Optical fiber sensing signal processing method, device, equipment and medium

Abstract

This application discloses a fiber optic sensing signal processing method, apparatus, device, and medium, applied in the field of CO2 geological storage technology. It acquires raw fiber optic sensing signals from a distributed fiber optic sensing system located in a CO2 storage monitoring area; inputs the raw fiber optic sensing signals into a multi-scale broadband noise suppression model; through a multi-scale broadband noise suppression layer in the model, broadband noise at different scales is suppressed on the raw fiber optic sensing signals to obtain intermediate fiber optic sensing signals; through a cross-scale broadband noise suppression layer in the model, cross-scale broadband noise is suppressed on the intermediate fiber optic sensing signals to obtain effective fiber optic sensing signals. By adapting multi-scale broadband noise suppression layers to different scales of broadband noise and learning the correlation between broadband noise at different scales, the broadband noise suppression coverage is improved, and the accuracy of extracting effective fiber optic sensing signals is enhanced.

Description

Technical Field

[0001] This application relates to the field of CO2 geological storage technology, and in particular to a fiber optic sensing signal processing method, apparatus, equipment and medium. Background Technology

[0002] Carbon dioxide geological storage (CCUS) technology is a key component of carbon capture and storage (CCS) systems, designed to inject industrially emitted carbon dioxide (CO2) into deep underground spaces for long-term storage. To ensure the safety of CO2 storage, the CO2 injection process and storage status must be monitored in real time to prevent CO2 leakage or the induction of destructive earthquakes.

[0003] Distributed fiber optic sensing technology, due to its advantages such as high sensitivity, full distribution, and resistance to high temperature and pressure, has become the mainstream technology for CCUS monitoring, used to monitor microseismic events induced by CO2 injection, abnormal caprock strain, and precursor signals of leakage. However, industrial storage sites are usually located in complex human environments (such as oilfield operation areas and industrial parks), and fiber optic sensing signals are susceptible to multi-scale broadband noise interference, leading to signal distortion and affecting the monitoring effect of CO2 injection process and storage status.

[0004] Therefore, there is an urgent need for an intelligent method that can simultaneously process multi-scale broadband noise and extract effective fiber optic sensing signals with high fidelity for complex CCUS operating conditions. Summary of the Invention

[0005] This application provides a method, apparatus, device, and medium for processing fiber optic sensing signals to solve the problem of severe distortion in fiber optic sensing signals affecting the monitoring effect of CO2 storage. The technical solution provided by this application is as follows: On the one hand, this application provides a fiber optic sensing signal processing method, including: Acquire raw fiber optic sensing signals collected by a distributed fiber optic sensing system located in the CO2 storage monitoring area; The original fiber optic sensing signal is input into a pre-constructed multi-scale broadband noise suppression model for multi-scale broadband noise suppression to obtain an effective fiber optic sensing signal. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer. In the multi-scale broadband noise suppression model, the original fiber optic sensing signal is subjected to broadband noise suppression at different scales through the multi-scale broadband noise suppression layer to obtain an intermediate fiber optic sensing signal. The intermediate fiber optic sensing signal is then subjected to cross-scale broadband noise suppression through the cross-scale broadband noise suppression layer to obtain an effective fiber optic sensing signal.

[0006] Optionally, the multi-scale broadband noise suppression layer includes multiple parallel CNN networks; the intermediate fiber optic sensing signal is obtained by performing broadband noise suppression at different scales on the original fiber optic sensing signal through the multi-scale broadband noise suppression layer, including: Intermediate fiber optic sensing signals are obtained by performing broadband noise suppression at different scales on the original fiber optic sensing signals in parallel using multiple CNN networks.

[0007] Optionally, the cross-scale broadband noise suppression layer includes an LSTM network connected to multiple CNN networks; the effective fiber optic sensing signal is obtained by performing cross-scale broadband noise suppression on the intermediate fiber optic sensing signal through the cross-scale broadband noise suppression layer, including: Effective fiber optic sensing signals are obtained by performing cross-scale broadband noise suppression on the intermediate fiber optic sensing signals output by multiple CNN networks using an LSTM network.

[0008] On the one hand, this application provides a fiber optic sensing signal processing method, including: Obtain a multi-scale broadband noise sample set; wherein, the multi-scale broadband noise sample set includes each historical fiber optic sensing signal with broadband noise at different scales, as well as the calibration intermediate fiber optic sensing signal and the calibration effective fiber optic sensing signal corresponding to each historical fiber optic sensing signal. A multi-scale broadband noise suppression model is trained based on a multi-scale broadband noise sample set; the multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer. During the training of the multi-scale broadband noise suppression model, a combined loss function of multi-scale weighted loss function and cross-scale suppression loss function is used to update the network parameters of the multi-scale broadband noise suppression model.

