A method, apparatus and electronic equipment for reservoir scheduling based on rainfall

Abstract

This solution relates to a reservoir scheduling method, apparatus, and electronic device based on rainfall. The rainfall-based reservoir scheduling method includes: 100. Obtaining forecast rainfall data for a first zone; calculating N deterministic coefficients for each rainfall data point among N rainfall data points of the first zone stored in a database; 200. Determining the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data; 300. Scheduling the first zone according to the determined scheduling scheme. This invention solves the problem in existing technologies where extensive computation is required to obtain a reservoir scheduling scheme during reservoir scheduling by storing historical reservoir scheduling schemes in a database and directly retrieving the scheme from the database during scheduling.

Description

Technical Field

[0001] This invention relates to the field of hydrology and water resources, and in particular to a reservoir scheduling method, apparatus and electronic equipment based on rainfall. Background Technology

[0002] Reservoirs, as important water conservancy projects for human utilization and management of water resources, effectively resolve the contradiction between water resource allocation and the needs of human socio-economic development. Reservoir scheduling schemes, as powerful tools guiding reservoir operation, are one of the key technologies for realizing the comprehensive benefits of reservoirs and ensuring the safety of downstream communities. Current reservoir scheduling research largely focuses on the scheduling algorithms themselves. When given rainfall conditions, optimization and accuracy of the algorithms are used to improve the accuracy of output results such as controlled outflow and compensated flow. However, historical scheduling schemes are not stored, and there is a lack of a comprehensive and knowledge-sharing reservoir scheduling database. Furthermore, methods for retrieving data from this database are lacking. Each time a reservoir scheduling scheme is determined, extensive calculations using algorithms are required to obtain the desired scheme. Summary of the Invention

[0003] The purpose of this invention is to at least solve one of the technical problems existing in the prior art. To this end, in one aspect, this invention proposes a reservoir scheduling method based on rainfall.

[0004] The first aspect of this solution proposes a reservoir scheduling method based on rainfall, including:

[0005] Obtain the forecast rainfall data for the first zone, which is any zone of the reservoir;

[0006] The deterministic coefficients of the predicted rainfall data and the N rainfall data of the first partition stored in the database are calculated to obtain N deterministic coefficients, where N is an integer greater than 1.

[0007] The scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients is determined as the scheduling scheme corresponding to the forecast rainfall data.

[0008] The first partition is scheduled according to the determined scheduling scheme, which includes the flow generation of the first partition, the flow inflow of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow.

[0009] Furthermore, it also includes: obtaining the current water level of the reservoir to determine the initial water level for regulation.

[0010] The step of determining the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data includes:

[0011] The largest deterministic coefficient among the N deterministic coefficients and the scheduling scheme corresponding to the first starting water level are determined as the scheduling scheme corresponding to the forecast rainfall data.

[0012] Furthermore, it also includes: determining the runoff generation and runoff concentration of the first partition based on N rainfall data of the first partition.

[0013] Determine M starting water levels, where M is an integer greater than 1.

[0014] Based on the N rainfall data, the runoff and runoff of the first zone, the M starting water levels, and a multi-objective optimization algorithm, N*M scheduling schemes for the first zone are determined, where the highest water level and maximum outflow of the reservoir corresponding to the scheduling scheme are minimized.

[0015] Store the N*M scheduling schemes of the first partition into the database.

[0016] Furthermore, determining the runoff generation and runoff concentration of the first zone based on N rainfall data points of the first zone includes:

[0017] The parameters of the runoff generation and confluence model are determined based on the historical rainfall and runoff data of the first zone, thus obtaining the first runoff generation and confluence model.

[0018] The runoff and runoff of the first partition are determined based on N rainfall data points from the first partition and the first runoff-confluence model.

[0019] Furthermore, it also includes: obtaining N images of the first partition and their corresponding feature values, wherein the images are rainfall images and the feature values ​​include the maximum rainfall, the minimum rainfall, and the batch average rainfall.

[0020] Rainfall data is extracted from the N images based on the feature values ​​to obtain N rainfall data points for the first partition.

[0021] Furthermore, the image contains X grid units, where X is an integer greater than 1.

[0022] The step of extracting rainfall data from the N images based on the feature values ​​to obtain N rainfall data points for the first partition includes:

[0023] Determine the pixel points and coordinate points of X grid units in the first image, where the first image is any one of the N images.

[0024] Extract the RGB values ​​of the pixel.

[0025] The rainfall amount corresponding to the grid cell is calculated based on the RGB value and the feature value, resulting in X rainfall amounts.

[0026] N rainfall data points for the first partition are determined based on the coordinates of the X grid cells and the X rainfall amounts.

[0027] Furthermore, determining the M starting water levels includes:

[0028] The M starting water levels are determined based on the maximum and minimum values ​​of the starting water level range of the first partition and a water level interval setting.

