A method and device for predicting reservoir water levels

Abstract

This invention discloses a method and device for predicting reservoir water levels. The method includes: for each rain gauge station associated with a target reservoir, determining the correlation between the rain gauge station and the target reservoir based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range; filtering the obtained correlations using a preset correlation threshold to determine the target rain gauge station corresponding to the filtered correlations; performing curve fitting using the cumulative daily rainfall of each target rain gauge station on a target date and the water level difference on the target date to obtain an initial fitting curve; wherein the target date is the date with the highest water level within the first preset time range; adjusting the initial fitting curve using a pre-determined rainfall change multiple to obtain a target fitting curve; and predicting the water level difference using the target fitting curve and the cumulative daily rainfall corresponding to the date to be predicted. This improves the accuracy of reservoir water level prediction.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and device for predicting reservoir water levels. Background Technology

[0002] During the flood season, it is necessary to monitor key reservoirs and other water bodies. Monitoring methods include manual inspections or video surveillance to detect potential hazards, or using sensors to obtain real-time water levels and comparing them with pre-set thresholds for early warning. These forecasting methods are relatively passive.

[0003] In addition, continuous heavy rainfall often occurs during the flood season, and the changes in reservoir water levels are affected by many factors. Relying solely on real-time water levels to determine whether to issue an early warning cannot achieve accurate prediction of reservoir water levels, thus resulting in low accuracy of early warnings.

[0004] Therefore, improving the accuracy of reservoir water level prediction is particularly important. Summary of the Invention

[0005] An exemplary embodiment of the present invention provides a method and device for predicting reservoir water levels, thereby improving the accuracy of reservoir water level prediction.

[0006] According to a first aspect of an exemplary embodiment, a method for predicting reservoir water levels is provided, the method comprising:

[0007] For each rain gauge station associated with the target reservoir, the correlation between the rain gauge station and the target reservoir is determined based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range.

[0008] The obtained correlations are filtered by applying a preset correlation threshold to determine the target rain gauge station corresponding to the filtered correlation.

[0009] An initial fitted curve is obtained by applying the cumulative daily rainfall on the target date and the water level difference on the target date to each target rain gauge station; wherein, the target date is the date with the highest water level within the first preset time range;

[0010] The initial fitting curve is adjusted by applying a predetermined multiple of rainfall change to obtain the target fitting curve;

[0011] The water level difference is predicted by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted.

[0012] The predicted water level is determined based on the obtained prediction difference and the current water level value.

[0013] According to a second aspect of an exemplary embodiment, a reservoir water level prediction device is provided, the prediction device comprising a processor and a memory, wherein:

[0014] The memory is configured as follows:

[0015] Store daily rainfall data and preset correlation thresholds for each rain gauge station;

[0016] The processor is configured to:

[0017] For each rain gauge station associated with the target reservoir, the correlation between the rain gauge station and the target reservoir is determined based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range.

[0018] The preset correlation threshold is used to filter the obtained correlations to determine the target rain gauge station corresponding to the filtered correlation.

[0019] An initial fitted curve is obtained by applying the cumulative daily rainfall on the target date and the water level difference on the target date to each target rain gauge station; wherein, the target date is the date with the highest water level within the first preset time range;

[0020] The initial fitting curve is adjusted by applying a predetermined multiple of rainfall change to obtain the target fitting curve;

[0021] The water level difference is predicted by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted.

[0022] The predicted water level is determined based on the obtained prediction difference and the current water level value.

[0023] According to a third aspect of an exemplary embodiment, a reservoir water level prediction device is provided, the device comprising:

[0024] The correlation determination module is used to determine the correlation between each rain gauge station associated with the target reservoir and the target reservoir based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range.

[0025] The filtering module is used to filter the obtained correlations by applying a preset correlation threshold and determine the target rain gauge corresponding to the filtered correlations.

[0026] The curve fitting module is used to perform curve fitting by applying the cumulative daily rainfall on the target date and the water level difference on the target date for each target rain gauge station to obtain an initial fitting curve; wherein, the target date is the date with the highest water level within the first preset time range;

[0027] The target curve determination module is used to adjust the initial fitting curve by applying a pre-determined multiple of rainfall change to obtain the target fitting curve;

[0028] The water level difference determination module is used to predict the water level difference by applying the target fitting curve and the cumulative daily rainfall corresponding to the date to be predicted;

[0029] The water level prediction module is used to determine the predicted water level based on the obtained prediction difference and the current water level value.

[0030] According to a fourth aspect of an exemplary embodiment, a computer storage medium is provided, the computer storage medium storing computer program instructions that, when executed on a computer, cause the computer to perform the reservoir water level prediction method as described in the first aspect.

[0031] In this embodiment, for each rain gauge station associated with a target reservoir, the correlation between the rain gauge station and the target reservoir is determined based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range. A preset correlation threshold is applied to filter the obtained correlations, identifying the target rain gauge station corresponding to the filtered correlation. This method of calculating the correlation of surrounding rain gauge stations and selecting those with high correlation for prediction improves the effectiveness of the prediction. Curve fitting is performed using the cumulative daily rainfall of each target rain gauge station on the target date and the water level difference on the target date to obtain an initial fitting curve, which provides a better fit. The initial fitting curve is adjusted using a predetermined rainfall change factor to account for changes in the prediction trend, resulting in a target fitting curve. The target fitting curve and the cumulative daily rainfall corresponding to the date to be predicted are used to predict the water level difference. The predicted water level change value is then summed with the current water level to obtain the predicted water level, improving the accuracy of water level prediction and assisting in early warning. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 An illustrative diagram shows an application scenario of a reservoir water level prediction method provided by an embodiment of the present invention.