[0009] Optionally, a combined loss function of multi-scale weighted loss function and cross-scale suppression loss function is used to update the network parameters of the multi-scale broadband noise suppression model, including: Based on the predicted intermediate fiber sensing signal and the calibrated intermediate fiber sensing signal of each historical fiber sensing signal, the multi-scale weighted loss value is calculated using a multi-scale weighted loss function; whereby the predicted intermediate fiber sensing signal is the output signal of the multi-scale broadband noise suppression layer. Based on the estimated effective fiber sensing signal and the calibrated effective fiber sensing signal from each historical fiber sensing signal, the cross-scale suppression loss value is calculated using the cross-scale suppression loss function; whereby the estimated effective fiber sensing signal is the output signal of the cross-scale broadband noise suppression layer. The total loss value is calculated based on the multi-scale weighted loss value and the cross-scale suppression loss value; The network parameters of the multi-scale broadband noise suppression model are updated based on the total loss value.

[0010] Optionally, the multi-scale weighted loss function is used to weight the suppression error of broadband noise at different scales according to its influence on the effective optical fiber sensing signal to obtain the multi-scale weighted loss value.

[0011] On the other hand, this application provides an optical fiber sensing signal processing device, comprising: The signal acquisition unit is used to acquire the raw fiber optic sensing signals collected by the distributed fiber optic sensing system set in the CO2 storage monitoring area. The noise suppression unit is used to input the original fiber optic sensing signal into a pre-trained multi-scale broadband noise suppression model for multi-scale broadband noise suppression to obtain an effective fiber optic sensing signal. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer. In the multi-scale broadband noise suppression model, the original fiber optic sensing signal is subjected to broadband noise suppression at different scales through the multi-scale broadband noise suppression layer to obtain an intermediate fiber optic sensing signal, and the intermediate fiber optic sensing signal is subjected to cross-scale broadband noise suppression through the cross-scale broadband noise suppression layer to obtain an effective fiber optic sensing signal.

[0012] On the other hand, this application provides an optical fiber sensing signal processing device, comprising: The sample acquisition unit is used to acquire a multi-scale broadband noise sample set; wherein, the multi-scale broadband noise sample set includes each historical fiber optic sensing signal with broadband noise at different scales, as well as the calibration intermediate fiber optic sensing signal and the calibration effective fiber optic sensing signal corresponding to each historical fiber optic sensing signal. The model training unit is used to train the multi-scale broadband noise suppression model based on a multi-scale broadband noise sample set. During the training process, a combination loss function of multi-scale weighted loss function and cross-scale suppression loss function is used to update the network parameters of the multi-scale broadband noise suppression model. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer.

[0013] On the other hand, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described fiber optic sensing signal processing method.

[0014] On the other hand, this application also provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the above-described fiber optic sensing signal processing method.

[0015] The beneficial effects of this application are as follows: On the one hand, this application accurately captures the characteristics of broadband noise at different scales through a multi-scale broadband noise suppression layer, achieving initial suppression of broadband noise at different scales and solving the problem of adapting to broadband noise at different scales. Subsequently, through a cross-scale broadband noise suppression layer, it accurately captures the cross-scale correlation characteristics between broadband noise at different scales (such as the coupling effect of slow low-scale temperature interference and mesoscale vibration noise), achieving coordinated suppression of cross-scale correlated broadband noise. This completes the comprehensive suppression from initial suppression at different scales to coordinated suppression across scales, thoroughly covering all-scale interference scenarios of broadband noise, improving the broadband noise suppression coverage, increasing the accuracy of extracting effective fiber optic sensing signals, and solving the problem of CO2 storage monitoring effects being affected by severe distortion of fiber optic sensing signals.

[0016] On the other hand, this application updates the network parameters of the multi-scale broadband noise suppression model by adopting a combination of multi-scale weighted loss function and cross-scale suppression loss function. This enables the multi-scale broadband noise suppression model to achieve both deep suppression of broadband noise and maximum preservation of key details of effective fiber optic sensing signals (such as minute strain changes in CO2 storage monitoring and weak acoustic signals in oil and gas well logging). This solves the contradiction between noise suppression and signal fidelity, and provides high-precision effective fiber optic sensing signal support for subsequent CO2 storage monitoring and analysis.

