Event-triggered roadway deformation reinforcing method and system

Abstract

The invention discloses an event-triggered roadway deformation reinforcing method, which comprises the following steps of: in a normal mode, adopting point type monitoring in a first sampling period, and preprocessing monitoring data; judging whether the monitoring data meets a pre-established mode switching trigger criterion condition, if so, switching from a normal mode to an event mode, starting three-dimensional point cloud monitoring while adopting point type monitoring in the event mode, and switching the sampling period to a second sampling period smaller than the first sampling period; in an event mode, performing preliminary risk level judgment according to a point type monitoring result, and performing risk check and correction in combination with a three-dimensional point cloud monitoring result to obtain a final risk level judgment result; and obtaining a support parameter recommendation combination matched with the risk level based on a pre-constructed reinforcement gear mapping rule base. According to the method, low-frequency acquisition and abstract uploading are executed in the stationary period, the scanning / uploading / computing strength is improved only when the event is triggered, and the bandwidth, the computing power and the storage overhead are remarkably reduced.

Description

Technical Field

[0001] This invention relates to the field of tunnel deformation reinforcement technology, specifically to an event-triggered tunnel deformation reinforcement method and system. Background Technology

[0002] In mining-affected roadways, field monitoring results show that surrounding rock deformation exhibits phased characteristics: a period of severe deformation occurs from 0 to 30 days, followed by a period of reduced deformation from 30 to 50 days, and stability is achieved after 50 days. Under specific conditions, roof subsidence can reach approximately 139.0–161.8 mm, sidewall convergence approximately 170.3–193.4 mm, and floor heave approximately 94.0–111.2 mm. These values ​​represent common deformation magnitudes observed in typical mining-affected roadways during field monitoring, illustrating the deformation amplitude and evolutionary stages of large-deformation roadways. On the other hand, optimizing support parameters is not always "the bigger the better." Existing research indicates that the anchor bolt installation angle exhibits extreme values ​​for roof displacement, and the improvement in roof subsidence due to anchor bolt / cable length and preload shows diminishing marginal returns. Therefore, a reasonable parameter range should be selected, and optimization should be combined with economic considerations. Furthermore, monitoring of deep-well, heavily mined soft rock roadways shows that the cumulative convergence of the two sides can reach approximately 1100 mm, the cumulative subsidence of the roof can reach approximately 230 mm, and the cumulative uplift of the floor can reach approximately 1175 mm. Factors affecting deformation include large burial depth, soft surrounding rock, secondary mining impact, and unreasonable support schemes. Studies on deep, high-stress soft rock roadways also indicate that large burial depth, low surrounding rock strength, and mismatch between the original support scheme and the working conditions can lead to a significant decrease in the stability of the surrounding rock.

[0003] Existing technologies can be mainly summarized into two types: 1) Regarding the full-section deformation monitoring scheme based on point cloud formation by 3D laser scanning, this technology often requires full-process or high-frequency scanning and point cloud registration, section extraction and deformation inversion. The data volume is large, the calculation and storage costs and bandwidth costs are high. In addition, the dust in the well can cause noise points to appear in the scan. In summary, this method is likely to lead to high equipment layout and maintenance costs.

[0004] 2) Based on fixed-frequency acquisition schemes using point-based monitoring such as cross-shaped points, displacement gauges, or convergence meters, the acquisition and transmission overhead is low, but the spatial coverage is limited, making it difficult to capture local asymmetric deformation and accelerated sudden changes. Fixed sampling periods can easily cause early warning delays.

[0005] In summary, there is a need for a technical solution that can operate in a low-resource mode during stable periods, enhance sampling, uploading, and calculation when roof settlement enters an accelerated or excessive risk phase, selectively enable or increase the frequency of full-section scanning to obtain high-precision deformation information, and directly map the monitoring results to executable support and reinforcement parameter levels. Summary of the Invention

[0006] This invention provides an event-triggered tunnel deformation reinforcement method and system to solve the above-mentioned problems in the prior art.

[0007] According to a first aspect, one embodiment provides an event-triggered tunnel deformation reinforcement method, the method comprising: In normal mode, point-based monitoring is used in the first sampling period, and the monitoring data is preprocessed; Determine whether the monitoring data meets the pre-established mode switching trigger criteria. If it does, switch from normal mode to event mode. In event mode, while using point monitoring, enable three-dimensional point cloud monitoring and switch the sampling period to a second sampling period that is shorter than the first sampling period. In event mode, a preliminary risk level is determined based on the point monitoring results, and the risk is verified and corrected by combining the 3D point cloud monitoring results to obtain the final risk level determination result. Based on a pre-built reinforcement level mapping rule library, a recommended combination of support parameters matching the risk level is obtained; Determine whether the monitoring data meets the pre-established event mode exit criteria. If it does, exit the event mode and restore the normal mode, and the sampling period is restored to the first sampling period.

[0008] Furthermore, under normal conditions, point-based monitoring is adopted in the first sampling cycle, and the monitoring data is preprocessed, specifically including: Discrete points are monitored by point-deployed sensors, including at least one of displacement gauges, convergence gauges, delamination gauges, stress gauges, or strain gauges. The first sampling period collects at least one of the following monitoring indicators: roof settlement, roof delamination, sidewall convergence, bottom slab heave, and stress / strain response of support components; the monitoring data undergoes preprocessing including outlier removal, missing data completion, and time alignment.

