Earthquake wave advanced forecasting method in tunnel for water conservancy investigation

Abstract

The invention relates to the technical field of tunnel abnormity early warning, in particular to an in-tunnel seismic wave advanced forecasting method for water conservancy investigation, and the method comprises the steps: detecting the flood discharge process of a flood discharge tunnel through the seismic wave advanced forecasting method, and obtaining tunnel structure data and tunnel geological data; dividing to obtain initial structure data, process structure data, end structure data, initial geological data, process geological data and end geological data; identifying to obtain a flood discharge parameter set, and obtaining tunnel sub-regions according to the initial structure data, the initial geological data and the flood discharge parameter set; obtaining a flood discharge impact index according to the tunnel sub-region process structure data, the process geological data and the flood discharge parameter set; obtaining a structure damage coefficient according to the flood discharge impact index, the initial structure data and the end structure data; and obtaining a sub-region early warning coefficient according to the structural damage coefficient, the initial geological data and the ending geological data. According to the method, the abnormality of the flood discharge tunnel is accurately identified and early warned through the sub-region early warning coefficient.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel anomaly early warning, and particularly to a method for seismic wave advanced prediction in a tunnel for water conservancy exploration. Background Technique

[0002] The seismic wave advanced prediction technology excites a seismic source and receives reflected wave signals, analyzes the wave field propagation characteristics, and realizes the accurate detection of the geological structure in front of the tunnel. In the field of water conservancy projects, as the core flood discharge structure of a water conservancy hub project, the flood discharge tunnel needs to quickly divert flood under high water level conditions to ensure the safety of the dam.

[0003] However, long-term flood discharge operations may cause deterioration of the mechanical properties of the surrounding rock. For example, continuous infiltration of high-pressure water into rock fractures reduces the shear strength of the structural plane, leading to the expansion of the broken zone or the activation of faults; the continuous scouring and erosion of the surrounding rock and lining structure by high-speed water flow may induce cavitation damage or lining peeling.

[0004] The existing seismic wave advanced prediction technology has significant limitations; traditional methods mostly focus on static detection during the tunnel construction period and lack a periodic dynamic monitoring mechanism for flood discharge tunnels during the operation period; at the same time, factors such as water flow pulsation pressure and sudden temperature change during the flood discharge process will change the wave velocity characteristics of the rock mass, and broadband noise and mechanical vibration during the flood discharge period seriously contaminate the effective seismic signals, resulting in missed or misjudged abnormal bodies.

[0005] Therefore, a method for seismic wave advanced prediction in a tunnel for water conservancy exploration is proposed. Summary of the Invention

[0006] The purpose of the present invention is to provide a method for seismic wave advanced prediction in a tunnel for water conservancy exploration. By detecting the flood discharge process of the flood discharge tunnel through the seismic wave advanced prediction method, tunnel structure data and tunnel geological data are obtained, and initial structure data, process structure data, end structure data, initial geological data, process geological data, and end geological data are divided; a flood discharge parameter set is identified, and tunnel sub-regions are obtained according to the initial structure data, initial geological data, and flood discharge parameter set; a flood discharge impact index is obtained according to the process structure data, process geological data, and flood discharge parameter set of the tunnel sub-region; a structure damage coefficient is obtained according to the flood discharge impact index, initial structure data, and end structure data; a sub-region early warning coefficient is obtained according to the structure damage coefficient, initial geological data, and end geological data. The present invention accurately identifies and warns of anomalies in the flood discharge tunnel through the sub-region early warning coefficient.

[0007] To achieve the above object, the present invention provides the following technical solution: A method for seismic wave advanced prediction in a tunnel for water conservancy exploration, including:

[0008] S10. Detect the flood discharge process of the flood discharge tunnel through the seismic wave advanced prediction method, including detecting the structure of the flood discharge tunnel to obtain tunnel structure data; detecting the geology around the flood discharge tunnel to obtain tunnel geology data;

[0009] S20. Divide the tunnel structure data according to the flood discharge process to obtain initial structure data, process structure data, and end structure data; divide the tunnel geology data according to the flood discharge process to obtain initial geology data, process geology data, and end geology data;

[0010] S30. Set up Internet of Things sensing devices to identify the flood discharge tunnel to obtain a set of flood discharge parameters, including flow velocity, flow rate, water pressure, vibration, and water temperature; divide the flood discharge tunnel according to the initial structure data, initial geology data, and the set of flood discharge parameters to obtain tunnel sub-areas;

[0011] S40. Identify the flood discharge impact index based on the process structure data, process geology data, and the set of flood discharge parameters of the tunnel sub-areas; identify the structure damage coefficient based on the flood discharge impact index, initial structure data, and end structure data of the tunnel sub-areas; identify the sub-area warning coefficient based on the structure damage coefficient, initial geology data, and end geology data of the tunnel sub-areas.

