Intelligent decision support system for rainfall-related tunnel deformation dynamic regulation and control

Abstract

The invention discloses an intelligent decision support system for rainfall relevance tunnel deformation dynamic regulation and control, particularly relates to the technical field of tunnel safety monitoring and regulation and control, and is used for solving the problem of local deformation prediction deviation caused by rainfall uniform distribution assumption in the prior art. Rainfall space distribution data and surrounding rock deformation monitoring data of a tunnel coverage area are acquired in real time through a data acquisition module, rainfall time modes are classified and dynamic weight coefficients are allocated by utilizing a classification weighting module, and a high-resolution space grid mapping relation between rainfall capacity and surrounding rock deformation is established in combination with a grid mapping module; furthermore, seepage guidance parameters and deformation diffusion risk parameters of each grid unit are quantified through a seepage analysis module, a health assessment module corrects a grading result based on a historical deformation mode reference library, and finally, a regulation and control generation module outputs a result matched with a risk grade, so that the tunnel safety management and control capability in a complex rainfall scene is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel safety monitoring and control, and more specifically, to an intelligent decision-making support system for dynamic regulation of rainfall-related tunnel deformation. Background Art

[0002] In the field of tunnel engineering safety monitoring, rainfall will affect the deformation of surrounding rock. In the prior art, the correlation between rainfall and tunnel deformation is based on the assumption of spatial uniformity of meteorological data, that is, it is defaulted that rainfall is evenly distributed within the tunnel coverage area, and the overall deformation trend is predicted by simplifying the hydro-mechanical model. For example, traditional monitoring systems usually adopt a fixed-threshold alarm mechanism, rely on the layout of limited sensors to obtain averaged data, and evaluate the tunnel stability by combining static geological parameters. Such methods have certain practicability in conventional rainfall scenarios, but ignore the strong heterogeneity of rainfall spatial distribution in actual engineering (such as local heavy rain, rainfall gradient differences caused by topography), resulting in a fundamental deviation in the monitoring and regulation logic.

[0003] There is a problem in the prior art that the local deformation prediction under non-uniform rainfall distribution is disconnected from the global tunnel safety assessment. That is, due to the significant spatial differences in rainfall in the axial, radial directions of the tunnel and geological weak areas (such as karst caves, fault zones), the traditional uniform distribution assumption cannot accurately describe the seepage and stress coupling effects of local surrounding rock, resulting in a serious deviation between the predicted deformation value and the actual mechanical response. For example, in the strong rainfall area of the tunnel vault, the pore water pressure may rise suddenly due to rapid seepage, while the stress state of the surrounding rock in the adjacent dry area is relatively stable. If this local deformation difference is not effectively identified, it will lead to insufficient pertinence of the support strategy and even the risk of local collapse. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides an intelligent decision-making support system for dynamic regulation of rainfall-related tunnel deformation to solve the problems proposed in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] An intelligent decision-making support system for dynamic regulation of rainfall-related tunnel deformation, comprising the following modules:

[0007] A data acquisition module for collecting rainfall spatial distribution data in the tunnel coverage area and surrounding rock deformation monitoring data in the corresponding area;

[0008] A classification and weighting module for classifying rainfall time patterns and assigning dynamic weight coefficients to different rainfall time patterns based on historical deformation data;

[0009] A grid mapping module, which is used to establish a spatial grid mapping relationship between the rainfall amount and the surrounding rock deformation amount in the tunnel coverage area based on the rainfall spatial distribution data, dynamic weight coefficients, and surrounding rock deformation monitoring data;

[0010] A seepage analysis module, which is used to perform seepage guidance analysis and deformation diffusion risk coupling analysis on each grid unit according to the spatial grid mapping relationship, and generate seepage guidance parameters and deformation diffusion risk parameters;

[0011] A health assessment module, which is used to perform a health status grading assessment on the grid unit based on the seepage guidance parameters and deformation diffusion risk parameters, and correct the health status grading assessment results in combination with a preset deformation mode reference library;

[0012] A regulation generation module, which is used to generate dynamic regulation instructions according to the corrected health status grading assessment results.

[0013] In a preferred embodiment, the rainfall spatial distribution data and the surrounding rock deformation monitoring data of the tunnel coverage area are collected, including:

[0014] Rain gauges are arranged along the axial direction of the tunnel at a preset interval, and the arrangement density is adjusted according to the geological structure risk level along the radial direction of the tunnel;

[0015] Displacement gauges are arranged at the arch bolt nodes, side wall joints, and invert monitoring points of the tunnel support structure, and the axial distance between adjacent displacement gauges is determined according to the tunnel diameter ratio;

[0016] The rainfall spatial distribution data is collected in real time by rain gauges and sensor noise is removed, and the surrounding rock deformation monitoring data is collected by displacement gauges and preprocessed by moving average filtering.

[0017] In a preferred embodiment, the rainfall time patterns are classified, and dynamic weight coefficients are assigned to different rainfall time patterns based on historical deformation data, including:

[0018] The rainfall time patterns are divided into continuous rainfall, intermittent rainfall, and burst rainfall events;

[0019] Based on historical deformation data, the mean and variance of the surrounding rock deformation rate corresponding to each rainfall time pattern are extracted. The weight coefficient of continuous rainfall is the ratio of the mean and variance of the deformation rate, and the weight coefficients of intermittent rainfall and burst rainfall are dynamically adjusted according to the difference between the mean deformation rate and the historical maximum deformation rate.

