Seabed seismic oscillation prediction method and device based on seismic exploration

Abstract

The invention relates to the technical field of earthquake prediction, in particular to a submarine seismic oscillation prediction method and device based on earthquake exploration, and the method comprises the following steps: selecting a monitoring region, collecting the submarine topographic data of the monitoring region, and drawing a topographic map; selecting a plurality of monitoring points according to the topographic map; geomagnetic field information, electric wave information, micro-seismic activity information and seabed deformation information of each monitoring point are collected and recorded; converting the collected information into acoustic signals and transmitting the acoustic signals to a sea surface buoy, and converting the acoustic signals into electric signals by the sea surface buoy and transmitting the electric signals to a ground data center; by optimizing site selection of monitoring points, integrating geomagnetic-electric wave-microseismic-deformation multi-source data, realizing high-efficiency data transmission by using an acoustic-electric staged communication chain, and finally realizing early recognition of earthquake precursor in combination with threshold analysis, the method can significantly improve the coverage range, data precision and early warning timeliness of seafloor earthquake monitoring, and has good application prospects. And a technical support is provided for marine geological disaster prevention and control.

Description

Technical Field

[0001] The present invention relates to the technical field of earthquake prediction, and particularly to a method and device for predicting submarine ground motion based on seismic exploration. Background Art

[0002] Submarine earthquakes are geological disasters caused by the movement of oceanic plates, fault activities, or stress accumulation in subduction zones. They are characterized by strong suddenness and great destructive power, and may trigger secondary disasters such as tsunamis and submarine landslides, posing a serious threat to the safety of coastal areas. However, due to the complex submarine environment and the difficulty of monitoring, traditional earthquake prediction methods often rely on land seismic networks and it is difficult to capture the precursor signals of submarine earthquakes in a timely manner, resulting in insufficient warning time. Therefore, developing a set of efficient and accurate submarine ground motion monitoring and prediction systems has important scientific significance and application value.

[0003] Currently, submarine earthquake monitoring mainly faces the following challenges: Difficult data acquisition: The high-pressure and highly corrosive submarine environment requires extremely high stability of sensors, and signal transmission is easily affected by seawater attenuation. Insufficient multi-source information fusion: Monitoring a single physical quantity (such as seismic waves) is difficult to comprehensively reflect the characteristics of fault activities, and it is necessary to combine multi-parameters such as geomagnetic fields, electric waves, and deformations for comprehensive analysis. Summary of the Invention

[0004] The present invention provides a method and device for predicting submarine ground motion based on seismic exploration, which can effectively solve the problems in the background art.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is: On the one hand, the present application provides a method for predicting submarine ground motion based on seismic exploration, including the following steps: Select a monitoring area, collect the submarine terrain data of the monitoring area and draw a topographic map. Select several monitoring points according to the topographic map. Collect the geomagnetic field information, electric wave information, microseismic activity information, and seabed deformation information of each monitoring point and record them. Convert the collected information into acoustic signals and transmit them to the sea surface buoy, and the sea surface buoy then converts each acoustic signal into an electrical signal and transmits it to the ground data center. The data center performs denoising processing on the received signals and converts them into digital signals. Classify the digital signals of the same type of each monitoring point and set a threshold. When at least one type of digital signal exceeds the threshold, an earthquake prediction alarm is issued.

[0006] In some embodiments of the present invention, the prediction method further includes using satellite detection to monitor the abnormal electron concentration and abnormal seafloor heat flow in the monitoring area.

[0007] In some embodiments of the present invention, the seabed terrain data is collected by a multibeam sonar.

[0008] In some embodiments of the present invention, the monitoring points are close to the seabed faults in the topographic map, and the ground of the monitoring points is flat.

[0009] In some embodiments of the present invention, the time of the information collected by each of the monitoring points is synchronized.

[0010] In some embodiments of the present invention, the acoustic signal has at least one of an acoustic modulation protocol, relay node deployment, or adaptive power control function during transmission.

