Flood early warning and forecasting method and system based on intelligent water conservancy

Abstract

The flood early warning and forecasting method based on the intelligent water conservancy comprises the following steps: determining a target area and a corresponding target water system; the target water system comprises a main stream and a branch stream; dividing the target area into a plurality of sub-areas according to a tributary basin; a plurality of rainfall sensors are arranged in each subarea, and rainfall data of the subareas are obtained; arranging a plurality of flow test points at the upstream and downstream of the branch to obtain branch flow change data; establishing a regression model of sub-region precipitation data and branch flow change data; according to the real-time precipitation amount of the sub-region, the branch flow forecast flow is evaluated through a regression model; and evaluating the main stream forecast flow according to the branch flow forecast flow. The invention relates to a flood early warning and forecasting method and system based on intelligent water conservancy, which are used for flood early warning and forecasting of small watershed surfaces and mountainous areas and establishing a relationship between rainfall of each collection plot and branch flow change data. And through multiple times of real-time data clustering correction, early warning and forecasting of the branch flow and the main flow can be completed more accurately.

Description

Technical Field

[0001] The present invention relates to the field of smart water conservancy and data processing technology, and in particular to a flood early warning and forecasting method and system based on smart water conservancy. Background Art

[0002] my country's rainstorm forecasting has made a lot of useful explorations and practices in terms of refinement. In the future, rainstorm forecasting will continue to develop in the direction of timing, fixed point and quantity. At present, judgments can be made one week in advance for heavy rainfall processes and weather trends, but if a higher forecast accuracy is required, it will take three days or less. Judging from the current rainstorm forecasting situation in my country, the forecast has a relatively good grasp of the rainstorm process, the scope and time of impact, but it is still difficult to forecast the area, magnitude and accurate time period of the heavy precipitation center. At present, the calculation of the average rainfall in a small watershed usually uses the real-time rainfall data collected by the rain collection station. Through various weighted calculations, the average rainfall in the small watershed is obtained. This method requires the division of specific watersheds according to the spatial location of the rain collection station.

[0003] 202211587354.1 A method and system for calculating the average rainfall in a small watershed based on raster analysis The present invention relates to a method and system for calculating the average rainfall in a small watershed based on raster analysis, which belongs to the field of hydrological technology and solves the problem of accurately calculating the average rainfall data in a small watershed while taking into account the timeliness and predictability of the calculation. The present invention performs rasterization and subdivision to make the distribution of the average rainfall data in the small watershed more reasonable. 201910160558.9 A real-time cumulative rainfall rainstorm and flash flood warning method based on rainfall runoff response The present invention provides a real-time cumulative rainfall rainstorm and flash flood warning method based on rainfall runoff response. Based on the intrinsic relationship between rainstorm and flash flood and rainfall process, the method fully considers the impact of the cumulative rainfall of the rainstorm and flash flood process on the steep rise of floods in small watersheds in mountainous areas, and uses the cumulative rainfall index of the rainstorm corresponding to the rainstorm and flash flood as the warning index of the rainstorm and flash flood disaster in small watersheds in mountainous areas to carry out rainstorm and flash flood warning. Summary of the invention

[0004] The purpose of the present invention is to provide a flood warning and forecasting method and system based on smart water conservancy, which is particularly suitable for flood warning and forecasting in small watersheds and mountainous areas.

[0005] The flood early warning and forecasting method based on smart water conservancy includes the following steps: S1, determining a target area and a corresponding target water system; the target water system includes a main stream and tributaries; dividing the target area into a plurality of sub-areas according to the tributary basins; S2, setting a plurality of rain sensors in each of the sub-areas to obtain precipitation data of the sub-areas; setting a plurality of flow test points upstream and downstream of the tributary to obtain flow change data of the tributary; S3, establishing a regression model of the sub-region precipitation data and the tributary flow change data; S4, evaluating the tributary forecast flow through the regression model according to the real-time precipitation in the sub-region; S5, evaluating the forecast flow of the main stream according to the forecast flow of the tributary.

[0006] As an optimization, each of the sub-areas is divided into a plurality of collection plots according to the terrain; and the collection plots are respectively provided with the rainfall sensors.