[0012] During the detection of the flood discharge tunnel structure, the first-frequency seismic wave is used; during the detection of the geology around the flood discharge tunnel, the second-frequency seismic wave is used.

[0013] The process of dividing the flood discharge tunnel to obtain tunnel sub-areas is as follows: Obtain the initial structure data, initial geology data, and the set of flood discharge parameters; perform clustering identification on the initial structure data, initial geology data, and the set of flood discharge parameters through a clustering algorithm, and divide the flood discharge tunnel according to the clustering identification results to obtain tunnel sub-areas.

[0014] The flood discharge impact index is obtained according to the flood discharge impact identification model; the flood discharge impact identification model includes a data denoising processing layer, a structure data identification layer, a geology data identification layer, and an impact index calculation layer;

[0015] The data denoising processing layer denoises the process structure data and process geology data according to the set of flood discharge parameters;

[0016] The structure data identification layer identifies the denoised process structure data, and according to the difference changes between the data, identifies the structure change characteristic data; the geology data identification layer identifies the denoised process geology data, and according to the difference changes between the data, identifies the geology change characteristic data; the impact index calculation layer identifies the flood discharge impact index according to the structure change characteristic data, geology change characteristic data, and the set of flood discharge parameters.

[0017] The process of obtaining the structural damage coefficient is as follows: Obtain the end structure data, denoise the end structure data to obtain optimized structure data; Identify the structural difference area based on the optimized structure data and the initial structure data; Then, obtain the structural difference coefficient according to the data difference degree of the structural difference area; Calculate the structural damage coefficient based on the structural difference area, the structural difference coefficient, and the flood discharge impact index.

[0018] The process of denoising the end structure data to obtain optimized structure data is as follows: Construct an environmental integration algorithm to denoise the end structure data. The environmental integration algorithm includes an environmental coupling correction module and a physical constraint optimization module; The environmental coupling correction module obtains the environmental difference parameters at the initial and end moments of flood discharge according to the flood discharge parameter set; Construct a wave velocity correlation correction model based on the environmental difference parameters, and correct the end structure data according to the wave velocity correlation correction model to obtain the first corrected data; The physical constraint optimization module is used to capture the spatio-temporal characteristic data of the first corrected data, and optimize and determine the spatio-temporal characteristic data to obtain optimized structure data.

[0019] The process of calculating the sub-region warning coefficient is as follows: Obtain the initial geological data and the end geological data; Preprocess the end geological data to obtain optimized geological data; Identify the geological difference area and the geological difference coefficient based on the optimized geological data and the initial geological data; Identify the positional relationship between the geological difference area and the structural difference area, and divide the geological difference area into an associated abnormal area and a non-associated abnormal area, and set weights respectively; Calculate the sub-region warning coefficient based on the associated abnormal area, the non-associated abnormal area, the geological difference coefficient, and the structural damage coefficient; The calculation formula of the sub-region warning coefficient is:

[0020]

[0021] where SUC represents the sub-region warning coefficient; loi represents the structural damage coefficient; α1 represents the weight of the associated area; aar i represents the volume of the associated abnormal area i; λ i represents the geological difference coefficient of the associated abnormal area i; AT1 represents the first threshold; α2 represents the weight of the non-associated area; nar j represents the volume of the non-associated abnormal area j; ω j represents the geological difference coefficient of the non-associated abnormal area j; AT2 represents the second threshold; exp represents the exponential function with the natural constant as the base; n represents the number of associated abnormal areas; m represents the number of non-associated abnormal areas.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0023] 1. The present invention constructs a flood discharge impact recognition model to identify the flood discharge impact index. First, denoise the process structure data and process geological data according to the flood discharge parameter set. Identify the denoised process structure data and process geological data to obtain structure change characteristic data and geological change characteristic data. Then, combine with the flood discharge parameter set for identification to obtain the flood discharge impact index. The flood discharge impact index can accurately measure the impact of water flow on the flood discharge tunnel during flood discharge.

[0024] 2. The present invention constructs an environmental integration algorithm to denoise the end structure data, and corrects the end structure data through the wave velocity correlation correction model in the environmental coupling correction module to obtain the first corrected data. Then, optimize the first corrected data through the physical constraint optimization module to obtain the optimized structure data. Identify the structure difference region and structure difference coefficient based on the optimized structure data and the initial structure data. Calculate the structure damage coefficient based on the structure difference region, structure difference coefficient, and flood discharge impact index. The structure damage coefficient can accurately measure the structural damage of the flood discharge tunnel before and after flood discharge.

