In-situ imaging and fracture identification method and system for borehole in geological exploration

Abstract

This invention belongs to the field of engineering geological exploration technology and discloses a method and system for in-situ borehole imaging and fracture identification in geological exploration. The method includes drilling multiple micro-holes in the borehole wall for geological exploration and placing an optical fiber in each micro-hole; emitting pulsed laser light into the optical fiber to acquire backscattered Rayleigh and Raman scattered light; determining the acoustic characteristics of the borehole wall based on the backscattered Rayleigh light and the temperature characteristics based on the Raman scattered light; and inputting the acoustic and temperature characteristics into a convolutional neural network to obtain a fracture probability map of the borehole wall. This invention relies on distributed optical fibers to synchronously acquire in-situ acoustic and temperature data of the borehole wall, eliminating the need for manual segment-by-segment inspection of the borehole wall and enabling rapid fracture identification across the entire borehole, significantly reducing identification time. This invention utilizes the above two complementary types of data to accurately capture minute fractures that are easily missed by conventional methods, reducing the probability of misjudgment and missed detection.

Description

Technical Field

[0001] This invention belongs to the field of engineering geological exploration technology, and specifically discloses a method and system for borehole in-situ imaging and fracture identification in geological exploration. Background Technology

[0002] Traditional geological surveys for pumped-storage power stations typically employ core drilling and artificial fracture statistics to identify potential geological hazards such as landslides, karst collapses, and earthquake liquefaction. However, conventional core drilling has a low core recovery rate, making it difficult to fully reveal the structural characteristics of deep rock masses. Artificial fracture statistics suffer from significant drawbacks, including low efficiency (logging a single borehole takes 4-6 hours) and poor data comparability (interpretations by different technicians can differ by up to 30%). More critically, in complex geological areas such as strongly weathered rock layers and fault fracture zones, the identification rate of hidden microfractures (aperture <1mm) and secondary fractures is less than 50%, and these structural surfaces are often key factors inducing engineering disasters such as reservoir leakage and cavern collapse.

[0003] Therefore, it is urgent to introduce new technologies to improve the efficiency and accuracy of crack identification, so as to provide reliable support for the site selection, design optimization and disaster prevention of pumped storage power stations. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for borehole in-situ imaging and fracture identification in geological exploration, so as to solve the technical problems of low core recovery rate, poor fracture identification efficiency, high rate of missed detection of microfractures and large error of manual interpretation in existing fracture identification methods.

[0005] The first aspect of the present invention provides a method for in-situ borehole imaging and fracture identification in geological exploration, comprising:

[0006] Step 1: Drill multiple microholes in the borehole wall for geological exploration and install optical fibers in each microhole;

[0007] Step 2: Emit pulsed laser light into the optical fiber to obtain backscattered Rayleigh light and Raman scattered light in the optical fiber;

[0008] Step 3: Determine the acoustic characteristics of the borehole wall based on the backscattered Rayleigh light, and determine the temperature characteristics of the borehole wall based on the Raman scattered light;

[0009] Step 4: Input the acoustic wave features and the temperature features into a convolutional neural network to obtain the crack probability map of the borehole wall.

[0010] Preferably, after step 4, the method further includes:

[0011] Step 5: Determine the characteristic frequency of water flow on the borehole wall based on the backscattered Rayleigh light;

[0012] Step 6: Construct a three-dimensional borehole wall image based on the fracture probability map and the water flow characteristic frequency.

[0013] Preferably, after step 6, the method further includes:

[0014] The preset crack probability threshold is compared with the crack probability value in the three-dimensional hole wall image, and a warning message is marked on the three-dimensional hole wall image based on the comparison result.

[0015] Preferably, the acoustic characteristics of the borehole wall are determined based on the backscattered Rayleigh light, specifically as follows:

[0016] The backscattered Rayleigh light is converted into a first electrical signal;

[0017] The first electrical signal is denoised to obtain a first denoised signal;

[0018] The first noise-reduced signal is subjected to a short-time Fourier transform to generate a time spectrum, and the acoustic characteristics of the borehole wall are determined based on the time spectrum.

[0019] Preferably, the acoustic wave characteristics include the reflection coefficient and the frequency domain energy ratio.

[0020] Preferably, the first electrical signal is denoised to obtain a first denoised signal, specifically as follows:

[0021] The first denoised signal is obtained by denoising the first electrical signal using wavelet transform.

[0022] Preferably, the temperature characteristics of the borehole wall are determined based on the Raman scattered light, specifically as follows:

[0023] The Raman scattered light is converted into a second electrical signal;

[0024] The second electrical signal is subjected to environmental drift elimination to obtain the second corrected signal;

[0025] Determine the axial temperature gradient and temperature-time derivative of the second corrected signal.

[0026] Preferably, before step 4, the method further includes:

[0027] The acoustic wave features and temperature features are aligned in time and location, and the spatial resolution is unified.

