Deep-buried tunnel blasting control instruction generation method and system based on cross-scale data

Abstract

The application relates to the technical field of data processing, and discloses a deep-buried tunnel blasting control instruction generation method and system based on cross-scale data, aiming to solve the problems that rock mass damage is uncontrollable, support failure risk is high, and parameter optimization is lagged in the prior art. The scheme mainly comprises the following steps: real-time monitoring of rock mass crack density, blasting vibration velocity, support structure strain and blasting thermal disturbance temperature rise, determination of rock mass breakage vibration frequency band, support structure resonance frequency band and power spectral density of support structure strain corresponding signals; calculation of rock mass damage index, rock mass breakage vibration energy, support structure vibration energy, peeling index and support time-varying stiffness; determination of a blasting parameter set based on a genetic algorithm, prediction of a predicted rock mass damage index at a future time; correction of the blasting parameters in the blasting parameter set and determination of corresponding engineering measures to generate a blasting control instruction. The application improves the accuracy of the control instruction, the safety of the tunnel construction and the timeliness of the control response.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a deep-buried tunnel blasting control instruction generation method and system based on cross-scale data. BACKGROUND

[0002] With the large-scale construction of deep-buried tunnel projects in the western plateau region, the full-face drilling and blasting method is widely used due to its strong geological adaptability. However, the problem of dynamic disaster chain in high ground stress environment is increasingly prominent: excavation unloading leads to tangential stress concentration of surrounding rock, and under the disturbance of blasting dynamic load, micro-crack propagation is induced, when the disturbance energy exceeds the energy storage threshold of rock mass, it will trigger time-delay rock burst, causing serious consequences such as overbreak rate and high support failure rate.

[0003] In the process of deep-buried tunnel blasting construction, blasting energy control is directly related to the stability of surrounding rock and the safety of supporting structure. The traditional blasting parameter design mainly relies on geological exploration experience and static mechanical model, which has the following key defects: first, the problem of multi-source data fragmentation: existing technologies usually monitor single parameters such as blasting vibration velocity, crack development or support strain independently, and lack of coordinated analysis of cross-scale dynamic response of rock mass-support system. For example, the key factors such as the weakening effect of blasting thermal disturbance on rock mass strength, rock mass damage induced by seepage field changes, and the coupling mechanism of support structure time-varying stiffness and blasting vibration are not systematically integrated, leading to one-sidedness of parameter decision. Second, the static defect of damage assessment: rock mass damage is mainly based on instantaneous vibration velocity threshold or empirical formula, without considering the dynamic interaction of blasting cumulative effect and environmental disturbance; the existing damage model ignores the correction of temperature on rock mass failure strain, leading to deviation of damage index calculation from actual working conditions, and aggravating the risk of support failure. Third, the lack of support-rock mass collaborative control: the stiffness of support structure is often regarded as a constant value, and its quantitative correlation with rock mass damage is not established; at the same time, the energy distribution of support structure resonance frequency band and rock mass fracture frequency band is not used for dynamic early warning, which is easy to cause structure resonance damage. Fourth, parameter optimization and blasting control lag: blasting parameter adjustment relies on manual trial and error, lacking intelligent optimization mechanism based on real-time data; existing methods cannot predict damage trend in advance through time series data, leading to slow response of engineering measures and difficulty in controlling crack propagation or peeling risk.

[0004] In summary, due to data fragmentation, static model, experience dependence and prediction deficiency, the existing technology has the problems of uncontrollable rock mass damage, high risk of support failure and lag of parameter optimization in blasting control. Therefore, it is urgent to establish a blasting control instruction generation method that integrates multi-source cross-scale data, dynamically couples rock mass-support response, integrates intelligent optimization and risk prediction, to realize safe and efficient construction of deep-buried tunnel. SUMMARY

[0005] The present application aims to solve the problems of uncontrollable rock mass damage, high risk of support failure and parameter optimization lag in the existing blasting control mode, and proposes a deep-buried tunnel blasting control instruction generation method and system based on cross-scale data.

[0006] The technical scheme adopted by the present application to solve the above technical problems is:

[0007] In a first aspect, the present application provides a deep-buried tunnel blasting control instruction generation method based on cross-scale data, which comprises:

[0008] Real-time monitoring of rock mass crack density, blasting vibration velocity, support structure strain and blasting thermal disturbance temperature rise, and determination of rock mass breakage vibration frequency band, support structure resonance frequency band and power spectral density of support structure strain corresponding signals;

[0009] According to the rock mass plastic strain, the rock mass damage index is calculated according to the rock mass plastic strain and the blasting thermal disturbance temperature rise, the rock mass breakage vibration energy is calculated according to the rock mass breakage vibration frequency band, the support structure vibration energy is calculated according to the support structure resonance frequency band, the stripping index is calculated according to the power spectral density, and the support time-varying stiffness is calculated;

[0010] Taking the minimum ratio of rock mass damage index to support time-varying stiffness as the optimization goal, the genetic algorithm is used to search for the optimal blasting parameter set in the blasting parameter space, and the predicted rock mass damage index at the future time is predicted according to the time series data of rock mass damage index, support time-varying stiffness and blasting vibration velocity and based on LSTM;

[0011] Based on the rock mass crack density, rock mass damage index, rock mass breakage vibration energy, support structure vibration energy, stripping index and predicted rock mass damage index, the blasting parameters in the blasting parameter set are corrected and the corresponding engineering measures are determined, and the blasting control instruction is generated according to the corrected blasting parameter set and the determined engineering measures.