[0011] In some embodiments of the present invention, the denoising process includes eliminating ocean current movement noise, biological activity noise, solar wind interference, lightning interference, and tidal interference.

[0012] In some embodiments of the present invention, the threshold is a dynamically adjusted threshold, and the dynamically adjusted threshold includes signal intensity, duration, and spatial range.

[0013] In some embodiments of the present invention, the data center is also used to store digital signals and construct a prediction model based on the digital signals and historical earthquake event data, and use the prediction model to predict the occurrence time and intensity of seabed earthquakes.

[0014] On the other hand, the present application also provides a seabed ground motion prediction device based on seismic exploration. The device is executed according to the seabed ground motion prediction method based on seismic exploration. The device includes: An environmental monitoring unit for selecting a monitoring area, collecting seabed terrain data of the monitoring area, and drawing a topographic map; A positioning unit for selecting a number of monitoring points according to the topographic map; An acquisition unit for collecting geomagnetic field information, radio wave information, microseismic activity information, and seabed deformation information of each monitoring point and recording them; A transmission unit for converting the collected information into acoustic signals and transmitting them to a sea surface buoy, and the sea surface buoy then converts the acoustic signals into electrical signals and transmits them to a ground data center; A data center for denoising the received signals and converting them into digital signals; An analysis unit for classifying the same type of digital signals of each monitoring point and setting a threshold, and when at least one type of digital signal exceeds the threshold, issuing an earthquake prediction alarm.

[0015] Through the technical solution of the present invention, the following technical effects can be achieved: By optimizing the selection of monitoring points, integrating multi-source data of geomagnetism, radio waves, microseisms, and deformation, and using an acoustic-electric phased communication link to achieve efficient data transmission, and finally combining threshold analysis to achieve early identification of earthquake precursors, this method can significantly improve the coverage, data accuracy, and warning timeliness of submarine earthquake monitoring, providing technical support for the prevention and control of marine geological disasters. Brief Description of the Drawings

[0016] Figure 1 is a flowchart of the present invention; Detailed Embodiments

[0017] In the description of the present application, those skilled in the art should know that the present application can be implemented as a method, a device, an electronic device, and a computer-readable storage medium. Therefore, the present application can be specifically implemented in the following forms: complete hardware, complete software (including firmware, resident software, microcode, etc.), and a combination of hardware and software. In addition, in some embodiments, the present application can also be implemented in the form of a computer program product in one or more computer-readable storage media, which contains computer program code.

[0018] The above computer-readable storage media can be any combination of one or more computer-readable storage media. Computer-readable storage media include: electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of computer-readable storage media include: portable computer disks, hard disks, random access memories, read-only memories, erasable programmable read-only memories, flash memories, optical fibers, compact disc read-only memories, optical storage devices, magnetic storage devices, or any combination of the above. In the present application, the computer-readable storage media can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component.

[0019] In the technical solution of the present application, the acquisition, storage, use, processing, etc. of data all comply with the relevant provisions of national laws.

[0020] The present application describes the provided method, device, and electronic device through flowcharts and / or block diagrams.

[0021] It should be understood that each block of the flowchart and / or block diagram, as well as the combinations of blocks in the flowchart and / or block diagram, can be implemented by computer-readable program instructions. These computer-readable program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable data processing devices, thereby producing a machine. These computer-readable program instructions are executed by a computer or other programmable data processing devices, resulting in a device that implements the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0022] These computer-readable program instructions can also be stored in a computer-readable storage medium that can cause a computer or other programmable data processing apparatus to work in a specific manner. In this way, the instructions stored in the computer-readable storage medium produce an instruction device product that includes instructions for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0023] The computer-readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other device, so that a series of operation steps are executed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby enabling the instructions executed on the computer or other programmable data processing apparatus to provide a process for implementing the functions / operations specified in the blocks of the flowchart and / or block diagram.

