A prediction method, device, readable storage medium and electronic equipment

Abstract

The specification discloses a prediction method, device, readable storage medium and electronic equipment. When a seismic wave propagates to a protected area, a seismic station detects a seismic wave signal, and then, based on the seismic wave signal and station information of the seismic station, a focus position and a magnitude of the seismic wave are predicted, and then, based on the predicted focus position and magnitude, prompt information is sent. It can be seen that the prediction method in the specification can predict the focus position and magnitude of the seismic wave based on the signal of the seismic wave and the station information of the seismic station before the seismic wave propagates to the protected position, and then, based on the predicted focus position and magnitude, prompt information is sent, which can reduce the damage degree of the protected area and protect the personal safety and property safety of users in the corresponding area of the protected area.

Description

Technical Field

[0001] This specification relates to the field of earthquake monitoring, and in particular to a prediction method, apparatus, readable storage medium, and electronic device. Background Technology

[0002] Earthquakes, also known as tectonic holes or seismic events, mainly occur at plate boundaries about 100 kilometers below the surface. Sudden sliding occurs between plate boundaries, releasing strain energy accumulated over a long period of time, thus triggering an earthquake and threatening people's lives and property.

[0003] Currently, since earthquakes typically cause significant damage to fortified areas, sending early warnings to these areas before the seismic waves reach them can reduce losses. These fortified areas can include densely populated areas, locations of important large-scale equipment, and construction sites that may be susceptible to severe secondary disasters caused by the earthquake.

[0004] Based on this, this specification provides a prediction method. Summary of the Invention

[0005] This specification provides a prediction method, apparatus, readable storage medium, and electronic device to partially solve the aforementioned problems existing in the prior art.

[0006] The following technical solution is adopted in this specification:

[0007] This specification provides a prediction method applied to an earthquake monitoring system, the method comprising:

[0008] Receive signals of the seismic waves to be predicted detected by seismic stations;

[0009] The signal and the station information of the seismic station are input into a pre-trained prediction model to obtain the source location and magnitude of the earthquake wave to be predicted output by the prediction model. The station information includes at least one of the historical earthquake source records of the seismic station and the location information of the seismic station.

[0010] The obtained source location and magnitude of the earthquake wave to be predicted are used as the prediction result, and a prompt message is sent according to the prediction result. The prompt message is used to indicate that an earthquake of the magnitude of the earthquake has occurred at the source location.

[0011] Optionally, the signal includes components in three directions: east-west, north-south, and vertical.

[0012] The signal and the seismic station information are input into a pre-trained prediction model, specifically including:

[0013] For each direction, the component of the signal in that direction is normalized to obtain the normalized result for that direction;

[0014] The normalized results corresponding to each direction are combined with the station information of the seismic station, and the combined result is input into the pre-trained prediction model.

[0015] Optionally, the signal and the seismic station information are input into a pre-trained prediction model, specifically including:

[0016] Based on the historical earthquake source records of the seismic station, determine the source location and earthquake magnitude of each reference earthquake detected by the seismic station in history;

[0017] Based on the reference earthquakes, their corresponding focal locations and magnitudes, the seismic stations, the distances between the focal locations of the reference earthquakes and the seismic stations, a focal distribution map is determined with reference earthquakes and seismic stations as nodes and the distances between the focal locations of the reference earthquakes and the seismic stations as edges.

[0018] Based on the source distribution map, the station information of the seismic stations is determined, and the signal and the determined station information of the seismic stations are input into the pre-trained prediction model.

[0019] Optionally, the signal and the station information of the seismic station are input into a pre-trained prediction model to obtain the source location and magnitude of the seismic wave to be predicted output by the prediction model, specifically including:

[0020] The signal and the station information of the seismic station are input into the feature extraction layer of the pre-trained prediction model to obtain the seismic features output by the feature extraction layer.

[0021] The earthquake features are input into the magnitude prediction layer, depth prediction layer, distance prediction layer, and azimuth prediction layer of the prediction model, respectively, to obtain the magnitude of the earthquake wave to be predicted output by the magnitude prediction layer, the vertical distance between the source of the earthquake wave to be predicted and the seismic station output by the depth prediction layer, the horizontal distance between the source of the earthquake wave to be predicted and the seismic station output by the distance prediction layer, and the azimuth angle between the source of the earthquake wave to be predicted and the seismic station output by the azimuth prediction layer.

[0022] The location of the seismic source of the earthquake wave to be predicted is determined based on the vertical distance between the source of the earthquake wave to be predicted and the seismic station, the horizontal distance between the source of the earthquake wave to be predicted and the seismic station, and the azimuth angle formed by the source of the earthquake wave to be predicted and the seismic station.

[0023] Optionally, the prediction model is trained in the following manner:

[0024] The signal of the sample seismic wave detected by the sample station and the station information of the sample station are determined as training samples, and the source location and magnitude of the sample seismic wave are determined as the first label.

[0025] The training samples are input into the prediction model to be trained, and the first prediction result of the training samples is obtained through the feature extraction layer and the source prediction layer of the prediction model.

[0026] Based on the difference between the first prediction result and the first annotation, a first loss is determined, and the prediction model is trained with minimizing the first loss as the optimization objective.

[0027] Optionally, the prediction model is trained with minimizing the first loss as the optimization objective, specifically including:

[0028] The time difference between the arrival times of the shear waves and the arrival times of the p-waves in the sample seismic waves is determined and used as a second label;

[0029] The training samples are input into the phase prediction layer of the prediction model to obtain the second prediction result output by the phase prediction layer.