[0012] Further, the detection method of the rock mass crack density comprises:

[0013] Real-time monitoring of microseismic energy, calculation of rock mass crack density according to the microseismic energy, and calculation formula as follows:

[0014] ;

[0015] Wherein, represents the rock mass crack density, represents the microseismic energy, represents the natural logarithm.

[0016] Further, the inversion formula of the rock mass plastic strain is as follows:

[0017] ;

[0018] The method for calculating the rock mass damage index comprises: determining the rock mass failure strain according to the uniaxial compressive strength of rock and a dynamic load constitutive model, correcting the rock mass failure strain according to the blasting thermal disturbance temperature rise, and calculating the rock mass damage index according to the rock mass plastic strain, the blasting thermal disturbance temperature rise and the corrected rock mass failure strain.

[0019] The correction formula of the rock mass failure strain is as follows:

[0020] ;

[0021] The calculation formula of the rock mass damage index is as follows:

[0022] ;

[0023] ;

[0024] ;

[0025] wherein, represents the rock mass damage index, represents the blasting vibration velocity, represents the rock mass plastic strain, represents the rock mass failure strain, represents the corrected rock mass failure strain, represents a temperature softening factor, represents the blasting thermal disturbance temperature rise, represents a seepage weakening factor, represents the current permeability coefficient, represents the initial permeability coefficient, represents a lithology scale coefficient, represents a lithology shape coefficient, represents a natural logarithm, represents an exponential function.

[0026] Further, the calculation formula of the support time-varying stiffness is as follows:

[0027] ;

[0028] wherein, represents the support time-varying stiffness at the moment, represents the support initial stiffness, represents a concrete age strength factor, represents the support age, represents a stiffness growth time constant, represents a natural logarithm.

[0029] Further, the set of blasting parameters is generated as follows:

[0030] ;

[0031] wherein, represents the set of blasting parameters, represents the amount of blasting explosive, represents the blast hole spacing, represents the blast time interval, represents the rock mass damage index, represents the time-varying stiffness of the support at the moment, represents finding a set of blasting parameters to minimize .

[0032] Further, the rock mass vibration frequency band is (100Hz, ∞), and the support structure resonance frequency band is (0, 50Hz);

[0033] The calculation formula of the rock mass fracture vibration energy is as follows:

[0034] ;

[0035] The calculation formula of the support structure vibration energy is as follows:

[0036] ;

[0037] wherein, represents the rock mass fracture vibration energy, represents the rock mass fracture vibration power spectral density, represents the support structure vibration energy, represents the support structure vibration power spectral density, represents the frequency.

[0038] Further, when calculating the stripping index, the stripping characteristic frequency band is (8Hz, 12Hz);

[0039] The calculation formula of the stripping index is as follows:

[0040] ;

[0041] wherein, represents the stripping index, represents the power spectral density of the support structure strain corresponding signal.

[0042] Further, the blasting parameters in the set of blasting parameters are corrected, including:

[0043] if the rock mass damage index or the predicted rock mass damage index is greater than the first threshold value, then the amount of blasting explosive is reduced according to a first preset proportion, and the blast hole spacing is reduced according to a second preset proportion;

[0044] if the rock mass fracture vibration energy is greater than the second threshold value, then the amount of blasting explosive is reduced according to a first preset proportion, the blast hole spacing is reduced according to a second preset proportion, and the blasting time interval is increased according to a preset time length;

[0045] if the rock mass fracture density is greater than the third threshold value, then the blast hole spacing is reduced according to a third preset proportion.

[0046] Further, corresponding engineering measures are determined, including:

[0047] if the predicted rock mass damage index is greater than the first threshold value or the rock mass fracture vibration energy is greater than the second threshold value, then the corresponding engineering measure is to start nano-silicon grouting;

[0048] if the rock mass fracture density is greater than the third threshold value, then the corresponding engineering measure is to shorten the single-cycle footage.

[0049] if the support structure vibration energy is greater than the fourth threshold value or the stripping index is greater than the fifth threshold value, then the corresponding engineering measure is to spray a concrete compensation layer;

[0050] if the blasting thermal disturbance temperature rise is greater than the sixth threshold value, then the corresponding engineering measure is to add a low-temperature adjusting agent to the grouting liquid.