[0025] 3. The present invention preprocesses the end geological data to obtain optimized geological data. Identify the geological difference region and geological difference coefficient based on the optimized geological data and the initial geological data. Divide the geological difference region into an associated abnormal region and a non-associated abnormal region according to the positional relationship between the geological difference region and the structure difference region. Calculate the sub-region warning coefficient based on the associated abnormal region, non-associated abnormal region, geological difference coefficient, and structure damage coefficient. The sub-region warning coefficient can accurately warn the flood discharge tunnel from two aspects of geological anomalies and structural losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a schematic flow chart of a seismic wave advanced prediction method in a tunnel for water conservancy exploration according to the present invention;

[0027] Figure 2 It is a schematic structural diagram of the flood discharge impact recognition model according to the present invention;

[0028] Figure 3 It is a schematic logical diagram of a seismic wave advanced prediction method in a tunnel for water conservancy exploration according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0030] Embodiment 1

[0031] The present invention proposes a method for earthquake wave advance prediction in a tunnel for water conservancy exploration, and the process is as Figure 1 shown, including:

[0032] S10. Detect the flood discharge process of the flood discharge tunnel through the earthquake wave advance prediction method, including detecting the structure of the flood discharge tunnel to obtain tunnel structure data; detecting the geology around the flood discharge tunnel to obtain tunnel geology data.

[0033] During the detection of the structure of the flood discharge tunnel, seismic waves with the first frequency are used; during the detection of the geology around the flood discharge tunnel, seismic waves with the second frequency are used.

[0034] The earthquake wave advance prediction technology excites an artificial seismic source and receives the reflected wave signal, analyzes the wave field propagation characteristics, and realizes the accurate detection of the geological structure in front of the tunnel. Its core mechanism lies in: when seismic waves propagate in a rock mass medium, when encountering an interface with wave impedance difference, such as geological anomalies such as faults, fracture zones, and karst caves, reflected waves, refracted waves, and scattered waves will be generated. By analyzing the travel time, amplitude, and spectrum characteristics of the received signal, the spatial position, geometric size, and physical properties of the geological interface can be inverted.

[0035] The flood discharge tunnel is often constructed with concrete. Concrete cracking will cause a decrease in local density and elastic modulus, change the seismic wave propagation characteristics, resulting in a decrease in wave velocity and an increase in reflection; at the same time, a wave impedance difference is formed at the crack interface, generating reflected wave and scattered wave signals. Therefore, the structural changes of concrete can be identified through seismic waves.

[0036] In order to accurately detect the structure of concrete, a high-frequency mechanical impact source, such as an electromagnetic hammer or a laser ultrasonic generator, is used to excite a short pulse wave with a main frequency of 1 - 5 kHz; a micro geophone array (spacing 0.5 - 1 m) is arranged along the lining surface, and the full waveform acquisition mode is adopted; a three-component sensor is added to capture the polarization change of the shear wave (S wave) caused by the crack. Reverse time migration or full waveform inversion is used to identify the cracks in the concrete. Cracks above 0.5 mm can be detected by frequencies above 1 kHz. Therefore, in this embodiment, the first frequency seismic wave is 1 - 5 kHz, which is a high-frequency seismic wave.

[0037] The second frequency seismic wave is 50 - 500 Hz to identify the geology around the tunnel.

[0038] S20. Divide the tunnel structure data according to the flood discharge process to obtain initial structure data, process structure data, and end structure data; divide the tunnel geology data according to the flood discharge process to obtain initial geology data, process geology data, and end geology data.

[0039] Among them, the initial structural data is the data obtained from the inspection of the tunnel structure before the flood discharge starts; the process structural data is the data obtained from the inspection of the tunnel structure during the flood discharge process; the end structural data is the data obtained from the inspection of the tunnel structure after the flood discharge ends. The initial geological data is the data obtained from the inspection of the tunnel geology before the flood discharge starts; the process geological data is the data obtained from the inspection of the tunnel geology during the flood discharge process; the end geological data is the data obtained from the inspection of the tunnel geology after the flood discharge ends.

[0040] S30. Set up Internet of Things sensing devices to identify the flood discharge tunnel, and obtain a set of flood discharge parameters, including flow velocity, flow rate, water pressure, vibration, and water temperature; divide the flood discharge tunnel according to the initial structural data, initial geological data, and the set of flood discharge parameters to obtain tunnel sub-regions.

[0041] The process of dividing the flood discharge tunnel to obtain tunnel sub-regions is as follows:

[0042] Obtain the initial structural data, initial geological data, and the set of flood discharge parameters;

[0043] Perform clustering identification on the initial structural data, initial geological data, and the set of flood discharge parameters through a clustering algorithm, and divide the flood discharge tunnel according to the clustering identification result to obtain tunnel sub-regions.

