Earthquake fracture activity monitoring method and system

Abstract

The invention discloses an earthquake fracture activity monitoring method and system, and belongs to the field of earthquake monitoring, and the method comprises the steps: obtaining a continuous vibration signal of a complex geological region through a distributed optical fiber sensing technology; analyzing and extracting strain change characteristics by adopting a fiber bragg grating to obtain a weak precursor signal set; applying time-frequency analysis to determine slow slip event distribution; when a threshold value is exceeded, micro-seismic events are classified through a convolutional neural network, and an event positioning data set is generated; after space coordinates are extracted, grid space interpolation is adopted to generate a fracture activity distribution diagram; dynamically updating the distribution map through a real-time data stream processing framework; when the response is abnormal, utilizing the pre-embedded optical fiber network to supplement data to generate enhanced strain distribution; a hidden fault zone boundary is determined through spectrum analysis; and extracting high-risk coordinates to generate a real-time earthquake risk distribution map. According to the invention, complex environment full-coverage monitoring and weak signal second-level early warning are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of earthquake monitoring, and in particular relates to a method and system for monitoring earthquake fault activity. Background Art

[0002] Monitoring earthquake fault activity is a core area of ​​earthquake disaster prevention and urban safety. The key lies in accurately capturing crustal deformation and microvibration signals to provide early warning of potential earthquake risks. Current monitoring relies primarily on seismometer networks and GPS stations. These traditional methods have significant limitations: Fixed-point deployment, constrained by terrain and urban environments, makes it difficult to fully cover hidden fault zones or complex geological areas, resulting in sparse monitoring points and insufficient spatial resolution. For example, in urban underground areas or traffic tunnels, equipment deployment difficulties create monitoring blind spots, preventing the capture of subtle deformation or microvibration events within the fault zone, typically key precursor signals such as slow slip or microseismic events. Furthermore, high equipment procurement and maintenance costs restrict the expansion of monitoring networks, and long data collection cycles result in poor real-time performance, making it difficult to meet the second-level response requirements for short-term warnings. In specific application scenarios, these limitations have serious consequences: when potential fault activity occurs, traditional systems, due to insufficient spatial coverage, may miss localized, subtle strain changes (such as precursor signals occurring only within a few meters). Furthermore, low resolution and delayed response times further hinder the system's ability to quickly locate abnormal signals. For example, during a certain fault activity, subtle strain changes were not detected due to the large distance between monitoring points, thus missing the opportunity for early warning. The root of these problems lies in the difficulty of existing technologies in achieving high-density continuous spatial monitoring in complex environments, and the lack of the ability to efficiently process massive amounts of data to quickly identify weak signals. Therefore, the field of earthquake monitoring faces two core challenges: one is how to break through terrain and environmental limitations and achieve a dense monitoring network with full coverage in key areas such as hidden fault zones; the other is how to establish a real-time data processing mechanism to accurately separate and locate weak precursor signals from the noise background. Solving these challenges is of great significance to improving the reliability of earthquake early warning and reducing the loss of life and property. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention proposes a method and system for monitoring earthquake fault activity to solve the problems existing in the above-mentioned prior art.

[0004] In a first aspect, to achieve the above-mentioned objectives, the present invention provides a method for monitoring earthquake fault activity, comprising the following steps:

[0005] Obtain continuous vibration signals in complex geological areas through distributed fiber optic sensing technology to generate high-density vibration data streams;

[0006] Fiber Bragg grating analysis is used to extract strain variation characteristics from the vibration data stream to obtain a weak precursor signal set;

[0007] Applying a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event;

[0008] If the time frequency distribution of the slow slip event exceeds a preset threshold, the microseismic events are classified using a convolutional neural network to generate an event location dataset;

[0009] Extracting spatial coordinate information from the event location dataset and generating a high-resolution fault activity distribution map using a gridded spatial interpolation method;

[0010] Dynamically update the fault activity distribution map through a real-time data stream processing framework to obtain a second-level positioning response result;

[0011] If the second-level positioning response result shows an abnormal area, the pre-buried optical fiber network in the infrastructure is used to obtain supplementary vibration data to generate an enhanced strain distribution;

[0012] Applying a spectral analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone;

[0013] The coordinates of high-risk areas are extracted from the boundary range to generate a real-time earthquake risk distribution map.

[0014] In a second aspect, the present invention further provides a system for monitoring earthquake fault activity, for implementing a method for monitoring earthquake fault activity, the system comprising:

[0015] Signal acquisition module, used to acquire continuous vibration signals in complex geological areas through distributed fiber optic sensing technology and generate high-density vibration data streams;

[0016] a strain feature extraction module for extracting strain variation features from the vibration data stream using fiber Bragg grating analysis to obtain a weak precursor signal set;

[0017] a slow slip identification module, configured to apply a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event;

[0018] A microseismic classification and positioning module is configured to classify microseismic events and generate an event positioning dataset using a convolutional neural network when the time frequency distribution of the slow slip event exceeds a preset threshold;

[0019] A spatial modeling module is used to extract spatial coordinate information from the event location dataset and generate a high-resolution fault activity distribution map using a gridded spatial interpolation method;

[0020] A dynamic response module is used to dynamically update the fault activity distribution map through a real-time data stream processing framework to obtain a second-level positioning response result;

[0021] an abnormality enhancement module, configured to acquire supplementary vibration data and generate an enhanced strain distribution using a pre-buried optical fiber network when the second-level positioning response result indicates an abnormal area;

[0022] A boundary analysis module, configured to apply a spectral analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone;

[0023] The risk output module is used to extract the coordinates of high-risk areas from the boundary range and generate a real-time earthquake risk distribution map.

[0024] In a third aspect, the present invention further provides a computer terminal device, comprising:

[0025] one or more processors;

[0026] a memory, coupled to the processor, for storing one or more programs;

[0027] When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the method for monitoring earthquake fault activity in the first aspect.