[0009] Furthermore, it is determined whether the monitoring data meets the pre-established mode switching trigger criteria, specifically including: The monitoring indicators are selected as the basis for triggering the judgment, and the characteristic values ​​of the corresponding monitoring indicators are calculated, including the rate of change, the acceleration of change, and the ratio of the average rate of change of the short window to the long window. The short window and the long window refer to the sliding time window or statistical time scale used to calculate the deformation characteristics. Set the trigger threshold for the corresponding monitoring indicator characteristic value. If any characteristic value exceeds the corresponding trigger threshold, the trigger condition of the trigger logic function Trigger(t) will be met. To suppress false triggering caused by noise or single anomalies, a trigger persistence criterion is introduced: when the trigger logic function Trigger(t) triggers in N consecutive samples, the mode switches from normal mode to event mode. The persistence criterion is expressed as follows: ≥N Where k is the summation index, representing the k-th sampling interval for backward backtracking; This indicates that index k is incremented from 0 to N-1; This represents the time interval between two consecutive calculations; N is the number of durations; 1{ } is an indicator function / indicator function, which takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

[0010] Furthermore, in the event mode, while employing point-based monitoring, 3D point cloud monitoring is enabled, and the sampling period is switched to a second sampling period shorter than the first sampling period. Specifically, it also includes: When link resources are limited, key feature values ​​should be uploaded first; when link resources are available, at least one of high-frequency sequence data and point cloud data should be uploaded.

[0011] Furthermore, in event mode, a preliminary risk level assessment is made based on the point-based monitoring results, and risk verification and correction are performed in conjunction with the 3D point cloud monitoring results to obtain the final risk level assessment result, which specifically includes: Risk level It includes at least three levels: low, medium, and high, and the judgment rules are expressed in the form of segmented thresholds:

[0012] Where L(t) = 0, 1, 2 correspond to low risk, medium risk, and high risk, respectively; , G(t) is the grading threshold; G(t) is the risk assessment indicator, which adopts at least one of the feature values ​​of the point-based monitoring indicators, or adopts a combination of feature values ​​to form a comprehensive assessment indicator. Based on the point cloud data obtained from 3D scanning, spatial features including cross-sectional contour changes, cross-sectional area changes or their rate of change, left-right asymmetric convergence degree, and local abnormal deformation are extracted and used to verify and conservatively correct the risk level L(t). When the point cloud spatial features indicate the presence of local abnormal exacerbation, significant asymmetric deformation, or rapid cross-sectional convergence, the risk level is raised by one level or the high-risk level is maintained.

[0013] Furthermore, based on a pre-built reinforcement level mapping rule base, recommended combinations of support parameters matching the risk level are obtained, specifically including:

[0014] Wherein, Φ( This is to strengthen the gear mapping rule base; The risk level is represented by Y(t), which is the recommended combination of support parameters, including the spacing between anchor bolts and cables s and the thickness of shotcrete x, the parameters p of the U-shaped steel support or scaffold, the pre-tightening force of the anchor cables, the grouting parameters, and the temporary support parameters. The reinforced gear position mapping rule base Φ( ) is a pre-established set of "risk level - support parameter" correspondences, used to map the risk level L(t) to a recommended combination of support parameters Y(t) that can be directly executed.

[0015] Furthermore, strengthen the gear mapping rule base Φ( The construction of ) includes: Constraint collection: Collect existing design parameters of the mine, relevant industry specifications / standards, previous support and reinforcement cases of similar roadways, and on-site constraints that can be implemented; Item Construction: Using risk level as the main index, construct several discrete reinforcement level items; each item should contain at least the corresponding recommended combination of support parameters; Verification and Consolidation: Safety and feasibility verification are performed on candidate item entries. After verification, the items are consolidated into an item library that can be directly looked up and output.

[0016] Furthermore, it is determined whether the monitoring data meets the pre-established event pattern exit criteria, specifically including: Set an exit threshold for the corresponding monitoring indicator characteristic values. If all characteristic values ​​are lower than the corresponding exit threshold, the exit condition of the exit logic function Exit(t) is met. The exit threshold is set to a certain proportion lower than the trigger threshold to form a hysteresis interval, thereby avoiding frequent jumps. To suppress false exits caused by noise or single anomalies, an exit persistence criterion is introduced: if the number of times Exit(t) is true within the most recent M samples is not less than M, then exit the event mode and return to the normal mode; The criteria for exiting persistence are expressed as follows:

[0017] in This indicates adding from k=0 to k=M-1, 1{ The parentheses represent an exponential / indicator function that takes the value 1 if the condition within the parentheses is true, and 0 otherwise. Exit(t) = 1 indicates that the exit condition is true at the t-th sampling. This indicates the time interval between two consecutive calculations.