[0034] Figure 2 A flowchart of a method for predicting reservoir water levels provided by an embodiment of the present invention is shown as an example;

[0035] Figure 3A flowchart illustrating a method for determining a first preset time range provided by an embodiment of the present invention is shown below;

[0036] Figure 4 An exemplary flowchart illustrates a method for determining the cumulative daily rainfall of a rain gauge station according to an embodiment of the present invention;

[0037] Figure 5 The flowchart illustrates an embodiment of the present invention of a method for determining the correlation degree between each rain gauge station and the target reservoir.

[0038] Figure 6 An exemplary flowchart of a curve fitting method provided by an embodiment of the present invention is shown;

[0039] Figure 7 An exemplary flowchart illustrates a method for determining the multiple of rainfall variation provided by an embodiment of the present invention;

[0040] Figure 8 An exemplary flowchart illustrates a method for predicting water level differences according to an embodiment of the present invention;

[0041] Figure 9 A flowchart of a method for predicting reservoir water levels provided by an embodiment of the present invention is shown as an example;

[0042] Figure 10 An exemplary illustration shows a real-time water level diagram of a reservoir provided by an embodiment of the present invention;

[0043] Figure 11 An exemplary diagram illustrating a resource analysis around a reservoir provided by an embodiment of the present invention is shown.

[0044] Figure 12 An exemplary schematic diagram of a reservoir water level prediction device provided in an embodiment of the present invention is shown;

[0045] Figure 13 An exemplary schematic diagram of a reservoir water level prediction device provided in an embodiment of the present invention is shown. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0047] During the flood season, it is necessary to monitor key reservoirs and other water bodies. Monitoring methods include manual inspections or video surveillance to detect potential hazards, or using sensors to obtain real-time water levels and comparing them with pre-set thresholds for early warning. These forecasting methods are relatively passive.

[0048] In addition, continuous heavy rainfall often occurs during the flood season, and the changes in reservoir water level are affected by many factors (rainfall interval, reservoir storage evaporation, upstream inflow, downstream outflow, etc.). Relying solely on real-time water level to determine whether to issue an early warning cannot achieve accurate prediction of reservoir water level, thus resulting in low accuracy of early warnings.

[0049] However, the aforementioned information is difficult to obtain, while rainfall data from rain gauges can be accurately obtained. Therefore, it is particularly important to learn how to use the obtained rain gauge data to scientifically predict reservoir water levels.

[0050] To address this, this application provides a method for predicting reservoir water levels. In this method, for each rain gauge station associated with a target reservoir, the correlation between the rain gauge station and the target reservoir is determined. Rain gauge stations with a correlation greater than a preset correlation threshold are then selected as target rain gauge stations. Curve fitting is performed using the cumulative daily rainfall on the target date and the water level difference on the target date for each target rain gauge station to obtain an initial fitted curve. The initial fitted curve is then optimized using a predetermined rainfall change factor to obtain a target fitted curve. The target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted are used to predict the water level difference. Finally, the predicted water level is determined based on the obtained predicted difference and the current water level. This improves the accuracy of reservoir water level prediction.

[0051] After introducing the design concept of the embodiments of this application, the following is a brief introduction to the application scenarios to which the technical solutions of the embodiments of this application can be applied. It should be noted that the application scenarios described below are only for illustrating the embodiments of this application and are not intended to limit the scope. In specific implementation, the technical solutions provided by the embodiments of this application can be flexibly applied according to actual needs.

[0052] refer to Figure 1 The diagram illustrates an application scenario for a reservoir water level prediction method. During periods of continuous heavy rainfall, the sudden surge in reservoir water levels due to high-intensity precipitation puts immense pressure on the reservoir, increasing the risk of dam failure. A dam failure would pose a significant threat to the lives and property of people downstream. Therefore, accurate and scientific reservoir water level prediction is crucial.

[0053] To further illustrate the technical solutions provided in the embodiments of this application, a detailed description is provided below in conjunction with the accompanying drawings and specific implementation methods. Although the embodiments of this application provide method operation steps as shown in the following embodiments or drawings, the method may include more or fewer operation steps based on conventional or non-inventive methods. In steps where there is no logically necessary causal relationship, the execution order of these steps is not limited to the execution order provided in the embodiments of this application.

[0054] The following is combined Figure 1 The application scenarios shown are for reference. Figure 2 The flowchart shown illustrates a method for predicting reservoir water levels, and explains the technical solution provided in the embodiments of this application.

[0055] First, the data storage process will be explained, so that when there is a need for data application later, it can be directly retrieved from the stored data.

[0056] Based on date, the water level data of the target reservoir for the current time (2022) and the previous two years (2020) are sorted in descending order and saved to the data table his_sstore. Each record includes the target reservoir code, name, address, latitude and longitude, water level measurement date, and daily water level difference. The cumulative daily rainfall data of the rain gauge stations is calculated, and each record includes the rain gauge code, name, address, latitude and longitude, target reservoir code, water level measurement date, and cumulative daily rainfall, and saved to the data table his_ystore. This is used for preprocessing the historical water level data of the target reservoir and the cumulative daily rainfall data of the rain gauge stations.