[0007] As an optimization, the regression model is a multiple regression model of the sub-region precipitation data and the tributary flow change data: ; The flow change data of the tributary; The area of ​​n plots of land collected in the sub-region; Collecting precipitation for n plots in the sub-region; is the regression coefficient of the n sampling plots in the sub-area.

[0008] As an optimization, the 24-hour forecast flow of the tributary is evaluated by the regression model based on the 24-hour precipitation forecast data of the sub-area.

[0009] As an optimization, before the real-time flow change data of the tributary reaches a preset value, the forecast flow of the tributary is evaluated by the regression model.

[0010] The real-time flow change data of the tributary reaches the preset value, which can be that the flow change data of the tributary at any flow test point reaches the preset value, or the flow change data of the tributary at the flow test point closest to the main stream reaches the preset value.

[0011] As an optimization, a second regression model was also included; ; is the preset value; The precipitation of n collection plots in the sub-area when the preset value is reached; is the second regression coefficient of the n sampling plots in the sub-area.

[0012] As an optimization, after the real-time flow change data of the tributary reaches the preset value, the forecast flow of the tributary is evaluated by the second regression model.

[0013] As an optimization, after the real-time flow change data of the tributary reaches the preset value, The real-time precipitation in the sub-area and the historical precipitation in the sub-area are calculated by clustering algorithm to obtain the historical precipitation with the greatest similarity; The historical tributary flow change data corresponding to the historical precipitation with the greatest similarity is used to evaluate the tributary forecast flow.

[0014] As an optimization, after the real-time flow change data of the tributary reaches the preset value, a plurality of historical precipitations with close similarity are obtained by performing clustering algorithm calculation on the real-time precipitation of each collection plot in the sub-region and the historical precipitation of each collection plot in the sub-region; The real-time change data of tributary flow and multiple historical tributary flow change data corresponding to multiple historical precipitations with similarities are calculated to obtain the closest historical tributary flow change data, which is used to evaluate the forecast tributary flow.

[0015] As an optimization, the preset value is set to 15% to 20% of the average flow of the tributaries 24 hours before rainfall.

[0016] A flood early warning and forecasting system based on smart water conservancy, comprising: Data acquisition module: used to obtain the precipitation data of the sub-region and the flow change data of the tributary in real time; Data storage module: used to store sub-region precipitation data and tributary flow change data; Data calculation module: used for calculating the parameters of the regression model and the second regression model; and evaluating the tributary forecast flow through the regression model, the second regression model and the clustering algorithm; Result display module: used for real-time display of the precipitation data of the sub-area, the flow change data of the tributary, the forecast flow of the tributary, and the forecast flow of the main stream.

[0017] The flood warning and forecasting method and system based on smart water conservancy of the present invention are used for flood warning and forecasting in small watersheds and mountainous areas. The tributary basin is divided into multiple sub-areas, and then each sub-area is divided into multiple collection plots according to the terrain; one or more rainfall sensors are set in each collection plot. Since the impact of rainfall in different collection plots in the tributary basin on the flood mainstream flow is quite different, the relationship between the rainfall in each collection plot and the tributary flow change data is established; and it is convenient to perform multiple real-time data clustering corrections, so as to more accurately complete the warning and forecast of tributary and mainstream flows. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments; Figure 1 This is a schematic diagram related to Example 5 and Example 6 of the present invention. DETAILED DESCRIPTION

[0019] The following embodiments are intended to further illustrate the present invention, but should not be construed as limiting the scope of protection of the present invention. Non-essential improvements and adjustments to the present invention made by those skilled in the art based on the contents of the present invention still fall within the scope of protection of the present invention.

[0020] Embodiment 1: A flood early warning and forecasting method based on smart water conservancy includes the following steps: S1, determine the target area and the corresponding target water system; the target water system includes the main stream and tributaries; divide the target area into multiple sub-areas according to the tributary basins.