[0028] Preferably, step 1 specifically includes:

[0029] A micro-water jet drill bit is used to drill a micro-hole at preset intervals on the borehole wall in a spiral arrangement.

[0030] Armored single-mode optical fiber is embedded in the micropore and acoustic coupling gel is injected into the micropore;

[0031] The micropores into which the acoustic coupling gel was injected were sealed using a curing resin.

[0032] A second aspect of the present invention provides a system based on the above-mentioned borehole in-situ imaging and fracture identification method in geological exploration, comprising:

[0033] An optical signal acquisition module is used to emit pulsed laser light into an optical fiber located in a micro-hole on the borehole wall, and to acquire backscattered Rayleigh light and Raman scattered light in the optical fiber.

[0034] A feature determination module is used to determine the acoustic characteristics of the borehole wall based on the backscattered Rayleigh light, and to determine the temperature characteristics of the borehole wall based on the Raman scattered light.

[0035] A crack identification module is used to input the acoustic features and the temperature features into a convolutional neural network to obtain a crack probability map of the borehole wall.

[0036] The borehole in-situ imaging and fracture identification method and system for geological exploration of the present invention have the following advantages compared with the prior art:

[0037] This invention relies on distributed optical fiber to synchronously collect acoustic and temperature data of the entire borehole wall, eliminating the need for manual inspection of the borehole wall segment by segment. It can quickly complete the screening of cracks throughout the entire borehole, significantly reducing the identification time.

[0038] This invention integrates two complementary physical features, sound waves and temperature, and uses a convolutional neural network to perform intelligent feature extraction. It can accurately capture subtle cracks that are easily missed by conventional methods, reducing the probability of misjudgment and missed judgment.

[0039] This invention utilizes a fiber optic sensing-based detection method with strong anti-interference capabilities, enabling stable operation in the complex underground environment of boreholes, ensuring the reliability of data acquisition, and ultimately outputting an intuitive fracture probability map. This provides clear judgment criteria for geological exploration and reliable technical support for the safe construction and operation of pumped storage power stations. Attached Figure Description

[0040] Figure 1 This is a flowchart of the borehole in-situ imaging and fracture identification method in geological exploration according to an embodiment of the present invention. Detailed Implementation

[0041] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the invention. However, those skilled in the art will understand that the invention can be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods are omitted so as not to obscure the description of the invention with unnecessary detail.

[0042] The first aspect of this invention provides a method for in-situ borehole imaging and fracture identification in geological exploration, such as... Figure 1 As shown, it includes:

[0043] Step 1: Drill multiple micro-holes in the borehole wall for geological exploration, and install an optical fiber in each micro-hole. Specifically:

[0044] Step 1.1: Using a micro water jet drill bit, drill a microhole at preset intervals on the borehole wall in a spiral pattern.

[0045] In this embodiment of the invention, before drilling microholes, a high-pressure water jet is used to flush the borehole wall to remove rock cuttings and mud, preventing them from affecting the subsequent fiber optic cable installation. Then, a fiber optic grating marker is placed every meter on the borehole wall for real-time compensation of fiber temperature and strain. Finally, a micro waterjet drill bit is used to drill microholes of a predetermined depth at predetermined intervals on the borehole wall in a spiral arrangement. These predetermined intervals can be 40cm to 60cm, and the predetermined depths can be 2cm to 3cm. For example, in this embodiment, a 3mm micro waterjet drill bit is used to drill microholes on the borehole wall at axial intervals of 50cm, with a depth of 2cm. The microholes are arranged in a spiral pattern to avoid stress concentration.

[0046] Step 1.2: Embed the armored single-mode fiber into the micropore and inject acoustic coupling gel into the micropore.

[0047] The embodiments of the present invention use armored single-mode optical fiber that is resistant to high temperature and high pressure, with a temperature resistance of over 150℃ and a tensile strength of ≥100kN.

[0048] For example, in this embodiment of the invention, a single-mode optical fiber with a wavelength of 1550nm and an attenuation of ≤0.2dB / km is selected as the core of the armored single-mode optical fiber; a polyimide coating is selected as the outer layer material of the armored single-mode optical fiber for coating, and the thickness of the polyimide coating is controlled within the range of 50μm-80μm to ensure that the optical fiber still has good mechanical properties and flexibility in a high-temperature environment; the coated optical fiber is subjected to high-temperature curing treatment, with the curing temperature set at 250℃-300℃ and the duration not less than 2 hours, to enhance the bonding force between the coating and the optical fiber substrate; after curing, the optical fiber is subjected to tensile strength test and temperature resistance performance verification to ensure that the armored single-mode optical fiber can withstand temperatures above 150℃, withstand pressure ≥20MPa, and has a tensile strength ≥100kN.