[0028] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for monitoring earthquake fault activity in the first aspect.

[0029] In a fifth aspect, the present invention further provides a computer program product, comprising a computer program, characterized in that when the computer program is executed by a processor, the steps of the method for monitoring earthquake fault activity in the first aspect are implemented.

[0030] Compared with the prior art, the present invention has the following advantages and technical effects:

[0031] The present invention provides a method and system for monitoring earthquake fault activity. This method uses distributed fiber optic sensing to achieve high-density acquisition of continuous vibration signals in complex geological areas. It utilizes fiber Bragg grating analysis to extract strain variation characteristics and accurately capture weak precursor signals. It combines a time-frequency analysis algorithm to determine the distribution of slow slip events. It uses a convolutional neural network to classify and locate microseismic events when a threshold is exceeded. It generates a fault activity distribution map based on gridded spatial interpolation. It achieves dynamic updates in seconds through a real-time data stream processing framework. It triggers pre-buried fiber optic networks in abnormal areas to supplement data and generate enhanced strain distributions. It applies spectral analysis to determine the boundaries of hidden fault zones. Finally, it outputs a real-time earthquake risk distribution map. This method overcomes the spatial coverage limitations of traditional monitoring, improves weak signal recognition capabilities and response speed, and enables accurate early warning of fault activity throughout the entire process. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The accompanying drawings, which constitute part of the present invention, are provided to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are provided to explain the present invention and do not constitute an undue limitation of the present invention. In the accompanying drawings:

[0033] Figure 1 Flowchart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0034] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0035] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] Example 1

[0037] like Figure 1 As shown, this embodiment provides a method for monitoring earthquake fault activity, including:

[0038] Obtain continuous vibration signals in complex geological areas through distributed fiber optic sensing technology to generate high-density vibration data streams;

[0039] Fiber Bragg grating analysis is used to extract strain variation characteristics from the vibration data stream to obtain a weak precursor signal set;

[0040] Applying a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event;

[0041] If the time frequency distribution of the slow slip event exceeds a preset threshold, the microseismic events are classified using a convolutional neural network to generate an event location dataset;

[0042] Extracting spatial coordinate information from the event location dataset and generating a high-resolution fault activity distribution map using a gridded spatial interpolation method;

[0043] Dynamically update the fault activity distribution map through a real-time data stream processing framework to obtain a second-level positioning response result;

[0044] If the second-level positioning response result shows an abnormal area, the pre-buried optical fiber network in the infrastructure is used to obtain supplementary vibration data to generate an enhanced strain distribution;

[0045] Applying a spectral analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone;

[0046] extracting high-risk area coordinates from the boundary range to generate a real-time seismic risk distribution map.

[0047] As an embodiment in this embodiment, the process of acquiring continuous vibration signals through distributed optical fiber sensing technology includes:

[0048] The optical fiber is laid in the target area, and the time series data of the first vibration signal is collected by using the Rayleigh scattering effect.

[0049] The wavelet transform algorithm is used for signal denoising for the time series data of the first vibration signal to generate the second vibration signal.

[0050] The amplitude, frequency and phase characteristics are extracted from the high-quality time series data of the second vibration signal to obtain the first feature set.

[0051] Step S101, acquire continuous vibration signals in a complex geological area through distributed optical fiber sensing technology to generate high-density vibration data stream.

[0052] The continuous vibration signals in the complex geological area are acquired through distributed optical fiber sensing, the optical fiber is laid in the target area, the Rayleigh scattering effect is used to collect the first vibration signal to obtain the time series data of the first vibration signal. For the time series data of the first vibration signal, the wavelet transform algorithm is used for signal denoising to generate the second vibration signal, and the high-quality time series data of the second vibration signal is determined. From the high-quality time series data of the second vibration signal, the time series analysis method is used to extract the amplitude, frequency and phase characteristics to obtain the first feature set. If the amplitude value in the first feature set is greater than a predetermined threshold, the first feature set is analyzed by clustering analysis to generate a vibration feature distribution map of the complex geological area, and the vibration mode is determined.

[0053] Specifically, the continuous vibration signals in the complex geological area are acquired through distributed optical fiber sensing, and the core is to realize high-precision signal acquisition by using the Rayleigh scattering effect of the optical fiber.

[0054] For example, in a single-mode optical fiber laid along an oil and gas pipeline, a laser emits pulsed light, and the backscattered light generated by Rayleigh scattering carries the vibration information of the stratum near the pipeline. The time series data of the scattered light is recorded by the collection device to form the time series of the first vibration signal, which usually contains amplitude, frequency and other information, and the time resolution can reach milliseconds and the spatial resolution can reach meters. This technology can monitor vibration events within a range of hundreds of kilometers in real time, and is suitable for seismic monitoring or analysis of underground structure changes. For the time series data of the first vibration signal, the wavelet transform algorithm is used for denoising to generate the second vibration signal.

[0055] In a possible implementation, the wavelet transform filters out high-frequency noise by decomposing the signal into different frequency subbands, and retains low-frequency signals related to geological vibration.

[0056] For example, in a certain oilfield monitoring scenario, the original signal may be disturbed by noise such as wind blowing and mechanical vibration. The wavelet transform can concentrate the noise energy in the high-frequency subband, and generate a high-quality second vibration signal by setting a threshold for rejection. The signal-to-noise ratio of the denoised signal is improved by about 20%, significantly improving the accuracy of subsequent analysis. From the high-quality time series data of the second vibration signal, the amplitude, frequency and phase characteristics are extracted by using the time series analysis method to generate the first feature set.

[0057] Specifically, the frequency distribution of the signal can be extracted by fast Fourier transform, and the amplitude and phase are calculated by combining time domain analysis.

