and the dynamic data are transmitted to a dynamic relational database; the instability model building module is used for analyzing the dynamic data to generate an instability model of the slope surface and generating a stress analysis report of the slope surface based on the instability model; a support interference field module is constructed, a support interference field is generated based on the stress analysis report, and an optimization scheme of flexible slope support is generated based on interaction of the support interference field and the instability model; constructing a visualization module for visually displaying the instability model and the optimization scheme of the flexible slope support; according to the scheme, an intelligent analysis closed loop with data driving and mechanical model depth coupling is constructed, and efficient optimization of slope support from passive response to active prediction is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of flexible slope support, in particular to a flexible slope support optimization design system. BACKGROUND

[0002] Currently, the optimization design of flexible slope support engineering mainly relies on periodic manual survey and static monitoring data of limited points, and the traditional optimization method is usually based on discrete, historical geological survey reports and simplified mechanical models; Then, the prior art has some deficiencies, the traditional monitoring system can only provide isolated physical measurement values, such as displacement or stress data of a single point, and lacks the ability to effectively associate and fuse multi-source, heterogeneous monitoring data with environmental factors, and the understanding of the slope state is insufficient, which leads to that the data analysis stays on the surface phenomenon, and the underlying mechanical mechanism and evolution process of the rock mass cannot be revealed, so that the judgment of the potential instability mode seriously depends on the personal experience of engineers, and the objectivity and systematicness are insufficient; at the same time, once the traditional support design scheme is determined, it is usually fixed during the construction period, and it is difficult to adjust in real time according to the dynamic response of the slope under the action of the external environment; the optimization process is an offline, non-continuous static calculation, and a dynamic model reflecting the interaction of the internal state of the slope and the propagation path of the instability risk cannot be constructed, so that it is difficult to realize the precise and active matching of the support structure and the slope instability process in time and space, which leads to that the support scheme may lack foresight and adaptability; Therefore, it is of great significance to develop an intelligent flexible slope support optimization design system. SUMMARY

[0003] The purpose of the present application is to provide a flexible slope support optimization design system to solve the problems in the background art.

[0004] In order to achieve the above purpose, the present application provides the following technical scheme: a flexible slope support optimization design system, comprising: A data acquisition module acquires dynamic data and image data of the slope surface based on a plurality of array sensors deployed on the slope surface, and transmits the dynamic data to a dynamic relational database; A instability model construction module connected with the data acquisition module is used to analyze the dynamic data to generate a instability model of the slope surface, and generate a stress analysis report of the slope surface based on the instability model; A support interference field module connected with the instability model construction module generates a support interference field based on the stress analysis report, and generates an optimization scheme of the flexible slope support based on the interaction of the support interference field and the instability model; A visualization module connected with the support interference field module is used to visually display the instability model and the optimization scheme of the flexible slope support.

[0005] In a preferred embodiment, the data acquisition module comprises: a multi-array sensor unit distributed on the surface and inside of the slope for collecting dynamic data of the slope and generating multi-source time-series dynamic data with time stamp and three-dimensional geographic coordinates; an environmental monitoring unit for synchronously collecting environmental state data of the environment where the slope is located; transmitting the dynamic data and the environmental state data to a dynamic relational database to generate event data groups.

[0006] In a preferred embodiment, the step of transmitting the dynamic data to the dynamic relational database to generate event data groups comprises: The dynamic relational database comprises a real-time data caching unit, an environmental modulation unit, and a dynamic relationship construction unit; The real-time data caching unit receives multi-source time-series dynamic data in real time and stores them in a cache space by constructing multiple sub-databases; The environmental modulation unit receives environmental state data and converts them into environmental state vectors; The dynamic relationship construction unit comprises a feature rule library, a relationship calculation sub-unit, and a structured event packaging unit; The feature rule library pre-stores feature extraction operators based on the principles of rock-soil mechanics and spatio-temporal correlation rules; The relationship calculation sub-unit continuously calls the feature extraction operators to perform parallel scanning and calculation on the multi-source time-series dynamic data and the environmental state vectors, generating relationship tuples; The relationship tuples contain multi-source time-series dynamic data, environmental state vectors, relationship type identifiers, and relationship strength quantification values; The structured event packaging unit analyzes the relationship tuples through a neural network, generates event data groups based on spatio-temporal correlation rules and relationship type identifiers.

[0007] In a preferred embodiment, the instability model construction module comprises: a slope digital model construction unit for constructing a three-dimensional slope digital model based on dynamic data and image data; a distributed event analysis unit for extracting three-dimensional geographic coordinate distribution features of event data groups and dynamically generating event analysis sub-units based on the three-dimensional geographic coordinate distribution features, which are used to receive and analyze event data groups to generate local instability feature vectors; a instability knowledge base unit pre-storing a set of instability mode rules based on the principles of rock-soil mechanics, which are used to define the mapping relationship between different combinations of local instability feature vectors and potential instability modes; The node cooperative reasoning unit is connected with each event analysis subunit and the instability knowledge base unit, cooperatively reasons the local instability feature vector, and synthesizes the instability model describing the slope stability state. Based on the instability model, a stress analysis report of the slope surface is generated.

