Accumulated absolute velocity prediction method, device, equipment, medium and program product

Abstract

The invention provides a cumulative absolute velocity prediction method, device and equipment, a medium and a program product. The method comprises the following steps: constructing an earthquake monitoring station group; the earthquake monitoring station group comprises a target monitoring station and peripheral stations; acquiring node features of each station in the earthquake monitoring station group, and acquiring edge features between the target monitoring station and the peripheral stations; constructing a historical station group feature map based on the node features and the edge features; the historical station group feature map is used for training a map neural network; and inputting the station group feature graph corresponding to the to-be-predicted station into the trained graph neural network to obtain a cumulative absolute speed prediction result of the to-be-predicted station. According to the method, the cumulative absolute velocity data of the region is predicted through the trained graph neural network, the historical seismic data and the actual measurement data of the surrounding stations are comprehensively analyzed, and the prediction accuracy of the cumulative absolute velocity is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of seismic data prediction, and in particular to a cumulative absolute velocity prediction method, device, equipment, medium and program product. Background Art

[0002] Cumulative Absolute Velocity (CAV) is important data for evaluating earthquakes. Existing prediction of CAV mainly combines interpolation methods and some earthquake prediction algorithms, making it difficult to comprehensively consider historical earthquake record data and measured data from seismic monitoring stations, resulting in inaccurate prediction of existing cumulative absolute velocity. Summary of the Invention

[0003] The present invention provides a cumulative absolute velocity prediction method, device, equipment, medium and program product to solve the defect of inaccurate prediction of existing cumulative absolute velocity and improve the prediction accuracy of cumulative absolute velocity.

[0004] The present invention provides a cumulative absolute velocity prediction method, including the following steps: Construct a seismic monitoring station group; the seismic monitoring station group includes a target monitoring station and surrounding stations; Obtain the node features of each station in the seismic monitoring station group, and obtain the edge features between the target monitoring station and the surrounding stations; Construct a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train a graph neural network; Input the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

[0005] According to the cumulative absolute velocity prediction method provided by the present invention, the construction of the seismic monitoring station group includes: Obtain the types and historical record numbers of historical seismic monitoring stations, and the distances between any two historical seismic monitoring stations; Determine the target monitoring station and surrounding stations among the historical seismic monitoring stations; the surrounding stations are the stations around the target monitoring station; the types of the target monitoring station and the surrounding stations are surface stations; the historical record numbers of the target monitoring station and the surrounding stations are greater than a first threshold; the distances between the target monitoring station and the surrounding stations are less than a second threshold; the number of surrounding stations is greater than a third threshold; Construct a seismic monitoring station group based on the target monitoring station and the surrounding stations.

[0006] According to a cumulative absolute velocity prediction method provided by the present invention, obtaining the node features of each station in the seismic monitoring station group includes: Obtaining the first node features of the target monitoring station; the first node features include the magnitude, epicentral distance, soil shear wave velocity, and focal depth in the historical records of the target monitoring station; Obtaining the second node features of the surrounding stations; the second node features include the soil shear wave velocity, epicentral distance, and cumulative absolute velocity in the historical records of the surrounding stations.

[0007] According to a cumulative absolute velocity prediction method provided by the present invention, obtaining the edge features between the target monitoring station and the surrounding stations includes: Based on the difference in site conditions, determining the first edge features between the target monitoring station and the surrounding stations; the difference in site conditions is determined based on the absolute difference in soil shear wave velocity between the target monitoring station and the surrounding stations; Based on the spatial relationship, determining the second edge features between the target monitoring station and the surrounding stations; the spatial relationship is determined based on the absolute difference in epicentral distance between the target monitoring station and the surrounding stations; Based on the distance between the target monitoring station and the surrounding stations, determining the third edge features between the target monitoring station and the surrounding stations.

