Earthquake early warning system based on Haizhengtong platform

Abstract

The invention relates to the field of earthquake early warning, and discloses an earthquake early warning system based on a Haizhengtong platform, a data acquisition module, used for obtaining earthquake information of historical earthquakes, the earthquake information including epicenter coordinates, epicenter depth, earthquake level, and corresponding intensity attenuation model and the earthquake zone to which it belongs; an intensity prediction module, used for constructing a generalized intensity prediction model based on the epicenter coordinates, epicenter depth and earthquake level, and performing differential correction on the generalized intensity prediction model according to the earthquake zone to which it belongs, so as to obtain a fitted intensity prediction model; the output of the fitted intensity prediction model represents the predicted intensity of any coordinates with respect to the epicenter coordinates; a region division module, used for dividing the earthquake region into a first warning region to an Nth warning region based on the epicenter depth and the earthquake level, with the epicenter coordinates as the center, in combination with the fitted intensity prediction model; and a warning sending module, used for allocating warning sending priorities to the first warning region to the Nth warning region according to the earthquake level.

Description

Technical Field

[0001] The present invention relates to the field of earthquake early warning, and more specifically, to an earthquake early warning system based on the Haizhengtong platform. Background Art

[0002] Earthquake early warning uses the fact that the non-destructive P waves arrive earlier than the S waves. However, traditional earthquake early warning uses the simultaneous transmission of early warning signals, which results in a slow speed of sending early warnings, and there is a problem that the early warning signals arrive later than the S waves in some areas.

[0003] For destructive earthquakes, when the warning signal arrives later than the S wave, the warning signal loses its original warning function. Therefore, the existing technology has the problems of large concurrency and insufficient warning effectiveness. Summary of the invention

[0004] The present invention provides an earthquake early warning system based on the Haizhengtong platform to solve the technical problems raised in the background technology.

[0005] The present invention provides an earthquake early warning system based on the Haizhengtong platform, comprising:

[0006] The data acquisition module is used to obtain the seismic information of historical earthquakes through the Haizhengtong platform. The seismic information includes the epicenter coordinates, epicenter depth, earthquake level, and the corresponding intensity attenuation model and the seismic zone to which it belongs; wherein the intensity attenuation model represents an attenuation model in which the earthquake intensity changes with the change of the straight-line distance from the epicenter coordinates;

[0007] The intensity prediction module is used to construct a generalized intensity prediction model based on the epicenter coordinates, epicenter depth and earthquake level, and to perform differential correction on the generalized intensity prediction model according to the seismic zone to which it belongs, so as to obtain a fitted intensity prediction model; the output of the fitted intensity prediction model represents the predicted intensity of any coordinates with respect to the epicenter coordinates;

[0008] The regional division module is used to divide the earthquake area into the first warning area to the second warning area based on the epicenter depth and earthquake level, with the epicenter coordinates as the center, combined with the fitting intensity prediction model. Early warning area;

[0009] The warning sending module is used to send warning signals from the first warning area to the second warning area according to the earthquake level. The warning areas are assigned the priority of sending warnings, as follows:

[0010] Determine the warning to be sent to the first warning area to the second warning area The time when the warning is sent in the warning area;

[0011] Determine the first warning area to the Warning area wave and The time between waves;

[0012] If the earthquake magnitude is less than the foreshock threshold, the first warning area to the second warning area will be Warnings are sent in order of warning areas;

[0013] If the earthquake magnitude is greater than the foreshock threshold, the first Warning area, and from Warning area to The warning areas are sent out in order, from the first warning area to the - Delayed warning sending in the region.

[0014] Furthermore, the formula of the generalized intensity prediction model is:

[0015] ;

[0016] in, represents the intensity prediction value of the target location, Basic parameters of earthquake intensity at the epicenter, represents the first proportionality coefficient, Indicates the magnitude of the earthquake, represents the first nonlinear parameter, Represents the proportional coefficient of distance attenuation, Indicates the straight-line distance between the target point and the epicenter, Indicates the distance adjustment offset. represents the second nonlinear parameter, represents the second proportionality coefficient, represents the natural base, represents the epicenter depth, Represents the attenuation parameter.

