A flood season runoff scenario targeting extraction method

Abstract

The application provides a flood period runoff scene targeted extraction method, collects long series daily natural flow processes of a hydrological station in a flood period; calculates annual evaluation indexes of the hydrological station, constructs a flood period flow characteristic matrix; adopts a synchronous back substitution reduction algorithm to reduce scenes of the flood period flow characteristic matrix, determines an optimal reduction scene number and a reduced flood period flow characteristic matrix based on an elbow principle; takes the flood period flow process corresponding to the reduced flood period flow characteristic matrix as a classification center, calculates the Euclidean distance between the flood period flow process of each year and each classification center, divides the flood period flow process into the classification center with the minimum distance, and obtains a clustered and decomposed flood period flow process matrix; calculates the evaluation index mean of the clustered and decomposed flood period flow characteristic matrix of each classification center, evaluates the clustered and decomposed flood period flow characteristic; determines the optimal reduction scene number and the reduced flood period flow characteristic matrix, and obtains the clustered and decomposed flood period flow process matrix.

Description

Technical Field

[0001] This invention relates to the field of flood season runoff feature recognition and scene classification, and in particular to a method for targeted extraction of flood season runoff scenes. Background Technology

[0002] Flood season runoff scenarios are important input data for flood control scheduling of cascade reservoirs. As the requirements for refined scheduling of cascade reservoirs continue to increase, the characteristics of flood season runoff scenarios are becoming increasingly important for the safety of flood control scheduling in the basin.

[0003] Generally, river runoff distribution varies significantly throughout the year, with the flood season often concentrating over 70% of the annual runoff and resulting in multiple flood events. Flood season runoff also varies considerably across different typical years, and due to global climate change and human activities, periods of drought during the flood season are not uncommon. Furthermore, natural flood season runoff data spans a long period, is complex, and exhibits variability, all of which increase the difficulty of identifying the characteristics of flood season runoff scenarios. Therefore, extracting flow scenarios that reflect the characteristics of flood season runoff processes and guiding targeted reservoir management is a crucial research task. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of the prior art by providing a targeted extraction method for runoff scenarios during the flood season, thus providing technical support for targeted scheduling of cascade reservoirs under specific flood season water conditions.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for targeted extraction of runoff scenarios during the flood season, comprising: S1. Collect daily natural flow data during the flood season from a series of hydrological station chiefs; S2. Calculate the annual evaluation indicators of hydrological stations and construct a flood season flow characteristic matrix; S3. The synchronous back-substitution reduction algorithm is used to reduce the flood season flow feature matrix into scenarios. Based on the elbow principle, the optimal number of reduced scenarios and the reduced flood season flow feature matrix are determined. S4. Using the flood season flow process corresponding to the reduced flood season flow feature matrix as the classification center, calculate the Euclidean distance between the flood season flow process of each year and each classification center, and divide the flood season flow process into the classification center with the smallest distance to obtain the flood season flow process matrix after clustering decomposition. S5. Calculate the mean of the evaluation index of the flood season flow characteristic matrix of each classification center after cluster decomposition, and evaluate the flood season flow characteristics after cluster decomposition.

[0006] Furthermore, in S1, the daily natural flow process during the flood season, collected from a series of hydrological stations, is specifically defined as: ; in, This represents a matrix showing the daily natural flow process during the flood season for a series of hydrological stations. Indicates the hydrological station number The first flood season of the year Daily natural flow; This indicates the total number of years to be calculated; This indicates the total number of days in the flood season.

[0007] Furthermore, in S2, the annual evaluation indicators for the hydrological station are calculated as follows: ; ; ; ; ; ; ; ; ; ; ; ; ; ; in, Indicates the first Average annual flood season flow; Indicates the first Annual flood season flow standard deviation; Indicates the first Annual flood season flow variation coefficient; Indicates the first Maximum annual flood season flow; Indicates the first Minimum annual flood season flow; Indicates the first Annual flood season flow peak-valley difference; Indicates the first Annual flood season flow skewness coefficient; Indicates the first Annual flood season flow peak coefficient; Indicates the first Average annual flow during the main flood season; Indicates the first Average flow during the post-holiday flood season; and These respectively indicate the start and end times of the main flood season; and These respectively indicate the start and end times of the post-flood season; Indicates the first The annual flood season saw continuous increases in maximum flow rate; Indicates the first The percentage increase in annual flood season flow; Indicates the first Year From the 1st to the 1st Daily traffic increase vector; Indicates the first Annual adjacent two-day flow rate variation vector; This indicates the number of positive numbers extracted from the flow rate variation vectors of two consecutive days; Indicates the first Maximum 3-day flood volume during the annual flood season; Indicates the first Year The daily flood volume for three consecutive days during the flood season; Indicates the first Maximum 7-day flood volume during the annual flood season; Indicates the first Year The daily flood volume for 7 consecutive days during the flood season; This indicates a long calculation period; In S2, the flood season flow characteristic matrix is ​​constructed as follows: ; in, This represents the constructed flood season flow characteristic matrix.

