Reservoir flood control peak shaving scheduling method, system and device based on timing dynamic target water level and analytic POA optimization and medium

Abstract

This invention relates to a reservoir flood control peak shaving scheduling method, system, equipment, and medium based on time-series dynamic target water level and analytical POA optimization, belonging to the field of reservoir flood control scheduling technology. The method includes: obtaining the start and end times of flood events; identifying and dividing the flood peak and valley generation interval sequence within the start and end times; constructing a piecewise linearly changing time-series dynamic target water level process line based on the interval sequence and performing storage compensation calculations to generate an initial reservoir capacity sequence; under the constraint of the boundary reservoir capacity of adjacent time periods, locally linearizing the reservoir capacity relationship within the boundary reservoir capacity, then gradually transforming the objective function into a quadratic function, and solving for the derivative of the quadratic function to obtain the optimal reservoir capacity for the current time period; iteratively solving for the optimal reservoir capacity for each time period and updating the reservoir capacity sequence until convergence, outputting a reservoir scheduling scheme generated from the optimized reservoir capacity sequence. This invention improves the flood control efficiency and emergency response capability of reservoirs in dealing with complex multi-peak floods.

Description

Technical Field

[0001] This invention relates to the field of reservoir flood control scheduling technology, and in particular to a reservoir flood control peak shaving scheduling method, system, equipment and medium based on time-series dynamic target water level and analytical POA optimization. Background Technology

[0002] Reservoir flood control scheduling is a crucial link in the basin's flood control and disaster reduction system. Its core task is to rationally allocate inflow floodwaters during floods by regulating the reservoir's discharge facilities, thereby reducing peak flow, ensuring downstream flood control safety, and simultaneously ensuring the safety of the reservoir itself. Existing reservoir flood control scheduling methods typically generate scheduling plans based on pre-established scheduling rules or optimization algorithms.

[0003] In terms of dispatching strategies, conventional flood control methods often use a fixed flood limit water level as the control target for the entire flood process. However, due to the significant time-varying and uncertain nature of flood processes, especially when encountering multi-peak or long-duration floods, a fixed target water level strategy is difficult to adapt to the dynamic evolution of the flood process. For example, if the flood limit water level is strictly maintained before the arrival of the flood peak, the reservoir will not be able to fully utilize its capacity to intercept floodwater, and the peak reduction effect will be limited. In the receding phase after the flood peak or during the interval between flood peaks, a fixed low target water level strategy cannot effectively guide the reservoir to increase discharge and pre-emptively create storage capacity in a timely manner, resulting in insufficient storage capacity when subsequent flood peaks arrive, forcing the release of large amounts of water.

[0004] In terms of flood process identification, existing scheduling methods have weak automatic analysis capabilities for complex flood processes. Some methods rely on human experience to judge the onset, peak, and recession stages of floods, lacking refined and automated means of identifying the peak and trough structure of floods.

[0005] In the optimization of reservoir discharge schemes, to achieve optimal peak shaving, the industry commonly employs mathematical optimization methods such as dynamic programming, stepwise optimization algorithms, and genetic algorithms to solve the reservoir discharge process. Among these, stepwise optimization algorithms are widely used due to their good adaptability to reservoir dispatching problems. However, conventional stepwise optimization algorithms typically require numerical iteration methods to repeatedly calculate decision variables to approximate the optimal solution when solving two-stage subproblems. Furthermore, reservoir dispatching involves complex nonlinear reservoir capacity curves and hydraulic constraints, resulting in a computationally intensive and slow convergence process that often takes minutes or even hours. Therefore, when facing sudden floods requiring rapid response, the computational efficiency of these methods often fails to meet the practical demands of real-time emergency dispatching for second-level decision-making.

[0006] Therefore, those skilled in the art urgently need a reservoir flood control and peak shaving scheduling method that can dynamically adjust the scheduling target according to the real-time evolution of the flood, automatically identify complex flood peak and valley structures, and improve computational efficiency while ensuring solution accuracy, so as to meet the dual requirements of real-time performance and accuracy in the event of a sudden flood. Summary of the Invention

[0007] (a) Technical problems to be solved

[0008] In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a reservoir flood control peak shaving scheduling method, system, equipment and medium based on time-series dynamic target water level and analytical POA optimization, which solves the technical problems of poor adaptability of scheduling targets, low degree of automation of flood peak and valley identification and insufficient optimization calculation efficiency in the existing reservoir flood control scheduling methods.

[0009] (II) Technical Solution

[0010] To achieve the above objectives, the main technical solutions adopted by the present invention include:

[0011] In a first aspect, embodiments of the present invention provide a reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization, comprising:

[0012] The baseflow is segmented from the acquired forecast inflow data of the reservoir basin to obtain the start and end times of flood events;

[0013] Based on the flood valley flow and interpeak duration of flood events, flood peaks and valleys are identified and divided within the start and end time, generating interval sequences that include flood peak intervals, flood valley intervals, and end intervals.