[0018] Furthermore, it is determined whether the monitoring data meets the pre-established mode switching trigger criteria, specifically including: When the roof subsidence is selected as the trigger criterion: Define the trigger logic function Trigger(t):

[0019] in, To increase the rate of settlement, The threshold for triggering settlement rate increase. For settlement acceleration, The trigger threshold for the rate of change in growth rate. This represents the ratio of the growth rate of the short window to the growth rate of the long window within the time window. The trigger threshold is the ratio of the growth rate of the short window to that of the long window; symbol " "" indicates the logical "OR", meaning that if either condition is true, then The triggering condition is deemed met when This indicates that the triggering condition is not met; Settlement growth rate :

[0020] Accelerated growth :

[0021] Where S(t) is the roof settlement, and T is the calculation time interval, i.e., T=T1 in normal mode and T=T2 in event mode, and T2 <T1;S(t T) represents time t T's top plate subsidence; The ratio of the growth rate of the short window to the long window in the time window :

[0022] Where S(t) is the amount of roof subsidence. For the short window length, The length of the long window, and The average growth rate of the short window and the average growth rate of the long window were respectively and .

[0023] Furthermore, it is determined whether the monitoring data meets the pre-established event pattern exit criteria, specifically including: When the roof subsidence is selected as the trigger criterion: The exit logic function Exit(t) is defined as follows:

[0024] in, To increase the rate of settlement, This is the exit threshold for the rate of increase in subsidence. For settlement acceleration, This is the exit threshold for settlement acceleration.

[0025] According to three aspects, one embodiment provides an electronic device, the device comprising: a processor and a memory; The memory is used to store one or more program instructions; The processor is configured to run one or more program instructions to perform the steps of an event-triggered tunnel deformation reinforcement method as described in any of the preceding claims.

[0026] According to a fourth aspect, one embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of an event-triggered tunnel deformation reinforcement method as described in any of the preceding claims.

[0027] This invention provides an event-triggered tunnel deformation reinforcement method, which has the following beneficial effects: 1) Compared with fixed-frequency point monitoring, this invention utilizes the speed-up and acceleration triggering mechanism to improve sampling and calculation in a timely manner during the deformation acceleration stage, thereby reducing early warning lag and minimizing missed reports; 2) Compared with high-frequency scanning of point clouds across the entire cross section, this invention only increases the scanning / uploading / computing intensity when an event is triggered, and collects data and uploads summaries at low frequency during the stable period, which significantly reduces bandwidth, computing power and storage overhead; 3) By mapping risk levels, risk levels are directly converted into executable combinations of support parameters (anchor bolt and cable spacing, preload, shotcrete thickness, and optional grouting anchors, U-shaped steel / scaffolding parameters, etc.), forming a closed loop of monitoring, assessment, and recommendations; 4) Employ trigger or exit hysteresis and persistence criteria to reduce false triggering and frequent switching, ensuring on-site execution stability; 5) In event mode, three-dimensional laser scanning can be enabled to form point clouds to obtain the deformation field of the entire cross section and perform cross section extraction and settlement inversion, which is used to spatially verify and correct the point monitoring results, thus balancing accuracy and resource consumption. Attached Figure Description

[0028] Figure 1 A flowchart of an event-triggered tunnel deformation reinforcement method provided in one embodiment of the present invention; Figure 2 This is a flowchart illustrating the specific implementation of an event-triggered tunnel deformation reinforcement method according to an embodiment of the present invention. Figure 3 This is a timing diagram of feature priority uploading and re-uploading in event mode in an event-triggered tunnel deformation reinforcement method provided in one embodiment of the present invention; Figure 4 This is a diagram showing the calculation of point cloud data-derived cross-sectional indicators and their rate of change in an event-triggered tunnel deformation reinforcement method provided in one embodiment of the present invention. Detailed Implementation

[0029] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to the present invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0030] Furthermore, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for the clear description of a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed.

[0031] The first embodiment of this invention provides an event-triggered tunnel deformation reinforcement method that increases sampling frequency, communication transmission, and computing resources only when triggering conditions are met, and outputs suggested support reinforcement parameter levels. The following is in conjunction with... Figure 1 and Figure 2 Please provide a detailed explanation.

[0032] like Figure 1 As shown, in step S100, point monitoring is performed in the first sampling period under normal mode, and the monitoring data is preprocessed.

[0033] Routine monitoring and preprocessing: In this embodiment, the top plate settlement data S(t) is collected in the first sampling period T1, and the monitoring quantities such as top plate delamination, top and bottom plate movement, sidewall convergence, bottom heave, and stress or strain of support components can be collected simultaneously. The monitoring data is preprocessed by removing outliers, filling in missing data and aligning time.

[0034] In normal operation mode, point monitoring such as displacement gauges / convergence meters is preferred to reduce costs. In event mode, 3D laser scanning to form point cloud monitoring or increasing its scanning frequency can be selectively enabled to obtain full-section, millimeter-level deformation field information for risk assessment and verification. Example of typical values: T1 = 24h or 6h.

[0035] Layered monitoring and on-demand activation strategy: In normal mode, point sensors such as displacement gauges and convergence meters are preferred for low-frequency data acquisition, generating and uploading only summary data or key features; in event mode, while maintaining point monitoring, three-dimensional laser scanning is selectively enabled to form point cloud monitoring or its scanning frequency is increased to extract spatial features such as cross-sectional contour, cross-sectional area change rate, and arch curvature, which are used to spatially verify and correct the risk classification results, thereby balancing monitoring accuracy and resource consumption.

