Flood prediction method, device, computer equipment and storage medium

Abstract

The present application relates to a flood prediction method, device, computer equipment and storage medium. The method comprises: by geographically gridding multiple basins contained in the study area, determining the seasonal index of each grid center, and then dividing all grids into multiple homogeneous regions according to the similarity between the seasonal indexes of all grid centers, and then for each homogeneous region, based on the historical flood data of the first type of basin in the homogeneous region, training a flood prediction model for the homogeneous region, and finally, inputting the meteorological characteristics and underlying surface characteristics of the second type of basin into the flood prediction model corresponding to the homogeneous region to which the basin to be predicted belongs, and obtaining the flood frequency of the second type of basin. By adopting this method, when the homogeneous regions are divided based on the seasonal index, the homogeneous regions can be divided more accurately, thereby improving the accuracy of the prediction results of the flood prediction model.

Description

Technical Field

[0001] The present application relates to the field of meteorological and hydrological technology, and in particular to a flood prediction method, device, computer equipment and storage medium. Background Art

[0002] With the development of water conservancy technology, when designing flood control structures, it is necessary to consider the size of the possible flood as an important design basis, so it is necessary to predict the flood in the basin. However, at present, the spatial distribution of runoff observation sites around the world is sparse and the coverage period is short. There is a particular lack of observations on small and medium-sized rivers in China, making flood prediction more difficult.

[0003] In traditional technology, regional flood frequency analysis is used to predict flood frequencies in basins with no or insufficient flood data. This analysis method first divides different basins into multiple hydrological homogeneity regions based on the meteorological characteristics and underlying surface characteristics of the basin. For a certain hydrological homogeneity region, a flood prediction model is established with basin meteorology (such as precipitation) and underlying surface characteristics (such as basin average elevation) as input. Finally, the model is used to predict floods in basins with no or insufficient flood data within the hydrological homogeneity region.

[0004] However, traditional technologies only consider the similarities between different river basins in terms of meteorology and underlying surface, but meteorology and underlying surface cannot fully and strictly represent the actual situation of flood characteristics in the corresponding river basin. The prediction accuracy of the flood prediction model based on this is low. Summary of the invention

[0005] Based on this, it is necessary to provide a flood prediction method, device, computer equipment and storage medium that can improve the accuracy of flood prediction in response to the above technical problems.

[0006] A flood prediction method, the method comprising:

[0007] The multiple river basins included in the study area are geographically gridded to determine the seasonal indicators of each grid center; the seasonal indicators include the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center;

[0008] All grids are divided into multiple homogeneous areas according to the similarity between the seasonal indicators of all grid centers;

[0009] For each homogeneous region, a flood prediction model for the homogeneous region is trained based on the historical flood data of the first type of basin in the homogeneous region; the first type of basin is a basin among multiple basins that meets the historical data record quantity condition;

[0010] The meteorological characteristics and underlying surface characteristics of the second type of basin are input into the flood prediction model corresponding to the homogeneous area to which the second type of basin belongs to obtain the flood frequency of the second type of basin; the second type of basin is the basin among multiple basins that does not meet the historical data record quantity condition.

[0011] In one embodiment, determining the seasonal index of each grid center includes: calculating the seasonal index of the geometric center of the first type of watershed based on the historical flood records of the first type of watershed; and calculating the seasonal index of all grid centers based on the seasonal index of the geometric center of the first type of watershed.

[0012] In one embodiment, a seasonal index of a geometric center of a first type of river basin is calculated based on historical flood records of the first type of river basin, including: converting multiple actual flood occurrence times of the first type of river basin based on an radian numerical processing algorithm to obtain multiple radian values, and determining an average flood occurrence time of the geometric center of the first type of river basin based on the multiple radian values; determining a flood occurrence concentration of the geometric center of the first type of river basin based on the multiple radian values; the flood occurrence concentration is used to characterize the probability of a flood occurring in the first type of river basin at a single time point within an observation period.

[0013] In one embodiment, determining the average occurrence time of flood in the geometric center of a first type of watershed based on multiple radian values ​​includes: performing cosine operation on the multiple radian values ​​respectively to obtain multiple cosine values, and summing the multiple cosine values ​​to obtain a first operation value; performing sine operation on the multiple radian values ​​respectively to obtain multiple sine values, and summing the multiple sine values ​​to obtain a second operation value; determining a third operation value based on the first operation value and the second operation value, and performing inverse tangent operation on the third operation value to obtain the average occurrence time of flood in the geometric center of the first type of watershed.

[0014] In one embodiment, the flood occurrence concentration of the geometric center of the first type of watershed is determined based on multiple radian values, including: squaring the first operation value and the second operation value respectively, and calculating the flood occurrence concentration of the geometric center of the first type of watershed based on the operation results and the number of unit time periods in the observation period.

[0015] In one embodiment, the seasonal index of all grid centers is calculated based on the seasonal index of the geometric center of the first type of watershed, including: performing spatial interpolation operation on the seasonal index of the geometric center of the first watershed to obtain the seasonal index of all grids.

[0016] In one of the embodiments, all grids are divided into a plurality of homogeneous regions according to the similarities between the seasonal indicators of all grid centers, including: determining the similarities between the seasonal indicators of all grid centers to obtain a plurality of similarity parameters; determining the maximum similarity parameter among the similarities less than a preset threshold among the plurality of similarity parameters as a target similarity parameter; clustering the seasonal indicators based on the target similarity parameter and a cluster recognition algorithm, and dividing all grids into a plurality of homogeneous regions according to the clustering results.

[0017] In one embodiment, the similarity between the seasonal indicators of all grid centers is determined, including: calculating the similarity between the average flood occurrence time in the first seasonal indicator and the average flood occurrence time in the second seasonal indicator to obtain a first similarity value; calculating the similarity between the flood occurrence concentration in the first seasonal indicator and the flood occurrence concentration in the second seasonal indicator to obtain a second similarity value; calculating the similarity between the geographical location of the grid corresponding to the first seasonal indicator and the geographical location of the grid corresponding to the second seasonal indicator to obtain a third similarity value; determining the similarity between the first seasonal indicator and the second seasonal indicator based on the first similarity value, the second similarity value and the third similarity value; wherein the first seasonal indicator and the second seasonal indicator are any two of the seasonal indicators of all grid centers.

[0018] A flood prediction device, comprising:

[0019] A determination module is used to geographically grid the multiple basins included in the study area and determine the seasonal index of each grid center; the seasonal index includes the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center;

[0020] A partitioning module is used to divide all grids into multiple homogeneous areas according to the similarity between the seasonal indicators of all grid centers;

[0021] A training module is used for training a flood prediction model of a homogeneous region based on historical flood data of a first-category river basin in the homogeneous region for each homogeneous region; the first-category river basin is a river basin among multiple river basins that meets the historical data record quantity condition;

[0022] The prediction module is used to input the meteorological characteristics and underlying surface characteristics of the second type of basin into the flood prediction model corresponding to the homogeneous area to which the second type of basin belongs, so as to obtain the flood frequency of the second type of basin; the second type of basin is a basin among multiple basins that does not meet the historical data record quantity condition.