[0014] Based on the interval sequence, a piecewise linearly varying time-series dynamic target water level process line is constructed, and storage compensation calculation is performed in conjunction with the predicted inflow process data to generate an initial reservoir capacity sequence;

[0015] Under the constraint of the boundary capacity of adjacent time periods, the capacity relationship within the boundary capacity range is locally linearized. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal capacity of the current time period.

[0016] The optimal reservoir capacity for each time period is solved by traversing all scheduling periods. The initial reservoir capacity sequence is replaced with the updated reservoir capacity sequence and multiple iterations are performed until the preset convergence condition is met. The reservoir scheduling scheme generated by the optimized reservoir capacity sequence is then output.

[0017] Optionally, baseflow segmentation is performed on the acquired forecast inflow data of the reservoir basin to obtain the start and end times of the flood event, including:

[0018] Acquire forecast inflow data for the reservoir basin and conduct basin size analysis to obtain scale characteristic parameters of the reservoir basin;

[0019] Based on the scale and characteristic parameters of the reservoir basin, the corresponding baseflow segmentation method is selected from the preset baseflow segmentation method library to calculate the baseflow segmentation threshold and obtain the segmentation flow threshold of the reservoir basin. The baseflow segmentation methods include the average flow method, the flood peak percentage method, and the change point detection method.

[0020] The forecast inflow data is compared with the segmented flow threshold. Based on the comparison results, the forecast inflow process of the reservoir basin is divided into baseflow and flood. The period when the flow rate is continuously higher than the segmented flow threshold is taken as the start and end time of the flood event.

[0021] Optionally, based on the flood valley flow and interpeak duration of the flood event, the flood peaks and valleys within the start and end times are identified and divided, generating an interval sequence including the flood peak interval, the flood valley interval, and the end interval, including:

[0022] Based on the flood valley flow and interpeak duration of flood events, flood peaks and valleys within the start and end times are identified and divided using flood merging and flood division criteria, generating an initial interval sequence containing flood peak intervals, flood valley intervals, and end intervals.

[0023] Based on the set constraints including minimum flood peak interval, minimum flood peak amplitude, and minimum flood peak duration, the flood peak intervals in the initial interval sequence are filtered out, and flood peak intervals that do not meet the constraints are removed to obtain the optimized interval sequence.

[0024] The flood merging criterion is as follows: if the valley flow between adjacent flood peaks is greater than a preset proportional threshold multiplied by the flow of the preceding flood peak, and the duration between peaks is less than a preset time threshold, then the adjacent flood peaks and the intermediate process will be merged into the same flood peak interval.

[0025] The flood segmentation criterion is as follows: if the flood valley flow drops to near the baseflow level or the interpeak duration exceeds the preset independent flood event time threshold, the flood process is segmented at the flood valley to define the flood valley interval.

[0026] Optionally, based on the interval sequence, a piecewise linearly varying time-series dynamic target water level process line is constructed, and storage compensation calculations are performed in conjunction with the predicted inflow process data to generate an initial reservoir capacity sequence including:

[0027] Based on the flood control level information of the reservoir, obtain the target water level range of the reservoir basin at each flood stage;

[0028] Based on the temporal characteristics of each flood peak interval, flood valley interval, and terminal interval in the interval sequence, the target water level of each interval is linearly adjusted to generate a time-series dynamic target water level process line, so as to guide the reservoir to perform peak interception and reduction in the flood peak interval, pre-release and empty the reservoir in the flood valley interval, and restore the reservoir level in the terminal interval.

[0029] Based on the forecasted inflow data and the time-series dynamic target water level process line, the water balance equation is used to calculate the storage compensation and generate the initial reservoir capacity sequence for all scheduling periods.

[0030] Optionally, based on the temporal characteristics of each flood peak interval, flood trough interval, and terminal interval in the interval sequence, the target water level of each interval is linearly adjusted to generate a time-series dynamic target water level process line, including:

[0031] Based on the start and end times t of the interval in which the current time t is located. start and t end Obtain the time progress factor ;

[0032] Based on the interval type and the time progress factor, the target water level is linearly adjusted:

[0033] Within the flood peak range, the target water level Z will be... t The water level rises linearly from the initial moment of the interval to the flood control high water level Z. max Dynamic target water level The calculation formula is:

[0034] ;

[0035] Within the flood valley area, the target water level Z will be... t The water level drops linearly from the initial moment of the interval to the flood control level Z. min Dynamic target water level The calculation formula is:

[0036] ;

[0037] Within the final interval, the target water level Z will be... t Linear regression from the water level at the beginning of the interval to the user-specified final water level Z of the scheduling target. tar Dynamic target water level The calculation formula is:

[0038] .

[0039] Optionally, under the constraint of the boundary storage capacity of adjacent time periods, the storage capacity relationship within the boundary storage capacity range is locally linearized. Then, the preset stepwise optimization objective function is transformed into a quadratic function with the storage capacity of the intermediate time period to be optimized as the variable. The quadratic function is then solved analytically to obtain the optimal storage capacity for the current time period, including:

[0040] Obtain the upstream and downstream boundary storage capacities of the current time period as boundary storage capacity constraints for adjacent time periods;

[0041] Under the constraint of reservoir capacity at the boundary of adjacent time periods, the objective function for stepwise optimization of the reservoir capacity in the current time period is constructed with the goal of minimizing the weighted sum of the water level deviation term and the outflow smoothing term. The water level deviation term is the squared difference between the water level corresponding to the reservoir capacity at the end of the current time period and the target water level in the time-series dynamic target water level process line in the current time period, and the outflow smoothing term is the squared difference between the outflow in the current time period and the outflow in the previous time period.