[0017] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a schematic flowchart of an optical fiber sensing signal processing method according to an embodiment of this application. Figure 2 This is a schematic flowchart of another fiber optic sensing signal processing method in the embodiments of this application; Figure 3 This is a schematic diagram of the composition structure of an optical fiber sensing signal processing device according to an embodiment of this application; Figure 4 This is a schematic diagram of the composition structure of another fiber optic sensing signal processing device in an embodiment of this application; Figure 5 This is a schematic diagram of the hardware structure of the electronic device in the embodiments of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and beneficial effects of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] The terms "first," "second," and similar words used in this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Words such as "including" or "comprising" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connected," "coupled," or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0021] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application. Furthermore, the embodiments and features in the embodiments of this application can be combined with each other without conflict.

[0022] This application provides a fiber optic sensing signal processing method, see below. Figure 1 As shown, the general flow of the fiber optic sensing signal processing method provided in this application embodiment is as follows: Step 101: Acquire the raw fiber optic sensing signal collected by the distributed fiber optic sensing system set in the CO2 storage monitoring area.

[0023] In this embodiment, a distributed fiber optic sensing system is deployed in the CO2 storage monitoring area (such as around the storage well or on the surface above the storage layer). This system uses phase-sensitive optical time-domain reflectometry (Φ-OTDR), with a fiber optic cable length of 5-10 km, a sampling frequency of 1 kHz, and a sampling duration of 1 hour. The system acquires raw fiber optic sensing signals (including strain signals, acoustic signals, superimposed with low-scale temperature drift noise, meso-scale formation micro-vibration noise, high-scale equipment electromagnetic interference noise, vehicle vibration, and non-geological noise from pump start-up and shutdown). After acquisition, the raw signals are preprocessed: extreme outliers are removed by mean filtering, and normalization is used to map the signal amplitude to the [0,1] interval, eliminating the influence of signal amplitude differences on subsequent processing.

[0024] Step 102: Input the original fiber optic sensing signal into the pre-constructed multi-scale broadband noise suppression model for multi-scale broadband noise suppression to obtain an effective fiber optic sensing signal; wherein, the multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer; in the multi-scale broadband noise suppression model, the original fiber optic sensing signal is subjected to broadband noise suppression at different scales through the multi-scale broadband noise suppression layer to obtain an intermediate fiber optic sensing signal, and the intermediate fiber optic sensing signal is subjected to cross-scale broadband noise suppression through the cross-scale broadband noise suppression layer to obtain an effective fiber optic sensing signal.

[0025] In this embodiment, the preprocessed original fiber optic sensing signal is input into a pre-trained multi-scale broadband noise suppression model. This model includes a multi-scale broadband noise suppression layer (parallel CNN network module) and a cross-scale broadband noise suppression layer (LSTM network module). The specific processing procedure is as follows: (1) Multi-scale broadband noise suppression layer processing (acquiring intermediate fiber sensing signals): The multi-scale broadband noise suppression layer consists of three parallel CNN subnetworks, corresponding to low, medium, and high-scale broadband noise processing, respectively: Low-scale CNN subnetwork: It adopts 7×7 convolutional kernels, with 3 convolutional layers (convolution stride of 1, padding method of same), 1 max pooling layer (pooling kernel of 2×2, stride of 2) and ReLU activation function to extract low-scale temperature drift noise features in the frequency range of 0-10Hz. The noise at this scale is initially filtered out through convolution operation and pooling processing, and the low-scale intermediate fiber optic sensing signal is output. Mesoscale CNN subnetwork: It adopts 5×5 convolutional kernel, with 3 convolutional layers, 1 max pooling layer and ReLU activation function. It is designed to enhance the feature extraction accuracy for mesoscale strata vibration noise in the frequency range of 10-100Hz (the effective signal is mainly distributed in this frequency band), and while initially suppressing noise, it retains the details of the effective signal and outputs the mesoscale intermediate fiber optic sensing signal. High-scale CNN subnetwork: It adopts 3×3 convolutional kernels, with 4 convolutional layers, 1 max pooling layer and ReLU activation function, to extract high-scale electromagnetic interference and equipment start-up and shutdown noise features in the frequency range of 100-1000Hz, accurately filter out high-frequency interference, and output high-scale intermediate fiber optic sensing signals.