[0008] In a preferred embodiment, the step of the node cooperative reasoning unit being connected with each event analysis subunit and the instability knowledge base unit, cooperatively reasoning the local instability feature vector, and synthesizing the instability model describing the slope stability state is: The node cooperative reasoning unit receives the local instability feature vector from each event analysis subunit, and performs spatiotemporal alignment processing on all the local instability feature vectors; The converged local instability feature vector set is matched with the instability mode rule set in the instability knowledge base unit, the corresponding potential instability mode is activated by calculating the matching degree of the feature vector and the instability mode rule set; Based on the three-dimensional geographic coordinates in the local instability feature vector, the correlation strength between different position local instability features is analyzed, and an association network representing the instability propagation path is constructed with each event analysis subunit as a node and the correlation strength as an edge; Combined with the activated potential instability mode and the association network, the instability model of the slope is synthesized through a weighted fusion algorithm; Based on the instability model, a stress analysis report of the slope surface is generated. The stress analysis report includes potential sliding surface geometric parameters, safety factor distribution, stress-strain field data, support resistance demand distribution, and dynamic evolution prediction.

[0009] In a preferred embodiment, the support intervention field module includes: The field mapping unit is used to receive the stress analysis report, and map the support resistance demand distribution, potential sliding surface geometric parameters, and stress-strain field data in the report into an initial intervention field overlaid on the instability model; The dynamic field generator generates a dynamic field generator based on the corresponding position of the three-dimensional geographic coordinates of the event analysis subunit on the initial intervention field, and is connected with the event analysis subunit through a dynamic channel; The field model interaction engine is constructed in the dynamic field generator, an optimization objective function is built in, the local instability feature vector from the event analysis subunit is received through the dynamic channel, and the local instability feature vector is input into the optimization objective function, and the parameters of the initial intervention field are dynamically corrected through an iterative adjustment algorithm; The channel monitoring unit continuously monitors the real-time load of all dynamic channels, and based on a preset super-fusion threshold, the dynamic channels are materialized to generate anchor nails of the initial intervention field. The modified parameters of the initial interference field and the anchor generating support interference field are obtained, and the parameters of the support interference field are discretized into the spatial layout coordinates, stress parameters and construction sequence of the flexible support member through the field streamline tracking technology.

[0010] In a preferred embodiment, the step of continuously monitoring the real-time load of all dynamic channels, and solidifying the dynamic channels based on a preset overload threshold, and generating the anchors of the flexible support net is: The channel monitoring unit continuously monitors the real-time load of all data channels, and sets a preset overload threshold; When it is found that the real-time load of one or more dynamic channels in a preset range continuously exceeds the preset overload threshold within a preset time, it is determined that the area is a high-energy core area; The channel monitoring unit sends a freezing application to the dynamic channels in the high-energy core area, and the dynamic channels receiving the freezing application are solidified to generate anchors and record the anchor parameters, and the anchor data is transmitted to the field model interaction engine; The anchor parameters include spatial positioning parameters, mechanical property parameters and dynamic correlation parameters.

[0011] In a preferred embodiment, the visualization module comprises: The model rendering unit is used for generating a three-dimensional visualization scene based on the instability model and the support interference field, and rendering the geometric shape, instability area and support structure of the slope in real time; The data superposition unit is used for superimposedly displaying the dynamic data, stress analysis report and optimization scheme in the form of layers; The early warning display unit dynamically marks the high-risk area based on the instability model, and highlights the early warning information in the form of color or animation.

[0012] In the above technical solution, the present application provides technical effects and advantages: 1. The present application continuously collects dynamic data and image data through the multiple array sensors deployed on the slope surface, and efficiently processes the data in combination with the dynamic relational database, so that the system can capture the subtle changes of the slope state in real time. The instability model construction module integrates multiple source data to generate a comprehensive instability model through the distributed event analysis unit and the node collaborative reasoning unit, and accurately reflects the stress state of the slope. The support interference field module converts the stress analysis into an initial interference field through the field mapping unit, and dynamically modifies the parameters through the interaction between the dynamic field generator and the event analysis subunit. The channel monitoring unit monitors the data load and automatically generates anchors to ensure the stability of the support structure. The mechanism generates an optimization scheme based on the intelligent algorithm of rock-soil mechanics principle, reduces the subjective error of manual design, improves the reliability and economy of the scheme, and greatly reduces the probability of slope instability accidents through the active monitoring and rapid intervention mechanism, thereby improving the intelligent level of flexible slope support. 2、The application extracts features by embedding prior knowledge of geotechnical mechanics, performs parallel calculation and correlation analysis on spatiotemporal sequence dynamic data and environmental state vectors, and generates relationship tuples with clear physical meaning; then, the application performs structured packaging through a neural network model to form an event data group representing a specific geotechnical mechanical process; this process essentially completes the paradigm shift from low-dimensional numerical perception to high-dimensional situation understanding, and upgrades discrete physical observation values to structured knowledge units containing causal logic; such knowledge units not only represent changes in the mechanical state of a local area, but also provide rich standardized inputs for subsequent distributed collaborative reasoning and instability mode identification of the system through their inherent spatiotemporal correlation attributes, thereby laying the foundation for the system's cognitive decision-making. BRIEF DESCRIPTION OF DRAWINGS

[0013] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art based on these drawings.

