Earthquake fault monitoring method and system based on Beidou

Abstract

The invention provides a Beidou-based seismic fault monitoring method and system, and relates to the technical field of seismic fault monitoring, and the method comprises the steps: obtaining Beidou satellite data of a plurality of monitoring points of a seismic fault within a preset sampling time; preprocessing each piece of Beidou satellite data, and inputting each piece of preprocessed Beidou satellite data into a preset prediction model for displacement prediction; calculating the sliding amount of each monitoring point in each time step in the future based on each predicted displacement amount in each time step; carrying out the risk early warning of the seismic fault based on the change of the sliding amount of each monitoring point in each time step in the future, and obtaining a risk early warning result. According to the method, the sliding amount of multiple monitoring positions of the seismic fault in multiple time steps in the future is predicted through the Beidou satellite data, the risk of the seismic fault is evaluated according to the change of the predicted sliding amount in the multiple time steps in the future, and the seismic fault can be monitored more accurately in real time.

Description

Technical Field

[0001] The present invention relates to the technical field of seismic fault monitoring. Specifically, it relates to a Beidou-based seismic fault monitoring method and system. Background Art

[0002] Seismic fault monitoring is an important means to detect the activity of faults and the risk of potential earthquakes in advance. The monitoring results of seismic faults are involved in aspects such as earthquake early warning, disaster assessment, and scientific research. Its ultimate goal is to minimize the impact of earthquake disasters on human society. Through long-term and real-time monitoring and analysis of seismic fault data, researchers can better understand the occurrence mechanism of earthquakes, improve earthquake prediction and emergency response capabilities, thereby providing strong support for ensuring the safety of people's lives and property and social stability.

[0003] Traditional seismic fault monitoring requires a large number of ground devices to be installed at specific locations, and data collection and processing are carried out manually. There are not only problems such as limited monitoring range and high dependence on ground facilities and manual maintenance, but also the real-time performance and accuracy of data acquisition are not ideal.

[0004] The global coverage and all-weather characteristics of the Beidou system can enable a wider monitoring range of seismic faults, especially for remote or inaccessible areas, thus filling the blind spots of traditional ground monitoring. Moreover, the millimeter-level positioning accuracy and almost delay-free real-time data provided by Beidou satellites can significantly improve the monitoring accuracy and response speed, thereby enhancing the capabilities of earthquake early warning and fault activity detection. At the same time, the cost of data acquisition and processing based on the Beidou satellite system is relatively low, reducing the dependence on ground facilities and manual maintenance.

[0005] Therefore, there is an urgent need for a method and system for seismic fault monitoring based on Beidou data to improve the accuracy and real-time performance of seismic fault monitoring while reducing the operating cost. Summary of the Invention

[0006] The purpose of the present invention is to provide a Beidou-based seismic fault monitoring method and system to improve the above problems. To achieve the above purpose, the technical solutions adopted by the present invention are as follows:

[0007] In a first aspect, the present application provides a Beidou-based seismic fault monitoring method, including:

[0008] Obtaining Beidou satellite data of multiple monitoring points of a seismic fault within a preset sampling time;

[0009] Preprocess each piece of the Beidou satellite data, and input each piece of the preprocessed Beidou satellite data into a preset prediction model for displacement prediction to obtain multiple predicted displacements. One predicted displacement includes the displacements of one monitoring point at multiple future time steps;

[0010] Calculate the sliding amount of each monitoring point at each future time step based on each predicted displacement at each time step to obtain multiple sliding amounts;

[0011] Based on the changes in the sliding amounts of each monitoring point at each future time step, perform risk early warning for the seismic fault to obtain a risk early warning result.

[0012] In a second aspect, the present application also provides a Beidou-based seismic fault monitoring system, including:

[0013] An acquisition module, configured to acquire Beidou satellite data of multiple monitoring points of a seismic fault within a preset sampling time;

[0014] A first processing module, configured to preprocess each piece of the Beidou satellite data, and input each piece of the preprocessed Beidou satellite data into a preset prediction model for displacement prediction to obtain multiple predicted displacements. One predicted displacement includes the displacements of one monitoring point at multiple future time steps;

[0015] A second processing module, configured to calculate the sliding amount of each monitoring point at each future time step based on each predicted displacement at each time step to obtain multiple sliding amounts;

[0016] A third processing module, configured to perform risk early warning for the seismic fault based on the changes in the sliding amounts of each monitoring point at each future time step to obtain a risk early warning result.