[0026] (2) Cross-scale broadband noise suppression layer processing (acquiring effective fiber optic sensing signals): The cross-scale broadband noise suppression layer is a shared LSTM network with 256 hidden layer units, configured with two LSTM layers and one fully connected layer, and the dropout coefficient is set to 0.2 (to prevent overfitting). Low-scale, mesoscale, and high-scale intermediate fiber sensing signals are input into the LSTM network. This network uses a gating mechanism to memorize the temporal characteristics of broadband noise signals at different scales, learns the cross-scale coupling relationships between low-scale temperature drift and mesoscale vibration, and between mesoscale vibration and high-scale electromagnetic interference, further filters out residual cross-scale correlated noise, and outputs effective fiber sensing signals.

[0027] In this embodiment, a multi-scale broadband noise suppression layer is used to accurately capture the characteristics of broadband noise at different scales, achieving initial suppression of broadband noise at different scales and solving the adaptation problem of broadband noise at different scales. Subsequently, a cross-scale broadband noise suppression layer is used to accurately capture the cross-scale correlation characteristics between broadband noise at different scales (such as the coupling effect of slow low-scale temperature interference and mesoscale vibration noise), achieving synergistic suppression of cross-scale correlated broadband noise. This completes the comprehensive suppression from initial suppression at different scales to synergistic suppression across scales, thoroughly covering the full-scale interference scenarios of broadband noise, improving the broadband noise suppression coverage, increasing the accuracy of extracting effective fiber optic sensing signals, and solving the problem of CO2 storage monitoring effects being affected by severe distortion of fiber optic sensing signals.

[0028] Based on the above embodiments, this application provides another fiber optic sensing signal processing method, see below. Figure 2 As shown, the general flow of the fiber optic sensing signal processing method provided in this application embodiment is as follows: Step 201: Obtain a multi-scale broadband noise sample set; wherein, the multi-scale broadband noise sample set includes each historical fiber optic sensing signal with broadband noise at different scales, as well as the calibration intermediate fiber optic sensing signal and the calibration effective fiber optic sensing signal corresponding to each historical fiber optic sensing signal.

[0029] In this embodiment, historical fiber optic sensor signals from CO2 sequestration monitoring scenarios are collected to construct a multi-scale broadband noise sample set. This multi-scale broadband noise sample set includes a low-scale sub-sample set, a mid-scale sub-sample set, and a high-scale sub-sample set. The low-scale sub-sample set corresponds to temperature drift and slow geological deformation signals (frequency < 10 Hz), the mid-scale sub-sample set corresponds to CO2 injection pump start-up / shutdown and transport vehicle vibration noise and effective micro-seismic signals (frequency 10-100 Hz), and the high-scale sub-sample set corresponds to high-frequency instrument noise and micro-fracture detail signals (frequency > 100 Hz). The multi-scale broadband noise sample set contains no fewer than 10,000 samples. Each sample includes: historical fiber optic sensor signals (containing broadband noise and non-geological noise), calibrated intermediate fiber optic sensor signals (intermediate signals manually labeled and filtered for corresponding scale noise using professional signal processing software), and calibrated effective fiber optic sensor signals (noise-free effective signals verified in the laboratory). Next, the sample set was preprocessed: data augmentation was performed using time shifting, random pruning, and additive Gaussian noise (signal-to-noise ratio 10-20dB) to expand the sample size to 20,000; the sample set was divided into training set, validation set, and test set in a 7:2:1 ratio for model training, validation, and performance evaluation.

[0030] Step 202: Train the multi-scale broadband noise suppression model based on the multi-scale broadband noise sample set; wherein, the multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer.

[0031] In this embodiment of the application, when training the multi-scale broadband noise suppression model based on a multi-scale broadband noise sample set, each historical fiber sensing signal can be input into the multi-scale broadband noise suppression model. In the multi-scale broadband noise suppression model, the estimated intermediate fiber sensing signal is obtained by performing broadband noise suppression at different scales on each historical fiber sensing signal through a multi-scale broadband noise suppression layer, and the estimated effective fiber sensing signal is obtained by performing cross-scale broadband noise suppression on each estimated intermediate fiber sensing signal through a cross-scale broadband noise suppression layer.