[0017] Furthermore, the generalized intensity prediction model is differentiated and corrected according to the seismic zone to which it belongs, including:

[0018] Step 1, adding several correction layers to the generalized intensity prediction model, each correction layer corresponds to a corresponding earthquake zone;

[0019] Step 2, determining a coordinate set for each seismic zone, each coordinate set is the coordinate within the corresponding seismic zone;

[0020] Step 3: When the target location is within the earthquake zone with the same coordinates as the epicenter, the corresponding correction layer is activated. The calculation formula of the correction layer is as follows:

[0021] ;

[0022] represents the corrected intensity prediction value of the target location, Indicates the geometric correction coefficient of the seismic zone. When the epicenter coordinates are located in the center of the seismic zone , when the epicenter coordinates are located at the two ends of the earthquake zone , represents the decay rate parameter, represents the directional correction coefficient, Represents the angle between the target location and the propagation direction of the earthquake. represents the third coefficient, express function, represents the fourth weight coefficient, Represents the differential operation.

[0023] Furthermore, the parameters of the fitted intensity prediction model are optimized through the intensity attenuation model, as follows:

[0024] Step 4, randomly encode the custom parameters in the fitted intensity prediction model and initialize and generate several individuals;

[0025] Step 5: Compare the predicted intensity of the fitted intensity prediction model corresponding to each individual with the actual intensity of the intensity attenuation model, and take the sum of the square differences between the predicted intensity and the actual intensity at each target location as the objective function value;

[0026] Step 6: Sort the individuals from small to large according to the objective function value of each individual to obtain the first feature sorting; retain a preset number of individuals from the first feature sorting from front to back, and update the remaining individuals as parents for random crossover and mutation;

[0027] Step 7: Repeat step 6 until the preset number of updates is reached, and then output the individual with the smallest objective function value.

[0028] Furthermore, the earthquake area is divided into the first early warning area to the second early warning area. Warning areas include:

[0029] Step 8, inserting a number of target positions at fixed intervals into the target area, predicting the intensity of each target position by fitting the intensity prediction model, and obtaining the intensity prediction value of each target position;

[0030] Step 9: Perform K-means clustering on the intensity prediction values ​​to obtain clusters; and sort the average intensity prediction values ​​of each cluster from small to large to obtain the second feature sorting;

[0031] Step 10: sort the target area according to the second feature The clusters are divided into warning areas in order. The average intensity prediction value of the first warning area is the largest. The average predicted intensity value in the warning area is the smallest.

[0032] Further, it is determined that the warning is sent to the first warning area to the second warning area. The warning sending time of the warning area; the calculation formula of the warning sending time is:

[0033] ;

[0034] in, represents the sending time of the i-th warning area, represents the Euclidean distance from the warning center to the i-th warning area, represents the speed of light, represents the area of ​​the ith sub-region, Indicates the concurrent time delay of warning sending per unit area, represents the exponential amplification parameter, , e represents the natural base, A weight parameter representing the area of ​​the sub-region.

[0035] Further, the first warning area to the Warning area wave and The interval time of the wave; the calculation formula of the interval time is:

[0036] ;

[0037] in, represents the i-th warning area wave and The arrival time difference of the waves, express The speed of wave propagation, express The speed at which waves travel; and is the density coefficient for S-wave and P-wave attenuation; and is the directional attenuation coefficient of S-wave and P-wave; is the crust density of the ith warning area; is the distance from the ith warning area to the earthquake source; It is the angle between the direction of seismic wave propagation and the crustal layer.

[0038] Furthermore, according to the warning sending time and interval time, the first Warning areas include:

[0039] Manually set the warning validity time, and determine the time difference between the warning arrival time and the S wave arrival time of the Kth warning area as the warning validity time. The calculation formula is as follows:

[0040] ;

[0041] in, Indicates the effective time of the warning. Indicates the epicenter coordinates to The Euclidean distance of the warning area, represents the Euclidean distance from the epicenter coordinates to the monitoring point, represents the Euclidean distance from the monitoring point to the early warning center, Indicates that the warning center has Distance to the warning area.