[0042] Substituting the water balance equation of the outflow smoothing term and the linear equation of the reservoir capacity curve for the current period into the objective function, we obtain a quadratic function of the reservoir capacity for the current period.

[0043] By utilizing the convexity of a quadratic function, the optimal storage capacity for the current time period can be obtained analytically by taking the derivative and setting it to zero.

[0044] Based on the closed interval formed by the dead storage capacity and flood control capacity of the reservoir, the optimal storage capacity is boundary-trimmed to obtain the feasible optimal storage capacity for the current period.

[0045] Optionally, the optimal reservoir capacity for each time period is solved by traversing all scheduling periods, and the initial reservoir capacity sequence is replaced with the obtained updated reservoir capacity sequence for multiple rounds of iteration until the preset convergence condition is met. The output reservoir scheduling scheme generated from the optimized reservoir capacity sequence includes:

[0046] Using the initial storage capacity sequence as the initial solution for the current iteration, the optimal storage capacity for each time period is solved sequentially in chronological order to obtain the updated storage capacity sequence after one complete traversal.

[0047] Get the maximum relative change between the updated storage capacity sequence and the storage capacity sequence before this traversal;

[0048] If the maximum relative change is not less than the preset convergence threshold, the updated storage capacity sequence will be used as the initial solution for the next traversal, and the optimization process will be repeated for all time periods.

[0049] If the maximum relative change is less than the preset convergence threshold, convergence is determined and iteration stops, and the current updated storage capacity sequence is output as the optimized storage capacity sequence;

[0050] The optimized reservoir capacity sequence is converted into the corresponding reservoir water level process line and outflow process line, generating a reservoir scheduling scheme that includes time-period water level control values ​​and discharge operation instructions.

[0051] Secondly, embodiments of the present invention provide a reservoir flood control and peak shaving scheduling system based on time-series dynamic target water level and analytical POA optimization, comprising:

[0052] The baseflow segmentation module is used to segment the acquired forecast inflow data of the reservoir basin to obtain the start and end times of flood events.

[0053] The flood peak and valley identification module is used to identify and divide the flood peaks and valleys within the start and end time based on the flood valley flow and inter-peak duration of a flood event, and generate an interval sequence containing the flood peak interval, flood valley interval and end interval.

[0054] The dynamic target water level and initial reservoir capacity generation module is used to construct a piecewise linearly changing time-series dynamic target water level process line based on the interval sequence, and to perform storage compensation calculations in conjunction with the forecast inflow process data to generate the initial reservoir capacity sequence.

[0055] The POA optimization module is used to perform local linearization of the storage capacity relationship within the boundary storage capacity range under the constraint of the boundary storage capacity of adjacent time periods. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the storage capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal storage capacity of the current time period.

[0056] The iterative convergence and scheduling output module is used to traverse all scheduling periods to solve for the optimal reservoir capacity for each period, and replace the initial reservoir capacity sequence with the obtained updated reservoir capacity sequence for multiple rounds of iteration until the preset convergence condition is met, and output the reservoir scheduling scheme generated by the optimized reservoir capacity sequence.

[0057] Thirdly, embodiments of the present invention provide an electronic device, comprising:

[0058] At least one processor;

[0059] and memory that is communicatively connected to at least one processor;

[0060] The memory stores instructions that can be executed by at least one processor. These instructions are executed by at least one processor to enable at least one processor to execute the reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization described above.

[0061] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing computer-executable instructions. When executed by a processor, the computer-executable instructions implement the reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization described above.

[0062] (III) Beneficial Effects

[0063] The present invention discloses a reservoir flood control and peak reduction scheduling method based on time-series dynamic target water level and analytical POA optimization. By adopting a time-series dynamic target water level process line with piecewise linear changes based on peak-valley identification, compared with the existing scheduling method with fixed target water level, it can dynamically adjust the scheduling target according to the real-time evolution of the flood. It guides the reservoir to intercept and reduce the peak during the flood peak period and guides the reservoir to pre-release water to make room during the flood valley period. This makes full use of the reservoir's flood control capacity, improves the peak reduction effect, and avoids insufficient reservoir capacity when subsequent flood peaks arrive.

[0064] Secondly, this invention transforms the subproblems of the stepwise optimization algorithm into univariate quadratic functions and directly solves for the optimal reservoir capacity using analytical methods. Compared with existing optimization methods that rely on numerical iteration, it can eliminate the iterative calculation process, reduce the complexity of single-step solution to the constant level, and achieve second-level generation of full-field scheduling schemes to meet the real-time emergency response requirements for sudden floods.