[0029] A second aspect of the present invention provides a reservoir scheduling device based on rainfall, comprising:

[0030] The acquisition unit is used to acquire the forecast rainfall data of the first zone, which is any zone of the reservoir.

[0031] The calculation unit is used to calculate the deterministic coefficient of each rainfall data in the forecast rainfall data and the N rainfall data of the first partition stored in the database to obtain N deterministic coefficients, where N is an integer greater than 1.

[0032] The determining unit is used to determine the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data.

[0033] The scheduling unit is used to schedule the first partition according to a determined scheduling scheme, which includes the flow generation of the first partition, the flow inflow of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow.

[0034] A third aspect of the present invention provides an electronic device, including a memory and a processor. The memory is used to store a database used in a rainfall-based reservoir scheduling method proposed in this solution, and the processor is used to execute the computer program to implement the steps of the rainfall-based reservoir scheduling method proposed in this solution.

[0035] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of a rainfall-based reservoir scheduling method proposed in this solution.

[0036] The beneficial effects of this invention are as follows: Historical scheduling schemes are stored in a database; the deterministic coefficients of each rainfall data point in the first partition stored in the database are calculated to obtain N deterministic coefficients, where N is an integer greater than 1; the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients is determined as the scheduling scheme corresponding to the forecast rainfall data; when determining the scheduling scheme, the scheduling scheme is directly determined from the database through rainfall data, without the need for complex algorithms, and the scheduling scheme can be determined through multiple calculations, thus making the determination process of the scheduling scheme simpler and faster, and by using historical scheduling schemes, the scheduling scheme of the current forecast rainfall is more in line with the actual situation. Attached Figure Description

[0037] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of the reservoir scheduling method according to the first embodiment of the present invention;

[0039] Figure 2 This is a line graph comparing rainfall data calculated according to the present invention;

[0040] Figure 3 This is a flowchart of the process for obtaining the current scheduling scheme according to the present invention;

[0041] Figure 4 This is a schematic diagram of the block storage of the scheduling scheme of the present invention;

[0042] Figure 5 This is a flowchart of the process for obtaining N rainfall data points according to the present invention;

[0043] Figure 6 This is a flowchart of the present invention for parsing N rainfall data points from rainfall images;

[0044] Figure 7 This is a schematic diagram of the image raster unit for storing rainfall data according to the present invention;

[0045] Figure 8 This is a flowchart of the reservoir scheduling method according to the second embodiment of the present invention;

[0046] Figure 9 This is a flowchart illustrating the storage of rainfall data according to the second embodiment of the present invention;

[0047] Figure 10 This is a schematic diagram of the reservoir scheduling device according to the third embodiment of the present invention;

[0048] Figure 11 This is a schematic diagram of the electronic device structure according to the fourth embodiment of the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this invention. Obviously, the described embodiments are only some embodiments of this invention, not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0050] To better understand the embodiments of the present invention, relevant concepts are described below. A runoff generation and confluence model is a type of hydrological model. The core content of reservoir runoff generation and confluence model research is the distribution and movement of water within a given area, between reservoirs, and along the watershed after rainfall. The runoff generation and confluence model reduces the rainfall-runoff formation process to two processes: runoff generation and runoff confluence. In this model, rainfall is generally used as the input condition, and flow rate is used as the output condition to obtain the runoff generation and runoff confluence of the watershed where the reservoir is located after a rainfall event. Runoff generation refers to the process of net rainfall after deducting losses; correspondingly, runoff volume refers to the portion of rainfall that forms runoff. Runoff confluence refers to the process of various components of runoff generated at various points along the watershed, flowing from slopes to streams, rivers, and finally to the watershed outlet. Watersheds can usually be divided into two basic parts: slopes and river networks. Therefore, watershed runoff confluence can also be divided into slope runoff confluence and river network runoff confluence. In this scheme, it mainly refers to river network runoff confluence.

[0051] In this plan, the rainfall data mainly includes the time of rainfall and the amount of rainfall at each time of rainfall.

[0052] Please see Figure 1 , Figure 1 This is a schematic flowchart of a reservoir scheduling method based on rainfall disclosed in an embodiment of the present invention. Figure 1 As shown, the rainfall-based reservoir scheduling method may include the following steps.

[0053] 100. Obtain the forecast rainfall data of the first partition, and calculate the deterministic coefficient of each rainfall data in the N rainfall data of the first partition stored in the database to obtain N deterministic coefficients.

[0054] The first partition is any partition of the reservoir, and N is an integer greater than 1.

[0055] 200. The scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients is determined as the scheduling scheme corresponding to the forecast rainfall data.

[0056] 300. The first partition is scheduled according to the determined scheduling scheme, which includes the flow generation of the first partition, the flow inflow of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow.

[0057] It should be noted that the water level threshold of the first zone is a preset reservoir water level threshold, which is an appropriate value that meets the reservoir's carrying capacity and the flow of irrigation water provided downstream by the reservoir, and can accept the flood discharge flow from upstream. This value can be changed according to the actual situation to adjust the reservoir scheduling plan.

