Typhoon "Time-Space-Form-Quantity" Similarity Forecasting Method and Device, and Electronic Equipment

Abstract

The present invention discloses a typhoon "time - space - shape - quantity" similarity prediction method and device, including: obtaining historical typhoon data and current real - time typhoon data, both of which at least include typhoon attribute characteristics; based on the current real - time typhoon data, screening historical typhoon data with similar time in the historical typhoon database, and then screening historical typhoon data with similar spatial positions according to the spatial positions, so as to obtain a typhoon data set with similar time and space; performing geometric shape feature analysis on the typhoon data set with similar time and space to obtain geometric features; performing proximity analysis on the typhoon data set with similar time and space to obtain proximity; selecting geometric features, proximity, and typhoon attribute characteristics to represent the main characteristics of typhoons, and constructing a typhoon similarity set - pair analysis matrix; forming set - pairs with the current real - time typhoon and the historical typhoons with similar time and space in the typhoon set - pair analysis matrix, and performing set - pair similarity analysis and decision - making optimization; according to the results of set - pair similarity analysis and optimization, performing similarity prediction.

Description

Technical Field

[0001] This application relates to the field of disaster prevention and mitigation, and particularly to a typhoon "time-space-shape-quantity" similarity prediction method and device, and an electronic device. Background Art

[0002] A typhoon is one of the most serious natural disasters in the world and is a highly destructive disaster weather system. Correctly predicting the future trajectory of a typhoon is of extremely important significance for minimizing typhoon disaster losses, making early preparations, and taking early defenses. Currently, typhoon prediction methods mainly include two categories: numerical prediction and statistical prediction. Numerical prediction predicts the movement trend and activity characteristics of a typhoon based on large-scale meteorological numerical calculations. Statistical prediction, on the basis of historical typhoon data, uses methods such as probability, similarity, regression analysis, and climatological persistence prediction to predict typhoons. However, due to the complexity and uncertainty of influencing factors, the relationship between typhoon movement and its influencing factors is a highly non-linear relationship, resulting in high complexity, great difficulty, and low reliability in prediction.

[0003] As a kind of statistical prediction method, the similarity prediction method mainly predicts the current typhoon through a similarity typhoon matching algorithm. Most of the current similarity criteria mainly use the average distance between corresponding control points as the main reference basis, without considering the similarity degree of the path geometric shape. The influencing factors considered by the method itself are not comprehensive, and at the same time, the dynamic changes of typhoon similarity characteristics are not considered. The reliability and accuracy of the similarity prediction results need to be improved. Summary of the Invention

[0004] The purpose of the embodiments of this application is to provide a typhoon "time-space-shape-quantity" similarity prediction method and device, and an electronic device, so as to solve the technical problem that the reliability and accuracy of the similarity prediction results need to be improved due to the insufficient utilization of typhoon historical data characteristics in the existing similarity prediction technology.

[0005] According to the first aspect of the embodiments of this application, a typhoon path "time-space-shape-quantity" similarity prediction method is provided, including:

[0006] S1. Obtain historical typhoon data and current real-time typhoon data. Both types of data at least include typhoon attribute characteristics. Store the historical typhoon data in a historical typhoon database, and store the current real-time typhoon data in a real-time typhoon database;

[0007] S2. Based on the current real-time typhoon data, screen historical typhoon data with similar time in the historical typhoon database, and then screen historical typhoon data with similar spatial positions from the screened historical typhoon data with similar time, so as to obtain a typhoon data set with similar time and space;

[0008] S3. Perform geometric feature analysis on the typhoon data set with spatio-temporal similarity to obtain geometric features;

[0009] S4. Perform closeness analysis on the typhoon data set with spatio-temporal similarity to obtain closeness;

[0010] S5. Select the geometric features, closeness, and typhoon attribute features to characterize the main features of typhoons, and construct a typhoon similarity set pair analysis matrix;

[0011] S6. In the typhoon set pair analysis matrix, form a set pair with the current real-time typhoon and the historical typhoons with spatio-temporal similarity, and conduct set pair similarity analysis and decision-making optimization;

[0012] S7. According to the results of the set pair similarity analysis and optimization, conduct similarity prediction on typhoons.