[0021] S2, multiple rain sensors are set in each of the sub-areas to obtain precipitation data of the sub-areas; the sub-areas are divided into multiple collection plots according to the terrain; the collection plots are respectively provided with rain sensors, and the point precipitation data can be converted into basin precipitation data by arithmetic mean method or Thiessen polygon method; Multiple flow test points are set up upstream and downstream of the tributaries to obtain the tributary flow change data; the calculation formula for the river flow at a certain section at a specific time can be derived based on the flow velocity and cross-sectional area.

[0022] S3, establishing a regression model for the precipitation data of each sampling plot in the sub-region and the tributary flow change data; The regression model is a multiple regression model of the sub-region precipitation data and the tributary flow change data: ; The flow change data of the tributary; The area of ​​n plots of land collected in the sub-region; Collecting precipitation for n plots in the sub-region; is the regression coefficient of the n sampling plots in the sub-area.

[0023] S4, evaluating the 24-hour forecast flow of the tributary by the regression model according to the 24-hour precipitation forecast data of the sub-region; After rainfall, before the real-time flow change data of the tributary reaches a preset value, the forecast flow of the tributary is evaluated by a regression model; The preset value is set to 15% to 20% of the average flow of the tributary 24 hours before rainfall. The real-time flow change data of the tributary reaches the preset value, which can be the tributary flow change data of any of the flow test points reaching the preset value, or the tributary flow change data of the flow test point closest to the main stream reaching the preset value; After the real-time flow change data of the tributary reaches the preset value, the forecast flow of the tributary is evaluated by the second regression model; Second regression model; ; is the preset value; The precipitation of n collection plots in the sub-area when the preset value is reached; is the second regression coefficient of the n sampling plots in the sub-area.

[0024] The flow change data of the tributary; The areas of n plots are collected for the sub-region.

[0025] S5, evaluate the forecast flow of the main stream based on the forecast flow of the tributaries. The total flow of the main stream in the near future can be calculated through the forecast data of each tributary flowing into the main stream, thereby evaluating the forecast flow of the main stream and determining whether to issue an early warning based on the situation.

[0026] Embodiment 2: A flood early warning and forecasting method based on smart water conservancy includes the following steps: S1, determine the target area and the corresponding target water system; the target water system includes the main stream and tributaries; divide the target area into multiple sub-areas according to the tributary basins.

[0027] S2, multiple rain sensors are set in each of the sub-areas to obtain precipitation data of the sub-areas; the sub-areas are divided into multiple collection plots according to the terrain; the collection plots are respectively provided with rain sensors to obtain precipitation data; multiple flow test points are set upstream and downstream of the tributaries to obtain tributary flow change data.

[0028] S3, using the collected historical data, a regression model of precipitation data of each collection plot in the sub-region and tributary flow change data is established; The regression model is a multiple regression model of the sub-region precipitation data and the tributary flow change data: ; The flow change data of the tributary; The area of ​​n plots of land collected in the sub-region; Collecting precipitation for n plots in the sub-region; is the regression coefficient of the n sampling plots in the sub-area.

[0029] S4, 24-hour forecast flow of tributaries: based on the 24-hour precipitation forecast data of the sub-region, the 24-hour forecast flow of tributaries is evaluated by regression model; Forecast flow of tributaries: After rainfall, before the real-time flow change data of tributaries reaches the preset value, the forecast flow of tributaries is evaluated through the regression model. The forecast flow of tributaries is predicted in a short period of time; Forecast flow of tributaries: after rainfall, when the real-time flow change data of tributaries reaches a preset value, the real-time precipitation of each collection plot in the sub-region is combined with the historical precipitation of each collection plot in the sub-region to obtain multiple historical precipitations with similarities; The real-time change data of tributary flow and a plurality of historical tributary flow change data corresponding to a plurality of historical precipitations with similarity are calculated to obtain the closest historical tributary flow change data, which is used to evaluate the forecast flow of the tributary; The closest historical tributary flow change data, that is, the real-time change data of tributary flow corresponding to the same precipitation situation divided by the historical tributary flow change data, is the item closest to 1.

[0030] S5, evaluate the forecast flow of the main stream based on the forecast flow of the tributaries. The total flow of the main stream in the near future can be calculated through the forecast data of each tributary flowing into the main stream, thereby evaluating the forecast flow of the main stream and determining whether to issue an early warning based on the situation.