[0032] In practical implementation, each historical fiber optic sensing signal in the training set can be input into the multi-scale broadband noise suppression model to be trained. In the multi-scale broadband noise suppression model, the three CNN sub-networks of the multi-scale broadband noise suppression layer process each historical fiber optic sensing signal and output the estimated low-scale intermediate fiber optic sensing signal, the estimated mesoscale intermediate fiber optic sensing signal, and the estimated high-scale intermediate fiber optic sensing signal, respectively. The LSTM network of the cross-scale broadband noise suppression layer processes the estimated low-scale intermediate fiber optic sensing signal, the estimated mesoscale intermediate fiber optic sensing signal, and the estimated high-scale intermediate fiber optic sensing signal and outputs the estimated effective fiber optic sensing signal.

[0033] Step 203: During the training of the multi-scale broadband noise suppression model, a combination loss function of multi-scale weighted loss function and cross-scale suppression loss function is used to update the network parameters of the multi-scale broadband noise suppression model.

[0034] In this embodiment, when updating the network parameters of the multi-scale broadband noise suppression model using a combined loss function of multi-scale weighted loss function and cross-scale suppression loss function, the total loss value can be calculated based on the estimated intermediate fiber sensing signal, the calibrated intermediate fiber sensing signal, the estimated effective fiber sensing signal, and the calibrated effective fiber sensing signal from each historical fiber sensing signal. The network parameters of the multi-scale broadband noise suppression model are then updated based on this total loss value. The estimated intermediate fiber sensing signal is the output signal of the multi-scale broadband noise suppression layer, and the estimated effective fiber sensing signal is the output signal of the cross-scale broadband noise suppression layer. The specific process is as follows: (1) Calculate the multi-scale weighted loss value: A multi-scale weighted loss function is employed, calculating the multi-scale weighted loss value based on the mean square error (MSE) between the estimated intermediate fiber sensing signal and the calibrated intermediate fiber sensing signal from various historical fiber sensing signals. In this multi-scale weighted loss function, weights are allocated according to the degree of influence of broadband noise at different scales on the effective fiber sensing signal. For example, the weight for the mesoscale (frequency band where the effective signal is located) is set to 0.5, the weight for the low-scale is set to 0.3, and the weight for the high-scale is set to 0.2 (the weights are determined through multiple experimental calibrations to ensure priority is given to mesoscale suppression accuracy). The formula for calculating the multi-scale weighted loss function is as follows:

[0035] in, This represents the multi-scale weighted loss value. , , The weights are assigned according to the degree of influence of broadband noise at different scales on the effective fiber optic sensing signal. To estimate the mean square error between the low-scale intermediate fiber sensing signal and the calibrated low-scale intermediate fiber sensing signal; To estimate the mean square error between the mesoscale intermediate fiber sensing signal and the calibrated mesoscale intermediate fiber sensing signal; To estimate the mean square error between the high-scale intermediate fiber sensing signal and the calibrated high-scale intermediate fiber sensing signal.

[0036] (2) Calculate the cross-scale suppression loss value: Using a cross-scale suppression loss function, the mean square error is calculated as the cross-scale suppression loss value based on the estimated effective fiber sensing signal and the calibrated effective fiber sensing signal from each historical fiber sensing signal. .

[0037] (3) Calculate the total loss and update the network parameters of the multi-scale broadband noise suppression model: Based on multi-scale weighted loss value and cross-scale suppression loss value Calculate the total loss value And based on the total loss value Update the network parameters of the multi-scale broadband noise suppression model. The formula for calculating the total loss is: ;in, This represents the multi-scale weighted loss value. This represents the cross-scale suppression loss value; and The weights are determined through validation set tuning, taking into account both multi-scale suppression and cross-scale correlation suppression. The Adam optimizer is used, with an initial learning rate of 0.001, which decays to 0.5 every 5 epochs. The goal is to minimize the total loss value, and all parameters (convolutional kernel weights, LSTM gating parameters, fully connected layer weights, etc.) of each CNN network and LSTM network are updated through backpropagation.

[0038] (4) Feature matching verification: Feature matching verification is performed on the effective fiber optic sensing signal output from the LSTM network: The time-domain features (peak value, mean, variance) and frequency-domain features (spectral peak value, characteristic frequency bandwidth) of the effective fiber optic sensing signal are extracted to construct a feature map. Similarity is calculated (using a cosine similarity algorithm) with a pre-calibrated feature map of effective CO2 storage monitoring signals (obtained under the same acquisition parameters and without noise interference). If the matching degree is greater than a set threshold (e.g., 95%), the signal is considered a qualified effective fiber optic sensing signal, and the multi-scale broadband noise suppression model training is complete. If the matching degree is less than or equal to the set threshold (e.g., 95%), the multi-scale broadband noise suppression model continues to be trained until the effective fiber optic sensing signal output from the LSTM network meets the matching requirements.