[0020] In a preferred embodiment, the determination condition for continuous rainfall is that the rainfall duration exceeds a preset threshold and the interruption interval is less than a set duration.

[0021] In a preferred embodiment, based on rainfall spatial distribution data, dynamic weight coefficients, and surrounding rock deformation monitoring data, a spatial grid mapping relationship between rainfall and surrounding rock deformation in the tunnel-covered area is established, including:

[0022] Taking the tunnel axis as the reference, grid cells are divided. The axial division density of the grid cells is dynamically adjusted according to the geological structure risk level, and the axial division density of the grid cells in the high-risk area is higher than that in the low-risk area;

[0023] The rainfall data of each grid cell is calculated by spatial interpolation through rain gauges covering its monitoring range, and the dynamic weight coefficient corresponding to the rainfall time pattern is superimposed during the interpolation process to correct the rainfall contribution degree;

[0024] The surrounding rock deformation data is collected by displacement meters and associated with the corresponding grid cells. Combining the geological permeability coefficient and the average deformation rate, a non-linear coupling analysis is performed on the correlation between rainfall and surrounding rock deformation to generate the deformation response coefficient of each grid cell;

[0025] Based on the deformation response coefficient and rainfall data, a spatial grid mapping relationship is constructed, and the weight distribution of the spatial grid mapping relationship is dynamically adjusted according to the geological permeability coefficient and the variance of the deformation rate.

[0026] In a preferred embodiment, based on the spatial grid mapping relationship, seepage guidance analysis and deformation diffusion risk coupling analysis are performed on each grid cell to generate seepage guidance parameters and deformation diffusion risk parameters, including:

[0027] The seepage guidance parameter is calculated based on the product of the rainfall gradient and the geological permeability coefficient of the grid cell. The rainfall gradient is the difference in rainfall between adjacent grid cells divided by the corresponding spacing;

[0028] The deformation diffusion risk parameter is calculated according to the covariance of the deformation rate of the current grid cell and the stress state of adjacent grid cells. The stress state is quantified by the change rate of the surrounding rock displacement monitored by displacement meters;

[0029] The seepage guidance parameter and the deformation diffusion risk parameter are superimposed.

[0030] In a preferred embodiment, based on the seepage guidance parameter and the deformation diffusion risk parameter, a health status grading evaluation is performed on the grid cells, and the health status grading evaluation result is corrected by combining a preset deformation mode reference library, including:

[0031] According to the comparison relationship between the seepage guidance parameter and the pore water pressure threshold, and the comparison relationship between the deformation diffusion risk parameter and the deformation rate threshold, a preliminary grading result of the health status is generated;

[0032] Based on a preset deformation mode reference library, calculate the similarity distance between the current deformation data and various deformation modes in the reference library to generate a mode deviation degree.

[0033] Modify the preliminary classification result according to the mode deviation degree.

[0034] In a preferred embodiment, the preliminary classification result includes a stable state, a risk warning state, and an emergency response state.

[0035] The correction rule for modifying the preliminary classification result is: when the mode deviation degree exceeds the deviation threshold, upgrade the stable state to the risk warning state, or upgrade the risk warning state to the emergency response state.

[0036] In a preferred embodiment, generate dynamic control instructions according to the corrected health state classification evaluation result, including:

[0037] Based on the emergency response state grid cells in the health state classification result, generate local support reinforcement instructions and drainage system startup instructions.

[0038] For the risk warning state grid cells, generate classification control instructions according to the weighted score ranking of the deformation rate variance and the mode deviation degree.

[0039] The triggering timing of the dynamic control instructions is dynamically adjusted according to the timing matching degree between the real-time monitoring data of the pore water pressure and the historical seepage loss control events.

[0040] In a preferred embodiment, the reinforcement position of the support reinforcement instruction is located according to the high-value area of the seepage orientation parameter; the timing matching degree is calculated by the similarity distance between the current pore water pressure curve and the historical event curve.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. By constructing a full-chain analysis system for the spatio-temporal heterogeneity of rainfall and the dynamic response of surrounding rock deformation, the accuracy and reliability of tunnel safety monitoring and control are significantly improved; by dynamically perceiving the spatio-temporal distribution characteristics of rainfall, classifying and weighting the mechanical contribution degrees of different rainfall patterns, and combining high-resolution spatial grid mapping technology, the accurate quantification of local seepage orientation and deformation diffusion risk is realized; the dynamic weight coefficient distribution mechanism can differentially evaluate the cumulative effects of continuous rainfall, intermittent rainfall and other patterns on surrounding rock deformation, and the grid mapping relationship establishes a local mechanical response model by fusing multi-source data (rainfall, geological permeability, real-time deformation), effectively overcoming the problem of insufficient adaptability of traditional methods to non-uniform rainfall scenarios; this enables the system to identify in advance the seepage loss control risks in geological weak parts such as fault zones and karst cave areas, providing a reliable basis for hierarchical early warning.