[0058] For example, when monitoring the expansion of underground fractures, the extracted amplitude may reflect the intensity of fracture expansion, the frequency distribution indicates the periodicity of vibration, and the phase change may be related to the direction of geological stress change. The feature set contains information such as amplitude peak 0.1-0.5mV and frequency range 0.1-10Hz, providing a basis for geological feature analysis. If the amplitude value in the first feature set is greater than a preset threshold, such as 0.3mV, a vibration feature distribution map is generated by cluster analysis.

[0059] In an embodiment, the K-means clustering algorithm is used to group the feature set by amplitude and frequency, and identify the vibration patterns of different geological regions.

[0060] For example, in a certain shale gas extraction area, cluster analysis finds that high-amplitude high-frequency signals are concentrated in fracture-dense areas, and low-amplitude low-frequency signals appear in stable strata, generating a distribution map to intuitively display the spatial heterogeneity of vibration patterns.

[0061] As an embodiment in the present embodiment, the process of analyzing and extracting strain change characteristics from the vibration data stream by using the fiber Bragg grating includes:

[0062] The weak signals related to strain change are classified and extracted by using the support vector machine algorithm.

[0063] Specifically, in step S102, the fiber Bragg grating is used to analyze and extract strain change characteristics from the vibration data stream to obtain a set of weak precursor signals.

[0064] The weak signals related to strain change are extracted from the vibration data stream by using the fiber Bragg grating to obtain the set of weak precursor signals.

[0065] Specifically, when the original vibration data stream is acquired by using the fiber Bragg grating, the strain measurement is realized by using the wavelength selective reflection of the grating periodic refractive index change on light.

[0066] For example, when a fiber Bragg grating (FBG) is deployed in a geological area, the period of the grating changes when the ground surface undergoes slight deformation, causing the wavelength of the reflected light to shift. This wavelength shift is directly related to the strain and can be accurately captured by a spectrometer, forming a raw vibration data stream.

[0067] In one possible implementation, optical fiber is laid along a geological fault zone, with a grating spacing of 1 meter and a wavelength offset resolution of 0.01 nanometer, sufficient to detect microscopic strain changes. This method is suitable for monitoring subtle formation deformations, such as stress adjustments caused by changes in groundwater levels. When using optical time-domain reflectometry to collect the first vibration signal, a laser pulse propagates along the fiber, reflects off the grating, and then returns. The time delay of the reflected signal reflects the grating position, and the intensity change reflects the magnitude of the strain.

[0068] For example, in monitoring ground subsidence, an optical time-domain reflectometer (OTDR) emits laser light in 10-nanosecond pulses. The collected time series data includes changes in reflected light intensity caused by strain, with a temporal resolution of microseconds and a spatial resolution of approximately 0.5 meters. This high-resolution data can precisely locate the location of strain, providing a reliable foundation for subsequent analysis. A Fourier transform algorithm decomposes the time series data of the first vibration signal into frequency components, filters out high-frequency noise, and generates a second vibration signal.

[0069] For example, in certain geological areas, the original signal may be mixed with high-frequency interference from surface vehicle vibrations. Fourier transform decomposes the signal, retaining low-frequency geological signals of 0.1-5 Hz while removing noise above 20 Hz. This denoising method effectively improves signal quality, generating a second vibration signal that clearly reflects formation strain and significantly improving the signal-to-noise ratio. From the high-quality time series data of the second vibration signal, a short-time Fourier transform extracts amplitude, frequency, and phase features through time-frequency analysis to form the first feature set.

[0070] It should be noted that this method can capture short-lived vibration events and is suitable for analyzing transient geological changes. If the amplitude feature in the first feature set is greater than 0.2 millivolts, the support vector machine algorithm is used to classify the feature set and extract weak precursor signals.

[0071] In one possible implementation, the algorithm uses historical strain data during training to optimize classification boundaries and improve the detection accuracy of weak signals. This approach can effectively separate weak precursor signals and provide data support for geological anomaly analysis.

[0072] Step S103 : applying a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event.

[0073] The time-frequency distribution of the slow slip event is determined by calculating the time-frequency distribution of the weak signal set using continuous wavelet transform.

[0074] In one embodiment, if the amplitude feature in the first feature set is greater than a preset threshold of 0.3 millivolts, a classification analysis is performed using a random forest algorithm. The random forest algorithm constructs multiple decision trees to comprehensively determine whether the signal in the feature set is related to a slow slip event.

[0075] For example, when monitoring a fault zone, the algorithm, trained on historical data, takes as input amplitude and frequency characteristics and outputs a signal classification result, identifying a collection of weak signals with an amplitude of 0.4 millivolts and a frequency of 3 hertz. These signals may indicate the occurrence of slow slip in the formation.

[0076] For example, for weak signal sets, the continuous wavelet transform is used to calculate the time-frequency distribution. The continuous wavelet transform accurately captures the time-frequency characteristics of the signal by adjusting the wavelet scale.

[0077] In one possible implementation, the Morlet wavelet basis function is used to analyze a time window of 0.5 seconds to obtain a frequency distribution diagram of 0.5-8 Hz, reflecting the time evolution characteristics of the slow slip event.

[0078] It should be noted that this method can reveal the dynamic changes of the signal and provide a basis for judging the time of occurrence and duration of the event.

[0079] In one embodiment, monitoring of slow slip events may be combined with geological context analysis.

[0080] For example, in a certain active fault region, continuous wavelet transform results show that a 2 Hz signal lasts approximately 10 seconds, potentially indicating a slow release of formation stress. Multi-dimensional feature analysis significantly improves signal reliability and event location accuracy. This multi-layered analysis approach, from denoising to feature extraction to classification and time-frequency analysis, forms a complete technical chain, effectively supporting the precise monitoring of slow slip events.

[0081] As an implementation in this embodiment, the process of classifying microseismic events using a convolutional neural network includes:

[0082] Perform classification analysis on the feature set through convolutional neural network;

[0083] According to the classification results, the grid search algorithm is used to locate and calculate the time-frequency distribution of the microseismic event set.

[0084] Specifically, in step S104, if the time frequency distribution of the slow slip event exceeds a preset threshold, the microseismic events are classified by a convolutional neural network to generate an event location dataset.

