Reservoir level prediction method and prediction system based on lead-lag relationship

Abstract

The invention provides a reservoir level prediction method and system based on a lead-lag relation, and relates to the field of hydrological prediction. The method provided by the invention comprises the following steps: acquiring original hydrological time sequence data (including water level, flow and the like), and analyzing and acquiring advanced hydrological elements and time periods thereof, and lagged hydrological elements and time periods thereof by adopting maximum mutual information number time-lag correlation; performing time alignment on the advanced hydrological elements according to the advanced time period number to construct an advanced data set; performing time alignment on the lagged hydrological elements according to the number of lagged time periods, and if the lagged hydrological elements which are not acquired exist, acquiring a predicted value for filling by adopting the hydrological element prediction model based on a single-model single-output multi-step rolling strategy, and constructing a lagged data set; and according to the advanced data set and the lagged data set, predicting the reservoir level by adopting a reservoir level prediction model based on a data driving method. According to the method, advanced hydrological elements and lagging hydrological elements are considered, and high-precision and high-timeliness prediction of the water level of the reservoir is achieved.

Description

Technical Field

[0001] The present invention relates to the field of hydrological prediction, and particularly to a reservoir water level prediction method and prediction system based on lead-lag relationships. Background Art

[0002] The change of reservoir water level is directly related to key fields such as water resources management, flood control scheduling, and power generation efficiency. Therefore, reservoir water level prediction is of decisive significance for ensuring the safety of water conservancy projects, realizing the rational utilization of water resources, and formulating flood control and disaster reduction measures. The reservoir water level is comprehensively affected by various meteorological and hydrological elements such as flow rate, rainfall, and evaporation. These elements have significant time lags, and their interaction presents complex non-linear characteristics, resulting in extremely complex water level change laws. In the practice of flood control, drought relief, and water resources management, prediction errors of flood season water levels may lead to incorrect flood control decisions, threatening the safety of people's lives and property; during water resources allocation, prediction deviations will cause water resources waste or supply shortages, affecting the sustainable development of the regional economy and society.

[0003] Traditional single-factor water level prediction models only rely on historical water level data for prediction. When the prediction period is long, their prediction results often show problems of insufficient timeliness and large errors; existing research has bottlenecks in dealing with the complex relationships of multiple elements, and it is difficult to accurately reveal the internal connections between each element and the water level, unable to meet the actual needs of high-precision water level prediction. Therefore, there is an urgent need to provide a solution to improve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide a reservoir water level prediction method and prediction system based on lead-lag relationships, which can improve the low accuracy and low timeliness of the prediction results of existing water level prediction methods.

[0005] In a first aspect, a reservoir water level prediction method based on lead-lag relationships provided by the present invention includes: Obtaining original hydrological time series data , and obtaining leading hydrological elements, leading time periods , lagging hydrological elements , and lagging time periods by using maximum mutual information time lag correlation analysis; the original hydrological time series data includes reservoir water level time series data ; ; Aligning the leading hydrological elements in time according to the leading time periods to construct a leading data set ; Aligning the lagging hydrological elements in time according to the lagging time periods Perform time alignment. If there are lag hydrological elements that have not been obtained, use a hydrological element prediction model to obtain predicted values based on a single-model single-output multi-step rolling strategy for filling, and construct a lag dataset. ; According to the leading dataset and the lag dataset , use a reservoir water level prediction model based on a data-driven method to predict the reservoir water level.

[0006] A reservoir water level prediction method based on the leading-lag relationship provided by the present invention obtains the original hydrological time series data, and uses the maximum mutual information time-delay correlation analysis to obtain the leading hydrological elements and their time periods, and the lag hydrological elements and their time periods; then perform time alignment on the leading hydrological elements and the lag hydrological elements; finally, according to the aligned hydrological elements, use a reservoir water level prediction model based on a data-driven method to achieve high-precision and high-timeliness prediction of the reservoir water level.