[0023] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0024] The multiple river basins included in the study area are geographically gridded to determine the seasonal indicators of each grid center; the seasonal indicators include the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center;

[0025] All grids are divided into multiple homogeneous areas according to the similarity between the seasonal indicators of all grid centers;

[0026] For each homogeneous region, a flood prediction model for the homogeneous region is trained based on the historical flood data of the first type of basin in the homogeneous region; the first type of basin is a basin among multiple basins that meets the historical data record quantity condition;

[0027] The meteorological characteristics and underlying surface characteristics of the second type of basin are input into the flood prediction model corresponding to the homogeneous area to which the second type of basin belongs to obtain the flood frequency of the second type of basin; the second type of basin is the basin among multiple basins that does not meet the historical data record quantity condition.

[0028] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the following steps:

[0029] The multiple river basins included in the study area are geographically gridded to determine the seasonal indicators of each grid center; the seasonal indicators include the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center;

[0030] All grids are divided into multiple homogeneous areas according to the similarity between the seasonal indicators of all grid centers;

[0031] For each homogeneous region, a flood prediction model for the homogeneous region is trained based on the historical flood data of the first type of basin in the homogeneous region; the first type of basin is a basin among multiple basins that meets the historical data record quantity condition;

[0032] The meteorological characteristics and underlying surface characteristics of the second type of basin are input into the flood prediction model corresponding to the homogeneous area to which the second type of basin belongs to obtain the flood frequency of the second type of basin; the second type of basin is the basin among multiple basins that does not meet the historical data record quantity condition.

[0033] The flood prediction method, device, computer equipment and storage medium described above, by geographically gridding the multiple basins contained in the study area, determine the seasonal index of each grid center, and then divide all grids into multiple homogeneous regions according to the similarity between the seasonal indexes of all grid centers, and then for each homogeneous region, based on the historical flood data of the first type of basin in the homogeneous region, train a flood prediction model for the homogeneous region, and finally, input the meteorological characteristics and underlying surface characteristics of the second type of basin into the flood prediction model corresponding to the homogeneous region to which the basin to be predicted belongs, and obtain the flood frequency of the second type of basin. The seasonal index in this method can characterize the mechanism of flood occurrence. Therefore, when dividing homogeneous regions based on seasonal indexes, homogeneous regions can be divided more accurately, and the model trained based on the flood data in the homogeneous region is more accurate, thereby improving the accuracy of the prediction results of the flood prediction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a structural diagram of a computer device in one embodiment;

[0035] Figure 2 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0036] Figure 3 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0037] Figure 4 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0038] Figure 5 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0039] Figure 6 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0040] Figure 7 A diagram of a complex network and cluster structure in one embodiment;

[0041] Figure 8 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0042] Fig. 9 A schematic diagram of a flood prediction method according to an embodiment of the present invention;

[0043] Fig.10 A distribution diagram of the geometric center of a watershed in one embodiment;

[0044] Fig.11 A contour map of flood seasonality indicators in one embodiment;

[0045] Fig.12 A schematic diagram of a homogeneous region division result in one embodiment;

[0046] Fig.13 A schematic diagram of flood frequency prediction accuracy evaluation in one embodiment;

[0047] Fig.14 is a structural block diagram of a flood prediction device in one embodiment;

[0048] Fig.15 FIG. 4 is a structural block diagram of a flood prediction device in one embodiment. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0050] The flood prediction method provided in this application can be applied to Figure 1 In the computer device shown. The computer device may be a server, and the processor in the computer device may execute a computer program to perform flood prediction. When the computer program is executed by the processor, flood data stored in the computer may be obtained, a corresponding flood prediction model may be trained for the flood data, and flood prediction may be performed using the flood model.

[0051] In one embodiment, Figure 2 As shown in the figure, a flood prediction method is provided, which is applied to Figure 1 The computer device in the example is used to illustrate, including the following steps:

[0052] S202. Grid the multiple river basins included in the research scope geographically and determine the seasonal indicators of each grid center; the seasonal indicators include the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center.

[0053] For example, this method can be applied to flood prediction of any basin within a large spatial scale (such as the whole country, the whole province, and the first-level basin). It is especially suitable for small and medium-sized basins that are less affected by human activities without significant dam regulation, large-scale irrigation, water diversion in the basin, land cover changes, etc. Among them, small and medium-sized basins can be basins with a catchment area of ​​less than 10,000 km. 2 For the convenience of expression, the research scope is used to replace the above large spatial scale regional scope.

[0054] Gridding can refer to dividing the research area geographically into several grids of the same size. The grid resolution can be set by technicians in this field according to the size of the watershed. For example, the grid resolution can be 10 km. 2 ~100km 2 . The location of each river basin within the research scope can be the location of its geographical center. It can be understood that the grid where the geographical center of each river basin is located can refer to the grid corresponding to the river basin. Seasonal indicators can be used to characterize the seasonal characteristics of floods. The seasonal indicator of a grid center in an embodiment of the present application can be the average occurrence time of floods in the grid center, or it can be the concentration of flood occurrence.

[0055] In a possible implementation, the average flood occurrence time can represent the average flood occurrence time of a geographical location within a unit time period. For example, it can be the average value of the time when the flood occurs in each unit time period of the location. Exemplarily, the observation period of basin x is 3 years, and the unit time period is 1 year. Among them, the flood occurs in March in the first year, in April in the second year, and in May in the third year. Then the average flood occurrence time of the geometric center of basin x is April. The above flood occurrence time can also be in units of days, ten days, or quarters, which are not limited here.

[0056] In one possible implementation, the flood concentration can be used to characterize the probability of a flood occurring at a geographical location at a single point in time. The higher the flood concentration of the location, the more concentrated the flood occurrence time of the location is at the same point in time. For example, based on the above example of basin x, the observation period of basin y is also 3 years, and floods occur in April every year. Therefore, the average flood occurrence time of basin y is the same as that of basin x. We can intuitively find from the data that the flood occurrence time of basin y is all concentrated in April. Compared with basin x, the concentration of floods in the geometric center of basin y is higher.

[0057] Specifically, the computer device can select small and medium-sized river basins with flood data that are less affected by human activities within the research scope, and confirm the maximum flow (instantaneous flow or daily average flow) in each unit period of multiple river basins within the scope as a flood event, and then extract the time when the floods occurred in multiple river basins. Furthermore, the computer device can determine the seasonal index of each grid center based on the time when the floods occurred in each of the above multiple river basins.