[0085] If the time-frequency distribution of slow slip events exceeds a preset threshold, a convolutional neural network is used to perform classification analysis to obtain a microseismic event set. Based on this microseismic event set, a grid search algorithm is used to calculate the location of the time-frequency distribution of the microseismic event set to obtain an event location dataset.

[0086] For example, within a target geological area, a high-sensitivity seismometer array is used to acquire real-time seismic signal streams to monitor micro-vibrations in the ground. The seismic signal stream contains vibration information from multiple sources, such as changes in ground stress, microseismic events, and background noise.

[0087] It's important to note that seismograph arrays are typically deployed along geological fault zones, with sampling frequencies up to 1000 Hz and a spatial resolution of approximately 10 meters, making them suitable for capturing signal changes caused by microseismic events. Wavelet transforms are used for denoising, decomposing the signal into frequency subbands, retaining low-frequency signals associated with microseismic events while filtering out high-frequency interference.

[0088] For example, in a fault zone monitoring scenario, where the original signal stream is contaminated by surface traffic noise, the wavelet transform uses the Symlet wavelet basis function, decomposing it to four layers, retaining the 0.2-15 Hz signal components to generate the first denoised signal. This denoising method effectively separates microseismic signals and reduces the impact of noise.

[0089] In one possible implementation, the first denoised signal is subjected to a short-time Fourier transform (SFT) to extract amplitude and phase features. The SFT analyzes the signal using a sliding time window to generate a time-frequency distribution that reflects the dynamic characteristics of the signal.

[0090] For example, with a time window set to 0.3 seconds, the analysis yielded a feature set with an amplitude range of 0.05-0.4 millivolts and a phase variation range of -π / 4 to π / 4. Amplitude features indicate the intensity of microseismic events, while phase features reflect the temporal changes in signal propagation, helping to identify the triggering moment of microseismic events.

[0091] For example, if the amplitude feature in the first feature set exceeds a preset threshold of 0.2 millivolts, a convolutional neural network is used for classification analysis. Through multiple layers of convolution and pooling operations, the convolutional neural network extracts deep patterns in the feature set and determines whether the signal is a microseismic event.

[0092] In one possible implementation, a grid search algorithm is used to calculate the time-frequency distribution of a microseismic event set. Grid search divides the time-frequency grid into segments, analyzes the spatial and temporal distribution of the signal, and determines the event location.

[0093] As an implementation manner in this embodiment, the process of generating a distribution map using the gridded spatial interpolation method includes:

[0094] Use spatial grid division algorithm to divide the spatial coordinate information into regions;

[0095] For the divided area set, the Kriging interpolation algorithm is used to perform grid interpolation calculation;

[0096] If the spatial density characteristics of the interpolated dataset exceed the preset threshold, it is smoothed using a two-dimensional convolution algorithm;

[0097] Based on the smoothed data set, a contour drawing algorithm is used to generate a high-resolution fault activity distribution map.

[0098] Specifically, in step S105 , spatial coordinate information is extracted from the event location dataset, and a gridded spatial interpolation method is used to generate a high-resolution fault activity distribution map.

[0099] Extract spatial coordinate information from the event location dataset, and use a spatial grid partitioning algorithm to divide the spatial coordinate information into regions to obtain a first partitioned region set. For the first partitioned region set, use a Kriging interpolation algorithm to perform gridded interpolation calculation on the spatial coordinate information, where the Kriging interpolation formula is: Where Z(s) is the interpolation result of the predicted point s, λ i is the weight coefficient, Z(s i ) is a known point s i The first interpolated dataset is obtained by calculating the observed values. If the spatial density characteristics of the first interpolated dataset exceed a preset threshold, the first interpolated dataset is smoothed using a two-dimensional convolution algorithm to obtain a second interpolated dataset. Based on the second interpolated dataset, a high-resolution fault activity distribution map is generated using a contour drawing algorithm to obtain a first distribution image.

[0100] For example, in the field of earthquake monitoring, extracting spatial coordinate information from event location datasets is a key step in analyzing the spatial distribution of microseismic events. Spatial coordinate information typically contains the three-dimensional coordinates of the microseismic event, such as longitude, latitude, and depth. By parsing the location dataset, the precise location of each event can be obtained.

[0101] For example, in a fault zone monitoring scenario, a positioning dataset recorded the coordinates of 1,000 microseismic events, covering an area of ​​10 km x 10 km x 5 km, with a coordinate accuracy of 0.1 meter. This high-precision coordinate system provides a reliable basis for subsequent regional delineation.

[0102] In one possible implementation, a spatial gridding algorithm divides coordinate information into regular grid cells to analyze the regional characteristics of microseismic events. Gridding can be based on a fixed grid size, for example, dividing the monitoring area into three-dimensional grid cells of 1 km x 1 km x 0.5 km. The number of events within each grid cell is counted to form a first set of partitioned regions.

[0103] For example, in the fault zone described above, after partitioning into 1,000 grid cells, some grid cells have significantly higher event counts than others, indicating areas of concentrated fault activity. This partitioning method facilitates the identification of hotspots of geological activity.

[0104] For example, for the first set of partitioned regions, the Kriging interpolation algorithm is used to generate a continuous spatial distribution model. Kriging interpolation uses observed values ​​at known points to predict data at unknown points, making it particularly well-suited for unevenly distributed geological data. Assuming the event density within a grid cell is the observed value, Kriging interpolation can predict the density value in the surrounding unobserved areas.

[0105] For example, in a certain fault zone, the event density at observation points ranges from 0.1 to 5 events per cubic kilometer. Interpolation generates a smoothed density distribution dataset. This method can effectively fill in data gaps and improve the accuracy of distribution analysis.

