Water level prediction method and device for water collecting area of pumped storage power station

Abstract

The application provides a water level prediction method and device for a water collection area of a pumped storage power station, and the method comprises the following steps: correcting rainfall prediction data to obtain rainfall input data; making a prediction based on the rainfall input data and basin water level characteristic parameters to obtain an initial water level prediction value; constructing a residual prediction model based on the residual between the measured monitoring water level value and the initial water level prediction value, determining boundary correction data of a monitoring blank period based on the initial water level prediction value of an effective monitoring period adjacent to the monitoring blank period and the residual prediction model; simulating the water level of the monitoring blank period based on the boundary correction data and the rainfall input data to obtain target water level data of the monitoring blank period; constructing a full-period continuous water level sequence based on the measured monitoring water level data and the target water level data of the monitoring blank period, making a water level prediction based on the full-period continuous water level sequence to obtain a water level prediction result; and making a water level warning according to the comparison result of the water level prediction result and a preset warning threshold.

Description

Technical Field

[0001] This invention relates to the field of hydrological forecasting and early warning technology, and in particular to a method and device for forecasting water levels in the catchment area of ​​a pumped storage power station. Background Technology

[0002] Pumped storage power stations are mostly located in small watersheds in mountainous areas. Their safe operation relies heavily on real-time monitoring and accurate forecasting of water levels and flow rates. Hydrological monitoring of the catchment area where the pumped storage power station is located is the core foundation for flood control scheduling and unit operation. However, since most pumped storage power stations are located in remote mountainous areas, hydrological observation of the catchment area faces several challenges. First, manual observation periods are limited, resulting in data gaps during non-observation periods and the potential to miss extreme values ​​such as flood peaks and maximum water levels. Second, the geographical constraints of pumped storage power stations also limit the deployment and maintenance of online monitoring equipment. Therefore, monitoring gaps are unavoidable in many cases, posing a challenge to the accuracy of water level forecasting.

[0003] In existing technologies, data filling for monitoring gaps often relies on simple interpolation methods, without constructing refined hydrological simulation models. This makes it impossible to accurately recreate the complete hydrological dynamic process of rainfall-runoff-confluence. At the same time, existing technologies do not comprehensively consider the multiple limitations of pumped storage power stations in small watersheds, such as engineering layout, operation and maintenance, and short-term construction economy. They separate manual observation from online monitoring, failing to form a complementary and compatible technical system. As a result, the problem of missing hydrological data during monitoring gaps remains unresolved, and the accuracy of water level forecasts is insufficient to meet the actual needs of power stations. Summary of the Invention

[0004] This invention provides a method and apparatus for predicting water levels in the catchment area of ​​a pumped storage power station, which solves the problem that data filling for monitoring gaps in the prior art does not involve the complete hydrological dynamic process, thus leading to insufficient accuracy in hydrological forecasting.

[0005] This invention provides a method for water level forecasting in the catchment area of ​​a pumped storage power station, comprising: acquiring and correcting rainfall forecast data to obtain rainfall input data; performing preliminary water level forecasting based on the rainfall input data and watershed water level characteristic parameters to obtain an initial water level forecast value; constructing a residual prediction model based on the residual sequence between the measured monitored water level value and the initial water level forecast value, and determining the boundary correction data of the monitoring blank period based on the initial water level forecast value of the effective monitoring period adjacent to the monitoring blank period and the residual prediction model; performing water level simulation on the monitoring blank period based on the boundary correction data and the rainfall input data to obtain the target water level data of the monitoring blank period; constructing a continuous water level sequence for the entire time period based on the monitored water level data of the effective monitoring period and the target water level data of the monitoring blank period, and performing water level forecasting based on the continuous water level sequence for the entire time period to obtain a water level forecast result; and issuing a water level warning based on the comparison result of the water level forecast result and a preset warning threshold.

[0006] According to the method for predicting water level in the catchment area of ​​a pumped storage power station provided by the present invention, the step of performing water level simulation on the monitoring blank period based on the boundary correction data and the rainfall input data to obtain target water level data includes: using the boundary correction data as the initial boundary condition for the monitoring blank period and the rainfall input data as the hydrological driving condition; performing runoff simulation based on the rainfall input data to obtain the runoff volume per time period; performing confluence simulation based on the runoff volume per time period to obtain the inflow volume per time period at a preset control section; and converting the inflow volume per time period into water level data per time period according to the water level-flow relationship of the preset control section to obtain the target water level data.

[0007] According to the method for predicting water level in the catchment area of ​​a pumped storage power station provided by the present invention, the step of performing runoff simulation based on the rainfall input data to obtain the runoff volume for each time period includes: calculating the runoff volume using a full-storage runoff model when the runoff parameters of the catchment area meet a first preset condition; and calculating the runoff volume using an over-permeability runoff model when the runoff parameters of the catchment area meet a second preset condition.

