A rainstorm similarity analysis method and terminal based on rain type and rain intensity

Abstract

The application discloses a rainstorm similarity analysis method and a terminal based on rain type and rain intensity, and comprises the following steps: acquiring real-time hourly precipitation in a current first preset time length, judging whether it is a rainstorm event, recording it as a real-time rainstorm event if it is, and calculating the rain type and the rain intensity of the real-time rainstorm event to obtain a real-time rain type and a real-time rain intensity; analyzing the similarity of the real-time rain type, the real-time hourly precipitation and the real-time rain intensity with the historical rain type, the historical hourly precipitation and the historical rain intensity of a plurality of historical rainstorm events; calculating the rainstorm comprehensive similarity ranking of the real-time rainstorm event and each historical rainstorm event based on the similarity; and identifying similar rainstorm events according to the rainstorm comprehensive similarity ranking. The application comprehensively considers the change characteristics of the precipitation with time, combines the rain intensity index of the rainstorm event, and can better identify the historical rainstorm events similar to the current real-time rainstorm event.

Description

Technical Field

[0001] The present invention relates to the technical field of disaster prevention and reduction and meteorological forecasting, and in particular to a rainstorm similarity analysis method based on rain type and rain intensity and a terminal. Background Art

[0002] The existing rainstorm similarity analysis mainly uses multi-factor characteristic indicators to comprehensively analyze rainstorm similarity, among which rainstorm characteristic indicators include:

[0003] Rainstorm center: the area where the intensity of rainstorm is concentrated.

[0004] Cumulative rainfall: the depth of rainfall that falls on a certain area of ​​the ground during a certain period of time.

[0005] Rainstorm intensity: the amount of precipitation per unit time.

[0006] Duration of heavy rainfall: the length of time when the hourly rainfall exceeds 5 mm.

[0007] Rainfall pattern: the changing characteristics of rainfall within a certain period of time.

[0008] According to the above characteristic indicators, the Euclidean distance of each indicator between the real-time rainstorm and the historical rainstorm events is calculated respectively, the distances of the indicators are added to obtain the comprehensive distance, and the comprehensive similarity of the rainstorm is obtained by sorting the distances from small to large.

[0009] Existing methods often ignore temporal variations in precipitation when selecting rainstorm characteristic indicators, resulting in significant discrepancies in the periods of heavy rainfall among similarly identified rainstorm events. Some methods incorporate rain pattern characteristics into rainstorm similarity analysis, which increases the reliability of the similarity analysis to a certain extent. However, current rain pattern identification methods still have some misjudgments, and there are rainstorm events with ambiguous rain patterns. Therefore, methods that only perform similarity analysis on rainstorm events with the same rain pattern are insufficient. Summary of the Invention

[0010] The technical problem to be solved by the present invention is to provide a rainstorm similarity analysis method and terminal based on rain type and rain intensity, comprehensively considering the temporal variation characteristics of precipitation and combining the rainstorm intensity index of rainstorm events to achieve better identification of similar rainstorm events.

[0011] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0012] A rainstorm similarity analysis method based on rain type and rain intensity comprises the following steps:

[0013] S1. Obtaining real-time hourly precipitation within a current first preset time period, determining whether it is a rainstorm event, and if so, recording it as a real-time rainstorm event, and calculating the rain type and rain intensity of the real-time rainstorm event to obtain the real-time rain type and real-time rain intensity. The real-time rain type is determined by using a fuzzy recognition algorithm based on the seven typical rain types summarized by Mologov;

[0014] S2. analyzing the similarity between the real-time rain pattern, the real-time hourly precipitation, and the real-time rain intensity and the historical rain patterns, historical hourly precipitation, and historical rain intensities of a plurality of historical rainstorm events;

[0015] S3. Obtaining a comprehensive similarity ranking of the real-time rainstorm event and each of the historical rainstorm events based on similarity calculation;

[0016] S4. Identify similar rainstorm events based on the comprehensive similarity ranking of the rainstorms.

[0017] In order to solve the above technical problems, another technical solution adopted by the present invention is:

[0018] A rainstorm similarity analysis terminal based on rain type and rain intensity comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When executing the computer program, the processor implements the steps of the above-mentioned rainstorm similarity analysis method based on rain type and rain intensity.