[0044] The present invention collects the initial structural data, initial geological data, and the set of flood discharge parameters; performs clustering identification on the initial structural data, initial geological data, and the set of flood discharge parameters through a clustering algorithm, and divides the flood discharge tunnel according to the clustering identification result to obtain tunnel sub-regions; thereby accurately dividing and identifying the flood discharge tunnel according to the initial structural data, initial geological data, and the set of flood discharge parameters, so that the parameters within each tunnel sub-region are consistent.

[0045] S40. Identify the flood discharge impact index according to the process structural data, process geological data, and the set of flood discharge parameters of the tunnel sub-region; identify the structural damage coefficient according to the flood discharge impact index, initial structural data, and end structural data of the tunnel sub-region; identify the sub-region warning coefficient according to the structural damage coefficient, initial geological data, and end geological data of the tunnel sub-region.

[0046] The flood discharge impact index is obtained according to a flood discharge impact identification model; the flood discharge impact identification model is constructed based on a deep learning network, and its structure is as Figure 2 shown, including a data denoising processing layer, a structural data identification layer, a geological data identification layer, and an impact index calculation layer;

[0047] The data denoising processing layer denoises the process structural data and process geological data according to the set of flood discharge parameters;

[0048] The structure data recognition layer recognizes the denoised process structure data, and based on the difference changes between the data, recognizes the structure change feature data;

[0049] The geological data recognition layer recognizes the denoised process geological data, and based on the difference changes between the data, recognizes the geological change feature data;

[0050] The flood discharge impact index calculation layer recognizes based on the structure change feature data, geological change feature data, and flood discharge parameter set to obtain the flood discharge impact index.

[0051] The calculation formula for the flood discharge impact index is:

[0052]

[0053] Among them, FII represents the flood discharge impact index; csp represents the structure process change coefficient; β1 represents the structure change weight; gpv represents the geological process change coefficient; β2 represents the geological change weight; CH represents the change threshold; η1 represents the water pressure weight; pw represents the maximum water pressure; pt represents the water pressure threshold; η2 represents the flow velocity weight; sv represents the maximum flow velocity; st represents the flow velocity threshold.

[0054] The structure process change coefficient is recognized based on the structure change feature data; the geological process change coefficient is recognized based on the geological change feature data.

[0055] The present invention constructs a flood discharge impact recognition model to recognize the flood discharge impact index; first, denoise the process structure data and process geological data according to the flood discharge parameter set; recognize the denoised process structure data and process geological data to obtain the structure change feature data and geological change feature data; then combine with the flood discharge parameter set for recognition to obtain the flood discharge impact index; through the flood discharge impact index, the impact of water flow on the flood discharge tunnel during flood discharge can be accurately measured.

[0056] The process for obtaining the structure damage coefficient is as follows:

[0057] Obtain the end structure data, denoise the end structure data to obtain the optimized structure data;

[0058] Recognize based on the optimized structure data and the initial structure data to obtain the structure difference region; then, based on the data difference degree of the structure difference region, obtain the structure difference coefficient;

[0059] Calculate the structure damage coefficient based on the structure difference region, structure difference coefficient, and flood discharge impact index.

[0060] The calculation formula for the structure damage coefficient is:

[0061]

[0062] Among them, loi represents the structural damage coefficient; FII represents the flood discharge impact index; csa c represents the volume of the structural difference region c; θ c represents the structural difference coefficient of the structural difference region c; FT represents the structural difference threshold; p represents the number of structural difference regions.

[0063] The process of obtaining volume data is as follows: using the amplitude and phase information of seismic waves, a three-dimensional wave velocity model is constructed to directly display the spatial morphology of abnormal bodies, and volume parameters are measured and obtained.

[0064] Before and after the flood discharge in the flood discharge tunnel, its environment will change significantly, affecting seismic waves; for example, humidity change: after the flood discharge, the water content of the surrounding rock of the tunnel increases, resulting in a decrease in wave velocity and an increase in signal attenuation; temperature fluctuation: the temperature difference of the flood discharge water flow causes thermal expansion and contraction of microcracks in the lining, changing the reflection characteristics of the wave impedance interface.

[0065] Therefore, in the process of identifying the structural differences between the initial structure data and the end structure data, it is necessary to denoise and optimize the end structure data according to the environmental changes of the flood discharge tunnel at different times;

[0066] Traditional methods mainly include wavelet transform, Kalman filter, and deep learning models, etc.; however, conventional wavelet transform, Kalman filter, deep learning models, etc. cannot distinguish systematic distortions caused by environmental factors from random noise. For this reason, the present invention proposes an environmental integration algorithm to denoise the end structure data.

