Method and System for Inverting Dynamic Loads on Surrounding Rock of Deep-Buried Tunnels under Blasting Disturbance

Abstract

This invention relates to the field of tunnel and underground engineering technology, and discloses a method and system for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance. This solves the problem that the dynamic load of surrounding rock in deep-buried tunnel blasting construction is difficult to measure directly, and that existing simulation methods have low accuracy. The inversion calculation method of this invention includes: deploying sensors at typical tunnel cross-sections to monitor the vibration velocity time history data of the surrounding rock during blasting construction; establishing a three-dimensional finite element numerical model based on the tunnel structure and surrounding rock characteristics; constructing an optimized inversion objective function with the goal of minimizing the error between measured vibration data and numerical simulation calculation data; and employing an intelligent optimization algorithm to iteratively invert the equivalent dynamic load time history acting on the surrounding rock. This invention can efficiently and accurately invert the true dynamic response of the surrounding rock, providing a scientific basis for the design optimization and construction safety control of tunnel support structures.

Description

Technical Field

[0001] This invention relates to the field of tunnel and underground engineering technology, specifically to a method and system for inverting the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance. Background Technology

[0002] In the construction of deep-buried tunnels in mountainous areas, the drill-and-blast method is widely used due to its strong adaptability and good economy. However, the strong impact load generated by blasting can disturb the original stress state of the surrounding rock, threatening the stability of the tunnel and the initial support structure. Therefore, accurately obtaining the dynamic loads acting on the tunnel surrounding rock during blasting is the key to scientific support design and safety assessment.

[0003] Currently, direct measurement of dynamic stress or load inside surrounding rock faces challenges such as difficulty in sensor deployment, low survival rate, and high cost, making it difficult to achieve efficient on-site monitoring.

[0004] In existing technologies, the simulation of blasting loads mainly relies on empirical formulas and simplified waveforms. For example, the Sadovsky formula is often used to predict blasting vibration velocities.

[0005] ;

[0006] in, Let be the velocity of the particle vibration. For the amount of explosives, For distance, and These are empirical coefficients related to the medium and blasting conditions. Although this formula can roughly estimate vibration propagation attenuation, it is based solely on the charge amount and distance, and cannot reflect the nonlinear dynamic response and radiation damping effect of the surrounding rock under complex geological conditions.

[0007] Furthermore, existing numerical simulations often use simplified triangular loads or exponentially decaying waveforms as equivalent blast load inputs. For example, the load is simplified to a linearly rising-linearly falling triangular pulse, or a single exponentially decaying function. These methods neglect the multidirectional propagation of blast shock waves in the surrounding rock, waveform reflection / scattering, and the interaction between the surrounding rock and the support structure. This leads to significant deviations between the simulated dynamic response of the surrounding rock and the actual response, especially under conditions of high ground stress and nonlinear materials in deeply buried tunnels, making it impossible to accurately assess the safety of the support structure.

[0008] Therefore, designing a theoretical method and system that can indirectly and accurately invert the dynamic load of surrounding rock based on easily measurable physical quantities has important theoretical value and engineering significance. Summary of the Invention

[0009] The technical problem to be solved by this invention is to provide a method and system for inverting and calculating the dynamic load of the surrounding rock in deep-buried tunnels under blasting disturbance, so as to solve the problem that the dynamic load of the surrounding rock in deep-buried tunnel blasting construction is difficult to measure directly, and the existing simulation methods have low accuracy.

[0010] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0011] On the one hand, the present invention provides a method for inverting and calculating the dynamic load of the surrounding rock of a deep-buried tunnel under blasting disturbance, comprising the following steps:

[0012] S1. Vibration velocity sensors are deployed at typical monitoring sections of deep-buried tunnels to collect time history data of vibration velocity at key parts of the surrounding rock during blasting operations.

[0013] S2. Based on the tunnel geometric parameters, surrounding rock mechanical parameters, and support parameters, establish a three-dimensional finite element numerical model that can reflect the interaction between the surrounding rock and the support structure;

[0014] S3. The objective function of the inversion analysis is established with the goal of iteratively converging the error norm between the measured vibration velocity time history data and the vibration velocity time history data calculated by the three-dimensional finite element numerical model under equivalent dynamic load to the global minimum.

[0015] S4. Using a global optimization algorithm, the objective function is iteratively solved until the objective function converges, thereby obtaining the time history of the equivalent dynamic load acting on the tunnel surrounding rock.