[0008] According to the method for predicting water level in the catchment area of ​​a pumped storage power station provided by the present invention, the step of converting the inflow rate over time period into water level data over time period based on the water level-flow relationship of the preset control section to determine the target water level data includes: converting the inflow rate over time period into a corresponding preliminary water level sequence based on the water level-flow relationship of the preset control section, and determining the water level change rate over time period based on the preliminary water level sequence; determining a time-series extension method based on the rainfall input data over time period and the water level change rate over time period, and extending the preliminary water level sequence over time period based on the time-series extension method to obtain time-series extended water level data over time period; and determining the target water level data based on the time-series extended water level data over time period and the preliminary water level sequence.

[0009] According to the method for predicting water levels in the catchment area of ​​a pumped storage power station provided by the present invention, the step of determining the time-series extension method based on the rainfall input data and the water level change rate for each time period includes: determining the time-series extension method as linear fitting when the rainfall intensity corresponding to the rainfall input data for each time period is less than or equal to a preset rainfall intensity threshold, and the water level change rate for each time period is less than or equal to a preset water level change rate threshold; and determining the time-series extension method as cubic spline interpolation when the rainfall intensity corresponding to the rainfall input data for each time period is greater than the preset rainfall intensity threshold, or the water level change rate for each time period is greater than the preset water level change rate threshold.

[0010] According to the method for predicting water level in the catchment area of ​​a pumped storage power station provided by the present invention, the step of determining the target water level data based on the time-series extended water level data and the preliminary water level sequence includes: determining the weights of the time-series extended water level data and the preliminary water level sequence based on the time-series rainfall input data; and determining the target water level data based on the time-series extended water level data, the preliminary water level sequence, and the corresponding weights.

[0011] According to the method for predicting water level in the catchment area of ​​a pumped storage power station provided by the present invention, the step of performing water level simulation on the monitoring blank period based on the boundary correction data and the rainfall input data to obtain the target water level data for the monitoring blank period further includes: selecting historical observation data with continuous water level and flow observation results as verification data; dividing the known period and the period to be verified based on the historical observation data; using the actual data of the known period as input to perform water level simulation on the period to be verified to obtain the target water level data corresponding to the period to be verified; comparing the target water level data corresponding to the period to be verified with the actual observation data corresponding to the period to be verified on a time-by-time basis to obtain the comparison result; and iteratively optimizing the water level simulation parameters used for the water level simulation based on the comparison result.

[0012] This invention also provides a water level forecasting device for the catchment area of ​​a pumped storage power station, comprising: a rainfall correction module for acquiring and correcting rainfall forecast data to obtain rainfall input data; an initial water level forecasting module for performing preliminary water level forecasting based on the rainfall input data and watershed water level characteristic parameters to obtain an initial water level forecast value; and a boundary correction module for constructing a residual prediction model based on the residual sequence between the measured monitored water level value and the initial water level forecast value, and determining the initial water level forecast value based on the initial water level forecast value of the effective monitoring period adjacent to the monitoring blank period and the residual prediction model. The system includes: a boundary correction data monitoring blank period; a water level simulation module for simulating water levels during the monitoring blank period based on the boundary correction data and the rainfall input data, to obtain target water level data for the monitoring blank period; an optimization module for constructing a continuous water level sequence for the entire time period based on the monitoring water level data of the effective monitoring period and the target water level data of the monitoring blank period, and for forecasting water levels based on the continuous water level sequence for the entire time period, to obtain water level forecast results; and a forecasting module for issuing water level warnings based on the comparison between the water level forecast results and a preset warning threshold.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the water level forecasting method for the catchment area of ​​a pumped storage power station as described above.

[0014] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the water level forecasting method for the catchment area of ​​a pumped storage power station as described above.