[0067] Construct an environmental integration algorithm to denoise the end structure data, and the environmental integration algorithm includes an environmental coupling correction module and a physical constraint optimization module;

[0068] The environmental coupling correction module obtains environmental difference parameters at the initial moment and the end moment of flood discharge according to the flood discharge parameter set; constructs a wave velocity correlation correction model according to the environmental difference parameters, and corrects the end structure data according to the wave velocity correlation correction model to obtain first corrected data;

[0069] Among them, the wave velocity correlation correction model corrects the wave velocity according to the humidity and temperature of the environment, and the formula is as follows:

[0070]

[0071] Among them, V p (H(t), T(t)) represents the corrected wave velocity; V p0 represents the reference wave velocity; γ represents the humidity correction coefficient; H(t) represents the dynamic humidity, reflecting the change of humidity with time t; represents the temperature correction coefficient; T(t) represents the dynamic temperature, reflecting the change of temperature with time t; T0 represents the reference temperature.

[0072] Subsequently, the scale parameter of wavelet transform is determined based on the corrected wave velocity and the reference wave velocity, and the formula is as follows:

[0073]

[0074] where, a (t) represents the wavelet transform scale parameter, which changes with time t; V p0 represents the reference wave velocity; V p (H(t), T(t)) represents the corrected wave velocity; it is calculated from the dynamic humidity H(t) and the dynamic temperature T(t); a0 is the initial scale parameter, which is determined according to the seismic wave frequency and resolution requirements.

[0075] In wavelet transform, the scale parameter a (t) controls the balance between frequency resolution and time resolution. A smaller a (t) corresponds to high-frequency components, and a larger a (t) corresponds to low-frequency components. The scale parameter of wavelet analysis can be adaptively corrected according to the wave velocity offset caused by real-time environmental changes to compensate for signal phase delay and dispersion effects.

[0076] The physical constraint optimization module is obtained by combining the U-Net+LSTM model; it is used to capture the spatio-temporal feature data of the first corrected data, optimize and judge the spatio-temporal feature data, and obtain the optimized structure data; it includes a physical constraint loss function, and the physical constraint loss function is composed of the energy value of the denoised signal, the theoretical attenuation value of high-frequency energy, the wave velocity identified in the denoised signal, and the corrected wave velocity measured by the environmental coupling correction module.

[0077] In order to verify the effect of the environmental integration algorithm of this application; construction methods 1, 2, 3, and 4 are constructed for comparative experiments; method 1 is the environmental integration algorithm of this application; method 2 is the conventional wavelet transform; method 3 is the Kalman filter; method 4 is the deep learning model CNN. After multiple data identifications, data table 1 is obtained.

[0078] Table 1 Effect verification table of environmental integration algorithm

[0079] Method Number Method Content Recognition Accuracy Method 1 Environmental Integration Algorithm 0.9358 Method 2 Wavelet Transform 0.9014 Method 3 Kalman Filter 0.8865 Method 4 Deep Learning Model CNN 0.8924

[0080] From the recognition accuracy rate in Table 1, it can be seen that the environmental integration algorithm constructed in this application has significant advantages in seismic wave denoising in tunnels.

[0081] The environmental integration algorithm of the present invention denoises the end - structure data, corrects the end - structure data through the wave - velocity correlation correction model in the environmental coupling correction module to obtain the first corrected data; then optimizes the first corrected data through the physical constraint optimization module to obtain the optimized structure data; identifies the structural difference region and the structural difference coefficient based on the optimized structure data and the initial structure data; calculates the structural damage coefficient based on the structural difference region, the structural difference coefficient and the flood - discharge impact index; the structural damage coefficient can accurately measure the structural damage of the flood - discharge tunnel before and after flood - discharge.

[0082] The calculation process of the sub - region warning coefficient is as follows:

[0083] Obtain the initial geological data and the end geological data; pre - process the end geological data to obtain the optimized geological data; identify the geological difference region and the geological difference coefficient based on the optimized geological data and the initial geological data;

[0084] Identify the positional relationship between the geological difference region and the structural difference region, and divide the geological difference region into an associated abnormal region and a non - associated abnormal region, and set weights respectively;

[0085] Calculate the sub - region warning coefficient based on the associated abnormal region, the non - associated abnormal region, the geological difference coefficient and the structural damage coefficient; the calculation formula of the sub - region warning coefficient is:

[0086]

[0087] where SUC represents the sub - region warning coefficient; loi represents the structural damage coefficient; α1 represents the weight of the associated region; aar i represents the volume of the associated abnormal region i; λ i represents the geological difference coefficient of the associated abnormal region i; AT1 represents the first threshold; α2 represents the weight of the non - associated region; nar j represents the volume of the non - associated abnormal region j; ω j represents the geological difference coefficient of the non - associated abnormal region j; AT2 represents the second threshold; exp represents the exponential function with the natural constant as the base; n represents the number of associated abnormal regions; m represents the number of non - associated abnormal regions.