[0043] 2. Through the real-time interactive analysis of seepage and deformation coupling parameters, a closed-loop control link from risk perception to dynamic decision-making is formed; based on the multi-dimensional evaluation of seepage path strength, deformation chain effect, and historical pattern deviation, control instructions that strictly match the local risk level are generated; in areas with high seepage directivity, directional drainage and bolt reinforcement are automatically triggered, while in areas with high risk of deformation diffusion, grouting reinforcement and secondary lifting are initiated; this hierarchical response mechanism can not only avoid waste of resources caused by over-support, but also accurately inhibit the spread of local collapse risks, especially applicable to tunnel engineering scenarios with complex geological conditions and significant rainfall variations. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic structural diagram of the intelligent decision-making support system for rainfall-related tunnel deformation dynamic control of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0046] Embodiment: Figure 1 A schematic structural diagram of the intelligent decision-making support system for rainfall-related tunnel deformation dynamic control of the present invention is given. The intelligent decision-making support system for rainfall-related tunnel deformation dynamic control includes the following modules:

[0047] A data acquisition module for collecting rainfall spatial distribution data in the tunnel-covered area and surrounding rock deformation monitoring data in the corresponding area;

[0048] A classification and weighting module for classifying rainfall time patterns and assigning dynamic weight coefficients to different rainfall time patterns based on historical deformation data;

[0049] A grid mapping module for establishing a spatial grid mapping relationship between rainfall amount and surrounding rock deformation amount in the tunnel-covered area based on rainfall spatial distribution data, dynamic weight coefficients, and surrounding rock deformation monitoring data;

[0050] A seepage analysis module for performing seepage directivity analysis and deformation diffusion risk coupling analysis on each grid unit according to the spatial grid mapping relationship, and generating seepage directivity parameters and deformation diffusion risk parameters;

[0051] A health assessment module for performing a health status classification assessment on the grid unit based on the seepage directivity parameters and deformation diffusion risk parameters, and correcting the health status classification assessment result in combination with a preset deformation mode reference library;

[0052] A regulation generation module, configured to generate a dynamic regulation instruction according to the corrected hierarchical evaluation result of the health status.

[0053] Collect the rainfall spatial distribution data of the tunnel-covered area and the surrounding rock deformation monitoring data of the corresponding area. The specific implementation is as follows:

[0054] Rain gauges are arranged along the axial direction of the tunnel at a preset interval. The determination method of the preset interval is as follows: divide the hydrological sensitive areas according to the tunnel engineering geological exploration report. The determination criteria for hydrological sensitive areas are that the historical maximum daily rainfall exceeds 50 mm or the rock mass permeability coefficient is greater than 1×10 -5 cm / s; within the hydrological sensitive area, the axial interval of the rain gauges is set to 10 m, and in the non-sensitive area, it is set to 20 m; when arranging the rain gauges along the radial direction of the tunnel, adjust the arrangement density according to the geological structure risk level, and the geological structure risk level is comprehensively determined by the fracture density, core permeability coefficient and historical water inflow in the geological exploration data; for example, the area where the fracture density is greater than 3 fractures / m and the permeability coefficient is greater than 1×10 -4 cm / s is defined as the high risk level, and the radial arrangement density is 3 rain gauges per section, and the arrangement density in the low risk area is 1 rain gauge per section.

[0055] Displacement gauges are arranged at the arch bolt nodes, side wall joints and invert monitoring points of the tunnel support structure. The fixed end installation position of the arch support bolts at the arch bolt nodes, and the side wall joints are the stress concentration points where the initial support steel frame contacts the surrounding rock. The invert monitoring points are located at the deformation sensitive positions on both sides of the invert concrete pouring joint; the axial interval between adjacent displacement gauges is determined according to the proportional relationship of the tunnel diameter. The calculation method of the proportional relationship is as follows: when the tunnel diameter is less than 10 m, the axial interval is 1.5 times the diameter; when the diameter is greater than or equal to 10 m, the axial interval is 1.2 times the diameter; for example, when the tunnel diameter is 8 m, the axial interval of the displacement gauges is 12 m; the displacement gauges are fixed on the surface of the support structure by expansion bolts, and the wiring terminals are encapsulated with waterproof sealant.

[0056] The rainfall spatial distribution data is collected in real time by the rain gauges, and the sensor noise is eliminated during the data transmission process. The method for eliminating the sensor noise is as follows: set the effective range of the single-shot collected data of the rain gauge to be 0.1 mm to 50 mm. When the single-shot collected data exceeds this range, it is determined as noise data and discarded; for example, if the collected data in a certain time is 52 mm, the system automatically marks it as invalid data; after the surrounding rock deformation monitoring data is collected by the displacement gauges, it is preprocessed by moving average filtering. The window size of the moving average filtering is dynamically adjusted according to the data collection frequency. The specific rule is as follows: when the data collection frequency is 1 time per minute, the window size is set to 5 data points, that is, the filtered displacement average data is output every 5 minutes; when the collection frequency is adjusted to 1 time per 30 seconds, the window size is synchronously adjusted to 10 data points.

[0057] The layout density of the rain gauges strictly corresponds to the grid division rules of the spatial grid mapping relationship in the subsequent steps. The surrounding rock deformation monitoring data after moving average filtering is used as the basic input data for the historical deformation data analysis in the classification weighting module to ensure the stability of the dynamic weight coefficient calculation.

[0058] The determination basis of the geological structure risk level includes: the fissure distribution map in the geological exploration report, the test results of the permeability coefficient of the core sampling, and the abnormal sections recorded in the advanced geological prediction during the construction period; for example, a certain tunnel fault zone area is marked as a high-risk level according to the geological exploration report, and the layout density of its radial rain gauges is increased to 3 per section, and the axial spacing is shortened to 10 meters; the layout position of the displacement gauges at the arch top bolt nodes is determined according to the bolt layout diagram in the support design drawing. For example, displacement gauges are arranged at the arch top bolt nodes corresponding to each steel frame, and the spacing between adjacent steel frames is 1 meter.