[0058] S204, dividing all grids into multiple homogeneous regions according to the similarity between the seasonal indicators of all grid centers.

[0059] Among them, similarity can be used to characterize the similarity between the flood characteristics of two grid centers. The higher the similarity, the more similar the flood occurrence patterns of the two grid centers are. A homogeneous region can be a collection of multiple grids divided based on the similarity between multiple grid centers. The similarity between grids in a homogeneous region is high and can be regarded as having the same flood characteristics.

[0060] Specifically, after determining the seasonal indicators of all grid centers, the computer device can determine the similarity between the seasonal indicators of all grid centers based on the above seasonal indicators. Further, the computer device can divide all grids into multiple homogeneous regions based on the above similarity. In another possible implementation, the computer device can also determine the above similarity based on the seasonal indicators of all grid centers and the spatial proximity between all grid centers. The method for determining the similarity is not limited here. When determining the similarity between grid centers, considering the spatial proximity between grid centers and the seasonal indicators at the same time can make the calculation result of the similarity more accurate, improve the accuracy of dividing homogeneous regions, and thus improve the accuracy of flood prediction.

[0061] S206. For each homogeneous region, a flood prediction model for the homogeneous region is trained based on the historical flood data of the first type of river basin in the homogeneous region; the first type of river basin is a river basin among multiple river basins that meets the historical data record quantity condition.

[0062] Among them, the first type of basin can be a basin with rich historical flood data and meets the historical data recording condition. The above historical data recording condition can be confirmed by a person skilled in the art according to the actual situation. For example, the above historical data recording condition can include historical flood data in the past 10 years. At this time, if the basin x includes historical flood data in the past 11 years, the basin x can be a first type of basin. The flood prediction model can be a random forest model, a linear regression model, etc., and the type of flood prediction model is not limited here. The historical flood data of the basin at least includes the annual average maximum flow of the basin, the daily average precipitation of the basin, the ratio of snowfall to total precipitation in the basin, the basin area, and the average slope of the basin.

[0063] For example, the computer device may establish a regression equation for the index flood of each homogeneous region. The index flood is a representative flood magnitude of a basin, and the annual average maximum flow μ (m 3 / s). For the basins in the same homogeneous area, computer equipment can establish the regression relationship between meteorological and underlying surface characteristics and index floods:

[0064]

[0065] Among them, Pmean is the average daily precipitation in the basin (mm), snow frac is the ratio of snowfall to total precipitation in the basin, and area is the area of ​​the basin (km 2 ), slope is the average slope of the basin. β0, β1, β2, β3 and β4 are the five parameters in the above regression relationship, which can be fitted by the historical flood data of the first type of basin in each homogeneous area.

[0066] Furthermore, the computer device can establish a flood frequency curve for each homogeneous area. Specifically, for the first type of basin x, the computer device can fit its dimensionless annual maximum flow through the generalized extreme value distribution The probability distribution curve of is:

[0067]

[0068] Where Q ix is the annual maximum flow of basin x in the i-th year. Assume that the curve parameters fitted by basin x are basin {u x ,σ x ,ξ x}, the basin has n x If there are annual maximum flow records, the parameters of the flood frequency curve in the homogeneous area are:

[0069]

[0070] S208. Input the meteorological characteristics and underlying surface characteristics of the second type of river basin into the flood prediction model corresponding to the homogeneous area to which the second type of river basin belongs, and obtain the flood frequency of the second type of river basin; the second type of river basin is a river basin among multiple river basins that does not meet the historical data record quantity condition.

[0071] Among them, the second type of basin may be a basin with a lack of historical flood data and does not meet the historical data recording condition. The above historical data recording condition can be confirmed by a person skilled in the art according to the actual situation. For example, the above historical data recording condition may include historical flood data in the last 10 years. At this time, if the basin x includes historical flood data in the last 9 years, the basin x may be a second type of basin. The meteorological characteristics of the basin may include the average daily precipitation in the basin and the ratio of snowfall to total precipitation in the basin. The underlying surface characteristics of the basin may include the basin area and the average slope of the basin. The flood frequency may be the flood magnitude corresponding to a certain recurrence period (such as once in a hundred years).

[0072] Specifically, for any data-free watershed within the research scope, the computer device can determine the homogeneous region to which the geometric center of the watershed belongs according to steps S202 and S204. Furthermore, the computer device can input the watershed daily average precipitation, the ratio of watershed snowfall to total precipitation, the watershed area, and the watershed average slope of the watershed according to the above expression (1) to obtain the index flood μ of the watershed. x If the flood flow rate of the homogeneous area with a return period of T years obtained by the frequency curve of the homogeneous area is defined as q (R) (T), then the T-year return flow of the basin is Q x (T) is:

[0073] Q x (T) = μ x q (R) (T) (4)

[0074] The above flood prediction method is to geographically grid the multiple basins contained in the study area, determine the seasonal index of each grid center, and then divide all grids into multiple homogeneous regions according to the similarity between the seasonal indexes of all grid centers. Then, for each homogeneous region, based on the historical flood data of the first type of basin in the homogeneous region, a flood prediction model for the homogeneous region is trained. Finally, the meteorological characteristics and underlying surface characteristics of the second type of basin are input into the flood prediction model corresponding to the homogeneous region to which the basin to be predicted belongs, and the flood frequency of the second type of basin is obtained. The seasonal index in this method can characterize the mechanism of flood occurrence. Therefore, when dividing homogeneous regions based on seasonal indexes, homogeneous regions can be divided more accurately, and the model trained based on flood data in homogeneous regions is more accurate, thereby improving the accuracy of the prediction results of the flood prediction model.

[0075] In one embodiment, Figure 3 As shown, based on the above embodiment, S202 includes:

[0076] S302. Calculate the seasonal index of the geometric center of the first type of watershed based on the historical flood records of the first type of watershed.

[0077] Specifically, for the first type of river basin that meets the historical data record quantity condition, its historical flood data is abundant. The computer equipment can directly calculate the seasonal index of the geometric center of the river basin based on the historical flood records of the first type of river basin. The richer the historical flood data, the more accurate the calculated seasonal index.

[0078] S304. Calculate the seasonal indexes of all grid centers according to the seasonal indexes of the geometric centers corresponding to the first type of watersheds.

[0079] Specifically, for all grid centers, the computer device can use interpolation or the like to calculate the seasonal indexes of the remaining grids according to the seasonal index of the geometric center of the first type of watershed. The specific interpolation method is not limited herein.