[0036] The information acquired by the point sensors differs from that acquired by the 3D scan in terms of data format and spatial coverage. Point sensors acquire deformation or stress response data at discrete measuring points, outputting a scalar or time series with a limited number of channels that varies over time. This data includes at least the roof subsidence S(t), and optionally includes the approach / convergence of the sidewalls, the floor heave, the roof delamination, and the stress or strain of the support components. 3D scanning acquires spatial morphological information of the surrounding rock surface of the tunnel, outputting point cloud data (i.e., a set of coordinates and attributes of multiple spatial points). By registering, extracting sections, and fitting contours of the point clouds at different times, spatial features such as cross-sectional contours, changes in cross-sectional area and their rate of change, and changes in the geometric shape of the arch can be derived. Preferably, point sensors are primarily used for low-resource monitoring in normal mode; in event mode, the frequency of 3D scanning can be selectively enabled or increased for spatial verification and auxiliary judgment of deformation distribution, thus balancing monitoring cost and timely risk response.

[0037] Point sensors (displacement gauges, convergence gauges, delamination gauges, stress / strain gauges, etc.) acquire response quantities at discrete measurement points. The output format is a scalar or time series data with a small number of channels that varies with time. It includes at least the roof subsidence S(t), and can optionally include the roof delamination D(t), the side convergence C(t), the bottom bulge F(t), the stress / strain response P(t) of the support components, etc. The advantages of this type of data are low acquisition and transmission overhead, good continuity, and suitability for routine monitoring with low resources.

[0038] The point cloud data generated by 3D laser scanning acquires the spatial morphology of the surrounding rock surface in the tunnel, and the output is a set of points {(x i ,y i ,z iThe system includes optional attributes (reflection intensity, confidence level, etc.); by registering point clouds, extracting cross sections and fitting contours at different times, spatial features such as cross section contours, cross section area Asec(t), cross section area change rate, arch curvature change, left and right asymmetric convergence degree, and local abnormal deformation patches can be derived for full cross section verification and local risk identification.

[0039] Feature index calculation: Calculate the settlement amount S(t) and settlement rate V(t) within the sliding time window: (1) And the rate of change (acceleration) A(t): (2) Where S(t) is the amount of roof settlement, T is the calculation time interval, V(t) is the settlement rate, and A(t) is the settlement acceleration.

[0040] It can also calculate the corresponding rate or rate of change index based on monitoring quantities such as top plate delamination, side wall convergence / retraction, bottom plate heave, and stress or strain of support components; when three-dimensional scanning is enabled to form point cloud monitoring, spatial feature indexes such as cross-sectional area change rate can also be derived from point cloud data for risk assessment or auxiliary verification.

[0041] At time t, the settlement rate V(t) and the rate of change of ... s The length of the long window is W l ,and The average growth rate of the short window and the average growth rate of the long window are respectively: V s (t), V l (t).

[0042] In this invention, "short window" and "long window" both refer to sliding time windows or statistical time scales used to calculate deformation characteristics. The short window reflects the immediate trend of surrounding rock deformation in a recent period, while the long window reflects the background change level of surrounding rock deformation over a longer period. The time span of the short window is shorter than that of the long window (for example, a short window can be a window with several sampling intervals, while a long window can be several times the length of the short window). Therefore, it can be used to construct a comparative index of "short window change intensity compared to long window change intensity" to identify the transition from a stable stage to an accelerated deformation stage.

[0043] Based on this, the ratio of the growth rate of the short window to the long window, R(t), is defined as shown in equation (3): (3) Where S(t) is the amount of roof subsidence, Ws W l This represents the length of the time window.

[0044] like Figure 1 As shown, in step S200, it is determined whether the monitoring data meets the pre-established mode switching trigger criterion conditions. If it does, the normal mode is switched to the event mode. In the event mode, three-dimensional point cloud monitoring is enabled while point monitoring is used, and the sampling period is switched to a second sampling period that is less than the first sampling period.

[0045] Trigger Criterion: Define a trigger logic function Trigger(t), which characterizes whether a trigger event "requires reinforcement / requires entering event mode" occurs at time t. Its value is a Boolean or binary quantity: when... "Time" indicates that at least one triggering condition is met at time t (e.g., the rate of increase or change in settlement exceeds a preset threshold, or the ratio of short window to long window features exceeds a threshold), and triggers the system to switch from normal mode to event mode after combining with the persistence criterion; when This indicates that the trigger condition is not met, and the system maintains normal monitoring and low-resource operation. By expressing multiple trigger conditions in a unified logical function form, it is convenient to realize modular calculation of trigger criteria and subsequent resource scheduling control.

[0046] Its logic function Trigger(t) is shown in equation (4): (4) Among them, V th As the threshold for settlement rate increase, A th R is the threshold for the rate of change of growth. th The ratio threshold; symbol " "" indicates the logical "OR", meaning that if either condition is true, Trigger(t) is considered true.

