Lightning early warning system
By combining atmospheric electric field, lightning positioning and radar data, and using cluster analysis and neural network technologies, the problems of high error rate and poor detection range accuracy of the lightning warning system are solved, and efficient and accurate lightning warning and risks are achieved. Level determination meets the lightning protection and disaster reduction needs of social units.
Patent Information
- Application Number
- CN202411060278.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-05
- Publication Date
- 2024-11-15
AI Technical Summary
The existing lightning warning system has high pre-alarm error rate and poor accuracy in lightning detection range, which cannot meet the needs of social units for lightning protection, disaster reduction and lightning protection safety.
The atmospheric electric field data acquisition module, lightning positioning data acquisition module, radar data acquisition module, thunderstorm occurrence judgment unit, cluster analysis module and thunderstorm movement trajectory prediction module are used to comprehensively judge the thunderstorm occurrence and movement trajectory, and use the cumulative difference method and clustering Technologies such as analysis, hierarchy and neural networks improve the accuracy and efficiency of lightning warnings.
Real-time and accurate lightning risk level judgment and early warning information transmission are achieved, the false reporting rate and underreport rate are reduced, and the effectiveness of the early warning system and the stability of the production work of the user unit are improved.
Smart Images

Figure CN118962261A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of lightning monitoring, and in particular relates to a lightning early warning system. Background Art
[0002] The existing lightning warning systems mainly include lightning warning systems based on atmospheric electric field meters and lightning warning systems based on weather radars.
[0003] The defect of the lightning warning system based on the atmospheric electric field meter is that the electric field measurement of the ground atmospheric electric field meter is greatly affected by the surrounding environment and other external interference factors, so the warning rate (POD) and missed warning rate (FTWR) of the lightning warning system based on the atmospheric electric field meter within the warning range appear to be acceptable on the surface, but its false alarm rate (FAR) is too high and the real effective alarm rate (EAR, the real warning accuracy rate) is too low, which makes the early warning system's advance notice and advance alarm misjudgment rate high and the real accuracy rate too low, resulting in the lightning warning system based on the atmospheric electric field meter cannot meet the social units' warning needs for lightning disaster reduction and lightning safety.
[0004] The defect of the lightning warning system based on weather radar is that since the lightning location system detects lightning that has already occurred, it has a large detection range (hundreds of kilometers) and can accurately locate the lightning strike site, but its detection efficiency varies with the detection range. When the detection range is within the 25km lightning warning range of the social unit, the detection efficiency is very low, so that a large number of real lightning within the lightning warning range cannot be detected. As a result, the lightning warning system based on the lightning location system cannot meet the actual needs of social units for lightning disaster reduction and lightning safety in terms of accurate advance notification and accurate advance warning. Summary of the invention
[0005] In view of the problems that the existing lightning warning system has a high misjudgment rate of pre-alarm and a poor accuracy of lightning detection range, the present invention provides a lightning warning system.
[0006] In order to achieve the above technical objectives, the technical solution adopted by the present invention is as follows:
[0007] A lightning warning system, comprising an atmospheric electric field data acquisition module, a lightning location data acquisition module, a radar data acquisition module, a thunderstorm occurrence judgment unit, a cluster analysis module and a thunderstorm movement trajectory prediction module;
[0008] Atmospheric electric field data acquisition module, used to obtain atmospheric electric field data;
[0009] A lightning location data acquisition module is used to obtain lightning location data;
[0010] Radar data acquisition module, used to obtain data on radar echo intensity, vertical liquid water content and echo top height;
[0011] The thunderstorm occurrence judgment unit comprehensively judges whether a thunderstorm has occurred through the collected atmospheric electric field data, lightning location data, radar echo intensity, vertical liquid water content and echo top height data;
[0012] Cluster analysis module, cluster analysis focuses on thunderstorm information in the area;
[0013] The thunderstorm relative motion information judgment module uses the atmospheric electric field time series data to judge the thunderstorm relative motion information;
[0014] The thunderstorm movement trajectory prediction module uses lightning location data to predict the thunderstorm movement trajectory.