[0088] Among them, the determination process of the associated abnormal region is: obtain the distribution positions of the geological abnormal region and the structural abnormal region, determine the highest correlation coefficient between the geological abnormal region and the structural abnormal region according to the distribution positions; make a judgment according to the correlation coefficient.

[0089] The present invention preprocesses the end geological data to obtain optimized geological data; identifies geological difference regions and geological difference coefficients based on the optimized geological data and the initial geological data; divides the geological difference regions into associated abnormal regions and unassociated abnormal regions according to the positional relationship between the geological difference regions and the structural difference regions, and calculates sub-region warning coefficients based on the associated abnormal regions, unassociated abnormal regions, geological difference coefficients, and structural damage coefficients; through the sub-region warning coefficients, the flood discharge tunnel can be accurately warned of geological anomalies and structural losses.

[0090] The present invention detects the flood discharge process of the flood discharge tunnel through the seismic wave advanced prediction method, obtains tunnel structure data and tunnel geological data, and divides them into initial structure data, process structure data, end structure data, initial geological data, process geological data, and end geological data; identifies a flood discharge parameter set, and obtains tunnel sub-regions based on the initial structure data, initial geological data, and flood discharge parameter set; obtains a flood discharge impact index based on the process structure data, process geological data, and flood discharge parameter set of the tunnel sub-region; obtains a structural damage coefficient based on the flood discharge impact index, initial structure data, and end structure data; obtains a sub-region warning coefficient based on the structural damage coefficient, initial geological data, and end geological data. The present invention accurately identifies and warns of anomalies in the flood discharge tunnel through the sub-region warning coefficient.

[0091] Embodiment 2

[0092] The present invention proposes a seismic wave advanced prediction method in a tunnel for water conservancy exploration, and its logic is as Figure 3 shown, including:

[0093] S10. Detect the flood discharge process of the flood discharge tunnel through the seismic wave advanced prediction method, including detecting the structure of the flood discharge tunnel to obtain tunnel structure data; detecting the geology around the flood discharge tunnel to obtain tunnel geological data.

[0094] During the process of detecting the structure of the flood discharge tunnel, seismic waves with the first frequency are used; during the process of detecting the geology around the flood discharge tunnel, seismic waves with the second frequency are used.

[0095] S20. Divide the tunnel structure data according to the flood discharge process to obtain initial structure data, process structure data, and end structure data; divide the tunnel geological data according to the flood discharge process to obtain initial geological data, process geological data, and end geological data.

[0096] S30. Set up Internet of Things sensing devices to identify the flood discharge tunnel to obtain a flood discharge parameter set, including flow velocity, flow rate, water pressure, vibration, and water temperature; divide the flood discharge tunnel according to the initial structure data, initial geological data, and flood discharge parameter set to obtain tunnel sub-regions.

[0097] The process of dividing the flood discharge tunnel to obtain the tunnel sub - regions is as follows:

[0098] Obtain the initial structure data, initial geological data, and flood discharge parameter set;

[0099] Perform clustering recognition on the initial structure data, initial geological data, and flood discharge parameter set through a clustering algorithm, and divide the flood discharge tunnel according to the clustering recognition result to obtain the tunnel sub - regions.

[0100] The present invention collects the initial structure data, initial geological data, and flood discharge parameter set; performs clustering recognition on the initial structure data, initial geological data, and flood discharge parameter set through a clustering algorithm, and divides the flood discharge tunnel according to the clustering recognition result to obtain the tunnel sub - regions; thereby accurately dividing and identifying the flood discharge tunnel according to the initial structure data, initial geological data, and flood discharge parameter set, so that the parameters within each tunnel sub - region are consistent.

[0101] S40. Identify the flood discharge impact index based on the process structure data, process geological data, and flood discharge parameter set of the tunnel sub - regions; identify the structural damage coefficient based on the flood discharge impact index, initial structure data, and end - structure data of the tunnel sub - regions; identify the sub - region early - warning coefficient based on the structural damage coefficient, initial geological data, and end - geological data of the tunnel sub - regions.

[0102] The flood discharge impact index is obtained according to the flood discharge impact recognition model; the flood discharge impact recognition model includes a data denoising processing layer, a structure data recognition layer, a geological data recognition layer, and an impact index calculation layer;

[0103] The data denoising processing layer denoises the process structure data and process geological data according to the flood discharge parameter set;

[0104] The structure data recognition layer recognizes the denoised process structure data, and according to the difference changes between the data, recognizes the structure change characteristic data;

[0105] The geological data recognition layer recognizes the denoised process geological data, and according to the difference changes between the data, recognizes the geological change characteristic data;

[0106] The impact index calculation layer recognizes according to the structure change characteristic data, geological change characteristic data, and flood discharge parameter set to obtain the flood discharge impact index.