[0019] The beneficial effect of the present invention is that after judging that the current real-time rainstorm is a rainstorm event according to the precipitation within a preset time period, similarity analysis is performed on the real-time rain type, real-time hourly precipitation and real-time rainfall intensity of the real-time rainstorm event and the historical rain type, historical hourly precipitation and historical rainfall intensity of historical rainstorm events, and a comprehensive rainstorm similarity ranking of the real-time rainstorm event and the historical rainstorm event is obtained to analyze similar rainstorm events, comprehensively considering the change characteristics of precipitation over time and combining the rainfall intensity index of the rainstorm event to achieve better identification of historical rainstorm events similar to the current real-time rainstorm event. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 Flowchart of a rainstorm similarity analysis method based on rain type and rain intensity according to an embodiment of the present invention;

[0021] Figure 2 There are seven typical rain types based on the summary of Mologovo;

[0022] Figure 3 2 is a schematic structural diagram of a rainstorm similarity analysis terminal based on rain type and rain intensity according to an embodiment of the present invention.

[0023] Description of labels:

[0024] 1. A rainstorm similarity analysis terminal based on rain type and rain intensity; 2. Memory; 3. Processor. DETAILED DESCRIPTION

[0025] To illustrate the technical content, achieved objectives and effects of the present invention in detail, the following description is given in conjunction with the embodiments and accompanying drawings.

[0026] Prior to this, the proper nouns appearing in this invention are explained as follows:

[0027] 1. Based on Molokov's seven typical rainfall patterns, which were proposed by Molokov in his book Storm Drains and Combined Sewers, published in March 1959.

[0028] 2. DTW: Dynamic Time Warping, a dynamic time warping algorithm that constructs a correspondence between sequence elements of two different lengths based on the principle of closest distance to evaluate the similarity between the two sequences.

[0029] Please refer to Figure 1 and Figure 2 , a rainstorm similarity analysis method based on rain type and rain intensity, comprising the steps of:

[0030] S1. Obtaining real-time hourly precipitation within a current first preset time period, determining whether it is a rainstorm event, and if so, recording it as a real-time rainstorm event, and calculating the rain type and rain intensity of the real-time rainstorm event to obtain the real-time rain type and real-time rain intensity. The real-time rain type is determined by using a fuzzy recognition algorithm based on the seven typical rain types summarized by Mologov;

[0031] S2. analyzing the similarity between the real-time rain pattern, the real-time hourly precipitation, and the real-time rain intensity and the historical rain patterns, historical hourly precipitation, and historical rain intensities of a plurality of historical rainstorm events;

[0032] S3. Obtaining a comprehensive similarity ranking of the real-time rainstorm event and each of the historical rainstorm events based on similarity calculation;

[0033] S4. Identify similar rainstorm events based on the comprehensive similarity ranking of the rainstorms.

[0034] From the above description, it can be seen that the beneficial effect of the present invention is that after judging the current real-time rainstorm as a rainstorm event based on the precipitation within a preset time period, the real-time rain type, real-time hourly precipitation and real-time rain intensity of the real-time rainstorm event are similar to the historical rain type, historical hourly precipitation and historical rain intensity of historical rainstorm events. The real-time rainstorm event and the historical rainstorm event are ranked by comprehensive similarity to analyze similar rainstorm events, comprehensively considering the change characteristics of precipitation over time and combining the rain intensity index of the rainstorm event to achieve better identification of historical rainstorm events similar to the current real-time rainstorm event.

[0035] Furthermore, before step S1, the following steps are further included:

[0036] S0. Obtain the historical hourly precipitation of a single rain gauge station, and filter out precipitation events with a duration greater than a first preset duration, recording them as historical rainstorm events;

[0037] The rainfall type and rainfall intensity of each of the historical rainstorm events are calculated to obtain the historical rainfall type and historical rainfall intensity.

[0038] From the above description, we can see that historical rainstorm events can be screened through the historical data of a single rainfall station to ensure the authenticity and reliability of the data, and further enable better identification of historical rainstorm events similar to the current real-time rainstorm event.