[0106] In one possible implementation, if the spatial density feature of the first interpolated dataset exceeds a preset threshold, such as 2 times / cubic kilometer, a two-dimensional convolution algorithm is used for smoothing. Two-dimensional convolution uses a convolution kernel to perform a weighted average on the data, reducing the impact of local noise.

[0107] For example, a 3×3 convolution kernel is used to smooth the density distribution and generate a second interpolated dataset. This smoothing process can eliminate abnormal peaks in the data and make the distribution characteristics more consistent with geological laws.

[0108] For example, based on the second interpolated dataset, a contouring algorithm generates a high-resolution fault activity distribution map. The contouring algorithm generates continuous density contours from the interpolated data, reflecting the activity intensity of the fault zone.

[0109] For example, setting the contour interval to 0.5 times per cubic kilometer produces a first distribution image that clearly shows high-density areas concentrated near the main slip surface of the fault. This image provides geologists with an intuitive reference for the distribution of fault activity, helping to identify areas of potential earthquake risk.

[0110] In a possible implementation, the contour drawing algorithm can also be combined with color coding to enhance the visualization effect.

[0111] For example, a red-yellow-blue gradient is used to represent areas with a density range from high to low. Red areas correspond to densities greater than 3 times per cubic kilometer, indicating areas of high activity. This visualization method facilitates quick identification of fault activity trends and supports further geological analysis.

[0112] As an implementation in this embodiment, the process of dynamically updating through the real-time data stream processing framework includes:

[0113] Use streaming algorithms to segment time series data;

[0114] For segmented data sets, a grid partitioning algorithm is used to divide the spatial coordinates into regions;

[0115] If the spatial density of the region exceeds the preset threshold, it is smoothed using the mean filter algorithm;

[0116] Based on the smoothed data set, the support vector machine algorithm is used to classify and locate the fault activity;

[0117] A dynamically updated fault activity distribution map is generated through a contour drawing algorithm.

[0118] Specifically, in step S106, the fault activity distribution map is dynamically updated through a real-time data stream processing framework to obtain a second-level positioning response result.

[0119] Time series data is acquired from a real-time data stream and segmented using a streaming algorithm to obtain a first segmented dataset. A gridding algorithm is used to segment the spatial coordinates of the first segmented dataset into regions. If the spatial density of the region exceeds a preset threshold, the region is smoothed using a mean filtering algorithm to obtain a second segmented dataset. Based on the second segmented dataset, a support vector machine algorithm is used to classify the fault activity and determine its location coordinates, obtaining a first location result set. A contour drawing algorithm is used to render the first location result set, generating a dynamically updated fault activity distribution map and obtaining a first distribution image.

[0120] For example, in the field of earthquake monitoring, obtaining time series data from real-time data streams is fundamental to analyzing the dynamic distribution of fault activity. Time series data typically contains the timestamps of microseismic events and associated spatial coordinates, such as longitude, latitude, and depth. Real-time data streams may originate from a network of sensors at earthquake monitoring stations, with data updated at a frequency of seconds or milliseconds.

[0121] For example, a certain fault zone monitoring system collects 1000 microseismic event coordinates and time information per second, covering an area of 20 km x 20 km x 10 km, with a coordinate accuracy of 0.2 meters. Such high-frequency and high-precision data provides reliable input for subsequent segmentation processing.

[0122] In one possible implementation, the streaming algorithm processes the time series data for segmentation to generate a first segmented data set. Streaming divides the data by time window, such as a 10-minute time window, and extracts microseismic event data within each window. Assuming that 5000 events are recorded in a certain 10-minute window, the algorithm divides the data into consecutive subsets based on the timestamp, with each subset containing time and spatial information. This segmentation method facilitates capturing short-term trends in fault activity.

[0123] For example, for the first segmented data set, the grid division algorithm divides the spatial coordinates into regular regions. Grid division can be based on fixed dimensions, such as dividing the monitoring area into 2 km x 2 km x 1 km three-dimensional grid cells. Each grid counts the number of microseismic events within it, forming a regional density distribution. Assuming that the number of events in a certain grid is 50, which is much higher than the average of 10, it indicates that this area may be a hotspot of fault activity. This regional division method facilitates subsequent analysis of the characteristics of high-density regions.

[0124] In one possible implementation, if the spatial density of a certain grid region exceeds a preset threshold, such as 30 times per cubic kilometer, a mean filter algorithm is used for smoothing. Mean filtering generates a second segmented data set by calculating the average density of the grid and its neighboring regions.

[0125] In one possible implementation, the contour drawing algorithm renders the first positioning result set to generate a dynamically updated fault activity distribution map. The algorithm draws contours based on the positioning coordinates and density values, such as setting an interval of 5 times per cubic kilometer, and generates a first distribution image that dynamically displays the changes in high-density regions over time.

[0126] Preferably, the image can be combined with color coding, such as using red to represent high-activity areas with a density greater than 40 times per cubic kilometer. This dynamic visualization method facilitates real-time monitoring of fault activity trends and provides intuitive references for earthquake risk assessment.

[0127] It can be understood that each step of the above method is closely related to the characteristics of real-time data, from time series segmentation to spatial grid division, to classification and visualization, forming a complete analysis chain. This method can effectively process high-frequency data and quickly generate dynamic distribution maps, providing real-time support for geological monitoring.

[0128] Step S107: If the second-level positioning response result shows an abnormal area, the pre-buried optical fiber network in the infrastructure is used to obtain supplementary vibration data to generate an enhanced strain distribution.

[0129] Vibration data is acquired from a fiber optic network and segmented using a time series segmentation algorithm to generate a first vibration dataset. A gridding algorithm is used to segment the spatial coordinates of the first vibration dataset into regions. If the vibration intensity of a region exceeds a preset threshold, the region is smoothed using a mean filtering algorithm to generate a second vibration dataset. Based on the second vibration dataset, a data fusion algorithm is used to integrate the vibration data with time series data in a real-time data stream to generate a first strain distribution dataset. A contour drawing algorithm is used to render the spatial coordinates of the abnormal region of the first strain distribution dataset, generating a dynamically updated enhanced strain distribution map and obtaining a first enhanced distribution image.