[0039] In this embodiment, by employing a combined loss function of multi-scale weighted loss function and cross-scale suppression loss function, the network parameters of the multi-scale broadband noise suppression model are updated. This enables the multi-scale broadband noise suppression model to achieve both deep suppression of broadband noise and maximum preservation of key details of effective fiber optic sensing signals (such as minute strain changes in CO2 storage monitoring and weak acoustic signals in oil and gas well logging). This resolves the contradiction between noise suppression and signal fidelity, providing high-precision effective fiber optic sensing signal support for subsequent CO2 storage monitoring and analysis.

[0040] Based on the above embodiments, this application provides an optical fiber sensing signal processing device, see below. Figure 3 As shown, the fiber optic sensing signal processing device 300 provided in this application embodiment includes at least: The signal acquisition unit 301 is used to acquire the original fiber optic sensing signal collected by the distributed fiber optic sensing system set in the CO2 storage monitoring area. The noise suppression unit 302 is used to input the original optical fiber sensing signal into a pre-trained multi-scale broadband noise suppression model for multi-scale broadband noise suppression to obtain an effective optical fiber sensing signal. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer. In the multi-scale broadband noise suppression model, the original optical fiber sensing signal is subjected to broadband noise suppression at different scales through the multi-scale broadband noise suppression layer to obtain an intermediate optical fiber sensing signal, and the intermediate optical fiber sensing signal is subjected to cross-scale broadband noise suppression through the cross-scale broadband noise suppression layer to obtain an effective optical fiber sensing signal.

[0041] In one possible implementation, the multi-scale broadband noise suppression layer includes multiple parallel CNN networks; the noise suppression unit 302 is used to perform broadband noise suppression at different scales on the original optical fiber sensing signal in parallel through multiple CNN networks to obtain an intermediate optical fiber sensing signal.

[0042] In one possible implementation, the cross-scale broadband noise suppression layer includes an LSTM network connected to multiple CNN networks; the noise suppression unit 302 is used to perform cross-scale broadband noise suppression on the intermediate fiber sensing signals output by the multiple CNN networks through the LSTM network to obtain an effective fiber sensing signal.

[0043] Based on the above embodiments, this application provides another fiber optic sensing signal processing device, see below. Figure 4 As shown, the fiber optic sensing signal processing device 400 provided in this application embodiment includes at least: The sample acquisition unit 401 is used to acquire a multi-scale broadband noise sample set; wherein, the multi-scale broadband noise sample set includes each historical fiber optic sensing signal with broadband noise at different scales, as well as the calibration intermediate fiber optic sensing signal and the calibration effective fiber optic sensing signal corresponding to each historical fiber optic sensing signal. The model training unit 402 is used to train a multi-scale broadband noise suppression model based on a multi-scale broadband noise sample set. During the training process of the multi-scale broadband noise suppression model, a combination loss function of multi-scale weighted loss and cross-scale suppression loss is used to update the network parameters of the multi-scale broadband noise suppression model. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer.

[0044] In one possible implementation, the model training unit 402 is used to calculate a multi-scale weighted loss value using a multi-scale weighted loss function based on the estimated intermediate fiber sensing signal and the calibrated intermediate fiber sensing signal from each historical fiber sensing signal; to calculate a cross-scale suppression loss value using a cross-scale suppression loss function based on the estimated effective fiber sensing signal and the calibrated effective fiber sensing signal from each historical fiber sensing signal; to calculate a total loss value based on the multi-scale weighted loss value and the cross-scale suppression loss value; and to update the network parameters of the multi-scale broadband noise suppression model based on the total loss value; wherein the estimated intermediate fiber sensing signal is the output signal of the multi-scale broadband noise suppression layer; and the estimated effective fiber sensing signal is the output signal of the cross-scale broadband noise suppression layer.

[0045] In one possible implementation, the multi-scale weighted loss function weights the suppression error of broadband noise at different scales according to its influence on the effective optical fiber sensing signal to obtain the multi-scale weighted loss value.

[0046] It should be noted that the principle of the fiber optic sensing signal processing device provided in this application embodiment to solve the technical problem is similar to that of the fiber optic sensing signal processing method provided in this application embodiment. Therefore, the implementation of the fiber optic sensing signal processing device provided in this application embodiment can refer to the implementation of the fiber optic sensing signal processing method provided in this application embodiment, and the repeated parts will not be described again.