[0007] Optionally, when using the maximum mutual information time-delay correlation analysis to obtain the leading hydrological elements , the leading time period , the lag hydrological elements and the lag time period , it includes: sequentially selecting the original hydrological time series data and the reservoir water level time series data to construct a two-dimensional dataset , and using a grid division algorithm to calculate the corresponding maximum mutual information for different time delays, obtaining its maximum information coefficient based on the maximum mutual information at different time delays, comparing the maximum information coefficients corresponding to all time delays, obtaining the maximum value of the maximum information coefficient and its corresponding time delay; if the time delay is less than 0, the current original hydrological time series data is the leading hydrological element , and the leading time period is the current time delay; if the time delay is greater than 0, the current original hydrological time series data is the lag hydrological element , and the lag time period is the current time delay.

[0008] Optionally, when performing time alignment on the leading hydrological elements , it includes: moving all the leading hydrological elements backward in time order by their respective leading time periods , to obtain the leading hydrological elements after time alignment : , Respectively, for each time-aligned leading hydrological element and the reservoir water level time series data construct a leading dataset : { D i ,t C } = [ Z 1 Z 2 ⋮ H i , 1 C ' H i , 2 C ' ⋮ Z T H i , T C ' ] , Merge all the leading datasets , and obtain a leading dataset containing all time-aligned leading hydrological elements : { D i ,t C ' } = [ Z 1 H 1 , 1 C ' H 2 , 1 C ' ⋯ H i , 1 C ' Z 2 H 1 , 2 C ' H 2 , 2 C ' ⋯ H i , 2 C ' ⋮ ⋮ ⋮ ⋮ Z T H 1 , T C ' H 2 , T C ' ⋯ H i , T C ' ] .

[0009] Optionally, when time-aligning the lagging hydrological elements , it includes: moving all the lagging hydrological elements forward in chronological order by their respective lagging time periods . If there are unobtained lagging hydrological elements, use a hydrological element prediction model to obtain predicted values according to the single-model single-output multi-step rolling strategy to fill the lagging hydrological elements. Respectively, for each filled lagging hydrological element and the reservoir water level time series data construct a lagging dataset : { D j ,t F } = [ Z 1 Z 2 ⋮ H j, 1 F ' H j, 2 F ' ⋮ Z T H j, 3 F ' ] , Merge all the lagging datasets , and obtain a lagging dataset containing all time-aligned lagging hydrological elements : { D j ,t F ' } = [ Z 1 H 1 , 1 F ' H 2 , 1 F ' ⋯ H j , 1 F ' Z 2 H 1 , 2 F ' H 2 , 2 F ' ⋯ H j , 2 F ' ⋮ ⋮ ⋮ ⋮ Z T H 1 , T F ' H 2 , T F ' ⋯ H j , T F ' ] .

[0010] Optionally, when using a hydrological element prediction model to obtain predicted values based on the single-model single-output multi-step rolling strategy, it includes: Step 1: Use the historical data of the latest time steps in the time-aligned lagging hydrological elements as the input sequence, input it into the hydrological element prediction model, and output the predicted value of the first future time step; where is the preset lagging time period; Step 2: Remove the historical data of the earliest time step in the input sequence, and add the predicted value of the first future time step to the end of the input sequence to form an updated input sequence; Step 3: Input the updated input sequence into the hydrological element prediction model to obtain the predicted value of the next future time step; Step 4: Repeat Step 2 to Step 3, sequentially remove the data at the earliest time step in the current input sequence, and add the latest obtained predicted value to the end of the input sequence until the predicted values for the required number of future time steps are obtained.

[0011] Optionally, the training method of the reservoir water level prediction model includes: defining an objective function, using the NSGA-II multi-objective optimization algorithm to optimize the parameters to be optimized and hyperparameters in the data-driven method to generate a Pareto optimal solution set; based on the TOPSIS multi-attribute decision-making method, calculate the closeness degrees of each solution in the Pareto optimal solution set to the ideal solution and the negative ideal solution respectively, sort according to the closeness degrees and select the optimal solution; construct a reservoir water level prediction model based on the optimal solution.