[0016] This scheme uses the readily measurable surface vibration velocity time history of the surrounding rock as the inversion constraint, a three-dimensional numerical model considering the interaction between the surrounding rock and the support as the computational carrier, a quantitative convergence criterion based on minimizing the error norm, and a global optimization algorithm to complete the nonlinear load time history optimization. This forms a dynamic inversion mechanism driven by measured constraints, forward transmission of the numerical model, evaluation of the objective function error, and iterative optimization by the intelligent algorithm. From a physical mechanism perspective, this approach bypasses the engineering bottlenecks faced by direct measurement of dynamic loads inside the surrounding rock, such as difficulties in sensor deployment, high costs, and low survival rates. At the same time, from a computational principle perspective, it overcomes the accuracy limitations of traditional empirical formulas that do not consider the nonlinearity and radiation damping of the surrounding rock, and simplified loads that do not consider the propagation / reflection of shock waves and the dynamic interaction between the structure and the rock. This allows the equivalent dynamic load obtained from the inversion to truly restore the characteristics and spatial distribution of blasting impacts, improving the accuracy and engineering applicability of dynamic load identification of the surrounding rock under complex geological conditions of deep burial and high ground stress. This provides reliable load data for dynamic response analysis, safety assessment, and dynamic design of tunnel support structures.

[0017] Furthermore, in step S1, the typical monitoring section is selected from the section with complex geological conditions or the section with the maximum burial depth. The key parts of the surrounding rock include the tunnel's arch, arch waist, and sidewalls. The section with complex geological conditions includes the section with fractured surrounding rock, the section with developed joints and fissures, the section with weak interlayers, or the section with fault zones.

[0018] The vibration velocity sensors are deployed as follows: they are placed along the tunnel arch, left and right arch waists, and left and right sidewalls; each vibration velocity sensor monitors vibration velocity components in at least three directions: radial, tangential, and axial.

[0019] In this scheme, sections with complex geological conditions and sections with the maximum burial depth exhibit high levels of surrounding rock stress and more significant dynamic responses. Selecting these sections allows for the acquisition of the most representative vibration data under the most unfavorable working conditions, ensuring the effectiveness of the inversion constraints. The arch crown, arch waist, and sidewalls are the most sensitive parts of the tunnel section to the impact of blasting and the most critical parts of the support structure under stress. Multi-location deployment can cover the spatial dynamic response distribution characteristics of the surrounding rock. Vibration velocity sensors monitoring the three-dimensional vibration components can completely capture the vibration transmission law in three orthogonal directions, avoiding inversion distortion caused by data from a single location or direction, and improving the completeness and reliability of the inversion results from the data source. Among them, radial direction is the direction along the radius of the tunnel section; tangential direction is the direction along the circumferential tangent on the tunnel cross section; and axial direction is the direction parallel to the longitudinal axis of the tunnel, i.e., the direction of tunnel extension.

[0020] Furthermore, the three-dimensional finite element numerical model uses viscoelastic boundaries to simulate the radiation damping effect of an infinite domain foundation, and employs a nonlinear dynamic constitutive model to describe the mechanical properties of the surrounding rock.

[0021] In this scheme, the viscoelastic boundary can absorb the scattered waves and radiated energy generated during the propagation of the blast shock wave, avoiding the distortion of the dynamic response calculation of the surrounding rock caused by wave reflection at the outer boundary of the model, and accurately restoring the real dynamic boundary properties of the surrounding rock in the infinite domain; the nonlinear dynamic constitutive model of the surrounding rock can reflect the plastic deformation and energy dissipation characteristics of the surrounding rock under high ground stress, which fits the real mechanical behavior of the surrounding rock of the deep-buried tunnel, making the load-response transfer relationship of the numerical model consistent with the actual engineering, and ensuring the basic accuracy of the inversion from the perspective of forward computing theory.

[0022] Furthermore, in step S3, the objective function for the inversion analysis is established as follows:

[0023] ;

[0024] in, The objective function is... The equivalent dynamic load time history to be inverted; For the first The first measuring point Time histories of vibration velocities measured in each direction; For a three-dimensional finite element numerical model under load The vibration velocity time history calculated for the corresponding measuring point and direction under the action; These are the weighting coefficients; This represents the number of measurement points. The number of directions.

[0025] In this scheme, the sum of squared errors is used to be more sensitive to the deviation between the measured and simulated responses. It can simultaneously constrain the vibration response fitting degree across all time periods, all measurement points, and all directions, ensuring that the equivalent dynamic load time history obtained by inversion is true and reliable in both the time domain and spatial distribution.

[0026] Furthermore, the weight coefficients in the objective function The settings should be adjusted based on the importance of the measurement point location, the signal-to-noise ratio, or the sensitivity to vibration direction.

[0027] The tunnel arch crown and arch waist are designated as primary important measuring points, and the sidewalls are designated as secondary important measuring points. The weight coefficients for primary important measuring points range from 1.5 to 2.0, and the weight coefficients for secondary important measuring points range from 0.8 to 1.0.