[0065] Furthermore, this invention employs a multi-criteria automatic identification method based on flood valley flow and inter-peak duration to divide flood peaks and valleys. Compared to methods that rely on human experience for judgment, it can achieve automated analysis of complex flood process structures, accurately define the timing of storage and release, and reduce the reliance on dispatchers' experience, thereby improving the scientific nature and consistency of dispatching decisions. Attached Figure Description

[0066] Figure 1 This is a flowchart illustrating a reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization, provided in an embodiment of the present invention.

[0067] Figure 2 This is a schematic diagram illustrating the comparison and analysis of outbound and inbound flow rates according to an embodiment of the present invention;

[0068] Figure 3 This is a schematic diagram illustrating the comparative analysis of reservoir water level processes according to an embodiment of the present invention.

[0069] Figure 4 This is a schematic diagram illustrating the comparative analysis of outbound flow processes provided in an embodiment of the present invention. Detailed Implementation

[0070] To better explain and facilitate understanding of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0071] refer to Figures 1 to 4 As shown in the embodiment of the present invention, a reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization is proposed. The method includes: dividing the baseflow of the acquired forecast inflow data of the reservoir basin to obtain the start and end times of the flood event; identifying and dividing the flood peaks and valleys within the start and end times based on the valley flow and inter-peak duration of the flood event, generating an interval sequence containing the peak interval, valley interval, and end interval; constructing a piecewise linearly varying time-series dynamic target water level process line based on the interval sequence, and performing storage compensation calculations in conjunction with the forecast inflow data to generate... The initial reservoir capacity sequence is generated. Under the constraint of the boundary reservoir capacity of adjacent time periods, the reservoir capacity relationship within the boundary reservoir capacity range is locally linearized. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the reservoir capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal reservoir capacity of the current time period. The optimal reservoir capacity of each time period is solved by traversing all scheduling time periods. The updated reservoir capacity sequence is then used to replace the initial reservoir capacity sequence for multiple rounds of iteration until the preset convergence condition is met. Finally, the reservoir scheduling scheme generated by the optimized reservoir capacity sequence is output.

[0072] This embodiment, by employing a segmented linearly varying time-series dynamic target water level process line based on peak-valley identification, can dynamically adjust the scheduling target according to the real-time evolution of the flood, compared to existing fixed target water level scheduling methods. It guides reservoirs to intercept and reduce peak flows during flood peak periods and to pre-release water during flood valley periods, thereby fully utilizing reservoir flood control capacity, improving peak reduction effects, and avoiding insufficient capacity during subsequent flood peaks. Secondly, this embodiment transforms the subproblems of the stepwise optimization algorithm into univariate quadratic functions and directly solves for the optimal reservoir capacity analytically. Compared to existing optimization methods that rely on numerical iteration, this eliminates the iterative calculation process, reducing the complexity of single-step solutions to constant levels, achieving second-level generation of scheduling plans for the entire flood event, and meeting the real-time emergency response requirements for sudden floods. Furthermore, this embodiment uses a multi-criteria automatic identification method based on flood valley flow and inter-peak duration to divide flood peaks and valleys. Compared to methods relying on human experience, this enables automated analysis of complex flood process structures, accurately defining the timing of water storage and release, reducing the reliance on dispatcher experience, and improving the scientific nature and consistency of scheduling decisions.

[0073] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention can be understood more clearly and thoroughly, and that the scope of the present invention can be fully conveyed to those skilled in the art.

[0074] Specifically, refer to Figure 1 As shown in the figure, the reservoir flood control and peak shaving scheduling method proposed in this embodiment based on time-series dynamic target water level and analytical POA optimization may include the following steps S100 to S500:

[0075] S100. Perform baseflow segmentation on the acquired forecast inflow data of the reservoir basin to obtain the start and end times of the flood event.

[0076] In this embodiment, step S100 may include the following sub-steps S110 to S130:

[0077] S110. Obtain the forecast inflow process data of the reservoir basin, and conduct basin scale analysis to obtain the scale characteristic parameters of the reservoir basin.

[0078] S120. Based on the scale and characteristic parameters of the reservoir basin, select the corresponding baseflow segmentation method from the preset baseflow segmentation method library to calculate the baseflow segmentation threshold and obtain the segmentation flow threshold of the reservoir basin. The baseflow segmentation methods include the average flow method, the flood peak percentage method, and the change point detection method.

[0079] Furthermore, this embodiment employs three adaptive baseflow segmentation methods: the average flow rate method, the peak flow percentage method, and the change point detection method. Average flow rate method: segmentation flow rate threshold. ,in, The average flow rate over the entire process. Peak flow percentage method: segmented flow threshold. Where p is the percentage of the flood peak, ranging from 10% to 30%, and Q is... peak Maximum instantaneous flow rate during a flood. Change point detection method: Determination of the starting point of the flood: Rate of change of flow rate. Q in The inflow rate is α, and the rate of change threshold is α; the endpoint of the drainage is determined as follows: Furthermore, the flow rate drops back to the baseflow level. When the reservoir basin is relatively small, a larger α and a shorter time threshold are used; when the reservoir basin is relatively small, a smaller α and the baseflow percentage method are used.