[0130] For example, in the field of earthquake monitoring, acquiring vibration data through fiber optic networks is a crucial tool for analyzing underground strain distribution. Utilizing distributed fiber optic sensing technology, fiber optic networks can capture minute vibration signals from the Earth's crust, generating high-frequency time series data. A time series segmentation algorithm then segments this data into segments, forming the first vibration dataset.

[0131] For example, the monitoring area can be divided into a 1 km x 1 km two-dimensional grid, and the average vibration intensity of each grid is calculated. For example, if the vibration intensity in a certain grid area is 0.5 milligal, exceeding the preset threshold of 0.3 milligal, it indicates possible signs of geological activity in that area. Grid division facilitates the identification of high-intensity areas, providing a basis for further processing.

[0132] In a possible implementation, a mean filter algorithm is used to perform smoothing processing on the region where the vibration intensity exceeds a threshold value to generate a second vibration data set.

[0133] Specifically, based on the second vibration data set, the data fusion algorithm integrates the vibration data with the time series data in the real-time data stream to generate a first strain distribution data set.

[0134] For example, a fusion algorithm combines vibration intensity data from a fiber optic network with time-series data from a seismic monitoring station to calculate strain distribution. If the vibration intensity and time-series data for a particular area indicate concentrated strain, the algorithm outputs the strain value for that area, such as 0.002 microstrain, to generate a strain distribution dataset. This fusion approach improves the comprehensiveness and reliability of the data.

[0135] As you can see, the above method forms a complete analysis chain, from vibration data acquisition to image rendering. Each step closely integrates the characteristics of optical fiber network data. Through segmentation, meshing, smoothing, data fusion, and visualization, efficient strain distribution analysis is achieved. This method fully utilizes the advantages of high-frequency data and provides reliable support for seismic monitoring.

[0136] Step S108 : applying a spectrum analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone.

[0137] Vibration data is acquired from the optical fiber network. A fast Fourier transform algorithm is used to perform spectral analysis on the time series of the vibration data, extracting frequency features to obtain a first frequency feature set. A gridding algorithm is used to divide the spatial coordinates into regions for the first frequency feature set. If the amplitude of the frequency feature exceeds a preset threshold, it is marked as an abnormal region, thereby obtaining a first abnormal region set. Based on the first abnormal region set, a data fusion algorithm is used to integrate the frequency features with the strain distribution data to generate a first strain boundary dataset containing the boundary range. A contour drawing algorithm is used to render the spatial coordinates of the abnormal region for the first strain boundary dataset, generating an image of the boundary range of the hidden fault zone and obtaining a first boundary distribution image.

[0138] For example, in earthquake monitoring, using fiber optic networks to acquire vibration data and perform spectral analysis is an important means of identifying underground geological activity. Fast Fourier transform algorithms convert time series data into the frequency domain to extract the frequency characteristics of vibrations.

[0139] In a possible implementation manner, a grid partitioning algorithm performs region partitioning on the spatial coordinates of the first frequency feature set.

[0140] For example, the monitoring area is divided into 0.5 km x 0.5 km grids, and the mean amplitude of the frequency signature is calculated for each grid. For example, if the mean amplitude of a particular grid is 0.7 milligal, exceeding the preset threshold of 0.4 milligal, it is marked as an anomaly, generating the first set of anomaly areas. This division method facilitates the identification of areas of potential geological activity.

[0141] It should be noted that the threshold setting is based on historical data and geological characteristics to ensure that the marked abnormal areas are of practical significance.

[0142] For example, for the first abnormal region set, the data fusion algorithm integrates the frequency characteristics with the strain distribution data to generate a first strain boundary data set.

[0143] Specifically, the fusion algorithm combines the frequency characteristics of the optical fiber network with strain data from seismic monitoring stations to calculate the strain boundaries of the anomaly area. For example, if the dominant frequency of an anomaly area is 10 Hz, corresponding to a strain value of 0.003 microstrain, the algorithm outputs the boundary of the area, such as a 1 km x 1 km rectangle. This fusion method improves the accuracy of the boundary data and provides a reliable foundation for subsequent analysis.

[0144] In a possible implementation, a contour drawing algorithm performs image rendering on the first strain boundary data set to generate a boundary range image of the hidden fault zone.

[0145] As you can see, the above method forms a complete analysis process through frequency feature extraction, gridding, data fusion, and image rendering. Each step fully leverages the characteristics of fiber optic network data, transforming vibration data into an image of the hidden fault zone boundary. This method effectively supports the identification of abnormal areas and the determination of their boundaries during seismic monitoring.

[0146] Step S109: extracting the coordinates of high-risk areas from the boundary range data to generate a real-time earthquake risk distribution map.

[0147] The time series of the spatial coordinates and vibration data are obtained from the boundary range data, and the spatial coordinates are divided into regions using a grid division algorithm. If the frequency characteristic amplitude of the vibration data exceeds a preset threshold, it is marked as a high-risk area to obtain a first high-risk area set. For the first high-risk area set, a data fusion algorithm is used to integrate the time series of the vibration data and the boundary range data to generate a first risk distribution data set containing risk levels. Based on the first risk distribution data set, the K-means clustering algorithm is used to classify the spatial coordinates of the high-risk area, and the risk level of each coordinate is determined to obtain a first risk coordinate set. For the first risk coordinate set, a contour drawing algorithm is used to render the spatial coordinates to generate a real-time earthquake risk distribution map to obtain a first distribution image.

[0148] For example, in the field of earthquake monitoring, when using vibration data and spatial coordinates obtained from fiber optic networks to identify high-risk areas, grid division algorithm is one of the key steps.