[0047] To suppress false triggering caused by noise or single anomalies, a persistence criterion is introduced: when the triggering logic function is true in N consecutive samples, a hardening event is determined to be triggered and the event mode is entered. This persistence criterion can be expressed by equation (5): ≥N(5) Where k is the summation index, representing the k-th sampling interval for backward backtracking; This indicates that index k is incremented from 0 to N-1; This represents the time interval between two consecutive calculations; N is the number of durations; 1{ } is an indicator function / indicator function, which takes the value 1 when the condition in parentheses is true, and 0 otherwise. When equation (5) is satisfied, a reinforcement event is triggered, the system switches from normal mode to event mode and performs risk level determination and support gear output.

[0048] To adapt to different deformation modes and support forms, in addition to the roof settlement index, indicators such as roof delamination, sidewall convergence, floor bulging, stress or strain of support components, and cross-sectional area can also be introduced as triggering criteria. Let the roof delamination be D(t), the sidewall convergence be C(t), the floor bulging be F(t), the stress or strain response of the support components be P(t), and the cross-sectional area be A. sec (t), the corresponding rate or rate of change indices are as follows: (6) Where X(t) represents the value of any selectable monitoring quantity at time t, as given above. V X (t) represents the rate or rate of change index corresponding to X(t); This indicates the time interval between two consecutive calculations.

[0049] Threshold setting: To avoid the difficulty in applying thresholds due to differences in absolute deformation rates across different mining areas or roadways, trigger and exit thresholds can be adaptively generated based on "baseline statistics during the stable period." Let the mean of the growth rate sequence within the stable period sliding window be μ. V The standard deviation is σ V The mean of the growth rate change sequence is μ A The standard deviation is σ A The trigger threshold can then be set to , Where k is a coefficient, for example, 2 to 3. The exit threshold can be set to... , η is the hysteresis coefficient, for example, taken as 0.6 to 0.8, to form a hysteresis interval and suppress frequent switching. As an alternative, the quantile method can also be used to set the threshold, for example, taking the high quantile of the stable period sequence, such as the 90% to 99% quantile, as the trigger threshold to enhance the ability to resist outliers.

[0050] Event-mode resource scheduling: After entering event mode, the sampling period is switched from the first sampling period T1 to the second sampling period T2, and the data upload and calculation frequency is increased; the sampling period switching rule is shown in equation (7), and satisfies (For example, T2 = 1h or 10min).

[0051] (7) In event mode, the frequency of monitoring point clouds formed by 3D laser scanning can be enabled or increased. For example... Figure 3 As shown, in order to reduce the communication bandwidth usage, a "feature-first upload" strategy is adopted: when the link resources are limited, the top plate sinking amount S(t) and its key features such as V(t) and A(t) are uploaded first; when the link resources are available, at least one of the high-frequency sequence data and point cloud data is uploaded. The upload strategy is shown in Equation (8).

[0052] (8) Among them, B avail (t) represents the available bandwidth at time t, B th The threshold is used for supplementary transmission; Series / PointCloud represents at least one of high-frequency sequence data and point cloud data. As an optional implementation, the point cloud scanning frequency switches between normal mode and event mode, as shown in Equation (9), and the event mode frequency is higher than the normal mode frequency.

[0053] (9) The purpose of enabling or increasing the frequency of 3D laser scanning to form point cloud monitoring in event mode is to obtain spatial morphological information of the roadway cross section and the surrounding rock surface, which is used to spatially verify the triggering results of point monitoring, and to provide auxiliary criteria for cross-sectional scale for risk level correction and support parameter output.

[0054] like Figure 1 As shown, in step S300, in event mode, a preliminary risk level is determined based on the point monitoring results, and risk verification and correction are performed in combination with the three-dimensional point cloud monitoring results to obtain the final risk level determination result.

[0055] Risk level determination and reinforcement grade output: In event mode, the risk level L(t) is determined based on the roof settlement S(t), settlement rate V(t), and rate of change A(t). The risk level includes at least three grades: low, medium, or high, and the determination rule can be expressed in the form of segmented thresholds as Equation (10): (10) Where L(t) = 0, 1, 2 correspond to low risk, medium risk, and high risk, respectively; G1 and G2 are the grading thresholds; G(t) is the risk assessment indicator, preferably at least one of V(t), A(t), and R(t), or a comprehensive assessment indicator composed of a combination of the above indicators. For example, G(t)... .

[0056] like Figure 1 As shown, in step S400, based on the pre-built reinforcement level mapping rule library, a recommended combination of support parameters matching the risk level is obtained.

[0057] After obtaining the risk level L(t), the preset reinforcement level mapping rule is called to output the recommended combination of support parameters, as shown in equation (11): (11) Wherein, Φ( ) is the gear mapping rule or gear library; Y(t) includes at least the row spacing s between anchor bolts and anchor cables and the thickness of shotcrete x, and may optionally include U-shaped steel support or scaffold parameters p, anchor cable preload, grouting parameters and temporary support parameters, etc.