[0008] According to a cumulative absolute velocity prediction method provided by the present invention, constructing a historical station group feature map based on the node features and the edge features includes: Transforming the difference in site conditions and the spatial relationship to obtain non-linear features; Based on the non-linear features and the third edge features, determining the connection relationship between the target monitoring station and the surrounding stations; Based on the connection relationship and the target topological structure, constructing a historical station group feature map; in the target topological structure, the target monitoring station is connected to the surrounding stations, and no two surrounding stations are connected to each other.

[0009] According to a cumulative absolute velocity prediction method provided by the present invention, the graph neural network includes a dynamic graph attention layer, a global average pooling layer, and a fully connected layer; the cumulative absolute velocity prediction method further includes: Inputting the historical station group feature map into the dynamic graph attention layer to obtain the relationship between nodes and the adjusted edge feature weights; Inputting the relationship between nodes and the adjusted edge feature weights into the global average pooling layer to obtain a global feature vector; Input the global feature vector into the fully connected layer to obtain an initial prediction result; Train the graph neural network based on the initial prediction result; the training process includes network architecture adjustment and activation function adjustment.

[0010] The present invention also provides an accumulated absolute velocity prediction device, including the following modules: An earthquake monitoring station group construction module, configured to construct an earthquake monitoring station group; the earthquake monitoring station group includes a target monitoring station and surrounding stations; A feature acquisition module, configured to acquire the node features of each station in the earthquake monitoring station group, and acquire the edge features between the target monitoring station and the surrounding stations; A historical station group feature map construction module, configured to construct a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train the graph neural network; An accumulated absolute velocity prediction module, configured to input the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the accumulated absolute velocity prediction result of the station to be predicted.

[0011] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and running on the processor, where when the processor executes the computer program, it implements the accumulated absolute velocity prediction method as described in any one of the above.

[0012] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the accumulated absolute velocity prediction method as described in any one of the above.

[0013] The present invention also provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the accumulated absolute velocity prediction method as described in any one of the above.

[0014] The cumulative absolute velocity prediction method, device, equipment, medium and program product provided by the present invention constructs a seismic monitoring station group including a target monitoring station and surrounding stations, extracts the node features of each monitoring station and the edge features between the target monitoring station and the surrounding stations through the historical monitoring data recorded by the seismic monitoring stations; then constructs a historical station feature map through the node features and edge features, and the constructed station feature map is used to train a graph neural network with a structure suitable for analyzing the station feature map; finally, the station group feature map corresponding to the station to be predicted in the area lacking the cumulative absolute velocity is input into the trained graph neural network to obtain the prediction result of the cumulative absolute velocity of the station to be predicted. The present application predicts the cumulative absolute velocity data of the area through the trained graph neural network, comprehensively analyzes the historical seismic data and the measured data of the surrounding stations, and improves the prediction accuracy of the cumulative absolute velocity. Description of the Drawings

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0016] Figure 1 It is one of the flow diagrams of the cumulative absolute velocity prediction method provided by the present invention.

[0017] Figure 2 It is the second flow diagram of the cumulative absolute velocity prediction method provided by the present invention.

[0018] Figure 3 It is a schematic diagram of the station group feature map provided by the present invention.

[0019] Figure 4 It is a schematic diagram of the structure of the cumulative absolute velocity prediction device provided by the present invention.

[0020] Figure 5 It is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Embodiments

[0021] To make the objectives, technical solutions and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments in the present invention belong to the scope of protection of the present invention.

[0022] The following will be combined with Figures 1-5Describe the cumulative absolute velocity prediction method, device, equipment, medium and program product of the present invention.

[0023] Figure 1 It is one of the flow schematic diagrams of the cumulative absolute velocity prediction method provided by the present invention. As Figure 1 shown, the method includes the following: Step 100: Construct a seismic monitoring station group; the seismic monitoring station group includes a target monitoring station and surrounding stations; Specifically, the main content of the cumulative absolute velocity prediction method provided by the present invention includes the following steps: Step 1: Select a target monitoring station from a large number of historical seismic monitoring stations. This station is set in an area where CAV monitoring data is missing, but the target monitoring station has historical CAV data available for training. Select multiple stations with measured data around the target monitoring station (i.e., the surrounding stations in this embodiment) to jointly form a station group (i.e., the seismic monitoring station group in this embodiment). Then, based on the historical earthquake database, establish a sample database containing multiple groups of seismic monitoring stations for training the graph neural network.