[0015] Further, the thunderstorm occurrence judgment unit includes a first thunderstorm occurrence judgment module, a second thunderstorm occurrence judgment module and a third thunderstorm occurrence judgment module;
[0016] The first thunderstorm occurrence judgment module calculates the cumulative difference value through the collected atmospheric electric field data to judge whether there is a thunderstorm in the warning area;
[0017] The second thunderstorm occurrence judgment module inputs the collected lightning location data into the cluster analysis module, and the cluster analysis focuses on the thunderstorm information in the area;
[0018] The third thunderstorm occurrence judgment module uses the collected radar echo intensity, vertical liquid water content and echo top height data to construct a judgment matrix using the hierarchical analysis method and calculate the maximum eigenvalue λ of the judgment matrix max The relative weight coefficients of indicators in the same layer are obtained by using the eigenvector corresponding to the maximum eigenvalue, and the correlation matrix of radar echo intensity, vertical liquid water content, and echo top height is used to judge the occurrence of thunderstorms.
[0019] Furthermore, the first thunderstorm occurrence judgment module uses the cumulative difference method to capture the electric field jump phenomenon before the occurrence of lightning, and the calculation formula is:
[0020]
[0021] Where E is the atmospheric electric field value, t0 and t1 are two adjacent moments, the interval is the sampling time interval of the atmospheric electric field instrument, t-t0=15min; when E(t) reaches the set threshold, it can be considered that a thunderstorm has occurred in the warning area, and the selection of the threshold needs to be determined according to local meteorological and geographical conditions.
[0022] Furthermore, the detailed steps of clustering and analyzing thunderstorm information in the focus area in the second thunderstorm occurrence judgment module include:
[0023] S201, input parameters: domain radius ε, threshold MinPts, lightning location dataset D;
[0024] S202, randomly select a point p in the data set D, and determine whether it is a core point;
[0025] S203, if point p is a core point, find all points in the neighborhood that are directly density-reachable;
[0026] S204: Whether the points in the data set D have been judged, if not, repeat steps S202 to S204;
[0027] S205, merge density-reachable points and expand the cluster;
[0028] S206: Output the target cluster set.
[0029] Furthermore, the detailed steps of judging the occurrence of thunderstorms using the correlation matrix of the third thunderstorm occurrence judgment module include:
[0030] Construct a judgment matrix:
[0031] The nine-level scaling method (1-9 scaling method) is used to compare the indicators in pairs, determine the relative importance of the indicators and give the corresponding ratios, and the resulting judgment matrix A;
[0032] Calculate the relative weight:
[0033] By calculating the maximum eigenvalue λ of the judgment matrix A max and the eigenvector W corresponding to the maximum eigenvalue, and the relative weight coefficient of the indicators at the same level is obtained;
[0034] Consistency check:
[0035] When the order of the judgment matrix obtained by pairwise comparison is greater than or equal to 3, a consistency check is required.
[0036] Consistency Test Index (CI):
[0037] According to the order of the judgment matrix, the corresponding RI is searched in the table.
[0038] According to the values of CI and RI, calculate the consistency ratio CR: CR = CI / RI
[0039] When CR is less than or equal to 0.1, the consistency of the judgment matrix A meets the requirements. Otherwise, the pairwise comparison values of the judgment matrix need to be adjusted.
[0040] Furthermore, the detailed steps of using atmospheric electric field time series data to determine thunderstorm relative motion information include:
[0041] Collect atmospheric electric field data, manually mark the data of thunderstorms close to and far from the warning area, and create a data set;
[0042] Establish and train a thunderstorm relative motion recognition model based on a one-dimensional convolutional neural network;
[0043] The atmospheric electric field data to be predicted is input into the recognition model to obtain the movement direction of the thunderstorm relative to the atmospheric electric field instrument.