[0149] Specifically, the gridding algorithm divides the monitoring area into specific spatial scales to facilitate analysis of the spatial distribution of vibration data. Assuming the monitoring area is 4 km x 4 km, the gridding algorithm divides it into 0.4 km x 0.4 km grids, with each grid containing vibration data corresponding to a specific spatial coordinate. Vibration data is analyzed spectrally to obtain frequency characteristics. Assuming the dominant frequency of vibration data within a particular grid is 12 Hz, the amplitude is 0.8 mGal, and the preset threshold is 0.5 mGal. Because the amplitude of this grid exceeds the threshold, the algorithm marks it as a high-risk area, forming the first high-risk area set. This method effectively locates potentially seismically active areas through spatial gridding.

[0150] In one possible implementation, a data fusion algorithm integrates the vibration data time series of the first high-risk area set with the boundary range data to generate a first risk distribution data set.

[0151] For example, a fusion algorithm combines time series data from fiber optic networks with geological boundary data to calculate the risk level of high-risk areas. For example, if the time series of vibration data for a high-risk area shows persistent high-frequency vibration, combined with the boundary data, the algorithm assesses the risk level as high and generates a spatially distributed dataset containing the risk level. This fusion approach improves the comprehensiveness of risk assessment by integrating data from multiple sources.

[0152] In one possible implementation, a contour drawing algorithm performs image rendering on the first risk coordinate set to generate a real-time earthquake risk distribution map.

[0153] For example, the algorithm draws contour lines based on risk level, with intervals set to 0.2 risk units. The risk distribution is color-coded, with red indicating high-risk areas and green indicating low-risk areas. For example, a map of a given area shows a high-risk zone running northwest, approximately 0.5 kilometers wide. This visualization method intuitively presents the spatial distribution of risk, enabling monitoring personnel to quickly identify areas of potential earthquake threat.

[0154] It's important to note that the above process, from meshing to image rendering, forms a complete risk analysis system. Each step fully leverages the spatiotemporal characteristics of fiber optic network data, transforming vibration data into risk distribution maps. This approach, through multi-level analysis and visualization, provides efficient support for earthquake monitoring.

[0155] Example 2

[0156] In this embodiment, a computer terminal device is provided, including:

[0157] one or more processors;

[0158] a memory, coupled to the processor, for storing one or more programs;

[0159] When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the above-mentioned method for monitoring earthquake fault activity.

[0160] In this embodiment, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for monitoring earthquake fault activity are implemented.

[0161] In this embodiment, an electronic device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps of the above-mentioned method for monitoring earthquake fault activity.

[0162] In this embodiment, a computer program product is also provided, including a computer program, which implements the steps of the above-mentioned method for monitoring earthquake fault activity when executed by a processor.

[0163] The above program can be run in the processor, or it can be stored in the memory (or computer-readable medium), which includes permanent and non-permanent, removable and non-removable media and can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0164] These computer programs can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps of the functions specified in one or more blocks can be implemented by different modules corresponding to different steps.

[0165] This embodiment provides such a device or system. The system is called a monitoring system for earthquake fault activity, and includes:

[0166] Signal acquisition module, used to acquire continuous vibration signals in complex geological areas through distributed fiber optic sensing technology and generate high-density vibration data streams;

[0167] a strain feature extraction module for extracting strain variation features from the vibration data stream using fiber Bragg grating analysis to obtain a weak precursor signal set;

[0168] a slow slip identification module, configured to apply a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event;

[0169] A microseismic classification and positioning module is configured to classify microseismic events and generate an event positioning dataset using a convolutional neural network when the time frequency distribution of the slow slip event exceeds a preset threshold;

[0170] A spatial modeling module is used to extract spatial coordinate information from the event location dataset and generate a high-resolution fault activity distribution map using a gridded spatial interpolation method;

[0171] A dynamic response module is used to dynamically update the fault activity distribution map through a real-time data stream processing framework to obtain a second-level positioning response result;

[0172] an abnormality enhancement module, configured to acquire supplementary vibration data and generate an enhanced strain distribution using a pre-buried optical fiber network when the second-level positioning response result indicates an abnormal area;

[0173] A boundary analysis module, configured to apply a spectral analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone;

[0174] The risk output module is used to extract the coordinates of high-risk areas from the boundary range and generate a real-time earthquake risk distribution map.

[0175] As an implementation method of this embodiment, the signal acquisition module includes:

[0176] An optical fiber sensing unit is configured to utilize an optical fiber deployed in a target area to collect time series data of a first vibration signal using a Rayleigh scattering effect;

[0177] a signal denoising unit, configured to perform signal denoising on the time series data of the first vibration signal using a wavelet transform algorithm and generate a second vibration signal;

[0178] The feature extraction unit is used to extract amplitude, frequency and phase features from the high-quality time series data of the second vibration signal to obtain a first feature set.

[0179] As an implementation in this embodiment, the strain feature extraction module includes:

[0180] an optical time domain reflectometer unit, configured to collect a first vibration signal based on a strain effect by using an optical time domain reflectometer technique;

[0181] a frequency domain denoising unit, configured to perform signal denoising on the time series data of the first vibration signal using a Fourier transform algorithm;

[0182] The time-frequency feature unit is used to extract amplitude features, frequency features, and phase features from the denoised high-quality time series data using short-time Fourier transform;

[0183] The weak signal classification unit is used to classify and extract weak signals related to strain changes through a support vector machine algorithm when the amplitude feature is greater than a preset threshold.

[0184] As an implementation method of this embodiment, the microseismic classification and positioning module includes:

[0185] A seismic signal denoising unit, used to denoise the real-time seismic signal stream using wavelet transform;

[0186] A phase analysis unit, configured to extract amplitude and phase features of the denoised signal using short-time Fourier transform;

[0187] A convolution classification unit, configured to perform classification analysis on a feature set using a convolutional neural network when the amplitude feature exceeds a preset threshold;

[0188] The grid positioning unit is used to perform positioning calculation on the time frequency distribution of the microseismic event set using a grid search algorithm based on the classification results.