[0008] Furthermore, S3 specifically refers to: S301. Index Normalization: Normalize the indexes of each column vector in the constructed flood season flow characteristic matrix. The calculation formula is as follows: ; in, Represents the normalized i-th Average annual flood season flow; This represents the maximum average flow rate during the flood season over the years. This represents the minimum average flow rate during the flood season over the years. Represents the normalized i-th Annual flood season flow standard deviation; This represents the maximum standard deviation of flow rate during the flood season over the years; This represents the minimum standard deviation of the annual flood season flow rate. Represents the normalized i-th Annual flood season flow variation coefficient; This represents the maximum value of the coefficient of variation of flow during the flood season over the years; This represents the minimum value of the coefficient of variation of flow during the flood season over the years; Represents the normalized i-th Maximum annual flood season flow; This represents the maximum value among the historical flood season maximum values; This represents the minimum value among the maximum values ​​of the flood season over the years; Represents the normalized i-th Minimum annual flood season flow; This represents the maximum value among the minimum values ​​of the flood season over the years; This represents the minimum of the historical minimum flood season discharge values. Represents the normalized i-th Annual flood season flow peak-valley difference; This represents the maximum value among the peak-to-valley differences in flow during the flood season over the years; This represents the minimum value among the peak and valley differences in flow during the flood season over the years; Represents the normalized i-th Annual flood season flow skewness coefficient; This represents the maximum value among the flow skewness coefficients during the flood season over the years; This represents the minimum value among the flow skewness coefficients during the flood season over the years; Represents the normalized i-th Annual flood season flow peak coefficient; This represents the maximum value among the kurtosis coefficients of the flood season over the years; This represents the minimum value among the kurtosis coefficients of flood seasons over the years; Represents the normalized i-th Average annual flow during the main flood season; This represents the maximum value among the average flows during the main flood season over the years; This represents the minimum average flow rate during the main flood season over the years. Represents the normalized i-th Average flow during the post-holiday flood season; This represents the maximum value among the average flow rates during the flood season over the years. This represents the minimum value among the average flood season flows over the years. Represents the normalized i-th The annual flood season saw continuous increases in maximum flow rate; This represents the maximum value among the consecutive maximum flows during the flood season over the years; This represents the minimum value among the maximum consecutive increases in flow rate during the flood season over the years; Represents the normalized i-th The percentage increase in annual flood season flow; This represents the maximum percentage increase in flow rate during the flood season over the years. This represents the minimum percentage increase in flow rate during the flood season over the years. Represents the normalized i-th Maximum 3-day flood volume during the annual flood season; This represents the maximum value among the three largest flood volumes during the flood season in previous years; This represents the minimum value among the maximum 3-day flood volumes during the flood season in previous years; Represents the normalized i-th Maximum 7-day flood volume during the annual flood season; This represents the maximum value among the largest 7-day flood volumes during the flood season in previous years; This represents the minimum value among the maximum 7-day flood volumes during the flood season in previous years; S302, Set the number of scenes after reduction and maximum number of scenes ,make ; S303. The synchronous back-substitution reduction algorithm is used to reduce the scene of the normalized flood season flow feature matrix. The calculation formula is as follows: ; in, This represents the normalized flood season flow characteristic matrix; This indicates the synchronous back-substitution reduction algorithm; and The reduced flood season flow characteristic scenario matrix and its corresponding probability matrix are shown below: ; ; in, Represents the normalized i-th Average flow rate during the flood season in the scenario; Represents the normalized i-th Standard deviation of flood season flow rate; Represents the normalized i-th Scenario-specific flow variation coefficient during flood season; Represents the normalized i-th Maximum flow rate during the flood season in this scenario; Represents the normalized i-th Minimum flow rate during the flood season in this scenario; Represents the normalized i-th Scene of peak and valley flow differences during the flood season; Represents the normalized i-th Scene flood season flow skewness coefficient; Represents the normalized i-th Peak flow coefficient during the flood season in this scenario; Represents the normalized i-th Average flow during the main flood season in the scenario; Represents the normalized i-th Average flow rate after the flood season in the scenario; Represents the normalized i-th The maximum flow rate in the scene has been rising continuously during the flood season; Represents the normalized i-th The percentage increase in traffic during the flood season in the relevant scenarios; Represents the normalized i-th Maximum 3-day flood volume during the flood season; Represents the normalized i-th Maximum 7-day flood volume during the flood season; Indicates the reduced number of... Scenario-corresponding probability; S304, Order Repeat step S303 until... The scene dataset is obtained as follows: ; in, D _ C Indicates inclusion K A scene dataset consisting of a reduced flood season flow characteristic scene matrix; S305. Apply the elbow rule to the scene dataset. Calculations are performed to determine the optimal scene reduction number, specifically: S3051. Calculate the weighted sum of squared errors between the original scene and the reduced scene: ; ; ; in, This represents the weighted squared error vector between the original scene and the reduced scene; Represents the original scene and the reduced scene. Weighted sum of squared errors between scenarios; The first element in the normalized flood season flow characteristic matrix represents the... Scene; This represents the reduced flood season flow characteristic scenario matrix. Scene; S3052, Drawing Compared with the reduced number of scenes Based on the correlation curves between them and the elbow points in the graph, determine the optimal number of scenes to reduce. ; S306. Based on the optimal reduction in the number of scenes Using synchronous back-substitution reduction technology, the reduced flood season flow characteristic matrix is ​​obtained, as follows: ; ; ; in, and This indicates the use of the optimal reduction number. Reduced flood season flow characteristic scenario matrix and its corresponding probability matrix; Indicates the reduced number of... Probability corresponding to a scenario.