[0015] The present invention provides a method and apparatus for water level forecasting in the catchment area of ​​a pumped storage power station. First, it acquires rainfall forecast data and actual rainfall monitoring data, and corrects the rainfall forecast data based on the actual rainfall monitoring data to obtain rainfall input data. Then, it performs preliminary water level forecasting based on the rainfall input data and watershed water level characteristic parameters to obtain initial water level forecast values ​​for each time period. Based on the residual sequence between the measured monitoring water level values ​​and the initial water level forecast values, it determines boundary correction data for the monitoring blank period, which serves as the initial boundary condition for water level simulation during the monitoring blank period. Further, it performs water level simulation for the monitoring blank period based on the boundary correction data and rainfall input data to obtain target water level data for the monitoring blank period. Finally, based on the monitoring water level data of the effective monitoring period and the target water level data of the monitoring blank period, it constructs a continuous water level sequence for the entire time period, and performs water level forecasting based on the continuous water level sequence to obtain the water level forecast result. Compared to existing technologies that rely solely on simple interpolation to fill data gaps before forecasting, this invention combines boundary correction and water level simulation to continuously fill in monitoring gaps, enabling water level forecasts to be based on continuous water level conditions, thereby improving the continuity and accuracy of water level forecasts. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the method for predicting water level in the catchment area of ​​a pumped storage power station according to the present invention; Figure 2 The flowchart illustrates an example of the present invention of performing water level simulation on a monitoring blank period based on boundary correction data and rainfall input data to obtain target water level data; Figure 3 The flowchart illustrates an example of the present invention, which converts the inflow rate over time period into water level data over time period based on the water level-flow relationship of a preset control section in order to determine the target water level data. Figure 4 The flowchart illustrates an example of the present invention, which uses boundary correction data and rainfall input data to simulate water levels during a monitoring blank period and obtain target water level data for the monitoring blank period. Figure 5 A structural block diagram of a pumped storage power station catchment area water level prediction device according to an embodiment of the present invention is shown; Figure 6This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0018] 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. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0019] Because water level monitoring in the catchment area of ​​pumped storage power stations often has monitoring gaps, and these gaps may contain extreme values ​​such as flood peaks and highest water levels that are related to flood control scheduling and unit operation of pumped storage power stations, attention needs to be paid to the monitoring gaps in water level prediction of the catchment area of ​​pumped storage power stations. However, in the existing technology, simple interpolation methods are often used to fill in the data, which leads to insufficient prediction accuracy in the overall water level prediction process.

[0020] In view of this, the present invention provides a method for predicting the water level in the catchment area of ​​a pumped storage power station.

[0021] Figure 1 This is a flowchart of the method for predicting the water level in the catchment area of ​​a pumped storage power station according to the present invention.

[0022] like Figure 1 As shown, the method includes operations S110~S160.

[0023] In operation S110, rainfall forecast data is acquired and corrected to obtain rainfall input data.

[0024] According to an embodiment of the present invention, rainfall forecast data can be obtained from the numerical rainfall forecast products issued by the meteorological department for a future period of time; the correction of rainfall forecast data can be carried out by adjusting the residual between the actual rainfall data measured by the rain gauge and the rainfall forecast data on a time-by-time basis to eliminate the systematic error of the forecast, thereby obtaining rainfall input data and providing relatively accurate rainfall input data for subsequent monitoring of water level simulation during blank periods.

[0025] In operation S120, preliminary water level forecasts are made based on rainfall input data and watershed water level characteristic parameters to obtain initial water level forecast values.

[0026] According to embodiments of the present invention, watershed hydrological data may include runoff generation coefficient, soil moisture content, runoff time, etc.

[0027] According to an embodiment of the present invention, rainfall input data can be input into a preliminary hydrological forecasting model (such as the Xin'anjiang model, TOPMODEL, etc.) so that the preliminary hydrological forecasting model can simulate the water level changes over time based on the watershed hydrological data to obtain the initial water level forecast values ​​for each time period. Furthermore, through the aforementioned preliminary hydrological forecasting model, the mapping relationship between rainfall and water level in the preset control section of the catchment area can also be clearly understood.

[0028] In operation S130, a residual prediction model is constructed based on the residual sequence between the measured monitored water level value and the initial water level forecast value. Based on the initial water level forecast value of the effective monitoring period adjacent to the monitoring blank period and the residual prediction model, the boundary correction data of the monitoring blank period is determined.

[0029] According to an embodiment of the present invention, after obtaining the initial water level forecast value, the initial water level forecast value is further corrected by using the residual between the measured monitoring water level value and the initial water level forecast value.

[0030] Specifically, within a historical period with actual monitored water level values, the initial water level forecast is compared with the measured water level values ​​for the same period to obtain the forecast residuals for each period. The residual for a given period can be expressed as:

[0031] in, Indicates time The measured water level value, Indicates time The initial water level forecast value.

[0032] Then, a residual prediction model is constructed based on the residual sequences from multiple historical time periods. The residual prediction model can employ an autoregressive moving average model to characterize the temporal variation of water level forecast errors and predict the forecast residuals at the current or target time. Based on the residual prediction model, the predicted residuals at the corresponding time can be obtained. .

[0033] Furthermore, the predicted residuals will be... The initial water level forecast value at the corresponding time Error compensation is performed to obtain the corrected water level data:

[0034] in, This indicates the corrected water level data.