[0042] The beneficial effect of the present invention is that by combining the intensity prediction model, dynamic area division and early warning sending optimization strategy, the accuracy and efficiency of earthquake early warning are achieved. By predicting the distribution characteristics of earthquake intensity, the early warning areas with clear priorities are divided, and early warnings are issued to the early warning areas according to different situations, so as to maximize the effectiveness of the early warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a module diagram of the earthquake early warning system based on the Haizhengtong platform of the present invention. DETAILED DESCRIPTION

[0044] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0045] like Figure 1 As shown in the figure, the earthquake early warning system based on the Haizhengtong platform includes:

[0046] The data acquisition module is used to obtain the seismic information of historical earthquakes through the Haizhengtong platform. The seismic information includes the epicenter coordinates, epicenter depth, earthquake level, and the corresponding intensity attenuation model and the seismic zone to which it belongs; wherein the intensity attenuation model represents an attenuation model in which the earthquake intensity changes with the change of the straight-line distance from the epicenter coordinates;

[0047] The intensity prediction module is used to construct a generalized intensity prediction model based on the epicenter coordinates, epicenter depth and earthquake level, and to perform differential correction on the generalized intensity prediction model according to the seismic zone to which it belongs, so as to obtain a fitted intensity prediction model; the output of the fitted intensity prediction model represents the predicted intensity of any coordinates with respect to the epicenter coordinates;

[0048] The regional division module is used to divide the earthquake area into the first warning area to the second warning area based on the epicenter depth and earthquake level, with the epicenter coordinates as the center, combined with the fitting intensity prediction model. Early warning area;

[0049] The warning sending module is used to send warning signals from the first warning area to the second warning area according to the earthquake level. The warning areas are assigned the priority of sending warnings, as follows:

[0050] Determine the warning to be sent to the first warning area to the second warning area The time when the warning is sent in the warning area;

[0051] Determine the first warning area to the Warning area wave and The time between waves;

[0052] If the earthquake magnitude is less than the foreshock threshold, the first warning area to the second warning area will be Warnings are sent in order of warning areas;

[0053] If the earthquake magnitude is greater than the foreshock threshold, the first Warning area, and from Warning area to The warning areas are sent out in order, from the first warning area to the - Delayed warning sending in the region.

[0054] Specifically, when the earthquake magnitude is less than the foreshock threshold (foreshocks are usually less than 5 on the Richter scale), it is judged that the earthquake may be a foreshock, and the foreshock is less destructive. Therefore, the first warning area (the area close to the epicenter) to the second warning area (the area close to the epicenter) are warned. Warnings are sent to warning areas in order, so that residents near the epicenter have enough time to evacuate. When the earthquake magnitude is greater than the foreshock threshold, the earthquake is judged to be a destructive earthquake (a large earthquake without foreshocks). At this time, giving priority warnings to residents near the epicenter has little effect (the warning information is sent after the earthquake). If a wave arrives, the warning priority area is changed so that residents in the subsequent warning areas have enough time to evacuate. This method reduces the delay of concurrent warnings and is conducive to providing effective warnings to as many residents as possible.

[0055] In one embodiment of the present invention, the formula of the generalized severity prediction model is:

[0056] ;

[0057] in, represents the intensity prediction value of the target location, Basic parameters of earthquake intensity at the epicenter, represents the first proportionality coefficient, Indicates the magnitude of the earthquake, represents the first nonlinear parameter, Represents the proportional coefficient of distance attenuation, Indicates the straight-line distance between the target point and the epicenter, Indicates the distance adjustment offset. represents the second nonlinear parameter, represents the second proportionality coefficient, represents the natural base, represents the epicenter depth, Represents the attenuation parameter.

[0058] Specifically, by constructing a generalized intensity prediction model, an accurate prediction of earthquake intensity in the target area is achieved. Based on key parameters such as epicenter location, earthquake depth and earthquake level, combined with distance attenuation characteristics and correction factors, the model dynamically calculates the earthquake intensity at any coordinate point, making up for the shortcomings of traditional intensity models in adapting to complex earthquake scenarios and differences in earthquake zones. By introducing nonlinear attenuation mechanisms and random error corrections, the model can more realistically reflect the intensity change laws of the epicenter and surrounding areas, provide a scientific basis for earthquake regional division and early warning strategies, significantly improve the accuracy and flexibility of earthquake intensity prediction, and has important application value in earthquake early warning and disaster response.