[0031] Embodiment 3: A flood early warning and forecasting method based on smart water conservancy includes the following steps: S1, determine the target area and the corresponding target water system; the target water system includes the main stream and tributaries; divide the target area into multiple sub-areas according to the tributary basins.

[0032] S2, multiple rain sensors are set in each of the sub-areas to obtain precipitation data of the sub-areas; the sub-areas are divided into multiple collection plots according to the terrain; the collection plots are respectively provided with rain sensors to obtain precipitation data; multiple flow test points are set upstream and downstream of the tributaries to obtain tributary flow change data.

[0033] S3, using the collected historical data, a regression model of precipitation data of each collection plot in the sub-region and tributary flow change data is established; The regression model is a multiple regression model of the sub-region precipitation data and the tributary flow change data: ; The flow change data of the tributary; The area of ​​n plots of land collected in the sub-region; Collecting precipitation for n plots in the sub-region; is the regression coefficient of the n sampling plots in the sub-area.

[0034] S4, 24-hour forecast flow of tributaries: based on the 24-hour precipitation forecast data of the sub-region, the 24-hour forecast flow of tributaries is evaluated by regression model; Forecast flow of tributaries: After rainfall, before the real-time flow change data of tributaries reaches the preset value, the forecast flow of tributaries is evaluated through the regression model. The forecast flow of tributaries is predicted in a short period of time; Forecast flow of tributaries: after rainfall, when the real-time flow change data of tributaries reaches a preset value, the real-time precipitation of each collection plot in the sub-region is combined with the historical precipitation of each collection plot in the sub-region to obtain the historical precipitation with the greatest similarity through clustering algorithm calculation; The historical tributary flow change data corresponding to the historical precipitation with the greatest similarity is used to evaluate the tributary forecast flow.

[0035] S5, evaluate the forecast flow of the main stream based on the forecast flow of the tributaries. The total flow of the main stream in the near future can be calculated through the forecast data of each tributary flowing into the main stream, thereby evaluating the forecast flow of the main stream and determining whether to issue an early warning based on the situation.

[0036] Embodiment 4: A flood early warning and forecasting system based on smart water conservancy, comprising: Data acquisition module: used to obtain sub-region precipitation data and tributary flow change data in real time; Data storage module: used to store sub-region precipitation data and tributary flow change data; Data calculation module: used to calculate the parameters of the regression model and the second regression model; and to evaluate the tributary forecast flow through the regression model, the second regression model, and the clustering algorithm; Result display module: used for real-time display of the precipitation data of the sub-area, tributary flow change data, tributary forecast flow, and main stream forecast flow.

[0037] Embodiment 5: A flood early warning and forecasting method based on smart water conservancy includes the following steps: S1, determine the target area and the corresponding target water system; Figure 1 As shown, the target water system includes the main stream G1 and five tributaries Z1, Z2, Z3, Z4, and Z5; the target area is divided into five sub-areas D1, D2, D3, D4, and D5 according to the five tributary basins.

[0038] S2, each sub-area is divided into multiple collection plots according to the terrain; each collection plot is provided with a rain sensor, and the point precipitation data can be converted into sub-area precipitation data by arithmetic mean method or Thiessen polygon method; Multiple flow test points are set up upstream and downstream of the tributaries to obtain the tributary flow change data; the calculation formula for the river flow at a certain section at a specific time can be derived based on the flow velocity and cross-sectional area.

[0039] S3, taking sub-area D1 as an example, it includes 6 collection plots. Then, based on the precipitation data of each flood in sub-area D1 and the flow change data of tributaries, a multivariate regression model of precipitation data of sub-area D1 and flow change data of corresponding tributary Z1 is established: ; This is the flow change data of tributary Z1; is the area of ​​the six collection plots in sub-area D1; The precipitation of the six sampling plots in sub-area D1; is the regression coefficient of the six sampling plots in sub-area D1.