[0057] Secondly, the method for predicting reservoir water levels in the embodiments of this application will be explained:

[0058] S201. For each rain gauge station associated with the target reservoir, determine the correlation between the rain gauge station and the target reservoir based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within the first preset time range.

[0059] S202. Apply a preset correlation threshold to filter the obtained correlations and determine the target rain gauge station corresponding to the filtered correlations.

[0060] S203. Apply the cumulative daily rainfall and water level difference of the target date for each target rain gauge station to perform curve fitting to obtain an initial fitting curve; wherein, the target date is the date with the highest water level within the first preset time range.

[0061] S204. Adjust the initial fitting curve using a predetermined multiple of rainfall change to obtain the target fitting curve.

[0062] S205. Use the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted to predict the water level difference.

[0063] S206. Determine the predicted water level based on the obtained prediction difference and the current water level value.

[0064] In this embodiment, for each rain gauge station associated with a target reservoir, the correlation between the rain gauge station and the target reservoir is determined based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range. A preset correlation threshold is applied to filter the obtained correlations, identifying the target rain gauge station corresponding to the filtered correlation. This method of calculating the correlation of surrounding rain gauge stations and selecting those with high correlation for prediction improves the effectiveness of the prediction. Curve fitting is performed using the cumulative daily rainfall of each target rain gauge station on the target date and the water level difference on the target date to obtain an initial fitting curve, which provides a better fit. The initial fitting curve is adjusted using a predetermined rainfall change factor to account for changes in the prediction trend, resulting in a target fitting curve. The target fitting curve and the cumulative daily rainfall corresponding to the date to be predicted are used to predict the water level difference. The predicted water level change value is then summed with the current water level to obtain the predicted water level, improving the accuracy of water level prediction and assisting in early warning.

[0065] Regarding S201, in practical applications, each reservoir is associated with at least one rain gauge station. This association is determined based on the degree of influence of rainfall measured by the rain gauge station on the reservoir's water level. This degree of influence is related to, for example, the relative position of the rain gauge station and the reservoir; the closer the rain gauge station is to the reservoir, the greater the influence. For example, the reservoir used in this embodiment is referred to as the target reservoir. In practical applications, each rain gauge station associated with the target reservoir is identified. Then, for each rain gauge station associated with the target reservoir, the correlation between the rain gauge station and the target reservoir is determined based on the cumulative daily rainfall of the rain gauge station within a first preset time range and the difference in water level between the rain gauge station and the target reservoir.

[0066] refer to Figure 3 The process of determining the first preset time range is explained below:

[0067] S301. Based on the highest water level of the target reservoir each day within the initial preset time range, filter a preset number of highest water level values.

[0068] The daily highest water level refers to the highest water level within a 24-hour period. The initial preset time range is, for example, 365 days in 2021, and the daily highest water level values ​​(Hmax1-Hmax) of the target reservoir during these 365 days are obtained. 365 If the preset quantity is 30, then the 365 highest water level values ​​are sorted from highest to lowest, and the top 30 highest water level values ​​are selected.

[0069] S302. Determine the dates corresponding to the preset number of highest water level values ​​selected to form a first preset time range; wherein, the maximum value of the water level for each day is the highest water level value for that day.

[0070] Specifically, the dates corresponding to these 30 highest water level values ​​are determined. For example, if the corresponding 30 days are distributed across 3 days in June, 15 days in July, and 12 days in August, then the first preset time range is these 30 days. It should be noted that these 30 days may or may not be consecutive; this is just an example and does not constitute a specific limitation.

[0071] refer to Figure 4 The process of determining the cumulative daily rainfall at each rain gauge station will be explained.

[0072] S401. For each reference date within the first preset time range, determine the daily rainfall data for the N days preceding that reference date; N is a preset integer.

[0073] The first preset time range still uses the aforementioned 30 days as an example. Each reference date is one day within those 30 days. Taking the first day of those 30 days (e.g., June 27th) as an example, with N set to 5, the rainfall data for each of the four days—June 22nd, June 23rd, June 24th, June 25th, and June 26th—is determined. N is determined based on actual needs; for example, if the first five days have a greater impact on reservoir water levels, then N is set to 5.

[0074] S402. Based on the daily rainfall data and corresponding weights, determine the cumulative daily rainfall for the reference date.

[0075] The Analytic Hierarchy Process (AHP) was used to calculate the weights of the daily rainfall on the water level for the first five days. The weights for the following days were determined to be 0.1 for day N-1, 0.2 for day N-2, 0.4 for day N-3, 0.2 for day N-4, and 0.1 for day N-5.

[0076] Thus, the cumulative daily rainfall for the reference date is determined as follows:

[0077] Dy=0.1*a1+0.2*a2+0.4*a3+0.2*a4+0.1*a5;

[0078] Where a1 is the rainfall on day N-1, a2 is the rainfall on day N-2, a3 is the rainfall on day N-3, a4 is the rainfall on day N-4, and a5 is the rainfall on day N-5.

[0079] The above process is illustrated using a reference date within a first preset time range. In this case, the dates of the previous N days are past dates, and the corresponding rainfall is obtained through rain gauges.