[0059] In one embodiment of the present invention, the generalized intensity prediction model is differentially corrected according to the seismic zone to which it belongs, including:

[0060] Step 1, adding several correction layers to the generalized intensity prediction model, each correction layer corresponds to a corresponding earthquake zone;

[0061] Step 2, determining a coordinate set for each seismic zone, each coordinate set is the coordinate within the corresponding seismic zone;

[0062] Step 3: When the target location is within the earthquake zone with the same coordinates as the epicenter, the corresponding correction layer is activated. The calculation formula of the correction layer is as follows:

[0063] ;

[0064] represents the corrected intensity prediction value of the target location, Indicates the geometric correction coefficient of the seismic zone. When the epicenter coordinates are located in the center of the seismic zone , when the epicenter coordinates are located at the two ends of the earthquake zone , represents the decay rate parameter, represents the directional correction coefficient, Represents the angle between the target location and the propagation direction of the earthquake. represents the third coefficient, express function, represents the fourth weight coefficient, Represents the differential operation.

[0065] Specifically, in order to improve the adaptability and accuracy of the intensity prediction model in different seismic zones, differentiated correction of the model is achieved by adding multiple correction layers. The specific method is: first, according to the geological characteristics of different seismic zones, an independent correction layer is added to each seismic zone, and each correction layer is responsible for the intensity correction of the corresponding seismic zone; secondly, a set of coordinate sets representing the characteristics of the seismic zone is determined for each seismic zone, and these coordinates are used as reference points for correction calculations; finally, when the target position and the epicenter position are located in the same seismic zone, the correction layer of the seismic zone is activated, and the intensity prediction value is adjusted by the calculation formula of the correction layer. The formula combines the geometric characteristics of the seismic zone, the directionality of earthquake propagation, the attenuation characteristics of the propagation speed, and the difference parameters between regions, and performs multi-dimensional correction on the intensity to ensure that the model can more realistically and accurately reflect the seismic intensity distribution characteristics of different seismic zones. This method significantly improves the regional adaptability and scientificity of the intensity prediction model, and provides a solid foundation for the precision of earthquake early warning.

[0066] In one embodiment of the present invention, the parameters of the fitted intensity prediction model are optimized through an intensity attenuation model, as follows:

[0067] Step 4, randomly encode the custom parameters in the fitted intensity prediction model and initialize and generate several individuals;

[0068] Step 5: Compare the predicted intensity of the fitted intensity prediction model corresponding to each individual with the actual intensity of the intensity attenuation model, and take the sum of the square differences between the predicted intensity and the actual intensity at each target location as the objective function value;

[0069] Step 6: Sort the individuals from small to large according to the objective function value of each individual to obtain the first feature sorting; retain a preset number of individuals from the first feature sorting from front to back, and update the remaining individuals as parents for random crossover and mutation;

[0070] Step 7: Repeat step 6 until the preset number of updates is reached, and then output the individual with the smallest objective function value.

[0071] Specifically, in order to improve the accuracy of the fitted intensity prediction model, the model parameters are optimized by combining the intensity attenuation model. The specific process is as follows: First, the key parameters (such as proportional coefficients, attenuation factors, etc.) in the fitted intensity prediction model are randomly initialized and encoded. The combination of these codes is called "individuals", and an initial population containing multiple individuals is generated. Next, the intensity prediction model corresponding to each individual is used to calculate the predicted intensity value of each location in the target area, and these predicted values ​​are compared with the actual intensity data. By calculating the sum of the square differences between the predicted values ​​and the actual values ​​of all target locations, the "objective function value" of the individual is obtained to measure the quality of the individual. Subsequently, all individuals are sorted from small to large according to the objective function value, and some individuals with better performance (smaller objective function values) are retained, and the rest of the individuals are eliminated. For the retained individuals, new individuals are generated through random crossover and mutation operations to form a new population, so as to further explore possible better parameter combinations. This process is repeated continuously, and after a preset number of iterations, the individual with the smallest objective function value is finally output, that is, the optimized model parameters. The core of this optimization process is to automatically search for the optimal combination of model parameters by simulating natural selection and genetic evolution, so that the output of the fitted intensity prediction model can be as close as possible to the actual data of the intensity attenuation model, thereby significantly improving the prediction accuracy and applicability of the model.

[0072] In one embodiment of the present invention, the earthquake region is divided into the first early warning area to the second early warning area. Warning areas include:

[0073] Step 8, inserting a number of target positions at fixed intervals into the target area, predicting the intensity of each target position by fitting the intensity prediction model, and obtaining the intensity prediction value of each target position;

[0074] Step 9: Perform K-means clustering on the intensity prediction values ​​to obtain clusters; and sort the average intensity prediction values ​​of each cluster from small to large to obtain the second feature sorting;

[0075] Step 10: sort the target area according to the second feature The clusters are divided into warning areas in order. The average intensity prediction value of the first warning area is the largest. The average predicted intensity value in the warning area is the smallest.