[0058] The maximum controlled outflow refers to the difference between the flood storage volume in the reservoir and the reservoir water level threshold when rainfall occurs. When the flood storage volume in the reservoir exceeds the optimal reservoir water level threshold, the reservoir needs to release floodwater to bring the water level back to the optimal reservoir water level threshold. At this time, the reservoir releases floodwater according to the maximum controlled outflow flow rate. Alternatively, it can be the maximum flow rate that the reservoir can provide when downstream flow compensation is required.

[0059] The compensation flow refers to the flow rate required from upstream reservoirs or runoff when the water level in the reservoir is lower than the reservoir's water level threshold. The number of associated reservoirs refers to the number of reservoirs that can exchange flow with the reservoirs in the first zone and mutually influence each other's runoff.

[0060] Specifically, in obtaining the rainfall data P from the weather forecast... x Then, calculate the N precipitation data points stored in the first partition of the database and the N deterministic coefficients of the precipitation data. Select the precipitation data point with the largest deterministic coefficient among the N deterministic coefficients as the precipitation data P that is most similar to the predicted precipitation data. s The most similar rainfall data P was obtained. s The corresponding scheduling plan will be used as the scheduling plan for this rainfall event.

[0061] The formula for calculating N deterministic coefficients can be as follows:

[0062]

[0063] Wherein d(P x ,P i ) is P x With P i The coefficient of certainty; P x (j) represents actual or predicted rainfall data; P i (j) represents rainfall data from a specific rainfall event in the rainfall reservoir; P iais the average rainfall in a certain rainfall event in the rainfall database; m is the length of the rainfall data sequence, which refers to the time period included in the rainfall data.

[0064] For example, please see the appendix. Figure 2 , attached Figure 2 This is a line graph comparing three rainfall events (P1, P2, P3). Specific rainfall data is shown in Appendix 1, which presents the details of the three rainfall events. P1 corresponds to rainfall event 1, P2 to rainfall event 2, and P3 to rainfall event 3. Using these three rainfall events as an example, the calculation principle for the most similar rainfall event is illustrated. In this example, rainfall data P1 can be used as the rainfall data for this event. The process of selecting the rainfall data most similar to P1 is as follows: 1) Calculate the mean of rainfall data P1 based on the data in Appendix 1. 1a =5.28; 2) Calculate the coefficients of determination between P1 and P2, and between P1 and P3 according to the above formula. The coefficient of determination between P1 and P2 is d(P1,P2) = 0.56, and the coefficient of determination between P1 and P3 is d(P1,P3) = 0.13; 3) Compare the two sets of coefficients of determination. d(P1,P2) > d(P1,P3). Therefore, P2 is selected as the rainfall data that is more similar to P1.

[0065] Appendix 1: Schematic diagram of similar rainfall data calculation

[0066]

[0067]

[0068]

[0069]

[0070]

[0071]

[0072]

[0073] In the first embodiment, please refer to Figure 3 , Figure 3 This is a flowchart illustrating the process of obtaining the current scheduling scheme according to an embodiment of the present invention. Figure 3 Also includes:

[0074] 210. Obtain the current water level of the reservoir to get the first starting water level.

[0075] The step of determining the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data includes:

[0076] 220. The largest deterministic coefficient among the N deterministic coefficients and the scheduling scheme corresponding to the first starting water level are determined as the scheduling scheme corresponding to the forecast rainfall data.

[0077] It is worth mentioning that step 210 above can also be: set a water level threshold in advance to obtain the first adjustment water level.

[0078] Specifically, the maximum certainty coefficient is the correlation coefficient between rainfall data. In the database, N rainfall data points correspond to different scheduling schemes. The maximum certainty coefficient corresponds to one of the N rainfall data points, which is the most similar rainfall data point. The most similar rainfall data point corresponds to M scheduling schemes. When confirming the scheduling scheme for this rainfall event from the database, the M scheduling schemes corresponding to the most similar rainfall data are first selected using the maximum certainty coefficient. Then, a unique scheduling scheme is determined from the M scheduling schemes using the first starting water level. This unique scheduling scheme corresponds to the most similar rainfall data point and the first starting water level. This unique scheduling scheme is then used as the scheduling scheme for this forecast rainfall event.

[0079] In the first embodiment, taking the rainfall data and scheduling scheme of the first zone as an example, the establishment of the database includes:

[0080] The runoff generation and runoff concentration of the first zone are determined based on N rainfall data points from the first zone.

[0081] Determine M starting water levels, where M is an integer greater than 1.

[0082] Based on the N rainfall data, the runoff and runoff of the first zone, the M starting water levels, and a multi-objective optimization algorithm, N*M scheduling schemes for the first zone are determined, where the highest water level and maximum outflow of the reservoir corresponding to the scheduling scheme are minimized.