[0030] Based on the difference between the second prediction result and the second label, a second loss is determined, and the prediction model is trained with the goal of minimizing the sum of the first loss and the second loss.

[0031] Optionally, the method further includes:

[0032] The prediction results are stored in the historical earthquake source record corresponding to the seismic station.

[0033] This specification provides a prediction device applied to an earthquake monitoring system, the device comprising:

[0034] The receiving module is used to receive the signal of the seismic wave to be predicted detected by the seismic station;

[0035] The prediction module is used to input the signal and the station information of the seismic station into a pre-trained prediction model to obtain the source location and magnitude of the seismic wave to be predicted output by the prediction model, wherein the station information includes at least one of the historical seismic source records of the seismic station and the location information of the seismic station.

[0036] The prompting module is used to take the source location and magnitude of the earthquake wave to be predicted as the prediction result, and send a prompting message according to the prediction result. The prompting message is used to indicate that an earthquake of the magnitude has occurred at the source location.

[0037] This specification provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described prediction method.

[0038] This specification provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the prediction method described above.

[0039] The above-mentioned technical solutions adopted in this specification can achieve the following beneficial effects:

[0040] Before seismic waves reach the fortified area, the signals of the seismic waves are detected by seismic stations. Based on the signals of the seismic waves and the station information, the location and magnitude of the seismic source are predicted, and then a warning message is sent based on the predicted location and magnitude.

[0041] As can be seen, the prediction method in this application can predict the source location and magnitude of seismic waves before they propagate to the fortified area based on the seismic wave signal and the station information of seismic stations. Then, based on the predicted source location and magnitude, it can send warning information, which can reduce the degree of damage to the fortified area and protect the personal and property safety of users in the fortified area. Attached Figure Description

[0042] The accompanying drawings, which are included to provide a further understanding of this specification and form part of this specification, illustrate exemplary embodiments and their descriptions, serving to explain this specification and do not constitute an undue limitation thereof.

[0043] In the picture:

[0044] Figure 1 This is a flowchart illustrating the prediction method provided in this specification.

[0045] Figure 2 This is a flowchart illustrating the prediction method provided in this specification.

[0046] Figure 3 This is a flowchart illustrating the prediction method provided in this specification.

[0047] Figure 4 This is a schematic diagram of the predictive model provided in this specification;

[0048] Figure 5This is a detailed structural diagram of the prediction model provided in this specification;

[0049] Figure 6 This is a schematic diagram of the predictive device provided in this specification;

[0050] Figure 7 The corresponding information provided in this specification Figure 1 A schematic diagram of an electronic device. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of them. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0052] The technical solutions provided in the various embodiments of this specification are described in detail below with reference to the accompanying drawings.

[0053] Earthquakes typically generate S-waves and P-waves. P-waves travel faster but are less destructive, while S-waves travel slower but are more destructive. Therefore, P-waves often reach the ground earlier than S-waves. If the location and magnitude of an earthquake's hypocenter can be predicted based on the P-waves before the S-waves reach the protected area, and if early warnings can be issued based on these predictions, effective earthquake monitoring can be achieved.

[0054] The designated earthquake-protected area can include areas where nuclear facilities, large reservoirs, rail transit systems, high-speed railways, key substations, oil and gas pipelines (stations), mines, petrochemical plants, major bridges and long tunnels, and other construction projects that may be affected by severe secondary disasters caused by earthquakes are located. It can also include densely populated areas such as schools, hospitals, large shopping malls, stadiums, train stations, and airports. Of course, the designated area can also include areas equipped with automatic earthquake early warning information receiving devices. The specific designation of this earthquake-protected area can be determined according to needs, and this manual does not impose any restrictions on it.

[0055] Figure 1 The flowchart of the prediction method provided in this specification is shown, and it specifically includes the following steps:

[0056] S100: Receives signals of seismic waves to be predicted detected by seismic stations.

[0057] This specification provides a prediction method, in which the prediction model can be pre-trained. The execution of this prediction method can be performed by an earthquake monitoring system used to predict the focal location and magnitude of earthquakes. This earthquake monitoring system can be deployed on electronic devices such as servers and terminals; this specification uses an earthquake monitoring system deployed on a server as an example. The electronic device executing this prediction method and the electronic device executing the training process of the prediction model can be the same electronic device or different electronic devices; this specification does not impose any restrictions on this.

[0058] Unlike current methods that can only predict earthquake magnitude and focal depth after the arrival of seismic waves, this method cannot predict these attributes before an earthquake strikes, leading to greater damage during earthquakes. This manual provides a new prediction method. Before seismic waves reach the fortified area, the method detects the signals at seismic stations and predicts the focal location and magnitude based on the signals and station information. Based on the predicted focal location and magnitude, a warning message is then sent. Therefore, this prediction method can predict the focal location and magnitude of seismic waves before they reach the fortified area, thus reducing the damage caused by seismic waves and protecting property in the fortified area.

[0059] Based on the above brief description of the prediction method in this specification, it can be seen that the prediction method in this specification can detect seismic wave signals through seismic stations.

[0060] Specifically, when an earthquake occurs, it typically generates transverse and longitudinal waves. Seismic stations, located on or below the ground, can collect the signals of these seismic waves. These underground seismic stations can be located in water or underground in mines; the specific location of the seismic station can be chosen as needed, and this manual does not impose any restrictions.