[0014] Figure 1 The system flowchart of the present application.

[0015] Figure 2 The logic block diagram of the present application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme of the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0017] Embodiment 1, please refer to Figure 1 and Figure 2 The flexible slope support optimization design system described in the present embodiment comprises: A data acquisition module acquires dynamic data and image data of the slope surface based on a plurality of array sensors deployed on the slope surface, and transmits the dynamic data to a dynamic relational database; A instability model construction module connected with the data acquisition module is used to analyze the dynamic data to generate a instability model of the slope surface, and generate a stress analysis report of the slope surface based on the instability model; Support interference field module: connected with the instability model construction module, based on the stress analysis report to generate the support interference field, based on the support interference field and the instability model interaction to generate the optimization scheme of the flexible slope support; Visualization module: connected with the support interference field module, for visualizing display of the instability model and the optimization scheme of the flexible slope support; Further, at present, the optimization design of the flexible slope support engineering mainly relies on periodic artificial survey and static monitoring data of limited points, and the traditional optimization method is usually established on the basis of discrete, historical geological survey report and simplified mechanical model; Then, the prior art has some deficiencies, the traditional monitoring system can only provide isolated physical quantity measurement values, such as displacement or stress data of a single point, lacks the ability to effectively associate and fuse multi-source, heterogeneous monitoring data and environmental factors, and the understanding depth of the slope state is insufficient, which leads to that the data analysis stays on the surface phenomenon, cannot reveal the inherent mechanical mechanism and evolution process of the rock mass hidden behind the data, makes the judgment of the potential instability mode seriously depend on the personal experience of the engineers, and the objectivity and systematization are insufficient; meanwhile, the traditional support design scheme is usually fixed once determined, and it is difficult to be adjusted in real time according to the dynamic response of the slope under the action of the external environment; the optimization process is an offline, non-continuous static calculation, and a dynamic model capable of truly reflecting the interaction of the internal state of the slope and the propagation path of the instability risk cannot be constructed, so it is difficult to realize the accurate and active matching of the support structure and the slope instability process in time and space, which leads to that the support scheme may lack foresight and adaptability; The present application can capture the subtle changes of the slope state in real time through the continuous collection of dynamic data and image data by the multi-array sensors deployed on the slope surface and the efficient processing by the dynamic relational database, the instability model construction module can generate a comprehensive instability model by integrating multi-source data through the distributed event analysis unit and the node collaborative reasoning unit, and accurately reflect the stress state of the slope; the support interference field module can convert the stress analysis into an initial interference field by using the field mapping unit, and dynamically correct the parameters by interacting with the event analysis subunit through the dynamic field generator; the channel monitoring unit can monitor the data load and automatically generate anchor nails to ensure the stability of the support structure, the mechanism can automatically generate an optimization scheme based on the intelligent algorithm of rock mechanics principle, reduce the subjective error of manual design, improve the reliability and economy of the scheme, and the active monitoring and rapid intervention mechanism can greatly reduce the probability of slope instability accidents and improve the intelligent level of the flexible slope support; The characteristic extraction operator embedded with geotechnical mechanics prior knowledge is used to perform parallel calculation and correlation analysis on the dynamic data of time-space sequence and the environmental state vector, to generate relationship tuples with clear physical meaning. Then, the neural network model is used for structured packaging to form event data groups representing specific geotechnical mechanics processes. This process essentially completes the paradigm shift from low-dimensional numerical perception to high-dimensional situation understanding, and improves discrete physical observation values to structured knowledge units containing causal logic. Such knowledge units not only represent the changes of the mechanical state of the local area, but also provide rich standardized inputs for subsequent distributed collaborative reasoning and instability mode identification of the system through their inherent spatio-temporal correlation attributes, laying the foundation for the whole system cognitive decision-making.

[0018] In one embodiment, the data acquisition module comprises: A multi-array sensing unit is distributedly deployed on the surface and inside of the slope to collect dynamic data of the slope, and to generate multi-source time-series dynamic data by labeling the dynamic data with time stamps and three-dimensional geographic coordinates; An environmental monitoring unit is used to synchronously collect environmental state data of the environment where the slope is located; The dynamic data and the environmental state data are transmitted to a dynamic relational database to generate event data groups; Further, the multi-array sensing unit is composed of geological sensor arrays arranged on the surface and inside the slope, specifically including micro-electro-mechanical system (MEMS) tilt sensors, fiber bragg grating (FBG) strain sensors, and resistance-type soil pressure cells, etc. These sensors communicate through low-power wide-area network (LPWAN) protocols. The multi-array sensing unit collects dynamic data of the slope, including deformation data, stress-strain data, and groundwater dynamic data. Deformation data is obtained by global navigation satellite system (GNSS) receivers and MEMS tilt sensors, which measure the absolute displacement and inclination angle changes of the ground surface. The array of buried inclinometers monitors the horizontal displacement of the deep layer. Stress-strain data is measured by distributed FBG sensors, which measure the continuous strain field changes along the fiber path inside the rock-soil mass. Resistance-type or vibrating wire soil pressure cells record the contact stress at key interfaces. Groundwater dynamic data is monitored by a network of osmometers, which monitor the spatial and temporal distribution and fluctuations of pore water pressure, reflecting the influence of hydrological conditions on slope stability. Each sensing node is equipped with a global positioning system (GPS) module and a miniature high-precision clock chip, which can automatically add nanosecond-level time stamps from the Beidou satellite navigation system and centimeter-level three-dimensional geographic coordinates based on spatial interpolation algorithms to the data packets while collecting dynamic physical quantities such as vibration, displacement, and stress. This generates a multi-source time-series dynamic data stream with a unified spatiotemporal reference. The environmental monitoring unit integrates weather stations, soil moisture meters, and osmometers to continuously collect environmental state data such as precipitation, temperature, humidity, and pore water pressure. After preliminary filtering and compression by the edge computing gateway, all sensing data is transmitted to the cloud dynamic relational database via the 5G dedicated slice network. The dynamic relational database adopts a fusion architecture of time series database and spatial database. Through the built-in rule engine and stream processing platform, real-time spatiotemporal registration, correlation analysis, and feature extraction are performed on the injected heterogeneous data. Finally, event data groups with complete semantic information are generated, providing standardized data input for the upper analysis module.