[0040] S4, based on the 24-hour precipitation forecast data of the sub-region, the 24-hour flow of the tributary can be calculated through the regression model to evaluate the 24-hour forecast flow of the tributary; After rainfall occurs, the change in tributary flow within half an hour to two hours can be calculated based on the real-time precipitation in the sub-area as the tributary forecast flow.

[0041] After rainfall, before the real-time flow change data of the tributary reaches a preset value, the forecast flow of the tributary is evaluated by a regression model; The preset value is set to 15% to 20% of the average flow of the tributary 24 hours before rainfall. The real-time flow change data of the tributary reaches the preset value, which can be the tributary flow change data of any of the flow test points reaching the preset value, or the tributary flow change data of the flow test point closest to the main stream reaching the preset value; After the real-time flow change data of the tributary reaches the preset value, taking sub-area D1 as an example, according to the real-time precipitation of the six acquisition plots, the flow change data of tributary Z1 can be calculated by the second regression model every five or ten minutes to evaluate the tributary forecast flow; Taking sub-region D1 as an example, the second regression model; ; Preset value for tributary Z1; The precipitation of the six collection plots in sub-area D1 when the preset value is reached; is the second regression coefficient of the six sampling plots in sub-area D1; The precipitation of the six sampling plots in sub-area D1; This is the flow change data of tributary Z1; It is the area of ​​6 collection plots in sub-area D1.

[0042] S5, evaluate the forecast flow of the main stream based on the forecast flow of the tributaries. The total flow of the five tributaries Z1, Z2, Z3, Z4, and Z5 flowing into the main stream can be calculated through the forecast data, thereby evaluating the forecast flow of the main stream and determining whether to issue an early warning based on the situation.

[0043] Embodiment 6: A flood early warning and forecasting method based on smart water conservancy includes the following steps: S1, determine the target area and the corresponding target water system; Figure 1 As shown, the target water system includes the main stream G1 and five tributaries Z1, Z2, Z3, Z4, and Z5; the target area is divided into five sub-areas D1, D2, D3, D4, and D5 according to the five tributary basins.

[0044] S2, each sub-area is divided into multiple collection plots according to the terrain; each collection plot is equipped with a rainfall sensor, which can convert the point precipitation data into sub-area precipitation data through the arithmetic mean method or Thiessen polygon method; multiple flow test points are set up upstream and downstream of the tributary to obtain the flow change data of the five tributaries Z1, Z2, Z3, Z4, and Z5.

[0045] S3, establish a regression model of precipitation data of each collection plot in the sub-region and tributary flow change data; take sub-region D1 as an example, it includes 6 collection plots, then according to the precipitation data of each flood in sub-region D1 and the tributary flow change data, establish a multivariate regression model of precipitation data of sub-region D1 and flow change data of corresponding tributary Z1: ; This is the flow change data of tributary Z1; is the area of ​​the six collection plots in sub-area D1; The precipitation of the six sampling plots in sub-area D1; is the regression coefficient of the six sampling plots in sub-area D1.

[0046] S4, based on the 24-hour precipitation forecast data of the sub-region, the 24-hour flow of the tributary can be calculated through the regression model to evaluate the 24-hour forecast flow of the tributary; After rainfall occurs, the change in tributary flow within half an hour to two hours can be calculated based on the real-time precipitation in the sub-area as the tributary forecast flow.

[0047] Forecast flow of tributaries: After rainfall, before the real-time flow change data of tributaries reaches the preset value, the forecast flow of tributaries is evaluated through the regression model. The forecast flow of tributaries is predicted in a short period of time; Forecast flow of tributaries: After rainfall, when the real-time flow change data of tributaries reaches the preset value, the real-time precipitation of each collection plot in the sub-region is combined with the historical precipitation of each collection plot in the sub-region to obtain multiple historical precipitations with similarity; the real-time change data S of tributary flow is combined with multiple historical tributary flow change data corresponding to multiple historical precipitations with similarity. The closest historical tributary flow change data is calculated to evaluate the tributary forecast flow. The above steps can be calculated periodically, for example, every ten minutes, half an hour or one hour, and through multiple real-time data clustering corrections, the early warning and forecast of tributary and main stream flows can be completed more accurately.