[0083] Store the N*M scheduling schemes of the first partition into the database.

[0084] In the above steps, N*M scheduling schemes for the first partition can also be determined based on N flow data, the flow generation and confluence of the first partition, the M starting water levels, and a multi-objective optimization algorithm; the N flow data correspond to the N rainfall data.

[0085] Specifically, N rainfall events of different magnitudes and typical processes collected from the first partition are mapped to N rainfall data points. These N rainfall data points, along with the runoff generation and confluence of the first partition and M initial reservoir states, are paired to form N*M scheduling scenarios. The M initial reservoir states correspond to M starting water levels, and each rainfall data point corresponds to a unique flow rate. For different scheduling scenarios, the minimum maximum water level and the minimum maximum outflow of the reservoir during the scheduling process are used as the scheduling optimization objectives. The corresponding reservoir scheduling schemes for each scenario are used as optimization parameters. Multi-objective optimization algorithms, such as Genetic Algorithm (GA) and Shuffled Complex Evolution Algorithm (SCE-UA), are used to calculate and obtain N*M scheduling schemes that minimize the scheduling objectives under different scheduling scenarios. These scheduling schemes, Proj(,L), are then stored in a database in blocks.

[0086] When optimizing N*M scheduling scenarios using a multi-objective optimization algorithm, the following formula is used to normalize the multi-objectives:

[0087]

[0088] Where nt is the number of targets, i = 1, 2, ..., nt; k i This refers to the conversion factor for each objective. Different objectives may have different weightings after normalization, so the conversion factor needs to be multiplied by the different objective values; t i These are target values ​​for different objectives. Furthermore, the optimization objectives are not limited to the highest water level or the maximum outflow during reservoir scheduling; the scheduling optimization objectives can be flexibly combined and set according to actual operational needs, thereby making the scheduling scheme better meet the actual operational needs of the reservoir.

[0089] Additionally, please see the appendix. Figure 4 , attached Figure 4 This is a schematic diagram of a block storage scheduling scheme in one embodiment of the present invention, according to the appendix. Figure 4 As shown, the method for storing the scheduling scheme in blocks in the database is as follows:

[0090] In the database, scheduling schemes are stored using a block-based storage approach. Index tables are created for the rainfall data P and the starting water level L corresponding to the N*M scheduling schemes, and a lookup table is created for the corresponding scheduling schemes. The database contains N rainfall data points and M starting water levels. Each rainfall data point is paired with one of the M starting water levels; that is, one rainfall data point corresponds to M scheduling schemes, and one rainfall data point and one starting water level correspond to one scheduling scheme. The M scheduling schemes are stored in a single address in the database. Therefore, N rainfall data points have N*M scheduling schemes, Proj(P,L), stored in N different storage addresses in the database. During queries, a block-based lookup approach enables fast searching of scheduling schemes for different scenarios. First, the most similar rainfall data is used to find the M scheduling schemes corresponding to the rainfall data. Then, the starting water level is used to search among the M scheduling schemes to find the unique scheduling scheme corresponding to both the rainfall data and the starting water level. This allows users to quickly find the scheduling scheme stored in the database using the two indexes of rainfall data and starting water level, thus avoiding complex calculations.

[0091] It should be noted that in the above scheme, the index can be set according to actual needs. For example, flow data and starting water level can be used as two indexes for retrieving scheduling schemes. One flow data corresponds to one rainfall data, that is, the most similar rainfall data also corresponds to the most similar flow data. By querying the most similar flow data, a unique scheduling scheme corresponding to the most similar rainfall data and the first starting water level can also be obtained.

[0092] In the first embodiment, determining the runoff generation and runoff concentration of the first zone based on N rainfall data points of the first zone includes:

[0093] The parameters of the runoff generation and confluence model are determined based on the historical rainfall and runoff data of the first zone, thus obtaining the first runoff generation and confluence model.

[0094] The runoff and runoff of the first partition are determined based on N rainfall data points from the first partition and the first runoff-confluence model.

[0095] Specifically, in practice, a region's hydrological model contains many unknown parameters. First, a hydrological model of a certain region is collected. Based on the measured rainfall and runoff data of the first sub-region, the unknown parameters of the hydrological model are calibrated to obtain a runoff generation and confluence model with determined parameters. The parameters include, but are not limited to, rainfall amount, rainfall time, and runoff size.

[0096] Using N rainfall data points as the rainfall conditions for the runoff generation and confluence model, the runoff generation and confluence P ~ Q = {Q} for different zones is calculated. i |i=0,1,2,…,n}, where the rainfall data P i With traffic data Qi One-to-one correspondence, the rainfall data P i This includes a sequence of rainfall data formed by rainfall amounts corresponding to different time zones. For example, if a rainfall data is divided into 24 time zones, then the rainfall data includes the rainfall amounts corresponding to these 24 time zones.