[0080] In this embodiment, the seasonal index of all grid centers is calculated based on the seasonal index of the geometric center of the first type of watershed, so that the seasonal characteristics of floods can cover all grid centers, thereby ensuring that any watershed without flood data or with insufficient flood data can also be divided into the homogeneous area to which it belongs, and ultimately realizing the flood frequency prediction of watersheds without flood data or with insufficient flood data.

[0081] In one embodiment, Figure 4 As shown, based on the above embodiment, S302 includes:

[0082] S402. Based on the radian numerical processing algorithm, multiple actual flood occurrence times of the first type of watershed are converted to obtain multiple radian values, and the average flood occurrence time of the geometric center of the first type of watershed is determined according to the multiple radian values.

[0083] Among them, the radian numerical processing algorithm is an algorithm that converts the occurrence time of floods into radian values.

[0084] Specifically, since the date of each year is cyclical, the radian system is usually used to represent the time of flood occurrence. For example, if the basin x is a first-class basin, the flood in the basin x in the i-th year occurs on the j-th day of the year. i days, and there are m i days, then the radian value θ of the flood occurrence time in the geometric center of the basin x in the i-th year i It can be expressed as:

[0085]

[0086] Furthermore, the computer device can convert multiple actual flood occurrence times of the geometric center of the first type of basin based on the radian numerical processing algorithm to obtain multiple radian values, and determine the average flood occurrence time of the geometric center of the first type of basin based on the multiple radian values. Exemplarily, for the above basin x, if the geometric center of the basin x includes i actual flood occurrence times, the computer device can obtain θ1, θ2, ..., θ based on the radian numerical processing algorithm. i There are i radian values ​​in total, and the average flood occurrence time of the geometric center of the first type of basin in i year is determined based on the above i radian values.

[0087] S404. Calculate the flood occurrence concentration of the geometric center of the first type of watershed according to multiple radian values; the flood occurrence concentration is used to characterize the probability of a flood occurring in the first type of watershed at a single time point within an observation period.

[0088] The single time point may be a quarterly time, a monthly time, or a date, which is not limited here.

[0089] Specifically, the computer device can calculate the flood concentration of the geometric center of the first type of basin based on multiple radian values, and the calculation method is not limited here. It should be pointed out that for those skilled in the art, without departing from the concept of this solution, several variations and improvements can be made to the calculation method. The flood concentration obtained by any calculation method, as long as it can characterize the probability of flood occurrence in the first type of basin at a single time point within an observation period, should be included in the protection scope of this solution.

[0090] In this embodiment, the computer device can convert multiple actual flood occurrence times of the geometric center of the first type of river basin based on the radian numerical processing algorithm to obtain multiple radian values, and determine the average flood occurrence time and flood occurrence concentration of the geometric center of the first type of river basin based on the multiple radian values, thereby providing a seasonal indicator data basis for the division of homogeneous areas. When dividing homogeneous areas based on this seasonal indicator, the accuracy of the homogeneous area division results can be improved, thereby improving the accuracy of the prediction results of the flood prediction model.

[0091] In one embodiment, Figure 5 As shown, based on the above embodiment, S402 includes:

[0092] S502 , respectively perform cosine operations on a plurality of radian values ​​to obtain a plurality of cosine values, and perform a sum operation on the plurality of cosine values ​​to obtain a first operation value.

[0093] Specifically, the computer device may take the cosine of the multiple radian values ​​in the first type of watershed and add them together to obtain the first operation value. For example, for the watershed x, if the first operation value is defined as C, then Among them, θ i is the radian value of the flood occurrence time in the i-th year of basin x, and n is the total number of floods in basin x (i.e. the number of years).

[0094] S504, respectively perform sine operation on the multiple radian values ​​to obtain multiple sine values, and perform a sum operation on the multiple sine values ​​to obtain a second operation value.

[0095] Specifically, the computer device may take the sine of multiple radian values ​​in the first type of watershed and add them together to obtain the first operation value. For example, for the watershed x, if the second operation value is defined as S, then Among them, θ i It can be the radian value of the flood occurrence time in the i-th year of basin x, and n can be the total number of floods in basin x.

[0096] S506: Determine a third operation value according to the first operation value and the second operation value, perform an arctangent operation on the third operation value, and obtain an average flood occurrence time of the geometric center of the first type of watershed.

[0097] Exemplarily, the third operation value may be defined as The average flood occurrence time of the geometric center of the first type of basin can be defined as The average flood occurrence time of the geometric center of the first type of basin is Can be:

[0098]

[0099] In this embodiment, the computer device can perform cosine operations on multiple radian values ​​respectively to obtain multiple cosine values, and perform sum operations on the multiple cosine values ​​to obtain a first operation value. The computer device can also perform sine operations on multiple radian values ​​respectively to obtain multiple sine values, and perform sum operations on the multiple sine values ​​to obtain a second operation value. Then, the computer device can determine a third operation value based on the first operation value and the second operation value, and perform an inverse tangent operation on the third operation value to obtain the average flood occurrence time of the geometric center of the first type of watershed. This embodiment determines the third operation value through the first operation value and the second operation value, and provides an algorithmic basis for the average flood occurrence time in the seasonal index, thereby providing a seasonal indicator data basis for the division of homogeneous areas. When dividing homogeneous areas based on this seasonal indicator, the accuracy of the homogeneous area division results can be improved, thereby improving the accuracy of the prediction results of the flood prediction model.

[0100] In one embodiment, based on the above embodiment, S404 includes:

[0101] The first operation value and the second operation value are squared respectively, and the flood occurrence concentration of the geometric center of the first type of watershed is calculated according to the operation results and the number of unit time periods in the observation period.

[0102] The number of unit time periods within the observation period can be the total number of floods within the basin observation period. For example, for basin x, if the observation period is 10 years and the unit time period is 1 year, the total number of floods in basin x within 10 years is 10 times.

[0103] For example, if the first operation value of the first type of basin is defined as C, the second operation value of the first type of basin is defined as S, the number of unit time periods within the observation period of the first type of basin is defined as n, and the flood occurrence concentration of the grid corresponding to the first type of basin is defined as R, then the expression of R is:

[0104]

[0105] Among them, the flood occurrence concentration R∈[0,1] of the geometric center of the first type of basin can represent the variability of the flood occurrence time of the first type of basin. For example, R=0 can represent that floods can occur equally on any day of the year in the first type of basin, and R=1 can represent that all floods in the first type of basin occur on the same day every year.

[0106] In this embodiment, the computer device can perform square operations on the first operation value and the second operation value respectively, and then calculate the flood occurrence concentration of the geometric center of the first type of basin according to the operation results and the number of unit time periods in the observation period, which provides an algorithm basis for the flood occurrence concentration in the seasonal index, thereby providing a seasonal index data basis for the division of homogeneous regions. When the homogeneous regions are divided based on this seasonal index, the accuracy of the homogeneous region division results can be improved, thereby improving the accuracy of the prediction results of the flood prediction model.