[0051] In a second aspect, the present application provides a deep-buried tunnel blasting control instruction generation system based on cross-scale data, which is used to realize the deep-buried tunnel blasting control instruction generation method based on cross-scale data as described in the first aspect, and the system comprises:

[0052] a multi-scale data acquisition unit, which is used to monitor the rock mass fracture density, the blasting vibration velocity, the support structure strain, and the blasting thermal disturbance temperature rise in real time, and determine the rock mass fracture vibration frequency band, the support structure resonance frequency band, and the power spectral density of the support structure strain corresponding signal;

[0053] a mutual feedback analysis unit, which is used to inverse the rock mass plastic strain according to the blasting vibration velocity, calculate the rock mass damage index according to the rock mass plastic strain and the blasting thermal disturbance temperature rise, calculate the rock mass fracture vibration energy according to the rock mass fracture vibration frequency band, calculate the support structure vibration energy according to the support structure resonance frequency band, calculate the stripping index according to the power spectral density, and calculate the support time-varying stiffness;

[0054] The decision application unit is used for minimizing the ratio of the rock mass damage index to the support time-varying stiffness as an optimization target, searching for an optimal blasting parameter set in a blasting parameter space based on a genetic algorithm, and predicting a predicted rock mass damage index at a future time according to time sequence data of the rock mass damage index, the support time-varying stiffness and the blasting vibration speed and based on an LSTM. The blasting parameter set is corrected and corresponding engineering measures are determined based on the rock mass crack density, the rock mass damage index, the rock mass breakage vibration energy, the support structure vibration energy, the stripping index and the predicted rock mass damage index, and blasting control instructions are generated according to the corrected blasting parameter set and the determined engineering measures.

[0055] The method and system for generating blasting control instructions for deep-buried tunnels based on cross-scale data provided by the present application fuse multi-scale data such as rock mass crack density, blasting vibration speed, support structure strain and blasting thermal disturbance temperature rise, cover internal rock mass damage, dynamic response and thermal effect, ensure that blasting control decisions are based on multi-source information and reduce one-sidedness, provide quantitative evaluation through calculation of damage indexes and vibration energy, help to adjust parameters in a timely manner, realize dynamic control, avoid cumulative damage and improve the safety of tunnel construction, minimize the ratio of the rock mass damage index to the support time-varying stiffness as an optimization target and use a genetic algorithm to search for an optimal blasting parameter set, balance rock mass damage and support structure stability, ensure that blasting is carried out within a safe range, prevent support failure or excessive rock mass damage, and the genetic algorithm can efficiently find a global optimal solution or an approximate optimal solution, improving blasting efficiency and economy, predict future rock mass damage indexes based on an LSTM model, realize preventive control, adjust blasting parameters in advance, avoid potential risks and enhance system robustness, correct blasting parameters and determine engineering measures based on multiple indexes, make blasting control more suitable for actual conditions, reduce unexpected events, ensure construction continuity and quality. In addition, the present application starts from real-time monitoring, goes through data calculation, optimization search, prediction correction and finally generates control instructions, each link is tightly coupled, monitoring data drives index calculation, index calculation supports the optimization target, the optimization result is input into the prediction model, the prediction output guides parameter correction, and the corrected instruction affects the next round of monitoring, thereby forming a control closed loop. At the same time, the comprehensiveness of cross-scale data, the global optimization ability of the genetic algorithm and the time sequence prediction advantage of the LSTM are utilized to ensure rapid and accurate response and avoid the data lag problem in traditional methods. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 A flowchart of the method for generating blasting control instructions for deep-buried tunnels based on cross-scale data provided by the present application is shown in the figure.

[0057] Figure 2 A structural diagram of the system for generating blasting control instructions for deep-buried tunnels based on cross-scale data provided by the present application is shown in the figure. DETAILED DESCRIPTION

[0058] Since the current blasting parameter design scheme mainly relies on the geological exploration experience and the static mechanical model, the rock mass damage is uncontrollable, the supporting failure risk is high, and the parameter optimization is lagged in the blasting control.

[0059] Therefore, the technical scheme of the present application is proposed. In the present application, first, multi-dimensional data such as rock mass crack density, blasting vibration velocity, supporting structure strain and blasting thermal disturbance temperature rise are monitored in real time, and rock mass fracture vibration frequency band, supporting structure resonance frequency band and power spectrum density are extracted therefrom. These data cover the internal damage, dynamic response and thermal effect of the rock mass, forming a cross-scale data basis, and avoiding the limitations of a single data source through multi-source fusion; at the same time, by calculating the rock mass fracture vibration energy, the supporting structure vibration energy and the peeling index, the original data are converted into quantitative engineering indexes, providing unified and comparable input for subsequent decision-making. Then, by calculating the rock mass damage index and the supporting time-varying stiffness, the rock mass damage index quantifies the cumulative damage degree of the rock mass caused by blasting, and the supporting time-varying stiffness reflects the dynamic bearing capacity of the supporting structure. The present application takes the minimum ratio of the two as the optimization goal, and the essence is to balance the minimum rock mass damage and the maximum supporting stability. The minimum ratio means that under the given blasting conditions, the optimal toughness state of the rock mass-supporting system is sought, so as to avoid the collapse risk caused by excessive rock mass damage and the resonance failure caused by insufficient supporting stiffness; at the same time, the genetic algorithm is used to solve the optimal blasting parameter set, improving the solving efficiency. Then, using the time series data of the rock mass damage index, the supporting time-varying stiffness and the blasting vibration velocity, the future rock mass damage index is predicted through the LSTM (Long Short Term Memory) model. By predicting the future state, the blasting parameters can be corrected in advance and engineering measures can be triggered, so as to avoid potential risks and enhance the system robustness; finally, based on the multi-indexes such as rock mass crack density, rock mass damage index, rock mass fracture vibration energy, supporting structure vibration energy, peeling index and predicted rock mass damage index, the blasting parameters are corrected and the engineering measures are determined, avoiding the one-sidedness of single parameter adjustment and ensuring the scientificity and operability of the control command.