[0097] It should be noted that the above-described reservoir scheduling method based on rainfall is based on only one partition. When there are multiple partitions, the same example for one partition can be used. Correspondingly, the database should contain more than N rainfall data points and more than N*M scheduling schemes for multiple partitions. The storage method also uses partition-based storage of rainfall data and scheduling schemes. When there are multiple partitions, the corresponding index table can include partitions, rainfall data, and starting water levels, or it can include partitions, flow data, and starting water levels. Correspondingly, during a query, a query can first be performed based on the partition to obtain N*M scheduling schemes for one partition, and then a search can be performed using the same method as the aforementioned single-partition retrieval method.

[0098] In the first embodiment, please refer to the appendix. Figure 5 , attached Figure 5 A flowchart for obtaining N rainfall data points is attached. Figure 5 As shown, the method also includes:

[0099] 110. Obtain N images and corresponding feature values ​​of the first partition, wherein the images are rainfall images and the feature values ​​include the maximum rainfall, the minimum rainfall, and the batch average rainfall.

[0100] 120. Extract rainfall data from the N images based on the feature values ​​to obtain N rainfall data for the first partition.

[0101] This solution converts rainfall data into images and stores them in the database as images, reducing the storage space occupied by rainfall data; however, the storage method of rainfall data can be determined according to the actual situation, or it can be stored directly as data.

[0102] In the first embodiment, please refer to the appendix. Figure 6 and attached Figure 7 , Figure 6 This is a flowchart of the present invention for parsing N rainfall data points from rainfall images; Figure 7 This is a schematic diagram of the image raster unit for storing rainfall data according to the present invention.

[0103] As attached Figure 7 As shown, the image contains X grid units, where X is an integer greater than 1.

[0104] The step of extracting rainfall data from the N images based on the feature values ​​to obtain N rainfall data points for the first partition includes:

[0105] 121. Determine the pixel points and coordinate points of X grid units in the first image, wherein the first image is any one of the N images.

[0106] 122. Extract the RGB values ​​of the pixel.

[0107] 123. Calculate the rainfall corresponding to the grid cell based on the RGB value and the feature value to obtain X rainfall amounts.

[0108] 124. Determine N rainfall data for the first partition based on the coordinates of the X grid cells and the X rainfall amounts.

[0109] Specifically, before obtaining N images and corresponding feature values ​​for the first partition (step 110), historical rainfall data for different partitions are sorted according to the Pearson III curve (P-III), and the sorting results are used to establish a joint distribution based on the Gaussian Copula distribution function. This simulates the rainfall situation of each partition under different rainfall scenarios. Finally, the rainfall situation of the partitions is scaled according to historical typical rainfall to form N rainfall data P = {P} under different levels that conform to the regional distribution. i |i=0,1,2,…,n}.

[0110] Record the extreme values ​​and other characteristic values ​​of rainfall data. These characteristic values ​​include, but are not limited to, the maximum, minimum, and average rainfall amounts, and are calculated according to the following formula:

[0111]

[0112] The rainfall data, which includes the corresponding rainfall amount at each time, is normalized (minimum value is 0, maximum value is 1). Then, through data image storage technology, the rainfall data is stored as image raster units, and the feature values ​​are stored, thereby reducing the storage space occupied by the rainfall data.

[0113] Among them, P i P represents the rainfall value at time i in the rainfall sequence, where i = 0, 1, 2…; min P is the minimum value of the rainfall sequence. max This represents the maximum value of the rainfall sequence.

[0114] When converting rainfall data into images and storing them as images, you can convert rainfall data from a single rainfall event in one partition into one image, or you can convert rainfall data from the same rainfall event in multiple partitions into one image, or you can convert multiple rainfall events in one partition into one image. There are no restrictions on the combination of methods for converting rainfall data into images.

[0115] When using the database to extract scheduling schemes in practice, please refer to the appendix. Figure 7 , attached Figure 7This is a schematic diagram of the image raster unit for storing rainfall data according to the present invention, attached. Figure 7 A combined approach was adopted, which transformed the data of the same rainfall event from multiple regions onto a single map, as shown in the attached figure. Figure 7 As shown: Appendix Figure 7 The data shows a 144-hour rainfall event across four zones (data values ​​are shown in Appendix 1). A 144*4 pixel image is generated, where each raster cell contains 4*144 pixels. Each pixel corresponds to a coordinate value, representing the rainfall amount at a specific moment in the rainfall data. When calculating the coefficient of determination, taking pixel I (coordinates (0,5)) as an example, step 110 involves querying the database to obtain the feature values ​​P for this rainfall event. max =30.85,P min =0, and the rainfall data image, corresponding to step 122: obtain the RGB value of the pixel at this point as (249, 169, 0). Corresponding step: 123: calculate the rainfall at pixel I using the formula. Corresponding to step 124: Combining the coordinates (0, 5) of the pixel, the rainfall in the 5th time period of partition 1 in this rainfall data is 0.08 mm. Similarly, the rainfall at other times in the 144-hour rainfall data can be obtained through other pixels in the rainfall image, forming the rainfall data sequence, and obtaining the N rainfall data. Then, the correlation between the rainfall data sequence of all rainfall data and the rainfall data sequence of the forecast rainfall is calculated to obtain the most similar rainfall data.