[0107] In one embodiment, based on the above embodiment, S304 includes:

[0108] The seasonal index of the geometric center of the first watershed is spatially interpolated to obtain the seasonal index of all grid centers.

[0109] Among them, the spatial interpolation operation can be used to convert the seasonal data of discrete points into continuous seasonal data. Exemplarily, for all grids in this application, the seasonal index of the geometric center of the first-class watershed can be directly calculated from the historical flood data of the first-class watershed, and the seasonal index of the grid center can be obtained by weighting the seasonal index of the geometric center of multiple first-class watersheds.

[0110] Exemplarily, the computer device may use the Kriging interpolation formula to estimate the seasonal index data of all grid centers. If the seasonal index data of the grid center to be estimated is defined as Z(x0), where x0 is the remaining grid position to be estimated, then:

[0111]

[0112] Among them, λ i It can be the weight parameter of the Kriging formula, which is used to characterize the weight of the geometric center position of the first type of watershed. iIt can be estimated by fitting the spatial variogram using seasonal index data of the geometric center of multiple first-class watersheds. i Represents the location of the geometric center of the first type of watershed close to the grid center, Z(x i ) represents the geometric center position x of the first type of watershed i seasonal index data at, N is the number of neighboring grids at the center of the grid x0. According to the universal kriging method, Z(x i ) can be expressed as a polynomial and contains linear terms of certain covariates. In this embodiment, Z(x i ) can be selected as the average elevation and average slope of the first type of watershed selected in the spatial interpolation calculation.

[0113] Specifically, the computer device can perform spatial interpolation operations on the seasonal index of the geometric center of the first watershed through the above-mentioned Kriging formula to obtain the seasonal index of all grid centers. Furthermore, the computer device can also perform spatial interpolation operations on the seasonal index of the geometric center of the first watershed through interpolation methods such as spline interpolation method and inverse distance weighted method to obtain the seasonal index of all grid centers. The interpolation calculation method is not limited here.

[0114] In this embodiment, the computer device can perform spatial interpolation operation on the seasonal index of the geometric center of the first watershed to obtain seasonal indexes of all grid centers. In the spatial interpolation operation, the seasonal index data Z(x i ) uses the average elevation and average slope of the watershed as covariates. That is to say, in the spatial interpolation operation, this scheme takes into account both the seasonal similarity of regional floods caused by large-scale climate and the difference in flood similarity caused by terrain differences, making the interpolation results more accurate and improving the accuracy of calculating seasonal indicators in all grid centers.

[0115] In one embodiment, Figure 6 As shown, based on the above embodiment, S204 includes:

[0116] S602: Determine similarity parameters between pairs of seasonal indicators of all grid centers to obtain multiple similarities.

[0117] Among them, the similarity can be used to characterize the similarity of flood characteristics between the seasonal indicators of each grid center.

[0118] Specifically, the computer device can determine the similarity between the seasonal indicators of all grid centers through the seasonal indicator data of each grid center, and obtain multiple similarity parameters. For example, for grid x, grid y, and grid z, the computer device can determine the similarity between the seasonal indicators of all grid centers based on the average flood occurrence time of the grid x, grid y, and grid z. The similarities D(x,y), D(x,z) and D(y,z) between the three grid centers are determined based on the flood occurrence concentration R. The method for determining the similarities is not limited here.

[0119] S604: Determine, among the similarities of the multiple similarity parameters that are smaller than a preset threshold, the largest similarity parameter as the target similarity parameter.

[0120] The target similarity parameter may be an indicator parameter that characterizes the similarity of the seasonal characteristics of floods between grid centers. The similarity may be positively correlated with the similarity of the seasonal characteristics of floods between grid centers, or negatively correlated with the similarity of the seasonal characteristics of floods between grid centers, which is not limited here. The preset threshold may be a value set by a person skilled in the art, or may be a value corresponding to a quantile of all similarity values ​​selected by a computer device.

[0121] For example, taking the negative correlation between the similarity and the similarity of the seasonal flood characteristics between the grid centers as an example, the computer device can select the similarity corresponding to the first 5% quantile after all similarity values ​​are arranged from small to large as the target similarity parameter for determining that the flood characteristics of the two grid centers are similar. Further, if the similarity between the two grid centers is less than the above target similarity, the computer device can connect the center positions of the two grids to characterize the similarity of the flood characteristics between the grids.

[0122] S606: Cluster each seasonal index based on the target similarity parameter and the cluster identification algorithm, and divide all grids into multiple homogeneous areas according to the clustering results.

[0123] Clustering may refer to the process of dividing a collection of all grids into a collection of multiple classes consisting of similar grids. The cluster identification algorithm may be a clustering algorithm used in clustering. In addition, a partitioning method, a density algorithm, etc. may also be used, which are not limited here.

[0124] Specifically, the computer device can cluster each seasonal indicator based on the target similarity parameter and the cluster identification algorithm, and divide all grids into multiple homogeneous regions according to the clustering results. Exemplarily, based on the above-mentioned embodiment in S604, the computer device can construct a grid complex network. The complex network is a network structure composed of a huge number of nodes and the connection relationship between nodes. It is a tool for describing the relationship between each grid. In this embodiment, the computer device can take each grid center as a node, and connect the grid centers to each other according to whether the similarity between the grid centers is less than the target similarity parameter, thereby constructing a grid complex network.

[0125] Furthermore, the computer device can divide the complex grid network into regions based on the cluster recognition algorithm. The cluster recognition algorithm in the grid complex network is a classification method for the network structure, and can automatically identify the optimal number of classifications, which is suitable for regional division. Among them, the cluster can be a collection of nodes in the above-mentioned grid network. The number of edges connecting the points in the same cluster is relatively large, while the number of edges connecting the points in different clusters is relatively small. The cluster recognition algorithm can use the value of maximizing the modularity M as an indicator to classify the nodes. The calculation formula of M can be:

[0126]

[0127] Among them l c is the number of edges inside cluster c, d c is the total number of edges owned by all nodes in cluster c, n c is the number of clusters, m is the total number of edges in the network, and modularity M is an indicator that can characterize the effect of clustering. The larger M is, the better the clustering effect is. Computer equipment can obtain the optimal number of classifications by continuously optimizing the value of modularity M. After all nodes are clustered and divided with the optimal number of classifications, the community composed of the grids corresponding to the nodes can be characterized as the homogeneous regions finally divided. An example of the region division of a grid complex network can be as follows: Figure 7 As shown, after the grid is divided into regions using the cluster recognition algorithm, each cluster region can represent a homogeneous region, and each grid can correspond to a homogeneous region to which it belongs.