[0107] The present invention constructs a flood discharge impact recognition model to recognize the flood discharge impact index; first, denoise the process structure data and process geological data according to the flood discharge parameter set; recognize the denoised process structure data and process geological data to obtain structure change characteristic data and geological change characteristic data; then combine with the flood discharge parameter set for recognition to obtain the flood discharge impact index; the flood discharge impact index can accurately measure the impact of water flow on the flood discharge tunnel during flood discharge.

[0108] The process of obtaining the structure damage coefficient is as follows:

[0109] Obtain the end structure data, denoise the end structure data to obtain optimized structure data;

[0110] Recognize according to the optimized structure data and the initial structure data to obtain the structure difference region; then obtain the structure difference coefficient according to the data difference degree of the structure difference region;

[0111] Calculate the structure damage coefficient according to the structure difference region, the structure difference coefficient and the flood discharge impact index.

[0112] The process of denoising the end structure data to obtain optimized structure data is as follows:

[0113] Construct an environment integration algorithm to denoise the end structure data, and the environment integration algorithm includes an environment coupling correction module and a physical constraint optimization module;

[0114] The environment coupling correction module obtains the environmental difference parameters at the initial moment and the end moment of flood discharge according to the flood discharge parameter set; constructs a wave velocity correlation correction model according to the environmental difference parameters, and corrects the end structure data according to the wave velocity correlation correction model to obtain the first corrected data;

[0115] The physical constraint optimization module is used to capture the spatio-temporal characteristic data of the first corrected data, and optimize and judge the spatio-temporal characteristic data to obtain optimized structure data.

[0116] The present invention constructs an environment integration algorithm to denoise the end structure data, corrects the end structure data through the wave velocity correlation correction model in the environment coupling correction module to obtain the first corrected data; then optimizes the first corrected data through the physical constraint optimization module to obtain optimized structure data; recognizes the structure difference region and the structure difference coefficient according to the optimized structure data and the initial structure data; calculates the structure damage coefficient according to the structure difference region, the structure difference coefficient and the flood discharge impact index; the structure damage coefficient can accurately measure the structural damage of the flood discharge tunnel before and after flood discharge.

[0117] The calculation process of the sub-region warning coefficient is as follows:

[0118] Obtain initial geological data and end geological data; preprocess the end geological data to obtain optimized geological data; identify geological difference regions and geological difference coefficients based on the optimized geological data and the initial geological data;

[0119] Identify the positional relationship between the geological difference regions and the structural difference regions, and divide the geological difference regions into associated abnormal regions and unassociated abnormal regions, and set weights respectively;

[0120] Calculate the sub-region warning coefficient based on the associated abnormal regions, unassociated abnormal regions, geological difference coefficients, and structural damage coefficients; the calculation formula for the sub-region warning coefficient is:

[0121]

[0122] where SUC represents the sub-region warning coefficient; loi represents the structural damage coefficient; α1 represents the weight of the associated region; aar i represents the volume of the associated abnormal region i; λ i represents the geological difference coefficient of the associated abnormal region i; AT1 represents the first threshold; α2 represents the weight of the unassociated region; nar j represents the volume of the unassociated abnormal region j; ω j represents the geological difference coefficient of the unassociated abnormal region j; AT2 represents the second threshold; exp represents the exponential function with the natural constant as the base; n represents the number of associated abnormal regions; m represents the number of unassociated abnormal regions.

[0123] The present invention preprocesses the end geological data to obtain optimized geological data; identifies geological difference regions and geological difference coefficients based on the optimized geological data and the initial geological data; divides the geological difference regions into associated abnormal regions and unassociated abnormal regions according to the positional relationship between the geological difference regions and the structural difference regions, and calculates the sub-region warning coefficient based on the associated abnormal regions, unassociated abnormal regions, geological difference coefficients, and structural damage coefficients; through the sub-region warning coefficient, the flood discharge tunnel can be accurately warned from two aspects of geological anomalies and structural losses.