[0019] In one embodiment, the step of transmitting dynamic data to the dynamic relational database to generate event data groups is: The dynamic relational database includes a real-time data caching unit, an environmental modulation unit, and a dynamic relationship construction unit. The real-time data caching unit receives multi-source time-series dynamic data in real time and stores it in the cache space by building multiple sub-databases. The environmental modulation unit receives environmental state data and converts it into an environmental state vector. The dynamic relationship construction unit includes a feature rule library, a relationship calculation sub-unit, and a structured event packaging unit. The feature rule library pre-stores feature extraction operators based on rock-soil mechanics principles and spatiotemporal correlation rules. The relationship calculation subunit continuously calls the feature extraction operator to perform parallel scanning and calculation on the multi-source time-series dynamic data and the environmental state vector, and generates a relationship tuple; The relationship tuple includes the multi-source time-series dynamic data, the environmental state vector, a relationship type identifier, and a relationship strength quantization value. The structured event packaging unit analyzes the relationship tuple through a neural network, and generates an event data group based on the spatiotemporal association rule and the relationship type identifier. Further, the generation of the event data group relies on a multi-level dynamic relationship type database architecture. The real-time data caching unit uses a distributed time series database such as InfluxDB as a core storage engine, and simultaneously constructs a Redis cluster as a high-speed cache sub-database. Through a streaming processing platform, the multi-source time-series dynamic data is received and temporarily cached at a high throughput. The environmental modulation unit converts the environmental state data into a numerical vector through feature engineering. Specifically, a hot encoding process is used to process discrete meteorological categories, and a normalization method is used to process continuous physical quantities. Finally, an environmental state vector with fixed dimensions is generated. The core of the dynamic relationship construction unit is a feature rule library, in which pre-stored feature extraction operators exist in the form of pre-compiled functions, such as a differential operator for calculating displacement rate changes, a spatial clustering operator for identifying strain sets, etc. The relationship calculation subunit is implemented based on the Apache Flink streaming processing engine. Through a parallel pipeline, these operators are continuously called to perform real-time scanning on the dynamic data and the environmental vector in the cache. The Pearson correlation coefficient algorithm is used to calculate the relationship strength, and a relationship tuple containing the original data segment, the relationship type identifier, and the quantized strength value is output. Based on the obtained relationship triple, a recurrent neural network model is used to identify the spatiotemporal pattern of the input relationship tuple sequence. When a feature sequence that meets a specific instability precursor is identified, the packaging process is triggered, and an event data group with complete semantic description is generated. For example, under the condition of continuous rainfall, when the system monitors that the pore water pressure in the slope region X increases from 30 kPa to 60 kPa within 6 hours, the surface displacement rate accelerates from 0.5 mm / day to 3.0 mm / day, and the strain value in the deep strain concentration area increases from 10 micro-strain to 25 micro-strain, the system identifies the coordinated evolution of these parameters in space and time through geotechnical mechanics operators: water pressure rise triggers deformation acceleration, and the strain concentration area completely coincides with the high-risk area. Based on this, a relationship tuple representing the "hydraulic-deformation-strain strong spatiotemporal coordination" is generated, and finally the neural network determines that the multi-parameter acceleration anomaly pattern meets the instability precursor characteristics, and a semantic event data group "rainfall-induced shallow slip instability precursor in region X" is packaged and generated.