[0116] Appendix 2: Schematic diagram of rainfall data for four zones

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124] In the first embodiment, determining the M starting water levels includes:

[0125] The M starting water levels are determined based on the maximum and minimum values ​​of the starting water level range of the first partition and a water level interval setting.

[0126] Specifically, the process of setting the M starting water levels is as follows: setting the reservoir starting water level range (l min ,l max Set water level interval d l According to the formula: Initial water level L = {L | L = l} min +d l *i; i = 0, 1, 2, ..., M; L <l max The M starting water levels are obtained, and the water level interval d l This is a value set according to actual needs.

[0127] To solve the above problems, this solution provides a second embodiment, with the remaining steps being the same as the first embodiment, except that: a threshold for adjusting the water level is set to obtain the first water level.

[0128] The step of determining the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data includes:

[0129] The largest deterministic coefficient among the N deterministic coefficients and the scheduling scheme corresponding to the first starting water level are determined as the scheduling scheme corresponding to the forecast rainfall data.

[0130] In addition, the second embodiment of this solution provides another method for reservoir scheduling based on rainfall.

[0131] See appendix Figure 8 , attached Figure 8 The flowchart of the reservoir scheduling method according to the second embodiment of the present invention is attached. Figure 8 As shown: A reservoir scheduling method based on rainfall includes:

[0132] 100. Obtain the forecast rainfall data of the first zone, and calculate the deterministic coefficient of each rainfall data in the N rainfall data of the first zone stored in the database to obtain N deterministic coefficients, where N is an integer greater than 1; the first zone is any zone of the reservoir.

[0133] 200. The scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients is determined as the scheduling scheme corresponding to the forecast rainfall data.

[0134] 300. The first partition is scheduled according to the determined scheduling scheme, which includes the flow generation of the first partition, the flow inflow of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow.

[0135] 400. Store the rainfall data and the scheduling plan for the current rainfall forecast in the database.

[0136] Steps 100, 200, and 300 are the same as those in the first embodiment described above; for a detailed explanation of steps 100, 200, and 300, please refer to the first embodiment.

[0137] The specific implementation method of step 400 is as follows: Please refer to the appendix for details. Figure 9 , attached Figure 9 To store the flowchart of this rainfall data,

[0138] Store the rainfall data and the scheduling plan for the current rainfall forecast in the database; including:

[0139] 410. Determine the runoff generation and runoff concentration of the first zone based on the rainfall data of the first zone;

[0140] Determine M starting water levels, where M is an integer greater than 1.

[0141] 420. Based on the rainfall data, the runoff and runoff of the first zone, the M starting water levels, and the multi-objective optimization algorithm, determine the M scheduling schemes for the first zone, wherein the highest water level and maximum outflow of the reservoir corresponding to the scheduling scheme are minimized.

[0142] 430. Store the M scheduling schemes of the first partition into the database.

[0143] Steps 410, 420, and 430 are based on the same principle as the process of establishing a database in the first embodiment. Please refer to the explanation of establishing a database in the first embodiment for further understanding.

[0144] After reservoir scheduling is carried out according to the determined scheduling plan, the forecasted rainfall data is converted into images according to the principles described in steps 121, 122, 123, and 124. The rainfall images and feature values ​​of the forecasted rainfall are stored in the database. Similar to the aforementioned N*M scheduling plan, based on the forecasted rainfall data, the runoff generation and runoff of the first zone, the M starting water levels, and a multi-objective optimization algorithm, M scheduling plans for the first zone corresponding to the forecasted rainfall data are determined. The scheduling plan corresponds to the minimum maximum water level and maximum outflow of the reservoir. The M scheduling plans are stored in the database, and the database is continuously updated and enriched to make it more complete and representative.

[0145] To address the above problems, the present invention provides a third embodiment, please refer to the appendix. Figure 10 A reservoir scheduling device based on rainfall, which is used to implement the steps of the reservoir scheduling method based on rainfall proposed in this scheme, includes: an acquisition unit 1001, a calculation unit 1002, a determination unit 1003, and a scheduling unit 1004.

[0146] The acquisition unit 1001 is used to acquire the forecast rainfall data of the first partition, which is any partition of the reservoir.

[0147] The calculation unit 1002 is used to calculate the deterministic coefficient of each rainfall data in the forecast rainfall data and the N rainfall data of the first partition stored in the database to obtain N deterministic coefficients, where N is an integer greater than 1.

[0148] The determining unit 1003 is used to determine the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data.

[0149] The scheduling unit 1004 is used to schedule the first partition according to a determined scheduling scheme, the scheduling scheme including the flow generation of the first partition, the flow inflow of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow.