[0128] In this embodiment, the computer device can determine the similarity between each seasonal index and obtain multiple similarity parameters, and then determine the maximum similarity parameter among the multiple similarity parameters that are less than the preset threshold as the target similarity parameter, and finally cluster each seasonal index based on the target similarity parameter and the cluster recognition algorithm, and divide the multiple watersheds into multiple homogeneous areas according to the clustering results. Since the cluster recognition algorithm does not need to pre-set the number of classifications, the influence of human factors is eliminated, so the clustering result is more accurate, and the division result of the homogeneous area is also more accurate.

[0129] In one embodiment, Figure 8 As shown, based on the above embodiment, S602 includes:

[0130] S802: Calculate the similarity between the average flood occurrence time in the first seasonal index and the average flood occurrence time in the second seasonal index to obtain a first similarity value.

[0131] The first seasonal index and the second seasonal index may be seasonal indexes of any two grid centers. The first similarity value may be used to represent the similarity between the two grid centers in terms of the average flood occurrence time.

[0132] Specifically, the computer device may obtain the average flood occurrence time in the seasonal index of each of the two grid centers, and calculate the first similarity value between the two grid centers. For example, for grid x and grid y, if the average flood occurrence time of the center of grid x is The average flood occurrence time at the center of grid y is Then the first similarity value between the center of grid x and the center of grid y is Can be:

[0133]

[0134] S804: Calculate the similarity between the flood occurrence concentration in the first seasonal index and the flood occurrence concentration in the second seasonal index to obtain a second similarity value.

[0135] The second similarity value may be used to characterize the similarity between the two grid centers in terms of flood occurrence concentration.

[0136] Specifically, the computer device may obtain the flood occurrence concentration in the seasonal index of each of the two grid centers, and calculate the second similarity value between the two grid centers. For example, for grid x and grid y, if the flood occurrence concentration of the center of grid x is R x , the flood concentration at the center of grid y is R y, then the second similarity value D between the center of grid x and the center of grid y R (x,y) can be:

[0137] D R (x,y)=R x -R y (11)

[0138] S806: Calculate the similarity between the geographical location of the grid corresponding to the first seasonal index and the geographical location of the grid corresponding to the second seasonal index to obtain a third similarity value.

[0139] The second similarity value may be used to represent the similarity between the two grids in terms of geographical locations.

[0140] Specifically, the computer device may obtain the geographical distance between the centers of the two grids under the assumption of the earth's spherical surface, and calculate the third similarity value between the centers of the two grids. For example, for grid x and grid y, if the geographical distance between the centers of the two grids under the assumption of the earth's spherical surface is defined as geodist(x, y), then the third similarity value D between the center of grid x and the center of grid y is loc (x,y) can be:

[0141]

[0142] S808: Determine the similarity between the first seasonal indicator and the second seasonal indicator according to the first similarity value, the second similarity value, and the third similarity value.

[0143] Among them, the third similarity value can be used to characterize the similarity between two grid centers after considering three factors: average flood occurrence time, flood occurrence concentration and geographical location.

[0144] Specifically, the computer device may determine the similarity between the first seasonal indicator and the second seasonal indicator according to the sum of the first similarity value, the second similarity value, and the third similarity value. Exemplarily, based on the above example, the similarity D(x, y) between the centers of the grid x and the grid y may be:

[0145]

[0146] In this embodiment, the computer device can respectively calculate three similarity values ​​between two grid centers in terms of average flood occurrence time, flood concentration and geographical location, and calculate the final similarity value after weighted summation. At the same time, the spatial proximity of the grid centers and the seasonal similarity of the floods are considered, thereby improving the accuracy of the similarity calculation and the accuracy of the division of homogeneous areas based on similarity.

[0147] In one embodiment, Fig. 9 As shown in FIG. 1 , based on the above embodiment, the historical flood data set of the entire United States is used as an example to deduce the flood frequency of basins with no data or insufficient flood data in the United States. The above data set includes 600 small and medium-sized basins (area 4km 2 ~8268km 2 ) from 1980 to 2014. The daily runoff data of the above 600 small and medium-sized river basins are less affected by human activities such as dams, water diversion, and land use changes, and can more accurately represent the natural runoff of the river basins. The above dataset also contains the average meteorological characteristic data of the river basin (such as rainfall and snowfall) and the underlying surface characteristic data (such as elevation and slope).

[0148] In this example, in order to evaluate the flood prediction method in this application, 100 basins can be randomly selected as the test set, that is, they are assumed to be basins with no data or basins with insufficient flood data, and the remaining 500 basins are used as the training set. The design floods of the 100 basins are predicted and compared with the design floods derived from the measured runoff data. The geometric center distribution diagram of the basins corresponding to the historical flood data in the above data set is shown in Figure 1. Fig.10 As shown, the hollow circle represents the geometric center of the corresponding basin of the training set, and the solid circle represents the geometric center of the basin corresponding to the test set. Now, using the flood prediction method in this application, the entire continental United States is geographically gridded, and the flood data of the basins corresponding to 500 hollow circles are used to predict the flood frequency of the basins corresponding to the remaining 100 solid circles.

[0149] Specifically, this embodiment may include the following steps:

[0150] S902. Obtain historical flood events in the training set basin within the research scope.

[0151] Specifically, for the river basins corresponding to the above 500 training sets, the computer device can extract the annual maximum flow value of each river basin and its corresponding date as its flood event.

[0152] S904. Calculate the flood seasonality index of the geometric center of the training set watershed.

[0153] Specifically, for each flood event in the training set basin, formula (5) is used to convert it into radians, and then formulas (6) and (7) are used to calculate the average flood occurrence time θ and occurrence concentration R of the geometric center of each training set basin, respectively.

[0154] S906. Use Kriging to spatially interpolate flood seasonality indicators for all grid centers.

[0155] Specifically, the computer device can use the NASA Shuttle Radar Topography Mission (SRTM) elevation data to download 50 km 2 High-resolution elevation and slope data, i.e. the grid resolution in this example is 50km 2 Then, the computer device can use the autoKrige function of the R language automap package to perform automatic parameter optimization Kriging interpolation, and input the training set watershed data represented by the geometric center: R, average elevation of the basin, average slope of the basin, output 50km 2 The following are the 4978 grids in the continental United States. and R. The interpolated flood seasonality index contour lines for the entire continental United States are as follows Fig.11 As shown, (a) is a contour map of the average flood occurrence time, with the contour lines in months. (b) is a contour map of flood occurrence concentration.

[0156] S908. Divide all grids into homogeneous regions using a complex network cluster identification algorithm.