[0058] Furthermore, such as Figure 4 As shown, when 3D laser scanning is enabled to acquire point cloud data in event mode, spatial features such as cross-sectional contour changes, cross-sectional area changes or their rate of change, left-right asymmetric convergence degree, and local abnormal deformation can be extracted from the point cloud. These features are used to verify and conservatively correct the risk level L(t). When the spatial features of the point cloud indicate the presence of local abnormal exacerbation, significant asymmetric deformation, or rapid cross-sectional convergence, it is preferable to raise the risk level by one level or maintain a higher risk level, and output the corresponding recommended combination of support parameters.

[0059] Wherein, the gear mapping rule Φ( The risk level (L(t)) or support parameter library is a pre-established set of "risk level - support parameter" correspondences used to map the risk level L(t) to a recommended combination of support parameters Y(t) that can be directly executed. The support parameter library is established as follows: based on the existing mine design specifications, past support experience of similar roadways, and on-site implementation conditions, combined with engineering constraints such as surrounding rock type, burial depth stress level, roadway cross-section, and degree of mining impact, several discrete reinforcement support levels are given in advance; each support level includes at least the values ​​or ranges of the anchor bolt and anchor cable spacing s and the shotcrete thickness x, and may optionally include U-shaped steel support or canopy parameters p, anchor cable preload, grouting parameters, and temporary support parameters, to form a set of parameter entries {L, s, x, p, ...}.

[0060] The recommended combination of support parameters corresponding to each risk level is preferably a pre-determined discrete level (lookup table output) to ensure that executable suggestions can be given quickly after the event is triggered. At the same time, in one embodiment, the threshold or parameter entries in the level library can be updated according to construction feedback or monitoring results to achieve adaptation to different mining areas and different surrounding rock conditions, but this update does not change the basic mechanism of "risk level - level mapping output".

[0061] The gear mapping rule Φ( The risk level or support parameter combination library is a pre-established set of correspondences between risk levels and support parameter combinations. Its establishment process includes: (1) Constraint collection: Collect existing design parameters of the mine, relevant industry specifications / standards, previous support and reinforcement cases of similar roadways, and on-site constraints (material specifications, construction technology, net cross-section requirements, transportation and installation conditions, etc.).

[0062] (2) Item construction: The risk level L is used as the main index (and engineering characteristics such as the surrounding rock type, burial depth stress level, roadway cross section, and degree of mining impact can be selected as secondary indexes) to construct several discrete reinforcement level items; each item must include at least the value or range of the spacing s between anchor bolts and anchor cables and the thickness x of shotcrete, and can optionally include U-shaped steel / scaffolding parameters p, pre-tightening force, grouting parameters and temporary support parameters, etc., to form parameter items.

[0063] (3) Verification and solidification: Safety verification and constructability verification of candidate slot items (based on past monitoring results, engineering experience comparison, and, if necessary, numerical analysis or field test section verification). After verification, the slots are solidified into a slot library that can be directly looked up and output.

[0064] The gear selection database obtained through the above methods not only has standardized and experiential sources, but also meets the requirements for on-site implementation, and supports iterative updates based on construction feedback and monitoring results during subsequent operation.

[0065] like Figure 1 As shown, in step S500, it is determined whether the monitoring data meets the pre-established event mode exit criteria. If it does, the event mode is exited and the normal mode is restored, and the sampling period is restored to the first sampling period.

[0066] Exit Criterion and Normalization: When the exit condition is met and at least M samples are taken, the system exits the event mode and returns to the normalization mode, and the sampling period is restored to the first sampling period T1. The exit condition can be represented by the logic function Exit(t), as shown in equation (12): (12) To ensure exit stability and avoid frequent switching due to fluctuations in the ratio indicator during the low-speed fluctuation phase, the exit criterion adopts the joint decline of V(t) and A(t) as the exit condition to ensure exit stability. The aforementioned short-window to long-window growth ratio R(t) and its trigger threshold Rth are mainly used to characterize "recent acceleration" in the trigger phase, and can be used as auxiliary verification indicators rather than necessary conditions in the exit phase.

[0067] To suppress false exits caused by noise or single anomalies, a persistence criterion is introduced: when the number of times Exit(t) is true within the most recent M samples is not less than M, the exit event mode is determined, as shown in equation (13): (13) in denotes the sum from k = 0 to k = M - 1, 1{ } denotes that the exponential function / indicator function takes 1 when the condition in the parentheses holds, and 0 otherwise. Exit(t) = 1 indicates that the exit condition holds at the t-th sampling; denotes the time interval between two adjacent calculations.

[0068] When the condition of Equation (12) is satisfied, the sampling period switches back from T2 to T1, as shown in Equation (14): (14) where M can take 2 or 3, and Δt is the sampling interval. The exit threshold V exit and A exit are preferably set to a certain proportion lower than the trigger threshold to form a hysteresis interval, thus avoiding frequent jumps.

[0069] The present invention proposes an event-triggered large deformation reinforcement suggestion and resource adaptive scheduling method for roadway surrounding rock. By calculating the roof subsidence S(t), settlement growth rate V(t), growth rate change rate A(t) and the optional short-window and long-window growth rate ratio R(t) in real time, and combining the "trigger criterion + persistence criterion + exit hysteresis criterion", it realizes the automatic identification of large deformation risks in the roadway, suppression of false triggers and stable switching.