[0020] In one embodiment, the instability model construction module includes: A slope digital model construction unit constructs a three-dimensional slope digital model based on dynamic data and image data. The distributed event analysis unit extracts the three-dimensional geographic coordinate distribution characteristics of the event data group, and dynamically generates an event analysis subunit based on the three-dimensional geographic coordinate distribution characteristics, which is used to receive and analyze the event data group to generate a local instability feature vector; The instability knowledge base unit pre-stores a set of instability mode rules based on the principles of geotechnical mechanics, and the set of instability mode rules is used to define the mapping relationship between different combinations of local instability feature vectors and potential instability modes; The node collaborative reasoning unit is connected with each event analysis subunit and the instability knowledge base unit, performs collaborative reasoning on the local instability feature vectors, and synthesizes an instability model describing the state of slope stability; Based on the instability model, a stress analysis report of the slope surface is generated; Further, the technical implementation of the instability model construction module starts from the slope digital model construction unit, which establishes a three-dimensional geological entity model with centimeter-level precision by fusing laser point cloud scanning data and multi-view oblique photogrammetry images, and maps the geotechnical mechanics parameters as additional attributes to the model grid elements; The distributed event analysis unit dynamically instantiates lightweight edge computing nodes as event analysis subunits at the corresponding positions of the slope digital model according to the spatial coordinate information of the event data group. Each subunit analyzes the local event data group through the built-in feature extraction algorithm, and quantitatively generates a local instability feature vector containing dimensions such as displacement acceleration, strain energy density, and water coupling coefficient; The instability knowledge base unit adopts an expert system architecture, and its set of instability mode rules defines the triggering conditions of various instability modes in the form of production rules, for example, when the combination of feature vectors simultaneously satisfies local shear strain, displacement acceleration index exceeds the threshold value, and is spatially coupled with a high-porosity water pressure zone, the circular arc sliding mode rule is activated; The node collaborative reasoning unit normalizes the feature vectors output by each subunit through a spatio-temporal alignment engine, constructs a topological network with analysis nodes as vertices and spatio-temporal correlation strength as edges using a graph neural network algorithm, and realizes feature propagation and aggregation through a multi-layer message passing mechanism. Finally, the fused global features are matched with the instability knowledge base, and an instability model containing potential sliding surface geometric parameters, safety factor field distribution, and stress-strain cloud diagram is output, and a detailed stress analysis report is generated accordingly.

[0021] In one embodiment, the step of connecting with each event analysis subunit and the instability knowledge base unit, performing collaborative reasoning on the local instability feature vectors, and synthesizing an instability model describing the state of slope stability is as follows: The node collaborative reasoning unit receives the local instability feature vectors from each event analysis subunit and performs spatio-temporal alignment processing on all local instability feature vectors; The converged local instability feature vector set is matched with the instability mode rule set in the instability knowledge base unit, a corresponding potential instability mode is activated by calculating the matching degree of the feature vector and the instability mode rule set; Based on the three-dimensional geographic coordinates in the local instability feature vector, the correlation strength between local instability features at different positions is analyzed, and a correlation network representing the instability propagation path is constructed with each event analysis subunit as a node and the correlation strength as an edge. Combined with the activated potential instability mode and the correlation network, a weighted fusion algorithm is used to synthesize the instability model of the slope; Based on the instability model, a stress analysis report of the slope surface is generated, including potential sliding surface geometric parameters, safety factor distribution, stress-strain field data, support resistance requirement distribution, and dynamic evolution prediction. Further, the node collaborative inference unit first starts the space-time alignment engine to standardize the local instability feature vectors uploaded by the distributed event analysis unit; the engine uses the dynamic time warping algorithm to eliminate the time phase differences of each sensor data stream, and simultaneously maps the feature vectors of discrete coordinates to a unified three-dimensional grid node through the Kriging space interpolation method, to establish a feature vector matrix with space-time consistency. In the rule matching stage, the node collaborative inference unit matches the local instability feature vectors processed by the space-time alignment with the pre-stored instability mode rule set in the instability knowledge base unit. This matching process is implemented through a multi-level calculation framework. First, the matching core uses a vector similarity measurement algorithm, specifically to calculate the cosine similarity between each local instability feature vector and the pre-defined standard instability mode feature vector in the rule set, and generates an initial matching degree score. Then, the system calls the composite logical judgment conditions defined in the rule set. These conditions are scripts written based on the principles of rock mechanics, for example, a rule may require that the displacement acceleration index be greater than a set threshold, the strain concentration degree exceed a critical level, and the abnormal area overlap with a high-pore water pressure area in space, and other multiple Boolean conditions. The system will evaluate these conditions one by one and make a weighted correction to the initial matching degree to generate a potential instability mode. Based on the three-dimensional geographic coordinates in the local instability feature vector and the generated potential instability mode, the system calculates the statistical dependence of feature vectors at different positions through the mutual information entropy algorithm, quantifies the spatial correlation strength by combining the gravity model, and constructs a directed graph network with the event analysis subunit as the node and the spatial correlation strength as the edge. This network accurately represents the propagation path of the instability risk. The system integrates the activated instability mode and the correlation network through a multi-stage fusion process. First, a dynamic weight distribution mechanism is used to calculate a comprehensive weight coefficient based on the confidence value obtained by each instability mode in the rule matching stage and the topological centrality index of the mode corresponding node in the correlation network. Then, a slope stability control equation is established based on the theory of rock plasticity mechanics, and the slope digital model is converted into a numerical calculation grid through finite element discretization. In the solving process, the system embeds the potential slip path analyzed by the correlation network into the iterative calculation as a nonlinear constraint condition, and uses the strength reduction method to cyclically adjust the shear strength parameters of the rock mass, to monitor the state change of the global stiffness matrix of the model in real time. When the system stiffness matrix shows singularity characteristics, it is determined that the slope has reached a critical instability state, and the spatial coordinate parameters and geometric morphological features of the three-dimensional slip surface are accurately output at this time. The system synchronously generates complete mechanical field data including safety factor contour map, plastic strain distribution, and principal stress direction, and generates an instability model. Based on this, a stress analysis report is generated, including the geometric parameters of the slip surface, safety factor distribution, stress-strain field, support resistance requirement, and dynamic evolution prediction.