[0024] The present application will be described in detail below with reference to the accompanying drawings in the present application. Embodiment

[0025] As Figure 1 shown, the method for predicting submarine ground motion based on seismic exploration of the present invention includes the following steps: S1. Select a monitoring area, collect submarine terrain data of the monitoring area, and draw a topographic map; This step is the primary link in submarine earthquake prediction. Its core objective is to determine potential high-risk earthquake areas and provide an accurate basis for the submarine geological structure for subsequent monitoring point layout; the monitoring area is selected based on the actual engineering needs or the sea area of densely populated areas, and reverse faults that may trigger tsunamis or submarine earthquakes are preferentially selected. The submarine terrain data can be collected in the following ways; Multi-beam bathymetric system (MBES): Mounted on a scientific research ship or AUV, it emits a fan-shaped sound beam (frequency 12 - 50 kHz), calculates the water depth through the echo time and angle, and generates a terrain grid with a resolution of 1m × 1m. This method has the advantages of full coverage, high precision, and the ability to identify fault cliffs, submarine landslides, etc.; Side-scan sonar (SSS): Provides information on the texture of the submarine surface (such as rock outcrops, sediment distribution), and assists in judging traces of fault activities (such as linear offset landforms); Light Detection and Ranging (LIDAR): Penetrates water bodies with blue-green lasers to obtain centimeter-level terrain details, and it is mainly applicable to shallow seas; Topographic map drawing: Import the collected data into a GIS platform (such as QGIS or ArcGIS), perform coordinate unification and grid interpolation to generate a digital elevation model (DEM) and a three-dimensional seabed topographic map, calculate derivative parameters such as slope, aspect, and curvature, mark steep slopes (>30°) and flat areas (<5°), detect linear structures through an automatic fault identification algorithm (such as the Hough transform or machine learning), verify the extensibility and activity of faults by combining seismic profile data (such as single-channel seismic), and then output a topographic map with comprehensive information such as topographic contour lines, fault lines, and historical epicenter overlay layers. At the same time, high, medium, and low-risk areas can be divided according to the terrain complexity and fault density; Of course, in some embodiments, seabed terrain data can also be collected by a multibeam sonar. The multibeam sonar system mainly includes a transmitting transducer array, a receiving transducer array, a motion sensor, a GNSS positioning system, and a sound velocity profiler; the functions of each component are as follows: Transmitting transducer array: Transmit a high-frequency fan-shaped sound wave beam (such as 30 kHz, beam angle 120°) to cover a strip area perpendicular to the course; Receiving transducer array: Receive the echo signal reflected from the seabed and record the arrival time and angle of each beam; Motion sensor (MRU): Real-time monitor the roll, pitch, and heave of the hull, and dynamically compensate for the geometric deviation of sound wave emission / reception; Sound velocity profile data: Combine the real-time sound velocity profile and correct the bending of the sound wave propagation path (ray tracing algorithm) to ensure accurate water depth calculation; GNSS positioning system: Synchronously record the emission position of each beam to provide spatial coordinates for subsequent data splicing.

[0026] In actual use, data collection can be carried out according to the following steps: Before navigation: The GNSS delimits 10 parallel survey lines (spacing 6 km), and the SVP measures the sound velocity profile; Collection: Transmit 30 kHz sound waves (coverage width 7 km), and the MRU compensates for the 5° roll of the hull in real time; Processing: The software eliminates the edge beams (opening angle >60°) and generates a 30 m grid DTM; Result: The topographic map shows that the height difference of the fault scarp is 200 m, which guides the layout of monitoring points.

[0027] The multibeam sonar realizes the accurate collection of the seabed terrain through the collaboration of multiple components (acoustics, positioning, attitude, sound velocity) and hierarchical data processing. Each component plays a specific role in the process and finally outputs a high-resolution terrain model, providing basic spatial data support for the seismic monitoring system.