[0049] In this embodiment of the invention, armored single-mode optical fiber is embedded in a micro-hole, and then an expandable rubber packer is used to tightly attach the armored single-mode optical fiber to the hole wall to ensure effective acoustic wave transmission; at the same time, an optical fiber grid is arranged along the annular space around the drilled micro-hole.

[0050] To ensure effective sound wave transmission, coupling enhancement is required. Specifically, polyurethane gel is used as the acoustic coupling gel, which is heat-resistant and hydrolysis-resistant. An expandable rubber packer is used to tightly attach the armored single-mode optical fiber to the hole wall, and the gel is pumped through the microchannel built into the expandable rubber packer to ensure uniform coverage. After the optical fiber is laid, periodic hydraulic pulses of 0.1MPa to 0.5MPa are applied to allow the gel to fully penetrate into the microcracks.

[0051] Step 1.3: Use curing resin to seal the micropores of the injected acoustic coupling gel.

[0052] In this embodiment of the invention, the micropores are sealed by covering them with a rapidly curing resin.

[0053] Step 2: Emit pulsed laser light into the optical fiber to obtain backscattered Rayleigh light and Raman scattered light in the optical fiber.

[0054] Step 3: Determine the acoustic characteristics of the borehole wall based on the backscattered Rayleigh light, and determine the temperature characteristics of the borehole wall based on the Raman scattered light; the acoustic characteristics include the reflection coefficient and the frequency domain energy ratio; the temperature characteristics include the axial temperature gradient and the temperature time derivative.

[0055] The above-mentioned determination of the acoustic characteristics of the borehole wall based on backscattered Rayleigh light is as follows:

[0056] The backscattered Rayleigh light is converted into a first electrical signal; the first electrical signal is denoised using wavelet transform to obtain a first denoised signal; the first denoised signal is subjected to short-time Fourier transform to generate a time spectrum, and the acoustic characteristics of the borehole wall are determined based on the time spectrum.

[0057] Specifically, in this embodiment of the invention, a distributed acoustic wave sensing system is composed of an acoustic wave hardware subsystem, an acoustic wave signal processing unit, and an acoustic wave auxiliary module.

[0058] The acoustic-hardware subsystem generates a pulsed laser, captures the backscattered Rayleigh light in the optical fiber, and converts the backscattered Rayleigh light into a first electrical signal.

[0059] The acoustic signal processing unit receives the first electrical signal through the acoustic high-speed data acquisition card and performs phase demodulation and spectrum analysis on the first electrical signal, outputting the vibration event characteristics (i.e. the subsequent water flow characteristic frequency).

[0060] The acoustic wave-auxiliary module is connected to the acoustic wave-hardware subsystem and the acoustic wave-signal processing unit through a combination of direct hardware connection and software interaction. The acoustic wave-auxiliary module is used to synchronize the acoustic wave-high-speed data acquisition card and the acoustic wave-signal processing unit, correct the thermal drift of the optical fiber, and compress the first electrical signal and vibration event characteristics through the edge computing node.

[0061] The acoustic wave hardware subsystem of this invention consists of an acoustic wave laser emission module, an acoustic wave fiber optic sensing network module, and an acoustic wave optical receiving and detection module. The acoustic wave signal processing unit consists of a real-time signal demodulation module, a vibration event detection module, and a spatial positioning and imaging module. The acoustic wave auxiliary module consists of an acoustic wave synchronization and control module, an acoustic wave data storage and communication module, and an acoustic wave calibration and diagnosis module.

[0062] 1. Sound wave - hardware subsystem composition

[0063] 1.1 Composition of the acoustic-laser emission module:

[0064] 1.1.1 Select a narrow linewidth laser with a wavelength of 1550nm (to be compatible with the low-loss window of optical fiber), a linewidth of <1kHz (to ensure coherence), and an output power of 50mW-200mW (to adapt to different distances);

[0065] 1.1.2 Select a pulse modulator with electro-optic modulation, a pulse width of 1ns-20ns (to correspond to a spatial resolution of 0.1m-2m), and a repetition frequency of 1kHz-10kHz (to balance with the detection distance);

[0066] 1.1.3 The acoustic-laser emission module consists of a narrow linewidth laser and a pulse modulator.

[0067] 1.2 The acoustic-fiber sensing network module consists of armored single-mode fiber and an optical path coupling assembly composed of a circulator and a beam splitter.

[0068] 1.3 Composition of the acoustic-optical receiving and detection module:

[0069] 1.3.1 Select a balanced detector with a bandwidth ≥ 200MHz (to cover the acoustic frequency band) and a common-mode rejection ratio > 40dB (to suppress laser intensity noise);

[0070] 1.3.2 Select a high-speed acoustic data acquisition card with a sampling rate ≥100MS / s (to meet the spatiotemporal resolution requirements) and a resolution of 16 bits (dynamic range >80dB);

[0071] 1.3.3 The acoustic wave-optical receiving and detection module consists of a balanced detector and an acoustic wave-high-speed data acquisition card.