[0060] The technical solutions in the embodiments will be clearly and completely described below with reference to the drawings in the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments.

[0061] Figure 1 A flowchart of a deep-buried tunnel blasting control instruction generation method based on cross-scale data is shown. Please refer to Figure 1 , which comprises the following steps:

[0062] Step 1, real-time monitoring of rock mass crack density, blasting vibration velocity, support structure strain and blasting thermal disturbance temperature rise, and determining the rock mass fracture vibration frequency band, support structure resonance frequency band and power spectrum density of support structure strain corresponding signal.

[0063] In this embodiment, the excavation surface can be directly scanned by a three-dimensional laser scanner after each cycle of blasting and the rock mass crack density can be counted by Hough transform. The rock mass crack density can also be counted in advance by a three-dimensional laser scanner and Hough transform, while the microseismic energy, i.e. the energy peak value within a preset time period after blasting, is monitored by a microseismic sensor array to construct a relationship between blasting energy and rock mass crack density; in actual application, the microseismic energy is monitored in real time by a microseismic sensor array, and then the rock mass crack density is calculated according to the microseismic energy, and the calculation formula is as follows:

[0064] ;

[0065] wherein, represents the rock mass crack density (unit: piece / cm²), represents the microseismic energy (unit: J), represents the natural logarithm, 0.18 represents the energy-density conversion coefficient (unit: piece / (cm³·lnJ)), and 2.4 represents the original crack density (unit: piece / cm²).

[0066] In this embodiment, the blasting vibration velocity is monitored in real time by a microseismic sensor array, which is arranged along the tunnel axis within the range behind the tunnel face; and the blasting thermal disturbance temperature rise is monitored in real time by an infrared thermal imager array.

[0067] In this embodiment, the FBG strain sensor arranged in a rhombus topology array is used to monitor the support structure strain, and the vertex spacing of the FBG strain sensor is less than or equal to 10 cm. The FBG strain sensor collects the original wavelength signal, which is converted into the support structure strain according to the FBG wavelength change corresponding to the original wavelength signal, and the formula is as follows:

[0068] ;

[0069] ;

[0070] wherein, represents the support structure strain (unit: ), represents the original strain (unit: ), represents the calibration coefficient (unit: / pm), represents the FBG wavelength change (unit: pm), represents the grating incident angle (unit: degree).

[0071] In the embodiment, the rock mass fracture vibration signal is a high-frequency fracture signal, and a corresponding rock mass fracture vibration frequency band is (100 Hz, ∞), that is, greater than 100 Hz; the supporting structure resonance signal is a low-frequency structure vibration, and a corresponding supporting structure resonance frequency band is (0, 50 Hz), that is, less than 50 Hz.

[0072] Step 2, according to the blasting vibration velocity, the rock mass plastic strain is inversed, according to the rock mass plastic strain and the blasting thermal disturbance temperature rise, the rock mass damage index is calculated, according to the rock mass fracture vibration frequency band, the rock mass fracture vibration energy is calculated, according to the supporting structure resonance frequency band, the supporting structure vibration energy is calculated, according to the power spectral density, the stripping index is calculated, and the supporting time-varying stiffness is calculated.

[0073] In the embodiment, the inversion formula of the rock mass plastic strain is as follows:

[0074] ;

[0075] wherein, represents the rock mass plastic strain (dimensionless), represents the blasting vibration velocity, 0.024 represents a vibration velocity-strain conversion coefficient (unit: ), and 0.18 represents a background plastic strain (dimensionless).

[0076] The calculation method of the rock mass damage index comprises the following steps: determining the rock mass failure strain according to the uniaxial compressive strength of rock and a dynamic load constitutive model, correcting the rock mass failure strain according to the blasting thermal disturbance temperature rise, and calculating the rock mass damage index according to the rock mass plastic strain, the blasting thermal disturbance temperature rise, and the corrected rock mass failure strain.

[0077] The correction formula of the rock mass failure strain is as follows:

[0078] ;

[0079] wherein, represents the rock mass failure strain (dimensionless), represents the corrected rock mass failure strain (dimensionless), represents the blasting thermal disturbance temperature rise (unit: ℃), represents a temperature rise weakening coefficient (unit: 1 / ℃), represents a reference temperature (unit: ℃).