[0061] Then, when the seismic station determines that an earthquake has occurred based on the detected seismic wave signals, it can send the detected seismic wave signals to a server deployed with an earthquake monitoring system. The seismic station can be configured with earthquake conditions such as "if the amplitude of the acquired signal is greater than a preset threshold, then the signal is a seismic wave signal." When the seismic wave meets these conditions, the seismic station can determine that an earthquake has occurred.

[0062] Finally, the server can receive the seismic wave signals sent by the seismic station and use these signals as the signals for the seismic waves to be predicted.

[0063] Of course, the seismic station can also send the signals of seismic waves collected at different time periods to the server according to a preset time interval. The server will then receive the seismic wave signals and use the received seismic wave signals as the signals of the seismic waves to be predicted.

[0064] The specific timing for when the seismic station sends seismic wave signals to the server can be configured as needed; this manual does not impose any restrictions on this.

[0065] S102: Input the signal and the station information of the seismic station into a pre-trained prediction model to obtain the source location and magnitude of the earthquake wave to be predicted output by the prediction model, wherein the station information includes at least one of the historical earthquake source records of the seismic station and the location information of the seismic station.

[0066] In one or more embodiments provided in this specification, generally, the probability of an earthquake occurring in a region where two tectonic plates meet is higher than the probability of an earthquake occurring in a region located between tectonic plates. Similarly, if the hypocenters of earthquakes historically detected by a seismic station are all due north of that seismic station, and assuming that the seismic wave detected by that station this time is a seismic wave to be predicted, then the hypocenter of the seismic wave to be predicted is also highly likely to be due north of that seismic station.

[0067] Obviously, the station information from seismic stations is also very helpful in determining the source location and magnitude of the seismic wave to be predicted. Therefore, the server can determine the source location and magnitude of the seismic wave to be predicted based on the signal of the seismic wave and the station information of the seismic stations that acquired the signal. The magnitude is the seismic grade of the seismic wave to be predicted.

[0068] Specifically, the server can determine the seismic station that acquired the signal of the seismic wave to be predicted from among the seismic stations that can communicate with the server, based on the station identifier that sent the signal of the seismic wave to be predicted.

[0069] Secondly, after identifying the seismic station that collected the signal of the seismic wave to be predicted, the server can determine at least one of the historical earthquake source records of the seismic station and the location information of the seismic station.

[0070] The historical earthquake source records of this seismic station include the source locations and magnitudes of seismic waves detected by the station throughout history. These records can be determined manually or based on the prediction method described in this manual. After predicting the source location and magnitude of the seismic wave to be predicted, these information is stored in the historical earthquake source records of the corresponding seismic station. The location information of the seismic station can be its latitude and longitude, or its specific location within the region, such as the three-dimensional coordinates of the seismic station in a three-dimensional coordinate system centered on the server. The specific methods for determining these historical earthquake source records, the form of the location information, and the determination of the seismic station's location can be configured as needed; this manual does not impose any restrictions on these aspects.

[0071] Then, the server can determine the station information of the seismic station based on at least one of the historical earthquake source records of the seismic station and the location information of the seismic station.

[0072] Finally, the server can use the seismic station information and the signal of the seismic wave to be predicted as inputs to a pre-trained prediction model to obtain the source location and magnitude of the seismic wave output by the model. Figure 2 As shown. The location of the earthquake source can be its latitude and longitude and its depth, or the three-dimensional coordinates corresponding to the location of the earthquake source in a three-dimensional coordinate system centered on the server's location. The specific form of the earthquake source location can be set as needed, and this specification does not impose any restrictions on it.

[0073] Figure 2 This is a flowchart illustrating the prediction method provided in this specification. The server can input the signal of the seismic wave to be predicted and the station information of the seismic wave into the prediction model, and obtain the source location and magnitude output by the prediction model. The server can then perform subsequent steps based on the determined source location and magnitude.

[0074] S104: The obtained source location and magnitude of the earthquake wave to be predicted are used as the prediction result, and a prompt message is sent according to the prediction result. The prompt message is used to indicate that an earthquake of the magnitude of the earthquake has occurred at the source location.

[0075] In one or more embodiments provided in this specification, the source location and magnitude of the seismic wave are predicted. The purpose is to provide early warnings based on the prediction results to protect the personal and property safety of users in the fortified area. Therefore, after determining the source location and magnitude of the seismic wave to be predicted, the server can issue an early warning.

[0076] Specifically, the server can use the source location and magnitude of the earthquake wave to be predicted as the prediction result.

[0077] Then, the server can determine a designated area containing the location of the epicenter based on the determined prediction result. This designated map can be a regular or irregular area centered on the epicenter location, or it can be a provincial-level administrative region, a municipal-level administrative region, etc., containing the epicenter location. The specific method for determining this designated area can be set as needed, and this manual does not impose any restrictions on it.

[0078] Finally, the server can send a notification message to the seismic stations in the specified area to indicate that an earthquake of that magnitude occurred at the hypocenter location in that area. This notification message is specifically for indicating that an earthquake of that magnitude occurred at the hypocenter location.

[0079] Of course, each designated region can be equipped with its own corresponding earthquake monitoring system. After determining the prediction result, the earthquake monitoring system deployed on the server can send graphic information to the earthquake monitoring system corresponding to the designated region based on the prediction result. The earthquake monitoring system corresponding to the designated region can then issue an earthquake early warning based on the prompt information.