[0124] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for advanced prediction of seismic waves in tunnels for water conservancy survey, characterized in that: include: S10. Detecting the flood discharge process of the flood discharge tunnel by using the seismic wave advance prediction method, including detecting the flood discharge tunnel structure to obtain tunnel structure data; Conduct geological tests around the flood discharge tunnel to obtain tunnel geological data; S20. Divide the tunnel structure data according to the flood discharge process to obtain initial structure data, process structure data and end structure data; The tunnel geological data is divided according to the flood discharge process to obtain initial geological data, process geological data and final geological data; S30. Setting an IoT sensor device to identify the flood discharge tunnel and obtain a flood discharge parameter set, including flow velocity, flow rate, water pressure, vibration and water temperature; Divide the flood discharge tunnel according to the initial structural data, initial geological data and flood discharge parameter set to obtain tunnel sub-areas; S40. Obtain the flood discharge impact index based on the process structural data, process geological data and flood discharge parameter set of the tunnel sub-area; obtain the structural damage coefficient based on the flood discharge impact index, initial structural data and final structural data of the tunnel sub-area; obtain the sub-area early warning coefficient based on the structural damage coefficient, initial geological data and final geological data of the tunnel sub-area.

2. The method for advanced prediction of earthquake waves in tunnels for water conservancy survey according to claim 1, characterized in that: In the process of detecting the flood discharge tunnel structure, the first frequency seismic wave is used; in the process of detecting the geology around the flood discharge tunnel, the second frequency seismic wave is used.

3. The method for advanced prediction of earthquake waves in tunnels for water conservancy survey according to claim 1, characterized in that: The process of dividing the flood discharge tunnel and obtaining tunnel sub-areas is as follows: Obtaining initial structural data, initial geological data and flood discharge parameter sets; The initial structural data, initial geological data and flood discharge parameter set are clustered and identified by a clustering algorithm. The flood discharge tunnel is divided according to the clustering identification results to obtain tunnel sub-areas.

4. The method for advanced prediction of earthquake waves in tunnels for water conservancy survey according to claim 1, characterized in that: The flood discharge impact index is obtained according to a flood discharge impact identification model; the flood discharge impact identification model includes a data denoising processing layer, a structural data identification layer, a geological data identification layer and an impact index calculation layer; The data denoising processing layer denoises the process structural data and the process geological data according to the flood discharge parameter set; The structural data recognition layer recognizes the denoised process structural data and obtains structural change feature data according to the difference changes between the data; The geological data recognition layer recognizes the de-noised process geological data and obtains geological change characteristic data according to the difference changes between the data; The impact index calculation layer is identified based on the structural change characteristic data, the geological change characteristic data and the flood discharge parameter set to obtain the flood discharge impact index.

5. The method for advanced prediction of earthquake waves in tunnels for water conservancy survey according to claim 1, characterized in that: The process of obtaining the structural damage coefficient is as follows: Acquire end structure data, and perform denoising on the end structure data to obtain optimized structure data; Identify according to the optimized structure data and the initial structure data to obtain a structure difference region; and then obtain a structure difference coefficient according to the degree of data difference in the structure difference region; The structural damage coefficient is calculated based on the structural difference area, structural difference coefficient and flood discharge impact index.

6. The method for advanced prediction of earthquake waves in tunnels for water conservancy survey according to claim 5, characterized in that: The process of denoising the final structure data to obtain the optimized structure data is as follows: Constructing an environment integration algorithm to denoise the end structure data, the environment integration algorithm comprising an environment coupling correction module and a physical constraint optimization module; The environmental coupling correction module obtains environmental difference parameters at the initial time of flood discharge and the end time of flood discharge according to the flood discharge parameter set; constructs a wave velocity correlation correction model according to the environmental difference parameters, and corrects the end structure data according to the wave velocity correlation correction model to obtain first correction data; The physical constraint optimization module is used to capture the spatiotemporal feature data of the first correction data, and optimize and determine the spatiotemporal feature data to obtain optimized structural data.

7. The method for advanced prediction of earthquake waves in tunnels for water conservancy survey according to claim 1, characterized in that: The calculation process of the sub-region early warning coefficient is as follows: Acquire initial geological data and final geological data; pre-process the final geological data to obtain optimized geological data; identify geological difference areas and geological difference coefficients based on the optimized geological data and the initial geological data; Identify the positional relationship between geological difference areas and structural difference areas, and divide the geological difference areas into associated anomaly areas and unassociated anomaly areas, and set weights for each; The sub-region warning coefficient is calculated based on the associated abnormal area, the unassociated abnormal area, the geological difference coefficient and the structural damage coefficient; the calculation formula of the sub-region warning coefficient is: Among them, SUC represents the sub-region warning coefficient; loi represents the structural damage coefficient; α1 represents the associated area weight; aar i represents the volume of the associated abnormal region i; i represents the geological difference coefficient of the associated abnormal area i; AT1 represents the first threshold; α2 represents the weight of the unassociated area; nar j represents the volume of the unassociated abnormal region j; ω j represents the geological difference coefficient of the unrelated abnormal area j; AT2 represents the second threshold; exp represents an exponential function with a natural constant as the base; n represents the number of related abnormal areas; m represents the number of unrelated abnormal areas.

Images