[0013] Optionally, based on the current real-time typhoon data, screen the historical typhoon data with time similarity in the historical typhoon database according to time, and then screen the historical typhoon data with spatial position similarity according to spatial position in the screened historical typhoon data with time similarity, so as to obtain a typhoon data set with spatio-temporal similarity, including:

[0014] S21: Taking the month when the current real-time typhoon in the real-time typhoon database occurs as a benchmark, screen in the historical typhoon database according to time to obtain a typhoon data set that meets time similarity;

[0015] S22: Apply the GIS spatial analysis function, use the wind circle radius of the current typhoon at the current moment in the real-time typhoon database as the buffer radius, and construct a line buffer area with the real-time path backtracking several time periods from the current moment. According to the line buffer area, conduct spatial intersection calculation on the typhoons with time similarity to obtain a typhoon data set with spatio-temporal similarity.

[0016] Optionally, perform geometric feature analysis on the typhoon data set with spatio-temporal similarity to obtain slope, curvature, and turning angle, including:

[0017] S31: Calculate the slope of the latest time period of the path of the current real-time typhoon T0 and the paths of the typhoon data set T with spatio-temporal similarity respectively to determine the parallel characteristics between the paths and identify whether the evolution space directions of the typhoon paths are consistent; i to identify whether the evolution space directions of the typhoon paths are consistent;

[0018] S32: Calculate the curvature of the curve (P0 - P1 - P2) formed by the latest 3 points in time of the path of the current typhoon T0 and the paths of the typhoon data set T with spatio-temporal similarity respectively to identify the convexity and concavity of the geometric curve, where the curvature at point P1 represents the curvature of the curve (P0 - P1 - P2); i where the curvature at point P1 represents the curvature of the curve (P0 - P1 - P2);

[0019] S33: Calculate the path of the current typhoon T0 and the typhoon dataset T similar in time and space respectively i The latest three points of the path form the turning angle of the curve (P0-P1-P2) to identify the typhoon evolution trend and the degree of turning angle.

[0020] Optionally, performing a closeness analysis on the typhoon datasets that are similar in time and space to obtain a closeness includes:

[0021] A plurality of closed areas are constructed respectively with the line buffer boundary, the current typhoon path and the historical path, and the area of ​​each closed area in the buffer is directly obtained by using the GIS spatial analysis function. The smaller the value of the closed area, the closer the historical typhoon is to the real-time typhoon, and the greater the degree of closeness.

[0022] Optionally, the geometric features, proximity, and typhoon attribute features are selected to characterize the main features of the typhoon, and a typhoon similarity set pair analysis matrix is ​​constructed, including:

[0023] The geometric features include slope, curvature and angle;

[0024] The typhoon attribute characteristics include wind force F, central air pressure, moving speed, maximum ground wind speed near the center, radius of the seventh-level wind circle, and radius of the tenth-level wind circle.

[0025] Optionally, in the typhoon set pair analysis matrix, the current real-time typhoon and the historical typhoons similar in time and space form a set pair, and perform set pair similarity analysis and decision optimization, including:

[0026] S61: Based on the geometric features, attribute features and minimum enclosed area value of the current real-time typhoon T0, the corresponding eigenvalues ​​in the set pair analysis matrix are graded and quantified according to the multiple ratio relationship, and the identity, difference and opposition are calculated;

[0027] S62: Calculate similarity based on the sameness, difference and opposition, and perform typhoon similarity decision optimization based on the similarity.

[0028] S63: The similarity r i Sort by size and select the top three typhoons in the typhoon dataset that are similar in time and space. i or i Historical typhoons with a Ratio greater than 0.75 are selected as preferred typhoons for similarity prediction and participate in similarity prediction calculations.

[0029] Optionally, according to the results of the similarity analysis and decision optimization of the set, similarity prediction of the typhoon is performed, including:

[0030] S71: Calculating a similarity weight according to the similarity of the preferred typhoon;

[0031] S72: Obtain the predicted eigenvalue and predicted position of the real-time typhoon by weighting the selected typhoons with similar weights.

[0032] Optionally, it further includes:

[0033] Repeat S2 - S7 for rolling forecasting.