[0017] The beneficial effects of the present invention are as follows:

[0018] By preprocessing the Beidou satellite data, and then predicting the sliding amounts of multiple monitoring positions of the seismic fault at multiple future time steps based on the preprocessed Beidou satellite data, and evaluating the risk of the seismic fault according to the changes in the predicted sliding amounts at multiple future time steps, the present invention can monitor the seismic fault more real-time and accurately.

[0019] Other features and advantages of the present invention will be described in the subsequent specification, and part of them will become obvious from the specification, or can be understood by implementing the embodiments of the present invention. Description of the Drawings

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0021] Figure 1 Schematic flow chart of a Beidou-based earthquake fault monitoring method described in the embodiments of the present invention;

[0022] Figure 2 Schematic structural diagram of a Beidou-based earthquake fault monitoring system described in the embodiments of the present invention. Detailed implementation manners

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Generally, the components of the embodiments of the present invention described and illustrated in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed present invention, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0024] It should be noted that: similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present invention, terms such as "first" and "second" are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0025] Embodiment 1:

[0026] This embodiment provides a Beidou-based earthquake fault monitoring method.

[0027] See Figure 1 , which shows that this method includes steps S1, S2, S3, and S4.

[0028] S1. Obtain Beidou satellite data of multiple monitoring points of the earthquake fault within a preset sampling time.

[0029] It is understandable that the Beidou satellite system has high-precision positioning capabilities and can provide centimeter-level accuracy, which is crucial for the precise monitoring and prediction of seismic fault displacement. At the same time, the Beidou system can conduct real-time observations, enabling the immediate acquisition of the movement of the earth's crust before and after an earthquake, improving the real-time nature of the prediction of the entire seismic fault displacement, and thus more quickly detecting the activity status and change trends of the fault. Moreover, based on the high spatial resolution of the satellite signal receiver, the subtle impact of seismic activities on crustal changes can be monitored more accurately, thereby providing more precise data for subsequent prediction of the displacement of seismic faults.

[0030] Specifically, in this embodiment, Beidou satellite signal receivers are deployed every 10 km along the seismic fault to be monitored. The position of each Beidou satellite signal receiver is used as a monitoring point, and the sampling frequency is set to 1 Hz, and the time window is 24 hours, that is, the sampling time for each time is 24 hours, to collect Beidou satellite data. It is understandable that the deployment interval, sampling frequency, and time window of the Beidou satellite signal receivers can all be adjusted to adapt to different monitoring requirements.

[0031] S2. Preprocess each of the Beidou satellite data, and input each preprocessed Beidou satellite data into a preset prediction model for displacement prediction to obtain a plurality of predicted displacements. One predicted displacement includes the displacements of one monitoring point at multiple future time steps.

[0032] It is understandable that the original Beidou satellite data mainly includes carrier phase, pseudorange, Doppler frequency shift, timestamp, and satellite ephemeris. These data can be used to calculate the real-time geographical location of the Beidou satellite signal receiver, and thus infer the changes in the surface of the receiver's location at different time steps based on the changes in the geographical location of the receiver. However, the original Beidou satellite data cannot be directly used to analyze the surface deformation of the monitoring point. Through data preprocessing, the original Beidou satellite data needs to be converted into a time series to facilitate processing and calculation and describe the change of the position of the monitoring point over time. Therefore, it is necessary to preprocess each of the Beidou satellite data. The specific steps include:

[0033] S211. Perform cycle slip detection and repair on the carrier phase data of each piece of the Beidou satellite data, and perform error correction on the Beidou satellite data after each carrier phase is repaired to obtain multiple pieces of the Beidou satellite data after error correction. By fitting the carrier phase data, detect abnormal residual points, and repair cycle slips through difference or re-initialization. After completing cycle slip repair, it is also necessary to eliminate systematic errors in the Beidou satellite data. In this embodiment, the ionospheric delay is calculated and eliminated through dual-frequency observations, then the existing Saastamoinen model is used for tropospheric calibration, and then the satellite orbit error and clock error are repaired through precise ephemeris and clock error products. Finally, multi-path interference is reduced through filtering. It can be understood that the above cycle slip detection and repair and error correction of the original Beidou satellite data are common processing methods for those skilled in the art.