[0047] After introducing the fiber optic sensing signal processing method and apparatus provided in the embodiments of this application, the electronic equipment provided in the embodiments of this application will be briefly introduced next.

[0048] See Figure 5 As shown, the electronic device 500 provided in this application embodiment includes at least a processor 501, a memory 502, and a computer program stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program, it implements the fiber optic sensing signal processing method provided in this application embodiment.

[0049] In one possible implementation, processor 501 can be a single processing element or a collective term for multiple processing elements. For example, processor 501 can be a central processing unit (CPU), or one or more integrated circuits configured to implement the fiber optic sensing signal processing method described in the embodiments of this application. Specifically, processor 501 can be a general-purpose processor, including but not limited to CPUs, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0050] In one possible implementation, memory 502 may include a readable medium in the form of volatile memory, such as random access memory (RAM) 5021 and / or cache memory 5022, and may further include read-only memory (ROM) 5023; memory 502 may also include a program tool 5025 having a set (at least one) of program modules 5024, including but not limited to: operating subsystem, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0051] In one possible implementation, the electronic device 500 provided in this application embodiment may further include a bus 503 connecting different components (including processor 501 and memory 502). The bus 503 represents one or more types of bus structures, including memory bus, peripheral bus, local area bus, etc.

[0052] In one possible implementation, the electronic device 500 can also communicate with one or more devices that enable a user to interact with the electronic device 500 (e.g., mobile phones, computers, etc.), and / or with external devices 504 such as devices that enable the electronic device 500 to communicate with one or more other electronic devices 500 (e.g., routers, modems, etc.). This communication can be performed via an input / output (I / O) interface 505. Furthermore, the electronic device 500 can also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 506. Figure 5As shown, network adapter 506 communicates with other modules of electronic device 500 via bus 503. It should be understood that, although... Figure 5 As not shown, other hardware and / or software modules may be used in conjunction with the electronic device 500, including but not limited to microcode, device drivers, redundant processors, external disk drive arrays, Redundant Arrays of Independent Disks (RAID) subsystems, tape drives, and data backup storage subsystems.

[0053] It should be noted that, Figure 5 The electronic device 500 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0054] Furthermore, this application also provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the fiber optic sensing signal processing method described above in this application. Specifically, the computer instructions may be built into or installed in a processor, enabling the processor to implement the fiber optic sensing signal processing method described above in this application by executing the built-in or installed computer instructions.

[0055] Of course, the fiber optic sensing signal processing method provided in the embodiments of this application can also be implemented as a program product, which includes program code. When the program code is executed by a processor, it implements the fiber optic sensing signal processing method provided in the embodiments of this application.

[0056] The program product provided in this application embodiment can be any combination of one or more readable media, wherein the readable media can be a readable signal medium or a readable storage medium, and the readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. Specifically, more specific examples of readable storage media (a non-exhaustive list) include: electrical connections with one or more wires, portable disks, hard disks, RAM, ROM, erasable programmable read-only memory (EPROM), optical fibers, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0057] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0058] Furthermore, although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0059] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0060] Obviously, those skilled in the art can make various modifications and variations to the embodiments of this application without departing from the spirit and scope of the embodiments of this application. Therefore, if these modifications and variations to the embodiments of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

Claims

1. A fiber optic sensing signal processing method, characterized in that, include: Acquire raw fiber optic sensing signals collected by a distributed fiber optic sensing system located in the CO2 storage monitoring area; The original fiber optic sensing signal is input into a pre-constructed multi-scale broadband noise suppression model for multi-scale broadband noise suppression to obtain an effective fiber optic sensing signal. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer. In the multi-scale broadband noise suppression model, the original fiber optic sensing signal is subjected to broadband noise suppression at different scales through the multi-scale broadband noise suppression layer to obtain an intermediate fiber optic sensing signal, and the intermediate fiber optic sensing signal is subjected to cross-scale broadband noise suppression through the cross-scale broadband noise suppression layer to obtain an effective fiber optic sensing signal.

2. The fiber optic sensing signal processing method as described in claim 1, characterized in that, The multi-scale broadband noise suppression layer comprises multiple parallel CNN networks; The intermediate fiber optic sensing signal is obtained by performing broadband noise suppression at different scales on the original fiber optic sensing signal through the multi-scale broadband noise suppression layer, including: The intermediate fiber optic sensing signal is obtained by performing broadband noise suppression at different scales on the original fiber optic sensing signal in parallel using multiple CNN networks.