[0080] The calculation formula of the rock mass damage index is as follows:

[0081] ;

[0082] ;

[0083] ;

[0084] wherein, represents rock mass damage index (dimensionless), range [0, 1], represents no damage, represents complete damage, represents temperature softening factor (dimensionless), represents seepage weakening factor (dimensionless), represents current permeability coefficient (dimensionless), represents initial permeability coefficient (dimensionless), represents lithology scale coefficient (dimensionless), represents lithology shape coefficient (dimensionless), and Calibration by indoor dynamic loading test Hopkinson pressure bar test, for example, granite Take 1.2, Take 2.5, represents natural logarithm, represents exponential function, 0.02 represents temperature softening index (unit: 1 / ℃), 0.12 represents seepage weakening index (dimensionless).

[0085] In the above formula, the rock mass damage index is corrected by temperature and permeability to obtain a damage more consistent with the actual thermal coupling state, which significantly improves the damage evaluation accuracy of high ground temperature and high permeability pressure conditions and avoids overbreak and underbreak accidents.

[0086] In the present embodiment, based on the rock mass vibration frequency band being (100 Hz, ∞) and the supporting structure resonance frequency band being (0, 50 Hz), the calculation formula of the rock mass breakage vibration energy is as follows:

[0087] ;

[0088] The calculation formula of the supporting structure vibration energy is as follows:

[0089] ;

[0090] wherein, represents rock mass breakage vibration energy (unit: J), represents rock mass breakage vibration power spectral density (unit: J / Hz), represents supporting structure vibration energy (unit: J), represents supporting structure vibration power spectral density (unit: J / Hz), represents frequency (unit: Hz).

[0091] The rock mass and the support can be frequency-synchronized by separating the rock mass fracture and the support resonance signal through the above formula, solving the problem of aliasing misjudgment.

[0092] In this embodiment, when calculating the peeling index, the peeling characteristic frequency band is (8Hz, 12Hz), and the calculation formula of the peeling index is as follows:

[0093] ;

[0094] wherein, represents the peeling index (dimensionless), represents the power spectral density of the support structure strain corresponding signal (unit: με² / Hz).

[0095] In this embodiment, the calculation formula of the support time-varying stiffness is as follows:

[0096] ;

[0097] wherein, represents the support time-varying stiffness at time t (unit: GPa), represents the support initial stiffness (unit: GPa), which is determined by the concrete grade or the steel arch section parameter, represents the concrete age strength factor (dimensionless), and the empirical value is 0.05-0.12, represents the support age (unit: days), represents the stiffness growth time constant (unit: days), and the default value is 1 day, represents the natural logarithm.

[0098] Step 3, taking the minimum ratio of the rock mass damage index to the support time-varying stiffness as the optimization goal, searching for the optimal blasting parameter set in the blasting parameter space based on the genetic algorithm, and predicting the predicted rock mass damage index at the future time according to the time sequence data of the rock mass damage index, the support time-varying stiffness and the blasting vibration speed and based on the LSTM.

[0099] In this embodiment, the blasting parameter set generation logic is as follows:

[0100] ;

[0101] wherein, represents the blasting parameter set, represents the blasting explosive quantity (unit: kg), represents the blasting hole distance (unit: m), represents the blasting time interval (unit: ms), represents the rock mass damage index, represents the support time-varying stiffness at time t, representing finding a set of blasting parameters to minimize .

[0102] Specifically, the embodiment searches for the optimal solution in the blasting parameter space through a genetic algorithm to minimize the damage stiffness ratio. Meanwhile, the historical window data of the rock mass damage index, the support time-varying stiffness, and the blasting vibration velocity are input into the pre-trained LSTM to obtain the predicted rock mass damage index at the future time (such as 200 ms). By predicting the future state, the scheme can prospectively adjust the parameters, and the predicted rock mass damage index obtained by prediction is combined with real-time data (such as rock mass crack density and stripping index) to correct the blasting parameter set. The comparison between the predicted value and the real-time data provides a basis for decision-making, ensuring that the corrected parameter set can timely alleviate potential risks, forming a closed loop of monitoring-prediction-correction, realizing the prevention and adaptability of blasting control, and enhancing the robustness of the system.

[0103] Step 4, based on the rock mass crack density, the rock mass damage index, the rock mass breakage vibration energy, the support structure vibration energy, the stripping index, and the predicted rock mass damage index, the blasting parameters in the blasting parameter set are corrected and the corresponding engineering measures are determined, and the blasting control instructions are generated according to the corrected blasting parameter set and the determined engineering measures.

[0104] In the embodiment, the blasting parameters in the blasting parameter set are corrected, including:

[0105] If the rock mass damage index or the predicted rock mass damage index is greater than the first threshold value, the blasting explosive quantity is reduced according to the first preset proportion, and the blasting hole spacing is reduced according to the second preset proportion; if the rock mass breakage vibration energy is greater than the second threshold value, the blasting explosive quantity is reduced according to the first preset proportion, the blasting hole spacing is reduced according to the second preset proportion, and the blasting time interval is increased according to the preset time length; if the rock mass crack density is greater than the third threshold value, the blasting hole spacing is reduced according to the third preset proportion.