[0059] In the sensor noise rejection method, the effective range threshold is dynamically adjusted according to the regional historical meteorological data. For example, in areas with frequent heavy rains, the upper limit threshold of the single rainfall amount is increased to 100 mm; the window size of the moving average filtering is dynamically adapted according to the tunnel deformation rate. For example, when the real-time monitored deformation rate exceeds 0.1 mm / h, the window size is automatically adjusted to 3 data points to improve the response speed.

[0060] Classify the rainfall time patterns and assign dynamic weight coefficients to different rainfall time patterns based on the historical deformation data. The specific implementation is as follows:

[0061] The rainfall time patterns are divided into continuous rainfall, intermittent rainfall and burst rainfall events. The determination condition for continuous rainfall is that the rainfall duration exceeds the preset threshold and the interruption interval is less than the set duration; the determination method of the preset threshold is: statistically analyze the duration data of all rainfall events in the target area in the past five years, and select the duration value corresponding to the cumulative frequency of 90% as the threshold. For example, the statistical results show that 90% of the rainfall events have a duration of no more than 10 hours, so the preset threshold is set to 10 hours; the set duration of the interruption interval is determined according to the tunnel engineering grade. For first-level high-risk tunnels, it is set to 0.5 hours, and for second-level ordinary tunnels, it is set to 1 hour; for example, the interruption interval of a certain high-risk tunnel is set to 0.5 hours. If the rainfall interruption time exceeds 0.5 hours, it is determined as an intermittent rainfall event.

[0062] Extract the mean and variance of the surrounding rock deformation rate corresponding to each rainfall time pattern based on historical deformation data, where the historical deformation data is the displacement meter monitoring data after preprocessing, and the preprocessing methods include moving average filtering and outlier removal; the mean calculation method of the surrounding rock deformation rate is: for all historical data under the same rainfall time pattern, calculate the arithmetic mean of the deformation rate within the corresponding rainfall duration; the variance calculation method is: the average of the squared deviations of the deformation rate from the mean under the same rainfall time pattern; for example, under the continuous rainfall pattern, the historical deformation rate mean is 0.4 mm / hour and the variance is 0.015; under the intermittent rainfall pattern, the mean is 0.3 mm / hour and the variance is 0.02.

[0063] The weight coefficient of continuous rainfall is the ratio of the mean and variance of the deformation rate. For example, the weight coefficient of continuous rainfall is calculated as 0.4 / 0.015 ≈ 26.7; the weight coefficients of intermittent rainfall and burst rainfall are dynamically adjusted according to the difference between the mean deformation rate and the historical maximum deformation rate, where the historical maximum deformation rate is the peak deformation rate of the historical monitoring data under the same rainfall time pattern; for example, the historical maximum deformation rate of burst rainfall is 2.0 mm / hour, and the current mean deformation rate is 1.8 mm / hour, then the difference is 0.2 mm / hour, and the weight coefficient is adjusted to the ratio of the difference to the historical maximum deformation rate, that is, 0.2 / 2.0 = 0.1.

[0064] The classification results of the rainfall time pattern and the calculation results of the weight coefficient are transmitted to the subsequent steps to correct the contribution difference of different rainfall patterns to local deformation in the spatial grid mapping relationship; for example, the high weight coefficient of continuous rainfall will increase its weight in the calculation of the rainfall amount in the grid cell, while the low weight coefficient of burst rainfall will suppress the abnormal influence of short-term heavy rainfall.

[0065] The basis for the engineering grade division of the preset threshold and the interruption interval is the geological risk grade standard in the tunnel design specification. For example, a tunnel with a fault zone or a highly permeable formation is defined as a first-level high-risk tunnel; the statistical period of the historical maximum deformation rate is the same as the tunnel operation period. For example, for a tunnel in operation for more than three years, the historical maximum value is updated with the monitoring data of the past three years.

[0066] Based on the rainfall spatial distribution data, dynamic weight coefficients, and surrounding rock deformation monitoring data, establish the spatial grid mapping relationship between the rainfall amount and the surrounding rock deformation amount in the tunnel coverage area. The specific implementation is as follows:

[0067] Divide the grid cells based on the tunnel axis. The axial division density of the grid cells is dynamically adjusted according to the geological structure risk grade, and the geological structure risk grade is comprehensively determined based on the fracture distribution map in the geological exploration report, the test results of the core permeability coefficient, and the advanced geological prediction data during the construction period; for example, the fracture density is greater than 3 fractures per meter and the permeability coefficient is greater than 1×10-5 The area with a velocity of cm / s is defined as the high-risk level, and the axial division density of its grid cells is one cell per 10 meters, while the low-risk area is divided into one cell per 20 meters; the adjustment rule for the division density is: the number of grid cells in the high-risk area is twice that of the low-risk area to ensure the matching of data acquisition accuracy and geological risks.

[0068] The rainfall data of each grid cell is calculated by spatial interpolation through rain gauges covering its monitoring range. The spatial interpolation calculation uses the inverse distance weighting method to calculate the rainfall weights at each point within the grid cell based on the spatial positions of the rain gauges; during the interpolation process, the dynamic weight coefficient corresponding to the rainfall time pattern is superimposed to correct the rainfall contribution degree. The dynamic weight coefficient is derived from the output results calculated based on the mean and variance of historical deformation rates in the classification weighting step; for example, the weight coefficient for continuous rainfall is 25, and during the interpolation calculation, the rainfall data of the corresponding grid cell is multiplied by 25 and then participates in the spatial average calculation, while the weight coefficient for burst rainfall is 0.1, and the rainfall data is multiplied by 0.1 and then participates in the calculation.