[0034] S212. Perform position calculation on each piece of the Beidou satellite data after error correction, and generate a surface deformation time series for each monitoring point according to the result of the position calculation to obtain multiple surface deformation time series. One surface deformation time series includes the position, velocity, and acceleration of a monitoring point at each time step during the sampling time. It can be understood that when generating the surface deformation time series for each monitoring point, first, the corrected Beidou satellite data needs to be converted into the three-dimensional position of the receiver, that is, the longitude, latitude, and altitude of each monitoring point where the receiver is located; then the initial position of the receiver is calculated through pseudo-range observations, and then differential processing is performed using the carrier phase data of the reference station and the rover station to eliminate common errors. Finally, the precise position of the receiver is calculated through Kalman filtering to obtain the three-dimensional coordinates of the receiver at each time step.

[0035] After obtaining the three-dimensional coordinates of the receiver at each time step, the velocity and acceleration of the receiver at each time step are calculated through time difference method, so that the three-dimensional coordinates, three-dimensional velocities, and three-dimensional accelerations of each receiver at each time step can be sorted by time stamp to obtain the surface deformation time series of each receiver.

[0036] S213. Generate a spatio-temporal matrix based on multiple surface deformation time series to obtain the spatio-temporal matrix. It can be understood that after obtaining the surface deformation sequences of each receiver, since these data will be input into a preset neural network for processing later, it is necessary to further construct a spatio-temporal matrix according to all the obtained surface deformation time series. The specific steps include:

[0037] First, the data of the receivers need to be resampled according to a unified timestamp to ensure that the time points of all surface deformation time series are aligned. Then, an adjacency matrix is generated based on the geographical locations of the receivers. The generated adjacency matrix is used to extract the spatial features of the spatio-temporal matrix subsequently. Finally, the sequences of all receivers are stacked into a three-dimensional matrix to obtain the spatio-temporal matrix, which contains the deformation features of all monitoring points within the time window.

[0038] After completing the data preprocessing, the obtained spatio-temporal matrix will be used as the input of a preset prediction model, and the prediction model will predict the displacement of each of the said monitoring points at multiple future time steps.

[0039] S3. Calculate the slip amount of each of the said monitoring points at each future time step based on each of the predicted displacement amounts at each time step to obtain a plurality of said slip amounts. It can be understood that the slip amount of a seismic fault is a necessary basis for evaluating the activity of the seismic fault. After predicting the displacements of multiple monitoring points of a seismic fault at multiple future time steps in the present invention, the displacement amount is further converted into a slip amount. The slip amounts and the change trend of the slip amounts at multiple time steps not only more intuitively display the activity conditions of each region of the seismic fault, but also provide a data basis for subsequent more comprehensive risk early warning, achieving a better monitoring effect on the seismic fault.

[0040] Specifically, step S3 includes:

[0041] S31. Obtain first information, where the first information includes the strike angle, dip angle, and locking depth of the seismic fault;

[0042] S32. Construct a local coordinate system with a preset reference point as the origin to obtain the local coordinate system. It can be understood that a reference point on the fault is usually selected as the center of the locked section or a known rupture point. The X-axis of the local coordinate system is the horizontal direction along the fault strike, the Y-axis of the local coordinate system is the direction perpendicular to the fault strike and pointing to the hanging wall of the fault, and the Z-axis of the local coordinate system is the direction perpendicular to the fault plane, which is determined by the right-hand rule. After establishing the local coordinate system, a coordinate transformation matrix is constructed using the unit vectors in each direction of the local coordinate system for subsequent coordinate transformation.