[0034] According to the second aspect of the embodiments of the present application, there is provided a typhoon track "time - space - shape - quantity" similarity forecasting device, including:

[0035] A database construction module, configured to obtain historical typhoon data and current real-time typhoon data, both types of data at least include typhoon attribute characteristics, store the historical typhoon data in the historical typhoon database, and store the real-time typhoon data in the real-time typhoon database;

[0036] A time - space screening module, configured to, based on the current real-time typhoon data, screen historical typhoon data with similar time in the historical typhoon database, and then screen historical typhoon data with similar spatial positions from the screened historical typhoon data with similar time, so as to obtain a typhoon data set with similar time and space;

[0037] A first analysis module, configured to perform geometric feature analysis on the typhoon data set with similar time and space to obtain geometric features;

[0038] A second analysis module, configured to perform proximity analysis on the typhoon data set with similar time and space to obtain proximity;

[0039] A construction module, configured to select the geometric features, proximity, and typhoon attribute characteristics to represent the main features of the typhoon, and construct a typhoon similarity set pair analysis matrix;

[0040] A third analysis module, configured to form a set pair with the current real-time typhoon and the historical typhoons with similar time and space in the typhoon set pair analysis matrix, and perform set pair similarity analysis and decision-making optimization;

[0041] A prediction module, configured to perform similarity prediction on the typhoon according to the results of the set pair similarity analysis and optimization.

[0042] According to the third aspect of the embodiments of the present application, there is provided an electronic device, characterized by including:

[0043] One or more processors;

[0044] A memory, configured to store one or more programs;

[0045] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in the first aspect.

[0046] The technical solutions provided by the embodiments of the present application may include the following beneficial effects:

[0047] As can be seen from the above embodiments, the present application adopts comprehensive integration technical means, taking into account the similarity of typhoon occurrence time, the similarity of typhoon path spatial regions, the similarity of typhoon path geometric forms, and the similarity of typhoon attribute characteristics, overcoming the problem of insufficient utilization of typhoon information in previous similarity forecasts. Furthermore, a new similarity criterion for comprehensive similarity of multi-dimensional elements of typhoons is formed. By comprehensively considering space, time, geometric features, and attribute features, it not only describes the shape similarity features but also includes the quantity similarity characteristics. The considered information is relatively comprehensive and can be dynamically iterated according to the evolution of typhoons, with high resolution ability, realizing the integrated similarity discrimination of typhoon path in terms of "time-space-shape-quantity" and the credibility and reliability of the selected similar paths.

[0048] By using GIS technology and the rolling forecast method of step-by-step screening and hierarchical optimization, it overcomes the problem that previous typhoon similarity forecasts did not consider the dynamic changes of typhoon similarity characteristics, and thus forms a step-by-step filtering similarity forecast method for "time-space-shape-quantity" integration. This method first conducts time-space secondary screening, that is, first uses time selection and then uses spatial buffer analysis technology to conduct buffer analysis on the current real-time typhoon sequence, and fully utilizes the GIS line buffer technology to quickly generate a typhoon data set with spatio-temporal similarity. On the basis of spatio-temporal similarity, further use the GIS spatial calculation ability to calculate the area to quickly calculate the closeness, and then conduct optimization through set pair analysis and mutation decision-making.

[0049] By using the method of set pair analysis and mutation decision-making optimization, it overcomes the limitation that previous typhoon similarity forecasts only use the average distance between corresponding control points as the similarity criterion, and thus forms a similarity evaluation method based on set pair same, different, and opposite analysis and decision-making optimization. That is, by constructing a similarity analysis set pair that simultaneously includes features such as (angle of rotation, slope, curvature), closeness, attributes (wind force, air pressure, moving speed, wind direction, wind circle radius), etc.; grading and processing the feature values according to the ratio method for set pair same, different, and opposite analysis avoids the normalization processing operation and is simpler and more convenient; regarding the set pair analysis result as a decision-making problem, using the mutation series to obtain the similarity as the priority ranking and optimization, and taking the weighted average of the similarities of the selected typhoons as the weight for similarity prediction avoids the subjectivity of the weight and is more objective and reliable, and can perform automatic rolling forecasts through cyclic iteration.

[0050] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments that conform to the present application, and are used together with the specification to explain the principles of the present application.

[0052] Figure 1 It is a flowchart of a "time - space - shape - quantity" similarity prediction method for typhoon paths shown according to an exemplary embodiment.

[0053] Figure 2 It is a spatial similarity line buffer diagram shown according to an exemplary embodiment.

[0054] Figure 3 It is a geometric similarity feature calculation diagram shown according to an exemplary embodiment.

[0055] Figure 4 It is a block diagram of an apparatus for a "time - space - shape - quantity" similarity prediction method for typhoon paths shown according to an exemplary embodiment. Detailed implementation manners

[0056] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present application. On the contrary, they are merely examples of apparatuses and methods consistent with some aspects of the present application as detailed in the appended claims.