[0009] Furthermore, in S4, the flood season flow process corresponding to the reduced flood season flow feature matrix is ​​used as the classification center, specifically defined as: ; in, This represents the reduced matrix of daily natural flow processes during the flood season for a series of hydrological stations. This indicates the number of hydrological stations after reduction. Scene Flood Season Daily natural flow; In step S4, the Euclidean distance between the annual flood season flow process and each classification center is calculated, and the flood season flow process is assigned to the classification center with the smallest distance, resulting in the flood season flow process matrix after clustering decomposition. Specifically: ; ; ; ; in, Indicates the first Annual flood season flow process and the reduced first Euclidean distance between classification centers; Indicates the first The annual flood season flow process is assigned to the subscript of the classification center that is closest to it; Indicates the division to the th The set of indices for the classification centers; Indicates the search function; This indicates that after cluster decomposition, the data is divided into the th... A series of daily natural flow process matrices during the flood season for hydrological station chiefs at the classification center; Indicates the division to the th Hydrological station chiefs at the classification center, Flood season series Daily natural flow.

[0010] Furthermore, S5 specifically includes: ; ; ; ; ; ; ; ; ; ; ; ; ; ; in, , , , , , , , , , , , , , They represent the first The average values ​​of evaluation indicators for the flood season flow characteristic matrix of the classification center, including the average flow during the flood season, standard deviation, coefficient of variation, maximum value, minimum value, peak-to-valley difference, skewness, kurtosis, average flow during the main flood season, average flow during the post-flood season, maximum continuous rise in flow, cumulative number of days of rise, maximum 3-day flood volume, and maximum 7-day flood volume. Indicates the division to the th Total number of scenarios in the classification center.

[0011] The beneficial effects of this invention are as follows: a flood season flow characteristic matrix is ​​constructed, and the synchronous back-substitution reduction algorithm and elbow principle are adopted to determine the optimal number of reduction scenarios and the reduced flood season flow characteristic matrix, and the flood season flow process matrix after cluster decomposition is obtained, which provides technical support for cascade hydropower stations to carry out scheduling based on flood season runoff characteristics. Attached Figure Description

[0012] Figure 1 A flowchart of a method for targeted extraction of runoff scenarios during the flood season; Figure 2 This is a series of daily natural flow processes during the flood season for hydrological station chiefs; Figure 3 This is a schematic diagram of the flood season flow characteristic matrix; Figure 4 This is a diagram illustrating the optimal reduction in the number of scenes; Figure 5 This is a schematic diagram of the reduced flood season flow process; Figure 6 A schematic diagram illustrating the classification of runoff scenarios during the flood season. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0014] Please see Figure 1 A targeted extraction method for flood season runoff scenarios, comprising: S1. Collect daily natural flow data during the flood season from a series of hydrological station chiefs; S2. Calculate the annual evaluation indicators of hydrological stations and construct a flood season flow characteristic matrix; The evaluation indicators include the average flow during the flood season, standard deviation, coefficient of variation, maximum value, minimum value, peak-to-valley difference, skewness, kurtosis, average flow during the main flood season, average flow during the post-flood season, maximum continuous rise in flow, cumulative number of days of rise, maximum 3-day flood volume, and maximum 7-day flood volume.