[0043] S33. Establish a fault geometry model based on the first information and the local coordinate system to obtain the fault geometry model. It can be understood that the fault geometry model includes the strike angle, dip angle, locking depth, and coordinate transformation matrix, providing a spatial framework for fault activity. Fault slip is the movement along the fault plane. It is necessary to establish a fault geometry model and clarify the three-dimensional geometric parameters of the fault to project the surface displacement onto the fault plane.

[0044] S34. Calculate the sliding amount based on the fault geometric model and the displacement amount of each monitoring point at each future time step, and obtain the sliding amount of each monitoring point at each future time step.

[0045] Specifically, after the establishment of the local coordinate system and the fault geometric model, it is first necessary to project the displacement of the receiver onto the local coordinate system of the fault, and then calculate the tangential sliding amount based on the displacement projected onto the local coordinate system. The calculation formula is:

[0046]

[0047] where Δs i is the sliding amount of the i-th receiver, is the displacement of the projection of the displacement of the i-th receiver in the X-axis direction of the local coordinate system, is the displacement of the projection of the displacement of the i-th receiver in the Y-axis direction of the local coordinate system.

[0048] S4. Conduct risk warning for the seismic fault based on the change of the sliding amount of each monitoring point at each future time step, and obtain the risk warning result.

[0049] Embodiment 2:

[0050] The difference between this embodiment and Embodiment 1 is only that, further, in step S2, inputting each of the preprocessed Beidou satellite data into a preset prediction model for displacement prediction includes the following steps:

[0051] S221. Extract local features of the spatio-temporal matrix based on a preset one-dimensional convolutional neural network to obtain a first output feature. The one-dimensional convolutional neural network extracts local features of each surface deformation time series along the time dimension, such as acceleration mutation.

[0052] S222. Extract the long-range time dependence of the first output feature based on a preset bidirectional long short-term memory network to obtain a time feature matrix. The bidirectional long short-term memory network extracts the long-range time dependence of the output of the one-dimensional convolutional neural network, thereby outputting a time feature matrix and completing the extraction of the time features of the spatio-temporal matrix.

[0053] S223. Generate a spatial adjacency matrix based on the positions of the monitoring points, and input the spatial adjacency matrix and the spatio-temporal matrix into a preset graph convolutional network for feature aggregation to obtain a spatial feature matrix. The graph convolutional network aggregates the features of adjacent receivers according to the spatio-temporal matrix and the adjacency matrix containing the geographical location information of the receivers, thereby extracting the spatial features in the spatio-temporal matrix.

[0054] S224. Perform weighted fusion of the time feature matrix and the spatial feature matrix based on the attention mechanism to obtain a spatio-temporal feature matrix. After obtaining the time features and spatial features, calculate the spatio-temporal feature weights through the attention mechanism and perform weighted fusion to obtain the fused spatio-temporal feature matrix.

[0055] S225. Input the spatio-temporal feature matrix into a preset temporal convolutional network to predict the displacement of each monitoring point at multiple future time steps, and obtain multiple predicted displacements. Input the weighted-fused spatio-temporal feature matrix into the temporal convolutional network to output the displacements at multiple future time steps. It can be understood that the temporal convolutional network can efficiently capture long-term dependencies and is suitable for the requirement of the present invention to predict the displacement of monitoring points in real time, which can improve the real-time performance of our monitoring of seismic faults.

[0056] Embodiment 3:

[0057] The difference between this embodiment and Embodiment 1 or 2 is only that the step S4 includes:

[0058] S41. Generate a meshed fault surface based on the preset grid parameters and the fault geometry model, and calculate the sliding amount of each grid point on the meshed fault surface at each time step based on the sliding amount of each monitoring point at each time step, to obtain multiple grid point sliding amounts. It can be understood that first, grid parameters are defined, including strike length, dip length, and grid resolution, and then grid division is performed along the strike and dip of the fault, and the coordinates of each grid point are generated, so as to discretize the fault surface into regular grids, which is convenient for subsequent sliding amount interpolation and locked section identification.