[0028] For measurement points with a signal-to-noise ratio (SNR) greater than or equal to a preset SNR threshold, the weighting coefficient ranges from 1.2 to 1.5; for measurement points with a SNR less than the preset SNR threshold, the weighting coefficient ranges from 0.5 to 0.8.

[0029] The radial direction is defined as the sensitive vibration direction of the blasting impact. The radial direction weighting coefficient ranges from 1.2 to 1.5, while the tangential and axial direction weighting coefficients range from 0.8 to 1.0.

[0030] In this scheme, weights are assigned based on the importance of measurement points, which can enhance the response contribution of parts such as the arch crown and arch waist that play a controlling role in the safety of the tunnel structure, so that the inverted load can be preferentially matched with the dynamic characteristics of key areas; weights are assigned based on the signal-to-noise ratio, which can reduce the adverse effects of data with large noise interference and poor waveform integrity on the inversion and improve the stability of the inversion; weights are assigned based on the sensitivity of vibration direction, which can highlight the components that respond more significantly to changes in blasting loads and improve the accuracy of the inversion in identifying load characteristics.

[0031] Furthermore, in step S4, during the iterative solution process, the current candidate equivalent dynamic load time history is applied to the face or excavation profile of the three-dimensional finite element numerical model in the form of surface pressure load to calculate the corresponding vibration velocity time history data.

[0032] In this scheme, the blasting impact essentially generates instantaneous surface forces on the tunnel face and the surrounding rock of the excavation outline in the form of shock waves and explosive gas pressure. During the iteration process, candidate loads are applied to the above-mentioned locations in the form of surface pressure, which conforms to the actual action mechanism and force transmission path of blasting loads. By continuously updating candidate loads and repeating calculations, the global optimal solution that minimizes the objective function is gradually approximated, ensuring the physical rationality and mathematical convergence of the inversion iteration.

[0033] Furthermore, the surface pressure load is uniformly distributed along the normal direction of the tunnel face, or spatially non-uniformly distributed according to the arrangement of the blasting holes; when the blasting adopts multi-stage delayed detonation, the equivalent dynamic load time history is the superposition of multiple sub-load time histories.

[0034] In this scheme, uniformly distributed loads are suitable for conventional full-section blasting, while non-uniform distributions can match the differences in borehole layout and charge space, improving the adaptability of load form and blasting design. Under multi-stage millisecond delay initiation, the time history superposition of multiple sub-loads can restore the timing characteristics of segmented initiation and the law of energy release in stages, which is consistent with the actual loading process of the on-site blasting network. This makes the equivalent dynamic load consistent with the actual blasting effect in both spatial distribution and time history, further improving the inversion accuracy.

[0035] Furthermore, in step S4, the global optimization algorithm adopts a genetic algorithm, a particle swarm optimization algorithm, or a simulated annealing algorithm.

[0036] In this scheme, the genetic algorithm, particle swarm optimization algorithm, and simulated annealing algorithm all have global search capabilities, do not depend on the selection of initial values, are not easily trapped in local optima, and can stably solve complex optimization problems such as load time history inversion that are high-dimensional, nonlinear, and without explicit expressions. The adaptive iterative characteristics of the algorithms can automatically balance the search breadth and convergence speed, ensuring that the equivalent dynamic load time history obtained by inversion is the global optimal solution, thus improving the stability of the results.

[0037] Furthermore, in step S4, the conditions for the convergence of the objective function include: the objective function value is less than a preset threshold, the improvement of the objective function after multiple consecutive iterations is less than a preset percentage, or the maximum number of iterations is reached.

[0038] In this scheme, the multi-condition combined convergence criterion can achieve a balance between inversion accuracy and computational efficiency: the objective function threshold ensures that the inversion results meet the engineering accuracy requirements, the iterative improvement magnitude threshold avoids invalid iterations and over-computation, and the maximum number of iterations prevents computational dead loops, thus taking into account both the speed of field application and the reliability of results.

[0039] On the other hand, the present invention also provides a dynamic load inversion calculation system for the surrounding rock of a deep-buried tunnel under blasting disturbance, for implementing the above-mentioned inversion calculation method. The system includes:

[0040] The data acquisition module is used to control the vibration velocity sensors deployed at typical monitoring sections of deep-buried tunnels to collect and store vibration velocity time history data during the blasting process.

[0041] The numerical simulation module is used to establish a three-dimensional finite element numerical model that reflects the interaction between the surrounding rock and the support structure, based on the tunnel's geometric parameters, surrounding rock mechanical parameters, and support parameters.

[0042] The inversion analysis module is used to establish an objective function for inversion analysis by iteratively converging the error norm between the measured vibration velocity time history data and the vibration velocity time history data calculated by the three-dimensional finite element numerical model under equivalent dynamic load to the global minimum value; and to call the global optimization algorithm to iteratively solve the objective function until the objective function converges, thereby obtaining the equivalent dynamic load time history acting on the tunnel surrounding rock.