[0015] S3. The synchronous back-substitution reduction algorithm is used to reduce the flood season flow feature matrix into scenarios. Based on the elbow principle, the optimal number of reduced scenarios and the reduced flood season flow feature matrix are determined. S4. Using the flood season flow process corresponding to the reduced flood season flow feature matrix as the classification center, calculate the Euclidean distance between the flood season flow process of each year and each classification center, and divide the flood season flow process into the classification center with the smallest distance to obtain the flood season flow process matrix after clustering decomposition. S5. Calculate the mean of the evaluation index of the flood season flow characteristic matrix of each classification center after cluster decomposition, and evaluate the flood season flow characteristics after cluster decomposition.

[0016] Among them, the mean of the evaluation indicators refers to the average flow during the flood season, standard deviation, coefficient of variation, maximum value, minimum value, peak-to-valley difference, skewness, kurtosis, average flow during the main flood season, average flow during the post-flood season, maximum continuous rise in flow, cumulative number of days of rise, maximum 3-day flood volume, and maximum 7-day flood volume.

[0017] In S1, the daily natural flow process during the flood season is collected from a series of hydrological stations, such as... Figure 2 As shown, it is specifically defined as: ; in, This represents a matrix showing the daily natural flow process during the flood season for a series of hydrological stations. Indicates the hydrological station number The first flood season of the year Daily natural flow; This indicates the total number of years to be calculated; This indicates the total number of days in the flood season.

[0018] In S2, the annual evaluation indicators for the hydrological station are calculated as follows: ; ; ; ; ; ; ; ; ; ; ; ; ; ; in, Indicates the first Average annual flood season flow; Indicates the first Annual flood season flow standard deviation; Indicates the first Annual flood season flow variation coefficient; Indicates the first Maximum annual flood season flow; Indicates the first Minimum annual flood season flow; Indicates the first Annual flood season flow peak-valley difference; Indicates the first Annual flood season flow skewness coefficient; Indicates the first Annual flood season flow peak coefficient; Indicates the first Average annual flow during the main flood season; Indicates the first Average flow during the post-holiday flood season; and These respectively indicate the start and end times of the main flood season; and These respectively indicate the start and end times of the post-flood season; Indicates the first The annual flood season saw continuous increases in maximum flow rate; Indicates the first The percentage increase in annual flood season flow; Indicates the first Year From the 1st to the 1st Daily traffic increase vector; Indicates the first Annual adjacent two-day flow rate variation vector; This indicates the number of positive numbers extracted from the flow rate variation vectors of two consecutive days; Indicates the first Maximum 3-day flood volume during the annual flood season; Indicates the first Year The daily flood volume for three consecutive days during the flood season; Indicates the first Maximum 7-day flood volume during the annual flood season; Indicates the first Year The daily flood volume for 7 consecutive days during the flood season; This indicates a long calculation period; In S2, a flood season flow characteristic matrix is ​​constructed, such as... Figure 3 As shown, specifically: ; in, This represents the constructed flood season flow characteristic matrix.