[0189] As an implementation method of this embodiment, the space modeling module includes:

[0190] A grid division unit, used for dividing the spatial coordinate information into regions using a spatial grid division algorithm;

[0191] Kriging interpolation unit, used to perform grid interpolation calculations using the Kriging interpolation algorithm for the divided area set;

[0192] The density smoothing unit is used to perform smoothing using a two-dimensional convolution algorithm when the spatial density feature of the interpolated data set exceeds a preset threshold;

[0193] The distribution map generating unit is used to generate a high-resolution fault activity distribution map based on the smoothed data set by using a contour drawing algorithm.

[0194] As an implementation method of this embodiment, the dynamic response module includes:

[0195] Stream segmentation unit, used to segment time series data using streaming processing algorithms;

[0196] Density partitioning unit, used to divide the spatial coordinates into regions using a grid partitioning algorithm for segmented data sets;

[0197] A filtering and smoothing unit is used to perform smoothing using a mean filtering algorithm when the spatial density of an area exceeds a preset threshold;

[0198] Activity classification unit, used to classify and locate fault activities based on the smoothed data set using the support vector machine algorithm;

[0199] A dynamic rendering unit is used to generate a dynamically updated fault activity distribution map through a contour drawing algorithm.

[0200] The system or device is used to implement the functions of the method in the above-mentioned embodiment. Each module in the system or device corresponds to each step in the method, which has been explained in the method and will not be repeated here.

[0201] Through the above implementation, the problem of monitoring earthquake fault activity in the related art is solved, thereby ensuring that the problems existing in the prior art are solved.

[0202] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A method for monitoring earthquake fault activity, characterized in that: The following steps are involved: Obtain continuous vibration signals in complex geological areas through distributed fiber optic sensing technology to generate high-density vibration data streams; Fiber Bragg grating analysis is used to extract strain variation characteristics from the vibration data stream to obtain a weak precursor signal set; Applying a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event; If the time frequency distribution of the slow slip event exceeds a preset threshold, the microseismic events are classified using a convolutional neural network to generate an event location dataset; Extracting spatial coordinate information from the event location dataset and generating a high-resolution fault activity distribution map using a gridded spatial interpolation method; Dynamically update the fault activity distribution map through a real-time data stream processing framework to obtain a second-level positioning response result; If the second-level positioning response result shows an abnormal area, the pre-buried optical fiber network in the infrastructure is used to obtain supplementary vibration data to generate an enhanced strain distribution; Applying a spectral analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone; The coordinates of high-risk areas are extracted from the boundary range to generate a real-time earthquake risk distribution map.

2. The method according to claim 1, characterized in that The process of obtaining continuous vibration signals by using distributed optical fiber sensing technology includes: Optical fibers are laid in the target area, and the Rayleigh scattering effect is used to collect time series data of the first vibration signal; Performing signal denoising using a wavelet transform algorithm on the time series data of the first vibration signal to generate a second vibration signal; Amplitude, frequency, and phase features are extracted from high-quality time series data of the second vibration signal to obtain a first feature set.

3. The method according to claim 1, characterized in that The process of extracting strain variation characteristics by using fiber Bragg grating analysis includes: The weak signals related to strain changes are extracted through classification using the support vector machine algorithm.

4. The method according to claim 1, wherein The process of classifying microseismic events by using a convolutional neural network includes: If the amplitude feature exceeds a preset threshold, the feature set is classified and analyzed using a convolutional neural network; According to the classification results, the grid search algorithm is used to locate and calculate the time-frequency distribution of the microseismic event set.

5. The method according to claim 1, wherein The process of generating a distribution map using the gridded spatial interpolation method includes: Use spatial grid division algorithm to divide the spatial coordinate information into regions; For the divided area set, the Kriging interpolation algorithm is used to perform grid interpolation calculation; If the spatial density characteristics of the interpolated dataset exceed the preset threshold, it is smoothed using a two-dimensional convolution algorithm; Based on the smoothed data set, a contour drawing algorithm is used to generate a high-resolution fault activity distribution map.

6. The method according to claim 1, characterized in that The process of dynamically updating through the real-time data stream processing framework includes: Use streaming algorithms to segment time series data; For segmented data sets, a grid partitioning algorithm is used to divide the spatial coordinates into regions; If the spatial density of the region exceeds the preset threshold, it is smoothed using the mean filter algorithm; Based on the smoothed data set, the support vector machine algorithm is used to classify and locate the fault activity; A dynamically updated fault activity distribution map is generated through a contour drawing algorithm.

7. A monitoring system for earthquake fault activity, characterized in that: The system comprises: Signal acquisition module, used to acquire continuous vibration signals in complex geological areas through distributed fiber optic sensing technology and generate high-density vibration data streams; a strain feature extraction module for extracting strain variation features from the vibration data stream using fiber Bragg grating analysis to obtain a weak precursor signal set; a slow slip identification module, configured to apply a time-frequency analysis algorithm to the weak precursor signal set to determine the time-frequency distribution of the slow slip event; A microseismic classification and positioning module is configured to classify microseismic events and generate an event positioning dataset using a convolutional neural network when the time frequency distribution of the slow slip event exceeds a preset threshold; A spatial modeling module is used to extract spatial coordinate information from the event location dataset and generate a high-resolution fault activity distribution map using a gridded spatial interpolation method; A dynamic response module is used to dynamically update the fault activity distribution map through a real-time data stream processing framework to obtain a second-level positioning response result; an abnormality enhancement module, configured to acquire supplementary vibration data and generate an enhanced strain distribution using a pre-buried optical fiber network when the second-level positioning response result indicates an abnormal area; A boundary analysis module, configured to apply a spectral analysis algorithm to the enhanced strain distribution to determine the boundary range of the hidden fault zone; The risk output module is used to extract the coordinates of high-risk areas from the boundary range and generate a real-time earthquake risk distribution map.

8. A computer terminal device, characterized in that: include: one or more processors; a memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images