[0024] Step 200: Obtain the node features of each station in the seismic monitoring station group, and obtain the edge features between the target monitoring station and the surrounding stations; Specifically, the main content of the cumulative absolute velocity prediction method provided by the present invention also includes the following steps: Step 2: Encode the physical characteristics of the constructed seismic monitoring station group to obtain the node features of each monitoring station and the edge features between the target monitoring station and the surrounding stations, and construct a feature map of the seismic monitoring station group based on the node features and edge features to capture the mutual relationships and characteristics between each seismic monitoring station.

[0025] Step 300: Construct a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train the graph neural network; Specifically, the main content of the cumulative absolute velocity prediction method provided by the present invention also includes the following steps: Step 3: Use the station group feature map of the seismic monitoring station group as the input of the graph neural network, and the CAV at the target monitoring station as the output to construct a deep learning model based on the graph neural network (Graph Neural Network, GNN). Through the constructed multiple historical station group feature maps for training, this deep learning model can predict the CAV at the target location (i.e., the location of the station to be predicted).

[0026] Step 400: Input the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

[0027] Specifically, the main content of the cumulative absolute velocity prediction method provided by the present invention further includes the following steps: Step 4: After constructing and training the model for predicting the CAV at the target location, input the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the CAV at the station to be predicted output by the trained graph neural network.

[0028] In this embodiment, a seismic monitoring station group including the target monitoring station and surrounding stations is constructed. Through the historical monitoring data recorded by the seismic monitoring stations, the node features of each monitoring station and the edge features between the target monitoring station and the surrounding stations are extracted; then, a historical station feature map is constructed based on the node features and edge features, and the constructed station feature map is used to train a graph neural network with a structure suitable for analyzing the station feature map; finally, the station group feature map corresponding to the station to be predicted in the area lacking the cumulative absolute velocity is input into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted. This application predicts the cumulative absolute velocity data of the area through the trained graph neural network, comprehensively analyzes the historical seismic data and the measured data of the surrounding stations, and improves the prediction accuracy of the cumulative absolute velocity.

[0029] Figure 2 is the second flowchart of the cumulative absolute velocity prediction method provided by the present invention. As Figure 2 shown, the method may further include: Step 110: Obtain the types and historical record quantities of historical seismic monitoring stations, and the distances between any two historical seismic monitoring stations. Step 120: Determine the target monitoring station and surrounding stations among the historical seismic monitoring stations; the surrounding stations are the stations around the target monitoring station; the types of the target monitoring station and the surrounding stations are surface stations; the historical record quantities of the target monitoring station and the surrounding stations are greater than the first threshold; the distances between the target monitoring station and the surrounding stations are less than the second threshold; the number of surrounding stations is greater than the third threshold. Step 130: Based on the target monitoring station and the surrounding stations, construct a seismic monitoring station group.

[0030] Specifically, when selecting the target monitoring station and its surrounding stations from historical earthquake monitoring stations, the screening conditions mainly include the type of the station, the historical record data of the station, and the distance between stations. For example, the types of the target monitoring station and its surrounding stations are surface stations; the historical record data of the target monitoring station and its surrounding stations should be more than a certain number (for example, 20, which is the first threshold in this embodiment); the number of surrounding stations less than 40 kilometers (i.e., the second threshold in this embodiment) away from the selected target monitoring station should be greater than 3 (i.e., the third threshold in this embodiment).

[0031] As Figure 3 shown, taking the number of surrounding stations of the target monitoring station as 4 as an example, when the number of surrounding stations less than the second threshold away from the target monitoring station is greater than 4, 4 surrounding stations of the target monitoring station can be selected in the order from near to far. Each constructed earthquake monitoring station group includes one target monitoring station and multiple surrounding stations.