[0080] based on Figure 1 The prediction method described herein detects seismic wave signals at seismic stations before the waves reach the fortified area. Based on the seismic wave signals and station information, it predicts the hypocenter and magnitude of the seismic waves, and then sends out warning messages based on the predicted hypocenter and magnitude. Therefore, this prediction method, based on seismic wave signals and seismic station information, can more accurately predict the hypocenter and magnitude of seismic waves before they reach the fortified area, and then send out warning messages based on the predicted hypocenter and magnitude. This can reduce the degree of damage in the fortified area and protect the personal and property safety of users in the fortified area.

[0081] Furthermore, the prediction method provided in this specification predicts the source location and magnitude of the seismic wave to be predicted based on at least one of the historical earthquake source records and the location information of the seismic station. Compared to predictions based solely on the signal of the seismic wave, the obtained prediction results fully consider the environmental information of the environment where the seismic station collecting the seismic wave is located, and are therefore more accurate. This environmental information may include the location of the seismic station and the geological information of the area where the seismic station is located.

[0082] Furthermore, under normal circumstances, if a three-dimensional coordinate system is established with the seismic station as the center, the seismic wave signal can be characterized by components in three directions.

[0083] Specifically, if the three directions are east-west, north-south, and vertical, then the signal of the seismic wave to be predicted contains components in these three directions.

[0084] Therefore, the seismic station can collect seismic waves in each direction, and after collecting each seismic wave, combine the seismic waves collected in the same time period in each direction, and input the combination result into the prediction model as the seismic wave to be predicted.

[0085] The orientation of each coordinate axis in this three-dimensional coordinate system can be set as needed, and this manual does not impose any restrictions on this.

[0086] Furthermore, the server can normalize the signal of the seismic wave to be predicted before inputting it into the prediction model.

[0087] Specifically, the server can also normalize the signal component in each direction to obtain a normalized result for that direction. The server can determine the normalization result by compressing the amplitude of the component in that direction, or by determining the difference between the amplitude at each position and the mean in that direction, and then using the ratio of this difference to the variance of the amplitude of the component in that direction as the normalization result.

[0088] With x wave As a seismic wave, x wave_mean As the mean value determined based on the amplitudes of the seismic waves, x wave_std Taking the variance determined based on the amplitudes of seismic waves as an example, the normalized result for each direction can be:

[0089]

[0090] Furthermore, to avoid excessive information loss in the normalized result compared to the original seismic wave signal, the input information of the prediction model can also include statistical data of the seismic wave signal. This statistical data can include the maximum and minimum values, mean, and standard deviation of the seismic wave signal. Similarly, to avoid a large difference between the statistical data and the normalized result, which could lead to poor prediction performance, the server can also perform logarithmic processing on the statistical data and include the processing result as part of the normalization result.

[0091] The server can then fuse the normalized results corresponding to each direction, use the fused result as the signal of the seismic wave to be predicted, and then combine the signal of the seismic wave to be predicted with the station information of the seismic station, using the combined result as the input of the prediction model.

[0092] Of course, the server can also directly combine the normalized results corresponding to each direction with the station information of the seismic station to determine the input of the prediction model. The specific method for normalizing the signal of the seismic wave to be predicted can be set as needed; this manual does not impose any restrictions on this.

[0093] In addition, knowledge graphs or graph structures typically contain more information than other structures. Therefore, the server can use a graph structure to represent the station information of the seismic station.

[0094] Specifically, the server can determine the historical earthquake source records of the seismic station, and based on these records, determine the source location and magnitude of each reference earthquake detected by the station in history. For each reference earthquake, the reference earthquake is the earthquake corresponding to the seismic waves detected by the station in its history.

[0095] Then, the server can determine a source distribution map based on each reference earthquake, the corresponding focal location and magnitude of each reference earthquake, the seismic station, and the distances between the focal locations of each reference earthquake and the seismic station. This map uses reference earthquakes and seismic stations as nodes, and the distances between the focal locations of reference earthquakes and seismic stations as edges. In other words, the determined source distribution map contains nodes for the seismic station and each reference earthquake, and the edges between any two nodes can be the distances between the focal locations of the reference earthquakes and the seismic stations.

[0096] Of course, the above source distribution map may also include edges showing the distances between each source. In other words, for each reference earthquake, the server can determine the distance between the source of other reference earthquakes and the source of this reference earthquake, and determine the edges between other reference earthquakes and this reference earthquake based on the determined distances.

[0097] The specific method for determining the earthquake source distribution map and the representation type of each element contained in the earthquake source distribution map, i.e., each node and each edge, can be set as needed, and this manual does not impose any restrictions on this.

[0098] Finally, the server can determine the station information of the seismic station based on the source distribution map, fuse the determined station information with the seismic wave signal, and input the fusion result into the pre-trained prediction model.

[0099] The server can directly use the determined source distribution map as the station information for the seismic station. Alternatively, it can determine the station characteristics of the seismic station based on the source distribution map, and use this information as the station information. It can also supplement the source distribution map based on the location information of the seismic station. Specifically, it determines the node corresponding to the location information of the seismic station, updates that node in the source distribution map, and connects that node to the node corresponding to the seismic station. Then, based on the updated source distribution map, it determines the station characteristics of the seismic station, using this information as the station information. The specific content included in the source distribution map, how the station information is determined, and the specific format of the station information can be set as needed; this manual does not impose any restrictions on this.