[0157] Specifically, the computer device can calculate the similarity between the centers of the 4978 grids according to formulas (10), (11), (12) and (13), and use the 5% quantile distance of all similarities as the target similarity. The R language igraph package is used to construct a complex network with 4978 grids as nodes, and the node pairs with similarity lower than the target similarity are connected by an edge. The cluster_louvain function of the igraph package is used to perform cluster identification based on the modularity greedy optimization algorithm (louvainalgorithm), and a total of 6 homogeneous regions from R1 to R6 are divided. The homogeneous region division results are shown in FIG. Fig.12 shown.

[0158] S910. Use the index flood method to establish a flood prediction model for homogeneous areas.

[0159] Specifically, for each training set basin, the computer device can calculate its annual average maximum flow and read the basin daily average precipitation P of each basin from the data set. mean , the ratio of snowfall to total precipitation in the basin frac, basin area area and basin average slope slope. Then the computer device can distribute all the training set basins to the multiple homogeneous areas divided in step S1008 according to the positions of their geometric centers. In each homogeneous area, the flood data of the training set basin contained in the homogeneous area is used as a sample, and the R language lm function is used to fit the regression formula (1), so as to obtain the regression relationship between the meteorological characteristics and the underlying surface characteristics and the index flood μ.

[0160] Furthermore, for each training set basin, the computer equipment can fit the extreme value probability distribution curve (2) of the dimensionless annual maximum flow sample through the fevd function in the extRemes package of the R language, and calculate the frequency curve of the homogeneous region as a whole in the homogeneous region (3), thereby obtaining the dimensionless T-year return flow of the region as q (R) (T).

[0161] S912. Predict flood frequency for the test set watershed using a flood prediction model corresponding to the homogeneous region.

[0162] Specifically, for any one of the 100 test set basins, the computer device can obtain the grid where its geometric center is located, and input the average daily precipitation P of the test set basin through the index flood regression formula (1) of the homogeneous area to which the grid belongs. mean , the ratio of snowfall to total precipitation in the basin frac , basin area and basin average slope slope, calculate the index flood μ x .

[0163] Furthermore, combined with the flood frequency curve of the homogeneous region, the T-year return flow Q of the basin x can be calculated by formula (4): x (T). The above is a flood prediction method that regards the basin with no data or the basin with insufficient flood data as the test set basin.

[0164] In this embodiment, in order to compare and evaluate with the actual design flood value, the 100 test sets use the annual maximum flow extracted from the measured runoff to fit the extreme value distribution (2) to calculate the actual T-year return flow of each test set basin. Fig.13 As shown, in this example, T = 2, 10, 50 are selected respectively, and the determination coefficient R 2 The results show that the R 2 =0.88, R of a 10-year flood 2 =0.84, R of a 50-year flood 2 =0.73, it can be seen that the scheme used to predict flood frequency has a high accuracy.

[0165] It should be understood that although Figure 2-6 as well as Figure 8-9 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2-6 as well as Figure 8-9 At least part of the steps may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0166] In one embodiment, Fig.14 As shown, a flood prediction device is provided, the device comprising:

[0167] A determination module 10 is used to geographically grid the multiple basins included in the research scope and determine the seasonal index of each grid center; the seasonal index includes the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center;

[0168] A division module 20, for dividing all grids into a plurality of homogeneous regions according to the similarity between the seasonal indicators of all grid centers;

[0169] The training module 30 is used to train the flood prediction model of the homogeneous region based on the historical flood data of the first type of basin in the homogeneous region for each homogeneous region; the first type of basin is a basin among the multiple basins that meets the historical data record quantity condition;

[0170] The prediction module 40 is used to input the meteorological characteristics and underlying surface characteristics of the second type of basin into the flood prediction model corresponding to the homogeneous area to which the second type of basin belongs, so as to obtain the flood frequency of the second type of basin; the second type of basin is a basin among multiple basins that does not meet the historical data record quantity condition.

[0171] In one embodiment, based on the above embodiment, Fig.15 As shown, the determination module 10 includes:

[0172] A first calculation unit 110 is used to calculate a seasonal index of a geometric center of a first type of watershed according to historical flood records of the first type of watershed;

[0173] The second calculation unit 120 is used to calculate the seasonal indexes of all grid centers according to the seasonal index of the geometric center of the first type of watershed.

[0174] In one of the embodiments, based on the above embodiments, the first calculation unit 110 is specifically used to: convert multiple actual flood occurrence times of the first type of river basin based on a radian numerical processing algorithm to obtain multiple radian values, and determine the average flood occurrence time of the geometric center of the first type of river basin based on the multiple radian values; determine the flood occurrence concentration of the grid corresponding to the first type of river basin based on the multiple radian values; the flood occurrence concentration is used to characterize the probability of a flood occurring in the first type of river basin at a single time point within an observation period.

[0175] In one embodiment, based on the above embodiment, the first calculation unit 110 is specifically used to: perform cosine operation on multiple radian values ​​respectively to obtain multiple cosine values, and sum the multiple cosine values ​​to obtain a first operation value; perform sine operation on multiple radian values ​​respectively to obtain multiple sine values, and sum the multiple sine values ​​to obtain a second operation value; determine a third operation value based on the first operation value and the second operation value, perform an inverse tangent operation on the third operation value, and obtain the average flood occurrence time of the geometric center of the first type of watershed.

[0176] In one of the embodiments, based on the above embodiment, the first calculation unit 110 is specifically used to: perform square operations on the first operation value and the second operation value, respectively, and calculate the flood occurrence concentration of the geometric center of the first type of river basin based on the operation results and the number of unit time periods within the observation period.

[0177] In one of the embodiments, based on the above embodiment, the second calculation unit 120 is specifically used to: perform spatial interpolation operation on the seasonal index of the geometric center of the first watershed to obtain the seasonal index of all grid centers.

[0178] In one of the embodiments, based on the above embodiment, the second calculation unit 120 is specifically used to: determine the similarity between the seasonal indicators of all grid centers, and obtain multiple similarity parameters; determine that among the multiple similarity parameters that are less than a preset threshold, the largest similarity parameter is the target similarity parameter; cluster each seasonal indicator based on the target similarity parameter and the cluster recognition algorithm, and divide all grids into multiple homogeneous areas according to the clustering results.

[0179] In one of the embodiments, based on the above embodiments, the second calculation unit 120 is specifically used to: calculate the similarity between the average occurrence time of floods in the first seasonal indicator and the average occurrence time of floods in the second seasonal indicator to obtain a first similarity value; calculate the similarity between the flood occurrence concentration in the first seasonal indicator and the flood occurrence concentration in the second seasonal indicator to obtain a second similarity value; calculate the similarity between the geographical location of the grid corresponding to the first seasonal indicator and the geographical location of the grid corresponding to the second seasonal indicator to obtain a third similarity value; determine the similarity between the first seasonal indicator and the second seasonal indicator based on the first similarity value, the second similarity value and the third similarity value; wherein the first seasonal indicator and the second seasonal indicator are any two of the seasonal indicators of all grid centers.