[0043] The results output and visualization module is used to display, store, and output the equivalent dynamic load time history and surrounding rock dynamic response cloud map obtained by inversion.

[0044] In this solution, the system integrates on-site data acquisition, parametric numerical modeling, automated dynamic calculation, intelligent optimization and inversion, and visualized output into a unified platform using a modular architecture. Data is seamlessly connected between modules, forming a fully automated closed loop from measured signals to final load results. This eliminates operational errors and subjectivity caused by manual modeling and trial calculations, lowers the barrier to entry for on-site technicians, and improves the efficiency and repeatability of inversion analysis, enabling high-precision load inversion technology to be applied on a large scale in actual tunnel engineering.

[0045] The beneficial effects of this invention are:

[0046] (1) Overcoming the challenges of engineering measurement:

[0047] This invention uses the vibration velocity response of the surrounding rock surface, which is easy to measure, to invert the dynamic load that is difficult to measure directly. It eliminates the need to install stress sensors inside the surrounding rock that are easily damaged by blasting. This effectively solves the technical bottlenecks of direct measurement of dynamic load in the surrounding rock of deep-buried tunnels, such as the difficulty of installation, low survival rate, and high cost. It has the advantages of strong engineering feasibility and wide applicability.

[0048] (2) Improve the reliability of dynamic calculations:

[0049] This invention uses viscoelastic boundaries to simulate the radiation damping effect of surrounding rock in an infinite domain, while also considering the nonlinear constitutive relationship of the surrounding rock dynamics. It realistically reflects the propagation, reflection, and scattering of blasting shock waves and the dynamic interaction between the surrounding rock and the support structure. It solves the calculation distortion problem caused by traditional fixed boundaries, linear elastic models, and simplified load inputs, making the load-response mapping relationship more consistent with engineering practice. It is suitable for deep-buried tunnels under complex geological conditions.

[0050] (3) High inversion accuracy:

[0051] This invention constructs a weighted objective function based on the sum of squared time history errors of vibration velocity. By allocating weights, it highlights key stress-bearing parts and high signal-to-noise ratio signals, suppressing noise interference and local waveform distortion. Combined with a global optimization algorithm, it achieves global optimization, avoids local optima, and can accurately restore the peak value, duration, and spatial distribution characteristics of blasting loads, significantly improving the accuracy of dynamic load inversion.

[0052] (4) Highly instructive:

[0053] The accurate dynamic load of the surrounding rock obtained by the inversion of this invention can more accurately assess the dynamic safety of the support structure, provide quantitative and reliable data support for the design of tunnel dynamic support, optimization of blasting parameters and formulation of safety control standards, and effectively improve the level of safety management and control in tunnel construction.

[0054] (5) High degree of systematization:

[0055] This invention integrates the inversion calculation method into a system, realizing an automated process from data acquisition, modeling, inversion to result output, improving analysis efficiency, reducing errors caused by manual intervention, and facilitating its application in engineering sites. Attached Figure Description

[0056] Figure 1 This is a flowchart of the dynamic load inversion calculation method for the surrounding rock of a deep-buried tunnel under blasting disturbance in this invention.

[0057] Figure 2 This is a structural diagram of the dynamic load inversion calculation system for the surrounding rock of a deep-buried tunnel under blasting disturbance in this invention.

[0058] Figure 3 This is a schematic diagram of the tunnel monitoring section sensor layout in an embodiment of the present invention. Detailed Implementation

[0059] This invention aims to provide a method and system for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance, solving the problem that the dynamic load of surrounding rock in deep-buried tunnel blasting construction is difficult to measure directly, while existing simulation methods have low accuracy. Its core idea is: based on the inherent correlation between load and vibration response in structural dynamics, the dynamic load of surrounding rock under deep-buried tunnel blasting conditions, which cannot be directly monitored, is transformed into a physical quantity that can be indirectly determined by the vibration velocity of the surrounding rock surface; by establishing a three-dimensional numerical model that can truly reflect the dynamic characteristics of the surrounding rock, the boundary radiation damping effect, and the dynamic interaction between the surrounding rock and the support structure, a target function that minimizes error is constructed based on measured vibration data, and automatically iteratively approximated using a global optimization algorithm, ultimately obtaining the equivalent dynamic load time history consistent with the actual blasting action; simultaneously, the entire inversion process is integrated into an automated system, realizing integrated implementation from on-site monitoring to load output, thereby providing accurate and reliable load basis data for tunnel blasting construction and dynamic support design.

[0060] In practical implementation, the method for inverting the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance provided by this invention is described in [reference needed]. Figure 1 The implementation steps include the following:

[0061] S1. Vibration velocity sensors are deployed at typical monitoring sections of deep-buried tunnels to collect time history data of vibration velocity at key parts of the surrounding rock during blasting operations.