[0070] The present invention proposes a resource adaptive scheduling mechanism for event patterns: when the trigger event holds, the sampling period is automatically switched from the first sampling period T1 to the second sampling period T2 (and T2 < T1), and the data upload and calculation frequencies are synchronously increased; and under the condition of limited link, the "feature first upload" strategy is adopted to preferentially upload the key features S, V, A to reduce the bandwidth occupation, and then supplement the high-frequency sequence data and / or point cloud data when the link permits, so as to significantly reduce the consumption of communication, calculation and storage resources while ensuring safe response.

[0071] The present invention establishes a risk level - support parameter gear mapping output mechanism: in the event mode, the risk level L(t) is divided into at least three gears of low / medium / high, and the directly executable recommended combination of support parameters {s, x} (spacing s between bolts and cables, shotcrete thickness x) is output through the preset gear mapping rule, and the U-shaped steel support / frame parameters, cable pre-tightening force, grouting parameters and temporary support parameters can be optionally output, realizing the automatic and standardized decision-making output from the monitoring results to the support plan.

[0072] In addition, embodiments of the present invention also provide an electronic device, the device comprising: a processor and a memory; the memory being used to store one or more program instructions; the processor being used to execute one or more program instructions to perform the steps of an event-triggered tunnel deformation reinforcement method as described in any of the preceding embodiments.

[0073] It should be noted that for a detailed description of the electronic device provided in the embodiments of the present invention, please refer to the relevant description of the event-triggered tunnel deformation reinforcement method provided in the embodiments of this application, which will not be repeated here.

[0074] In addition, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of an event-triggered tunnel deformation reinforcement method as described in any of the preceding claims.

[0075] It should be noted that for a detailed description of the computer-readable storage medium provided in the embodiments of the present invention, please refer to the relevant description of the event-triggered tunnel deformation reinforcement method provided in the embodiments of this application, which will not be repeated here.

[0076] The above examples illustrate the present invention only to aid in understanding it and are not intended to limit the scope of the invention. Those skilled in the art can make various simple deductions, modifications, or substitutions based on the principles of this invention.

Claims

1. An event-triggered tunnel deformation reinforcement method, characterized in that, The method includes: In normal mode, point-based monitoring is used in the first sampling period, and the monitoring data is preprocessed; Determine whether the monitoring data meets the pre-established mode switching trigger criteria. If it does, switch from normal mode to event mode. In event mode, while using point monitoring, enable three-dimensional point cloud monitoring and switch the sampling period to a second sampling period that is shorter than the first sampling period. In event mode, a preliminary risk level is determined based on the point monitoring results, and the risk is verified and corrected by combining the 3D point cloud monitoring results to obtain the final risk level determination result. Based on a pre-built reinforcement level mapping rule library, a recommended combination of support parameters matching the risk level is obtained; Determine whether the monitoring data meets the pre-established event mode exit criteria. If it does, exit the event mode and restore the normal mode, and the sampling period is restored to the first sampling period.

2. The event-triggered tunnel deformation reinforcement method as described in claim 1, characterized in that, In normal operation, point-based monitoring is used in the first sampling period, and the monitoring data is preprocessed, specifically including: Discrete points are monitored by point-deployed sensors, including at least one of displacement gauges, convergence gauges, delamination gauges, stress gauges, or strain gauges. The first sampling period collects at least one of the following monitoring indicators: roof settlement, roof delamination, sidewall convergence, bottom slab heave, and stress / strain response of support components; the monitoring data undergoes preprocessing including outlier removal, missing data completion, and time alignment.

3. The event-triggered tunnel deformation reinforcement method as described in claim 1, characterized in that, Determine whether the monitoring data meets the pre-established mode switching trigger criteria, specifically including: The monitoring indicators are selected as the basis for triggering the judgment, and the characteristic values ​​of the corresponding monitoring indicators are calculated, including the rate of change, the acceleration of change, and the ratio of the average rate of change of the short window to the long window. The short window and the long window refer to the sliding time window or statistical time scale used to calculate the deformation characteristics. Set the trigger threshold for the corresponding monitoring indicator characteristic value. If any characteristic value exceeds the corresponding trigger threshold, the trigger condition of the trigger logic function Trigger(t) will be met. To suppress false triggering caused by noise or single anomalies, a trigger persistence criterion is introduced: when the trigger logic function Trigger(t) triggers in N consecutive samples, the mode switches from normal mode to event mode. The persistence criterion is expressed as follows: ≥N Where k is the summation index, representing the k-th sampling interval for backward backtracking; This indicates that index k is incremented from 0 to N-1; This represents the time interval between two consecutive calculations; N is the number of durations; 1{ } is an indicator function / indicator function, which takes the value 1 when the condition inside the parentheses is true, and 0 otherwise.

4. The event-triggered tunnel deformation reinforcement method as described in claim 1, characterized in that, In the event mode, while employing point monitoring, 3D point cloud monitoring is also enabled, and the sampling period is switched to a second sampling period that is shorter than the first sampling period. Specifically, it also includes: When link resources are limited, key feature values ​​should be uploaded first; when link resources are available, at least one of high-frequency sequence data and point cloud data should be uploaded.