[0180] Specifically, the device for flood prediction provided above can execute the flood prediction method embodiment described above, and its implementation principle and technical effect are similar, which will not be described in detail here.

[0181] For the specific definition of the flood prediction device, please refer to the definition of the flood prediction method above, which will not be repeated here. Each module in the above-mentioned flood prediction device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0182] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 1 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used for hydrological data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a flood prediction method is implemented.

[0183] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0184] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor can implement the steps in the above-mentioned method embodiments when executing the computer program.

[0185] The implementation principle and technical effects of the computer device provided in this embodiment are similar to those of the above method embodiments, and will not be repeated here.

[0186] A computer-readable storage medium stores a computer program, which can implement the steps in the above-mentioned method embodiments when executed by a processor.

[0187] The computer-readable storage medium provided in this embodiment has similar implementation principles and technical effects to those of the above method embodiments, and will not be described in detail here.

[0188] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0189] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0190] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A flood prediction method, characterized in that: include: Gridding multiple river basins within the research scope according to their sizes, and determining seasonal indicators for each grid center; the seasonal indicators include the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center; Dividing all grids into a plurality of homogeneous regions according to the spatial proximity between the grid centers and the similarity between the seasonal indicators of all the grid centers; For each of the homogeneous regions, a regression relationship between meteorological characteristics and underlying surface characteristics and index floods is established based on historical flood data of a first type of river basin in the homogeneous region, and a flood frequency curve of the homogeneous region is established according to the regression relationship; the first type of river basin is a river basin among the multiple river basins that exceeds a preset amount of historical data records; The historical flood data include the annual average maximum flow of the basin, the average daily precipitation of the basin, the ratio of snowfall to total precipitation in the basin, the area of ​​the basin, and the average slope of the basin; The meteorological characteristics and underlying surface characteristics of the second type of river basin are input into the flood frequency curve corresponding to the homogeneous area to which the second type of river basin belongs to obtain the flood frequency of the second type of river basin; the second type of river basin is the river basin among the multiple river basins that does not exceed the preset historical data record amount.

2. The method according to claim 1, characterized in that The seasonal indicators for determining each grid center include: Calculating a seasonal index of the geometric center of the first type of watershed based on historical flood records of the first type of watershed; The seasonal indexes of all the grid centers are calculated based on the seasonal index of the geometric center of the first type of watershed.

3. The method according to claim 2, characterized in that The calculating of the seasonal index of the geometric center of the first type of watershed according to the historical flood records of the first type of watershed comprises: Based on the radian numerical processing algorithm, multiple actual flood occurrence times of the first type of watershed are converted to obtain multiple radian values, and the average flood occurrence time of the geometric center of the first type of watershed is determined according to the multiple radian values; The flood occurrence concentration of the geometric center of the first type of river basin is determined according to the multiple radian values; the flood occurrence concentration is used to characterize the probability of flooding in the first type of river basin at a single time point within an observation period.

4. The method according to claim 3, characterized in that The step of determining the average flood occurrence time of the geometric center of the first type of river basin according to the multiple radian values ​​includes: Performing cosine operations on the multiple radian values ​​respectively to obtain multiple cosine values, and performing a sum operation on the multiple cosine values ​​to obtain a first operation value; Performing sine operation on the multiple radian values ​​respectively to obtain multiple sine values, and performing sum operation on the multiple sine values ​​to obtain a second operation value; A third operation value is determined according to the first operation value and the second operation value, and an arc tangent operation is performed on the third operation value to obtain an average flood occurrence time of the geometric center of the first type of watershed.

5. The method according to claim 4, characterized in that The step of calculating the flood occurrence concentration of the geometric center of the first type of river basin according to the multiple radian values ​​includes: The first operation value and the second operation value are squared respectively, and the flood occurrence concentration of the geometric center of the first type of river basin is calculated according to the operation results and the number of unit time periods in the observation period.

6. The method according to claim 2, characterized in that The calculating the seasonal index of all the grid centers according to the seasonal index of the geometric center of the first type of watershed comprises: A spatial interpolation operation is performed on the seasonal index of the geometric center of the first type of watershed to obtain the seasonal index of all the grid centers.

7. The method according to claim 1, characterized in that According to the similarity between the seasonal indicators of all the grid centers, all the grids are divided into a plurality of homogeneous regions, including: Determine the similarity between the seasonal indicators of all the grid centers, and obtain multiple similarity parameters; Determine the largest similarity parameter among the plurality of similarity parameters whose similarities are less than a preset threshold as a target similarity parameter; The seasonal indicators are clustered based on the target similarity parameter and the cluster identification algorithm, and all the grids are divided into a plurality of homogeneous regions according to the clustering result.

8. The method according to claim 7, characterized in that Determining the similarity between the seasonal indicators of all the grid centers comprises: Calculating the similarity between the average flood occurrence time in the first seasonal index and the average flood occurrence time in the second seasonal index to obtain a first similarity value; Calculating the similarity between the flood occurrence concentration in the first seasonal index and the flood occurrence concentration in the second seasonal index to obtain a second similarity value; Calculating the similarity between the geographical location of the grid corresponding to the first seasonal index and the geographical location of the grid corresponding to the second seasonal index to obtain a third similarity value; Determining the similarity between the first seasonal indicator and the second seasonal indicator according to the first similarity value, the second similarity value, and the third similarity value; The first seasonal index and the second seasonal index are any two of the seasonal indexes of all the grid centers.

9. A flood prediction device, characterized in that: The device comprises: A determination module is used to perform grid processing on the multiple basins included in the research scope according to the basin size geographically, and determine the seasonal index of each grid center; the seasonal index includes the average flood occurrence time of the grid center and the flood occurrence concentration of the grid center; A division module, used for dividing all grids into a plurality of homogeneous regions according to the spatial proximity between the grid centers and the similarity between the seasonal indicators of all the grid centers; A training module is used to establish, for each of the homogeneous regions, a regression relationship between meteorological characteristics and underlying surface characteristics and index floods based on historical flood data of a first type of river basin in the homogeneous region, and to establish a flood frequency curve of the homogeneous region according to the regression relationship; the first type of river basin is a river basin among the multiple river basins that exceeds a preset amount of historical data records; the historical flood data includes the annual average maximum flow of the river basin, the daily average precipitation of the river basin, the ratio of snowfall to total precipitation in the river basin, the area of ​​the river basin, and the average slope of the river basin; A prediction module is used to input the meteorological characteristics and underlying surface characteristics of the second type of river basin into the flood frequency curve corresponding to the homogeneous area to which the second type of river basin belongs, so as to obtain the flood frequency of the second type of river basin; the second type of river basin is the river basin among the multiple river basins that does not exceed the preset historical data record amount.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images