[0028] S2. Select several monitoring points according to the topographic map; In this method, the core principle for selecting monitoring points is the priority of tectonic activities, that is, in the near - field of the fault. The monitoring points need to be deployed within 1 - 3 times the fault width on both sides of the submarine fault trace (for example, for a 10 - km - wide fault, it is deployed in the range of 10 - 30 km). This is because the fault zone is a stress - concentration area, and seismic precursor signals (such as microseisms and geomagnetic anomalies) are more significant in this area. Moreover, for plate - boundary faults (such as subduction zones), points need to be symmetrically arranged on the hanging wall and footwall to capture differential deformation; At the same time, the seabed surface needs to be flat, that is, select areas with a seabed slope < 5° (verified by calculating the slope - aspect matrix from multibeam topographic data). This can ensure stable contact of the monitoring equipment, avoid slippage (when the inclination > 10°, the risk of equipment overturning increases by 80%), and reduce the shielding effect of terrain on acoustic signal transmission (steep slopes cause multi - path interference of sound waves); The specific selection process is as follows: Fault spatial analysis: Use GIS tools to extract the fault line from the topographic map (such as through curvature analysis or manual interpretation), and generate a fault buffer zone. The buffer zone width = fault slip rate (mm / year) × coefficient (such as 1 km / mm / yr).

[0029] Flat area screening: Input multibeam DEM data, calculate the slope map and roughness index, extract areas with a slope < 5° and roughness < 0.1, and combine with side - scan sonar images to exclude pseudo - flat areas with gravel or bio - mounds on the surface; Optimal point determination: Overlay the fault buffer zone and the flat - area layer, take the intersection area, and confirm the basement lithology through close - range AUV photography (preferably select hard basements such as basalt); The selection of monitoring points is a process of balancing tectonic sensitivity and engineering feasibility, which requires comprehensive geological interpretation, terrain quantitative analysis, and on - site verification. This standard can ensure that the observation equipment can not only effectively capture precursor signals but also work stably for a long time.

[0030] S3. Collect the geomagnetic field information, radio wave information, microseismic activity information, and seabed deformation information of each monitoring point and record them; In this step, geomagnetic field information can be obtained by an Ocean Bottom Magnetometer (OBM), which mainly generates a piezomagnetic effect (a change of 0.1 - 10 nT) through fault friction and can reflect the rock stress state; radio wave information can be obtained by an Ocean Bottom Electromagnetic instrument (OBEM), whose measurement range is ultra - low - frequency electric field (0.001 - 10 Hz), and its source is the electromagnetic signal induced by pore fluid flow (10 - 100 μV / km), and this information can indicate the change of fault permeability; microseismic activity information can be obtained by a broadband Ocean Bottom Seismometer (OBS), whose measurement range is 0.01 - 50 Hz, which can capture foreshocks / slow earthquakes (ML≥0.5) and can locate fault micro - ruptures; seabed deformation information can be obtained by a pore pressure gauge / tilt meter, which mainly measures vertical displacement and tilt angle, and seabed deformation is mainly caused by seabed uplift / subsidence due to fault creep; During acquisition, each monitoring point needs to be carried out simultaneously, and the synchronization error ≤ 1 ms (corresponding to the seismic wave propagation distance of about 5 m). The specific measurement method can adopt the pulse test method, where the master node emits an acoustic / optical trigger signal and calculates the dispersion of the response time of each node; or adopt the natural event method, using the time difference of the arrival of teleseismic P - waves at each station to inversely calculate the clock deviation; This technical system realizes multi - parameter three - dimensional observation of tectonic activities through multi - sensor collaborative networking and high - precision time synchronization. The time synchronization error is controlled at the millisecond to microsecond level, meeting the correlation analysis requirements of earthquake precursor signals (such as the electromagnetic - deformation coupling phenomenon accompanied by slow slip events), and providing an evidence chain at the physical mechanism level for earthquake prediction.