[0069] The surrounding rock deformation data is collected by displacement gauges and associated with the corresponding grid cells. The association method is: the deformation data of the displacement gauge monitoring points is mapped to the center point of the grid cell to which it belongs through spatial interpolation; a non-linear coupling analysis is performed on the correlation between rainfall and surrounding rock deformation amount by combining the geological permeability coefficient and the mean deformation rate. The specific method of non-linear coupling analysis is: for each grid cell, calculate the product of the geological permeability coefficient and the mean deformation rate, and then perform a ratio operation with the rainfall data to generate a deformation response coefficient; for example, for a grid cell with a geological permeability coefficient of 2×10 -5 cm / s, a deformation rate of 0.5 mm / h, and a rainfall of 50 mm, the deformation response coefficient is calculated as (2×10 -5 ×0.5) / 50 = 2×10 -7 ; the geological permeability coefficient is derived from the core permeability test data in the geological exploration report, and the mean deformation rate is obtained by calculating the historical displacement gauge monitoring data.

[0070] A spatial grid mapping relationship is constructed based on the deformation response coefficient and the rainfall data. The weight distribution of the spatial grid mapping relationship is dynamically adjusted according to the geological permeability coefficient and the deformation rate variance. The specific adjustment rule is: the higher the geological permeability coefficient and the larger the deformation rate variance of the grid cell, the higher the weight distribution ratio; for example, for a grid cell with a geological permeability coefficient of 1×10 -4 cm / s and a variance of 0.03, the weight distribution ratio is 0.8, while for a permeability coefficient of 5×10 -6Cells with a velocity of cm / s and a variance of 0.01 have a weight ratio of 0.2; the weight distribution ratio is used for risk level determination in subsequent health assessment steps. For example, grid cells with a high weight ratio will trigger support reinforcement instructions preferentially.

[0071] The product operation in the non - linear coupling analysis is used to quantify the interaction effect between geological conditions and deformation rate. For example, the rainfall in areas with a high permeability coefficient has a significant impact on deformation, and the product result amplifies the deformation response coefficient; in spatial interpolation calculations, the spatial distribution density of rain gauges matches the grid division density. For example, a 10 - meter grid cell corresponds to a rain gauge installed every 10 meters, and a 20 - meter grid cell in a low - risk area corresponds to a rain gauge installed every 20 meters.

[0072] Perform seepage orientation analysis and deformation diffusion risk coupling analysis on each grid cell according to the spatial grid mapping relationship, and generate seepage orientation parameters and deformation diffusion risk parameters. The specific implementation is as follows:

[0073] Calculate the seepage orientation parameter based on the product of the rainfall gradient and the geological permeability coefficient of the grid cell. The rainfall gradient is the difference in rainfall between adjacent grid cells divided by the corresponding spacing, and the spacing is the axial distance between the center points of adjacent grid cells. For example, if the rainfall in two adjacent grid cells is 50 mm and 30 mm respectively, and the spacing is 10 m, then the rainfall gradient is calculated as (50 - 30) / 10 = 2 mm / m; the geological permeability coefficient is obtained from the core permeability test data in the geological exploration report. For example, the geological permeability coefficient of a certain grid cell is 1×10 -5 cm / s; the seepage orientation parameter is calculated as the product of the rainfall gradient and the geological permeability coefficient, that is, 2 mm / m×1×10 -5 cm / s = 2×10 -5 cm / s·mm / m.

[0074] Calculate the deformation diffusion risk parameter according to the covariance between the deformation rate of the current grid cell and the stress state of adjacent grid cells. The deformation rate is obtained by calculating the change rate of surrounding rock displacement collected by displacement gauges, and the stress state is quantified by the change rate of surrounding rock displacement of adjacent grid cells. For example, the deformation rate of the current grid cell is 0.5 mm / h, and the deformation rate of an adjacent grid cell is 0.3 mm / h. The covariance is calculated as a statistic of the deviation of the product of the two deformation rates from the mean; the covariance calculation result is used to characterize the intensity of the chain risk of deformation diffusion. For example, when the covariance value is 0.05, the deformation diffusion risk parameter is marked as a medium - risk level.

[0075] Superimpose the seepage orientation parameter and the deformation diffusion risk parameter. The superimposition method is linear weighted summation, where the weight coefficient of the seepage orientation parameter is 0.6 and the weight coefficient of the deformation diffusion risk parameter is 0.4; for example, the seepage orientation parameter is 2×10-5 cm / s·mm / m, and the deformation diffusion risk parameter is 0.05, then the superposition result is 2×10 -5 ×0.6 + 0.05×0.4 = 0.020012; the superposition result is input into the hierarchical evaluation step to correct the chain risk level of local seepage loss control and deformation diffusion. For example, when the superposition result exceeds the threshold of 0.02, it is determined as a high risk level and a drainage system start instruction is triggered.

[0076] In the calculation process of the seepage directivity parameter, the data of the rainfall gradient and the geological permeability coefficient are from the output result of the spatial grid mapping relationship establishment step; in the calculation of the deformation diffusion risk parameter, the stress state data of adjacent grid cells are obtained through the spatial interpolation result of the displacement meter. For example, when there is no displacement meter arranged in adjacent grid cells, its stress state is estimated by the inverse distance weighted method.