[0019] Specifically, S3 is: S301. Index Normalization: Normalize the indexes of each column vector in the constructed flood season flow characteristic matrix. The calculation formula is as follows: ; in, Represents the normalized i-th Average annual flood season flow; This represents the maximum average flow rate during the flood season over the years. This represents the minimum average flow rate during the flood season over the years. Represents the normalized i-th Annual flood season flow standard deviation; This represents the maximum standard deviation of flow rate during the flood season over the years; This represents the minimum standard deviation of the annual flood season flow rate. Represents the normalized i-th Annual flood season flow variation coefficient; This represents the maximum value of the coefficient of variation of flow during the flood season over the years; This represents the minimum value of the coefficient of variation of flow during the flood season over the years; Represents the normalized i-th Maximum annual flood season flow; This represents the maximum value among the historical flood season maximum values; This represents the minimum value among the maximum values ​​of the flood season over the years; Represents the normalized i-th Minimum annual flood season flow; This represents the maximum value among the minimum values ​​of the flood season over the years; This represents the minimum of the historical minimum flood season discharge values. Represents the normalized i-th Annual flood season flow peak-valley difference; This represents the maximum value among the peak-to-valley differences in flow during the flood season over the years; This represents the minimum value among the peak and valley differences in flow during the flood season over the years; Represents the normalized i-th Annual flood season flow skewness coefficient; This represents the maximum value among the flow skewness coefficients during the flood season over the years; This represents the minimum value among the flow skewness coefficients during the flood season over the years; Represents the normalized i-th Annual flood season flow peak coefficient; This represents the maximum value among the kurtosis coefficients of the flood season over the years; This represents the minimum value among the kurtosis coefficients of flood seasons over the years; Represents the normalized i-th Average annual flow during the main flood season; This represents the maximum value among the average flows during the main flood season over the years; This represents the minimum average flow rate during the main flood season over the years. Represents the normalized i-th Average flow during the post-holiday flood season; This represents the maximum value among the average flow rates during the flood season over the years. This represents the minimum value among the average flood season flows over the years. Represents the normalized i-th The annual flood season saw continuous increases in maximum flow rate; This represents the maximum value among the consecutive maximum flows during the flood season over the years; This represents the minimum value among the maximum consecutive increases in flow rate during the flood season over the years; Represents the normalized i-th The percentage increase in annual flood season flow; This represents the maximum percentage increase in flow rate during the flood season over the years. This represents the minimum percentage increase in flow rate during the flood season over the years. Represents the normalized i-th Maximum 3-day flood volume during the annual flood season; This represents the maximum value among the three largest flood volumes during the flood season in previous years; This represents the minimum value among the maximum 3-day flood volumes during the flood season in previous years; Represents the normalized i-th Maximum 7-day flood volume during the annual flood season; This represents the maximum value among the largest 7-day flood volumes during the flood season in previous years; This represents the minimum value among the maximum 7-day flood volumes during the flood season in previous years; S302, Set the number of scenes after reduction and maximum number of scenes =50, let ; S303. The synchronous back-substitution reduction algorithm is used to reduce the scene of the normalized flood season flow feature matrix. The calculation formula is as follows: ; in, This represents the normalized flood season flow characteristic matrix; This indicates the synchronous back-substitution reduction algorithm; and The reduced flood season flow characteristic scenario matrix and its corresponding probability matrix are shown below: ; ; in, Represents the normalized i-th Average flow rate during the flood season in the scenario; Represents the normalized i-th Standard deviation of flood season flow rate; Represents the normalized i-th Scenario-specific flow variation coefficient during flood season; Represents the normalized i-th Maximum flow rate during the flood season in this scenario; Represents the normalized i-th Minimum flow rate during the flood season in this scenario; Represents the normalized i-th Scene of peak and valley flow differences during the flood season; Represents the normalized i-th Scene flood season flow skewness coefficient; Represents the normalized i-th Peak flow coefficient during the flood season in this scenario; Represents the normalized i-th Average flow during the main flood season in the scenario; Represents the normalized i-th Average flow rate after the flood season in the scenario; Represents the normalized i-th The maximum flow rate in the scene has been rising continuously during the flood season; Represents the normalized i-th The percentage increase in traffic during the flood season in the relevant scenarios; Represents the normalized i-th Maximum 3-day flood volume during the flood season; Represents the normalized i-th Maximum 7-day flood volume during the flood season; Indicates the reduced number of... Scenario-corresponding probability; S304, Order Repeat step S303 until... The scene dataset is obtained as follows: ; in, D _ C Indicates inclusion K A scene dataset consisting of a reduced flood season flow characteristic scene matrix; S305. Apply the elbow rule to the scene dataset. Calculations are performed to determine the optimal scene reduction number, specifically: S3051. Calculate the weighted sum of squared errors between the original scene and the reduced scene: ; ; ; in, This represents the weighted squared error vector between the original scene and the reduced scene; Represents the original scene and the reduced scene. Weighted sum of squared errors between scenarios; The first element in the normalized flood season flow characteristic matrix represents the... Scene; This represents the reduced flood season flow characteristic scenario matrix. Scene; S3052, Drawing Compared with the reduced number of scenes Correlation curves between them, such as Figure 4 As shown in the figure, the optimal number of scenes to be reduced is determined based on the elbow points. ; S306. Based on the optimal reduction in the number of scenes Using synchronous back-substitution reduction technology, the reduced flood season flow characteristic matrix was obtained, as shown in Table 1. Specifically: ; ; ; in, and This indicates the use of the optimal reduction number. Reduced flood season flow characteristic scenario matrix and its corresponding probability matrix; Indicates the reduced number of... Probability corresponding to a scenario.