[0059] Specifically, the step of calculating the sliding amount of each grid point on the meshed fault surface at each time step based on the sliding amount of each monitoring point at each time step includes:

[0060] S411. Obtain the geographical coordinates of each monitoring point, and convert the geographical coordinates of each monitoring point into first coordinates in the local coordinate system to obtain multiple first coordinates;

[0061] S412. Calculate the grid point sliding amount of each grid point at each time step based on the Kriging interpolation method, the first coordinates, and the sliding amount of each monitoring point at each time step, to obtain multiple grid point sliding amounts.

[0062] It can be understood that after obtaining the meshed fault surface, the local coordinates and sliding amounts of each monitoring point, the fault sliding amount distribution matrix S map is generated by Kriging interpolation, and the calculation formula is:

[0063]

[0064] Among them, S map (m,n) represents the grid point sliding amount of the grid point with coordinates (m,n), N is the number of monitoring points, and Δs i is the sliding amount of the i-th monitoring point, and ω i is the Kriging weight corresponding to the i-th monitoring point.

[0065] S42. Identify the fault locked segments based on the grid point sliding amounts of each grid point at each time step, and obtain multiple fault locked segments at each time step. It can be understood that the locked segments represent areas where sliding has not occurred for a long time, and the stress in these areas gradually accumulates and may eventually trigger violent sliding. Therefore, by dividing the grid and calculating the grid point sliding amounts of each grid point, the locked segments can be accurately identified based on the grid point sliding amounts of the grid points in each area, the prediction of the locked segment area and activity can be completed, thereby effectively monitoring the activity of the seismic fault and conducting risk assessment and early warning in advance.

[0066] Specifically, step S42 includes:

[0067] S421. Select the locked points at each time step based on a preset locked threshold and the grid point sliding amounts of each grid point at each time step, and obtain multiple locked points. It can be understood that the locked threshold set in this example is 20% of the long-term average sliding amount of the seismic fault. Through binary marking, the values of the points in the sliding amount distribution matrix that exceed the locked threshold are set to 0, and the values of the points below the locked threshold are set to 1, thereby obtaining a binary matrix for marking the locked points in the grid.

[0068] S422. Cluster each of the locked points at each time step based on a preset clustering algorithm, and merge the adjacent locked points after clustering at each time step to obtain multiple fault locked segments at each time step. It can be understood that in this embodiment, the neighborhood radius is first set to define the maximum distance between adjacent points, then the minimum number of grid points forming a locked segment is set, and then the DBSCAN clustering algorithm is used to merge the locked points in the binary matrix obtained in step S421 to obtain multiple locked segments. Each obtained locked segment includes the coordinate range of the grid points and the locked segment area, where the locked segment area is the product of the number of grid points and the resolution.

[0069] S43. Calculate the rupture probability of each of the fault locked segments based on the stress state of each of the fault locked segments, and obtain the rupture probability of each of the fault locked segments at each time step. It can be understood that as time goes by, the plate movement and crustal deformation in the area around the locked segment will continuously exert pressure on the locked segment, causing the stress in these areas to gradually increase. Once the stress accumulates to the critical point, the locked segment may rupture and release energy, causing an earthquake. Therefore, evaluating the rupture probability of the locked segment helps to understand the likelihood of an earthquake occurring in these areas in the future, making the monitoring results of the seismic fault more intuitive and accurate.

[0070] Specifically, step S43 includes:

[0071] S431. Obtain the second information and the third information, where the second information includes the friction coefficient, rock elastic modulus, Poisson's ratio, and rock density of the seismic fault, and the third information includes the average recurrence period of historical earthquakes corresponding to the seismic fault and the time of the most recent earthquake;

[0072] S432. Calculate the average sliding amount of each of the fault locked segments based on the grid point sliding amount of each of the locking points of each of the fault locked segments;

[0073] S433. Calculate the Coulomb rupture potential of each of the fault locked segments based on the average sliding amount of each of the fault locked segments and the second information. It can be understood that to calculate the Coulomb rupture potential of the locked segment, it is necessary to first calculate the shear stress and normal stress of the locked segment. Among them, the calculation formula for shear stress is:

[0074]

[0075] where τ k is the shear stress of the k-th locked segment, E is the elastic modulus, v is Poisson's ratio, D is the locking depth, and Δs avg,k is the average sliding amount of the k-th locked segment.