[0035] Since the monitoring blank period lacks actual monitored water level values, it is impossible to directly perform residual correction within this period. Therefore, the aforementioned error compensation can be prioritized for the effective monitoring period adjacent to the monitoring blank period, thereby obtaining the boundary correction data corresponding to the starting boundary of the monitoring blank period. The boundary correction data can be used as the initial boundary conditions for subsequent water level simulations during the monitoring blank period, thereby improving the accuracy of subsequent runoff generation and confluence simulations during the monitoring blank period.

[0036] In operation S140, based on boundary correction data and rainfall input data, water level simulation is performed for the monitoring blank period to obtain the target water level data for the monitoring blank period.

[0037] According to an embodiment of the present invention, the boundary correction data is used as the initial boundary condition for the water level in the water level simulation during the monitoring blank period. Based on this, runoff generation simulation and confluence simulation are performed sequentially based on the rainfall input data to obtain the target water level data during the monitoring blank period.

[0038] In operation S150, based on the monitored water level data during the effective monitoring period and the target water level data during the monitoring blank period, a continuous water level sequence for the entire time period is constructed, and water level forecasting is performed based on the continuous water level sequence for the entire time period to obtain the water level forecasting results.

[0039] According to an embodiment of the present invention, the monitoring water level data of the effective monitoring period and the target water level data of the monitoring blank period can be connected in chronological order to form a continuous water level sequence of the control section throughout the entire time period. Then, the continuous water level sequence throughout the entire time period is used as the basic characterization of the current water level status of the basin, and combined with the rainfall forecast input of the future period, the water level change process of the future period is predicted and simulated to obtain the water level forecast result of the corresponding future period.

[0040] In operation S160, a water level warning is issued based on the comparison between the water level forecast result and the preset warning threshold.

[0041] Specifically, multiple water level warning thresholds at different levels can be set by combining the operational safety standards of pumped storage power stations, the flood control capacity of the basin, and historical hydrological extreme value data. After obtaining the water level forecast results for future periods, the water level forecast results at each time point are compared with the corresponding preset warning thresholds. When the water level forecast result reaches or exceeds a certain preset warning threshold, the corresponding level of warning signal is triggered. In this way, water level risks in future periods can be identified in advance based on water level forecast results, thereby providing early warning support for the flood control scheduling and operational safety of the power station.

[0042] By employing the aforementioned setup, a more reliable rainfall input is first obtained based on rainfall forecast data and actual rainfall monitoring data. Then, combined with the basin's water level characteristic parameters, initial water level forecast values ​​for each time period are obtained. The residual sequence between the measured monitoring water level values ​​and the initial water level forecast values ​​is used to correct the adjacent boundaries of the monitoring gap periods, thereby improving the accuracy of the initial state of the monitoring gap periods. Furthermore, water level simulation is performed on the monitoring gap periods based on boundary correction data and rainfall input data to obtain target water level data for the monitoring gap periods. This is then combined with the monitoring water level data from the effective monitoring periods to construct a continuous water level sequence for the entire time period. Water level forecasting and early warning are then performed based on this sequence. Compared to existing technologies that rely solely on simple interpolation to fill in the gap data before forecasting, this invention simulates and fills in the monitoring gap periods, ensuring the continuous and predictable water level status throughout the entire time period. This improves the continuity, stability, and accuracy of water level forecasting, making it more suitable for all-time forecasting and early warning applications in scenarios where monitoring gaps exist in the catchment areas of pumped storage power stations.

[0043] Figure 2 The flowchart illustrates an example of the present invention of performing water level simulation on a monitoring blank period based on boundary correction data and rainfall input data to obtain target water level data.

[0044] like Figure 2 As shown, operation S130 includes operations S210~S240.

[0045] In operation S210, boundary correction data is used as the initial boundary condition for the monitoring blank period, and rainfall input data is used as the hydrological driving condition.

[0046] In embodiments of the present invention, since reliable actual water level data is lacking within the monitoring blank period, subsequent water level simulations need to determine the simulation starting point based on the water level status corresponding to the effective monitoring period adjacent to the monitoring blank period. Boundary correction data is obtained based on the initial water level forecast value and after error compensation using a residual prediction model. Compared to the uncorrected initial water level forecast value, it is closer to the actual water level status of the control section at the start of the simulation. Therefore, using boundary correction data as the starting boundary condition for the monitoring blank period can improve the accuracy of the simulation starting point, thereby reducing the accumulated error in the subsequent time-by-time water level simulation process and improving the reliability of the target water level data for the monitoring blank period.

[0047] When operating S220, runoff simulation is performed based on rainfall input data to obtain the runoff volume for each time period.