[0020] Table 1. Reduced flood season flow characteristic matrix

[0021] In S4, the flood season flow process corresponding to the reduced flood season flow feature matrix is ​​used as the classification center, such as... Figure 5 As shown, it is specifically defined as: ; in, This represents the reduced matrix of daily natural flow processes during the flood season for a series of hydrological stations. This indicates the number of hydrological stations after reduction. Scene Flood Season Daily natural flow; In step S4, the Euclidean distance between the annual flood season flow process and each classification center is calculated, and the flood season flow process is assigned to the classification center with the smallest distance, resulting in the flood season flow process matrix after clustering decomposition. Specifically: ; ; ; ; in, Indicates the first Annual flood season flow process and the reduced first Euclidean distance between classification centers; Indicates the first The annual flood season flow process is assigned to the subscript of the classification center that is closest to it; Indicates the division to the th The set of indices for the classification centers; Indicates the search function; This indicates that after cluster decomposition, the data is divided into the th... A series of daily natural flow process matrices during the flood season for hydrological station chiefs at the classification center; Indicates the division to the th Hydrological station chiefs at the classification center, Flood season series Daily natural flow.

[0022] like Figure 6 As shown in Table 2, S5 specifically includes: ; ; ; ; ; ; ; ; ; ; ; ; ; ; in, , , , , , , , , , , , , , They represent the first The average values ​​of evaluation indicators for the flood season flow characteristic matrix of the classification center, including the average flow during the flood season, standard deviation, coefficient of variation, maximum value, minimum value, peak-to-valley difference, skewness, kurtosis, average flow during the main flood season, average flow during the post-flood season, maximum continuous rise in flow, cumulative number of days of rise, maximum 3-day flood volume, and maximum 7-day flood volume. Indicates the division to the th Total number of scenarios in the classification center.

[0023] Table 2. Classification Feature Values ​​of Runoff Scenarios During Flood Season

[0024] As shown in Table 2, the characteristic values ​​of the indicators vary considerably across different scenarios. Scenario 1 is characterized by a large proportion of inflow during the post-flood season, with the main peak occurring during the post-flood season, resulting in rapid flow increases and large flood volumes. Scenario 2 is characterized by a single-peak process, with concentrated flood volumes and rapid flow increases. Scenario 3 is characterized by a large proportion of inflow during the post-flood season, with smaller fluctuations in the continuous flow increase compared to scenarios 1 and 2. Scenario 4 is characterized by a large overall inflow during the flood season, with a predominance of inflow during the post-flood season, and large maximum 3-day and 7-day flood volumes. Scenario 5 is characterized by a relatively flat inflow during the flood season, with roughly equal inflows during the main and post-flood seasons. Scenario 6 is characterized by the highest frequency of occurrence, with generally small maximum flow values ​​during the flood season, smaller maximum 3-day and 7-day flood volumes, and moderate flow fluctuations. Scenario 7 is characterized by a large proportion of inflow during the main flood season, with the average flood season flow being slightly higher than the multi-year average. Scenario 8 is characterized by a relatively low inflow during the flood season, with a large proportion of inflow during the post-flood season, and the maximum flood season flow concentrated in the post-flood season.

[0025] The embodiments described above are merely illustrative of implementation methods of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for targeted extraction of runoff scenarios during the flood season, characterized in that, include: S1. Collect daily natural flow data during the flood season from a series of hydrological station chiefs; S2. Calculate the annual evaluation indicators of hydrological stations and construct a flood season flow characteristic matrix; S3. The synchronous back-substitution reduction algorithm is used to reduce the flood season flow feature matrix into scenarios. Based on the elbow principle, the optimal number of reduced scenarios and the reduced flood season flow feature matrix are determined. S4. Using the flood season flow process corresponding to the reduced flood season flow feature matrix as the classification center, calculate the Euclidean distance between the flood season flow process of each year and each classification center, and divide the flood season flow process into the classification center with the smallest distance to obtain the flood season flow process matrix after clustering decomposition. S5. Calculate the mean of the evaluation index of the flood season flow characteristic matrix of each classification center after cluster decomposition, and evaluate the flood season flow characteristics after cluster decomposition.

2. The method for targeted extraction of runoff scenarios during the flood season according to claim 1, characterized in that: In S1, the daily natural flow process during the flood season, collected from the hydrological station chiefs, is specifically defined as follows: ； in, This represents a matrix showing the daily natural flow process during the flood season for a series of hydrological stations. Indicates the hydrological station number The first flood season of the year Daily natural flow; This indicates the total number of years to be calculated; This indicates the total number of days in the flood season.