[0072] 2. Composition of the acoustic wave-signal processing unit

[0073] 2.1 Composition of the real-time signal demodulation module:

[0074] 2.1.1 The phase change of backscattered Rayleigh light is detected by using a phase-sensitive light temporal reflection constructed by an I / Q demodulation algorithm and a differential phase extraction algorithm;

[0075] 2.1.2 In the time domain, the sound wave spectrum (1Hz~10kHz) is first analyzed by short-time Fourier transform, and then the vibration signals of different frequency bands are separated by wavelet transform.

[0076] 2.2 Composition of the vibration event detection module:

[0077] 2.2.1 Extract the peak amplitude and duration of the acoustic signal in the time domain, and extract the spectral entropy and dominant frequency component of the acoustic signal in the frequency domain;

[0078] 2.2.2 The extracted features are used to train the convolutional neural network and the long short-term memory network, and finally the micro-vibration, water flow noise and mechanical interference are classified.

[0079] 2.3 Composition of the Spatial Positioning and Imaging Module:

[0080] 2.3.1 The vibration location is calibrated using the results obtained from the vibration event detection module;

[0081] 2.3.2 Mapping vibration intensity to a three-dimensional visualization of the borehole wall.

[0082] 3. Composition of the acoustic wave-auxiliary module

[0083] 3.1 Composition of the acoustic wave synchronization and control module:

[0084] 3.1.1 Data alignment across multiple systems using atomic clocks;

[0085] 3.1.2 Thermal drift is corrected using a fiber optic grating reference point, and temperature and strain effects are distinguished using a dual-wavelength method;

[0086] 3.2 Composition of the acoustic wave-data storage and communication module:

[0087] 3.2.1 Real-time processing of low-latency signal demodulation and data compression using FPGA with only storage of abnormal event fragments;

[0088] 3.3 Composition of the Acoustic Wave Calibration and Diagnostic Module:

[0089] 3.3.1 A reference signal of known frequency is generated using a piezoelectric ceramic vibrator to calibrate the signal source;

[0090] 3.3.2 Optical power monitoring is used to detect fiber optic breaks or aging for health monitoring.

[0091] This invention utilizes a distributed acoustic wave sensing system to record the acoustic wave signal reflected from the orifice wall and the characteristic frequency of the water flow in real time, and uses wavelet transform to remove mechanical vibration interference from the acoustic wave signal to obtain an acoustic spectrum, including:

[0092] The first electrical signal was recorded using a distributed acoustic wave sensing system at a sampling rate of 10 kHz, and a GPS clock was used to ensure that the acquisition time error between the distributed acoustic wave sensing system and the subsequent distributed temperature sensing system was <1 μs.

[0093] A first denoised signal is obtained by using a distributed acoustic wave sensing system to perform noise reduction processing on the first electrical signal and suppress isotropic environmental noise.

[0094] A distributed acoustic wave sensing system is used to generate a time spectrum by performing a short-time Fourier transform on the first noise-reduced signal, and the reflection coefficient and frequency domain energy ratio are extracted to generate an acoustic spectrum.

[0095] The detailed method for obtaining the above spectrograms is as follows:

[0096] 1. Distributed acoustic signal acquisition

[0097] 1.1 A distributed acoustic wave sensing system is used to record the original acoustic wave vibration signal in real time in the time domain with a sampling rate of 10kHz, in order to capture instantaneous events such as rock mass fracturing.

[0098] 1.2 The distributed acoustic wave sensing system is used to analyze the acoustic wave spectrum (1Hz~10kHz) in the frequency domain through short-time Fourier transform in order to distinguish different types of vibration sources;

[0099] 1.3 A GPS clock is used to ensure that the acquisition time of the distributed acoustic wave sensing system is aligned with that of the subsequent distributed temperature sensing system, and the acquisition time error is guaranteed to be less than 1μs.

[0100] 2. Distributed acoustic signal noise reduction process

[0101] 2.1 The Db4 wavelet basis was selected to perform wavelet transform on the acoustic signal after short-time Fourier transform, and the detail coefficients were divided into 5 levels.

[0102] 2.2 A maximum frequency band threshold of 50Hz is used to filter out acoustic signals caused by drilling mechanical vibration;

[0103] 2.3 Polarization filtering is used to suppress isotropic environmental noise, retaining only the reflection signal from the crack.

[0104] 3. Feature Extraction

[0105] 3.1 Calculate the reflection coefficient using the following formula:

[0106] (1)

[0107] in, Represents depth, Represents the reflection coefficient. This represents the power of the reflected sound wave at depth z. This represents the incident sound wave power at depth z.

[0108] 3.2 The condition for determining cracks is when When the value is >-40dB, it is marked as a potential crack.