[0057] The terms used in the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The singular forms "a", "the", and "said" used in the present application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0058] Figure 1 It is a flowchart of a "time - space - shape - quantity" similarity prediction method for typhoon paths shown according to an exemplary embodiment. As Figure 1 shown, the method may include the following steps:

[0059] S1. Obtain historical typhoon data and current real - time typhoon data. Both types of data include at least typhoon attribute characteristics. Store the historical typhoon data in a historical typhoon database and store the real - time typhoon data in a real - time typhoon database;

[0060] S2. Based on the current real - time typhoon data, screen the historical typhoon data with similar time in the historical typhoon database, and then screen the historical typhoon data with similar spatial positions among the screened historical typhoon data with similar time, so as to obtain a typhoon data set with similar time and space;

[0061] S3. Conduct geometric feature analysis on the typhoon data sets with similar time and space to obtain geometric features;

[0062] S4. Conduct proximity analysis on the typhoon data sets with similar time and space to obtain proximity;

[0063] S5. Select the geometric features, proximity, and typhoon attribute features to characterize the main features of typhoons, and construct a set pair analysis matrix for typhoon similarity;

[0064] S6. In the typhoon set pair analysis matrix, form a set pair with the current real-time typhoon and the historical typhoons with similar time and space, and conduct set pair similarity analysis and decision-making optimization;

[0065] S7. According to the results of the set pair similarity analysis and decision-making optimization, conduct similarity prediction on typhoons.

[0066] As can be seen from the above technical solutions, this application comprehensively considers time, space, geometric features, and typhoon attribute features, describes both shape similarity features and quantity similarity features, considers relatively comprehensive information, can be dynamically iterated according to the evolution of typhoons, has high resolution ability, and realizes the integrated similarity discrimination of typhoon paths in terms of "time-space-shape-quantity" and the credibility and reliability of the selected similar paths.

[0067] In the specific implementation of S1, obtain historical typhoon data and current real-time typhoon data. Both types of data include at least typhoon attribute features. Store the historical typhoon data in the historical typhoon database and store the real-time typhoon data in the real-time typhoon database;

[0068] In the specific implementation of S2, based on the current real-time typhoon data, screen the historical typhoon data with similar time in the historical typhoon database, and then screen the historical typhoon data with similar spatial positions among the screened historical typhoon data with similar time, so as to obtain a typhoon data set with similar time and space, including:

[0069] S21: Based on the month when the current real-time typhoon occurs in the real-time typhoon database, screen in the historical typhoon database according to time to obtain a typhoon data set that meets the time similarity;

[0070] Specifically, based on the month when the current real-time typhoon occurs in the real-time typhoon database, screen in the historical typhoon database according to time, and select a time period range of one month, with half a month before and after the current typhoon, to obtain a typhoon data set that meets the time similarity.

[0071] S22: Apply the GIS spatial analysis function to construct a line buffer area with the wind circle radius of the current typhoon at the current moment in the real-time typhoon database as the buffer radius and the real-time path backtracking several time periods from the current moment. According to the line buffer area, perform a spatial intersection calculation on the typhoons with similar time to obtain a dataset of typhoons with spatio-temporal similarity.

[0072] Specifically, apply the GIS spatial analysis function to construct a line buffer area with the wind circle radius (R7 or R10) of the current typhoon (T0) at the current moment in the real-time typhoon database as the buffer radius and the real-time path backtracking (3 or 5) time periods from the current time period (take 3 when the real-time typhoon is within the 48-hour warning line and 5 when it is outside the 48-hour warning line). Perform a spatial intersection calculation on the typhoons with similar time to obtain a dataset of typhoons with spatio-temporal similarity Ti at the current time period, where i = 1, 2,... n.

[0073] In the specific implementation of S3, perform a geometric feature analysis on the dataset of typhoons with spatio-temporal similarity to obtain the slope, curvature, and turning angle, including:

[0074] S31: Calculate the slope of the latest time period P0(x0, y0)-P1(x1, y1) of the path of the current real-time typhoon T0 and the paths in the dataset of typhoons with spatio-temporal similarity T i to determine the parallel feature between the paths to identify whether the evolving spatial directions of the typhoon paths are consistent;

[0075]

[0076] In the formula: is the slope of the line segment of the path P0(x0, y0)-P1(x1, y1) in the latest time period, (x0, y0) is the coordinate of P0, and (x1, y1) is the coordinate of P1;

[0077] S32: Calculate the curvature of the curve P0(x0, y0)-P1(x1, y1)-P2(x2, y2) formed by the latest 3 points in time of the path of the current real-time typhoon T0 and the paths in the dataset of typhoons with spatio-temporal similarity T i to identify the convexity and concavity of the geometric curve, where the curvature at point P1 represents the curvature of the curve P0(x0, y0)-P1(x1, y1)-P2(x2, y2);