[0012] Optionally, the objective function includes: mean absolute error, root mean square error, mean square error, interquartile range related index, Nash-Sutcliffe efficiency coefficient, coefficient of determination, and maximum error.

[0013] In a second aspect, the present invention also provides a reservoir water level prediction system based on the lead-lag relationship, including: A data acquisition module that acquires original hydrological time series data and obtains leading hydrological elements, leading time periods, lagging hydrological elements, and lagging time periods by using the maximum mutual information time-delay correlation analysis; the original hydrological time series data includes reservoir water level time series data; A leading alignment module that aligns the leading hydrological elements in time according to the leading time periods to construct a leading data set; A lagging alignment module that aligns the lagging hydrological elements in time according to the lagging time periods. If there are lagging hydrological elements that have not been obtained, use a hydrological element prediction model to obtain predicted values based on the single-model single-output multi-step rolling strategy for filling and construct a lagging data set; A water level prediction module that predicts the reservoir water level by using a reservoir water level prediction model based on a data-driven method according to the leading data set and the lagging data set. A leading alignment module that aligns the leading hydrological elements in time according to the leading time periods to construct a leading data set; A lagging alignment module that aligns the lagging hydrological elements in time according to the lagging time periods. If there are lagging hydrological elements that have not been obtained, use a hydrological element prediction model to obtain predicted values based on the single-model single-output multi-step rolling strategy for filling and construct a lagging data set; A water level prediction module that predicts the reservoir water level by using a reservoir water level prediction model based on a data-driven method according to the leading data set and the lagging data set. Description of the Drawings

[0014] Figure 1 ​​​​​​Flow chart of a reservoir water level prediction method based on lead-lag relationship provided by an embodiment of the present invention.

[0015] Figure 2 Structural diagram of a reservoir water level prediction system based on lead-lag relationship provided by an embodiment of the present invention. Detailed implementation manners

[0016] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Apparently, the described embodiments are some but not all of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein shall have the ordinary meanings as understood by those of ordinary skill in the art in the field to which the present invention belongs.

[0017] Refer to Figure 1 , the present invention provides a reservoir water level prediction method based on lead-lag relationship, including the following steps: S1. Obtain the original hydrological time series data , and use the maximum mutual information time delay correlation analysis to obtain the leading hydrological elements , the leading time period number , the lagging hydrological elements and the lagging time period number ; the original hydrological time series data includes the reservoir water level time series data ; S2. Align the leading hydrological elements according to the leading time period number to construct a leading data set ; S3. Align the lagging hydrological elements according to the lagging time period number . If there are lagging hydrological elements that have not been obtained, use the hydrological element prediction model to obtain the predicted values based on the single-model single-output multi-step rolling strategy for filling, and construct a lagging data set ; S4. According to the leading data set and the lagging data set , use the reservoir water level prediction model based on the data-driven method to predict the reservoir water level.

[0018] In fact, the reservoir water level prediction method provided by the present invention obtains the original hydrological time series data, and uses the maximum mutual information time-delay correlation analysis to obtain the leading hydrological elements, the number of leading time periods, the lagging hydrological elements and the number of lagging time periods. Then, the hydrological elements are time-aligned according to the number of time periods. When aligning the lagging hydrological elements, if there are missing parts, the hydrological element prediction model can be used to obtain the predicted values according to the single-model single-output multi-step rolling strategy for filling. Finally, a data set is constructed based on the aligned hydrological elements, and a reservoir water level prediction model based on the data-driven method is used to predict the reservoir water level. Therefore, high-timeliness and low-error prediction results can be obtained, and the purpose of high-precision prediction of the reservoir water level is achieved.

[0019] In some embodiments, in step S1, the original hydrological time series data and the reservoir water level time series data are collected from the historical hydrological data related to the Xiajiang Project in the middle reaches of the Ganjiang River from December 2013 to August 2020. The historical hydrological data covers time series data such as the outflow discharge, the flow at the Xiajiang Station, the flow at the Dongbei Station, the flow at the Ji'an Station, the rainfall at the Xiajiang Station, the rainfall at the Dongbei Station, the rainfall at the Ji'an Station, and the evaporation at the Xiajiang Station; among them, the original hydrological time series data includes water level, flow, etc.