[0100] Of course, the seismic wave signal fused with the station information can be the waveform of the seismic wave signal or the seismic features obtained by feature extraction of the seismic wave signal. The specific form of the seismic wave signal can be set as needed, and this manual does not impose any restrictions on it.

[0101] Furthermore, taking the determined hypocenter location as a three-dimensional coordinate system constructed with the seismic station as the origin as an example, if the distance between the hypocenter location and the origin is short, the accuracy of the hypocenter location is high; if the distance between the hypocenter location and the origin is far, the accuracy of the hypocenter location is low. Therefore, to avoid the above situation, the server may not directly determine the three-dimensional coordinates corresponding to the hypocenter location, but instead determine the distance, depth, and azimuth angle of the hypocenter location relative to the seismic station to determine the hypocenter location.

[0102] Specifically, the prediction model can be a multi-task model. This prediction model includes a feature extraction layer, a magnitude prediction layer, a depth prediction layer, a distance prediction layer, and an azimuth prediction layer.

[0103] Therefore, the server can input the signal and the station information of the seismic station into the feature extraction layer of the pre-trained prediction model to obtain the seismic features output by the feature extraction layer.

[0104] Then, the server can take the earthquake characteristics as input, and input the magnitude prediction layer, depth prediction layer, distance prediction layer and azimuth prediction layer of the prediction model to obtain the magnitude of the earthquake wave to be predicted output by the magnitude prediction layer, the vertical distance between the source of the earthquake wave to be predicted and the seismic station output by the depth prediction layer, the horizontal distance between the source of the earthquake wave to be predicted and the seismic station output by the distance prediction layer, and the azimuth angle between the source of the earthquake wave to be predicted and the seismic station output by the azimuth prediction layer.

[0105] Finally, the server can determine the location of the seismic wave source based on the vertical distance between the source of the seismic wave and the seismic station, the horizontal distance between the source of the seismic wave and the seismic station, and the azimuth angle between the source of the seismic wave and the seismic station. Figure 3 As shown.

[0106] Figure 3 This is a flowchart illustrating the prediction method provided in this specification. The server takes the signal of the earthquake to be predicted and the station information of the seismic station as input, and inputs them into the feature extraction layer of the prediction model to obtain earthquake features. These earthquake features are then input into the magnitude prediction layer, depth prediction layer, distance prediction layer, and azimuth prediction layer of the prediction model, respectively, to obtain the magnitude of the earthquake wave to be predicted output by the magnitude prediction layer, the vertical distance between the source of the earthquake wave to be predicted and the seismic station output by the depth prediction layer, the horizontal distance between the source of the earthquake wave to be predicted and the seismic station output by the distance prediction layer, and the azimuth angle between the source of the earthquake wave to be predicted and the seismic station output by the azimuth prediction layer. Based on the magnitude, horizontal distance, vertical distance, and azimuth angle determined above, the source location and magnitude of the earthquake wave to be predicted are determined.

[0107] Wherein, the vertical direction is the direction from the Earth's center to the ground, the horizontal direction is the direction parallel to the ground, and the azimuth angle can be the angle corresponding to the straight line formed between the seismic station and the seismic source.

[0108] It should be noted that, generally, the predicted earthquake magnitude is greater than a preset first threshold, the focal depth is greater than a preset second threshold, the focal distance is greater than a preset third threshold, and the azimuth falls within a preset fourth range. The first threshold can be magnitude 0, the second threshold can be the opposite of the highest elevation in the region, the third threshold can be 0, and the fourth range can be [0, 2π]. Of course, the first, second, third, and fourth thresholds can be set as needed, and this specification does not impose any restrictions on this.

[0109] Furthermore, the prediction model in this prediction method can be trained in the following manner.

[0110] Specifically, the server can randomly select any one of the various seismic stations as the sample station.

[0111] Secondly, the server can determine the sample seismic wave from the historical earthquake source record corresponding to the sample station, and determine the signal of the sample seismic wave based on the sample seismic wave.

[0112] Therefore, the server can determine the signal of the sample seismic wave and the station information of the sample station as training samples, and determine the source location and magnitude of the sample seismic wave as the first label.

[0113] Then, the server can use the training sample as input to the feature extraction layer of the prediction model to be trained, obtain the sample features corresponding to the training sample, and then input the sample features into the source prediction layer of the prediction model to obtain the source location and magnitude corresponding to the training sample, which is used as the first prediction result.

[0114] Finally, the server can determine the first loss based on the difference between the first prediction result and the first label, and then train the prediction model with the minimization of the first loss as the optimization objective.

[0115] It should be noted that the aforementioned source prediction layer may correspond to only one task, namely, the task of predicting the source location and magnitude of the seismic wave to be predicted. Alternatively, it may correspond to multiple tasks, such as predicting the vertical distance between the source of the seismic wave and the seismic station, predicting the horizontal distance between the source of the seismic wave and the seismic station, predicting the azimuth angle between the source of the seismic wave and the seismic station, and predicting the magnitude of the seismic wave to be predicted. In other words, the source prediction layer includes the aforementioned magnitude prediction layer, depth prediction layer, distance prediction layer, and azimuth prediction layer. The specific task that the source prediction layer corresponds to can be set as needed, and this specification does not impose any restrictions on this.

[0116] In addition, when determining training samples, the server can also perform data augmentation on the signals of the determined sample seismic waves, such as adding noise, event offset, random zeroing, and random amplitude modulation, and use the data augmentation results as training samples to enhance the robustness of the prediction model trained based on these training samples.