[0062] This step aims to obtain the actual dynamic response time history data of key parts of the tunnel surrounding rock under blasting excavation, so as to provide real, reliable and strongly constrained measured input conditions for subsequent dynamic load inversion.

[0063] Specifically, high-precision vibration velocity sensors are deployed at key locations in the surrounding rock (arch crown, arch waist, and sidewalls) of typical monitoring sections of deeply buried tunnels (areas with complex geological conditions or maximum burial depth). During tunnel blasting and excavation, vibration velocity time-history data in three orthogonal directions at these monitoring points are collected simultaneously. .

[0064] S2. Based on the tunnel geometric parameters, surrounding rock mechanical parameters, and support parameters, establish a three-dimensional finite element numerical model that can reflect the interaction between the surrounding rock and the support structure;

[0065] This step aims to construct a dynamic calculation model of the surrounding rock-support structure that is consistent with the actual engineering situation, accurately reproduce the propagation law of blasting load in the surrounding rock and the interaction with the structure, and establish the calculation relationship between load and response.

[0066] Specifically, based on the tunnel engineering design drawings and geological survey reports, the physical and mechanical parameters of the surrounding rock and the parameters of the support structure (shotcrete thickness, anchor bolt parameters, etc.) are determined. Using finite element software or a self-developed program, a three-dimensional finite element numerical model is established, including the tunnel excavation outline, support structure, and a sufficiently large area of ​​surrounding rock. The model boundary uses viscoelastic boundaries to absorb scattered waves, simulating an infinite domain, and the material model considers the dynamic nonlinear characteristics of the surrounding rock.

[0067] S3. To minimize the error norm between the measured vibration velocity time history data and the vibration velocity time history data calculated by the three-dimensional finite element numerical model under equivalent dynamic load, an objective function for inversion analysis is established.

[0068] This step aims to establish a quantitative error evaluation criterion between the measured vibration response and the numerical simulation response, transforming the inversion problem into a solvable optimization problem, and providing a clear convergence objective and calculation basis for iterative optimization.

[0069] Specifically, an objective function for the inversion analysis is defined, which characterizes the overall difference between the measured vibration data and the numerical simulation data. Its preferred mathematical expression is:

[0070] ;

[0071] in, The equivalent dynamic load time history to be inverted (discretized into vector form); This represents the total number of measurement points. The number of measurement directions (usually 3); For monitoring duration; Weighting coefficients are assigned to data from different measurement points and directions to emphasize key measurement points or data with high signal-to-noise ratios.

[0072] objective function The smaller the value, the more likely the inverted load is to be applied. This makes the numerical simulation results closer to the measured data.

[0073] The equivalent dynamic load time history Specifically, this refers to the time history of the surface pressure load acting on the tunnel face (or the current excavation profile). This load is assumed to be a surface force uniformly distributed along the normal direction of the tunnel face or spatially distributed according to the blasting design, used to equivalently simulate the instantaneous loading effect of the blasting shock wave on the surrounding rock. This is achieved through inversion... By applying dynamic time history analysis to the face of the tunnel (or the free face of the excavation) in the numerical model, the dynamic response of the surrounding rock that closely matches the measured vibration velocity can be obtained.

[0074] The rationale for this equivalent load application method lies in the fact that the blasting load mainly acts rapidly on the surrounding rock at the tunnel face through gas expansion and shock waves near the blast source, manifesting as an instantaneous surface pressure pulse. However, directly measuring this load at monitoring sections far from the blast source is difficult. Therefore, using an equivalent surface load simplifies the inversion process while preserving the main impact characteristics of the load (such as peak value, rise time, and duration). If the blasting design involves multiple millisecond delays initiation, the equivalent load can be further discretized into multiple sub-load time histories and superimposed on the corresponding blasting area.

[0075] The weighting coefficient This value is used to adjust the relative importance of the contributions of different measuring points and directions to the objective function, and its preferred range is 0.5 to 2.0. An exemplary selection principle is as follows:

[0076] (1) Based on the importance of the measuring point location:

[0077] The measurement points at key stress-bearing parts such as the arch crown and arch waist have higher weights (e.g., W=1.5~2.0), while the measurement points at sidewalls or secondary parts have lower weights (e.g., W=0.8~1.0), in order to highlight areas that have a greater impact on tunnel stability.

[0078] (2) Based on signal quality:

[0079] Measurement points with high signal-to-noise ratio, complete waveform, and no obvious interference have higher weights (e.g., W=1.2~1.5), while measurement points with higher noise or severe signal attenuation have lower weights (e.g., W=0.5~0.8).

[0080] (3) Based on the sensitivity to vibration direction:

[0081] The radial vibration component is more sensitive to the blasting impact response, so its weight can be increased accordingly.