[0080] refer to Figure 5The process of determining the correlation between each rain gauge station associated with the target reservoir and the target reservoir is explained:

[0081] S501, for each rain gauge station associated with the target reservoir, obtain the first sequence number determined by sorting the cumulative daily rainfall of the rain gauge station from high to low within the first preset time range; obtain the second sequence number determined by sorting the water level difference of the target reservoir from high to low within the first preset time range.

[0082] Taking the rain gauge Y1 associated with the target reservoir as an example, the cumulative daily rainfall of this rain gauge over a 30-day period within a first preset time range is obtained and sorted from highest to lowest, with the first sequence number being 1 to 30. The daily water level difference (the difference between the highest and lowest water levels) of the target reservoir during these 30 days is obtained and sorted from highest to lowest, with the second sequence number being 1 to 30.

[0083] S502. For each first serial number, determine the second serial number corresponding to the date to which the first serial number belongs, and determine the difference between the first serial number and the second serial number.

[0084] For each first sequence number, such as first sequence number 2, its date is determined to be July 4th within 30 days. The second sequence number corresponding to the reservoir water level on July 4th is then determined, for example, second sequence number 5. The difference between first sequence number 2 and second sequence number 5 is determined to be the sequence number difference 3.

[0085] S503. Determine the correlation between the rain gauge and the target reservoir based on the sum of squares of each difference and the number of rain gauges.

[0086] The above example uses the first serial number to obtain a serial number difference. When the first preset range is 30 days, 30 serial number differences are obtained. The correlation between the rain gauge and the target reservoir is calculated as follows:

[0087]

[0088] Where ρ is the relevance, and the difference in ordinal numbers is represented by d. i This means that n is the number of days included in the first preset time range, such as 30 in the example above.

[0089] Regarding S202, a preset correlation threshold is applied to filter the obtained correlations and determine the target rain gauge station corresponding to the filtered correlations.

[0090] For example, if there are 25 rain gauge stations associated with a target reservoir, and the preset correlation threshold is 0.6, then among the 25 correlation scores obtained from these 25 rain gauge stations, those with a correlation score greater than 0.6 are selected and designated as target rain gauge stations. In this example, for instance, 15 target rain gauge stations are selected.

[0091] Regarding S203, curve fitting is performed using the cumulative daily rainfall and water level difference on the target date for each target rain gauge station to obtain the initial fitted curve.

[0092] For example, the target date is the date with the highest water level within a first preset time range. For instance, if the first preset time range is 30 days, then the target date is the day with the highest water level (e.g., July 6th). (See reference) Figure 6 The process of fitting the curve is explained.

[0093] S601. Determine the curve to be fitted.

[0094] Assume the curve to be fitted is a cubic curve, y = a + bt + ct 2 +dt 3 , where a, b, c, and d are undetermined constant parameters.

[0095] S602. Using the cumulative daily rainfall of each target rain gauge station on the target date as the independent variable and the water level difference on the target date as the dependent variable, perform curve fitting to obtain the constant parameters in the curve to be fitted.

[0096] The independent variable t in the curve to be fitted is the cumulative daily rainfall of each target rain gauge on the target date (e.g., July 6th), and the dependent variable y in the curve to be fitted is the water level difference on the target date.

[0097] During the fitting process, the following prediction model using the cubic curve method was determined:

[0098] ∑y=na+b∑t+c∑t 2 +d∑t 3

[0099] ∑ty=a∑t+b∑t 2 +c∑t 3 +d∑t 4

[0100] ∑t 2 y=a∑t 2 +b∑t 3 +c∑t 4 +d∑t 5

[0101] ∑t 3 y=a∑t 3 +b∑t 4 +c∑t 5 +d∑t 6

[0102] The summation symbol represents the summation of 15 data points from 15 target rain gauge stations. Then, by performing the inverse solution, we obtain the values ​​of a, b, c, and d, for example, a = -0.0054, b = -0.00800, c = 0.3083, and d = 0.1047.

[0103] S603. Apply constant parameters to adjust the curve to be fitted to obtain the initial fitted curve.

[0104] Substitute the values ​​of a, b, c, and d into the curve to be fitted to obtain the initial fitted curve, for example, y = -0.0054 - 0.00800t + 0.3083t. 2 +0.1047t 3 .

[0105] Regarding S204, in the example above, the initial fitting curve is obtained using historical data within a first time range, such as data from last year 2021 at the current time. Therefore, in order to obtain a target fitting curve that can be used to predict the reservoir water volume this year, the initial fitting curve is adjusted by a predetermined multiple of rainfall change to obtain the target fitting curve.

[0106] Among them, reference Figure 7 The process of determining the predetermined multiple of rainfall variation is explained below:

[0107] 701. For each target rain gauge station, determine the ratio of a first parameter for a first preset time range to a second parameter for a second preset time range; wherein, the first parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the first preset time range to the sum of the water level differences of the target reservoir; the second parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the second preset time range to the sum of the water level differences of the target reservoir.

[0108] The first preset time range is, for example, 2021 (last year of the current time), and the second preset time range is 2020 (the year before the current time). Within the first preset time range, there are 30 days. The cumulative daily rainfall of the target rain gauge stations over these 30 days is summed to obtain H1. Then, the water level differences of the target reservoirs over these 30 days are summed to obtain H2. The ratio B1 to H2 is then determined. Similarly, within the second preset time range, there are also 30 days. The cumulative daily rainfall of the target rain gauge stations over these 30 days is summed to obtain H3. Then, the water level differences of the target reservoirs over these 30 days are summed to obtain H4. The ratio B2 to H3 is then determined. Finally, the ratio S1 of B1 to B2 is determined as the ratio corresponding to the target rain gauge station.