[0117] Furthermore, when determining the annotations for each training sample, if the source prediction layer corresponds to multiple tasks, the server needs to determine the magnitude, source depth, source distance, and azimuth of the sample seismic wave, respectively, corresponding to the first magnitude annotation, first depth annotation, first distance annotation, and first azimuth annotation for that training sample. The first magnitude annotation is the magnitude of the sample seismic wave; the first depth annotation is the vertical distance between the source of the sample seismic wave and the sample station; and the first distance annotation is the horizontal distance between the source of the sample seismic wave and the sample station. The first azimuth annotation can be obtained in the following way:

[0118] First, determine the latitude and longitude of the earthquake source and the sample station.

[0119] Then, determine the preset azimuth angle determination formula.

[0120] Finally, the first azimuth angle label is determined based on the azimuth angle determination formula, the latitude and longitude of the seismic source, and the latitude and longitude of the sample station.

[0121] Furthermore, for earthquakes, the time difference between the arrival times of the shear wave and the p-wave can be used to characterize how long it takes for the corresponding shear wave to arrive after the p-wave is detected. Therefore, this time difference can also be considered one of the properties of an earthquake. Thus, the server can use this time difference to assist in training the prediction model.

[0122] Specifically, the server can determine the time difference between the arrival times of the shear waves and the arrival times of the p-waves in the sample seismic wave, as a second annotation.

[0123] Then, the server can input the training sample into the phase prediction layer of the prediction model to obtain the second prediction result output by the phase prediction layer.

[0124] Finally, the server can determine a second loss based on the difference between the second prediction result and the second label, and train the prediction model with the optimization objective of minimizing the sum of the first loss and the second loss.

[0125] It should be noted that this phase prediction layer can be used to assist in the model training phase of the prediction model, or it can be applied to the prediction method to predict how long after the arrival of the P-wave will the S-wave arrive, and send a warning message based on the arrival time of the S-wave. The specific use of this phase prediction layer can be configured as needed, and this manual does not impose any restrictions on it.

[0126] Furthermore, after obtaining the prediction results, the server can also store the prediction results in the historical earthquake source record corresponding to the seismic station, so that the source location and magnitude of the next earthquake wave to be detected by the seismic station can be predicted based on the historical earthquake source record corresponding to the seismic station.

[0127] Following the same approach, this specification provides a schematic diagram of the predictive model structure, such as... Figure 4 As shown.

[0128] Figure 4This is a schematic diagram of the prediction model provided in this specification. The prediction model includes a seismic wave encoding module, a geolocation information embedding module, a phase decoding module, and an output decoding module. The seismic wave encoding module encodes the seismic wave signal input to the prediction model; the geolocation information embedding module encodes the station information input to the prediction model; the input decoding module determines the magnitude and source location of the seismic wave to be predicted; and the phase decoding module determines the time difference between the arrival times of the shear wave and the p-wave of the seismic wave to be predicted.

[0129] Therefore, the server can input the seismic wave signal of the earthquake to be predicted into the prediction model, and obtain the seismic characteristics through the seismic wave encoding module. Simultaneously, the server can input the address information of the seismic station into the prediction model's address information embedding module to obtain the station's characteristics. Then, based on the seismic characteristics, the server can determine the time difference between the arrival times of the S-wave and P-wave of the earthquake to be predicted through the phase decoding module. Finally, based on the seismic characteristics and station characteristics, the server can determine the source location and magnitude of the earthquake to be predicted through the output decoding module.

[0130] Similarly, based on the same idea, this specification provides a detailed structural diagram of a prediction model, such as... Figure 5 As shown.

[0131] Figure 5 This is a detailed structural diagram of the prediction model provided in this specification. The prediction model includes a seismic wave coding module, an address information embedding module, a phase decoding module, and an output decoding module. The seismic coding module includes three fully connected layers and one convolutional layer; the address information embedding module contains three convolutional layers; the output decoding module contains one decoder; and the phase decoding module contains three fully connected layers and one long short-term memory (LSTM) layer.

[0132] Therefore, the server can input the signal of the seismic wave to be predicted into the seismic wave coding layer. Through two connection layers and one convolutional layer in the seismic wave coding module, the characteristics of the seismic wave signal are obtained. Then, through a fully connected layer, the seismic characteristics are determined based on the seismic wave signal's characteristics and waveform statistics. The phase decoding module can then obtain the time difference between the arrival times of the shear wave and the arrival times of the p-wave of the seismic wave to be predicted, based on these seismic characteristics and its own network structure. The address information embedding module can extract features from the source distribution map and the station identifier, obtaining a first vector and a second vector respectively. The first vector, the second vector, and the seismic characteristics are then fused, and through its three convolutional layers, the station characteristics of the seismic station are obtained. The output decoding module can fuse the station characteristics and the seismic characteristics, and through the decoder, determine the source location and magnitude of the seismic wave to be predicted based on the fusion result.

[0133] Following the same line of thought, this specification also provides a prediction device, such as... Figure 6 As shown.

[0134] Figure 6 This is a schematic diagram of the predictive device provided in this specification, which is applied to an earthquake monitoring system.

[0135] in:

[0136] The receiving module 200 is used to receive the signal of the seismic wave to be predicted detected by the seismic station.