[0078] Specifically, the curvature of P1 can be approximately obtained by using the method of fitting approximation:

[0079]

[0080] x' = x1 - x0 x" = x0 - 2x1 + x2

[0081] y' = y1 - y0, y" = y0 - 2y1 + y2

[0082] is the curvature of the curve P0(x0, y0)-P1(x1, y1)-P2(x2, y2);

[0083] When taking 5 points, the curvature of the curve P0(x0, y0)-P1(x1, y1)-P2(x2, y2)-P3(x3, y3)-P4(x4, y4)-P5(x5, y5) is represented by the curvature at the point P2(x2, y2), and the curvature of P2 can be obtained by curve fitting approximation:

[0084]

[0085]

[0086]

[0087] S33: Calculate the paths of the current real-time typhoon T0 and the typhoon dataset T with spatiotemporal similarity i The rotation angle of the curve P0(x0, y0)-P1(x1, y1)-P2(x2, y2) formed by the latest 3 points in time of the paths , to identify the typhoon evolution trend and the degree of turning.

[0088] Specifically, the rotation angle is represented by the included angle between P1P2 and P1P0:

[0089]

[0090] Or

[0091]

[0092] In the specific implementation of S4, a closeness analysis is performed on the spatiotemporally similar typhoon dataset to obtain the closeness, including:

[0093] Using the line buffer boundary, the path of the current real-time typhoon T0 and the path of the historical typhoon T i Separate multiple closed regions are constructed, and the area of each closed region within the buffer is directly obtained using the GIS spatial analysis function. The smaller the value of the closed region area, the closer the historical typhoon is to the real-time typhoon, and the greater the closeness.

[0094] Specifically, using the line buffer boundary generated by S2, the path of the current real-time typhoon T0 and the paths of the historical typhoons T i (i = 1, 2,..., n) to construct n closed regions respectively, and the area S of each closed region within the buffer is directly obtained using the GIS spatial analysis function i , Si The smaller the value of (i=1, 2, ..., n), the closer the historical typhoon is to the real-time typhoon, and the greater the degree of closeness, where S0=0.

[0095] In the specific implementation of S5, the geometric features, proximity, and typhoon attribute features are selected to characterize the main features of the typhoon, and a typhoon similarity set pair analysis matrix is ​​constructed;

[0096] Specifically, the typhoon geometric characteristic slope K obtained by calculation is selected Ti , curvature CK Ti 、Rotation angle θ Ti , closeness S i The main characteristics of a typhoon are characterized by the corresponding attribute characteristics such as wind force F, central air pressure, moving speed, maximum ground wind speed near the center, radius of the seventh-level wind circle, radius of the tenth-level wind circle, etc., namely:

[0097] T = (slope k, curvature ck, angle θ, area S, wind force F, central air pressure CP, moving speed V1, maximum ground wind speed near the center V2, radius of the seventh-level wind circle R7, radius of the tenth-level wind circle R10), thus obtaining the set pair analysis matrix describing the typhoon:

[0098]

[0099] The current real-time typhoon T0 and the historical typhoon T i The geometric characteristics (K Ti , CK Ti ,θ Ti ) is calculated by S3, and the closeness S i The attribute characteristic values ​​(F0, CP0, V10, V20, R70, R100) of T0 calculated by S4 are taken as the attribute values ​​of the current typhoon at the current time, T i The attribute characteristic value (F i , CP i , V1 i 、V2 i 、R7 i 、R10 i ) is obtained by linearly interpolating the corresponding attribute characteristic values ​​of the time period corresponding to the trajectory point of the historical typhoon at the current moment.

[0100] In the specific implementation of S6, the current real-time typhoon T0 and the historical typhoon T0 similar in time and space are used in the typhoon set pair analysis matrix. i (i=1,2,...n) forms a set pair analysis matrix to perform set pair similarity analysis and decision optimization, including:

[0101] S61: Based on the geometric features, attribute features and the smaller the enclosed area value of the current real-time typhoon T0, perform hierarchical quantization according to the ratio relationship to obtain the same degree, difference degree and opposition degree;