[0039] Furthermore, the historical rainfall type of each historical rainstorm event is calculated based on the seven typical rainfall types summarized by Mologov and judged using a fuzzy recognition algorithm. The specific steps are:

[0040] S02. Counting the total precipitation duration T, total precipitation P, cumulative precipitation duration t, and cumulative precipitation Pt of each historical rainstorm event, dividing the cumulative precipitation duration t by the total precipitation duration T and performing normalization processing to obtain the normalized cumulative precipitation duration curve for each historical rainstorm event; and dividing the cumulative precipitation Pt by the total precipitation P and performing normalization processing to obtain the normalized cumulative precipitation duration curve for each historical rainstorm event.

[0041] S03, repeating step S02 to calculate the cumulative precipitation duration curves of the seven typical rain types;

[0042] S04. For all normalized accumulated precipitation duration curves, a point is taken every 0.01 on the horizontal coordinate to obtain the accumulated precipitation value on the corresponding vertical coordinate on the curve;

[0043] S05. Calculate the Euclidean distance between each of the historical rainstorm events and the seven typical rain patterns through all the points selected in step S03, select the typical rain pattern with the smallest distance as the corresponding rain pattern of the historical rainstorm event, and obtain the historical rain pattern corresponding to each of the historical rainstorm events.

[0044] From the above description, we can see that the seven typical rainfall types summarized by Mologov are currently commonly used in heavy rain analysis. By using the fuzzy recognition algorithm to determine which of the seven typical rainfall types the historical heavy rain events belong to, we can ensure the accuracy of historical rainfall type judgment.

[0045] Furthermore, the historical rainfall intensity of each of the historical rainstorm events includes the historical average rainfall intensity and the historical maximum rainfall intensity;

[0046] The calculation of the historical rainfall intensity of each historical rainstorm event is specifically as follows:

[0047] S06. Calculate the average and maximum values ​​of the historical hourly precipitation corresponding to each of the historical rainstorm events to obtain the historical average rainfall intensity and the historical maximum rainfall intensity of each of the historical rainstorm events.

[0048] From the above description, it can be seen that the average rainfall intensity and maximum rainfall intensity of each historical rainstorm event are analyzed at the same time, that is, not only the average rainfall intensity index but also the maximum rainfall intensity index is considered, so as to more accurately identify similar rainstorm events through the results of multiple indicators.

[0049] Furthermore, before step S1, the following steps are further included:

[0050] The historical rain type, the historical hourly precipitation, the historical average rainfall intensity and the historical maximum rainfall intensity are used as the four indicators of the historical rainstorm events, and the four indicators of each historical rainstorm event are saved in the historical rainstorm event data set.

[0051] From the above description, it can be seen that each screened historical rainstorm event and the four indicators calculated are stored, so that when the similarity of real-time rainstorm events is subsequently analyzed, historical data can be directly obtained through the historical rainstorm event dataset for similarity analysis, without having to obtain historical data from the site to re-determine the historical rainstorm events and recalculate the four indicators, which effectively improves the efficiency of subsequent rainstorm similarity analysis.

[0052] Furthermore, the step S1 is specifically as follows:

[0053] S11, obtaining the real-time hourly precipitation within the first preset time period, determining whether it is a rainstorm event, and if so, recording it as a real-time rainstorm event;

[0054] S12. The rainfall type and rainfall intensity of the real-time rainstorm event are calculated in the same manner as the historical rainfall type and the historical rainfall intensity, and the real-time rainfall type, the real-time average rainfall intensity and the real-time maximum rainfall intensity are obtained.

[0055] From the above description, it can be seen that the calculation of the real-time rain type and real-time rain intensity of the real-time rainstorm event is the same as the calculation of the historical rain type and historical rain intensity, which also ensures the accurate judgment of the rain type of the real-time rainstorm event.

[0056] Furthermore, the step S2 is specifically as follows:

[0057] S21. Based on the rain type index, obtain, from the historical rainstorm event dataset, the historical rainstorm events corresponding to the same historical rain type as the real-time rain type, and assign a coefficient K of a first preset value to the historical rainstorm events; and assign a coefficient K of a second preset value to the other historical rainstorm events, where the first preset value is less than the second preset value.