[0032] In this embodiment, earthquake monitoring station groups are constructed through the data information of each historical earthquake monitoring station, so that the model trained based on the earthquake monitoring station groups can comprehensively learn the relationships and characteristics between stations.

[0033] In one embodiment, the cumulative absolute velocity prediction method provided by the embodiments of the present invention may further include: Step 210, obtaining the first node feature of the target monitoring station; the first node feature includes the magnitude, epicentral distance, soil shear wave velocity, and focal depth in the historical record of the target monitoring station; Step 220, obtaining the second node feature of the surrounding stations; the second node feature includes the soil shear wave velocity, epicentral distance, and cumulative absolute velocity in the historical record of the surrounding stations.

[0034] Specifically, the physical characteristics of the earthquake monitoring station group are encoded, and the encoded characteristics are designed as the graph features of the earthquake monitoring station group, specifically including: the node features, edge features, and topological structure of the graph of the earthquake monitoring station group are composed into graph features that can be input into the graph neural network; the node features are used to describe the attributes of the earthquake monitoring stations, where the node features of the target monitoring station are composed of attributes such as magnitude, epicentral distance, soil shear wave velocity, and focal depth, and the node features of multiple surrounding stations are composed of soil shear wave velocity, epicentral distance, and cumulative absolute velocity.

[0035] In this embodiment, the node features for constructing the station group feature graph are obtained by encoding the physical characteristics of the earthquake monitoring station group.

[0036] In one embodiment, the cumulative absolute velocity prediction method provided by the embodiments of the present invention may further include: Step 230: Determine the first edge feature between the target monitoring station and the surrounding stations based on the site condition differences; the site condition differences are determined based on the absolute difference in soil shear wave velocity between the target monitoring station and the surrounding stations; Step 240: Determine the second edge feature between the target monitoring station and the surrounding stations based on the spatial relationship; the spatial relationship is determined based on the absolute difference in epicentral distance between the target monitoring station and the surrounding stations; Step 250: Determine the third edge feature between the target monitoring station and the surrounding stations based on the distance between the target monitoring station and the surrounding stations.

[0037] Specifically, the edge feature is mainly used to quantify the relationship between the target monitoring station and the surrounding stations and is constructed based on multiple types of physical property differences. One of the physical property differences is the absolute difference in soil shear wave velocity, which represents the site condition differences between the target monitoring station and the surrounding stations; the second physical property difference is the absolute difference in epicentral distance, which reflects the spatial relationship between the stations; the third physical property difference is the distance between the surrounding stations and the target monitoring station.

[0038] In this embodiment, by quantifying the relationship between the target monitoring station and the surrounding stations, the edge feature for constructing the station group feature map is obtained.

[0039] In one embodiment, the cumulative absolute velocity prediction method provided by the embodiments of the present invention may further include: Step 310: Transform the site condition differences and the spatial relationship to obtain non-linear features; Step 320: Determine the connection relationship between the target monitoring station and the surrounding stations based on the non-linear features and the third edge feature; Step 330: Construct a historical station group feature map based on the connection relationship and the target topological structure; in the target topological structure, the target monitoring station is connected to the surrounding stations, and no connection exists between any two surrounding stations.

[0040] Specifically, in order to enhance the modeling ability of non-linear relationships, logarithmic transformation and square root transformation are performed on the above-mentioned difference features to generate extended non-linear features, that is, the non-linear features and the third edge feature in this embodiment. As Figure 3 shown, in the topological structure design of the historical station group feature map, a star topology structure can be adopted, where the target monitoring station is used as the central node and multiple surrounding stations are used as peripheral nodes. The connection method of the graph adopts bidirectional connection, that is, the central node is interconnected with each peripheral node. This star topology structure simplifies the station group feature map while considering the local relationship of seismic monitoring stations, reducing the structural complexity of the station group feature map.

[0041] In this embodiment, by constructing graph features and graph topologies, the structural complexity of the station group feature map is reduced.