[0150] The acquisition unit 1001 is also used to acquire the current water level of the reservoir and obtain the first starting water level.

[0151] Correspondingly, the determining unit 1003 is also used to determine the largest deterministic coefficient among the N deterministic coefficients and the scheduling scheme corresponding to the first starting water level as the scheduling scheme corresponding to the forecast rainfall data.

[0152] The determining unit 1003 is further configured to determine the runoff generation and runoff confluence of the first partition based on N rainfall data of the first partition; and to determine M starting water levels, where M is an integer greater than 1.

[0153] The calculation unit 1002 is also used to determine N*M scheduling schemes for the first partition based on the N rainfall data, the runoff and runoff of the first partition, the M starting water levels and the multi-objective optimization algorithm, wherein the highest water level and maximum outflow of the reservoir corresponding to the scheduling scheme are minimized.

[0154] The calculation unit 1002 is further configured to determine the parameters of the runoff generation and confluence model based on the historical rainfall data and historical runoff data of the first zone, thereby obtaining the first runoff generation and confluence model; and to determine the runoff generation and confluence of the first zone based on N rainfall data of the first zone and the first runoff generation and confluence model.

[0155] In the third embodiment, the acquisition unit 1001 is further configured to acquire N images of the first partition and corresponding feature values, and to extract rainfall data from the N images according to the feature values ​​to obtain N rainfall data of the first partition, wherein the images are rainfall images, and the feature values ​​include maximum rainfall, minimum rainfall, and batch average rainfall.

[0156] The acquisition unit 1001 is further configured to determine the pixel points and coordinate points of X grid units in the first image, and extract the RGB values ​​of the pixel points; wherein, the first image is any one of the N images.

[0157] The calculation unit 1002 is also used to calculate the rainfall corresponding to the grid unit based on the RGB value and the feature value, so as to obtain X rainfall amounts.

[0158] The determining unit 1004 is further configured to determine N rainfall data for the first partition based on the coordinate points of the X grid cells and the X rainfall amounts.

[0159] In the third embodiment, the calculation unit is further configured to determine the M starting water levels based on the maximum and minimum values ​​of the starting water level range of the first partition and a water level interval setting.

[0160] In the third embodiment, the calculation unit 1002 is further configured to determine the runoff generation and runoff confluence of the first partition based on the current rainfall data of the first partition; and to determine M starting water levels, where M is an integer greater than 1.

[0161] The calculation unit 1002 is also used to determine M scheduling schemes for the first partition based on the rainfall data, the runoff generation and runoff of the first partition, the M starting water levels and the multi-objective optimization algorithm, wherein the highest water level and maximum outflow of the reservoir corresponding to the scheduling scheme are minimized.

[0162] The third embodiment provides a rainfall-based reservoir scheduling device for executing the corresponding steps of the rainfall-based reservoir scheduling method provided in the first embodiment. Therefore, for a detailed explanation of the relevant functional implementation of the device in the third embodiment, please refer to the explanation of the corresponding steps of the method in the first embodiment.

[0163] In the third embodiment, the reservoir scheduling device based on rainfall should also include a storage unit for storing a database containing N*M scheduling schemes and N rainfall data. The storage unit is also used to store the M scheduling schemes for the current rainfall in the first partition into the database.

[0164] To address the above problems, the present invention provides a fourth embodiment, an electronic device, please refer to the appendix. Figure 11 It includes a memory 2001 and a processor 2002. The memory is used to store the database used in the computer program proposed in this solution, which is a rainfall-based reservoir scheduling method. The processor is used to execute the computer program to implement the steps of the rainfall-based reservoir scheduling method proposed in this solution.

[0165] To address the aforementioned problems, the present invention provides a fifth embodiment: a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of the rainfall-based reservoir scheduling method proposed in this solution.