[0080] S130. Compare the forecast inflow process data with the segmented flow threshold. Based on the comparison results, divide the forecast inflow process of the reservoir basin into a baseflow part and a flood part, and take the period when the flow rate is higher than the segmented flow threshold as the start and end time of the flood event.

[0081] S200: Based on flood event valley flow and interpeak duration, identify and divide flood peaks and valleys within the start and end time, and generate interval sequences containing flood peak intervals, valley intervals and end intervals.

[0082] This embodiment addresses the technical problem in existing technologies where the division of peak and valley intervals in complex flood processes relies on manual experience and lacks automated identification methods, leading to inaccurate timing of flood control operations. This embodiment employs a multi-criteria automatic identification strategy that integrates flood valley flow and peak duration to perform refined analysis of the peak and valley structure of the flood process. It automatically determines whether adjacent flood peaks belong to the same complex flood event or should be separated into independent flood events, thereby accurately defining the timing of water storage for peak shaving and water release for reservoir emptying. Specifically, step S200 includes the following sub-steps S210 to S220:

[0083] S210. Based on the flood valley flow and interpeak duration of the flood event, the flood merging criterion and flood division criterion are used to identify and divide the flood peaks and valleys within the start and end time, and generate an initial interval sequence containing the flood peak interval, flood valley interval and end interval.

[0084] Furthermore, when a flood event comprises multiple flood peaks, it is necessary to determine whether it is a single complex flood event or multiple independent events: the flood merging criterion is that if the trough discharge Q between adjacent flood peaks... valley The value greater than the preset proportional threshold β (β=0.6~0.8, no significant decline) multiplied by the preceding peak flow Q peak，prev Q valley >β×Q peak，prev And the duration of the peak is T inter-peak Less than the preset time threshold T threshold (T threshold (This can be the confluence time of the watershed or a specified value), T inter-pea <T threshold Then, adjacent flood peaks and intermediate processes are merged into the same flood peak interval. The flood division criterion is that if the flood valley flow drops back to near the baseflow level Q... base , If the duration between peaks exceeds the preset independent flood event time threshold, the flood process will be divided at the flood valley to define the flood valley interval.

[0085] S220. Based on the set constraints including minimum peak spacing, minimum peak amplitude, and minimum peak duration, the peak intervals in the initial interval sequence are filtered out, and the peak intervals that do not meet the constraints are removed to obtain the optimized interval sequence.

[0086] For example, the minimum peak spacing is used to eliminate peak noise interference caused by short-term flow fluctuations. For instance, the time difference between adjacent peaks can be set to be greater than 3-6 hours, or greater than 0.5 times the basin confluence time. Peak intervals that do not meet this spacing requirement are eliminated. The minimum peak amplitude is used to eliminate small flow oscillations with excessively small peak-to-valley differences. For instance, the difference between the peak flow and the adjacent valley flow can be set to be greater than 5% or 10% of the peak flow. Peak intervals with peak-to-valley differences below this amplitude threshold are eliminated. The minimum peak duration is used to eliminate instantaneous flow disturbances with excessively short durations. For instance, the duration for which the flow within a peak interval remains above a preset threshold can be set to be greater than a predetermined time value. Instantaneous peaks whose duration does not meet this requirement are eliminated. Through the filtering of these constraints, noise interference during the flood process can be effectively filtered out, ensuring that the identified peak intervals truly reflect the main rise and fall characteristics of the flood, providing a more reliable basis for the subsequent segmentation of dynamic target water levels.

[0087] S300. Based on the interval sequence, construct a segmented linearly changing time-series dynamic target water level process line, and combine it with the forecast inflow process data to perform storage compensation calculation and generate an initial reservoir capacity sequence.

[0088] This embodiment addresses the technical problem in existing technologies where fixed target water levels are difficult to adapt to the dynamic evolution of flood processes, leading to improper timing of peak shaving and reservoir pre-release. This embodiment employs a piecewise linear interpolation strategy based on flood peak and valley identification. This strategy dynamically adjusts the scheduling target according to the real-time evolution of the flood, guiding reservoirs to intercept and reduce peak flows during peak periods and guiding them to pre-release water during valley periods. This fully utilizes the reservoir's flood control capacity, improves peak shaving effectiveness, and avoids insufficient capacity when subsequent flood peaks arrive. Specifically, step S300 includes the following sub-steps S310 to S330:

[0089] S310. Based on the flood control level information of the reservoir, obtain the target water level range of the reservoir basin at each flood stage.

[0090] The flood control level information includes the high flood control level, the flood limit level, and the final control level. The high flood control level corresponds to the highest control level that can be reached during the flood peak period, the flood limit level corresponds to the target pre-discharge level that is expected to be lowered during the flood valley period, and the final control level corresponds to the final control level that needs to be returned to during the final control period.

[0091] S320. Based on the temporal characteristics of each flood peak interval, flood valley interval, and terminal interval in the interval sequence, the target water level of each interval is linearly adjusted to generate a time-series dynamic target water level process line, so as to guide the reservoir to perform peak interception and reduction in the flood peak interval, pre-release and empty the reservoir in the flood valley interval, and restore the reservoir level in the terminal interval.