[0020] In some embodiments, when obtaining the original hydrological time series data in step S1, it is necessary to correct the missing values and normalize the original hydrological time series data.

[0021] Specifically, the linear interpolation method is used to correct the missing values of the original hydrological time series data, and the formula is as follows: , where, is the missing point to be interpolated, is the estimated value of the missing point to be interpolated, , are the two known data points before and after the missing point.

[0022] Specifically, the min-max normalization method is used to map the range of all original hydrological time series data to the [0,1] standardization interval, and the formula is as follows: , where, is the normalized value, is the original data, is the minimum value of the original data, is the maximum value of the original data.

[0023] In some embodiments, when the maximum mutual information number time lag correlation analysis is used in step S1 to obtain the leading hydrological elements, the number of leading time periods, the lagging hydrological elements, and the number of lagging time periods, a certain original hydrological time series data and the reservoir water level time series data are sequentially selected to construct a two-dimensional data set, and the corresponding maximum mutual information is calculated using a grid partitioning algorithm for different time lags. The maximum information coefficient is obtained based on the maximum mutual information under different time lags, and the maximum information coefficient corresponding to all time lags is compared to obtain the maximum value of the maximum information coefficient and its corresponding time lag; if the time lag is less than 0, the current original hydrological time series data is the leading hydrological element, and the number of leading time periods is the current time lag; if the time lag is greater than 0, the current original hydrological time series data is the lagging hydrological element, and the number of lagging time periods is the current time lag. The specific process is as follows: S11, set the range of the number of leading time periods to [ - n ,-1 ] , the range of the number of lag periods is [ 1, m ] ; Select a certain original hydrological time series data , and combined with reservoir water level time series data Construct a sample set based on a given two-dimensional random variable , , .in, and is the maximum number of leading periods and the maximum number of lagging periods set in advance, is the time series length of the data; S12. Time series data and Analyze and determine its maximum and minimum values respectively. 、 In the range of values, according to the arithmetic progression, indivual, different values, so the sample space can be divided into List Gridding rows to generate totals A two-dimensional grid with grid cells. and The two-dimensional random variable sample set , count the number of sample data points falling into each grid, and calculate the number of sample data points falling into the first grid by these numbers. Column and The frequencies in the row grid are used to deduce the sample set of two-dimensional random variables exist 、 The marginal probability distribution on and , and the joint probability distribution ; S13. According to the edge probability distribution and joint probability distribution obtained above, the mutual information index of the current partition is derived : , Normalize the mutual information index to obtain the maximum mutual information : , Two-dimensional random variable sample set The maximum information coefficient Defined as the normalized maximum mutual information under this partition: , S14: Within the range of the set number of leading and lagging time periods, change the number of lag days in sequence ( exist [ - n ,-1 ] and [ 1, m ] For each time lag , the original hydrological time series data According to the time lag Perform time offset, reconstruct the sample set, repeat S12 to S13, and obtain the reservoir water level time series data With the original hydrological time series data of value; S15. Compare all calculated value, find the maximum value among them . The corresponding time lag , which is the time lag between the hydrological element and the reservoir water level. , then the hydrological element is ahead of the reservoir water level, recorded as the number of leading periods ;like , then the hydrological element lags behind the reservoir water level, recorded as the lag period number The above operation is performed for each hydrological element to identify the leading hydrological elements with a time lag of less than 0 with the reservoir water level. and the number of leading periods , and delayed hydrological elements with time lag greater than 0 and the number of lag periods .