[0044] Furthermore, the detailed steps of using lightning location data to predict the movement trajectory of thunderstorms include:
[0045] According to the thunderstorm information output results in the area of interest from cluster analysis, the centroid of the lightning cluster is calculated using formula (1) to establish the thunderstorm movement path dataset:
[0046]
[0047] Establish and train a thunderstorm path prediction model based on GRU recurrent neural network;
[0048] The data is input into the thunderstorm path prediction model to predict the movement path of the thunderstorm;
[0049] The moving path is corrected using the calculation results of the moving direction of the thunderstorm relative to the atmospheric electric field instrument.
[0050] A lightning early warning method comprises the following steps:
[0051] S1. Obtain atmospheric electric field data, calculate the cumulative difference value to determine whether a thunderstorm occurs in the warning area, as one of the criteria for determining the occurrence of a thunderstorm;
[0052] S2, obtain lightning location data, and cluster analyze thunderstorm information in the area of interest;
[0053] S3, obtain the data of radar echo intensity, vertical liquid water content, and echo top height, use the hierarchical analysis method to construct a judgment matrix, and calculate the maximum eigenvalue λ of the judgment matrix max The relative weight coefficients of indicators in the same layer are obtained by using the eigenvector corresponding to the maximum eigenvalue, and the correlation matrix of radar echo intensity, vertical liquid water content, and echo top height is used to judge the occurrence of thunderstorms.
[0054] S4. Use atmospheric electric field time series data to determine thunderstorm relative motion information;
[0055] S5. Use the lightning location data to predict the thunderstorm movement trajectory, and use the result of step S4 to correct it.
[0056] Compared with the prior art, the present invention has the following beneficial effects:
[0057] Using lightning monitoring, radar, and real-time atmospheric electric field monitoring data, the cloud platform automatically calculates and determines the lightning risk level in the target area in real time according to different weight coefficients. When the risk level increases, the system can automatically send level 1, 2, and 3 lightning warning information to users in the target area. At the same time, if the system detects that all values have returned to normal and lasts for more than 30 minutes, it will automatically release information to lift the warning status, effectively avoiding the impact on the production work of user units due to untimely warnings and untimely lifting of warnings. BRIEF DESCRIPTION OF THE DRAWINGS
[0058] Figure 1 This is a structural block diagram of a lightning warning system in an embodiment of the present invention;
[0059] Figure 2 The present invention is a flowchart of a lightning warning method according to an embodiment of the present invention. DETAILED DESCRIPTION
[0060] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.
[0061] like Figure 1 As shown, this embodiment provides a lightning warning system, including an atmospheric electric field data acquisition module, a lightning location data acquisition module, a radar data acquisition module, a thunderstorm occurrence judgment unit, a cluster analysis module and a thunderstorm movement trajectory prediction module;
[0062] The atmospheric electric field data acquisition module is used to obtain atmospheric electric field data; specifically, an atmospheric electric field instrument is used.
[0063] The lightning location data acquisition module is used to obtain lightning location data; specifically, a lightning locator is used.
[0064] The radar data acquisition module is used to obtain data on radar echo intensity, vertical liquid water content and echo top height; specifically, radar is used.
[0065] The thunderstorm occurrence judgment unit comprehensively judges whether a thunderstorm has occurred through the collected atmospheric electric field data, lightning location data, radar echo intensity, vertical liquid water content and echo top height data;
[0066] Cluster analysis module, cluster analysis focuses on thunderstorm information in the area;
[0067] The thunderstorm relative motion information judgment module uses the atmospheric electric field time series data to judge the thunderstorm relative motion information;
[0068] The thunderstorm movement trajectory prediction module uses lightning location data to predict the thunderstorm movement trajectory.
[0069] The thunderstorm occurrence judgment unit includes a first thunderstorm occurrence judgment module, a second thunderstorm occurrence judgment module and a third thunderstorm occurrence judgment module;
[0070] The first thunderstorm occurrence judgment module calculates the cumulative difference value through the collected atmospheric electric field data to judge whether there is a thunderstorm in the warning area;
[0071] The second thunderstorm occurrence judgment module inputs the collected lightning location data into the cluster analysis module, and the cluster analysis focuses on the thunderstorm information in the area;
[0072] The third thunderstorm occurrence judgment module uses the collected radar echo intensity, vertical liquid water content and echo top height data to construct a judgment matrix using the hierarchical analysis method and calculate the maximum eigenvalue λ of the judgment matrix max The relative weight coefficients of indicators in the same layer are obtained by using the eigenvector corresponding to the maximum eigenvalue, and the correlation matrix of radar echo intensity, vertical liquid water content, and echo top height is used to judge the occurrence of thunderstorms.