[0092] Furthermore, for the peak flood range, the target water level is set to rise linearly from the current water level at the start of the range to the flood control high water level; for the valley flood range, the target water level is set to fall linearly from the current water level at the start of the range to the flood control limit water level; for the final range, the target water level is set to linearly regress from the current water level at the start of the range to the final dispatch water level. The above linear adjustment is calculated using interpolation based on a time progress factor, which is defined as the proportion of the current time t within its respective range. , t start and t end This sets the start and end times of the current interval, and then adjusts the target water level linearly based on the interval type and time progress factor:

[0093] Within the flood peak range, the target water level Z will be... t The water level rises linearly from the initial moment of the interval to the flood control high water level Z. max Dynamic target water level The calculation formula is:

[0094] .

[0095] Within the flood valley area, the target water level Z will be... t The water level drops linearly from the initial moment of the interval to the flood control level Z. min Dynamic target water level The calculation formula is:

[0096] .

[0097] Within the final interval, the target water level Z will be... t Linear regression from the water level at the beginning of the interval to the user-specified final water level Z of the scheduling target. tar Dynamic target water level The calculation formula is:

[0098] .

[0099] S330. Based on the forecasted inflow process data and the time-series dynamic target water level process line, the water balance equation is used to perform storage compensation calculations to generate the initial reservoir capacity sequence for all scheduling periods.

[0100] Furthermore, firstly, based on the reservoir water level Z... t and the final water level of the scheduling target Z tar Calculate the storage capacity difference ΔV:

[0101] ;

[0102] In the formula, V tar For the corresponding Z tar Reservoir capacity, V tFor the corresponding Z t The reservoir's capacity.

[0103] Subsequently, based on the reservoir capacity difference and the predicted inflow rate I for that period, a water balance calculation is performed according to a preset scheduling calculation step size to determine the outflow rate Q for that period.

[0104] ;

[0105] In the formula, step is the scheduling calculation step size.

[0106] Finally, the calculation is performed time-by-time in the manner described above. After the calculation of each time period is completed, the reservoir capacity status is updated, and the updated reservoir capacity is used as the starting reservoir capacity for the next time period. This process is repeated until the calculation of all scheduling time periods is completed, thereby obtaining the initial reservoir capacity sequence that satisfies the time-series dynamic target water level process line constraint.

[0107] S400. Under the constraint of the boundary storage capacity of adjacent time periods, the storage capacity relationship within the boundary storage capacity range is locally linearized. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the storage capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal storage capacity of the current time period.

[0108] This embodiment addresses the technical problem in existing reservoir scheduling optimization algorithms that suffer from high computational complexity, slow convergence, and difficulty in meeting real-time emergency response requirements due to reliance on numerical iteration. This embodiment transforms the subproblems of the Progressive Optimization Algorithm (POA) into univariate quadratic functions and directly solves for the optimal reservoir capacity analytically. Compared to existing optimization methods that rely on numerical iteration, this eliminates the iterative calculation process, reduces the single-step solution complexity to constant level, and achieves second-level generation of the entire scheduling scheme, meeting the real-time emergency response requirements for sudden floods. Specifically, step S400 includes the following sub-steps S410 to S450:

[0109] S410. Obtain the upstream and downstream boundary storage capacities of the current time period as boundary storage capacity constraints for adjacent time periods.

[0110] Furthermore, the successive optimization algorithm (POA) is adopted to transform the global optimization into a two-stage optimization problem: fixing the boundary storage capacity V of adjacent time periods. t V t+2 Optimize intermediate time storage capacity V t+1 .

[0111] S420. Under the constraint of reservoir capacity at the boundary of adjacent time periods, construct a stepwise optimization objective function for the reservoir capacity in the current time period with the goal of minimizing the weighted sum of the water level deviation term and the outflow smoothing term. The water level deviation term is the squared difference between the water level corresponding to the reservoir capacity at the end of the current time period and the target water level in the time-series dynamic target water level process line in the current time period, and the outflow smoothing term is the squared difference between the outflow in the current time period and the outflow in the previous time period.

[0112] Furthermore, the objective function is gradually optimized as follows:

[0113] ;

[0114] In the formula, the first term is the water level deviation term (tracking the dynamic target water level), the second term is the outflow smoothing term (avoiding drastic flow fluctuations), ω1 and ω2 are target weight coefficients, and Q t+1 Q t Z represents the reservoir outflow at times t+1 and t. t Let t be the target water level of the reservoir.

[0115] S430. Substitute the water balance equation of the outflow smoothing term and the linear equation of the reservoir capacity curve for the current period into the objective function to obtain a quadratic function of the reservoir capacity for the current period.

[0116] Furthermore, The nonlinear storage capacity relationship within the interval is approximated by a quadratic function in one variable:

[0117] ;

[0118] In the formula, V t+2 V t Z represents the reservoir capacity at times t+2 and t; t+2 The target water level of the reservoir at time t+2.