[0048] S5, evaluate the forecast flow of the main stream based on the forecast flow of the tributaries. The total flow of the five tributaries Z1, Z2, Z3, Z4, and Z5 flowing into the main stream can be calculated through the forecast data, thereby evaluating the forecast flow of the main stream and determining whether to issue an early warning based on the situation.

Claims

1. A flood early warning and forecasting method based on smart water conservancy, characterized in that: The steps include: S1, determining a target area and a corresponding target water system; the target water system includes a main stream and tributaries; dividing the target area into a plurality of sub-areas according to the tributary basins; S2, setting a plurality of rain sensors in each of the sub-areas to obtain precipitation data of the sub-areas; setting a plurality of flow test points upstream and downstream of the tributary to obtain flow change data of the tributary; S3, establishing a regression model of the sub-region precipitation data and the tributary flow change data; S4, evaluating the tributary forecast flow through the regression model according to the real-time precipitation in the sub-region; S5, evaluating the forecast flow of the main stream according to the forecast flow of the tributary.

2. The flood early warning and forecasting method based on smart water conservancy according to claim 1 is characterized in that: Each of the sub-areas is divided into a plurality of collection plots according to the terrain; the collection plots are respectively provided with the rainfall sensors.

3. The flood early warning and forecasting method based on smart water conservancy according to claim 2 is characterized in that: The regression model is a multiple regression model of the sub-region precipitation data and the tributary flow change data: ; The flow change data of the tributary; The area of ​​n plots of land collected in the sub-region; Collecting precipitation for n plots in the sub-region; is the regression coefficient of the n sampling plots in the sub-area.

4. The flood early warning and forecasting method based on smart water conservancy according to claim 3 is characterized in that: Before the real-time flow change data of the tributary reaches a preset value, the forecast flow of the tributary is evaluated by the regression model.

5. The flood early warning and forecasting method based on smart water conservancy according to claim 4 is characterized in that: A second regression model was also included; ； is the preset value; The precipitation of n collection plots in the sub-area when the preset value is reached; is the second regression coefficient of the n sampling plots in the sub-area.

6. The flood early warning and forecasting method based on smart water conservancy according to claim 5 is characterized in that: After the real-time flow change data of the tributary reaches the preset value, the forecast flow of the tributary is evaluated by the second regression model.

7. The flood early warning and forecasting method based on smart water conservancy according to claim 4 is characterized in that: After the real-time flow change data of the tributary reaches the preset value, The real-time precipitation in the sub-area and the historical precipitation in the sub-area are calculated by clustering algorithm to obtain the historical precipitation with the greatest similarity; The historical tributary flow change data corresponding to the historical precipitation with the greatest similarity is used to evaluate the tributary forecast flow.

8. The flood early warning and forecasting method based on smart water conservancy according to claim 4 is characterized in that: After the real-time flow change data of the tributary reaches the preset value, The real-time precipitation of each collection plot in the sub-region and the historical precipitation of each collection plot in the sub-region are calculated by clustering algorithm to obtain a plurality of historical precipitations with close similarity; The real-time change data of tributary flow and multiple historical tributary flow change data corresponding to multiple historical precipitations with similarities are calculated to obtain the closest historical tributary flow change data, which is used to evaluate the forecast tributary flow.

9. The flood early warning and forecasting method based on smart water conservancy according to claim 4 is characterized in that: The preset value is set to 15% to 20% of the average flow of the tributaries in the 24 hours before rainfall.

10. A flood warning and forecasting system based on smart water conservancy, used to implement the flood warning and forecasting method based on smart water conservancy as described in any one of claims 1 to 9, characterized in that: include: Data acquisition module: used to obtain the precipitation data of the sub-region and the flow change data of the tributary in real time; Data storage module: used to store sub-region precipitation data and tributary flow change data; Data calculation module: used for calculating the parameters of the regression model and the second regression model; and evaluating the tributary forecast flow through the regression model, the second regression model and the clustering algorithm; Result display module: used for real-time display of the precipitation data of the sub-area, the flow change data of the tributary, the forecast flow of the tributary, and the forecast flow of the main stream.

Images