[0073] Furthermore, the first thunderstorm occurrence judgment module uses the cumulative difference method to capture the electric field jump phenomenon before the occurrence of lightning, and the calculation formula is:
[0074]
[0075] Where E is the atmospheric electric field value, t0 and t1 are two adjacent moments, the interval is the sampling time interval of the atmospheric electric field instrument, t-t0=15min; when E(t) reaches the set threshold, it can be considered that a thunderstorm has occurred in the warning area, and the selection of the threshold needs to be determined according to local meteorological and geographical conditions.
[0076] The detailed steps of clustering and analyzing thunderstorm information in the focus area in the second thunderstorm occurrence judgment module include:
[0077] S201, input parameters: domain radius ε, threshold MinPts, lightning location dataset D;
[0078] S202, randomly select a point p in the data set D, and determine whether it is a core point;
[0079] S203, if point p is a core point, find all points in the neighborhood that are directly density-reachable;
[0080] S204: Whether the points in the data set D have been judged, if not, repeat steps S202 to S204;
[0081] S205, merge density-reachable points and expand the cluster;
[0082] S206: Output the target cluster set.
[0083] The detailed steps of the third thunderstorm occurrence judgment module correlation matrix for judging the occurrence of thunderstorms include:
[0084] Construct a judgment matrix:
[0085] The nine-level scaling method (1-9 scaling method) is used to compare the indicators in pairs, determine the relative importance of the indicators and give the corresponding ratios. The resulting judgment matrix A is shown in Table 1.
[0086] Table 1 Pairwise comparison assignment table
[0087]
[0088] The judgment matrix A obtained by the above process is:
[0089]
[0090] Among them: a ii =1,a ji =1 / a ij
[0091] Calculate the relative weight:
[0092] By calculating the maximum eigenvalue λ of the judgment matrix A max and the eigenvector W corresponding to the maximum eigenvalue, and the relative weight coefficient of the indicators at the same level is obtained;
[0093] Consistency check:
[0094] When the order of the judgment matrix obtained by pairwise comparison is greater than or equal to 3, a consistency check is required.
[0095] Consistency Test Index (CI):
[0096] According to the order of the judgment matrix, the corresponding RI is searched in the table.
[0097] According to the values of CI and RI, calculate the consistency ratio CR: CR = CI / RI
[0098] When CR is less than or equal to 0.1, the consistency of the judgment matrix A meets the requirements. Otherwise, the pairwise comparison values of the judgment matrix need to be adjusted. See Table 2 below:
[0099] Table 2 Average random consistency index value table
[0100]
[0101] The detailed steps of using atmospheric electric field time series data to determine thunderstorm relative motion information include:
[0102] Collect atmospheric electric field data, manually mark the data of thunderstorms close to and far from the warning area, and create a data set;
[0103] Establish and train a thunderstorm relative motion recognition model based on a one-dimensional convolutional neural network;
[0104] The atmospheric electric field data to be predicted is input into the recognition model to obtain the movement direction of the thunderstorm relative to the atmospheric electric field instrument.
[0105] The detailed steps of using lightning location data to predict thunderstorm movement tracks include:
[0106] According to the thunderstorm information output results in the area of interest from cluster analysis, the centroid of the lightning cluster is calculated using formula (1) to establish the thunderstorm movement path dataset:
[0107]
[0108] Establish and train a thunderstorm path prediction model based on GRU recurrent neural network;
[0109] The data is input into the thunderstorm path prediction model to predict the movement path of the thunderstorm;
[0110] The moving path is corrected using the calculation results of the moving direction of the thunderstorm relative to the atmospheric electric field instrument.