3. The fiber optic sensing signal processing method as described in claim 2, characterized in that, The cross-scale broadband noise suppression layer includes an LSTM network, which is connected to multiple CNN networks. Effective fiber optic sensing signals are obtained by performing cross-scale broadband noise suppression on the intermediate fiber optic sensing signal through the cross-scale broadband noise suppression layer, including: The effective fiber optic sensing signal is obtained by performing cross-scale broadband noise suppression on the intermediate fiber optic sensing signals output by multiple CNN networks using the LSTM network.

4. A fiber optic sensing signal processing method, characterized in that, include: A multi-scale broadband noise sample set is obtained; wherein, the multi-scale broadband noise sample set includes each historical fiber optic sensing signal with broadband noise at different scales, as well as the calibration intermediate fiber optic sensing signal and the calibration effective fiber optic sensing signal corresponding to each historical fiber optic sensing signal. Based on the aforementioned multi-scale broadband noise sample set, a multi-scale broadband noise suppression model is trained; wherein, the multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer; During the training of the multi-scale broadband noise suppression model, a combined loss function of multi-scale weighted loss function and cross-scale suppression loss function is used to update the network parameters of the multi-scale broadband noise suppression model.

5. The fiber optic sensing signal processing method as described in claim 4, characterized in that, The network parameters of the multi-scale broadband noise suppression model are updated using a combined loss function of multi-scale weighted loss function and cross-scale suppression loss function, including: Based on the estimated intermediate fiber sensing signal and the calibrated intermediate fiber sensing signal of each historical fiber sensing signal, a multi-scale weighted loss value is calculated using a multi-scale weighted loss function; wherein, the estimated intermediate fiber sensing signal is the output signal of the multi-scale broadband noise suppression layer. Based on the estimated effective fiber sensing signal and the calibrated effective fiber sensing signal of each historical fiber sensing signal, the cross-scale suppression loss value is calculated using the cross-scale suppression loss function; wherein, the estimated effective fiber sensing signal is the output signal of the cross-scale broadband noise suppression layer. The total loss value is calculated based on the multi-scale weighted loss value and the cross-scale suppression loss value; Based on the total loss value, the network parameters of the multi-scale broadband noise suppression model are updated.

6. The fiber optic sensing signal processing method as described in claim 5, characterized in that, The multi-scale weighted loss function weights the suppression error of broadband noise at different scales according to its influence on the effective optical fiber sensing signal to obtain the multi-scale weighted loss value.

7. A fiber optic sensing signal processing device, characterized in that, include: The signal acquisition unit is used to acquire the raw fiber optic sensing signals collected by the distributed fiber optic sensing system set in the CO2 storage monitoring area. The noise suppression unit is used to input the original optical fiber sensing signal into a pre-trained multi-scale broadband noise suppression model for multi-scale broadband noise suppression to obtain an effective optical fiber sensing signal. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer. In the multi-scale broadband noise suppression model, the original optical fiber sensing signal is subjected to broadband noise suppression at different scales through the multi-scale broadband noise suppression layer to obtain an intermediate optical fiber sensing signal, and the intermediate optical fiber sensing signal is subjected to cross-scale broadband noise suppression through the cross-scale broadband noise suppression layer to obtain an effective optical fiber sensing signal.

8. A fiber optic sensing signal processing device, characterized in that, include: The sample acquisition unit is used to acquire a multi-scale broadband noise sample set; wherein, the multi-scale broadband noise sample set includes each historical fiber optic sensing signal with broadband noise at different scales, as well as the calibration intermediate fiber optic sensing signal and the calibration effective fiber optic sensing signal corresponding to each historical fiber optic sensing signal. The model training unit is used to train the multi-scale broadband noise suppression model based on the multi-scale broadband noise sample set. During the training process of the multi-scale broadband noise suppression model, a combination loss function of multi-scale weighted loss function and cross-scale suppression loss function is used to update the network parameters of the multi-scale broadband noise suppression model. The multi-scale broadband noise suppression model includes a multi-scale broadband noise suppression layer and a cross-scale broadband noise suppression layer.

9. An electronic device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the fiber optic sensing signal processing method as described in any one of claims 1-3, or implements the fiber optic sensing signal processing method as described in any one of claims 4-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which, when executed by a processor, implement the fiber optic sensing signal processing method as described in any one of claims 1-3, or the fiber optic sensing signal processing method as described in any one of claims 4-6.

Images