[0058] S22, based on the hourly precipitation index, using the dynamic time warping (DTW) algorithm to calculate the curve similarity distance between the real-time hourly precipitation and each of the historical hourly precipitation in the historical rainstorm event dataset, and sorting the curves in a forward or reverse order according to the curve similarity distance to obtain multiple curve similarity rankings L1;

[0059] S23, based on the average rainfall intensity index, calculating the difference between the real-time average rainfall intensity and each of the historical average rainfall intensities in the historical rainstorm event dataset, and sorting the differences in a forward or reverse order to obtain multiple average rainfall intensity similarity rankings L2;

[0060] S24. Based on the maximum rainfall intensity index, calculate the difference between the real-time maximum rainfall intensity and each of the historical maximum rainfall intensities in the historical rainstorm event dataset, and sort the differences in a forward or reverse order to obtain multiple maximum rainfall intensity similarity rankings L3.

[0061] From the above description, it can be seen that the coefficient K is assigned to different historical rainstorm events through the rain type index. The role of the coefficient K is to reduce the distance of historical rainstorm events with the same rain type as the real-time rainstorm event when calculating the comprehensive similarity distance, that is, to a certain extent, improve the ranking of historical rainstorm events with the same rain type as the real-time rainstorm event; at the same time, the curve similarity adopts the DTW dynamic time warping algorithm to fully consider the changing characteristics of precipitation over time. Finally, the similarity of the average rainfall intensity and the maximum rainfall intensity is calculated by the difference method to obtain the ranking of the four indicator results of the real-time rainstorm event and each historical rainstorm event, so as to calculate the comprehensive similarity distance of the similarity ranking.

[0062] Furthermore, the step S3 is specifically as follows:

[0063] S31, calculating the comprehensive similarity distance of the similarity ranking between the real-time rainstorm event and each of the historical rainstorm events in the historical rainstorm event dataset according to the formula L=K*(L1+L2+L3);

[0064] S32, ranking each of the comprehensive similarity distances in a forward or reverse direction to obtain a rainstorm comprehensive similarity ranking.

[0065] Further, the step S4 is specifically:

[0066] According to the rainstorm comprehensive similarity ranking, if the ranking is forward, the historical rainstorm event corresponding to the minimum L value is taken as the similar rainstorm event of the real-time rainstorm event;

[0067] If the ranking is reverse, the historical rainstorm event corresponding to the maximum L value is taken as the similar rainstorm event of the real-time rainstorm event.

[0068] As can be seen from the above description, each of the real-time rainstorm event and the historical rainstorm event has a comprehensive similarity distance, and the historical rainstorm event most similar to the real-time rainstorm event can be obtained by ranking each of the comprehensive similarity distances, that is, the rapid and accurate rainstorm similarity analysis is realized.

[0069] Please refer to Figure 3 A rainstorm similarity analysis terminal based on rain type and rain intensity, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned rainstorm similarity analysis method based on rain type and rain intensity when executing the computer program.

[0070] As can be seen from the above description, the beneficial effects of the present application are that based on the same technical concept, in combination with the above-mentioned multi-modal data feature method based on a near-electricity bracelet, a rainstorm similarity analysis terminal based on rain type and rain intensity is provided, after judging that the current real-time rainstorm is a rainstorm event according to the precipitation in a preset time period, the real-time rain type, the real-time hourly precipitation and the real-time rain intensity of the real-time rainstorm event and the historical rain type, the historical hourly precipitation and the historical rain intensity of the historical rainstorm event are all subjected to similarity analysis, the rainstorm comprehensive similarity ranking of the real-time rainstorm event and the historical rainstorm event is obtained to analyze the similar rainstorm event, the change characteristics of the precipitation with time are comprehensively considered, and the rain intensity index of the rainstorm event is combined to better identify the historical rainstorm event similar to the current real-time rainstorm event.

[0071] The rainstorm similarity analysis method and terminal based on rain type and rain intensity provided by the present application are suitable for rainstorm similarity analysis of a real-time rainstorm event to judge a similar rainstorm event in the scene of corresponding meteorological warning, protection and the like.