[0137] The prediction module 202 is used to input the signal and the station information of the seismic station into a pre-trained prediction model to obtain the source location and magnitude of the seismic wave to be predicted output by the prediction model, wherein the station information includes at least one of the historical seismic source records of the seismic station and the location information of the seismic station.

[0138] The prompting module 204 is used to take the source location and magnitude of the earthquake wave to be predicted as the prediction result, and send a prompting message according to the prediction result. The prompting message is used to indicate that an earthquake of the magnitude has occurred at the source location.

[0139] Optionally, the signal contains components in three directions: east-west, north-south, and vertical. The prediction module 202 is used to normalize the components of the signal in each direction to obtain a normalized result for that direction, combine the normalized results corresponding to each direction with the station information of the seismic station, and input the combined result into a pre-trained prediction model.

[0140] Optionally, the prediction module 202 is used to determine the source location and earthquake magnitude of each reference earthquake detected by the seismic station in history based on the historical earthquake source records of the seismic station; to determine a source distribution map with reference earthquakes and seismic stations as nodes and the distance between the source of the reference earthquakes and the seismic station as edges based on the reference earthquakes, the source locations and earthquake magnitudes of each reference earthquake, the seismic station, the source of each reference earthquake and the distance between the seismic station and the seismic station; to determine the station information of the seismic station based on the source distribution map; and to input the signal and the determined station information of the seismic station into a pre-trained prediction model.

[0141] Optionally, the prediction module 202 is used to input the signal and the station information of the seismic station into the feature extraction layer of a pre-trained prediction model to obtain the seismic features output by the feature extraction layer. The seismic features are then input into the magnitude prediction layer, depth prediction layer, distance prediction layer, and azimuth prediction layer of the prediction model to obtain the magnitude of the seismic wave to be predicted output by the magnitude prediction layer, the vertical distance between the source of the seismic wave to be predicted and the seismic station output by the depth prediction layer, the horizontal distance between the source of the seismic wave to be predicted and the seismic station output by the distance prediction layer, and the azimuth angle formed by the source of the seismic wave to be predicted and the seismic station output by the azimuth prediction layer. Based on the vertical distance between the source of the seismic wave to be predicted and the seismic station, the horizontal distance between the source of the seismic wave to be predicted and the seismic station, and the azimuth angle formed by the source of the seismic wave to be predicted and the seismic station, the location of the source of the seismic wave to be predicted is determined.

[0142] The device further includes:

[0143] The training module 206 is used to train the prediction model in the following manner: determining the signal of the sample seismic wave detected by the sample station and the station information of the sample station as training samples, and determining the source location and magnitude of the sample seismic wave as the first label; inputting the training samples into the prediction model to be trained; obtaining the first prediction result of the training samples through the feature extraction layer and the source prediction layer of the prediction model; determining the first loss based on the difference between the first prediction result and the first label; and training the prediction model with the minimization of the first loss as the optimization objective.

[0144] Optionally, the training module 206 is used to determine the time difference between the arrival time of the shear wave and the arrival time of the p-wave of the sample seismic wave as a second label, input the training sample into the phase prediction layer in the prediction model, obtain the second prediction result output by the phase prediction layer, determine the second loss based on the difference between the second prediction result and the second label, and train the prediction model with the optimization objective of minimizing the sum of the first loss and the second loss.

[0145] Optionally, the prompting module 204 is used to store the prediction results in the historical earthquake source record corresponding to the seismic station.

[0146] This specification also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 The prediction method provided.

[0147] This instruction manual also provides Figure 7 The diagram shows a schematic structural representation of the electronic device. Figure 7 At the hardware level, the electronic device includes a processor, internal bus, network interface, memory, and non-volatile memory, and may also include other hardware required for the business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then runs it to achieve the above-mentioned functions. Figure 1 The prediction method described above. Of course, in addition to software implementation, this specification does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0148] In the 1990s, improvements to a technology could be clearly distinguished as either hardware improvements (e.g., improvements to the circuit structure of diodes, transistors, switches, etc.) or software improvements (improvements to the methodology). However, with technological advancements, many methodological improvements today can be considered direct improvements to the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved methodology into the hardware circuit. Therefore, it cannot be said that a methodological improvement cannot be implemented using hardware physical modules. For example, a Programmable Logic Device (PLD) (such as a Field Programmable Gate Array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. Designers can program and "integrate" a digital system onto a PLD themselves, without needing chip manufacturers to design and manufacture dedicated integrated circuit chips. Furthermore, nowadays, instead of manually manufacturing integrated circuit chips, this programming is mostly implemented using "logic compiler" software. Similar to the software compiler used in program development, the original code before compilation must be written in a specific programming language, called a Hardware Description Language (HDL). There are many HDLs, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, and RHDL (Ruby Hardware Description Language). Currently, the most commonly used are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should understand that by simply performing some logic programming on the method flow using one of these hardware description languages ​​and programming it into an integrated circuit, the hardware circuit implementing the logical method flow can be easily obtained.

[0149] The controller can be implemented in any suitable manner. For example, it can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicon Labs C8051F320. A memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also recognize that, in addition to implementing the controller in purely computer-readable program code form, the same functionality can be achieved by logically programming the method steps to make the controller take the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the means included therein for implementing various functions can also be considered as structures within the hardware component. Alternatively, the means for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0150] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0151] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware components.