[0082] In preliminary analysis or when there is no prior information, all weight coefficients can be simplified. Furthermore, the weighting coefficients can be adjusted through pre-analysis (such as sensitivity analysis) or empirical iteration to further improve the robustness and accuracy of the inversion results.

[0083] S4. Using a global optimization algorithm, the objective function is iteratively solved until the objective function converges, thereby obtaining the time history of the equivalent dynamic load acting on the tunnel surrounding rock.

[0084] This step aims to automatically search for the optimal solution in the complete solution space using an intelligent optimization algorithm, so that the simulated response and the measured response achieve the best match, and finally obtain the equivalent dynamic load time history of the surrounding rock that conforms to the actual blasting mechanism.

[0085] Specifically, a global optimization algorithm is used as the solution engine to invert the loads to be retrieved. As optimization variables, the objective function The fitness function is used for iterative solving. An exemplary optimization process is as follows:

[0086] a) Initialize a set of candidate load vectors (population);

[0087] b) For each candidate load By substituting the values ​​into the numerical model and performing dynamic time history analysis, the simulated vibration velocity at each measuring point was calculated. ;

[0088] c) Calculate the objective function value for each candidate load based on the simulated and measured values. ;

[0089] d) Optimization algorithm based on The size (fitness) is used to perform selection, crossover, and mutation operations to generate a new generation of candidate loads;

[0090] e) Repeat steps bd until the objective function value is less than the preset threshold or the maximum number of iterations is reached. At this point, the optimal candidate load is the equivalent dynamic load time history obtained by inversion.

[0091] The structure of the dynamic load inversion calculation system for the surrounding rock of deep-buried tunnels under blasting disturbance provided by this invention is shown in the figure. Figure 2 It comprises several modules: data acquisition, numerical simulation, inversion analysis, and results output and visualization. The specific functions of each module are as follows:

[0092] Data acquisition module: Responsible for communicating with the field sensor array, preferably a high-precision triaxial vibration velocity sensor. This module connects to the sensors via wired or wireless means, supports multi-channel synchronous triggering acquisition, and has data preprocessing and remote real-time transmission functions. The acquired data is stored in binary or CSV format for easy retrieval later.

[0093] Numerical Simulation Module: Integrates the kernel of finite element analysis software, supporting parametric modeling. Automation is achieved through a script interface: based on input geological parameters, support parameters, and model dimensions, it automatically generates meshes, sets viscoelastic artificial boundaries, defines material nonlinear constitutive models, and performs explicit or implicit dynamic time-history analysis. The module supports batch submission of calculation tasks and outputs vibration velocity time-history data for specified measurement points.

[0094] Inversion Analysis Module: This is the core of the system. The main control program is implemented using a high-level programming language, integrating an optimization algorithm library. This module features automated iteration capabilities: it automatically reads measured data from the data acquisition module, calls the numerical simulation module to perform multiple dynamic calculations, calculates the objective function value in real time, and updates load assumptions until convergence; it supports parallel computing to accelerate the iteration process.

[0095] The results output and visualization module is implemented based on a graphical user interface, supporting the dynamic display of the equivalent dynamic load time history curves obtained from the inversion, the surrounding rock vibration velocity comparison curves, and stress / displacement / velocity field contour animations. It also supports data export functions, including generating analysis reports and data files in PDF or Word formats.

[0096] Example:

[0097] This embodiment takes the application of the present invention to a deep-buried railway tunnel project as an example, and its implementation is as follows:

[0098] On-site data monitoring:

[0099] like Figure 3 As shown, at the selected monitoring section, three-dimensional vibration velocity sensors were installed at five locations: the tunnel arch (1#), the left and right arch waists (2#, 3#), and the left and right sidewalls (4#, 5#). During a full-section blasting excavation, the vibration velocity time history of all sensor channels was recorded simultaneously, with the recording time continuously covering the entire blasting vibration process.

[0100] Numerical model establishment:

[0101] Based on the Class IV surrounding rock conditions of the tunnel (elastic modulus 2 GPa, Poisson's ratio 0.3, density 2500 kg / m³) and the initial support parameters (C25 shotcrete, 25 cm thick), a three-dimensional model of a circular tunnel with a diameter of 12 m was established. The model's range was more than five times the tunnel diameter to eliminate boundary effects. Fixed constraints were used at the bottom, and viscoelastic boundaries were set around the perimeter and top. An elastoplastic dynamic constitutive model was adopted for the surrounding rock.

[0102] Objective function construction: In this example, the objective function is defined as the sum of the squares of the time history residuals of vibration velocities in the three directions at each measuring point:

[0103] ;

[0104] To simplify calculations, all weight coefficients are temporarily assumed. =1.