[0109] 3.3 After performing a short-time Fourier transform on the time-domain signal (window length 100 ms, overlap 50%), the energy proportion of the characteristic frequency band (100Hz-500Hz) is calculated using the following formula:

[0110] (2)

[0111] in, The energy percentage representing the characteristic frequency band (100-500Hz), This represents the signal power spectral density.

[0112] 3.4 The criterion for determining seepage noise is when A value greater than 0.3 is considered a seepage characteristic.

[0113] In this embodiment of the invention, the temperature characteristics of the borehole wall are determined based on Raman scattered light, specifically by: converting the Raman scattered light into a second electrical signal; performing environmental drift elimination on the second electrical signal to obtain a second corrected signal; and determining the axial temperature gradient and temperature-time derivative of the second corrected signal.

[0114] This invention utilizes a temperature-hardware subsystem, a temperature-signal processing unit, and a temperature-auxiliary module to form a distributed temperature sensing system, including:

[0115] After generating pulsed laser light using a temperature-hardware subsystem, Raman scattered light in an optical fiber is captured and converted into a second electrical signal.

[0116] The temperature-signal processing unit receives the second electrical signal through the temperature-high-speed data acquisition card, performs intensity ratio calculation and time-domain positioning on the second electrical signal, and outputs the temperature distribution.

[0117] The temperature-auxiliary module is connected to the temperature-hardware subsystem and the temperature-signal processing unit through a combination of direct hardware connection and software interaction.

[0118] The temperature-auxiliary module synchronizes the pulsed laser and the temperature-high-speed data acquisition card to provide a GPS time reference, transmits the temperature compensation coefficient, and compresses the second electrical signal and temperature distribution through the edge computing node.

[0119] The temperature hardware subsystem consists of a temperature-laser emission module, a temperature-fiber optic sensing network module, and a temperature-optical receiving and detection module. The temperature signal processing unit consists of a temperature demodulation module, a noise suppression and data optimization module, and a temperature field visualization module. The temperature auxiliary module consists of a temperature synchronization and control module, a temperature data storage and communication module, and a temperature calibration and diagnostic module.

[0120] 1. Temperature - Hardware Subsystem Composition

[0121] 1.1 Composition of the Temperature-Laser Emission Module:

[0122] 1.1.1 Select a pulsed laser with a wavelength of 905nm (Raman scattering sensitive band), a pulse width of 5ns~10ns (to correspond to a spatial resolution of 0.5m~1m), and a peak power of 5W~10W (to ensure the Raman scattering signal intensity);

[0123] 1.1.2 Select an optical path modulation device consisting of an acousto-optic modulator for controlling the pulse frequency and width and an optical isolator for preventing reflected light from damaging the laser;

[0124] 1.1.3 The temperature-laser emission module consists of a pulsed laser and an optical path modulation device.

[0125] 1.2 The temperature-fiber sensing network module consists of armored single-mode fiber and an optical path coupling assembly composed of a fiber circulator and a Raman beam splitter.

[0126] 1.3 Composition of the temperature-optical receiving and detection module:

[0127] 1.3.1 A dual-channel detector is selected, consisting of an InGaAs detector with a wavelength of 1050nm~1100nm to detect the Stokes channel and a Si-APD detector with a wavelength of 900~950nm to detect the anti-Stokes channel.

[0128] 1.3.2 Select a high-speed temperature data acquisition card with a sampling rate ≥100MS / s (to meet the requirements of high spatial resolution) and a resolution >16 bits (to ensure temperature accuracy);

[0129] 1.3.3 The temperature-optical receiving and detection module consists of a dual-channel detector and a high-speed temperature data acquisition card.

[0130] 2. Temperature-Signal Processing Unit Composition

[0131] 2.1 Composition of the temperature demodulation module:

[0132] 2.1.1 The Raman scattering intensity ratio is calculated using the Raman scattering intensity ratio calculation algorithm. The calculation formula is as follows:

[0133] (3)

[0134] in, The ratio of Raman scattering intensity represents Represents anti-Stokes light intensity, Represents Stokes light intensity, and These represent different calibration coefficients.

[0135] 2.1.2 The location of the temperature point is calculated based on the round-trip time of the light pulse. The calculation formula is as follows:

[0136] (4)

[0137] in, Represents the location of the temperature point. Represents the speed of light. This represents the measured temperature. Represents the refractive index of the optical fiber. ≈1.47.

[0138] 2.2 Composition of the noise suppression and data optimization module:

[0139] 2.2.1 Reduce random noise by performing multiple samplings through cumulative averaging;

[0140] 2.2.2 High-frequency noise and effective temperature signal are separated by wavelet transform;

[0141] 2.2.3 Correct system errors using reference temperature points.

[0142] 2.3 Composition of the temperature field visualization module:

[0143] 2.3.1 Plot a one-dimensional temperature curve along the length of the optical fiber to show the temperature distribution.