[0072] Please refer to Figure 1 The embodiment one of the present application is:

[0073] A rainstorm similarity analysis method based on rain type and rain intensity, such as Figure 1As shown, the steps include:

[0074] S1. Obtain the real-time hourly precipitation within the current first preset time period, determine whether it is a rainstorm event, and if so, record it as a real-time rainstorm event, and calculate the rain type and rain intensity of the real-time rainstorm event to obtain the real-time rain type and real-time rain intensity. The real-time rain type is determined by using the seven typical rain types summarized by Mologov and a fuzzy recognition algorithm.

[0075] The first preset duration can be set to 24 hours.

[0076] S2. Analyze the similarity between the real-time rain pattern, real-time hourly precipitation and real-time rain intensity and the historical rain pattern, historical hourly precipitation and historical rain intensity of multiple historical rainstorm events.

[0077] S3. Based on the similarity calculation, the comprehensive similarity ranking of the real-time rainstorm event and each historical rainstorm event is obtained.

[0078] S4. Identify similar rainstorm events based on the comprehensive similarity ranking of rainstorms.

[0079] That is, in this embodiment, after judging that the current real-time rainstorm is a rainstorm event based on the precipitation within a preset time period, a similarity analysis is performed on the real-time rain type, real-time hourly precipitation and real-time rainfall intensity of the real-time rainstorm event and the historical rain type, historical hourly precipitation and historical rainfall intensity of historical rainstorm events, and a comprehensive similarity ranking of the real-time rainstorm event and the historical rainstorm events is obtained to analyze similar rainstorm events. The change characteristics of precipitation over time are comprehensively considered, combined with the rainfall intensity index of the rainstorm event, so as to better identify historical rainstorm events similar to the current real-time rainstorm event.

[0080] Please refer to Figure 2 , the second embodiment of the present invention is:

[0081] A rainstorm similarity analysis method based on rain type and rain intensity, based on the above embodiment 1, in this embodiment, before step S1, further includes:

[0082] S0. Obtain the historical hourly precipitation of a single rain gauge station, filter out precipitation events with a precipitation duration greater than a first preset duration, record them as historical rainstorm events, and then calculate the rain type and rain intensity of each historical rainstorm event to obtain the historical rain type and historical rain intensity.

[0083] That is, screening historical rainstorm events through the historical data of a single rain gauge station can ensure the authenticity and reliability of the data, and further enable better identification of historical rainstorm events similar to current real-time rainstorm events.

[0084] like Figure 2The figure shows seven typical rainfall types summarized by Molokov. These are commonly used rainfall types in current rainstorm analysis. By determining which of the seven typical rainfall types the acquired historical rainstorm event belongs to, it is possible to ensure accurate determination of the historical rainfall type. In this embodiment, the historical rainfall type of each historical rainstorm event is calculated based on the seven typical rainfall types summarized by Molokov and determined using a fuzzy recognition algorithm. The specific steps are as follows:

[0085] S02. Count the total precipitation duration T, total precipitation P, cumulative precipitation duration t, and cumulative precipitation Pt of each historical rainstorm event. Divide the cumulative precipitation duration t by the total precipitation duration T and then normalize it as the horizontal axis. Divide the cumulative precipitation Pt by the total precipitation P and then normalize it as the vertical axis to obtain the normalized cumulative precipitation duration curve of each historical rainstorm event.

[0086] S03. Repeat step S02 to calculate the cumulative precipitation duration curves of seven typical rain types.

[0087] S04. For all normalized cumulative precipitation duration curves, a point is taken every 0.01 on the horizontal coordinate to obtain the corresponding cumulative precipitation value on the vertical coordinate of the curve.

[0088] S05. Calculate the Euclidean distance between each historical rainstorm event and the seven typical rain patterns through all the points selected in step S03, select the typical rain pattern with the smallest distance as the rain pattern of the corresponding historical rainstorm event, and obtain the historical rain pattern corresponding to each historical rainstorm event.

[0089] At the same time, in addition to calculating the historical rainfall type, the historical rainfall intensity also needs to be calculated. In this embodiment, the historical rainfall intensity of each historical rainstorm event includes the historical average rainfall intensity and the historical maximum rainfall intensity. The calculation of the historical rainfall intensity of each historical rainstorm event is specifically as follows:

[0090] S06. Calculate the average and maximum historical hourly precipitation corresponding to each historical rainstorm event to obtain the historical average rainfall intensity and the historical maximum rainfall intensity of each historical rainstorm event.