[0031] S4. Convert the collected information into acoustic signals and transmit them to the sea - surface buoy, and then the sea - surface buoy converts each acoustic signal into an electrical signal and transmits it to the ground data center; Since electromagnetic waves, electrical signals, etc. attenuate extremely fast in water, while the attenuation of sound waves is relatively small, sound waves are the main means of underwater long - distance communication. Converting the collected information into acoustic signals can help with information transmission; in air, the transmission of electrical signals is more convenient and its attenuation is relatively small. Therefore, when the signal reaches the sea - surface buoy, the acoustic signal can be converted into an electrical signal for transmission. In this way, signal transmission functions with different paths are realized in water and air, and the signal transmission efficiency is maximized. The buoy can transmit the electrical signal to the ground station through radio (such as satellite, 4G / 5G) or optical fiber and finally reach the data center; In the process of acoustic transmission, the following technical means can also be used to enhance the signal transmission quality: Acoustic modulation protocols, which define how sound waves encode data (such as FSK, PSK, OFDM, etc.), solve problems such as underwater multipath and Doppler effects. For example, adaptive modulation is adopted to dynamically adjust the modulation method according to the channel quality to improve reliability; Relay node deployment. In long-distance transmission, relay nodes (such as autonomous underwater vehicles, fixed repeaters) are deployed to relay and forward signals to avoid signal attenuation. For example, in deep-sea monitoring, the single-hop communication distance is limited (usually several kilometers), and multi-hop relay is required. Adaptive power control. Dynamically adjust the transmission power according to environmental noise, distance, etc. to balance energy consumption and communication quality. Its main advantage is to save energy (which is crucial for underwater devices with battery constraints) and reduce multi-user interference. In actual use, it can at least possess one of the above three functions (or a combination) to address various challenges in underwater communication. Through underwater acoustic-radio cross-media communication, combined with technologies such as modulation protocols, relay deployment, or power control, reliable transmission of underwater data to the ground center is achieved, which is an important part of ocean informatization.

[0032] S5. The data center performs denoising processing on the received signal and converts it into a digital signal. The signal received by the data center from the sea surface buoy is an analog electrical signal after "underwater acoustic → electrical signal" conversion. This signal may contain valid data (such as temperature, salinity, sonar images collected by sensors) and environmental noise (interferences such as ocean currents, organisms, solar wind, lightning, tides, etc.). Due to the complex underwater acoustic channel, the signal may have been affected by multipath effects, Doppler frequency shift, attenuation, etc. Therefore, denoising and digitization processing are required first.

[0033] For different types of noise, different denoising methods can be adopted, as follows: Ocean current movement noise, whose source is the scattering of sound waves caused by the flow of seawater. The denoising method is to dynamically adjust the filter parameters, track the changing noise, and separate the noise in different frequency bands. Biological activity noise, whose source is the sound emitted by marine organisms such as whales and shrimps (such as "shrimp chirping"). The denoising method is to train a deep learning model (such as CNN) to identify and filter out biological sound patterns, or suppress it in a specific frequency band (such as 2 - 20 kHz). Solar wind interference, whose source is the coupling of electromagnetic noise caused by solar activities into the underwater acoustic channel. The denoising method is to use electromagnetic shielding to reduce coupling interference, or adopt wavelet denoising to eliminate high-frequency pulse noise. Lightning interference, whose source is the transient high-voltage pulse conducted by atmospheric lightning through water. The denoising method is to adopt transient noise suppression (TNR) to detect and eliminate sudden spikes, or use non-linear filtering (such as median filtering) to smooth outliers. Tidal interference, whose source is the low-frequency pressure fluctuation caused by periodic water level changes. The denoising method is to adopt synchronous averaging to use the tidal periodicity for noise cancellation, or use high-pass filtering to filter out low-frequency fluctuations (<1 Hz). Since computers generally process data as digital signals when dealing with data, it is necessary to convert the denoised signal into a digital signal. Specifically, according to the Nyquist sampling theorem, the sampling frequency ≥ 2 times the highest signal frequency (for example, a sampling rate of 10 - 50 kHz is commonly used in underwater acoustic communication). The continuous amplitude is discretized (such as a 16-bit ADC with a dynamic range of 96 dB), and then it is converted into binary data (such as PCM encoding) for easy storage and analysis; Denoising and digitization in the data center are the last link in the underwater information chain. It is necessary to integrate technologies such as signal processing (filtering / transformation), hardware (ADC), and AI (noise recognition) to cope with the extreme complexity of the marine environment. This process directly determines the usability of the data and is the core technical support in fields such as ocean observation, resource exploration, and national defense.