[0144] 2.3.2 Reconstruct the temperature field of the measured object by combining data from multiple optical fibers to perform three-dimensional imaging.

[0145] 3. Temperature-Auxiliary Module Composition

[0146] 3.1 Composition of the temperature synchronization and control module:

[0147] 3.1.1 Data alignment across multiple systems is achieved using a high-precision clock;

[0148] 3.1.2 The detector sensitivity is dynamically adjusted through automatic gain control to adapt to signal strength at different distances.

[0149] 3.2 Temperature-Data Storage and Communication Module Composition:

[0150] 3.2.1 Utilize edge computing nodes to process temperature data in real time and compress and store it;

[0151] 3.2.2 Upload data to the monitoring platform using wireless transmission.

[0152] 3.3 Temperature-Calibration and Diagnostic Module Composition:

[0153] 3.3.1 Temperature sensitivity is periodically calibrated using a blackbody furnace.

[0154] 3.3.2 Optical fiber breakage or aging can be detected by optical time domain reflectometer for health monitoring.

[0155] This invention utilizes a distributed temperature sensing system to record the temperature signal of an optical fiber in real time and eliminates environmental thermal drift of the temperature signal to obtain a temperature gradient map, including:

[0156] A distributed temperature sensing system is used to acquire a second electrical signal on the borehole wall every 5 minutes, and the drift is corrected by a fiber optic grating reference point system.

[0157] A second corrected signal is obtained by using a distributed temperature sensing system to perform environmental thermal drift elimination on the second electrical signal through an environmental drift elimination formula. The environmental drift elimination formula is as follows:

[0158] (5)

[0159] in, This represents the second correction signal. Represents the original measured temperature. Represents the thermal diffusivity compensation coefficient. Represents historical reference temperature. This represents the reference temperature of the distributed temperature sensing system.

[0160] A distributed temperature sensing system is used to calculate the axial temperature gradient and temperature-time derivative of the temperature signal and output a temperature gradient map. The formula for calculating the axial temperature gradient is:

[0161] (6)

[0162] in, Represents the axial temperature gradient. This represents the temperature measurement value at depth z. Represents the spatial step size. =0.5m.

[0163] The formula for calculating the temperature-time derivative is:

[0164] (7)

[0165] in, Represents the time derivative of temperature. This represents the temperature measurement at depth z at time t. Represents the sampling time interval. =5min.

[0166] Step 4: Input the acoustic wave features and temperature features into the convolutional neural network to obtain the probability map of cracks in the borehole wall.

[0167] Before step 4, the present invention further includes:

[0168] The acoustic and temperature features are aligned in time and location, and the spatial resolution is unified.

[0169] In this embodiment of the invention, the first noise-reduced signal obtained by wavelet transform processing of the distributed acoustic wave sensing system is represented in the form of time-depth-frequency to obtain the third acoustic wave signal, and the second correction signal obtained by eliminating environmental thermal drift using the distributed temperature sensing system is represented in the form of time-depth-temperature to obtain the fourth temperature signal.

[0170] The third acoustic signal and the fourth temperature signal were strictly aligned using GPS timestamps and the depth coordinates of the micropores, and the spatial resolution was unified to 0.5m using linear interpolation.

[0171] The reflection coefficient and frequency domain energy ratio of the third acoustic signal are extracted as acoustic features, and the axial temperature gradient and temperature-time derivative of the fourth temperature signal are extracted as temperature features.

[0172] The acoustic wave features and temperature features are fused into multimodal features. After the multimodal features are normalized, they are input into a two-branch convolutional neural network, and the multimodal features are 4-dimensional features.

[0173] A two-branch convolutional neural network is trained to output the fracture probability, aperture estimate, and seepage activity score, and a three-dimensional borehole wall image is drawn based on the fracture probability, aperture estimate, and seepage activity score.

[0174] The above steps are as follows:

[0175] 1. Feature Alignment and Stitching

[0176] 1.1 Align the reflection coefficient calculated in the distributed acoustic wave sensing system with the axial temperature gradient calculated in the distributed temperature sensing system according to depth, and use linear interpolation to compensate for resolution differences.

[0177] 1.2 Fusion of 4-dimensional feature vectors:

[0178] (8)

[0179] in, This is the fused 4-dimensional feature vector.

[0180] 2. Calculate the crack probability using the following formula:

[0181] (9)

[0182] in, Represents the probability of a crack. Represents the Sigmoid function. The weights of the reflection coefficients represent the weights of the distributed acoustic wave sensing system. The temperature gradient weights represent the weights of a distributed temperature sensing system. This represents the bias, = -1.5.