3. The method for targeted extraction of runoff scenarios during the flood season according to claim 2, characterized in that, In S2, the annual evaluation indicators for the hydrological station are calculated as follows: ； ； ； ； ； ； ； ； ； ； ； ； ； ； in, Indicates the first Average annual flood season flow; Indicates the first Annual flood season flow standard deviation; Indicates the first Annual flood season flow variation coefficient; Indicates the first Maximum annual flood season flow; Indicates the first Minimum annual flood season flow; Indicates the first Annual flood season flow peak-valley difference; Indicates the first Annual flood season flow skewness coefficient; Indicates the first Annual flood season flow peak coefficient; Indicates the first Average annual flow during the main flood season; Indicates the first Average flow during the post-holiday flood season; and These respectively indicate the start and end times of the main flood season; and These respectively indicate the start and end times of the post-flood season; Indicates the first The annual flood season saw continuous increases in maximum flow rate; Indicates the first The percentage increase in annual flood season flow; Indicates the first Year From the 1st to the 1st Daily traffic increase vector; Indicates the first Annual adjacent two-day flow rate variation vector; This indicates the number of positive numbers extracted from the flow rate variation vectors of two consecutive days; Indicates the first Maximum 3-day flood volume during the annual flood season; Indicates the first Year The daily flood volume for three consecutive days during the flood season; Indicates the first Maximum 7-day flood volume during the annual flood season; Indicates the first Year The daily flood volume for 7 consecutive days during the flood season; This indicates a long calculation period; In S2, the flood season flow characteristic matrix is ​​constructed as follows: ； in, This represents the constructed flood season flow characteristic matrix.

4. The method for targeted extraction of runoff scenarios during the flood season according to claim 3, characterized in that, Specifically, S3 is: S301. Index Normalization: Normalize the indexes of each column vector in the constructed flood season flow characteristic matrix. The calculation formula is as follows: ； in, Represents the normalized i-th Average annual flood season flow; This represents the maximum average flow rate during the flood season over the years. This represents the minimum average flow rate during the flood season over the years. Represents the normalized i-th Annual flood season flow standard deviation; This represents the maximum standard deviation of flow rate during the flood season over the years; This represents the minimum standard deviation of the annual flood season flow rate. Represents the normalized i-th Annual flood season flow variation coefficient; This represents the maximum value of the coefficient of variation of flow during the flood season over the years; This represents the minimum value of the coefficient of variation of flow during the flood season over the years; Represents the normalized i-th Maximum annual flood season flow; This represents the maximum value among the historical flood season maximum values; This represents the minimum value among the maximum values ​​of the flood season over the years; Represents the normalized i-th Minimum annual flood season flow; This represents the maximum value among the minimum values ​​of the flood season over the years; This represents the minimum of the historical minimum flood season discharge values. Represents the normalized i-th Annual flood season flow peak-valley difference; This represents the maximum value among the peak-to-valley differences in flow during the flood season over the years; This represents the minimum value among the peak and valley differences in flow during the flood season over the years; Represents the normalized i-th Annual flood season flow skewness coefficient; This represents the maximum value among the flow skewness coefficients during the flood season over the years; This represents the minimum value among the flow skewness coefficients during the flood season over the years; Represents the normalized i-th Annual flood season flow peak coefficient; This represents the maximum value among the kurtosis coefficients of the flood season over the years; This represents the minimum value among the kurtosis coefficients of flood seasons over the years; Represents the normalized i-th Average annual flow during the main flood season; This represents the maximum value among the average flows during the main flood season over the years; This represents the minimum average flow rate during the main flood season over the years. Represents the normalized i-th Average flow during the post-holiday flood season; This represents the maximum value among the average flow rates during the flood season over the years. This represents the minimum value among the average flood season flows over the years. Represents the normalized i-th The annual flood season saw continuous increases in maximum flow rate; This represents the maximum value among the consecutive maximum flows during the flood season over the years; This represents the minimum value among the maximum consecutive increases in flow rate during the flood season over the years; Represents the normalized i-th The percentage increase in annual flood season flow; This represents the maximum percentage increase in flow rate during the flood season over the years. This represents the minimum percentage increase in flow rate during the flood season over the years. Represents the normalized i-th Maximum 3-day flood volume during the annual flood season; This represents the maximum value among the three largest flood volumes during the flood season in previous years; This represents the minimum value among the maximum 3-day flood volumes during the flood season in previous years; Represents the normalized i-th Maximum 7-day flood volume during the annual flood season; This represents the maximum value among the largest 7-day flood volumes during the flood season in previous years; This represents the minimum value among the maximum 7-day flood volumes during the flood season in previous years; S302, Set the number of scenes after reduction and maximum number of scenes ,make ; S303. The synchronous back-substitution reduction algorithm is used to reduce the scene of the normalized flood season flow feature matrix. The calculation formula is as follows: ； in, This represents the normalized flood season flow characteristic matrix; This indicates the synchronous back-substitution reduction algorithm; and The reduced flood season flow characteristic scenario matrix and its corresponding probability matrix are shown below: ； ； in, Represents the normalized i-th Average flow rate during the flood season in the scenario; Represents the normalized i-th Standard deviation of flood season flow rate; Represents the normalized i-th Scenario-specific flow variation coefficient during flood season; Represents the normalized i-th Maximum flow rate during the flood season in this scenario; Represents the normalized i-th Minimum flow rate during the flood season in this scenario; Represents the normalized i-th Scene of peak and valley flow differences during the flood season; Represents the normalized i-th Scene flood season flow skewness coefficient; Represents the normalized i-th Peak flow coefficient during the flood season in this scenario; Represents the normalized i-th Average flow during the main flood season in the scenario; Represents the normalized i-th Average flow rate after the flood season in the scenario; Represents the normalized i-th The maximum flow rate in the scene has been rising continuously during the flood season; Represents the normalized i-th The percentage increase in traffic during the flood season in the relevant scenarios; Represents the normalized i-th Maximum 3-day flood volume during the flood season; Represents the normalized i-th Maximum 7-day flood volume during the flood season; Indicates the reduced number of... Scenario-corresponding probability; S304, Order Repeat step S303 until... The scene dataset is obtained as follows: ； in, D _ C Indicates inclusion K A scene dataset consisting of a reduced flood season flow characteristic scene matrix; S305. Apply the elbow rule to the scene dataset. Calculations are performed to determine the optimal scene reduction number, specifically: S3051. Calculate the weighted sum of squared errors between the original scene and the reduced scene: ； ； ； in, This represents the weighted squared error vector between the original scene and the reduced scene; Represents the original scene and the reduced scene. Weighted sum of squared errors between scenarios; The first element in the normalized flood season flow characteristic matrix represents the... Scene; This represents the reduced flood season flow characteristic scenario matrix. Scene; S3052, Drawing Compared with the reduced number of scenes Based on the correlation curves between them and the elbow points in the graph, determine the optimal number of scenes to reduce. ; S306. Based on the optimal reduction in the number of scenes Using synchronous back-substitution reduction technology, the reduced flood season flow characteristic matrix is ​​obtained, as follows: ； ； ； in, and This indicates the use of the optimal reduction number. Reduced flood season flow characteristic scenario matrix and its corresponding probability matrix; Indicates the reduced number of... Probability corresponding to a scenario.