[0076] The calculation formula for normal stress is:

[0077] σ k = ρgh·sinφ;

[0078] where σ k is the normal stress of the k-th locked segment, ρ is the rock density, g is the acceleration due to gravity, h is the fault burial depth, and φ is the fault dip angle.

[0079] After calculating the shear stress and normal stress, calculate the Coulomb rupture potential according to the Coulomb rupture criterion. The calculation formula is:

[0080] CFF k = τ k - μσk ;

[0081] Among them, CFF k is the Coulomb rupture potential of the k-th locked segment, τ k is the shear stress of the k-th locked segment, σ k is the normal stress of the k locked segments, and μ is the friction coefficient.

[0082] S434. Calculate the rupture probability of each fault locked segment based on the Brownian process time model, the third information, and each of the Coulomb rupture potentials, obtaining a plurality of the rupture probabilities. It can be understood that when calculating the rupture probability of the locked segment, historical earthquake data needs to be combined, and the calculation formula is:

[0083]

[0084] Among them, P k is the rupture probability of the k-th locked segment, T is the average recurrence period of historical earthquakes, t is the time interval between the current time and the last earthquake, CFF k is the Coulomb rupture potential of the k-th locked segment, CFF crit critical Coulomb stress.

[0085] S44. Conduct risk warning for the seismic fault based on each fault locked segment and each rupture probability at each time step, obtaining the risk warning result.

[0086] Specifically, step S44 includes:

[0087] S441. Obtain the fourth information, where the fourth information includes the area of each fault locked segment at each time step;

[0088] S442. Calculate the spatial overlapping area of each fault locked segment within consecutive time steps based on the fourth information, obtaining a plurality of the spatial overlapping areas;

[0089] S443. Generate the first risk warning information based on each of the spatial overlapping areas, obtaining the first risk warning information. It can be understood that if the overlapping area of a certain locked segment reaches 80% within multiple time steps, it indicates that the area corresponding to this locked segment has been in a state of low sliding amount for a long time, which will generate a large amount of stress accumulation and has a greater risk. Therefore, the present invention judges the survival duration of the locked segment through the overlapping area of the locked segment and conducts risk warning for the locked segment according to the survival duration.

[0090] Further, the possible situations for each locked section include continuous locking, intermittent locking, and newly emerging locking. Continuous locking means that a certain area is identified as a locked section in all predicted time steps, with a high risk. Intermittent locking means that a certain area only meets the locking conditions in some time steps and requires continuous attention and auxiliary judgment. Newly emerging locking means that it suddenly becomes a locked section in the later stage of the predicted time step, and it is necessary to extend the predicted time step or comprehensively consider other information related to this locked section for discrimination.

[0091] S444. Generate second risk warning information based on the rupture probability of each of the fault locked sections at each time step to obtain the second risk warning information. It can be understood that the rupture probability directly reflects the risk level of the existence of the locked section. Therefore, in this embodiment, the warning levels and corresponding thresholds are divided. The warning levels include yellow warning, orange warning, and red warning. Among them, if the rupture probability exceeds 0.6, it indicates that this locked section has a medium risk, issue a yellow warning, and continuously monitor the locked section. If the rupture probability exceeds 0.8, it indicates that this locked section has a high risk, issue an orange warning, and verify it in combination with other relevant data. If the rupture probability exceeds 0.9, it indicates that this locked section has an extremely high risk, issue a red warning, and immediately respond to emergency measures.

[0092] S445. Generate third risk warning information based on the change amount of the rupture probability of each of the fault locked sections at each time step to obtain the third risk warning information. It can be understood that the change trend of the rupture probability of the locked section is analyzed through the change amount of the rupture probability of each of the fault locked sections at each time step. If the change trend is an upward trend, more detailed investigation and analysis need to be carried out on this area. If the change trend is a downward trend, the monitoring of this area can be simplified, and more work focus can be allocated to high-risk areas.