[0183] 3. Calculate the opening estimate

[0184] 3.1 Calculation of aperture based on acoustic wave signal;

[0185] 3.1.1 Basic opening calculation of the opening model, the calculation formula is as follows:

[0186] (10)

[0187] in, The basic opening size represents the opening size in the opening size model. and Both represent the calibrated parameters. =0.01, =0.05.

[0188] 3.1.2 Frequency Domain Attenuation Compensation: High-frequency sound waves (>1kHz) attenuate more strongly in wide gaps. The opening is corrected by adjusting the spectral slope. The calculation formula is as follows:

[0189] (11)

[0190] in, This represents the corrected opening degree. The power spectrum slope represents the frequency band 100-500 Hz. Represents an empirical coefficient. =0.2.

[0191] 3.2 Calculation of opening degree based on temperature signal;

[0192] 3.2.1 The rate of temperature change caused by seepage is affected by the fracture aperture. The thermal diffusion constraint is calculated by inverting the heat conduction equation. The calculation formula is as follows:

[0193] (12)

[0194] in, Represents the opening degree of the temperature system. Represents the thermal diffusivity of the rock mass. =1×10 -6 m 2 / s, Represents the initial temperature. This represents the current temperature.

[0195] 3.2.2 Joint optimization: Utilizing the minimization of the weighted difference between the acoustic wave and temperature opening estimates, the calculation formula is as follows:

[0196] (13)

[0197] in, This represents the final aperture estimate. Represents a weight of 1. =0.7, Represents a weight of 2. =0.3.

[0198] 4. Calculate the seepage activity score

[0199] 4.1 Dynamic feature extraction;

[0200] 4.1.1 Calculation of seepage noise index for distributed acoustic wave sensing system, the calculation formula is as follows:

[0201] (14)

[0202] in, This represents the seepage noise index. Represents the energy of the characteristic frequency band of seepage. This represents the time-domain variance of the reflection coefficient.

[0203] 4.1.2 Calculation of thermal anomaly index for distributed temperature sensing system, the calculation formula is as follows:

[0204] (15)

[0205] in, Represents thermal anomaly indicators, This represents the step function.

[0206] 4.2 Scoring fusion and normalization;

[0207] 4.2.1 Calculate the overall score using the following formula:

[0208] (16)

[0209] in, Represents the overall score. and Represents two parameters, =1.2, =1.0.

[0210] 4.2.2 Dynamic threshold adjustment: The threshold is set to 0.7 during normal seasons. When the comprehensive score is >0.7, it is judged as active seepage.

[0211] The threshold for the rainy season is set at 0.6. When the comprehensive score is greater than 0.6, it is considered to be active seepage.

[0212] Following step 4, this embodiment of the invention further includes:

[0213] Step 5: Determine the characteristic frequency of water flow on the borehole wall based on the backscattered Rayleigh light.

[0214] Step 6: Construct a three-dimensional borehole wall image based on the fracture probability map and water flow characteristic frequency.

[0215] The preset crack probability threshold is compared with the crack probability value in the three-dimensional borehole wall image, and the result of the comparison determines whether to mark the warning information on the three-dimensional borehole wall image.

[0216] For example, a preset crack probability threshold is set to 0.7. When a region with a crack probability value greater than the preset crack probability threshold appears in the 3D borehole wall image, an early warning is issued. The early warning includes the 3D borehole wall image at the location exceeding the preset crack probability threshold (highlighted in red) and the crack probability value exceeding the preset crack probability threshold.

[0217] A second aspect of the present invention provides a system based on the above-mentioned borehole in-situ imaging and fracture identification method in geological exploration, comprising:

[0218] The optical signal acquisition module is used to emit pulsed laser light into the optical fiber located in the micro-hole on the borehole wall, and to acquire the backscattered Rayleigh light and Raman scattered light in the optical fiber.

[0219] The feature determination module is used to determine the acoustic characteristics of the borehole wall based on backscattered Rayleigh light and the temperature characteristics of the borehole wall based on Raman scattering light.

[0220] The fracture identification module is used to input acoustic and temperature features into a convolutional neural network to obtain fracture probability, aperture estimation, and seepage activity score. Based on the fracture probability, aperture estimation, and seepage activity score, a three-dimensional borehole wall image is drawn to obtain a fracture probability map of the borehole wall.

[0221] The in-situ borehole imaging and fracture identification method of this invention for geological exploration firstly uses a micro waterjet drill bit to precisely prepare micropores in the borehole wall and implants an optical fiber sensing network, ensuring sensing stability through sealing treatment. Subsequently, a distributed acoustic wave sensing system and a distributed temperature sensing system are constructed to collect acoustic wave signals and temperature signals reflected from the borehole wall, respectively. Then, a dual-branch convolutional neural network is designed to perform feature fusion analysis on acoustic spectrum features and temperature features, generating a high-precision fracture probability map, and reconstructing a three-dimensional borehole wall image by combining water flow characteristic frequencies. Finally, potential risk areas are automatically identified and early warnings are triggered by setting a preset threshold. This invention solves the problems of low core recovery rate, poor fracture identification efficiency, high microfracture missed detection rate, and large human interpretation errors in existing technologies. This improves the timeliness and accuracy of engineering decision-making, ensuring the safe and economical operation of the power plant throughout its entire life cycle.