[0105] In another embodiment, according to Figure 3 The sensor placement locations are shown. The weights for the three directions of the arch top measuring point (1#) are set to 1.5, the weights for the arch waist measuring points (2#, 3#) are set to 1.2, and the weights for the side wall measuring points (4#, 5#) are set to 1.0. If the signal-to-noise ratio (SNR) in a certain direction is lower than the threshold (e.g., SNR < 10), the corresponding weight is reduced to 0.6.

[0106] After adopting the above weights, the matching degree (correlation coefficient) between the load time history obtained by inversion and the measured vibration velocity is improved by about 8% to 15% compared with simplifying each weight coefficient to 1, and the dynamic response simulation of surrounding rock is more accurate.

[0107] Dynamic load inversion:

[0108] A genetic algorithm with real-number encoding was used as the optimization tool. The equivalent dynamic load time history (duration 0.1s, sampling interval 0.001s, total 100 points) was discretized into a 100-dimensional vector P. The population size was set to 50, and the maximum number of generations was set to 200. The optimization process was executed automatically by a computer program.

[0109] For each individual in each generation (one load assumption) The program automatically applies it as a surface force load to the excavation surface of the numerical model, runs dynamic analysis, extracts the simulated vibration velocity at each measuring point, and calculates the objective function value. The genetic algorithm performs selection, crossover, and mutation based on J(P) to generate a new generation of population. After 150 generations of iteration, the objective function value converges to below the stability threshold, and the best individual at this point is the optimal equivalent dynamic load time history obtained through inversion.

[0110] In other embodiments, the global optimization algorithm may also employ particle swarm optimization (PSO) or simulated annealing (SA).

[0111] When using the particle swarm optimization algorithm, the equivalent dynamic load vector will be... Considering particle positions, the population size is set to 40-60 particles, with inertial weights. The learning factor linearly decays from 0.9 to 0.4. The optimization process includes: initializing particle positions and velocities; for each particle, setting its position... Substitute the values ​​into the numerical model to calculate the simulated vibration velocity and evaluate the objective function. As fitness; update particle velocity and position based on individual optimality and global optimality; repeat iterations until convergence conditions are met.

[0112] When using the simulated annealing algorithm, starting from an initial load vector Begin by setting the initial temperature. Higher (e.g., 1000), cooling coefficient Multiple perturbations are performed at each temperature to generate new solutions. ,like Accept if necessary, otherwise proceed by probability. Accept; gradually reduce the temperature until it falls below the termination threshold.

[0113] The convergence condition is preferably one or a combination of the following:

[0114] (1) Objective function value Less than a preset threshold (e.g.) (or the relative error relative to the initial value is less than 5%).

[0115] (2) The root mean square error (RMSE) between the measured vibration velocity and the simulated vibration velocity is less than the preset value (e.g., 0.05 cm / s).

[0116] (3) The improvement in the objective function over multiple generations (e.g., 10 generations) is less than 1%;

[0117] (4) Reach the maximum number of iterations (e.g., 300 generations).

[0118] By applying the above convergence conditions, we can ensure that the inversion process terminates stably, avoid over-calculation, and guarantee the accuracy of the inversion.

[0119] In this embodiment, for the dynamic load inversion calculation system of deeply buried tunnel surrounding rock under blasting disturbance, the data acquisition module uses a 941B triaxial velocity sensor, connected to an industrial-grade data acquisition instrument (sampling rate 2048Hz) via RS485 bus, supporting 4G remote transmission. The numerical simulation module uses ABAQUS Explicit as its kernel, employing Python scripts to implement parametric modeling and batch processing submission (each dynamic analysis takes approximately 30 minutes). The inversion analysis module runs in the MATLAB R2023a environment, integrating the Global Optimization Toolbox. The results output module displays load curves and ABAQUS post-processing contour plots through the MATLAB GUI, and can export an engineering report containing all results with one click.

[0120] Through this embodiment, the peak dynamic load and time history of the surrounding rock under the blasting action were successfully inverted. Compared with the result estimated based on the empirical formula of explosive quantity, the load obtained by the present invention rises more steeply and decays faster, and can better reflect the impact characteristics of the near-blast source.

[0121] Although embodiments of the present invention have been described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present invention, and all such changes and alterations shall not depart from the protection scope of the present invention.

Claims

1. A method for calculating dynamic load inversion of surrounding rock of a deep-buried tunnel under blasting disturbance, characterized in that, Includes the following steps: S1. Vibration velocity sensors are deployed at typical monitoring sections of deep-buried tunnels to collect time history data of vibration velocity at key parts of the surrounding rock during blasting operations. S2. Based on the tunnel geometric parameters, surrounding rock mechanical parameters, and support parameters, establish a three-dimensional finite element numerical model that can reflect the interaction between the surrounding rock and the support structure; S3. The objective function of the inversion analysis is established with the goal of iteratively converging the error norm between the measured vibration velocity time history data and the vibration velocity time history data calculated by the three-dimensional finite element numerical model under equivalent dynamic load to the global minimum. The objective function for the established inversion analysis is: ； in, The objective function is... The equivalent dynamic load time history to be inverted; For the first The first measuring point Time histories of vibration velocities measured in each direction; For a three-dimensional finite element numerical model under load The vibration velocity time history calculated for the corresponding measuring point and direction under the action; These are the weighting coefficients; This represents the number of measurement points. Number of directions; S4. Using a global optimization algorithm, the objective function is iteratively solved until the objective function converges, thereby obtaining the time history of the equivalent dynamic load acting on the tunnel surrounding rock.