[0024] In some embodiments, when performing time alignment on leading hydrological elements in step S2, all leading hydrological elements are moved backward by their respective leading time periods in chronological order. Each time-aligned leading hydrological element is respectively combined with the reservoir water level time series data to construct a leading dataset, and all leading datasets are merged to obtain a leading dataset containing all time-aligned leading hydrological elements. The specific process is as follows: S21. Denote the reservoir water level time series data as , the leading hydrological element as and its corresponding leading time period as . Denote the time-aligned leading hydrological element as , where is the length of the time series of the data. For the leading hydrological element , since it leads the reservoir water level time periods, it is necessary to move the leading hydrological element backward by positions in chronological order, as follows: , S22. Combine the time-aligned leading hydrological element and the reservoir water level time series data to construct a leading dataset . can be represented as a matrix, as follows: { D i ,t C } = [ Z 1 Z 2 ⋮ H i , 1 C ' H i , 2 C ' ⋮ Z T H i , T C ' ] , S23. At this time, each row of data in the leading dataset represents the data values of the reservoir water level and the time-aligned leading hydrological elements at the same time point. Repeat S21 to S22 for the remaining leading hydrological elements to perform time alignment according to their respective leading time periods, and merge all leading datasets to obtain a leading dataset containing all time-aligned leading hydrological elements, as follows: { D i ,t C ' } = [ Z 1 H 1 , 1 C ' H 2 , 1 C ' ⋯ H i , 1 C ' Z 2 H 1 , 2 C ' H 2 , 2 C ' ⋯ H i , 2 C ' ⋮ ⋮ ⋮ ⋮ Z T H 1 , T C ' H 2 , T C ' ⋯ H i , T C ' ] .

[0025] In some embodiments, when performing time alignment on lagging hydrological elements in step S3, all lagging hydrological elements are moved forward by their respective lagging time periods in chronological order. If there are unobtained lagging hydrological elements, the hydrological element prediction model is used to obtain predicted values according to the single-model single-output multi-step rolling strategy to fill the lagging hydrological elements. Each filled lagging hydrological element is respectively combined with the reservoir water level time series data to construct a lagging dataset, and all lagging datasets are merged to obtain a lagging dataset containing all time-aligned lagging hydrological elements. The specific process is as follows: S31. Denote the time series data of the reservoir water level as , the lagged hydrological element as and its corresponding lag period number as . Denote the lagged hydrological element after time alignment as , where is the length of the time series of the data; for the lagged hydrological element , since it lags behind the reservoir water level periods, the lagged hydrological element needs to be moved forward positions in chronological order, as follows: , . Among them, if the lagged hydrological element in the time period is not obtained, the single-model single-output multi-step rolling strategy is used to obtain the predicted value for filling . The single-model single-output multi-step rolling strategy is as follows: S311. Take the historical data of the last time steps of the lagged hydrological element after time alignment as the input sequence , and input the input sequence into the trained hydrological element prediction model to obtain the predicted value of the first future time step. If a data-driven prediction model is constructed using a long short-term memory network (LSTM), it can be expressed as ; S312. Add the predicted value to the input sequence , replace the historical data of the earliest time step in the input sequence to form the updated input sequence . If the original input sequence is , after adding the predicted value ; S313. Input the updated input sequence into the hydrological element prediction model to obtain the predicted value of the next future time step; S314. Repeat steps S312 to S313, add the latest obtained predicted value to the input sequence to replace the data of its earliest time step until the predicted values of the required number of future time steps are obtained. Let the input sequence at the nth prediction be , input it into the hydrological element prediction model to obtain the predicted value of the nth future time step, that is: ; S32. Combine the filled lagged hydrological elements with the time series data of the reservoir water level according to the same chronological order and structural requirements to construct a lagged data set , as follows: { D j ,t F } = [ Z 1 Z 2 ⋮ H j, 1 F ' H j, 2 F ' ⋮ Z T H j, 3 F ' ] , S33. At this time Each line of data in represents the reservoir water level and the aligned lag hydrological element data values at the same time point. Repeat S31 to S32 for the remaining lag hydrological elements to perform time alignment according to their respective corresponding lag periods and predicted values, and merge all lag data sets to obtain a lag data set containing all time-aligned lag hydrological elements , as follows: { D j ,t F ' } = [ Z 1 H 1 , 1 F ' H 2 , 1 F ' ⋯ H j , 1 F ' Z 2 H 1 , 2 F ' H 2 , 2 F ' ⋯ H j , 2 F ' ⋮ ⋮ ⋮ ⋮ Z T H 1 , T F ' H 2 , T F ' ⋯ H j , T F ' ] .