[0077] The superposition weight coefficient of the seepage directivity parameter and the deformation diffusion risk parameter is dynamically adjusted according to the engineering geological conditions. For example, in the area where the rock mass permeability coefficient is greater than 1×10 -4 cm / s, the weight coefficient of the seepage directivity parameter is increased to 0.8; the setting of the superposition threshold is based on the statistical result of seepage loss control events in historical monitoring data. For example, if the average value of the superposition result when seepage loss control events occurred in historical data is 0.025, the threshold is set to 0.02 to reserve a safety margin.

[0078] In the covariance calculation, the correlation between the deformation rate and the stress state is realized through the time series analysis of the monitoring data. For example, analyze the synchronous fluctuation degree of the deformation rate of adjacent grid cells. The higher the synchronous fluctuation frequency, the larger the covariance value; the calculation result of the seepage directivity parameter is used to identify the preferential seepage path. For example, the grid cells with a seepage directivity parameter greater than 1×10 -5 cm / s·mm / m are marked as potential seepage preferential channels.

[0079] Based on the seepage directivity parameter and the deformation diffusion risk parameter, the health state of grid cells is hierarchically evaluated, and the health state hierarchical evaluation result is corrected in combination with the preset deformation mode reference library. The specific implementation is as follows:

[0080] Based on the comparison relationship between the seepage guidance parameter and the pore water pressure threshold, and the comparison relationship between the deformation diffusion risk parameter and the deformation rate threshold, a preliminary classification result of the health status is generated. The pore water pressure threshold is determined according to the pore water pressure distribution under the stable state of the surrounding rock in historical monitoring data. For example, the pore water pressure values in the stable state within three years are statistically analyzed, and the pressure value corresponding to the cumulative probability of 95% is taken as the threshold. The deformation rate threshold is determined according to the mean and variance of the deformation rate during the historical safe operation period. For example, the threshold corresponding to the mean deformation rate of 0.3 mm / h is 0.5 mm / h. When the seepage guidance parameter exceeds the pore water pressure threshold and the deformation diffusion risk parameter exceeds the deformation rate threshold, it is determined as the emergency response state. When only one parameter exceeds the limit, it is determined as the risk warning state. When neither parameter exceeds the limit, it is determined as the stable state.

[0081] Based on the preset deformation mode reference library, calculate the similarity distance between the current deformation data and various deformation modes in the reference library to generate the mode deviation degree. The preset deformation mode reference library is constructed through the clustering analysis of historical surrounding rock deformation data, including the normal deformation mode, the local collapse mode, and the creep mode. The similarity distance is calculated by the dynamic time warping algorithm. After aligning the time series of the current deformation curve and the curve of each mode in the reference library, the difference value is calculated. For example, the difference value between the current deformation curve and the local collapse mode is 0.12, and the difference value between the current deformation curve and the normal mode is 0.05, then the mode deviation degree is marked as 0.12.

[0082] Modify the preliminary classification result according to the mode deviation degree. The modification rule is: when the mode deviation degree exceeds the deviation threshold, upgrade the stable state to the risk warning state, or upgrade the risk warning state to the emergency response state. The deviation threshold is determined according to the statistical distribution of the mode deviation degree in historical abnormal events. For example, the mean value of the mode deviation degree when the seepage out-of-control event occurs is statistically analyzed as 0.1, then the threshold is set to 0.08 for early warning. For example, the preliminary classification of a grid unit is the stable state, but the mode deviation degree is 0.12 (exceeding the threshold of 0.08), then it is modified to the risk warning state. If the preliminary classification is the risk warning state and the mode deviation degree exceeds 0.15, then it is modified to the emergency response state.

[0083] During the determination process of the pore water pressure threshold, the abnormal values in the construction disturbance period need to be excluded from the historical monitoring data. For example, the pore water pressure fluctuation data during the tunnel excavation stage are not involved in the statistics. The dynamic adjustment rule of the deformation rate threshold is: according to the extension of the tunnel operation stage, recalculate the mean and variance based on the newly added monitoring data every six months. For example, after one year of operation, the mean deformation rate is updated to 0.28 mm / h, then the threshold is synchronously adjusted to 0.48 mm / h.

[0084] In the dynamic time warping algorithm, the specific method of time series alignment is as follows: The current deformed curve and the reference mode curve are stretched or compressed and matched along the minimum cumulative distance path, and the difference value is calculated as the sum of the squares of the deformation rate differences of the corresponding points on the matching path. For example, if the length of the current deformed curve is 24 hours and the reference mode curve is 12 hours, the difference value is calculated after stretching and matching.

[0085] The correction result of the mode deviation degree is related to the spatial distribution of the seepage guiding parameter and the deformation diffusion risk parameter. For example, if a grid cell with a high risk level is located in a high-value area of the seepage guiding parameter at the same time, a synchronous drainage and support instruction is triggered; the update period of the preset deformation mode reference library is once a year, and the newly added historical data is iteratively optimized for the reference mode after cluster analysis.

[0086] Generate dynamic control instructions according to the corrected health status grading evaluation results. The specific implementation is as follows:

[0087] Based on the grid cells in the emergency response status in the health status grading results, generate local support reinforcement instructions and drainage system startup instructions. The reinforcement position of the support reinforcement instructions is located according to the high-value area of the seepage guiding parameter. The determination criterion for the high-value area of the seepage guiding parameter is: grid cells with parameter values greater than the 90th percentile in historical statistics; for example, the 90th percentile of the seepage guiding parameter in historical data is 1.5×10 -5 cm / s·mm / m, and the parameter of a current grid cell is 2×10 -5 cm / s·mm / m, then it is determined as a high-value area; the support reinforcement instructions include the bolt arrangement spacing and the grouting volume. The bolt spacing is dynamically adjusted according to the area of the high-value area. For example, when the area of the area is less than 5 square meters, the spacing is set to 0.5 meters, and when the area is greater than 5 square meters, the spacing is set to 1 meter; the drainage system startup instructions include the drainage pump operating power and the duration. The operating power is set according to the current pore water pressure value. For example, when the pore water pressure exceeds 0.5 MPa, the power is set to full load operation.