[0109] S702. Add up the ratios corresponding to each target rain gauge station to obtain the predetermined multiple of rainfall change.

[0110] The above describes the process of determining the ratio corresponding to a target rain gauge station. The 15 ratios obtained from the 15 target rain gauge stations are added together to obtain the predetermined rainfall change multiple Q.

[0111] Regarding S205, as mentioned above, the initial fitted curve y is adjusted by applying a pre-determined multiple of rainfall change Q to obtain the target fitted curve sw = Qy. The water level difference is then predicted using the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted.

[0112] refer to Figure 8 The process of predicting water level differences is explained below:

[0113] S801. For each target rain gauge station, the predicted value of the water level difference corresponding to the target rain gauge station is determined by applying the cumulative daily rainfall corresponding to the date to be predicted for the target rain gauge station and the target fitted curve.

[0114] The target fitting curve obtained in the above example is:

[0115] y = -0.0054 - 0.00800t + 0.3083t 2 +0.1047t 3

[0116] For each target rain gauge station, the independent variable is the cumulative daily rainfall at that station on the date to be predicted, and the dependent variable is the predicted result y1 of the water level difference for that target rain gauge station.

[0117] It should be noted that the calculation of the cumulative daily rainfall for the predicted date can refer to the process of determining the cumulative daily rainfall within the first preset time range. The difference is that when determining the daily rainfall for the N days before the reference date, if the dates of the N days before the reference date are past dates, the corresponding rainfall is obtained from rain gauges; if the dates of the N days before the reference date are future dates, the corresponding rainfall is obtained from meteorological forecast data.

[0118] S802. Based on the predicted values ​​of each water level difference and their corresponding weights, determine the predicted difference of the target reservoir; where the corresponding weights represent the correlation between the target rain gauge station and the target reservoir.

[0119] For example, with 15 target rain gauges, 15 prediction results can be obtained, and each prediction result has a corresponding weight. For instance, the corresponding weight can be the correlation between the target rain gauge and the target reservoir. In this way, by weighted summation, the predicted difference in water level of the target reservoir can be obtained.

[0120] S206 involves determining the predicted water level based on the obtained prediction difference and the current water level value.

[0121] The predicted water level is obtained by adding the predicted difference to the current water level. For example, if the predicted date is one day from the current date, the current water level is added to the predicted difference to get the predicted water level for that day. This process is repeated to obtain the predicted water level for the reservoir for 2, 3, 4 days, etc., based on how many millimeters of rain will fall each day in the next few days.

[0122] After obtaining the predicted water level, the predicted reservoir water level is compared with the preset flood control limit water level to issue reservoir warning information in advance. Based on GIS (Geographic Information System) maps, spatial analysis can be used to obtain information such as the danger sources around the reservoir and rescue forces. The above map provides intuitive guidance for leadership decision-making, assists leaders in making typhoon prevention arrangements in advance, and effectively avoids the occurrence of dangerous situations.

[0123] To further improve the technical solution of this application, a complete flowchart is provided below. (Reference) Figure 9 The flowchart shows a method for predicting reservoir water levels.

[0124] S901. For each rain gauge station associated with the target reservoir, obtain the first sequence number determined by sorting the cumulative daily rainfall of the rain gauge station from high to low within the first preset time range; obtain the second sequence number determined by sorting the water level difference of the target reservoir from high to low within the first preset time range.

[0125] S902. For each first serial number, determine the second serial number corresponding to the date to which the first serial number belongs, and determine the difference between the first serial number and the second serial number.

[0126] S903. Determine the correlation between the rain gauge and the target reservoir based on the sum of squares of each difference and the number of rain gauges.

[0127] S904. Apply a preset correlation threshold to filter the obtained correlations and determine the target rain gauge station corresponding to the filtered correlations.

[0128] S905. Determine the curve to be fitted.

[0129] S906. Using the cumulative daily rainfall on the target date for each target rain gauge as the independent variable and the water level difference on the target date as the dependent variable, perform curve fitting to obtain the constant parameters in the curve to be fitted.

[0130] S907. Apply constant parameters to adjust the curve to be fitted to obtain the initial fitted curve.

[0131] S908. Adjust the initial fitting curve using a predetermined multiple of rainfall change to obtain the target fitting curve.

[0132] S909. For each target rain gauge station, the predicted value of the water level difference corresponding to the target rain gauge station is determined by applying the cumulative daily rainfall corresponding to the date to be predicted for the target rain gauge station and the target fitted curve.

[0133] S910. Based on the predicted values ​​of each water level difference and their corresponding weights, determine the predicted difference of the target reservoir.

[0134] S911. Determine the predicted water level based on the obtained prediction difference and the current water level value.