[0048] According to an embodiment of the present invention, a coupled model of full-saturation runoff and infiltration runoff can be used to simulate the runoff process in a watershed. By combining parameters such as the runoff coefficient of the catchment area, soil moisture content, and land use type, the runoff volume during the monitoring blank period can be calculated time by time.

[0049] In operation S230, a flow simulation is performed based on the time-period production flow to obtain the time-period inflow flow of the preset control section.

[0050] According to an embodiment of the present invention, the process of confluence simulation can be divided into two stages: slope confluence simulation and river confluence simulation, which sequentially simulate the entire process of the flow from the slope into the river and from the upstream river into the downstream control section.

[0051] For example, the process of simulating slope runoff can be combined with the following parameters of the catchment area: the slope of the watershed, the roughness of the watershed, the type of land cover, etc., to calculate the slope runoff velocity and runoff time, and convert the time-by-time runoff obtained from the runoff process simulation into the time-by-time inflow of the preset control section.

[0052] Preferably, the river confluence simulation can be based on the Muskingu method for river confluence calculation.

[0053] In operation S240, the inflow rate for each time period is converted into water level data for each time period according to the water level-flow relationship of the preset control section in order to obtain the target water level data.

[0054] According to embodiments of the present invention, a water level-flow relationship for a preset control section can be established in advance based on historical measured water level data and historical measured flow data. The water level-flow relationship is used to characterize the correspondence between flow rate changes and water level changes at the preset control section.

[0055] After obtaining the inflow rate of the preset control section for each time period, the inflow rate of each time period can be input into the water level-flow relationship to determine the water level value corresponding to the inflow rate of each time period, thereby obtaining the water level data of the preset control section for each time period, so as to obtain the target water level data for the monitoring blank period.

[0056] In one illustrative embodiment, runoff simulation is performed based on rainfall input data to obtain time-period runoff volumes including: If the runoff parameters in the catchment area meet the first preset conditions, the runoff volume is calculated using the full-capacity runoff model.

[0057] When the runoff parameters in the catchment area meet the second preset condition, the runoff volume is calculated using the super-permeable runoff model.

[0058] According to embodiments of the present invention, runoff parameters may include parameters such as runoff coefficient, soil moisture content and land use type, field water holding capacity, soil infiltration capacity and initial loss.

[0059] According to an embodiment of the present invention, when the soil moisture content in the catchment area reaches the field capacity, the runoff parameters can be considered to meet the first preset condition. At this time, the full runoff model is used to calculate the runoff exceeding the infiltration capacity during rainfall.

[0060] According to an embodiment of the present invention, when the soil moisture content in the catchment area does not reach the field capacity and the rainfall intensity is greater than the soil infiltration capacity, the runoff parameters can be considered to meet the second preset condition. At this time, the super-infiltration runoff model is used to calculate the difference between the actual infiltration amount and the rainfall amount as the runoff volume.

[0061] Furthermore, if the soil moisture content does not reach the field capacity and the rainfall intensity is not greater than the soil infiltration capacity, it can be determined that there will be no runoff in the current period, and the runoff volume can be determined as zero.

[0062] By using the above methods, appropriate runoff generation models can be rationally selected based on the different soil water storage conditions and rainfall infiltration relationships in the catchment area, thereby improving the accuracy of time-period runoff generation calculation results.

[0063] Figure 3 The flowchart illustrates an example of the present invention, which converts the inflow rate over time into water level data over time according to the water level-flow relationship of a preset control section in order to determine the target water level data.

[0064] like Figure 3 As shown, operation S240 includes operations S310 to S330.

[0065] In operation S310, the inflow rate for each time period is converted into a corresponding preliminary water level sequence based on the water level-flow relationship of the preset control section, and the water level change rate for each time period is determined based on the preliminary water level sequence.

[0066] According to an embodiment of the present invention, the inflow rate of each time period is converted into water level data for each time period based on the water level-flow relationship of a preset control section, and then the water level data of each time period are spliced ​​together in chronological order to obtain a preliminary water level sequence.

[0067] In operation S320, the time series extension method is determined based on the rainfall input data and the water level change rate for each time period, and the preliminary water level sequence is extended according to the time series extension method to obtain the time series extended water level data for each time period.

[0068] According to an embodiment of the present invention, since the rate of water level change varies due to different rainfall input data in different time periods, using only a single time-series extension method may result in a significant difference between the final time-series extended water level data and the actual situation. Therefore, it is necessary to determine the time-series extension method based on the rainfall input data and the rate of water level change in each time period to ensure the accuracy of time-series extension under different conditions.

[0069] In operation S330, the target water level data is determined based on the time-series extended water level data and the preliminary water level sequence.