[0111] like Figure 2 As shown, a lightning warning method comprises the steps of:
[0112] S1. Obtain atmospheric electric field data, calculate the cumulative difference value to determine whether a thunderstorm occurs in the warning area, as one of the criteria for determining the occurrence of a thunderstorm;
[0113] S2, obtain lightning location data, and cluster analyze thunderstorm information in the area of interest;
[0114] S3, obtain the data of radar echo intensity, vertical liquid water content, and echo top height, use the hierarchical analysis method to construct a judgment matrix, and calculate the maximum eigenvalue λ of the judgment matrix max The relative weight coefficients of indicators in the same layer are obtained by using the eigenvector corresponding to the maximum eigenvalue, and the correlation matrix of radar echo intensity, vertical liquid water content, and echo top height is used to judge the occurrence of thunderstorms.
[0115] S4. Use atmospheric electric field time series data to determine thunderstorm relative motion information;
[0116] S5. Use the lightning location data to predict the thunderstorm movement trajectory, and use the result of step S4 to correct it.
[0117] Compared with the prior art, the present invention has the following beneficial effects:
[0118] Using lightning monitoring, radar, and real-time atmospheric electric field monitoring data, the cloud platform automatically calculates and determines the lightning risk level in the target area in real time according to different weight coefficients. When the risk level increases, the system can automatically send level 1, 2, and 3 lightning warning information to users in the target area. At the same time, if the system detects that all values have returned to normal and lasts for more than 30 minutes, it will automatically release information to lift the warning status, effectively avoiding the impact on the production work of user units due to untimely warnings and untimely lifting of warnings.
[0119] Since weather radar data has a high temporal and spatial resolution, it can effectively identify and track thunderstorm clouds, and the occurrence of lightning has a good correlation with the strong echo characteristics of weather radar. Therefore, weather radar data provides relatively complete spatial structure and sufficiently dense data in time for lightning warning and forecasting. In addition, the development of a new generation of weather radar puzzle systems has realized high-frequency, high-precision and multi-product real-time networking puzzles, expanding the application scope of the data. The use of "weather radar + lightning location network + atmospheric electric field instrument" monitoring data is closely integrated with weather forecast business work, playing an important role in improving regional short-term severe weather prediction and warning capabilities. Therefore, radar data has become an important basic data in the research of lightning warning technology.
[0120] The above is a detailed introduction to a lightning warning system provided by the present application. The description of the specific embodiments is only used to help understand the method and its core idea of the present application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.
Claims
1. A lightning warning system, characterized in that: It includes an atmospheric electric field data acquisition module, a lightning location data acquisition module, a radar data acquisition module, a thunderstorm occurrence judgment unit, a cluster analysis module and a thunderstorm movement trajectory prediction module; Atmospheric electric field data acquisition module, used to obtain atmospheric electric field data; A lightning location data acquisition module is used to obtain lightning location data; Radar data acquisition module, used to obtain data on radar echo intensity, vertical liquid water content and echo top height; The thunderstorm occurrence judgment unit comprehensively judges whether a thunderstorm has occurred through the collected atmospheric electric field data, lightning location data, radar echo intensity, vertical liquid water content and echo top height data; Cluster analysis module, cluster analysis focuses on thunderstorm information in the area; The thunderstorm relative motion information judgment module uses the atmospheric electric field time series data to judge the thunderstorm relative motion information; The thunderstorm movement trajectory prediction module uses lightning location data to predict the thunderstorm movement trajectory.
2. A lightning warning system according to claim 1, characterized in that: The thunderstorm occurrence judgment unit includes a first thunderstorm occurrence judgment module, a second thunderstorm occurrence judgment module and a third thunderstorm occurrence judgment module; The first thunderstorm occurrence judgment module calculates the cumulative difference value through the collected atmospheric electric field data to judge whether there is a thunderstorm in the warning area; The second thunderstorm occurrence judgment module inputs the collected lightning location data into the cluster analysis module, and the cluster analysis focuses on the thunderstorm information in the area; The third thunderstorm occurrence judgment module uses the collected radar echo intensity, vertical liquid water content and echo top height data to construct a judgment matrix using the hierarchical analysis method and calculate the maximum eigenvalue λ of the judgment matrix max The relative weight coefficients of indicators in the same layer are obtained by using the eigenvector corresponding to the maximum eigenvalue, and the correlation matrix of radar echo intensity, vertical liquid water content, and echo top height is used to judge the occurrence of thunderstorms.