2. The method for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance as described in claim 1, characterized in that, In step S1, the typical monitoring section is selected from the section with complex geological conditions or the section with the maximum burial depth. The key parts of the surrounding rock include the tunnel's arch, arch waist, and sidewalls. The section with complex geological conditions includes the section with fractured surrounding rock, the section with developed joints and fissures, the section with weak interlayers, or the section with fault zones. The vibration velocity sensors are deployed as follows: they are placed along the tunnel arch, left and right arch waists, and left and right sidewalls; each vibration velocity sensor monitors vibration velocity components in at least three directions: radial, tangential, and axial.

3. The method for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance as described in claim 1, characterized in that, The three-dimensional finite element numerical model uses viscoelastic boundaries to simulate the radiation damping effect of an infinite domain foundation and employs a nonlinear dynamic constitutive model to describe the mechanical properties of the surrounding rock.

4. The method for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance as described in claim 1, characterized in that, weighting factors in the objective function Depending on the importance of the measurement point location, the signal signal-to-noise ratio or the sensitivity of the vibration direction: The tunnel arch crown and arch waist are designated as primary important measuring points, and the sidewalls are designated as secondary important measuring points. The weight coefficients for primary important measuring points range from 1.5 to 2.0, and the weight coefficients for secondary important measuring points range from 0.8 to 1.

0. For measurement points with a signal-to-noise ratio (SNR) greater than or equal to a preset SNR threshold, the weighting coefficient ranges from 1.2 to 1.5; for measurement points with a SNR less than the preset SNR threshold, the weighting coefficient ranges from 0.5 to 0.

8. The radial direction is defined as the sensitive vibration direction of the blasting impact. The radial direction weighting coefficient ranges from 1.2 to 1.5, while the tangential and axial direction weighting coefficients range from 0.8 to 1.

0.

5. The method for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance as described in claim 1, characterized in that, In step S4, during the iterative solution process, the current candidate equivalent dynamic load time history is applied to the face or excavation profile of the three-dimensional finite element numerical model in the form of surface pressure load to calculate the corresponding vibration velocity time history data.

6. The method for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance as described in claim 5, characterized in that, The surface pressure load is uniformly distributed along the normal direction of the tunnel face, or spatially non-uniformly distributed according to the arrangement of the blasting holes; when the blasting adopts multi-stage delayed detonation, the equivalent dynamic load time history is the superposition of multiple sub-load time histories.

7. The method for inverting and calculating the dynamic load of surrounding rock in deep-buried tunnels under blasting disturbance as described in claim 1, characterized in that, In step S4, the global optimization algorithm adopts a genetic algorithm, a particle swarm optimization algorithm, or a simulated annealing algorithm.

8. The method according to any one of claims 1 to 7, wherein the method is characterized by, In step S4, the conditions for the convergence of the objective function include: the objective function value is less than a preset threshold, the improvement of the objective function after multiple consecutive iterations is less than a preset percentage, or the maximum number of iterations is reached.

9. The system for calculating the dynamic load of the surrounding rock of a deep buried tunnel under blasting disturbance, which is used for realizing the method for calculating the dynamic load of the surrounding rock of a deep buried tunnel under blasting disturbance according to any one of claims 1-8, characterized in that, The system includes: The data acquisition module is used to control the vibration velocity sensors deployed at typical monitoring sections of deep-buried tunnels to collect and store vibration velocity time history data during the blasting process. The numerical simulation module is used to establish a three-dimensional finite element numerical model that reflects the interaction between the surrounding rock and the support structure, based on the tunnel's geometric parameters, surrounding rock mechanical parameters, and support parameters. The inversion analysis module is used to establish an objective function for inversion analysis by iteratively converging the error norm between the measured vibration velocity time history data and the vibration velocity time history data calculated by the three-dimensional finite element numerical model under equivalent dynamic load to the global minimum value; and to call the global optimization algorithm to iteratively solve the objective function until the objective function converges, thereby obtaining the equivalent dynamic load time history acting on the tunnel surrounding rock. The results output and visualization module is used to display, store, and output the equivalent dynamic load time history and surrounding rock dynamic response cloud map obtained by inversion.

Images