[0034] S6. Classify the same type of digital signals at each monitoring point and set a threshold. When at least one type of digital signal exceeds the threshold, a seismic prediction alarm is issued; Specifically, group the monitoring data of the same type of digital signals (such as geomagnetic field, radio waves, microseismic activities, and seabed deformation) by type. For example, Category A: geomagnetic field signals; Category B: radio wave signals; Category C: microseismic activity signals; Category D: seabed deformation signals. Set the threshold according to the distribution positions of various signals and monitoring points in the topographic map. This threshold can be the threshold of various signals at each monitoring point or the overall threshold of a single type of signal in the topographic map. When the digital signal exceeds the threshold, a seismic prediction alarm can be sent via text message or through methods such as sound, light, and electricity; In actual use, the threshold is not a fixed value but is dynamically adjusted based on three dimensions: signal intensity, duration, and spatial range. For example, if the intensity of Category A signals in a certain area exceeds 3 times the standard deviation of the historical mean, lasts for 120 seconds, and 5 monitoring points are triggered simultaneously, it is determined as a seismic precursor; The adjustment process of the dynamic threshold includes: In the basic data preparation stage, the system continuously stores monitoring data for many years, classifies and archives them according to different geological regions. Each monitoring device uploads the latest readings per second to form a continuous data stream, and simultaneously obtains background information such as ocean meteorology and ocean current activities; The threshold dynamic update process continuously monitors the environmental noise level. When the background interference increases (such as during a storm), the intensity threshold is automatically increased, and the threshold is lowered during a calm period to improve sensitivity; Multi-dimensional collaborative judgment. In the intensity dimension, compare the deviation degree of the current signal from the dynamic baseline; in the time dimension, track the change in the duration of the abnormal signal; in the spatial dimension, monitor the geographical diffusion range of the abnormal signal; Threshold application. Generate a new set of threshold parameters every 5 minutes, and at the same time retain 3 sets of historical thresholds as verification references. When the signal exceeds the threshold, trigger a review mechanism, and after confirmation, upgrade it to a warning event; This dynamic adjustment mechanism effectively reduces the false alarm rate while ensuring the timeliness of early warning through continuous environmental adaptation and multi-party verification; When at least one type of digital signal exceeds the threshold, an earthquake prediction alarm is issued, enabling a multi-condition joint determination method to avoid phenomena such as missed alarms (e.g., high sensitivity is required for tsunami early warning) and false alarms. At the same time, the early warning rating can be determined according to the number of signal types exceeding the threshold. For example: yellow early warning, when a single type of signal exceeds the threshold, the corresponding measure is to strengthen monitoring and conduct manual review; orange early warning, when two types of signals exceed the threshold for 5 minutes, the response measure is to notify the coastal emergency department; red early warning, when multiple types of signals exceed the threshold + large-scale synchronous anomalies + continuous for 10 minutes, the corresponding measure is to issue a public alarm and initiate evacuation; This system realizes efficient earthquake prediction in complex marine environments through dynamic multi-dimensional threshold criteria and multi-source signal fusion. Its core lies in: Adaptive threshold: Avoid the failure of fixed thresholds in a changing environment; Space-time-intensity joint analysis: Extract real precursor signals from noise; Hierarchical response mechanism: Balance the sensitivity of early warning and the cost of false alarms.