[0026] In some embodiments, the training method of the reservoir water level prediction model in step S4 includes: defining multiple objective functions to comprehensively evaluate the performance of the reservoir water level prediction model, using the NSGA-II multi-objective optimization algorithm to optimize the objective functions, and precisely defining the parameters and hyperparameters to be optimized in the data-driven method; after optimization by NSGA-II, generating a Pareto optimal solution set, using the TOPSIS multi-attribute decision-making method, considering multiple attribute dimensions, calculating the closeness degrees of each solution in the Pareto optimal solution set to the ideal solution and the negative ideal solution respectively, sorting the alternative solutions according to the closeness degrees, and selecting the solution with the highest closeness degree as the optimal solution; finally, constructing a reservoir water level prediction model based on the data-driven method according to the optimal solution. Among them, the objective functions can be the mean absolute error (MAE), root mean square error (RMSE), mean square error (MSE), interquartile range related index (QR), Nash-Sutcliffe efficiency coefficient (NS), coefficient of determination (R²), and maximum error (MIE). In addition, the hydrological element prediction model can also be trained with reference to this training method

[0027] See Figure 2 , the present invention provides a reservoir water level prediction system based on the lead-lag relationship, including the following steps: Data acquisition module 100, which acquires the original hydrological time series data , and obtains the leading hydrological elements by using the maximum mutual information time lag correlation analysis , the leading period number , the lag hydrological elements and the lag period number ; the original hydrological time series data includes the reservoir water level time series data ; Leading alignment module 200, which performs time alignment on the leading hydrological elements according to the leading period number to construct a leading data set ; Lag alignment module 300, according to the number of lag periods Align the lag hydrological elements in terms of time. If there are unobtained lag hydrological elements, a prediction value is obtained by using a hydrological element prediction model based on a single-model single-output multi-step rolling strategy for filling, and a lag data set is constructed ; Water level prediction module 400, according to the leading data set and the lag data set , predicts the reservoir water level by using a reservoir water level prediction model based on a data-driven method

[0028] Although the embodiments of the present invention have been described in detail above, it is obvious to those skilled in the art that various modifications and changes can be made to these embodiments. However, it should be understood that such modifications and changes are all within the scope and spirit of the present invention described in the claims. Moreover, the present invention described herein can have other embodiments and can be implemented or realized in various ways

Claims

1. A reservoir water level prediction method based on lead-lag relationship, characterized in that including: Obtain the original hydrological time series data , and use the maximum mutual information time-delay correlation analysis to obtain leading hydrological elements , the number of leading time periods , lagging hydrological elements and the number of lagging time periods ; the original hydrological time series data includes reservoir water level time series data ; According to the number of leading time periods perform time alignment on the leading hydrological elements to construct a leading data set ; According to the number of lag periods perform time alignment on the lag hydrological elements If there are unobtained lag hydrological elements, use a hydrological element prediction model to obtain predicted values based on a single-model single-output multi-step rolling strategy for filling, and construct a lag data set ; Based on the leading dataset and the lagging dataset , a reservoir water level prediction model using a data-driven method is adopted to predict the reservoir water level.

2. The reservoir water level prediction method according to claim 1, characterized in that Obtaining leading hydrological elements by using maximum mutual information time-delay correlation analysis , the number of leading time periods , lagging hydrological elements and the number of lagging time periods When doing so, it includes: successively selecting the original hydrological time series data and the reservoir water level time series data to construct a two-dimensional data set , and using a grid division algorithm to calculate the corresponding maximum mutual information for different time delays, obtaining its maximum information coefficient based on the maximum mutual information at different time delays, comparing the maximum information coefficients corresponding to all time delays, obtaining the maximum value of the maximum information coefficient and its corresponding time delay; if the time delay is less than 0, the current original hydrological time series data is the leading hydrological element , and the number of leading time periods is the current time delay; if the time delay is greater than 0, the current original hydrological time series data is the lagging hydrological element , and the number of lagging time periods is the current time delay.