[0135] In summary, in this embodiment, correlation calculations are performed on surrounding rain gauge stations, and stations with high correlation are selected for prediction, which improves the effectiveness of the prediction. The cumulative daily rainfall of each rain gauge station is calculated based on different weights, and this cumulative daily rainfall is used to predict the difference in reservoir water levels, which improves the accuracy of the prediction. Furthermore, since the prediction of reservoir water levels is for areas with a distinct flood season, using a cubic curve equation provides a better fit. As the trend of predicted water level changes with the year, the curve fitting is corrected using multiples of the curve's trend, taking into account the issue of changing prediction trends. Additionally, the data from the selected rain gauge stations is used to predict the difference in reservoir water levels based on the expected rainfall over the next few days. The predicted reservoir water levels from multiple rain gauge stations are then weighted, summed, and averaged. The weights are calculated using the correlation of the rain gauge stations, taking into account the correlation of the predicted rain gauge stations and making the prediction results more accurate. The predicted water level change is summed with the current water level, and compared with the reservoir's flood control limit level to determine whether the water level will exceed the flood control limit level in the next few days. This improves the accuracy of water level prediction and assists leaders in making typhoon prevention arrangements in advance.

[0136] In a specific example Figure 10 A schematic diagram of real-time water level in a reservoir is shown. In this example, the dam crest elevation is 84.4m, the preset flood limit water level is 81m, the current water level is 73.78m, and the predicted water level is 75.01m. Figure 11 A schematic diagram of resource analysis around a reservoir is shown, in which the distribution of water network includes the distribution of rain gauges. In addition, the location and quantity of each resource distribution are only for illustration and do not constitute specific limitations.

[0137] like Figure 12 As shown, based on the same inventive concept, this embodiment of the invention provides a reservoir water level prediction device, which includes a correlation determination module 121, a screening module 122, a curve fitting module 123, a target curve determination module 124, a water level difference determination module 125, and a water level prediction module 126.

[0138] Among them, the correlation determination module 121 is used to determine the correlation between the rain gauge station and the target reservoir for each rain gauge station associated with the target reservoir, based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within a first preset time range.

[0139] The filtering module 122 is used to filter the obtained correlations by applying a preset correlation threshold and determine the target rain gauge station corresponding to the filtered correlation.

[0140] The curve fitting module 123 is used to perform curve fitting by applying the cumulative daily rainfall and the water level difference of the target date for each target rain gauge station to obtain an initial fitting curve; wherein, the target date is the date with the highest water level within the first preset time range.

[0141] The target curve determination module 124 is used to adjust the initial fitting curve by applying a pre-determined multiple of rainfall change to obtain the target fitting curve;

[0142] The water level difference determination module 125 is used to predict the water level difference by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted;

[0143] The water level prediction module 126 is used to determine the predicted water level based on the obtained prediction difference and the current water level value.

[0144] In some exemplary embodiments, the relevance determination module 121 is specifically used for:

[0145] For each rain gauge station associated with the target reservoir, obtain the first sequence number determined by sorting the cumulative daily rainfall of the rain gauge station from high to low within the first preset time range; obtain the second sequence number determined by sorting the water level difference of the target reservoir from high to low within the first preset time range.

[0146] For each first sequence number, determine the second sequence number corresponding to the date to which the first sequence number belongs, and determine the difference between the first sequence number and the second sequence number;

[0147] The correlation between the rain gauges and the target reservoir is determined by the sum of squares of the differences and the number of rain gauges.

[0148] In some exemplary embodiments, the curve fitting module 123 is specifically used for:

[0149] Determine the curve to be fitted;

[0150] The cumulative daily rainfall on the target date for each target rain gauge is used as the independent variable, and the water level difference on the target date is used as the dependent variable. Curve fitting is performed to obtain the constant parameters in the curve to be fitted.

[0151] The curve to be fitted is adjusted by applying constant parameters to obtain the initial fitted curve.

[0152] In some exemplary embodiments, the water level difference determination module 125 is specifically used for:

[0153] For each target rain gauge station, the predicted water level difference value for the target rain gauge station is determined by applying the cumulative daily rainfall corresponding to the date to be predicted and the target fitted curve.

[0154] Based on the predicted values ​​of each water level difference and their corresponding weights, the predicted difference of the target reservoir is determined; where the corresponding weights represent the correlation between the target rain gauge station and the target reservoir.

[0155] In some exemplary embodiments, a rainfall change multiple determination module is also included, which determines the rainfall change multiple in the following manner:

[0156] For each target rain gauge station, the ratio of a first parameter for a first preset time range to a second parameter for a second preset time range is determined; wherein, the first parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the first preset time range to the sum of the water level differences of the target reservoir; the second parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the second preset time range to the sum of the water level differences of the target reservoir.

[0157] The ratios corresponding to each target rain gauge station are added together to obtain the predetermined multiple of rainfall change.

[0158] In some exemplary embodiments, a rainfall determination module is also included, which determines the cumulative daily rainfall over a reference time range in the following manner:

[0159] For each reference date within the reference time range, determine the daily rainfall data for the N days preceding that reference date; the reference time range includes a first preset time range, a second preset time range, or a time range to be predicted; N is a pre-set integer;

[0160] Based on the daily rainfall data and corresponding weights, determine the cumulative daily rainfall for the reference date;

[0161] Where the dates of the previous N days are past dates, the corresponding rainfall is obtained from rain gauges; where the dates of the previous N days are future dates, the corresponding rainfall is obtained from meteorological forecast data.

[0162] In some exemplary embodiments, a first preset time range determination module is also included, used to determine the first preset time range:

[0163] Based on the highest water level of the target reservoir each day within the initial preset time range, a preset number of highest water level values ​​are selected;

[0164] The dates corresponding to the selected preset number of highest water level values ​​constitute the first preset time range; among them, the maximum value of the water level for each day is the highest water level value for that day.