[0042] In one embodiment, the cumulative absolute velocity prediction method provided by the embodiments of the present invention may further include: Step 410: Input the historical station group feature map into the dynamic graph attention layer to obtain the relationship between nodes and the adjusted edge feature weights; Step 420: Input the relationship between nodes and the adjusted edge feature weights into the global average pooling layer to obtain a global feature vector; Step 430: Input the global feature vector into the fully connected layer to obtain an initial prediction result; Step 440: Train the graph neural network based on the initial prediction result; the training process includes network architecture adjustment and activation function adjustment.

[0043] Specifically, the steps of constructing a deep learning network based on a graph neural network (i.e., the graph neural network in this embodiment) include: Design the architecture of the deep learning network based on GNN. The first and second layers of the network can adopt a graph attention network based on the multi-head attention mechanism to dynamically adjust the edge feature weights and capture the complex relationships between nodes; the third layer adopts graph convolution based on Transformer (encoder-decoder) to further improve the ability to model global dependencies. After multi-layer feature extraction, the network aggregates all node features into a fixed-length global feature vector through a global average pooling (Global Mean Pooling, GMP) operation. Finally, a two-layer fully connected network is adopted.

[0044] The input of the network is the station group feature map obtained above, and the output of the network is the CAV at the target monitoring station; by adjusting the network architecture, activation function and optimization method, a network architecture suitable for predicting the CAV at the target monitoring station is constructed.

[0045] Specifically, aiming at the characteristics of CAV prediction at the target monitoring station, a network architecture and optimization strategy are designed. Combining the distribution characteristics of the CAV data of the monitoring station, the key parameters of the graph neural network are adjusted; based on the complexity of the station group feature map, the network depth and activation function type of the graph neural network are optimized; based on the prediction accuracy and physical rationality of CAV at the target monitoring station, a network hyperparameter combination suitable for predicting CAV at the target monitoring station is determined. During the training process of the graph neural network, by introducing specific training criteria that adapt to the CAV prediction requirements at the target monitoring station, the error of the network prediction is optimized; finally, the verification effect of the trained model on the verification data set is determined, and the model can accurately predict the CAV at the target monitoring station location.

[0046] In this embodiment, by constructing a GNN-based network architecture and training a model for predicting CAV at the station to be predicted, the CAV prediction accuracy in the area lacking monitoring data is improved for the station to be predicted lacking CAV data.

[0047] Next, the cumulative absolute velocity prediction device provided by the present invention will be described. The cumulative absolute velocity prediction device described below can be mutually referred to with the cumulative absolute velocity prediction method described above.

[0048] Please refer to Figure 4 , the present invention also provides a cumulative absolute velocity prediction device, including: An earthquake monitoring station group construction module 401 for constructing an earthquake monitoring station group; the earthquake monitoring station group includes a target monitoring station and surrounding stations; A feature acquisition module 402 for acquiring the node features of each station in the earthquake monitoring station group and acquiring the edge features between the target monitoring station and the surrounding stations; A historical station group feature map construction module 403 for constructing a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train the graph neural network; A cumulative absolute velocity prediction module 404 for inputting the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

[0049] Optionally, the earthquake monitoring station group construction module includes: An acquisition unit for acquiring the types and historical record numbers of historical earthquake monitoring stations, and the distances between any two historical earthquake monitoring stations; A station determination unit for determining a target monitoring station and surrounding stations among historical earthquake monitoring stations; the surrounding stations are stations around the target monitoring station; the types of the target monitoring station and the surrounding stations are surface stations; the historical record quantities of the target monitoring station and the surrounding stations are greater than a first threshold; the distances between the target monitoring station and the surrounding stations are less than a second threshold; the quantity of the surrounding stations is greater than a third threshold. An earthquake monitoring station group construction unit for constructing an earthquake monitoring station group based on the target monitoring station and the surrounding stations.

[0050] Optionally, the feature acquisition module includes: A first node feature acquisition unit for acquiring a first node feature of the target monitoring station; the first node feature includes the magnitude, epicentral distance, soil shear wave velocity, and focal depth in the historical record of the target monitoring station. A second node feature acquisition unit for acquiring a second node feature of the surrounding stations; the second node feature includes the soil shear wave velocity, epicentral distance, and cumulative absolute velocity in the historical record of the surrounding stations.