[0106] In actual application, when the rock mass damage index or the predicted rock mass damage index is greater than 0.7, or the rock mass breakage vibration energy is greater than J, it indicates that the rock mass damage prediction value is too high or the rock mass high-frequency vibration energy is out of limit, at this time the blasting explosive quantity is reduced by 40% to reduce the total blasting energy, and the blasting hole spacing is increased by 20% to uniformly distribute the blasting energy; when the rock mass breakage vibration energy is greater than J, the blasting time interval is also increased by 15 ms to avoid energy superposition; when the rock mass crack density is greater than 5 cracks / cm², it indicates that the crack density is too high, at this time the blasting hole spacing is additionally increased by 15% to avoid the risk caused by crack penetration. Through the above process of blasting parameter optimization, the support cost and accident rate can be significantly reduced under the premise of ensuring construction safety.

[0107] In the present embodiment, corresponding engineering measures are determined, including:

[0108] If the predicted rock mass damage index is greater than the first threshold value or the rock mass fracture vibration energy is greater than the second threshold value, the corresponding engineering measure is to start nano-silicon slurry injection; if the rock mass fracture density is greater than the third threshold value, the corresponding engineering measure is to shorten the single cycle footage; if the support structure vibration energy is greater than the fourth threshold value or the stripping index is greater than the fifth threshold value, the corresponding engineering measure is to spray a concrete compensation layer; if the blasting thermal disturbance temperature rise is greater than the sixth threshold value, the corresponding engineering measure is to inject a low-temperature regulator into the grouting liquid.

[0109] In actual application, when the predicted rock mass damage index is greater than 0.7, or the rock mass fracture vibration energy is greater than J, it indicates that the rock mass damage prediction value is too high or the rock mass high-frequency vibration energy is out of limit, at this time nano-silicon slurry injection is also started to pre-fill the fractures to inhibit damage expansion and block the vibration energy transmission path; when the rock mass fracture density is greater than 5 lines / cm², it indicates that the fracture density is too high, at this time the single cycle footage is shortened to 1.5 m to reduce the blasting exposed surface and prevent the fractures from penetrating through; when the support structure vibration energy is greater than 200 J or the stripping index is greater than 0.15, the concrete compensation layer spraying is performed to increase the cross-sectional moment of inertia, improve the bending stiffness, and fill the void area to restore the contact stress; when the blasting thermal disturbance temperature rise is greater than 15℃, the low-temperature regulator is added to the grouting liquid to inhibit the temperature stress caused by the cement hydration heat.

[0110] The above process corrects the blasting parameters and determines the engineering measures based on multiple indexes such as rock mass fracture density, rock mass damage index, rock mass fracture vibration energy, support structure vibration energy, stripping index, and predicted rock mass damage index. The indexes are interrelated: fracture density and damage index reflect the current state of the rock mass, vibration energy (fracture and support) indicates dynamic risk, stripping index (based on power spectral density) warns of structure surface failure, and predicted damage index introduces the future state. When correcting the blasting parameters, high vibration energy may require reducing the blasting scale, and high stripping index may require strengthening the support. At the same time, the engineering measures address specific problems to ensure that the corrected blasting parameter set is executable on site. The blasting control instructions generated based on the corrected blasting parameter set and the determined engineering measures are not only accurate and efficient, but also actively respond to the uncertainty of deep tunneling, significantly reduce engineering risks, and improve overall economic benefits.

[0111] In summary, the deep-buried tunnel blasting control instruction generation method based on cross-scale data provided in this embodiment covers the internal damage of the rock mass, dynamic response and thermal effect by fusing multi-scale data such as rock mass crack density, blasting vibration velocity, supporting structure strain and blasting thermal disturbance temperature rise, ensures that the blasting control decision is based on multi-source information, and reduces one-sidedness; quantitative evaluation is provided by calculating damage index, vibration energy and other indicators to help timely adjust parameters, realize dynamic control, avoid cumulative damage, and improve the safety of tunnel construction; when calculating the rock mass damage index, the damage more consistent with the actual thermal coupling state is obtained through temperature and permeability correction, which significantly improves the damage evaluation accuracy of high-temperature and high-permeability conditions, and avoids overbreak or underbreak accidents; the ratio of the rock mass damage index to the supporting time-varying stiffness is minimized as the optimization target, and the genetic algorithm is used to search for the optimal blasting parameter set, which balances the rock mass damage and the stability of the supporting structure, ensures that the blasting is carried out within the safety range, prevents supporting failure or excessive damage to the rock mass, and the genetic algorithm can efficiently find the global optimal solution or approximate optimal solution, improving the blasting efficiency and economy; the future rock mass damage index is predicted based on the LSTM model to realize preventive control, adjust the blasting parameters in advance, avoid potential risks, and enhance the system robustness; the blasting parameters are corrected and the engineering measures are determined based on multiple indicators, so that the blasting control is more suitable for actual conditions, reduces unexpected events, and ensures construction continuity and quality. In addition, this embodiment starts from real-time monitoring, goes through data calculation, optimization search, prediction correction, and finally generates control instructions, each link is closely coupled, monitoring data drives index calculation, index calculation supports optimization target, optimization results input prediction model, prediction output guides parameter correction, and the corrected instruction affects the next round of monitoring, thereby forming a control closed loop, while taking advantage of the comprehensiveness of cross-scale data, the global optimization ability of genetic algorithm and the time series prediction advantage of LSTM, to ensure rapid and accurate response and avoid the data lag problem in traditional methods.