3. A lightning warning system according to claim 2, characterized in that: In the first thunderstorm occurrence judgment module, the cumulative difference method is used to capture the electric field jump phenomenon before lightning occurs. The calculation formula is: Where E is the atmospheric electric field value, t0 and t1 are two adjacent moments, and the interval is the sampling time interval of the atmospheric electric field instrument.
4. A lightning warning system according to claim 3, characterized in that: The detailed steps of clustering and analyzing thunderstorm information in the focus area in the second thunderstorm occurrence judgment module include: S201, input parameters: domain radius ε, threshold MinPts, lightning location dataset D; S202, randomly select a point p in the data set D, and determine whether it is a core point; S203, if point p is a core point, find all points in the neighborhood that are directly density-reachable; S204: Whether the points in the data set D have been judged, if not, repeat steps S202 to S204; S205, merge density-reachable points and expand the cluster; S206: Output the target cluster set.
5. A lightning warning system according to claim 4, characterized in that: The detailed steps of the third thunderstorm occurrence judgment module correlation matrix for judging the occurrence of thunderstorms include: Construct a judgment matrix: The nine-level scaling method is used to compare the indicators in pairs, determine the relative importance of the indicators and give the corresponding ratios, and the resulting judgment matrix A; Calculate the relative weight: By calculating the maximum eigenvalue λ of the judgment matrix A max and the eigenvector W corresponding to the maximum eigenvalue, and the relative weight coefficient of the indicators at the same level is obtained; Consistency check: When the order of the judgment matrix obtained by pairwise comparison is greater than or equal to 3, a consistency check is required.
6. A lightning warning system according to claim 5, characterized in that: Furthermore, the detailed steps of using atmospheric electric field time series data to determine thunderstorm relative motion information include: Collect atmospheric electric field data, manually mark the data of thunderstorms close to and far from the warning area, and create a data set; Establish and train a thunderstorm relative motion recognition model based on a one-dimensional convolutional neural network; The atmospheric electric field data to be predicted is input into the recognition model to obtain the movement direction of the thunderstorm relative to the atmospheric electric field instrument.
7. A lightning warning system according to claim 6, characterized in that: Furthermore, the detailed steps of using lightning location data to predict the movement trajectory of thunderstorms include: According to the thunderstorm information output results in the area of interest from cluster analysis, the centroid of the lightning cluster is calculated using formula (1) to establish the thunderstorm movement path dataset: Establish and train a thunderstorm path prediction model based on GRU recurrent neural network; The data is input into the thunderstorm path prediction model to predict the movement path of the thunderstorm; The moving path is corrected using the calculation results of the moving direction of the thunderstorm relative to the atmospheric electric field instrument.
8. A lightning early warning method, characterized in that: Includes steps: S1. Obtain atmospheric electric field data, calculate the cumulative difference value to determine whether a thunderstorm occurs in the warning area, as one of the criteria for determining the occurrence of a thunderstorm; S2, obtain lightning location data, and cluster analyze thunderstorm information in the area of interest; S3, obtain the data of radar echo intensity, vertical liquid water content, and echo top height, use the hierarchical analysis method to construct a judgment matrix, and calculate the maximum eigenvalue λ of the judgment matrix max and the eigenvector corresponding to the maximum eigenvalue, and the relative weight coefficient of the indicators in the same layer is obtained. The correlation matrix of radar echo intensity, vertical liquid water content, and echo top height is used to judge the occurrence of thunderstorms. S4. Use atmospheric electric field time series data to determine thunderstorm relative motion information; S5. Use the lightning location data to predict the thunderstorm movement trajectory, and use the result of step S4 to correct it.
Citation Information
Cited By
Fixed-point thunder and lightning early warning method based on multi-source data
CN120178384A