5. The method for targeted extraction of runoff scenarios during the flood season according to claim 4, characterized in that: In step S4, the flood season flow process corresponding to the reduced flood season flow feature matrix is ​​used as the classification center, specifically: ； in, This represents the reduced matrix of daily natural flow processes during the flood season for a series of hydrological stations. This indicates the number of hydrological stations after reduction. Scene Flood Season Daily natural flow; In step S4, the Euclidean distance between the annual flood season flow process and each classification center is calculated, and the flood season flow process is assigned to the classification center with the smallest distance, resulting in the flood season flow process matrix after clustering decomposition. Specifically: ； ； ； ； in, Indicates the first Annual flood season flow process and the reduced first Euclidean distance between classification centers; Indicates the first The annual flood season flow process is assigned to the subscript of the classification center that is closest to it; Indicates the division to the th The set of indices for the classification centers; Indicates the search function; This indicates that after cluster decomposition, the data is divided into the th... A series of daily natural flow process matrices during the flood season for hydrological station chiefs at the classification center; Indicates the division to the th Hydrological station chiefs at the classification center, Flood season series Daily natural flow.

6. The method for targeted extraction of runoff scenarios during the flood season according to claim 5, characterized in that, Specifically, S5 is: ； ； ； ； ； ； ； ； ； ； ； ； ； ； in, , , , , , , , , , , , , , They represent the first The average values ​​of evaluation indicators for the flood season flow characteristic matrix of the classification center, including the average flow during the flood season, standard deviation, coefficient of variation, maximum value, minimum value, peak-to-valley difference, skewness, kurtosis, average flow during the main flood season, average flow during the post-flood season, maximum continuous rise in flow, cumulative number of days of rise, maximum 3-day flood volume, and maximum 7-day flood volume. Indicates the division to the th Total number of scenarios in the classification center.

Images