[0076] Specifically, in order to accurately divide the earthquake area, the target area is first divided into multiple target location points at fixed intervals, and the intensity value of each target location is calculated using the fitted intensity prediction model to obtain the intensity distribution in the entire area. Then, using the K-means clustering method, these intensity values ​​are divided into multiple clusters according to similarity, and each cluster corresponds to an intensity range. Next, the average intensity value of each cluster is calculated and sorted from large to small. Finally, based on the sorting results, the target area is divided into multiple warning areas, where the cluster with the largest average intensity value is defined as the first warning area, and the cluster with the smallest average intensity value is defined as the first warning area. Warning area. According to the distance attenuation of intensity, the first warning area is located at the epicenter, the second The warning area is located at the edge of the target area. The warning areas to the first warning area are nested layer by layer.

[0077] Specifically, the detailed steps of K-means clustering of intensity prediction values ​​are as follows: Initialize cluster centers: First, according to the required number of clusters, randomly select the intensity prediction values ​​of several target locations as the initial cluster centers. Each cluster center represents the core point of a cluster and is used to measure the distance between the intensity value of the target location and it. Calculate the distance to assign target locations: Compare the intensity prediction values ​​of all target locations with each cluster center, and assign the target location to the corresponding cluster based on the principle of the closest distance. The distance calculation method can be an absolute difference or other measurement method based on actual needs. Update cluster centers: For each cluster, calculate the average value of the intensity values ​​of all target locations in the cluster, and use the average value as the new cluster center. The new cluster center will be closer to the center position of the data in the current cluster. Repeat the allocation and update: Based on the new cluster center, repeat the above steps, redistribute the intensity values ​​of all target locations to the cluster with the closest distance, and update the center point of each cluster. This process will continuously optimize the clustering results, making the intensity values ​​within each cluster more similar. Determine whether it has converged: When the cluster center no longer changes significantly or reaches the preset number of iterations, stop the iteration and finally obtain a stable clustering result. Output clustering results: Divide all target locations in the target area into multiple clusters according to the cluster allocation results, and calculate the average intensity value of each cluster to provide a basis for subsequent regional division and priority setting. Through the above steps, the K-means clustering method can efficiently divide the intensity prediction values ​​of the target area into several groups, each group represents an area with a different intensity range, thereby providing accurate data support for earthquake regional division and the implementation of early warning strategies.

[0078] In one embodiment of the present invention, it is determined that the warning is sent to the first warning area to the second warning area. The warning sending time of the warning area; the calculation formula of the warning sending time is:

[0079] ;

[0080] in, represents the sending time of the i-th warning area, represents the Euclidean distance from the warning center to the i-th warning area, represents the speed of light, represents the area of ​​the ith sub-region, Indicates the concurrent time delay of warning sending per unit area, represents the exponential amplification parameter, , e represents the natural base, A weight parameter representing the area of ​​the sub-region.

[0081] Specifically, in order to reasonably determine the time when the warning information is sent to each warning area, a time calculation method that comprehensively considers multiple factors is designed. Specifically, the sending time of each warning area consists of the following parts: first, the time for the warning information to propagate from the warning center to the target area, which depends on the distance between the warning center and the target area and the signal propagation speed; second, due to the different sizes of the warning areas, the concurrent sending of warning information within the area may cause a certain delay, so it is necessary to adjust according to the size of the area and the concurrent sending delay parameter per unit area; finally, in order to cope with complex actual scenarios, such as uneven distribution density of equipment in the area, an exponential amplification parameter is introduced to dynamically correct the time. In this way, it can ensure that the warning information reaches each area accurately according to the regional priority and actual conditions, thereby realizing an efficient and reasonable warning distribution strategy.

[0082] In one embodiment of the present invention, the first warning area to the second warning area are respectively determined. Warning area wave and The interval time of the wave; the calculation formula of the interval time is:

[0083] ;

[0084] in, represents the i-th warning area wave and The arrival time difference of the waves, express The speed of wave propagation, express The speed at which waves travel; and is the density coefficient for S-wave and P-wave attenuation; and is the directional attenuation coefficient of S-wave and P-wave; is the crust density of the ith warning area; is the distance from the ith warning area to the earthquake source; It is the angle between the direction of seismic wave propagation and the crustal layer.