[0102] Specifically, in the typhoon set pair analysis matrix, n sets of set pairs (T0, T i )(i = 1, 2,...n) are formed by the current real-time typhoon T0 and the historical typhoons Ti (i = 1, 2,...n) that are similar in time and space for set pair analysis. Based on the eigenvalue (slope k, curvature ck, rotation angle θ, wind force F, air pressure P, speed V, wind circle radius R) of the current real-time typhoon T0 and min(S i )(i = 1, 2...n), perform hierarchical quantization on the corresponding eigenvalues in the set pair analysis matrix according to the ratio relationship. When the ratio is in [0.5, 1.5], it is considered the same. When the ratio is greater than or equal to 2, it is considered opposite, and the rest are considered different. Count the numbers N1, N2, N3 of the same, different and opposite; where the number of features N = N1 + N2 + N3, and obtain:

[0103] Same degree: a i = N1 / N, i = 1, 2...n

[0104] Difference degree: b i = N2 / N, i = 1, 2...n

[0105] Opposition degree: c i = N3 / N, i = 1, 2...n

[0106] S62: According to the same degree, difference degree and opposition degree, perform decision optimization to calculate the similarity.

[0107] Specifically, the feature similarity of each set pair (T0, T i ) is represented by r i . In order to facilitate comparison, regarding the same degree, difference degree and opposition degree among the above set pairs as a complementary optimization decision problem, the normalization formula of the mutation system is used to directly calculate the mutation series.

[0108]

[0109] Calculate the similarity r i according to the complementary decision based on the mutation series;

[0110]

[0111] Performing set pair same, different and opposite analysis by grading the eigenvalues according to the ratio method avoids the normalization operation and is simpler and more convenient; regarding the set pair analysis result as a decision optimization problem to obtain the similarity as the basis for priority ranking and optimization.

[0112] S63: Sort the similarity r i in descending order, and select the top three r in the typhoon dataset with spatio-temporal similarity i or r i Greater than 0.75 (assuming there are m) historical typhoons are used as the preferred typhoons for similarity prediction and participate in the similarity prediction calculation.

[0113] In the specific implementation of S7, according to the preferred typhoons, similarity prediction of typhoons is performed, including:

[0114] S71: Calculate the similarity weight according to the similarity of the preferred typhoons;

[0115]

[0116] S72: Weight the preferred typhoons with the similarity weight to obtain the predicted characteristic values and predicted positions of the real-time typhoon.

[0117] Specifically, in T i dataset, select the corresponding historical typhoons of w i and weight them with w i to obtain the predicted characteristic values and predicted positions of the real-time typhoon.

[0118]

[0119] It may also include: repeating S2 to S7 for rolling forecasting.

[0120] Corresponding to the embodiments of the typhoon track "time-space-shape-quantity" similarity forecasting method described above, the present application also provides embodiments of a typhoon track "time-space-shape-quantity" similarity forecasting device.

[0121] Figure 4 is a block diagram of a typhoon track "time-space-shape-quantity" similarity forecasting device shown according to an exemplary embodiment. Refer to Figure 4 , the device includes:

[0122] Database construction module 1, configured to obtain historical typhoon data and current real-time typhoon data, both types of data at least include typhoon attribute characteristics, store the historical typhoon data in the historical typhoon database, and store the real-time typhoon data in the real-time typhoon database;

[0123] Spatio-temporal screening module 2, configured to, based on the current real-time typhoon data, screen historical typhoon data with similar time in the historical typhoon database, and then screen historical typhoon data with similar spatial positions from the screened historical typhoon data with similar time, so as to obtain a typhoon dataset with spatio-temporal similarity;

[0124] The first analysis module 3 is configured to perform geometric feature analysis on the typhoon data set with spatio-temporal similarity to obtain geometric features;

[0125] The second analysis module 4 is configured to perform proximity analysis on the typhoon data set with spatio-temporal similarity to obtain proximity;

[0126] The construction module 5 is configured to select the geometric features, proximity, and typhoon attribute features to characterize the main features of the typhoon, and construct a typhoon similarity set pair analysis matrix;

[0127] The third analysis module 6 is configured to form a set pair with the current real-time typhoon and the historical typhoons with spatio-temporal similarity in the typhoon set pair analysis matrix, and perform set pair similarity analysis and decision-making optimization;

[0128] The prediction module 7 is configured to perform similarity prediction on the typhoon according to the results of the set pair similarity analysis and decision-making optimization.

[0129] Regarding the device in the above embodiments, the specific manners in which each module performs operations have been described in detail in the embodiments related to the method, and will not be elaborated herein.

[0130] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can refer to the partial descriptions of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of the present application. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0131] Correspondingly, the present application further provides an electronic device, including: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement a typhoon path "time-space-shape-quantity" similarity prediction method as described above.