[0165] In some exemplary embodiments, an early warning module is also included, which is used to issue an early warning when the predicted water level is greater than the preset flood limit water level after the predicted water level is determined based on the obtained prediction difference and the current water level value.

[0166] Since this device is the same as the device in the method of this invention, and the principle of the device in solving the problem is similar to that of the method, the implementation of the device can be referred to the implementation of the method, and the repeated parts will not be described again.

[0167] like Figure 13 As shown, based on the same inventive concept, this embodiment of the invention provides a reservoir water level prediction device, which includes a processor 131 and a memory 132.

[0168] Memory 132 is configured as follows:

[0169] Store daily rainfall data and preset correlation thresholds for each rain gauge station;

[0170] Processor 131 is configured as follows:

[0171] For each rain gauge station associated with the target reservoir, the correlation between the rain gauge station and the target reservoir is determined based on the cumulative daily rainfall of the rain gauge station and the water level difference of the target reservoir within the first preset time range.

[0172] The obtained correlations are filtered by applying a preset correlation threshold to determine the target rain gauge station corresponding to the filtered correlation.

[0173] The initial fitted curve is obtained by using the cumulative daily rainfall and the water level difference on the target date for each target rain gauge station; where the target date is the date with the highest water level within the first preset time range.

[0174] The initial fitting curve is adjusted by applying a predetermined multiple of rainfall change to obtain the target fitting curve;

[0175] Water level difference is predicted by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted.

[0176] The predicted water level is determined based on the obtained prediction difference and the current water level value.

[0177] In some exemplary embodiments, processor 131 is further configured to:

[0178] For each rain gauge station associated with the target reservoir, obtain the first sequence number determined by sorting the cumulative daily rainfall of the rain gauge station from high to low within the first preset time range; obtain the second sequence number determined by sorting the water level difference of the target reservoir from high to low within the first preset time range.

[0179] For each first sequence number, determine the second sequence number corresponding to the date to which the first sequence number belongs, and determine the difference between the first sequence number and the second sequence number;

[0180] The correlation between the rain gauges and the target reservoir is determined by the sum of squares of the differences and the number of rain gauges.

[0181] In some exemplary embodiments, processor 131 is further configured to:

[0182] Determine the curve to be fitted;

[0183] The cumulative daily rainfall on the target date for each target rain gauge is used as the independent variable, and the water level difference on the target date is used as the dependent variable. Curve fitting is performed to obtain the constant parameters in the curve to be fitted.

[0184] The curve to be fitted is adjusted by applying constant parameters to obtain the initial fitted curve.

[0185] In some exemplary embodiments, processor 131 is further configured to:

[0186] For each target rain gauge station, the predicted water level difference value for the target rain gauge station is determined by applying the cumulative daily rainfall corresponding to the date to be predicted and the target fitted curve.

[0187] Based on the predicted values ​​of each water level difference and their corresponding weights, the predicted difference of the target reservoir is determined; where the corresponding weights represent the correlation between the target rain gauge station and the target reservoir.

[0188] In some exemplary embodiments, processor 131 is also configured to determine the rainfall change factor in the following manner:

[0189] For each target rain gauge station, the ratio of a first parameter for a first preset time range to a second parameter for a second preset time range is determined; wherein, the first parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the first preset time range to the sum of the water level differences of the target reservoir; the second parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the second preset time range to the sum of the water level differences of the target reservoir.

[0190] The ratios corresponding to each target rain gauge station are added together to obtain the predetermined multiple of rainfall change.

[0191] In some exemplary embodiments, processor 131 is also configured to determine the cumulative daily rainfall over a reference time range in the following manner:

[0192] For each reference date within the reference time range, determine the daily rainfall data for the N days preceding that reference date; the reference time range includes a first preset time range, a second preset time range, or a time range to be predicted; N is a pre-set integer;

[0193] Based on the daily rainfall data and corresponding weights, determine the cumulative daily rainfall for the reference date;

[0194] Where the dates of the previous N days are past dates, the corresponding rainfall is obtained from rain gauges; where the dates of the previous N days are future dates, the corresponding rainfall is obtained from meteorological forecast data.

[0195] In some exemplary embodiments, the processor 131 is further configured to determine a first preset time range in the following manner:

[0196] Based on the highest water level of the target reservoir each day within the initial preset time range, a preset number of highest water level values ​​are selected;

[0197] The dates corresponding to the selected preset number of highest water level values ​​constitute the first preset time range; among them, the maximum value of the water level for each day is the highest water level value for that day.

[0198] In some exemplary embodiments, the processor 131 is also configured to issue an early warning if the predicted water level is greater than a preset flood limit water level.

[0199] This invention also provides a computer storage medium storing computer program instructions, which, when executed on a computer, cause the computer to perform the steps of the above-described method for predicting reservoir water levels.