[0166] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A reservoir scheduling method based on rainfall, characterized in that, include: Obtain the forecast rainfall data for the first zone, which is any zone of the reservoir; The deterministic coefficients of the predicted rainfall data and the N rainfall data of the first partition stored in the database are calculated to obtain N deterministic coefficients, where N is an integer greater than 1; The scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients is determined as the scheduling scheme corresponding to the forecast rainfall data; The first partition is scheduled according to the determined scheduling scheme, which includes the runoff generation of the first partition, the runoff collection of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow. Also includes: Obtain the current water level of the reservoir to determine the initial water level for the first water adjustment. The step of determining the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data includes: The largest deterministic coefficient among the N deterministic coefficients and the scheduling scheme corresponding to the first starting water level are determined as the scheduling scheme corresponding to the forecast rainfall data. Also includes: The runoff generation and runoff concentration of the first zone are determined based on N rainfall data points from the first zone. Determine M starting water levels, where M is an integer greater than 1; Based on the N rainfall data, the runoff and runoff of the first zone, the M starting water levels, and a multi-objective optimization algorithm, N*M scheduling schemes for the first zone are determined, where the highest water level and maximum outflow of the reservoir corresponding to the scheduling scheme are minimized. Store the N*M scheduling schemes of the first partition into the database; In the database, scheduling schemes are stored in blocks. Index tables are created by indexing the rainfall data P and the starting water level L corresponding to N*M scheduling schemes, and a lookup table is created for the corresponding scheduling schemes. The database contains N rainfall data points and M starting water levels. Each rainfall data point is paired with one of the M starting water levels. One rainfall data point corresponds to M scheduling schemes, and one rainfall data point and one starting water level correspond to one scheduling scheme. The M scheduling schemes are stored in one address in the database. N rainfall data points have N*M scheduling schemes, Proj(P,L), stored in N storage addresses in the database. During queries, a block-based search method enables fast lookup of scheduling schemes for different scenarios. First, the most similar rainfall data is used to find the M scheduling schemes corresponding to the rainfall data. Then, the starting water level is used to search among the M scheduling schemes to find the unique scheduling scheme corresponding to both the rainfall data and the starting water level. This allows users to quickly find the scheduling schemes stored in the database based on the two indexes: rainfall data and starting water level.

2. The method according to claim 1, characterized in that, The step of determining the runoff generation and runoff concentration of the first zone based on N rainfall data points of the first zone includes: The parameters of the runoff generation and confluence model are determined based on the historical rainfall and runoff data of the first zone, thus obtaining the first runoff generation and confluence model; The runoff and runoff of the first partition are determined based on N rainfall data points from the first partition and the first runoff-confluence model.

3. The method according to claim 1, characterized in that, Also includes: Obtain N images and their corresponding feature values ​​from the first partition. The images are rainfall images, and the feature values ​​include the maximum rainfall, the minimum rainfall, and the batch average rainfall. Rainfall data is extracted from the N images based on the feature values ​​to obtain N rainfall data points for the first partition.

4. The method according to claim 3, characterized in that, The image contains X grid cells, where X is an integer greater than 1; The step of extracting rainfall data from the N images based on the feature values ​​to obtain N rainfall data points for the first partition includes: Determine the pixel points and coordinate points of X grid units in the first image, where the first image is any one of the N images; Extract the RGB values ​​of the pixels; The rainfall amount corresponding to the grid cell is calculated based on the RGB value and the feature value to obtain X rainfall amounts; N rainfall data points for the first partition are determined based on the coordinates of the X grid cells and the X rainfall amounts.

5. The method according to claim 1, characterized in that, Determining the M initial water levels includes: The M starting water levels are determined based on the maximum and minimum values ​​of the starting water level range of the first partition and a water level interval setting.

6. A reservoir scheduling device based on rainfall, characterized in that, include: The acquisition unit is used to acquire the forecast rainfall data of the first partition, which is any partition of the reservoir; The calculation unit is used to calculate the deterministic coefficient of each rainfall data in the forecast rainfall data and the N rainfall data of the first partition stored in the database to obtain N deterministic coefficients, where N is an integer greater than 1; The determining unit is used to determine the scheduling scheme corresponding to the largest deterministic coefficient among the N deterministic coefficients as the scheduling scheme corresponding to the forecast rainfall data; The scheduling unit is used to schedule the first partition according to a determined scheduling scheme, which includes the flow generation of the first partition, the flow inflow of the first partition, the water level threshold of the first partition, the maximum controlled outflow, the number of associated reservoirs, and the compensation flow. The database stores scheduling schemes in a block-based manner: Index tables are created for the rainfall data P and the starting water level L corresponding to N*M scheduling schemes, and a lookup table is created for the corresponding scheduling schemes. The N*M scheduling schemes are stored in blocks within the database, which contains N rainfall data points and M starting water levels. Each rainfall data point is paired with one of the M starting water levels, resulting in M ​​scheduling schemes for each rainfall data point. A rainfall data point and a starting water level correspond to one scheduling scheme. These M scheduling schemes are stored in a single address within the database. Therefore, N rainfall data points result in N*M scheduling schemes, each named Proj(P,L), stored in N different database addresses. During queries, a block-based lookup approach enables rapid retrieval of scheduling schemes for different scenarios. First, the most similar rainfall data is used to find the M scheduling schemes corresponding to the rainfall data. Then, the starting water level is used to retrieve the unique scheduling scheme corresponding to both the rainfall data and the starting water level from the M scheduling schemes. This allows users to quickly retrieve the scheduling schemes stored in the database based on the two indexes: rainfall data and starting water level. This method is used to implement the rainfall-based reservoir scheduling method as described in claim 1.

7. An electronic device comprising a memory and a processor, characterized in that, The memory is used to store a computer program and the database as described in any one of claims 1-5, and the processor is used to execute the steps of the computer program to implement the method as described in any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by a processor to implement the steps of the method as described in any one of claims 1-5.

Images