[0132] Correspondingly, the present application further provides a computer-readable storage medium, on which computer instructions are stored, and when the instructions are executed by a processor, a typhoon path "time-space-shape-quantity" similarity prediction method as described above is implemented.

[0133] Other embodiments of the present application will be readily contemplated by those skilled in the art upon consideration of the specification and practice of the disclosure herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include well-known knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and examples are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the claims.

[0134] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A typhoon "time - space - shape - quantity" similarity forecasting method, characterized in that, Including: S1. Obtain historical typhoon data and current real-time typhoon data. Both types of data include at least typhoon attribute characteristics. Store the historical typhoon data in the historical typhoon database and store the current real-time typhoon data in the real-time typhoon database. S2. Based on the current real-time typhoon data, screen for historical typhoon data with similar time in the historical typhoon database, and then screen for historical typhoon data with similar spatial positions among the screened historical typhoon data with similar time, so as to obtain a typhoon data set with similar time and space. S3. Conduct geometric feature analysis on the typhoon data set with similar time and space to obtain geometric features. S4. Conduct proximity analysis on the typhoon data set with similar time and space to obtain proximity. S5. Select the geometric features, proximity, and typhoon attribute characteristics to characterize the features of typhoons, and construct a typhoon similar set pair analysis matrix. S6. In the typhoon set pair analysis matrix, form set pairs with the current real-time typhoon and the historical typhoons with similar time and space, and conduct set pair similarity analysis and decision-making optimization. S7. According to the results of the set pair similarity analysis and optimization, conduct similarity prediction on typhoons. Among them, conducting geometric feature analysis on the typhoon data set with similar time and space to obtain slope, curvature, and rotation angle includes: S31: Calculate the slopes of the latest time periods of the paths of the current real-time typhoon T0 and the typhoon dataset T with spatiotemporal similarity respectively, so as to distinguish the parallel characteristics between the paths to identify whether the evolving spatial directions of the typhoon paths are consistent; i and identify whether the evolving spatial directions of the typhoon paths are consistent by distinguishing the parallel characteristics between the paths; S32: Calculate the curvature of the curve (P0 - P1 - P2) formed by the latest three points in time of the path of the current real-time typhoon T0 and the paths of the typhoon dataset T i that is spatiotemporally similar, to identify the convexity and concavity of the geometric curve, where the curvature at point P1 represents the curvature of the (P0 - P1 - P2) curve; S33: Calculate the angles formed by the latest three points in time of the paths of the current real-time typhoon T0 and the typhoon dataset T with spatiotemporal similarity respectively to identify the typhoon evolution trend and the degree of the angle; i The angle of the curve (P0 - P1 - P2) formed by the latest three points in time of the paths of the typhoon dataset T with spatiotemporal similarity is used to identify the typhoon evolution trend and the degree of the angle; Conducting proximity analysis on the typhoon data set with similar time and space to obtain proximity includes: Construct multiple closed regions with the line buffer boundary, the path of the current real-time typhoon T0, and the paths of historical typhoons Ti respectively. Use the GIS spatial analysis function to directly obtain the areas of each closed region within the buffer. The smaller the value of the closed region area, the closer the historical typhoon is to the real-time typhoon, and the greater the proximity. In the typhoon set pair analysis matrix, forming set pairs with the current real-time typhoon and the historical typhoons with similar time and space, and conducting set pair similarity analysis and decision-making optimization includes: S61: Based on the characteristics of the current real-time typhoon T0 and the minimum closed region area value, grade and quantify the corresponding eigenvalue in the set pair analysis matrix according to the ratio relationship, and calculate the identity degree, difference degree, and opposition degree. S62: Calculate the similarity degree according to the identity degree, difference degree, and opposition degree. S63: Sort the similarity r i in descending order, and select the top three r values in the typhoon dataset with spatio-temporal similarity i or r i Historical typhoons with a value greater than 0.75 are used as preferred typhoons for similarity prediction and participate in the similarity prediction calculation.

2. The method according to claim 1, wherein Based on the current real-time typhoon data, screening for historical typhoon data with similar time in the historical typhoon database, and then screening for historical typhoon data with similar spatial positions among the screened historical typhoon data with similar time, so as to obtain a typhoon data set with similar time and space includes: S21: In the historical typhoon database, screen according to time with the month when the current real-time typhoon occurs in the real-time typhoon database as the benchmark, and obtain a typhoon data set that meets the time similarity. S22: Apply the GIS spatial analysis function. Use the wind circle radius of the current typhoon at the current moment in the real-time typhoon database as the buffer radius, and construct a line buffer region with the real-time path backtracking several time periods from the current moment. According to the line buffer region, conduct spatial intersection calculation on the typhoons with similar time to obtain a typhoon data set with similar time and space.