[0035] In some embodiments of the present invention, the prediction method further includes using satellite detection to monitor the abnormal electron concentration and abnormal seafloor heat flow in the monitoring area; During the crustal stress accumulation in the earthquake gestation stage, the increase in radon gas escaping from the surface will be triggered. The decay of radioactive elements releases α particles, which ionize atmospheric molecules, resulting in an abnormal increase in the electron density of the local ionosphere. 6 - 72 hours before the earthquake, a sharp increase in electron concentration appears at a height of 300 km above the epicenter, with a typical increase of 20 - 35%, and the coverage range can reach 500 km around the epicenter. The electron concentration is inverted using the delay amount (TEC value) when the signal of the GPS / Beidou satellite navigation system passes through the ionosphere, or a dedicated satellite is used for electron concentration detection; The reason for the abnormal seafloor heat flow is that the pre-seismic slow slip causes the expansion of fractures and the upwelling of deep hydrothermal fluids. At the same time, the increase in micro-fractures in the rock increases the heat conduction efficiency; This technical system marks a paradigm shift in earthquake prediction from "predicting earthquakes based on earthquakes" to "identifying precursors in multiple physical fields", effectively extending the prediction time and improving the prediction accuracy.

[0036] In some embodiments of the present invention, the data center is also used to store digital signals and construct a prediction model based on the digital signals and historical earthquake event data, and use the prediction model to predict the occurrence time and intensity of submarine earthquakes; In the present invention, the data center not only undertakes the function of data storage, but also constructs a prediction model by integrating multi-source monitoring data (such as digital signals of geomagnetism, radio waves, micro-seismic activities, seabed deformation, etc.) and historical earthquake event data to predict the occurrence time and intensity of submarine earthquakes; The digital signals stored in the data center include: Real-time monitoring data: geomagnetic field signals (such as piezomagnetic effects caused by fault friction), radio wave signals (such as electromagnetic changes caused by pore fluid flow), micro-seismic activity signals (such as magnitudes and frequencies of foreshocks or slow earthquakes), seabed deformation signals (such as vertical displacements, changes in tilt angles); Historical earthquake event data: magnitudes, epicenter locations, occurrence times, focal mechanisms, etc. of historical earthquakes, and historical earthquake precursor signals (such as spatio-temporal characteristics of geomagnetic anomalies, radio wave changes); When storing data, it is stored in a standardized format (such as CSV, HDF5, or database tables) to ensure the integrity of timestamps (accurate to the millisecond level) and spatial coordinates (such as longitude, latitude, depth). For example, a time series database (such as InfluxDB) or a distributed storage system (such as Hadoop) is used to process massive data; The process of constructing the prediction model includes: Data preprocessing, removing noise (such as ocean current interference, biological sound patterns, tidal fluctuations), using wavelet denoising, median filtering, or deep learning models (such as CNN) to identify and remove outliers, and synchronizing time series data to ensure the time consistency of multi-source signals (error ≤ 1ms); Feature extraction, extracting statistical features (such as mean, variance, slope, mutation points) from time series, extracting frequency domain features (such as spectral energy after Fourier transform), and spatial features (such as deformation gradients in fault activity areas, spatial correlations of multiple parameters); Model selection and training, selecting a time prediction model (such as ARIMA, used to predict earthquake occurrence events) or an intensity prediction model (such as dividing magnitudes into several levels) according to the prediction target; The training strategy is to divide historical data into a training set (70%), a validation set (15%), and a test set (15%), and adopt time series cross-validation (such as the rolling window method) to avoid future data leakage, and adjust model parameters (such as learning rate, tree depth, number of neural network layers) through grid search or Bayesian optimization; Model evaluation and optimization, evaluating the model using the mean square error (MSE). For example, in the time prediction model, the accuracy of the predicted time can be measured, and in the intensity prediction model, the accuracy of the predicted intensity can be evaluated; The optimization methods can be ensemble learning: combining multiple models (such as LSTM + random forest) to improve robustness; Online learning: updating the model in real time to adapt to changes in data distribution (such as background noise fluctuations); Transfer learning: using earthquake data from other regions to pre-train the model to reduce the need for labeled data in new regions. Embodiment