[0091] That is, not only the average rainfall intensity index but also the strongest rainfall intensity index is considered, so as to more accurately identify similar rainstorm events through the results of multiple indicators.

[0092] In addition, in this embodiment, before step S1, the following steps are also included:

[0093] Historical rainfall type, historical hourly precipitation, historical average rainfall intensity and historical maximum rainfall intensity are used as the four indicators of historical rainstorm events, and the four indicators of each historical rainstorm event are saved in the historical rainstorm event dataset.

[0094] Each historical rainstorm event and the four calculated indexes thereof are stored, so that when analyzing the similarity of a real-time rainstorm event, historical data can be directly obtained from the historical rainstorm event data set for similarity analysis, without the need to re-determine historical rainstorm events from historical data and re-calculate the four indexes, thereby effectively improving the efficiency of subsequent rainstorm similarity analysis.

[0095] Embodiment three of the present application is:

[0096] A rainstorm similarity analysis method based on rain type and rain intensity, in the above embodiment two, in the present embodiment, step S1 is specifically:

[0097] S11, obtain the real-time hourly precipitation in the current first preset time length, and determine whether it is a rainstorm event, if so, mark it as a real-time rainstorm event.

[0098] S12, the real-time rain type and the real-time rain intensity of the real-time rainstorm event are calculated in the same way as the historical rain type and the historical rain intensity, and the real-time rain type, the real-time average rain intensity and the real-time maximum rain intensity are obtained.

[0099] That is, in the present embodiment, the real-time rain type and the real-time rain intensity of the real-time rainstorm event are calculated in the same way as the historical rain type and the historical rain intensity, to ensure the accuracy of the real-time rainstorm event rain type judgment.

[0100] Meanwhile, step S2 is specifically:

[0101] S21, based on the rain type index, the historical rainstorm event corresponding to the historical rain type same as the real-time rain type is obtained from the historical rainstorm event data set, and a coefficient K with a first preset value is assigned to it, and other historical rainstorm events are assigned a coefficient K with a second preset value, wherein the first preset value is less than the second preset value.

[0102] It is worth noting that the coefficient K has no specific meaning, and its role is to reduce the distance of the historical rainstorm event with the same rain type as the real-time rainstorm event when calculating the comprehensive similarity distance, that is, to improve the ranking of the historical rainstorm event with the same rain type as the real-time rainstorm event to a certain extent, in the present embodiment, the first preset value can be set to 0.5, and the second preset value can be set to 1, in order to facilitate calculation.

[0103] S22, based on the hourly precipitation index, the dynamic time warping DTW algorithm is used to calculate the curve similarity distance between the real-time hourly precipitation and each historical hourly precipitation in the historical rainstorm event data set, and the curve similarity distance is sorted in a positive or negative direction to obtain a plurality of curve similarity sorting L1.

[0104] That is, in the present embodiment, the curve similarity uses the DTW dynamic time warping algorithm, which can fully consider the change characteristics of precipitation over time

[0105] S23. Based on the average rainfall intensity index, calculate the difference between the real-time average rainfall intensity and each historical average rainfall intensity in the historical rainstorm event data set, and sort the differences in a forward or reverse order to obtain multiple average rainfall intensity similarity rankings L2.

[0106] S24. Based on the maximum rainfall intensity index, calculate the difference between the real-time maximum rainfall intensity and each historical maximum rainfall intensity in the historical rainstorm event data set, and sort the differences in a forward or reverse order to obtain multiple maximum rainfall intensity similarity rankings L3.

[0107] That is, finally, the similarity between the average rainfall intensity and the maximum rainfall intensity is directly calculated by the difference method, and together with the above-mentioned L1 and K, it is used as the four indicators of the real-time rainstorm event and each historical rainstorm event. Subsequently, the comprehensive ranking of the four indicators is performed to calculate the comprehensive similarity distance of the similarity ranking. That is, step S3 is specifically as follows:

[0108] S31. Calculate the comprehensive similarity distance of the similarity ranking between the real-time rainstorm event and each historical rainstorm event in the historical rainstorm event dataset according to the formula L=K*(L1+L2+L3).

[0109] S32. Sort each comprehensive similarity distance in a forward or reverse order to obtain a comprehensive similarity ranking of the rainstorm.