[0200] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0201] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0202] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0203] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0204] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method of predicting a water level of a reservoir, characterized by, include: For each rain gauge station associated with the target reservoir, obtain the first sequence number determined by sorting the cumulative daily rainfall of the rain gauge station from high to low within a first preset time range; obtaining a second serial number of the target reservoir water level difference from high to low in each day within the first preset time range; for each first serial number, determining the second serial number corresponding to the date to which the first serial number belongs, and determining the difference between the first serial number and the second serial number; according to the square sum of each difference value and the number of rainfall stations, the correlation degree of the rainfall station and the target reservoir is determined; wherein the calculation formula of the correlation degree of the rainfall station and the target reservoir is: ; wherein, is the correlation degree, d i is the i-th difference value, and n is the number of days included in the first preset time range; The obtained correlations are filtered by applying a preset correlation threshold to determine the target rain gauge station corresponding to the filtered correlation. An initial fitted curve is obtained by applying the cumulative daily rainfall on the target date and the water level difference on the target date to each target rain gauge station; wherein, the target date is the date with the highest water level within the first preset time range; The initial fitting curve is adjusted by applying a predetermined multiple of rainfall change to obtain the target fitting curve; The water level difference is predicted by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted. The predicted water level is determined based on the obtained prediction difference and the current water level value; The predetermined rainfall change multiple is determined as follows: For each target rain gauge station, the ratio of a first parameter for a first preset time range and a second parameter for a second preset time range is determined; wherein the first parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the first preset time range to the sum of the water level differences of the target reservoir; the second parameter is the ratio of the sum of the cumulative daily rainfall of the target rain gauge station within the second preset time range to the sum of the water level differences of the target reservoir; the predetermined rainfall change multiple is obtained by summing the ratios corresponding to each target rain gauge station.

2. The method according to claim 1, characterized in that, The method involves applying the cumulative daily rainfall and the water level difference on the target date for each target rain gauge station to perform curve fitting, resulting in an initial fitted curve, including: Determine the curve to be fitted; Using the cumulative daily rainfall of each target rain gauge station on the target date as the independent variable and the water level difference on the target date as the dependent variable, curve fitting is performed to obtain the constant parameters in the curve to be fitted. The constant parameters are applied to adjust the curve to be fitted to obtain an initial fitted curve.

3. The method according to claim 1, characterized in that, The method of predicting water level difference by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted includes: For each target rain gauge station, the predicted water level difference value corresponding to the target rain gauge station is determined by applying the cumulative daily rainfall corresponding to the date to be predicted and the target fitting curve. The predicted difference of the target reservoir is determined based on the predicted values ​​of each water level difference and their corresponding weights; wherein the corresponding weights are the correlation between the corresponding target rain gauge station and the target reservoir.

4. The method according to claim 1, characterized in that, The cumulative daily rainfall within the reference time range is determined as follows: For each reference date within the reference time range, determine the daily rainfall data for the N days preceding the reference date; the reference time range includes a first preset time range, a second preset time range, or a time range to be predicted; N is a pre-set integer; Based on the daily rainfall data and corresponding weights, the cumulative daily rainfall for the reference date is determined; Where the dates of the previous N days are past dates, the corresponding rainfall is obtained from rain gauges; where the dates of the previous N days are future dates, the corresponding rainfall is obtained from meteorological forecast data.

5. The method according to claim 1, characterized in that, The first preset time range is determined in the following way: Based on the highest water level of the target reservoir each day within the initial preset time range, a preset number of highest water level values ​​are selected; The dates corresponding to the selected preset number of highest water level values ​​constitute a first preset time range; wherein, the maximum value of the water level for each day is the highest water level value for that day.

6. The method according to any one of claims 1 to 5, characterized in that, After determining the predicted water level based on the obtained prediction difference and the current water level value, the method further includes: If the predicted water level is higher than the preset flood limit water level, an early warning will be issued.

7. A reservoir water level prediction device, characterized in that, Including processor and memory: The memory is configured as follows: Store daily rainfall data and preset correlation thresholds for each rain gauge station; The processor is configured to: For each rain gauge station associated with the target reservoir, a first sequence number is determined by sorting the cumulative daily rainfall of the rain gauge station from high to low within a first preset time range; a second sequence number is determined by sorting the water level difference of the target reservoir from high to low within the first preset time range; for each first sequence number, a second sequence number corresponding to the date to which the first sequence number belongs is determined, and the difference between the first sequence number and the second sequence number is determined; based on the sum of the squares of each difference and the number of rain gauge stations, the correlation between the rain gauge station and the target reservoir is determined; wherein, the formula for calculating the correlation between the rain gauge station and the target reservoir is: ;in, For relevance, d i Let n be the i-th difference, and n be the number of days included in the first preset time range; The preset correlation threshold is used to filter the obtained correlations to determine the target rain gauge station corresponding to the filtered correlation. An initial fitted curve is obtained by applying the cumulative daily rainfall on the target date and the water level difference on the target date to each target rain gauge station; wherein, the target date is the date with the highest water level within the first preset time range; The initial fitting curve is adjusted by applying a predetermined multiple of rainfall change to obtain the target fitting curve; The water level difference is predicted by applying the target fitted curve and the cumulative daily rainfall corresponding to the date to be predicted. The predicted water level is determined based on the obtained prediction difference and the current water level value; The processor is further configured to determine the predetermined rainfall change factor in the following manner: For each target rain gauge station, a ratio is determined between a first parameter for a first preset time range and a second parameter for a second preset time range. The first parameter is the ratio of the sum of the cumulative daily rainfall at the target rain gauge station within the first preset time range to the sum of the water level differences at the target reservoir. The second parameter is the ratio of the sum of the cumulative daily rainfall at each target rain gauge station within the second preset time range to the sum of the water level differences at the target reservoir. The ratios corresponding to each target rain gauge station are summed to obtain the predetermined rainfall change multiple.

Images