5. The event-triggered tunnel deformation reinforcement method as described in claim 1, characterized in that, In event mode, a preliminary risk level assessment is made based on point-based monitoring results, and then risk verification and correction are performed in conjunction with 3D point cloud monitoring results to obtain the final risk level assessment result, which specifically includes: Risk level It includes at least three levels: low, medium, and high, and the judgment rules are expressed in the form of segmented thresholds: Where L(t) = 0, 1, 2 correspond to low risk, medium risk, and high risk, respectively; , G(t) is the grading threshold; G(t) is the risk assessment indicator, which adopts at least one of the feature values ​​of the point-based monitoring indicators, or adopts a combination of feature values ​​to form a comprehensive assessment indicator. Based on the point cloud data obtained from 3D scanning, spatial features including cross-sectional contour changes, cross-sectional area changes or their rate of change, left-right asymmetric convergence degree, and local abnormal deformation are extracted and used to verify and conservatively correct the risk level L(t). When the point cloud spatial features indicate the presence of local abnormal exacerbation, significant asymmetric deformation, or rapid cross-sectional convergence, the risk level is raised by one level or the high-risk level is maintained.

6. The event-triggered tunnel deformation reinforcement method as described in claim 1, characterized in that, Based on a pre-built reinforcement level mapping rule library, recommended combinations of support parameters matching the risk level are obtained, specifically including: Wherein, Φ( This is to strengthen the gear mapping rule base; The risk level is represented by Y(t), which is the recommended combination of support parameters, including the spacing between anchor bolts and cables s and the thickness of shotcrete x, the parameters p of the U-shaped steel support or scaffold, the pre-tightening force of the anchor cables, the grouting parameters, and the temporary support parameters. The reinforced gear position mapping rule base Φ( ) is a pre-established set of "risk level - support parameter" correspondences, used to map the risk level L(t) to a recommended combination of support parameters Y(t) that can be directly executed.

7. The event-triggered tunnel deformation reinforcement method as described in claim 6, characterized in that, Strengthen gear mapping rule base Φ( The construction of ) includes: Constraint collection: Collect existing design parameters of the mine, relevant industry specifications / standards, previous support and reinforcement cases of similar roadways, and on-site constraints that can be implemented; Item Construction: Using risk level as the main index, construct several discrete reinforcement level items; each item should contain at least the corresponding recommended combination of support parameters; Verification and Consolidation: Safety and feasibility verification are performed on candidate item entries. After verification, the items are consolidated into an item library that can be directly looked up and output.

8. The event-triggered tunnel deformation reinforcement method as described in claim 3, characterized in that, Determine whether the monitoring data meets the pre-established event pattern exit criteria, specifically including: Set an exit threshold for the corresponding monitoring indicator characteristic values. If all characteristic values ​​are lower than the corresponding exit threshold, the exit condition of the exit logic function Exit(t) is met. The exit threshold is set to a certain proportion lower than the trigger threshold to form a hysteresis interval, thereby avoiding frequent jumps. To suppress false exits caused by noise or single anomalies, an exit persistence criterion is introduced: if the number of times Exit(t) is true within the most recent M samples is not less than M, then exit the event mode and return to the normal mode; The criteria for exiting persistence are expressed as follows: in This indicates adding from k=0 to k=M-1, 1{ The parentheses represent an exponential / indicator function that takes the value 1 if the condition within the parentheses is true, and 0 otherwise. Exit(t) = 1 indicates that the exit condition is true at the t-th sampling. This indicates the time interval between two consecutive calculations.

9. The event-triggered tunnel deformation reinforcement method as described in claim 3, characterized in that, Determine whether the monitoring data meets the pre-established mode switching trigger criteria, specifically including: When the roof subsidence is selected as the trigger criterion: Define the trigger logic function Trigger(t): in, To increase the rate of settlement, The threshold for triggering settlement rate increase. For settlement acceleration, The trigger threshold for the rate of change in growth rate. This represents the ratio of the growth rate of the short window to the growth rate of the long window within the time window. The trigger threshold is the ratio of the growth rate of the short window to that of the long window; symbol " "Represents the logical "OR", meaning if either condition is true, then The triggering condition is deemed met when This indicates that the triggering condition is not met; Settlement growth rate : Acceleration of growth : Where S(t) is the roof settlement, and T is the calculation time interval, i.e., T=T1 in normal mode and T=T2 in event mode, and T2 <T1;S(t T) represents time t T's top plate subsidence; The ratio of the growth rate of the short window to the long window in the time window : Where S(t) is the amount of roof subsidence. For the short window length, The length of the long window, and The average growth rate of the short window and the average growth rate of the long window were respectively and .

10. The event-triggered tunnel deformation reinforcement method as described in claim 9, characterized in that, Determine whether the monitoring data meets the pre-established event pattern exit criteria, specifically including: When the roof subsidence is selected as the trigger criterion: The exit logic function Exit(t) is defined as follows: in, To increase the rate of settlement, This is the exit threshold for the rate of increase in subsidence. For settlement acceleration, This is the exit threshold for settlement acceleration.

Images