[0119] Subsequently, ,Will as well as Substituting into the objective function for successive optimization, and expanding and rearranging, we arrive at a standard quadratic form:

[0120] ;

[0121] In the formula, I t+1 This represents the predicted inflow at time t+1.

[0122] S440. By utilizing the convexity of a quadratic function, the optimal storage capacity for the current time period can be analytically obtained by taking the derivative and setting it to zero.

[0123] Furthermore, regarding Setting the derivative to 0, we directly obtain the analytical solution as the optimal storage capacity for the current time period:

[0124] .

[0125] S450. Utilizing the convexity of a quadratic function, the optimal storage capacity for the current time period is obtained analytically by taking the derivative and setting it to zero.

[0126] Furthermore, the formula for calculating boundary clipping is:

[0127] ;

[0128] In equation (10), V min V max For Z respectively min Z max The corresponding reservoir capacity.

[0129] S500: Iterate through all scheduling periods to find the optimal reservoir capacity for each period, and replace the initial reservoir capacity sequence with the obtained updated reservoir capacity sequence for multiple rounds of iteration until the preset convergence condition is met, and output the reservoir scheduling scheme generated by the optimized reservoir capacity sequence.

[0130] In this embodiment, the approximation error introduced by the local linearization of the storage capacity curve is gradually corrected through multiple iterations, so that the optimized solution sequence gradually approaches the globally optimal scheduling scheme. Meanwhile, since the sub-problems of each time period in each iteration are solved directly using analytical methods, the computational complexity of a single iteration is low, and the convergence condition is usually met after a finite number of iterations, achieving rapid generation of the scheduling scheme for the entire event. Specifically, step S500 includes the following sub-steps S510 to S540:

[0131] S510. Using the initial storage capacity sequence as the initial solution for the current iteration, solve for the optimal storage capacity for each time period in chronological order to obtain the updated storage capacity sequence after one complete traversal.

[0132] Furthermore, from the beginning of the first scheduling period to the end of the last scheduling period, the boundary capacity of adjacent time periods is fixed for each time period. The analytical solution and boundary pruning described in step S400 are executed, and the capacity value at the corresponding position in the capacity sequence is updated with the optimization result of the current time period until all time periods have been traversed once.

[0133] S520. Obtain the maximum relative change between the updated storage capacity sequence and the storage capacity sequence before this traversal.

[0134] Furthermore, for each time period, the ratio of the absolute value of the difference between the updated storage capacity and the storage capacity before the update to the storage capacity before the update is calculated. The maximum value of this ratio in all time periods is taken as the maximum relative change, which is used to measure the convergence of this round of iteration.

[0135] S530a. If the maximum relative change is not less than the preset convergence threshold, the updated storage capacity sequence will be used as the initial solution for the next traversal, and the optimization process will be repeated for all time periods.

[0136] S530b: If the maximum relative change is less than the preset convergence threshold, convergence is determined and iteration stops. The current updated storage capacity sequence is output as the optimized storage capacity sequence.

[0137] S540. Convert the optimized reservoir capacity sequence into the corresponding reservoir water level process line and outflow process line to generate a reservoir scheduling scheme that includes time-by-time water level control values ​​and discharge operation instructions.

[0138] Furthermore, based on the reservoir's capacity-water level relationship curve, the optimized time-period reservoir capacity values ​​are converted into corresponding time-period water level values; and based on the water balance equation, the time-period outflow values ​​are inversely calculated from the time-period reservoir capacity changes and the predicted inflow. The final output is a scheduling scheme that can be directly executed by reservoir operation and management personnel or the automated control system.

[0139] In one specific embodiment, taking a typical double-peak flood event in a certain watershed as an example, the method of this embodiment is applied to reservoir flood control peak reduction scheduling, and compared and verified with the traditional fixed target water level scheduling method. The flood event selected in this embodiment lasts for approximately 350 calculation periods, including two distinct flood peak events. The first flood peak occurs around the 50th period, with a peak inflow of approximately 16,200 m³ / h. 3 / s; Subsequently, the floodwaters receded, forming a flood valley around the 100th period, with an inflow rate of approximately 4000 m³ / s at the bottom of the valley. 3 / s; The second flood peak occurred around the 150th period, with a peak inflow of approximately 16,000 m³ / s. 3 / s.

[0140] The method described in this embodiment is used to perform scheduling calculations for this flood event. First, steps S100 to S200 are executed to perform baseflow segmentation and peak-valley identification on the predicted inflow process. After identification, the flood event is divided into a first flood peak interval, a flood valley interval, a second flood peak interval, and a final interval. Among them, since the ratio of the flood valley flow to the first flood peak flow is approximately 0.25, which is less than the preset merging threshold of 0.6, and the flood valley duration exceeds 24 hours, the segmentation criterion is met. Therefore, the two flood peaks are determined to be relatively independent flood events, and a pre-release reservoir emptying strategy is implemented during the first flood peak receding phase.