[0022] In one embodiment, the support intervention field module comprises: a field mapping unit for receiving the force analysis report and mapping the support resistance demand distribution, potential slip surface geometric parameters and stress-strain field data in the report as an initial interference field covering on the instability model; a dynamic field generator for generating a dynamic field generator based on the corresponding position of the three-dimensional geographic coordinates of the event analysis subunit on the initial interference field and connecting with the event analysis subunit through a dynamic channel; an interactive engine for constructing a field model in the dynamic field generator, built-in optimization objective function, receiving the local instability feature vector from the event analysis subunit through the dynamic channel, and inputting the local instability feature vector into the optimization objective function to dynamically correct the parameters of the initial interference field through the iterative adjustment algorithm; a channel monitoring unit for continuously monitoring the real-time load of all dynamic channels, based on the preset super-fusion threshold to entity the dynamic channel, and generating the anchor nail of the initial interference field; obtaining the correction parameters of the initial interference field and the anchor nail to generate the support interference field, and discretizing the continuous support interference field into the spatial layout coordinates, force parameters and construction sequence of the flexible support member through the field streamline tracking technology; Further, the construction process of the initial interference field is a key step of converting abstract mechanical data into concrete spatial force field. The field mapping unit first analyzes the multi-dimensional data of the force analysis report, extracts the key parameters of the field strength, and converts the support resistance demand distribution into scalar field strength data. The geometric parameters of the potential slip surface are used to define the direction reference of the vector field, and the stress-strain field data are used as the basis input of the tensor field. In the specific implementation, the system converts the discrete support resistance demand data into continuous field strength distribution based on the three-dimensional grid structure of the instability model through spatial interpolation algorithm. This process uses the improved Kriging interpolation method, which not only considers the spatial distance relationship, but also introduces the geotechnical mechanical parameters as the collaborative variables to ensure that the field strength distribution conforms to the geomechanics law. For the establishment of the vector field, the system generates the field direction distribution covering the entire model by using the tangent direction of the potential slip surface as the main reference benchmark and combining with the principal stress direction data through the vector synthesis algorithm. In order to achieve the real coverage effect, the system uses the hierarchical mapping technology to first establish the two-dimensional field distribution on the surface of the instability model, and then extends to the depth direction according to the geological stratification data to form a three-dimensional field. The field parameters of each grid element are determined through weighted calculation, and the weight depends on the relative position of the element to the slip surface and its importance in the stress field. This depth-coupled mapping method ensures that the initial interference field is completely consistent with the instability model in geometry and mechanics, laying a solid foundation for subsequent dynamic optimization. The initial interference field generated based on this not only completely covers each geometric element of the instability model, but also realizes the continuous distribution of mechanical parameters at the physical level, forming a virtual support force field that accurately corresponds to the slope instability state. The dynamic field generator then instantiates multiple field generator instances at the spatial positions corresponding to the initial interference field according to the three-dimensional geographic coordinates of the event analysis sub-unit. Each instance establishes a dynamic channel connection with the corresponding event analysis sub-unit through a dedicated data link. In the field model interaction engine built inside the field generator, the optimization algorithm with the multi-objective function of minimizing the support structure weight, maximizing the safety factor, and optimizing the deformation coordination is implanted. The core of the field model interaction engine is a multi-objective optimization system, which aims to simultaneously achieve the minimization of the support structure weight, the maximization of the safety factor, and the optimization of the deformation coordination. The engine receives the local instability feature vector from the event analysis sub-unit in real time through the dynamic channel as the input data of the optimization process. The sequence quadratic programming algorithm based on sensitivity analysis is used for iterative calculation: the algorithm takes the field strength and direction parameters of the initial interference field as the design variables, calculates the sensitivity of the objective function to the variables in each iteration, constructs and solves the quadratic programming sub-problem to determine the parameter update direction, and then checks the safety constraints and adjusts the field distribution with new data.This process is executed in a loop until the optimal interference field parameters that satisfy the multi-objective balance are converged. The channel monitoring unit monitors the transmission load of all dynamic channels in parallel through packet rate detection and buffer depth monitoring. When the load of a specific area channel continuously exceeds the preset threshold, the channel materialization process is triggered, and the high-load channel is converted into an anchor structure with specific mechanical properties. The system finally integrates the corrected interference field parameters and anchor spatial distribution, solves the potential energy field gradient to develop field streamline tracking, and automatically generates the spatial topology of the flexible support net according to the streamline density. The design inclination of the anchor rod is calculated according to the streamline curvature, the support member specification parameters are determined according to the field strength value, and the phased construction timing is deduced in reverse according to the instability risk propagation path, completing the digital conversion from the continuous field model to the discrete support scheme and generating the design scheme of the flexible support.