[0085] Specifically, in order to determine the interval time of seismic waves between the first warning area and the last warning area, a calculation method based on the propagation speed of seismic waves is designed. Specifically, seismic waves are divided into two types, fast waves and slow waves, which propagate in the crust at different speeds. For each warning area, the fast wave usually arrives earlier than the slow wave, and the time interval between the two is directly related to the distance from the warning center to the area. By calculating the propagation time difference between the fast wave and the slow wave, the time interval for the two waves to arrive at the area can be obtained, thereby providing a scientific basis for the earthquake warning strategy. The calculation of this time interval comprehensively considers the propagation speed of fast waves and slow waves, as well as the geographical differences of different warning areas, to ensure that the warning information can be effectively sent before the slow wave arrives and prompt users to take risk avoidance measures.

[0086] In one embodiment of the present invention, the first Warning areas include:

[0087] Manually set the warning validity time, and determine the time difference between the warning arrival time and the S wave arrival time of the Kth warning area as the warning validity time. The calculation formula is as follows:

[0088] ;

[0089] in, Indicates the effective time of the warning. Indicates the epicenter coordinates to The Euclidean distance of the warning area, represents the Euclidean distance from the epicenter coordinates to the monitoring point, represents the Euclidean distance from the monitoring point to the early warning center, Indicates that the warning center has Distance to the warning area.

[0090] Specifically, the effective warning time refers to the time difference between the earthquake warning information reaching the target warning area and the time left for people in the area to take risk avoidance measures. If the wave arrives after 3 seconds, then these 3 seconds are the effective time of the warning, which provides valuable reaction time for people to evacuate or take shelter. When calculating the effective time of the warning, it is necessary to comprehensively consider the distance from the epicenter to the target warning area, the distance from the epicenter to the monitoring point, the distance from the monitoring point to the warning center, and the distance from the warning center to the target warning area. These distances are combined with the propagation speed of seismic waves and the time when the warning information is sent to determine the time difference from the arrival of the warning information to the arrival of the seismic wave, ensuring that the time is sufficient to meet the effective risk avoidance needs.

[0091] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. The earthquake early warning system based on the Haizhengtong platform is characterized by: include: The data acquisition module is used to obtain the seismic information of historical earthquakes through the Haizhengtong platform. The seismic information includes the epicenter coordinates, epicenter depth, earthquake level, and the corresponding intensity attenuation model and the seismic zone to which it belongs; wherein the intensity attenuation model represents an attenuation model in which the earthquake intensity changes with the change of the straight-line distance from the epicenter coordinates; The intensity prediction module is used to construct a generalized intensity prediction model based on the epicenter coordinates, epicenter depth and earthquake level, and to perform differential correction on the generalized intensity prediction model according to the seismic zone to which it belongs, so as to obtain a fitted intensity prediction model; the output of the fitted intensity prediction model represents the predicted intensity of any coordinates with respect to the epicenter coordinates; Among them, the formula of the generalized intensity prediction model is as follows: ； in, represents the intensity prediction value of the target location, Basic parameters of earthquake intensity at the epicenter, represents the first proportionality coefficient, Indicates the magnitude of the earthquake, represents the first nonlinear parameter, Represents the proportional coefficient of distance attenuation, Indicates the straight-line distance between the target point and the epicenter, Indicates the distance adjustment offset, represents the second nonlinear parameter, represents the second proportionality coefficient, represents the natural base, represents the epicenter depth, represents the attenuation parameter; the regional division module is used to divide the earthquake area into the first warning area to the second warning area based on the epicenter depth and earthquake level, with the epicenter coordinates as the center, combined with the fitting intensity prediction model Warning area; The warning sending module is used to send warning signals from the first warning area to the second warning area according to the earthquake level. The warning areas are assigned the priority of sending warnings, as follows: Determine the warning to be sent to the first warning area to the second warning area The time when the warning is sent in the warning area; Determine the first warning area to the Warning area wave and The time between waves; If the earthquake magnitude is less than the foreshock threshold, the first warning area to the second warning area will be Warnings are sent in order of warning areas; If the earthquake magnitude is greater than the foreshock threshold, the first Warning area, and from Warning area to The warning areas are sent out in order, from the first warning area to the - Delayed warning sending in the region.