[0070] According to an embodiment of the present invention, the weights of the time-series extended water level data and the weights of the preliminary water level sequence can be determined based on the time-series rainfall input data. The two weights are then fused to obtain the target water level data. This approach preserves the physical simulation capability of the preliminary water level sequence for actual hydrological processes while utilizing the time-series extended water level data to smooth and correct water level change trends. Consequently, the final target water level data achieves both physical plausibility and temporal continuity across different rainfall scenarios.

[0071] In one illustrative embodiment, determining the time-series extension method based on time-period rainfall input data and water level change rate includes: If the rainfall intensity corresponding to the rainfall input data for each time period is less than or equal to the preset rainfall intensity threshold, and the water level change rate for each time period is less than or equal to the preset water level change rate threshold, the time series extension method is determined to be linear fitting.

[0072] If the rainfall intensity corresponding to the rainfall input data for each time period is greater than the preset rainfall intensity threshold, or the water level change rate for each time period is greater than the preset water level change rate threshold, the time series extension method is determined to be cubic spline interpolation.

[0073] In one illustrative embodiment, the target water level data is determined based on time-series extended water level data and a preliminary water level sequence, including: Based on the rainfall input data for each time period, the weights of the time-series extended water level data and the weights of the preliminary water level sequence are determined; The target water level data is determined based on the extended time-series water level data, the preliminary water level sequence, and the corresponding weights.

[0074] According to an embodiment of the present invention, when the rainfall intensity is low, it indicates that the water level change process is weakly driven by short-term heavy rainfall and the time series change is relatively stable. At this time, the contribution of the time series extension result to the target water level data can be relatively large. When the rainfall intensity is high, it indicates that the water level change process is more significantly driven by rainfall. The preliminary water level sequence obtained based on the coupling of runoff generation, confluence and water level-discharge can better reflect the actual hydrological process. Therefore, the weight corresponding to the preliminary water level sequence can be relatively increased.

[0075] For example, the weights of the time-series stretching results Data can be input based on rainfall. I Determine as follows:

[0076] Accordingly, the weights of the preliminary water level sequence are: .

[0077] This allows us to determine the target water level data. :

[0078] in, This represents time-series extended water level data. This indicates the preliminary water level sequence.

[0079] Figure 4 The flowchart illustrates an example of the present invention, which uses boundary correction data and rainfall input data to simulate water levels during a monitoring blank period and obtain target water level data for that monitoring blank period.

[0080] like Figure 4 As shown, operation S130 also includes operations S410~S450.

[0081] In operation S410, historical observation data with continuous water level and flow rate observation results are selected as verification data.

[0082] According to an embodiment of the present invention, water level data and flow rate data that are continuously recorded in the time dimension can be selected from the historical monitoring data corresponding to the catchment area as verification data.

[0083] In some embodiments, historical observation data may include continuous monitoring data corresponding to multiple rainfall events, so that subsequent verification and parameter optimization processes cover the hydrological response characteristics under different rainfall intensities and different inflow conditions, thereby improving the adaptability of parameter optimization results to the target watershed.

[0084] When operating S420, based on historical observation data, the known time period and the time period to be verified are divided.

[0085] According to an embodiment of the present invention, the known time period is used to provide input conditions for the subsequent water level simulation process, while the time period to be verified is regarded as an artificially constructed monitoring blank period during the verification process, so as to test the ability of the water level simulation to simulate the water level change process during the period when there is a lack of measured monitoring data.

[0086] In operation S430, the actual data of the known time period is used as input to perform water level simulation for the time period to be verified, so as to obtain the target water level data corresponding to the time period to be verified.

[0087] According to an embodiment of the present invention, the actual water level data, actual flow data and corresponding rainfall data of a known time period can be used as input conditions. The actual water level and flow data at the end of the known time period can be used as the starting boundary conditions for the water level simulation of the time period to be verified, and the rainfall data corresponding to the time period to be verified can be used as hydrological driving conditions. Then, according to the aforementioned water level simulation steps, runoff generation simulation, runoff confluence simulation and water level-flow coupling conversion are performed on the time period to be verified to obtain the target water level data corresponding to the time period to be verified.

[0088] In operation S440, the target water level data corresponding to the period to be verified is compared with the actual observation data corresponding to the period to be verified time by time to obtain the comparison results.

[0089] According to embodiments of the present invention, the target water level data corresponding to the period to be verified can be compared with the actual observed water level data originally retained for that period at the same time resolution, hour by hour, to analyze the deviation between the simulation results and the actual observation results. In some embodiments, the comparison results may include at least one of the following: hour-by-hour error, maximum water level deviation, relative deviation of peak flow, and Nash efficiency coefficient.