[0037] The submarine ground motion prediction device based on seismic exploration of the present invention, the device is executed according to the submarine ground motion prediction method based on seismic exploration, and the device includes: An environmental monitoring unit for selecting a monitoring area, collecting submarine terrain data of the monitoring area and drawing a topographic map; A positioning unit for selecting a number of monitoring points according to the topographic map; An acquisition unit for collecting geomagnetic field information, radio wave information, microseismic activity information and seabed deformation information of each monitoring point and recording them; A transmission unit for converting the collected information into acoustic signals and transmitting them to a sea surface buoy, and the sea surface buoy then converts the acoustic signals into electrical signals and transmits them to a ground data center; A data center for denoising the received signals and converting them into digital signals; An analysis unit for classifying the digital signals of the same type of each monitoring point and setting a threshold value, and when at least one type of digital signal exceeds the threshold value, issuing a seismic prediction alarm.

[0038] In addition, the present application also provides an electronic device, including a bus, a transceiver, a memory, a processor, and a computer program stored on the memory and executable on the processor. The transceiver, the memory, and the processor are respectively connected through the bus. When the computer program is executed by the processor, it realizes each process of the method embodiment for controlling and outputting data, and can achieve the same technical effect. To avoid repetition, it will not be described in detail here.

[0039] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present invention, several improvements and modifications can be made, and these improvements and modifications should also be regarded as the protection scope of the present invention.

Claims

1. A method for predicting seafloor ground motion based on seismic exploration, characterized in that: The steps include: Select the monitoring area, collect the seabed topography data of the monitoring area and draw a topographic map; Select several monitoring points based on the topographic map; Collect and record the geomagnetic field information, radio wave information, microseismic activity information and seabed deformation information at each monitoring point; The collected information is converted into acoustic signals and transmitted to the sea surface buoy, which then converts the acoustic signals into electrical signals and transmits them to the ground data center; The data center denoises the received signal and converts it into a digital signal; The same type of digital signals at each monitoring point are classified and thresholds are set. When at least one type of digital signal exceeds the threshold, an earthquake prediction alarm is issued.

2. The method for predicting seabed ground motion based on seismic exploration according to claim 1, characterized in that: The prediction method also includes using satellite detection to monitor regional electron concentration anomalies and seabed heat flow anomalies.

3. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The seabed topography data is collected by multi-beam sonar.

4. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The monitoring point is close to the submarine fault in the topographic map, and the ground at the monitoring point is flat.

5. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The information collected by each monitoring point is synchronized in time.

6. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The acoustic signal has at least one of an acoustic modulation protocol, a relay node deployment or an adaptive power control function during transmission.

7. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The denoising process includes eliminating ocean current movement noise, biological activity noise, solar wind interference, lightning interference and tidal interference.

8. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The threshold is a dynamically adjusted threshold, and the dynamically adjusted threshold includes signal strength, duration, and spatial range.

9. The method for predicting seafloor ground motion based on seismic exploration according to claim 1, characterized in that: The data center is also used to store digital signals and build a prediction model based on the digital signals and historical earthquake event data, and use the prediction model to predict the time and intensity of submarine earthquakes.

10. A device for predicting seafloor earthquake motion based on seismic exploration, characterized in that: The device is implemented according to the method for predicting seabed seismic motion based on seismic exploration as claimed in any one of claims 1 to 9, and the device comprises: Environmental monitoring unit, used to select monitoring areas, collect seabed topographic data of the monitoring areas and draw topographic maps; A positioning unit, used to select several monitoring points according to the topographic map; An acquisition unit is used to collect and record the geomagnetic field information, radio wave information, microseismic activity information and seabed deformation information of each monitoring point; The transmission unit is used to convert the collected information into acoustic signals and transmit them to the sea surface buoy, and the sea surface buoy then converts the acoustic signals into electrical signals and transmits them to the ground data center; A data center for denoising the received signal and converting it into a digital signal; The analysis unit classifies the same type of digital signals from each monitoring point and sets thresholds. When at least one type of digital signal exceeds the threshold, an earthquake prediction alarm is issued.

Images