[0051] Optionally, the feature acquisition module further includes: A first edge feature acquisition unit for determining a first edge feature between the target monitoring station and the surrounding stations based on the difference in site conditions; the difference in site conditions is determined based on the absolute difference in soil shear wave velocity between the target monitoring station and the surrounding stations. A second edge feature determination unit for determining a second edge feature between the target monitoring station and the surrounding stations based on the spatial relationship; the spatial relationship is determined based on the absolute difference in epicentral distance between the target monitoring station and the surrounding stations. A third edge feature determination unit for determining a third edge feature between the target monitoring station and the surrounding stations based on the distance between the target monitoring station and the surrounding stations.

[0052] Optionally, the historical station group feature map construction module includes: A non - linear feature determination unit for transforming the difference in site conditions and the spatial relationship to obtain non - linear features. A connection relationship determination unit for determining the connection relationship between the target monitoring station and the surrounding stations based on the non - linear features and the third edge feature. A historical station group feature map construction unit for constructing a historical station group feature map based on the connection relationship and the target topological structure; in the target topological structure, the target monitoring station is connected to the surrounding stations, and no two surrounding stations are connected to each other.

[0053] Optionally, the cumulative absolute velocity prediction device further includes: An edge feature weight adjustment module, configured to input the historical station group feature map into the dynamic graph attention layer to obtain the relationship between nodes and the adjusted edge feature weights; A global average pooling module, configured to input the relationship between nodes and the adjusted edge feature weights into the global average pooling layer to obtain a global feature vector; An initial prediction result determination module, configured to input the global feature vector into the fully connected layer to obtain an initial prediction result; A graph neural network training module, configured to train the graph neural network based on the initial prediction result; the training process includes network architecture adjustment and activation function adjustment.

[0054] Figure 5 An example of the physical structure diagram of an electronic device is shown in Figure 5 As shown, the electronic device may include: a processor 510, a communication interface 520, a memory 530, and a communication bus 540. Among them, the processor 510, the communication interface 520, and the memory 530 complete communication with each other through the communication bus 540. The processor 510 can call the logical instructions in the memory 530 to execute the cumulative absolute velocity prediction method, which includes: constructing a seismic monitoring station group; the seismic monitoring station group includes a target monitoring station and surrounding stations; obtaining the node features of each station in the seismic monitoring station group, and obtaining the edge features between the target monitoring station and the surrounding stations; constructing a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train a graph neural network; inputting the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

[0055] In addition, when the logical instructions in the above-mentioned memory 530 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0056] On the other hand, the present invention also provides a computer program product. The computer program product includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the cumulative absolute velocity prediction method provided by the above-mentioned various methods. The method includes: constructing a seismic monitoring station group; the seismic monitoring station group includes a target monitoring station and surrounding stations; obtaining the node features of each station in the seismic monitoring station group, and obtaining the edge features between the target monitoring station and the surrounding stations; constructing a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train a graph neural network; inputting the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

[0057] In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it realizes the cumulative absolute velocity prediction method provided by the above-mentioned various methods. The method includes: constructing a seismic monitoring station group; the seismic monitoring station group includes a target monitoring station and surrounding stations; obtaining the node features of each station in the seismic monitoring station group, and obtaining the edge features between the target monitoring station and the surrounding stations; constructing a historical station group feature map based on the node features and the edge features; the historical station group feature map is used to train a graph neural network; inputting the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

[0058] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. A person of ordinary skill in the art can understand and implement it without creative labor.