3. The reservoir water level prediction method according to claim 1, characterized in that, When performing time alignment on the advanced hydrological elements it includes: moving all the advanced hydrological elements backward in chronological order by their respective advanced time periods to obtain the advanced hydrological elements after time alignment : The leading hydrological elements after time alignment for each are respectively and the reservoir water level time series data to construct a leading dataset : Merge all leading datasets to obtain a leading dataset containing all leading hydrological elements after time alignment : 。 4. The reservoir water level prediction method according to claim 1, wherein When performing time alignment on the lagged hydrological elements it includes: moving all the lagged hydrological elements forward in chronological order by their respective lag periods . If there are unobtained lagged hydrological elements, a hydrological element prediction model is used to obtain predicted values according to the single-model single-output multi-step rolling strategy to fill the lagged hydrological elements. Respectively, each filled lagged hydrological element and the reservoir water level time series data are used to construct a lagged dataset as follows: Merge all lagged datasets to obtain a lagged dataset containing all lagged hydrological elements after time alignment : 。 5. The reservoir water level prediction method according to claim 1, characterized in that, When obtaining prediction values by using a hydrological element prediction model based on a single-model single-output multi-step rolling strategy, it includes: Step 1. Use the historical data of the latest time steps in the lagged hydrological elements after time alignment as the input sequence, and input it into the hydrological element prediction model to output the predicted value of the first future time step; where is the preset number of lag periods; is the preset number of lag periods;​ Step 2: Remove the historical data of the earliest time step in the input sequence, and add the prediction value of the first future time step to the end of the input sequence to form an updated input sequence; Step 3: Input the updated input sequence into the hydrological element prediction model to obtain the prediction value of the next future time step; Step 4: Repeat Step 2 to Step 3, sequentially remove the data of the earliest time step in the current input sequence, and add the latest obtained prediction value to the end of the input sequence until the prediction values of the required number of future time steps are obtained.

6. The reservoir water level prediction method according to claim 1, wherein The training method of the reservoir water level prediction model includes: defining an objective function, using the NSGA-II multi-objective optimization algorithm to optimize the parameters and hyperparameters to be optimized in the data-driven method to generate a Pareto optimal solution set; based on the TOPSIS multi-attribute decision-making method, calculate the closeness degrees of each solution in the Pareto optimal solution set to the ideal solution and the negative ideal solution respectively, sort according to the closeness degrees and select the optimal solution; construct a reservoir water level prediction model based on the optimal solution.

7. The reservoir water level prediction method according to claim 6, characterized in that, The objective function includes: mean absolute error, root mean square error, mean square error, interquartile range related index, Nash-Sutcliffe efficiency coefficient, coefficient of determination, and maximum error.

8. A reservoir water level prediction system based on lead-lag relationship, characterized in that, including: Data acquisition module, which obtains the original hydrological time series data , and uses the maximum mutual information time-delay correlation analysis to obtain leading hydrological elements , the number of leading time periods , lagging hydrological elements and the number of lagging time periods ; The original hydrological time series data includes reservoir water level time series data ; An early alignment module, according to the number of early time periods performs time alignment on the early hydrological elements to construct an early data set ; Lag alignment module, according to the number of lag periods Perform time alignment on the lag hydrological elements If there are unobtained lag hydrological elements, a hydrological element prediction model is used to obtain predicted values based on the single-model single-output multi-step rolling strategy for filling, and a lag data set is constructed ; Water level prediction module, based on the leading dataset and the lagging dataset , uses a reservoir water level prediction model based on a data-driven method to predict the reservoir water level.

Images