2. The earthquake early warning system based on the Haizhengtong platform according to claim 1 is characterized in that: According to the seismic zone to which it belongs, the generalized intensity prediction model is differentiated and corrected, including: Step 1, adding several correction layers to the generalized intensity prediction model, each correction layer corresponds to a corresponding earthquake zone; Step 2, determining a coordinate set for each seismic zone, each coordinate set is the coordinate within the corresponding seismic zone; Step 3: When the target location is within the earthquake zone with the same coordinates as the epicenter, the corresponding correction layer is activated. The calculation formula of the correction layer is as follows: ； represents the corrected intensity prediction value of the target location, Indicates the geometric correction coefficient of the seismic zone. When the epicenter coordinates are located in the center of the seismic zone , when the epicenter coordinates are located at the two ends of the earthquake zone , represents the decay rate parameter, represents the directional correction coefficient, Represents the angle between the target location and the propagation direction of the earthquake. represents the third coefficient, express function, represents the fourth weight coefficient, Represents the differential operation.

3. The earthquake early warning system based on the Haizhengtong platform according to claim 2 is characterized in that: The parameters of the fitted intensity prediction model are optimized through the intensity attenuation model, as follows: Step 4, randomly encode the custom parameters in the fitted intensity prediction model and initialize and generate several individuals; Step 5: Compare the predicted intensity of the fitted intensity prediction model corresponding to each individual with the actual intensity of the intensity attenuation model, and take the sum of the square differences between the predicted intensity and the actual intensity at each target location as the objective function value; Step 6: Sort the individuals from small to large according to the objective function value of each individual to obtain the first feature sorting; retain a preset number of individuals from the first feature sorting from front to back, and update the remaining individuals as parents for random crossover and mutation; Step 7: Repeat step 6 until the preset number of updates is reached, and then output the individual with the smallest objective function value.

4. The earthquake early warning system based on the Haizhengtong platform according to claim 3 is characterized in that: The earthquake zone is divided into the first warning area to the second warning area. Warning areas include: Step 8, inserting a number of target positions at fixed intervals into the target area, predicting the intensity of each target position by fitting the intensity prediction model, and obtaining the intensity prediction value of each target position; Step 9: Perform K-means clustering on the intensity prediction values ​​to obtain clusters; and sort the average intensity prediction values ​​of each cluster from small to large to obtain the second feature sorting; Step 10: sort the target area according to the second feature The clusters are divided into warning areas in order. The average intensity prediction value of the first warning area is the largest. The average predicted intensity value in the warning area is the smallest.

5. The earthquake early warning system based on the Haizhengtong platform according to claim 4 is characterized in that: Determine the warning to be sent to the first warning area to the second warning area The warning sending time of the warning area; the calculation formula of the warning sending time is: ； in, represents the sending time of the i-th warning area, represents the Euclidean distance from the warning center to the i-th warning area, represents the speed of light, represents the area of ​​the ith sub-region, Indicates the concurrent time delay of warning sending per unit area, represents the exponential amplification parameter, , e represents the natural base, A weight parameter representing the area of ​​the sub-region.

6. The earthquake early warning system based on the Haizhengtong platform according to claim 5 is characterized in that: Determine the first warning area to the Warning area wave and The interval time of the wave; the calculation formula of the interval time is: ； in, represents the i-th warning area wave and The arrival time difference of the waves, express The speed of wave propagation, express The speed at which waves travel; and is the density coefficient for the attenuation of S and P waves; and is the directional attenuation coefficient of S-wave and P-wave; is the crust density of the ith warning area; is the distance from the ith warning area to the earthquake source; It is the angle between the direction of seismic wave propagation and the crustal layer.

7. The earthquake early warning system based on the Haizhengtong platform according to claim 6 is characterized in that: According to the warning sending time and interval time, the Warning areas include: Manually set the warning validity time, and determine the time difference between the warning arrival time and the S wave arrival time of the Kth warning area as the warning validity time. The calculation formula is as follows: ； in, Indicates the effective time of the warning. Indicates the epicenter coordinates to The Euclidean distance of the warning area, represents the Euclidean distance from the epicenter coordinates to the monitoring point, represents the Euclidean distance from the monitoring point to the early warning center, Indicates that the warning center has Distance to the warning area.

Images