[0093] Embodiment 4:

[0094] As Figure 2 shown, this embodiment provides a Beidou-based seismic fault monitoring system, and the system includes an acquisition module 901, a first processing module 902, a second processing module 903, and a third processing module 904:

[0095] The acquisition module 901 is used to acquire Beidou satellite data of multiple monitoring points of the seismic fault within a preset sampling time;

[0096] The first processing module 902 is used to preprocess each of the Beidou satellite data and input each of the preprocessed Beidou satellite data into a preset prediction model for displacement prediction to obtain a plurality of predicted displacements. One predicted displacement includes the displacements of one monitoring point at multiple future time steps;

[0097] A second processing module 903, configured to calculate the sliding amount of each of the monitoring points at each future time step based on each of the predicted displacement amounts at each time step, to obtain a plurality of the sliding amounts;

[0098] A third processing module 904, configured to perform risk early warning of a seismic fault based on the change in the sliding amount of each of the monitoring points at each future time step, to obtain a risk early warning result.

[0099] It should be noted that, regarding the system in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0100] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

[0101] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present invention, and all of them should be covered by the protection scope of the present invention.

Claims

1. A Beidou-based seismic fault monitoring method, characterized in that, Including: Obtaining Beidou satellite data of multiple monitoring points of a seismic fault within a preset sampling time; Preprocessing each of the Beidou satellite data, and inputting each preprocessed Beidou satellite data into a preset prediction model for displacement prediction to obtain multiple predicted displacements, where one predicted displacement includes the displacements of one monitoring point at multiple future time steps; Calculating the slip amount of each monitoring point at each future time step based on each of the predicted displacements at each time step to obtain multiple slip amounts; Performing risk warning on the seismic fault based on the changes in the slip amounts of each monitoring point at each future time step to obtain a risk warning result.

2. The method for monitoring earthquake faults based on Beidou according to claim 1, wherein , The preprocessing of each of the Beidou satellite data includes: Performing cycle slip detection and repair on the carrier phase data of each of the Beidou satellite data, and performing error correction on each Beidou satellite data after carrier phase repair to obtain multiple Beidou satellite data after error correction; Performing position solution on each of the Beidou satellite data after error correction, and generating a surface deformation time series for each monitoring point according to the position solution result to obtain multiple surface deformation time series, where one surface deformation time series includes the position, velocity, and acceleration of one monitoring point at each time step during the sampling time; Generating a spatio-temporal matrix based on multiple surface deformation time series to obtain the spatio-temporal matrix.

3. The method for monitoring earthquake faults based on Beidou according to claim 2, characterized in that , The inputting of each preprocessed Beidou satellite data into a preset prediction model for displacement prediction includes: Extracting local features of the spatio-temporal matrix based on a preset one-dimensional convolutional neural network to obtain a first output feature; Extracting the long-range time dependence of the first output feature based on a preset bidirectional long short-term memory network to obtain a time feature matrix; Generating a spatial adjacency matrix based on the positions of the monitoring points, and inputting the spatial adjacency matrix and the spatio-temporal matrix into a preset graph convolutional network for feature aggregation to obtain a spatial feature matrix; Performing weighted fusion of the time feature matrix and the spatial feature matrix based on an attention mechanism to obtain a spatio-temporal feature matrix; Inputting the spatio-temporal feature matrix into a preset temporal convolutional network for displacement prediction of each monitoring point at multiple future time steps to obtain multiple predicted displacements.

4. The method for monitoring earthquake faults based on Beidou according to claim 1, characterized in that , The calculation of the slip amount of each monitoring point at each future time step based on each of the predicted displacements at each time step includes: Obtaining first information, where the first information includes the strike angle, dip angle, and locking depth of the seismic fault; Constructing a local coordinate system with a preset reference point as the origin to obtain the local coordinate system; Establishing a fault geometry model based on the first information and the local coordinate system to obtain the fault geometry model; Calculating the slip amount based on the fault geometry model and the displacements of each monitoring point at each future time step to obtain the slip amount of each monitoring point at each future time step.