[0152] Those skilled in the art will understand that embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0153] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable lesion detection device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable lesion detection device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0154] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable lesion detection device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0155] These computer program instructions can also be loaded onto a computer or other programmable lesion detection device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0156] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0157] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0158] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0159] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0160] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, this specification may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this specification may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0162] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0163] The above description is merely an embodiment of this specification and is not intended to limit this specification. Various modifications and variations can be made to this specification by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of the claims of this specification.

Claims

1. A prediction method, characterized in that, The prediction method is applied to an earthquake monitoring system, and the method includes: Receive signals of the seismic waves to be predicted detected by seismic stations; The signal and the station information of the seismic station are input into a pre-trained prediction model to obtain the source location and magnitude of the earthquake wave to be predicted output by the prediction model. The station information includes the historical earthquake source records of the seismic station and the location information of the seismic station. The source location and magnitude of the earthquake wave to be predicted are used as the prediction result, and a prompt message is sent according to the prediction result. The prompt message is used to indicate that an earthquake of the magnitude has occurred at the source location. The step of inputting the signal and the seismic station information into the pre-trained prediction model specifically includes: Based on the historical earthquake source records of the seismic station, determine the source location and earthquake magnitude of each reference earthquake detected by the seismic station in history; Based on the reference earthquakes, their corresponding focal locations and magnitudes, the seismic stations, the distances between the focal locations of the reference earthquakes and the seismic stations, a focal distribution map is determined with reference earthquakes and seismic stations as nodes and the distances between the focal locations of the reference earthquakes and the seismic stations as edges. Based on the source distribution map, the station information of the seismic stations is determined, and the signal and the determined station information of the seismic stations are input into the pre-trained prediction model.

2. The method as described in claim 1, characterized in that, The signal contains components in three directions: east-west, north-south, and vertical. The signal and the seismic station information are input into a pre-trained prediction model, specifically including: For each direction, the component of the signal in that direction is normalized to obtain the normalized result for that direction; The normalized results corresponding to each direction are combined with the station information of the seismic stations, and the combined results are input into the pre-trained prediction model.

3. The method as described in claim 1, characterized in that, The signal and the station information of the seismic station are input into a pre-trained prediction model to obtain the source location and magnitude of the seismic wave to be predicted, as output by the prediction model. Specifically, this includes: The signal and the station information of the seismic station are input into the feature extraction layer of the pre-trained prediction model to obtain the seismic features output by the feature extraction layer. The earthquake features are input into the magnitude prediction layer, depth prediction layer, distance prediction layer, and azimuth prediction layer of the prediction model, respectively, to obtain the magnitude of the earthquake wave to be predicted output by the magnitude prediction layer, the vertical distance between the source of the earthquake wave to be predicted and the seismic station output by the depth prediction layer, the horizontal distance between the source of the earthquake wave to be predicted and the seismic station output by the distance prediction layer, and the azimuth angle between the source of the earthquake wave to be predicted and the seismic station output by the azimuth prediction layer. The location of the seismic source of the earthquake wave to be predicted is determined based on the vertical distance between the source of the earthquake wave to be predicted and the seismic station, the horizontal distance between the source of the earthquake wave to be predicted and the seismic station, and the azimuth angle formed by the source of the earthquake wave to be predicted and the seismic station.

4. The method as described in claim 1, characterized in that, The prediction model was trained in the following manner: The signal of the sample seismic wave detected by the sample station and the station information of the sample station are determined as training samples, and the source location and magnitude of the sample seismic wave are determined as the first label. The training samples are input into the prediction model to be trained, and the first prediction result of the training samples is obtained through the feature extraction layer and the source prediction layer of the prediction model. Based on the difference between the first prediction result and the first annotation, a first loss is determined, and the prediction model is trained with minimizing the first loss as the optimization objective.

5. The method as described in claim 4, characterized in that, The prediction model is trained with minimizing the first loss as the optimization objective, specifically including: The time difference between the arrival times of the shear waves and the arrival times of the p-waves in the sample seismic waves is determined and used as a second label; The training samples are input into the phase prediction layer of the prediction model to obtain the second prediction result output by the phase prediction layer. Based on the difference between the second prediction result and the second label, a second loss is determined, and the prediction model is trained with the goal of minimizing the sum of the first loss and the second loss.

6. The method as described in claim 1, characterized in that, The method further includes: The prediction results are stored in the historical earthquake source record corresponding to the seismic station.

7. A prediction device, characterized in that, The prediction device is applied to an earthquake monitoring system, and the device includes: The receiving module is used to receive the signal of the seismic wave to be predicted detected by the seismic station; The prediction module is used to input the signal and the station information of the seismic station into a pre-trained prediction model to obtain the source location and magnitude of the seismic wave to be predicted output by the prediction model. The station information includes historical seismic source records and the location information of the seismic station. Specifically, the prediction module is used to determine the source location and magnitude of each reference earthquake detected historically by the seismic station based on the historical seismic source records; to determine a source distribution map with reference earthquakes and seismic stations as nodes and the distance between the source of each reference earthquake and the seismic station as edges based on the reference earthquakes, their corresponding source locations and magnitudes, the seismic station, and the distance between the source of each reference earthquake and the seismic station; to determine the station information of the seismic station based on the source distribution map; and to input the signal and the determined station information of the seismic station into the pre-trained prediction model. The prompting module is used to take the source location and magnitude of the earthquake wave to be predicted as the prediction result, and send a prompting message according to the prediction result. The prompting message is used to indicate that an earthquake of the magnitude has occurred at the source location.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method described in any one of claims 1 to 6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 6.

Images