[0023] In one embodiment, the step of continuously monitoring the real-time load of all dynamic channels, materializing dynamic channels based on a preset super-fusion threshold, and generating anchors for the flexible support net is: The channel monitoring unit continuously monitors the real-time load of all data channels and sets a preset overload threshold. When it is found that the real-time load of one or more dynamic channels within a preset range continuously exceeds the preset overload threshold within a preset time, it is determined that the area is a high-energy core area. The channel monitoring unit sends a freezing application to the dynamic channels in the high-energy core area, and the dynamic channels that receive the freezing application are materialized to generate anchors and record anchor parameters. The anchor data is transmitted to the field model interaction engine. The anchor parameters include spatial positioning parameters, mechanical property parameters, and dynamic correlation parameters. Further, the channel monitoring unit continuously collects transmission state indicators of each dynamic channel through the probe program deployed in the data link layer. These indicators include but are not limited to data packet transmission rate, bit error rate and buffer queue depth. The system preset load threshold is a multi-dimensional dynamic threshold value, the value of which is determined by the channel base bandwidth and the real-time transmitted data type. When the monitoring unit detects that more than half of the dynamic channels in a certain geographical area continuously exceed the upper limit of the threshold value for three sampling periods, the area will be marked as a high-energy core area. At this time, the monitoring unit will send a freezing instruction frame containing a time stamp to the target channel, and the channel receiving the instruction will immediately start the materialization process: first, it stops the normal transmission of data flow, and then converts the channel state parameters at the current time into the mechanical properties of anchor nails. The spatial positioning parameters are determined by analyzing the three-dimensional coordinates of the channel terminal device, and the mechanical performance parameters are converted into tensile strength and anchoring force according to the historical load peak value of the channel. The dynamic correlation parameters are inherited from the original topological connection relationship of the channel. These quantified anchor nail parameter sets are packaged as structured data packets and transmitted in real time to the field model interaction engine through a special interface, providing key spatial anchoring point data for dynamic optimization of flexible support schemes.

[0024] In one embodiment, the visualization module comprises: a model rendering unit for generating a three-dimensional visualization scene based on the instability model and the support interference field, and rendering the geometric shape, instability area and support structure of the slope in real time; a data superimposition unit for superimposing and displaying dynamic data, stress analysis report and optimization scheme in the form of layers; a warning display unit for dynamically marking high-risk areas based on the instability model and highlighting warning information in the form of color or animation; Further, the visualization module constructs a three-dimensional scene using a physics-based rendering engine, which generates a slope visualization model with real lighting and material properties by fusing the triangular mesh data of the instability model with the vector field data of the support interference field. The model rendering unit dynamically adjusts the model accuracy at different viewing distances using level-of-detail techniques, and implements semi-transparent rendering of the instability area based on depth information through shader programming, and parameterized instantiation rendering of the support structure. The data overlay unit uses a multi-layer management architecture, aligns the real-time sensor data with the contour maps in the stress analysis report using a spatial registration algorithm, displays dynamic data changes using an interactive heat map layer, overlays the mechanical parameter labels of the support members using a vector layer, and realizes visual comparison of different optimization schemes through a time axis controller. The early warning display unit analyzes the safety factor field data in the instability model in real time, uses an elevation gradient-based color mapping technique to layer the risk areas, automatically triggers a pulse warning animation when the safety factor falls below the critical threshold, simulates the propagation path of the instability trend using a particle system, and generates a dynamic sign containing a risk level text description. All early warning elements support real-time interaction and perspective focusing operations in three-dimensional space.

[0025] The above merely provides a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A flexible slope protection optimization design system, characterized in that, Data acquisition module: Collects dynamic and image data of the slope based on multi-array sensors deployed on the slope, and transmits the dynamic data to a dynamic relational database; Instability Model Construction Module: Connected to the data acquisition module, it is used to analyze dynamic data to generate an instability model of the slope and generate a stress analysis report of the slope based on the instability model; Support Interference Field Module: Connected to the instability model construction module, it generates a support interference field based on the stress analysis report, and generates an optimized scheme for flexible slope support based on the interaction between the support interference field and the instability model; Visualization module: Connected to the support interference field module, it is used to visualize the instability model and the optimization scheme of flexible slope support.

2. The flexible slope support optimization design system according to claim 1, characterized in that, The data acquisition module includes: Multi-array sensing units are distributed on and inside the slope surface to collect dynamic data of the slope and generate multi-source time-series dynamic data by annotating the dynamic data with timestamps and three-dimensional geographic coordinates. The environmental monitoring unit is used to synchronously collect environmental status data of the environment where the slope is located. Dynamic data and environmental status data are transferred to a dynamic relational database to generate event data clusters.