3. The method according to claim 1, characterized in that, The geometric features include slope, curvature, and rotation angle. The typhoon attribute features include wind force F, central pressure, moving speed, maximum ground wind speed near the center, radius of the 7th level wind circle, and radius of the 10th level wind circle.

4. The method according to claim 1, characterized in that, According to the results of the set pair similarity analysis and decision-making optimization, similarity prediction is performed on the typhoon, including: S71: Calculate the similarity weight according to the similarity. S72: Weight the optimized typhoon with the similarity weight to obtain the predicted characteristic values and predicted positions of the real-time typhoon.

5. The method according to claim 1, wherein It also includes: Repeat S2~S7 for rolling forecasting.

6. A typhoon track "time-space-shape-quantity" similarity prediction device, characterized in that, Including: A database construction module, which is used to obtain historical typhoon data and current real-time typhoon data. Both types of data at least include typhoon attribute features. The historical typhoon data is stored in the historical typhoon database, and the real-time typhoon data is stored in the real-time typhoon database. A space-time screening module, which is used to screen historical typhoon data with similar time in the historical typhoon database based on the current real-time typhoon data, and then screen historical typhoon data with similar spatial positions from the screened historical typhoon data with similar time, so as to obtain a set of typhoons with similar space-time. A first analysis module, which is used to perform geometric feature analysis on the set of typhoons with similar space-time to obtain geometric features. A second analysis module, which is used to perform closeness analysis on the set of typhoons with similar space-time to obtain closeness. A construction module, which is used to select the geometric features, closeness, and typhoon attribute features to characterize the features of the typhoon, and construct a typhoon similarity set pair analysis matrix. A third analysis module, which is used to form a set pair with the current real-time typhoon and the historical typhoons with similar space-time in the typhoon set pair analysis matrix, and perform set pair similarity analysis and decision-making optimization. A prediction module, which is used to perform similarity prediction on the typhoon according to the results of the set pair similarity analysis and optimization. Among them, performing geometric feature analysis on the set of typhoons with similar space-time to obtain slope, curvature, and rotation angle includes: Calculate the paths of the current real-time typhoon T0 and the typhoon dataset T with spatiotemporal similarity respectively i for the slope of the latest time period of the paths, so as to distinguish the parallel characteristics between the paths and identify whether the evolving spatial directions of the typhoon paths are consistent; Calculate the paths of the current real-time typhoon T0 and the typhoon dataset T with spatiotemporal similarity respectively i The curvature of the curve formed by the latest three points in time of the paths of (P0 - P1 - P2) is calculated to identify the convexity and concavity of the geometric curve, where the curvature at point P1 represents the curvature of the (P0 - P1 - P2) curve; Calculate the paths of the current real-time typhoon T0 and the typhoon dataset T with spatiotemporal similarity respectively i to form the turning angle of the curve (P0 - P1 - P2) of the latest three points in time of the path, so as to identify the typhoon evolution trend and the degree of turning angle; Performing closeness analysis on the set of typhoons with similar space-time to obtain closeness includes: Construct multiple closed regions with the line buffer boundary, the path of the current real-time typhoon T0, and the paths of historical typhoons Ti respectively. Use the GIS spatial analysis function to directly obtain the areas of each closed region within the buffer. The smaller the value of the closed region area, the closer the historical typhoon is to the real-time typhoon, and the greater the closeness. Forming a set pair with the current real-time typhoon and the historical typhoons with similar space-time in the typhoon set pair analysis matrix, and performing set pair similarity analysis and decision-making optimization includes: Taking the characteristics of the current real-time typhoon T0 and the minimum closed region area value as the benchmark, grading and quantifying the corresponding characteristic values in the set pair analysis matrix according to the ratio relationship, and calculating the identity, difference degree, and opposition degree. Calculate the similarity according to the identity, difference degree, and opposition degree. For the similarity r i Sort by size, and select the top three r in the typhoon dataset with spatio-temporal similarity i Or r i Historical typhoons with r greater than 0.75 are used as preferred typhoons for similarity prediction and participate in the similarity prediction calculation.

7. An electronic device, characterized in that, Including: One or more processors; A memory, which is used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1-5.

Images