[0088] For grid cells in the risk warning state, generate hierarchical control instructions according to the weighted score ranking of the deformation rate variance and the mode deviation degree. In the weighted score, the weight of the deformation rate variance is 0.6, and the weight of the mode deviation degree is 0.4. The weight distribution is based on the statistical results of the contribution degrees of the two types of parameters in historical collapse events; for example, statistics show that the deformation rate variance is the main inducing factor in 80% of the events, so the weight is set to 0.6; the grading includes the monitoring frequency increase instruction and the local grouting instruction. The monitoring frequency increase instruction shortens the data acquisition interval from 1 hour to 15 minutes. The grouting volume of the local grouting instruction is set proportionally according to the deformation rate variance value. For example, when the variance is 0.02, the grouting volume is 50 liters, and when the variance is 0.05, the grouting volume is increased to 100 liters.

[0089] The triggering timing of the dynamic control instruction is dynamically adjusted according to the temporal matching degree between the real-time monitoring data of pore water pressure and historical seepage out-of-control events. The temporal matching degree is calculated by the similarity distance between the current pore water pressure curve and the historical event curve. The calculation method of the similarity distance is: after aligning the current curve and the historical curve along the time axis, calculate the sum of the squares of the pressure differences at the corresponding time points. For example, if the similarity distance between the current curve and the curve of a certain historical seepage out-of-control event is 0.1, which is lower than the threshold of 0.15, it is determined as a high matching degree and the drainage instruction is triggered immediately. The historical seepage out-of-control event curve is derived from the past event data stored in the database, and each curve contains the pore water pressure change sequence in the 24 hours before seepage out-of-control.

[0090] During the positioning process of the high-value area of the seepage guidance parameter, if there are multiple high-value areas in the same tunnel section, the area with the highest seepage guidance parameter and the deformation diffusion risk parameter exceeding the threshold is preferentially selected for support. For example, in a certain tunnel section, the seepage guidance parameter of area A is 2.5×10 -5 cm / s·mm / m and the deformation diffusion risk parameter is 0.08, and the parameters of area B are 2×10 -5 cm / s·mm / m and 0.12 respectively, then area B is preferentially supported.

[0091] After the hierarchical control instruction generated by weighted score ranking is generated, the instructions are aggregated according to the spatial adjacency relationship of the grid cells. For example, the grouting instructions of three adjacent risk warning grid cells are merged into a continuous grouting belt instruction, and the grouting volume is superimposed according to the number of cells. The implementation scope of the monitoring frequency increase instruction is adjusted according to the spatial gradient of the deformation rate variance. For example, in the area where the variance gradient is greater than 0.01 / m, the monitoring scope is expanded to 5 meters upstream and downstream.

[0092] The threshold of the temporal matching degree is set based on the average time interval from the matching degree reaching the standard to seepage out-of-control in historical events. For example, statistics show that on average, seepage out-of-control occurs 6 hours after the matching degree reaches 0.15, so the threshold is set to 0.15 to reserve disposal time. The triggering logic of the dynamic control instruction is: when the matching degree exceeds the threshold and the current pore water pressure curve shows an upward trend, the instruction is triggered immediately; if the matching degree exceeds the threshold but the curve is stable or decreasing, it is re-evaluated after a 1-hour delay.

[0093] In the local support and reinforcement instruction, the arrangement direction of the anchor rods is adjusted according to the spatial vector direction of the seepage guidance parameter. For example, when the seepage guidance parameter shows that the dominant seepage direction is the axial direction, the anchor rods are arranged along the radial direction to block the seepage path. The duration of the drainage system startup instruction is dynamically adjusted according to the real-time rainfall prediction data. For example, when the weather forecast shows that the rainfall in the next 3 hours exceeds 50 mm, the drainage pump runs continuously until the rainfall ends.

[0094] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain a formula that is closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0095] It should be noted that the present invention can be deployed on the device itself to achieve embedded applications, or can also run on a PC or other terminals with a user interface, so as to meet various hardware environments and usage requirements.

[0096] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0097] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and the inventive constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0098] In addition, in each embodiment of the present application, the various functional modules can be integrated in a processing module, or each module can exist physically alone, or two or more modules can be integrated in one module.

[0099] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection can be through some interfaces, and the indirect coupling or communication connection of the device or module can be in an electrical, mechanical or other form.

[0100] As described above, only the specific implementation manners of the present application are provided, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by 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.

[0101] Finally, the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent decision support system for dynamic regulation of tunnel deformation related to rainfall, characterized in that, It includes the following modules: A data acquisition module for acquiring rainfall spatial distribution data of the tunnel covered area and surrounding rock deformation monitoring data of the corresponding area; A classification and weighting module for classifying rainfall time patterns and assigning dynamic weight coefficients to different rainfall time patterns based on historical deformation data; A grid mapping module for establishing a spatial grid mapping relationship between rainfall amount and surrounding rock deformation amount in the tunnel covered area based on rainfall spatial distribution data, dynamic weight coefficients, and surrounding rock deformation monitoring data; A seepage analysis module for performing seepage orientation analysis and deformation diffusion risk coupling analysis on each grid unit according to the spatial grid mapping relationship to generate seepage orientation parameters and deformation diffusion risk parameters; A health assessment module for grading and assessing the health status of grid units based on seepage orientation parameters and deformation diffusion risk parameters, and correcting the health status grading and assessment results in combination with a preset deformation mode reference library; A regulation generation module for generating dynamic regulation instructions according to the corrected health status grading and assessment results.

2. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 1, characterized in that Acquire rainfall spatial distribution data of the tunnel covered area and surrounding rock deformation monitoring data of the corresponding area, including: Rain gauges are arranged along the tunnel axis at a preset interval, and the arrangement density is adjusted according to the geological structure risk level along the tunnel radius; Displacement gauges are arranged at the arch top bolt nodes, side wall joints, and invert monitoring points of the tunnel support structure, and the axial spacing between adjacent displacement gauges is determined according to the tunnel diameter ratio; Rainfall spatial distribution data is collected in real time by rain gauges and sensor noise is removed, and surrounding rock deformation monitoring data is collected by displacement gauges and preprocessed by moving average filtering.

3. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 1, characterized in that Classify rainfall time patterns and assign dynamic weight coefficients to different rainfall time patterns based on historical deformation data, including: Classify rainfall time patterns into continuous rainfall, intermittent rainfall, and burst rainfall events; Extract the mean and variance of the surrounding rock deformation rate corresponding to each rainfall time pattern based on historical deformation data. The weight coefficient of continuous rainfall is the ratio of the mean and variance of the deformation rate. The weight coefficients of intermittent rainfall and burst rainfall are dynamically adjusted according to the difference between the mean deformation rate and the historical maximum deformation rate.

4. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 3, characterized in that, The determination condition for continuous rainfall is that the rainfall duration exceeds a preset threshold and the interruption interval is less than a set duration.

5. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 1, wherein, Based on rainfall spatial distribution data, dynamic weight coefficients, and surrounding rock deformation monitoring data, establish a spatial grid mapping relationship between rainfall amount and surrounding rock deformation amount in the tunnel covered area, including: Divide grid units based on the tunnel axis. The axial division density of grid units is dynamically adjusted according to the geological structure risk level. The axial division density of grid units in high-risk areas is higher than that in low-risk areas; The rainfall data of each grid unit is calculated by spatial interpolation through rain gauges covering its monitoring range, and the dynamic weight coefficient of the corresponding rainfall time pattern is superimposed during the interpolation process to correct the rainfall contribution degree; Surrounding rock deformation data is collected by displacement gauges and associated with the corresponding grid units. The non-linear coupling analysis of the correlation between rainfall amount and surrounding rock deformation amount is performed in combination with the geological permeability coefficient and the mean deformation rate to generate the deformation response coefficient of each grid unit. Construct a spatial grid mapping relationship based on the deformation response coefficient and rainfall data, and the weight distribution of the spatial grid mapping relationship is dynamically adjusted according to the geological permeability coefficient and the variance of the deformation rate.

6. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 1, characterized in that, Conduct seepage guidance analysis and deformation diffusion risk coupling analysis on each grid cell according to the spatial grid mapping relationship, and generate seepage guidance parameters and deformation diffusion risk parameters, including: Calculate the seepage guidance parameter based on the product of the rainfall gradient and the geological permeability coefficient of the grid cell. The rainfall gradient is the difference in rainfall between adjacent grid cells divided by the corresponding spacing. Calculate the deformation diffusion risk parameter according to the covariance of the deformation rate of the current grid cell and the stress state of adjacent grid cells. The stress state is quantified by the change rate of the surrounding rock displacement monitored by a displacement meter. Overlay the seepage guidance parameter and the deformation diffusion risk parameter.

7. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 1, characterized in that Conduct a health status grading assessment on the grid cells based on the seepage guidance parameter and the deformation diffusion risk parameter, and correct the health status grading assessment results in combination with a preset deformation mode reference library, including: Generate a preliminary grading result of the health status according to the comparison relationship between the seepage guidance parameter and the pore water pressure threshold, and the comparison relationship between the deformation diffusion risk parameter and the deformation rate threshold. Based on the preset deformation mode reference library, calculate the similarity distance between the current deformation data and various deformation modes in the reference library, and generate a mode deviation degree. Correct the preliminary grading result according to the mode deviation degree.

8. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 7, characterized in that, The preliminary grading result includes a stable state, a risk warning state, and an emergency response state. The correction rule for correcting the preliminary grading result is: when the mode deviation degree exceeds the deviation threshold, upgrade the stable state to the risk warning state, or upgrade the risk warning state to the emergency response state.

9. The intelligent decision-making support system for dynamic regulation of tunnel deformation related to rainfall according to claim 1, wherein Generate dynamic control instructions according to the corrected health status grading assessment results, including: Generate local support reinforcement instructions and drainage system startup instructions based on the grid cells in the emergency response state in the health status grading results. For the grid cells in the risk warning state, generate graded control instructions according to the weighted score ranking of the deformation rate variance and the mode deviation degree. The triggering timing of the dynamic control instructions is dynamically adjusted according to the timing matching degree between the real-time monitoring data of the pore water pressure and the historical seepage out-of-control events.

10. The intelligent decision support system for dynamic regulation of tunnel deformation related to rainfall according to claim 9, characterized in that, The reinforcement position of the support reinforcement instruction is located according to the high-value area of the seepage guidance parameter; the timing matching degree is calculated by the similarity distance between the current pore water pressure curve and the historical event curve.

Images