[0112] Based on the above technical solutions, the embodiment further provides a deep-buried tunnel blasting control instruction generation system based on cross-scale data, which is used to implement the deep-buried tunnel blasting control instruction generation method based on cross-scale data described in the embodiment, and the system comprises:

[0113] A multi-scale data acquisition unit is configured to monitor rock mass crack density, blasting vibration velocity, supporting structure strain and blasting thermal disturbance temperature rise in real time, and determine the power spectral density of the rock mass breakage vibration frequency band, the supporting structure resonance frequency band and the supporting structure strain corresponding signal;

[0114] a mutual feedback analysis unit, configured to calculate rock mass plastic strain according to the blasting vibration velocity, calculate rock mass damage index according to the rock mass plastic strain and blasting thermal disturbance temperature rise, calculate rock mass fracture vibration energy according to the rock mass fracture vibration frequency band, calculate support structure vibration energy according to the support structure resonance frequency band, calculate stripping index according to the power spectral density, and calculate support time-varying stiffness;

[0115] a decision application unit, configured to minimize the ratio of the rock mass damage index to the support time-varying stiffness as an optimization objective, search for an optimal blasting parameter set in a blasting parameter space based on a genetic algorithm, simultaneously predict a predicted rock mass damage index at a future time according to time sequence data of the rock mass damage index, the support time-varying stiffness and the blasting vibration velocity and based on LSTM, correct the blasting parameters in the blasting parameter set and determine corresponding engineering measures based on the rock mass fracture density, the rock mass damage index, the rock mass fracture vibration energy, the support structure vibration energy, the stripping index and the predicted rock mass damage index, and generate a blasting control instruction according to the corrected blasting parameter set and the determined engineering measures.

[0116] It can be understood that, since the deep-buried tunnel blasting control instruction generation system based on cross-scale data described in the embodiment is a system for implementing the deep-buried tunnel blasting control instruction generation method based on cross-scale data described in the embodiment, the system disclosed in the embodiment is relatively simple because it corresponds to the method disclosed in the embodiment, and the relevant parts can be referred to the part of the method description.

Claims

1. A method for generating blasting control commands for deep-buried tunnels based on cross-scale data, characterized in that, The method includes: Real-time monitoring of rock mass fracture density, blasting vibration velocity, support structure strain, and blasting thermal disturbance temperature rise; and determination of the rock mass fracture vibration frequency band, support structure resonance frequency band, and power spectral density of the signal corresponding to support structure strain. Based on the blasting vibration velocity, the rock mass plastic strain is inverted; based on the rock mass plastic strain and the temperature rise caused by blasting thermal disturbance, the rock mass damage index is calculated; based on the rock mass fracture vibration frequency band, the rock mass fracture vibration energy is calculated; based on the support structure resonance frequency band, the support structure vibration energy is calculated; based on the power spectral density, the stripping index is calculated; and the support time-varying stiffness is calculated. With the optimization objective of minimizing the ratio of rock mass damage index to support time-varying stiffness, the optimal blasting parameter set is searched in the blasting parameter space based on a genetic algorithm. At the same time, the rock mass damage index is predicted based on the time series data of rock mass damage index, support time-varying stiffness and blasting vibration velocity and LSTM to predict future moments. Based on the rock mass fracture density, rock mass damage index, rock mass fracture vibration energy, support structure vibration energy, stripping index, and predicted rock mass damage index, the blasting parameters in the blasting parameter set are corrected and corresponding engineering measures are determined. Blasting control commands are generated based on the corrected blasting parameter set and the determined engineering measures.

2. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, The method for detecting the density of rock mass fractures includes: Real-time monitoring of microseismic energy, and calculation of rock mass fracture density based on the microseismic energy, using the following formula: ； in, Indicates the density of rock fractures. This represents the energy of microseismic events. It represents the natural logarithm.

3. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, The inversion formula for the plastic strain of the rock mass is as follows: ； The method for calculating the rock mass damage index includes: determining the rock mass failure strain based on the uniaxial compressive strength of the rock and the dynamic load constitutive model; correcting the rock mass failure strain based on the temperature rise caused by blasting thermal disturbance; and calculating the rock mass damage index based on the rock mass plastic strain, the temperature rise caused by blasting thermal disturbance, and the corrected rock mass failure strain. The corrected formula for the rock mass failure strain is as follows: ； The formula for calculating the rock mass damage index is as follows: ； ； ； in, Indicates the rock mass damage index. Indicates the blasting vibration velocity. Indicates the plastic strain of the rock mass. Indicates the rock mass failure strain. This represents the corrected rock mass failure strain. Indicates the temperature softening factor. This indicates the temperature rise due to the thermal disturbance during the explosion. Indicates the seepage weakening factor. This represents the current permeability coefficient. Represents the initial permeability coefficient. Indicates the lithological scale coefficient. Indicates the lithological morphology coefficient. Represents the natural logarithm. This represents an exponential function.

4. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, The formula for calculating the time-varying stiffness of the support is as follows: ； in, express Constantly maintain time-varying stiffness. Indicates the initial stiffness of the support. Indicates the strength factor of concrete at different ages. Indicates the age of support, This represents the time constant for stiffness growth. It represents the natural logarithm.

5. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, The logic for generating the blasting parameter set is as follows: ； in, This represents the set of explosive parameters. Indicates the amount of explosives used in blasting. Indicates the distance between blasting holes. Indicates the time interval between blasts. Indicates the rock mass damage index. express Constantly maintain time-varying stiffness. This means finding a set of explosive parameters to minimize .

6. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, The rock mass vibration frequency band is (100Hz, ∞), and the support structure resonance frequency band is (0, 50Hz). The formula for calculating the vibration energy of rock mass fracture is as follows: ； The formula for calculating the vibration energy of the support structure is as follows: ； in, This represents the vibrational energy of rock mass fracture. This represents the power spectral density of rock mass fracture vibration. This represents the vibration energy of the support structure. This represents the vibration power spectral density of the support structure. Indicates frequency.

7. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, When calculating the stripping index, the stripping characteristic frequency band is (8Hz, 12Hz). The formula for calculating the stripping index is as follows: ； in, Indicates the stripping index. This represents the power spectral density of the signal corresponding to the strain of the support structure.

8. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, The blasting parameters in the blasting parameter set are corrected, including: If the rock mass damage index or the predicted rock mass damage index is greater than the first threshold, the amount of blasting explosive is reduced according to the first preset ratio, and the blasting hole spacing is reduced according to the second preset ratio. If the vibration energy of rock mass fracture is greater than the second threshold, the amount of blasting explosive is reduced according to the first preset ratio, the blasting hole spacing is reduced according to the second preset ratio, and the blasting time interval is increased according to the preset duration. If the rock mass fracture density is greater than the third threshold, the blasting hole spacing will be reduced according to the third preset ratio.

9. The method for generating blasting control commands for deep-buried tunnels based on cross-scale data according to claim 1, characterized in that, Determine the corresponding engineering measures, including: If the predicted rock mass damage index is greater than the first threshold or the rock mass fracture vibration energy is greater than the second threshold, the corresponding engineering measure is to initiate nano-silicon injection slurry. If the rock mass fracture density is greater than the third threshold, the corresponding engineering measure is to shorten the advance by single-cycle advance. If the vibration energy of the support structure is greater than the fourth threshold or the stripping index is greater than the fifth threshold, the corresponding engineering measure is to spray a concrete compensation layer. If the temperature rise due to blasting thermal disturbance exceeds the sixth threshold, the corresponding engineering measure is to add a low-temperature regulator to the grouting liquid.

10. A deep-buried tunnel blasting control command generation system based on cross-scale data, characterized in that, The system is used to implement the method for generating blasting control commands for deep-buried tunnels based on cross-scale data as described in any one of claims 1 to 9, the system comprising: The multi-scale data acquisition unit is used to monitor rock fracture density, blasting vibration velocity, support structure strain and blasting thermal disturbance temperature rise in real time, and to determine the rock fracture vibration frequency band, support structure resonance frequency band and power spectral density of the signal corresponding to support structure strain. The feedback analysis unit is used to invert the rock mass plastic strain based on the blasting vibration velocity, calculate the rock mass damage index based on the rock mass plastic strain and the temperature rise caused by blasting thermal disturbance, calculate the rock mass fracture vibration energy based on the rock mass fracture vibration frequency band, calculate the support structure vibration energy based on the support structure resonance frequency band, calculate the stripping index based on the power spectral density, and calculate the support time-varying stiffness. The decision application unit is used to minimize the ratio of rock mass damage index to support time-varying stiffness as the optimization objective. Based on a genetic algorithm, it searches for the optimal blasting parameter set in the blasting parameter space. Simultaneously, based on the time-series data of rock mass damage index, support time-varying stiffness, and blasting vibration velocity, it predicts the rock mass damage index for future times using LSTM. Based on the rock mass fracture density, rock mass damage index, rock mass fracture vibration energy, support structure vibration energy, stripping index, and predicted rock mass damage index, it corrects the blasting parameters in the blasting parameter set and determines corresponding engineering measures. Based on the corrected blasting parameter set and the determined engineering measures, it generates blasting control commands.

Images