5. A Beidou-based earthquake fault monitoring method according to claim 4, characterized in that , Performing risk warning on the seismic fault based on the changes in the slip amounts of each monitoring point at each future time step includes: Generate a meshed fault surface based on preset grid parameters and the fault geometry model, and calculate the displacement of each grid point on the meshed fault surface at each time step based on each displacement at each time step, obtaining displacements of multiple grid points; Identify fault locked segments based on the displacements of each grid point at each time step, obtaining multiple fault locked segments at each time step; Calculate the rupture probability of each fault locked segment based on the stress state of each fault locked segment, obtaining the rupture probability of each fault locked segment at each time step; Conduct risk warning of seismic faults based on each fault locked segment and each rupture probability at each time step, obtaining the risk warning result.

6. The method for monitoring earthquake faults based on Beidou according to claim 5, wherein , The calculating the displacement of each grid point on the meshed fault surface at each time step based on each displacement at each time step includes: Obtain the geographical coordinates of each monitoring point, and convert the geographical coordinates of each monitoring point into first coordinates in a local coordinate system, obtaining multiple first coordinates; Calculate the displacement of each grid point at each time step based on Kriging interpolation method, the first coordinates and the displacements of each monitoring point at each time step, obtaining displacements of multiple grid points.

7. The method for monitoring earthquake faults based on Beidou according to claim 6, characterized in that , The identifying fault locked segments based on the displacements of each monitoring point at each future time step includes: Select locked points at each time step based on a preset locking threshold and the displacements of each grid point at each time step, obtaining multiple locked points; Cluster each locked point at each time step based on a preset clustering algorithm, and merge adjacent locked points after clustering at each time step, obtaining multiple fault locked segments at each time step.

8. A Beidou-based seismic fault monitoring method according to claim 6, characterized in that , The calculating the rupture probability of each fault locked segment based on the stress state of each fault locked segment includes: Obtain second information and third information, where the second information includes the friction coefficient, rock elastic modulus, Poisson's ratio and rock density of the seismic fault, and the third information includes the average recurrence period of historical earthquakes corresponding to the seismic fault and the time of the most recent earthquake; Calculate the average displacement of each fault locked segment based on the displacements of each locked point of each fault locked segment; Calculate the Coulomb rupture potential of each fault locked segment based on the average displacement of each fault locked segment and the second information; Calculate the rupture probability of each fault locked segment based on the Brownian process time model, the third information and each Coulomb rupture potential, obtaining multiple rupture probabilities.

9. A Beidou-based seismic fault monitoring method according to claim 6, characterized in that , The conducting risk warning of seismic faults based on each fault locked segment and each rupture probability at each time step includes: Obtain fourth information, where the fourth information includes the area of each fault locked segment at each time step; Calculate the spatial overlapping area of each fault locked segment within consecutive time steps based on the fourth information, obtaining multiple spatial overlapping areas; Generate a first risk warning message based on each of the spatial overlap areas to obtain the first risk warning message; Generate a second risk warning message based on the rupture probability of each of the fault locking segments at each time step to obtain the second risk warning message; Generate a third risk warning message based on the change amount of the rupture probability of each of the fault locking segments at each time step to obtain the third risk warning message.

10. A Beidou-based seismic fault monitoring system, characterized in that, Comprising: An acquisition module, configured to acquire Beidou satellite data of multiple monitoring points of a seismic fault within a preset sampling time; A first processing module, configured to preprocess each of the Beidou satellite data, and input each of the preprocessed Beidou satellite data into a preset prediction model for displacement prediction to obtain a plurality of predicted displacements, where one predicted displacement includes the displacements of one of the monitoring points at multiple future time steps; A second processing module, configured to calculate the sliding amount of each of the monitoring points at each future time step based on each of the predicted displacements at each time step to obtain a plurality of the sliding amounts; A third processing module, configured to perform risk warning on the seismic fault based on the change of the sliding amount of each of the monitoring points at each future time step to obtain a risk warning result.

Images