[0222] This invention relies on distributed optical fiber to synchronously collect acoustic and temperature data of the entire borehole wall, eliminating the need for manual inspection of the borehole wall segment by segment. It can quickly complete the screening of cracks throughout the entire borehole, significantly reducing the identification time.

[0223] This invention integrates two complementary physical features, sound waves and temperature, and uses a convolutional neural network to perform intelligent feature extraction. It can accurately capture subtle cracks that are easily missed by conventional methods, reducing the probability of misjudgment and missed judgment.

[0224] This invention utilizes a fiber optic sensing-based detection method with strong anti-interference capabilities, enabling stable operation in the complex underground environment of boreholes, ensuring the reliability of data acquisition, and ultimately outputting an intuitive fracture probability map. This provides clear judgment criteria for geological exploration and reliable technical support for the safe construction and operation of pumped storage power stations.

[0225] The above descriptions are merely a few embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any modifications or alterations made by those skilled in the art without departing from the scope of the technical solution of the present invention using the disclosed technical content are equivalent to equivalent implementation cases and fall within the scope of the technical solution.

Claims

1. A method for in-situ borehole imaging and fracture identification in geological exploration, characterized in that, include: Step 1: Drill multiple microholes in the borehole wall for geological exploration and install optical fibers in each microhole; Step 2: Emit pulsed laser light into the optical fiber to obtain backscattered Rayleigh light and Raman scattered light in the optical fiber; Step 3: Determine the acoustic characteristics of the borehole wall based on the backscattered Rayleigh light, and determine the temperature characteristics of the borehole wall based on the Raman scattered light; Step 4: Input the acoustic wave features and the temperature features into a convolutional neural network to obtain the crack probability map of the borehole wall.

2. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 1, characterized in that, Following step 4, the following is also included: Step 5: Determine the characteristic frequency of water flow on the borehole wall based on the backscattered Rayleigh light; Step 6: Construct a three-dimensional borehole wall image based on the fracture probability map and the water flow characteristic frequency.

3. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 2, characterized in that, Following step 6, the following is also included: The preset crack probability threshold is compared with the crack probability value in the three-dimensional hole wall image, and a warning message is marked on the three-dimensional hole wall image based on the comparison result.

4. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 1, characterized in that, The acoustic characteristics of the borehole wall are determined based on the backscattered Rayleigh light, specifically as follows: The backscattered Rayleigh light is converted into a first electrical signal; The first electrical signal is denoised to obtain a first denoised signal; The first noise-reduced signal is subjected to a short-time Fourier transform to generate a time spectrum, and the acoustic characteristics of the borehole wall are determined based on the time spectrum.

5. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 4, characterized in that, The acoustic wave characteristics include the reflection coefficient and the frequency domain energy ratio.

6. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 4, characterized in that, The first denoised signal is obtained by denoising the first electrical signal, specifically as follows: The first denoised signal is obtained by denoising the first electrical signal using wavelet transform.

7. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 1, characterized in that, The temperature characteristics of the borehole wall are determined based on the Raman scattered light, specifically as follows: The Raman scattered light is converted into a second electrical signal; The second electrical signal is subjected to environmental drift elimination to obtain the second corrected signal; Determine the axial temperature gradient and temperature-time derivative of the second corrected signal.

8. The borehole in-situ imaging and fracture identification method in geological exploration according to claim 1, characterized in that, Before step 4, the following is also included: The acoustic wave features and temperature features are aligned in time and location, and the spatial resolution is unified.

9. The method for in-situ borehole imaging and fracture identification in geological exploration according to claim 1, characterized in that, Step 1 is as follows: A micro-water jet drill bit is used to drill a micro-hole at preset intervals on the borehole wall in a spiral arrangement. Armored single-mode optical fiber is embedded in the micropore and acoustic coupling gel is injected into the micropore; The micropores into which the acoustic coupling gel was injected were sealed using a curing resin.

10. A system based on the borehole in-situ imaging and fracture identification method in geological exploration according to any one of claims 1-9, characterized in that, include: An optical signal acquisition module is used to emit pulsed laser light into an optical fiber located in a micro-hole on the borehole wall, and to acquire backscattered Rayleigh light and Raman scattered light in the optical fiber. A feature determination module is used to determine the acoustic characteristics of the borehole wall based on the backscattered Rayleigh light, and to determine the temperature characteristics of the borehole wall based on the Raman scattered light. A crack identification module is used to input the acoustic features and the temperature features into a convolutional neural network to obtain a crack probability map of the borehole wall.

Images