[0090] When operating the S450, the water level simulation parameters used for water level simulation are iteratively optimized based on the comparison results.

[0091] According to an embodiment of the present invention, the parameters affecting the accuracy of water level simulation can be adjusted based on the comparison results, and the water level simulation and results comparison for the period to be verified can be re-executed after the parameter adjustment until the comparison results meet the preset accuracy requirements or the relevant deviation indicators are minimized.

[0092] Through the above-described setup, the embodiments of the present invention do not directly correct the water level simulation process based on data from actual monitoring blank periods. Instead, they select historical observation data with continuous water level and flow observation results, divide the period into known periods and periods to be verified, and construct verification samples corresponding to the actual monitoring blank scenarios. Based on this, the period to be verified is treated as an artificially constructed monitoring blank period, and the actual data from the known periods is used as input to perform water level simulation for the period to be verified. The simulated target water level data is then compared with the actual observation data corresponding to the period to be verified on a time-by-time basis, and the water level simulation parameters are iteratively optimized based on the comparison results. Therefore, the water level simulation process can be calibrated offline under the support of real observation results, making the water level simulation parameters more closely match the actual hydrological response characteristics of the pumped storage power station's catchment area. This improves the simulation accuracy of target water level data during subsequent monitoring blank periods and further enhances the accuracy and reliability of the all-time water level forecast results.

[0093] The following describes the water level forecasting device for the catchment area of ​​a pumped storage power station provided by the present invention. The water level forecasting device for the catchment area of ​​a pumped storage power station described below can be referred to in correspondence with the water level forecasting method for the catchment area of ​​a pumped storage power station described above.

[0094] Figure 5 A structural block diagram of a pumped storage power station catchment area water level prediction device according to an embodiment of the present invention is shown.

[0095] like Figure 5 As shown, the pumped storage power station catchment area water level forecasting device 500 includes a rainfall correction module 510, an initial water level forecasting module 520, a boundary correction module 530, a water level simulation module 540, an optimization module 550, and a forecasting module 560.

[0096] The rainfall correction module 510 is used to acquire and correct rainfall forecast data to obtain rainfall input data; The initial water level forecast module 520 is used to make a preliminary water level forecast based on rainfall input data and watershed water level characteristic parameters to obtain the initial water level forecast value; The boundary correction module 530 is used to construct a residual prediction model based on the residual sequence between the measured monitoring water level value and the initial water level forecast value, and to determine the boundary correction data for the monitoring blank period based on the initial water level forecast value of the effective monitoring period adjacent to the monitoring blank period and the residual prediction model. The water level simulation module 540 is used to perform water level simulation for the monitoring blank period based on boundary correction data and rainfall input data, and obtain the target water level data for the monitoring blank period. The optimization module 550 is used to construct a continuous water level sequence for the entire time period based on the monitoring water level data during the effective monitoring period and the target water level data during the monitoring blank period, and to perform water level forecasting based on the continuous water level sequence for the entire time period to obtain the water level forecasting results. The forecast module 560 is used to issue water level warnings based on the comparison between the water level forecast results and the preset warning threshold.

[0097] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 610, a communication interface 620, a memory 630, and a communication bus 640. The processor 610, communication interface 620, and memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions from the memory 630 to execute a method for predicting the water level in the catchment area of ​​a pumped-storage power station.

[0098] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0099] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the water level forecasting method for the catchment area of ​​a pumped storage power station provided by the above methods.

[0100] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0101] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0102] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting water level in the catchment area of ​​a pumped storage power station, characterized in that, include: Acquire and correct rainfall forecast data to obtain rainfall input data; Based on the rainfall input data and the watershed water level characteristic parameters, a preliminary water level forecast is made to obtain the initial water level forecast value; A residual prediction model is constructed based on the residual sequence between the measured monitored water level value and the initial water level forecast value. Boundary correction data for the monitoring blank period is determined based on the initial water level forecast value of the effective monitoring period adjacent to the monitoring blank period and the residual prediction model. Based on the boundary correction data and the rainfall input data, water level simulation is performed on the monitoring blank period to obtain the target water level data for the monitoring blank period; Based on the monitored water level data during the effective monitoring period and the target water level data during the monitoring blank period, a continuous water level sequence for the entire time period is constructed, and water level forecasting is performed based on the continuous water level sequence for the entire time period to obtain the water level forecasting result. Water level warnings are issued based on the comparison between the water level forecast results and the preset warning threshold.