[0059] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments or equivalently replace some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cumulative absolute velocity prediction method, characterized in that: include: Constructing an earthquake monitoring station group; the earthquake monitoring station group includes a target monitoring station and surrounding stations; Obtaining node features of each station in the seismic monitoring station group, and obtaining edge features between the target monitoring station and the surrounding stations; Constructing a historical station group feature graph based on the node features and the edge features; the historical station group feature graph is used to train a graph neural network; The station group feature map corresponding to the station to be predicted is input into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

2. The cumulative absolute velocity prediction method according to claim 1, characterized in that: The construction of the earthquake monitoring station group comprises: Obtain the type and number of historical earthquake monitoring stations, as well as the distance between any two historical earthquake monitoring stations; Determine a target monitoring station and surrounding stations among historical earthquake monitoring stations; the surrounding stations are stations around the target monitoring station; the types of the target monitoring station and the surrounding stations are surface stations; the number of historical records of the target monitoring station and the surrounding stations is greater than a first threshold; the distance between the target monitoring station and the surrounding stations is less than a second threshold; the number of the surrounding stations is greater than a third threshold; Based on the target monitoring station and the surrounding stations, an earthquake monitoring station group is constructed.

3. The cumulative absolute velocity prediction method according to claim 2, characterized in that: The step of obtaining the node characteristics of each station in the seismic monitoring station group includes: Acquire the first node feature of the target monitoring station; the first node feature includes the magnitude, epicenter distance, soil shear wave velocity and focal depth in the historical records of the target monitoring station; The second node characteristics of the surrounding stations are obtained; the second node characteristics include soil shear wave velocity, epicenter distance and cumulative absolute velocity in the historical records of the surrounding stations.

4. The cumulative absolute velocity prediction method according to claim 1, characterized in that: The acquiring of edge features between the target monitoring station and the surrounding stations comprises: Determining a first edge feature between the target monitoring station and the surrounding stations based on a site condition difference; the site condition difference is determined based on an absolute difference in soil shear wave velocity between the target monitoring station and the surrounding stations; Determine a second edge feature between the target monitoring station and the surrounding stations based on a spatial relationship; the spatial relationship is determined based on an absolute difference in epicentral distance between the target monitoring station and the surrounding stations; Based on the distance between the target monitoring station and the surrounding stations, a third edge feature between the target monitoring station and the surrounding stations is determined.

5. The cumulative absolute speed prediction method according to claim 4, characterized in that: The constructing of the historical station group feature graph based on the node feature and the edge feature comprises: Transforming the site condition difference and the spatial relationship to obtain nonlinear characteristics; Determining a connection relationship between the target monitoring station and the surrounding stations based on the nonlinear feature and the third edge feature; Based on the connection relationship and the target topological structure, a historical station group characteristic graph is constructed; in the target topological structure, the target monitoring station is connected to the surrounding stations, and any two surrounding stations are not connected.

6. The cumulative absolute velocity prediction method according to claim 1, characterized in that: The graph neural network includes a dynamic graph attention layer, a global average pooling layer and a fully connected layer; the cumulative absolute speed prediction method also includes: Inputting the historical station group feature graph into the dynamic graph attention layer to obtain the node-to-node relationship and the adjusted edge feature weights; Inputting the inter-node relationship and the adjusted edge feature weights into the global average pooling layer to obtain a global feature vector; Inputting the global feature vector into the fully connected layer to obtain an initial prediction result; The graph neural network is trained based on the initial prediction results; the training process includes network architecture adjustment and activation function adjustment.

7. A cumulative absolute speed prediction device, characterized in that: include: An earthquake monitoring station group construction module is used to construct an earthquake monitoring station group; the earthquake monitoring station group includes a target monitoring station and surrounding stations; A feature acquisition module, used to acquire node features of each station in the seismic monitoring station group, and to acquire edge features between the target monitoring station and the surrounding stations; A historical station group feature graph construction module, used to construct a historical station group feature graph based on the node features and the edge features; the historical station group feature graph is used to train a graph neural network; The cumulative absolute velocity prediction module is used to input the station group feature map corresponding to the station to be predicted into the trained graph neural network to obtain the cumulative absolute velocity prediction result of the station to be predicted.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the cumulative absolute speed prediction method according to any one of claims 1 to 6 is implemented.

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

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

Images