3. The flexible slope support optimization design system according to claim 2, characterized in that, The steps for transmitting dynamic data to a dynamic relational database to generate event data clusters are as follows: Dynamic relational databases include real-time data caching units, environment modulation units, and dynamic relation construction units; The real-time data caching unit constructs multiple sub-databases to receive multi-source time-series dynamic data in real time and store it in the cache space; The environment modulation unit is used to receive environmental state data and convert the environmental state data into an environmental state vector. The dynamic relationship construction unit includes a feature rule base, a relationship calculation subunit, and a structured event encapsulation unit; The feature rule base contains pre-stored feature extraction operators based on the principles of rock and soil mechanics and spatiotemporal correlation rules; The relation computation subunit continuously calls the feature extraction operator to perform parallel scanning and computation on multi-source time-series dynamic data and environmental state vectors to generate relation tuples; Relational tuples contain multi-source time-series dynamic data, environment state vectors, relation type identifiers, and relation strength quantification values; The structured event encapsulation unit analyzes relation tuples through neural networks and generates event data clusters based on spatiotemporal association rules and relation type identifiers.

4. The flexible slope support optimization design system according to claim 1, characterized in that, The instability model construction module includes: The slope digital model building unit constructs a three-dimensional slope digital model based on dynamic data and image data; The distributed event analysis unit extracts the three-dimensional geographic coordinate distribution characteristics of event data clusters and dynamically generates event analysis sub-units based on the three-dimensional geographic coordinate distribution characteristics. These sub-units are used to receive and analyze event data clusters to generate local instability feature vectors. The instability knowledge base unit pre-stores a set of instability mode rules based on the principles of geotechnical mechanics. The instability mode rule set is used to define the mapping relationship between different combinations of local instability feature vectors and potential instability modes. The node collaborative reasoning unit is connected to each event analysis subunit and the instability knowledge base unit to perform collaborative reasoning on local instability feature vectors and synthesize an instability model describing the stability state of the slope. A stress analysis report of the slope surface is generated based on the instability model.

5. The flexible slope support optimization design system according to claim 4, characterized in that, The steps for connecting with each event analysis subunit and the instability knowledge base unit to perform collaborative reasoning on local instability feature vectors and synthesize an instability model describing the slope stability state are as follows: The node collaborative reasoning unit receives local instability feature vectors from each event analysis subunit and performs spatiotemporal alignment processing on all local instability feature vectors. The aggregated set of local instability feature vectors is matched with the set of instability mode rules in the instability knowledge base unit. By calculating the matching degree between the feature vectors and the set of instability mode rules, the corresponding potential instability modes are activated. Based on the three-dimensional geographic coordinates in the local instability feature vector, the correlation strength between local instability features at different locations is analyzed. Using each event analysis sub-unit as a node and the correlation strength as an edge, a correlation network representing the instability propagation path is constructed. By combining activated potential instability modes and related networks, a slope instability model is synthesized using a weighted fusion algorithm; The stress analysis report of the slope is generated based on the instability model. The stress analysis report includes the geometric parameters of the potential slip surface, the distribution of the safety factor, the stress-strain field data, the distribution of support resistance requirements, and the dynamic evolution prediction.

6. The flexible slope support optimization design system according to claim 1, characterized in that, The support interference field module includes: The field mapping unit is used to receive the stress analysis report and map the support resistance demand distribution, potential slip surface geometric parameters and stress-strain field data in the report into an initial interference field overlaid on the instability model. The dynamic field generator generates a dynamic field based on the corresponding position of the three-dimensional geographic coordinates of the event analysis subunit on the initial interferometric field, and is connected to the event analysis subunit through a dynamic channel; A field model interaction engine is built in the dynamic field generator, with a built-in optimization objective function. The local instability feature vectors from the event analysis subunit are received through a dynamic channel, and the local instability feature vectors are input into the optimization objective function. The parameters of the initial interference field are dynamically corrected through an iterative adjustment algorithm. The channel monitoring unit continuously monitors the real-time load of all dynamic channels, materializes the dynamic channels based on the preset super-fusion threshold, and generates anchors for the initial interference field. The correction parameters of the initial interference field and the anchor-generated support interference field are obtained. The parameters of the support interference field are discretized into the spatial layout coordinates, stress parameters and construction sequence of the flexible support components through field flow line tracing technology.

7. The flexible slope support optimization design system according to claim 6, characterized in that, The steps of continuously monitoring the real-time load of all dynamic channels, materializing the dynamic channels based on a preset over-fusion threshold, and generating anchors for the flexible support network are as follows: The channel monitoring unit continuously monitors the real-time load of all data channels and presets overload thresholds; If it is found that the real-time load of one or more dynamic channels within a preset range continuously exceeds the preset overload threshold within a preset time, then the area is determined to be a high-energy core area. The channel monitoring unit sends a freeze request to the dynamic channel in the high-energy core area. Upon receiving the freeze request, the dynamic channel materializes the anchor and records the anchor parameters, and transmits the anchor data to the field model interaction engine. The anchor parameters include spatial positioning parameters, mechanical performance parameters, and dynamic correlation parameters.

8. The flexible slope support optimization design system according to claim 1, characterized in that, The visualization module includes: The model rendering unit is used to generate a three-dimensional visualization scene based on the instability model and the support interference field, and to render the geometry, instability area and support structure of the slope in real time. The data overlay unit is used to display dynamic data, stress analysis reports, and optimization schemes in an overlay format. The early warning display unit dynamically marks high-risk areas based on the instability model and highlights the early warning information in the form of color or animation.

Images