[0110] Finally, in this embodiment, step S4 is specifically as follows:

[0111] According to the comprehensive similarity ranking of rainstorms, if the ranking is positive, the historical rainstorm event corresponding to the smallest L value is taken as the similar rainstorm event of the real-time rainstorm event;

[0112] If the sorting is in reverse, the historical rainstorm event corresponding to the largest L value is taken as the similar rainstorm event of the real-time rainstorm event.

[0113] That is, the real-time rainstorm event and each historical rainstorm event have a comprehensive similarity distance. By sorting each comprehensive similarity distance, we can get the historical rainstorm event that is most similar to the most real-time rainstorm event, thus achieving fast and accurate rainstorm similarity analysis.

[0114] Please refer to Figure 3 , the fourth embodiment of the present invention is:

[0115] A rainstorm similarity analysis terminal 1 based on rain type and rain intensity includes a memory 2, a processor 3, and a computer program stored in the memory 2 and executable on the processor 3. In this embodiment, when the processor 3 executes the computer program, the steps in any one of the above-mentioned embodiments 1 to 5 are implemented.

[0116] In summary, the rainstorm similarity analysis method and terminal based on rain type and rain intensity provided by the present invention have the following beneficial effects:

[0117] 1. Screen historical rainstorm events based on hourly precipitation data from the site (meteorological station or rain gauge) to ensure data authenticity;

[0118] 2. The fuzzy recognition method is used to calculate the rain type of rainstorm events. At the same time, the hourly precipitation, average rainfall intensity and maximum rainfall intensity of the rainstorm events are recorded to comprehensively calculate the rainstorm similarity. For real-time rainstorm events, the coefficient K is obtained based on the rain type, hourly precipitation, average rainfall intensity and maximum rainfall intensity. The DTW algorithm is used to calculate the curve similarity ranking L1, the average rainfall intensity similarity ranking L2 and the maximum rainfall intensity similarity ranking L3. The above four indicators are used to calculate the comprehensive similarity distance and obtain the comprehensive similarity ranking of rainstorms, so as to better identify historical rainstorm events similar to real-time rainstorm events.

[0119] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's description and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A rainstorm similarity analysis method based on rain type and rain intensity, characterized by: Including steps: S1. Obtaining real-time hourly precipitation within a current first preset time period, determining whether it is a rainstorm event, and if so, recording it as a real-time rainstorm event, and calculating the rain type and rain intensity of the real-time rainstorm event to obtain the real-time rain type and real-time rain intensity. The real-time rain type is determined by using a fuzzy recognition algorithm based on the seven typical rain types summarized by Mölököv; S2. analyzing similarities between the real-time rain pattern, the real-time hourly precipitation, and the real-time rain intensity and historical rain patterns, historical hourly precipitation, and historical rain intensities of multiple historical rainstorm events; S3. Obtaining a comprehensive similarity ranking of the real-time rainstorm event and each of the historical rainstorm events based on similarity calculation; S4. Identifying similar rainstorm events based on the comprehensive similarity ranking of the rainstorms; The step S2 is specifically as follows: S21. Based on the rain type index, obtain, from the historical rainstorm event dataset, the historical rainstorm events corresponding to the same historical rain type as the real-time rain type, and assign a coefficient K of a first preset value to the historical rainstorm events. For other historical rainstorm events, assign a coefficient K of a second preset value, where the first preset value is less than the second preset value. S22, based on the hourly precipitation index, using the dynamic time warping (DTW) algorithm to calculate the curve similarity distance between the real-time hourly precipitation and each of the historical hourly precipitation in the historical rainstorm event dataset, and sorting the curves in a forward or reverse order according to the curve similarity distance to obtain multiple curve similarity rankings L1; S23, based on the average rainfall intensity index, calculating the difference between the real-time average rainfall intensity and each historical average rainfall intensity in the historical rainstorm event dataset, and sorting the differences in a forward or reverse order to obtain multiple average rainfall intensity similarity rankings L2; S24, based on the maximum rainfall intensity index, calculating the difference between the real-time maximum rainfall intensity and each historical maximum rainfall intensity in the historical rainstorm event dataset, and sorting the differences in a forward or reverse order to obtain multiple maximum rainfall intensity similarity rankings L3; The step S3 is specifically as follows: S31, calculating the comprehensive similarity distance of the similarity ranking between the real-time rainstorm event and each of the historical rainstorm events in the historical rainstorm event dataset according to the formula L=K*(L1+L2+L3); S32. Sort each of the comprehensive similarity distances in a forward or reverse order to obtain a rainstorm comprehensive similarity ranking.