2. The method for predicting water level in the catchment area of ​​a pumped storage power station according to claim 1, characterized in that, The step of simulating water levels during the monitoring blank period based on the boundary correction data and the rainfall input data to obtain target water level data includes: The boundary correction data is used as the initial boundary condition for the monitoring blank period, and the rainfall input data is used as the hydrological driving condition. Runoff simulation is performed based on the rainfall input data to obtain the runoff volume for each time period; Based on the time-by-time flow rate, a flow-in simulation is performed to obtain the time-by-time inflow rate of the preset control section; Based on the water level-flow relationship of the preset control section, the inflow rate for each time period is converted into water level data for each time period to obtain the target water level data.

3. The method for predicting water level in the catchment area of ​​a pumped storage power station according to claim 2, characterized in that, The process of performing runoff simulation based on the rainfall input data to obtain the runoff volume for each time period includes: If the runoff parameters of the water collection area meet the first preset condition, the runoff volume is calculated using the full-capacity runoff model; When the flow parameters of the water collection area meet the second preset condition, the flow rate is calculated using the super-permeable flow model.

4. The method for predicting water level in the catchment area of ​​a pumped storage power station according to claim 2, characterized in that, The step of converting the inflow rate over time period into water level data over time period based on the water level-flow relationship of the preset control section to determine the target water level data includes: Based on the water level-flow relationship of the preset control section, the inflow rate for each time period is converted into a corresponding preliminary water level sequence, and the water level change rate for each time period is determined based on the preliminary water level sequence. The time-series extension method is determined based on the rainfall input data and the water level change rate for each time period, and the preliminary water level sequence is extended according to the time-series extension method to obtain the time-series extended water level data for each time period. The target water level data is determined based on the time-series extended water level data and the preliminary water level sequence.

5. The method for predicting water level in the catchment area of ​​a pumped storage power station according to claim 4, characterized in that, The method of determining the time-series extension based on time-period rainfall input data and water level change rate includes: If the rainfall intensity corresponding to the rainfall input data for each time period is less than or equal to a preset rainfall intensity threshold, and the water level change rate for each time period is less than or equal to a preset water level change rate threshold, then the time-series extension method is determined to be linear fitting. If the rainfall intensity corresponding to the rainfall input data for each time period is greater than the preset rainfall intensity threshold, or the water level change rate for each time period is greater than the preset water level change rate threshold, the time series extension method is determined to be cubic spline interpolation.

6. The method for predicting water level in the catchment area of ​​a pumped storage power station according to claim 4, characterized in that, The step of determining the target water level data based on the time-series extended water level data and the preliminary water level sequence includes: Based on the rainfall input data for each time period, the weights of the time-series extended water level data and the weights of the preliminary water level sequence are determined; The target water level data is determined based on the time-series extended water level data, the preliminary water level sequence, and the corresponding weights.

7. The method for predicting water level in the catchment area of ​​a pumped storage power station according to claim 1, characterized in that, The step of simulating water level during the monitoring blank period based on the boundary correction data and the rainfall input data to obtain the target water level data for the monitoring blank period also includes: Historical observation data with continuous water level and flow rate observation results were selected as validation data; Based on the historical observation data, the known time period and the time period to be verified are divided; Using the actual data of the known time period as input, water level simulation is performed on the time period to be verified to obtain the target water level data corresponding to the time period to be verified. The target water level data corresponding to the period to be verified is compared with the actual observation data corresponding to the period to be verified on a time-by-time basis to obtain the comparison results; Based on the comparison results, the water level simulation parameters used for the water level simulation are iteratively optimized.

8. A water level forecasting device for the catchment area of ​​a pumped storage power station, characterized in that, include: The rainfall correction module is used to acquire and correct rainfall forecast data to obtain rainfall input data; The initial water level forecast module is used to perform preliminary water level forecasts based on the rainfall input data and watershed water level characteristic parameters to obtain initial water level forecast values. The boundary correction module is used to construct a residual prediction model based on the residual sequence between the measured monitored water level value and the initial water level forecast value, and to determine the boundary correction data for the monitoring blank period based on the initial water level forecast value of the effective monitoring period adjacent to the monitoring blank period and the residual prediction model. The water level simulation module is used to perform water level simulation for the monitoring blank period based on the boundary correction data and the rainfall input data, so as to obtain the target water level data for the monitoring blank period. The optimization module is used to construct a continuous water level sequence for the entire time period based on the monitoring water level data during the effective monitoring period and the target water level data during the monitoring blank period, and to perform water level forecasting based on the continuous water level sequence for the entire time period to obtain the water level forecasting result. The forecast module is used to issue water level warnings based on the comparison between the water level forecast results and the preset warning threshold.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for predicting the water level in the catchment area of ​​a pumped storage power station as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for predicting the water level in the catchment area of ​​a pumped storage power station as described in any one of claims 1 to 7.

Images