2. A rainstorm similarity analysis method based on rain type and rain intensity according to claim 1, characterized in that: Before step S1, the following steps are also included: S0. Obtain the historical hourly precipitation of a single rain gauge station, and filter out precipitation events with a duration greater than a first preset duration, recording them as historical rainstorm events; The rainfall type and rainfall intensity of each of the historical rainstorm events are calculated to obtain the historical rainfall type and historical rainfall intensity.

3. A rainstorm similarity analysis method based on rain type and rain intensity according to claim 2, characterized in that: The historical rainfall type of each historical rainstorm event is calculated based on the seven typical rainfall types summarized by Mölököv and judged using a fuzzy recognition algorithm. The specific steps are as follows: S02. Counting the total precipitation duration T, total precipitation P, cumulative precipitation duration t, and cumulative precipitation Pt of each historical rainstorm event, dividing the cumulative precipitation duration t by the total precipitation duration T and performing normalization processing to obtain the normalized cumulative precipitation duration curve for each historical rainstorm event; and dividing the cumulative precipitation Pt by the total precipitation P and performing normalization processing to obtain the normalized cumulative precipitation duration curve for each historical rainstorm event. S03, repeating step S02 to calculate the cumulative precipitation duration curves of the seven typical rain types; S04. For all normalized accumulated precipitation duration curves, a point is taken every 0.01 on the horizontal coordinate to obtain the accumulated precipitation value on the corresponding vertical coordinate on the curve; S05. Calculate the Euclidean distance between each of the historical rainstorm events and the seven typical rain patterns through all the points selected in step S03, select the typical rain pattern with the smallest distance as the corresponding rain pattern of the historical rainstorm event, and obtain the historical rain pattern corresponding to each of the historical rainstorm events.

4. A rainstorm similarity analysis method based on rain type and rain intensity according to claim 2, characterized in that: The historical rainfall intensity of each historical rainstorm event includes the historical average rainfall intensity and the historical maximum rainfall intensity; The calculation of the historical rainfall intensity of each historical rainstorm event is specifically as follows: S06. Calculate the average and maximum values ​​of the historical hourly precipitation corresponding to each of the historical rainstorm events to obtain the historical average rainfall intensity and the historical maximum rainfall intensity of each of the historical rainstorm events.

5. A rainstorm similarity analysis method based on rain type and rain intensity according to claim 4, characterized in that: Before step S1, the following steps are also included: The historical rain type, the historical hourly precipitation, the historical average rainfall intensity and the historical maximum rainfall intensity are used as the four indicators of the historical rainstorm events, and the four indicators of each historical rainstorm event are saved in the historical rainstorm event data set.

6. A rainstorm similarity analysis method based on rain type and rain intensity according to claim 5, characterized in that: The step S1 is specifically as follows: S11, obtaining the real-time hourly precipitation within the first preset time period, determining whether it is a rainstorm event, and if so, recording it as a real-time rainstorm event; S12. Calculate the rain type and rain intensity of the real-time rainstorm event in the same manner as described in the calculation method of the historical rain type and the historical rain intensity, and obtain the real-time rain type, real-time average rain intensity and real-time maximum rain intensity.

7. The rainstorm similarity analysis method based on rain type and rain intensity according to claim 1, characterized in that: The step S4 is specifically as follows: According to the rainstorm comprehensive similarity ranking, if the ranking is positive, the historical rainstorm event corresponding to the smallest L value is used as a similar rainstorm event to the real-time rainstorm event; If the sorting is in reverse, the historical rainstorm event corresponding to the largest L value is used as a similar rainstorm event to the real-time rainstorm event.

8. A rainstorm similarity analysis terminal based on rain type and rain intensity, characterized by: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps in the method for rainstorm similarity analysis based on rain type and rain intensity as described in any one of claims 1 to 7 are implemented.

Images