[0141] Subsequently, step S300 is executed to construct a time-series dynamic target water level process line based on the identified interval sequence. During the first flood peak interval, the target water level linearly rises from 105 meters at the start of the rise to the flood control high water level of 107 meters; during the flood valley interval, the target water level linearly decreases from 107 meters to the flood control limit water level of 104 meters; during the second flood peak interval, the target water level again linearly rises from 104 meters to 107 meters; and during the final interval, the target water level gradually returns from 107 meters to the final dispatch water level of 105 meters. Based on the dynamic target water level process line, an initial reservoir capacity sequence is generated through storage compensation calculations.

[0142] Next, steps S400 to S500 are executed, employing an analytical stepwise optimization algorithm to iteratively optimize the initial reservoir capacity sequence. The weight coefficient for the water level deviation term is set to 1, the weight coefficient for the outflow smoothing term is set to 0.1, and the convergence threshold is set to 0.001. After six iterations, the maximum relative change in the reservoir capacity sequence decreases to 0.0008, satisfying the convergence condition, and the optimized scheduling scheme is output.

[0143] To verify the beneficial effects of this embodiment, the calculation results of the method in this embodiment are compared with the calculation results of the traditional fixed target water level method and measured data. Figures 2 to 4 As shown.

[0144] refer to Figure 2 The diagram showing the comparison and analysis of outbound and inbound flow is provided by [the relevant authority / organization]. Figure 2 Data shows that during the first flood peak period, the inflow reached its peak of approximately 16,200 m³ around the 50th period. 3 / s, while the calculated outflow rate during this period is approximately 10,000 m³. 3 / s, with a peak reduction rate of approximately 38%; in the flood valley section, the inflow drops back to approximately 4000-5000 m³ / s. 3 The outflow rate remained at a high level, exceeding the inflow rate at times, reflecting the proactive pre-emptive release of water to lower the reservoir's capacity. During the second flood peak period, the inflow rate reached a peak again around the 150th period, approximately 16,000 m³ / s. 3 / s, the calculated outflow is approximately 9500 m³ / s. 3 / s, with a peak reduction rate of approximately 41%. Comparing the calculated and measured data, it can be seen that the calculated outflow process and the measured outflow process have the same trend, and the deviation of the flow values ​​of each major scheduling node is within the acceptable range for engineering, further verifying the calculation accuracy and engineering practicality of the method in this embodiment.

[0145] refer to Figure 3The diagram showing the comparison and analysis of reservoir water level processes illustrates that the water level process calculated using the method in this embodiment matches the measured water level process well. During the first flood peak period, the calculated water level gradually rose from approximately 105 meters to approximately 107 meters, indicating that the reservoir effectively intercepted floodwaters during this stage, fully utilizing its flood control capacity to reduce peak flows. During the flood valley period, the calculated water level significantly decreased from approximately 107 meters to approximately 104 meters, indicating that the reservoir increased its discharge capacity during this stage, successfully pre-emptively creating storage capacity to prepare for the second flood peak. During the second flood peak period, the calculated water level rose again to approximately 107 meters, indicating that the reservoir again played a role in intercepting and reducing peak flows during the second flood peak. In the final period, the calculated water level steadily dropped back to the final dispatch level of approximately 105 meters. Throughout the entire dispatch process, the water level changes were smooth without drastic fluctuations, verifying the good operability of the dispatch scheme generated by the method in this embodiment.

[0146] refer to Figure 4 The diagram illustrating the comparative analysis of the outflow process shows that the outflow process calculated by the method in this embodiment maintains good consistency with the measured outflow process in terms of overall trend. During the peak flood period, when the inflow reaches its peak, the calculated outflow responds appropriately, reflecting the peak-shaving strategy. During the trough flood period, the inflow drops significantly, but the calculated outflow remains at a high level, actively increasing discharge to prepare for reservoir capacity. During the final stage of the receding flood, the calculated outflow decreases steadily, gradually returning to the target water level. Throughout the entire process, the trends of the calculated and measured values ​​are highly consistent, and the flow deviations at each major node are within the acceptable range for engineering purposes, verifying that the scheduling scheme generated by the method in this embodiment is reasonable and operable.

[0147] Secondly, this embodiment also provides a reservoir flood control and peak shaving scheduling system based on time-series dynamic target water level and analytical POA optimization, including:

[0148] The baseflow segmentation module is used to segment the baseflow of the acquired forecast inflow data of the reservoir basin to obtain the start and end times of flood events.

[0149] The flood peak and valley identification module is used to identify and divide the flood peaks and valleys within the start and end time based on the flood valley flow and inter-peak duration of a flood event, and generate an interval sequence containing the flood peak interval, flood valley interval and end interval.

[0150] The dynamic target water level and initial reservoir capacity generation module is used to construct a piecewise linearly changing time-series dynamic target water level process line based on the interval sequence, and to perform storage compensation calculations in conjunction with the forecast inflow process data to generate the initial reservoir capacity sequence.

[0151] The analytical POA optimization module is used to perform local linearization of the storage capacity relationship within the boundary storage capacity range under the constraint of the boundary storage capacity of adjacent time periods. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the storage capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal storage capacity of the current time period.

[0152] The iterative convergence and scheduling output module is used to traverse all scheduling periods to solve for the optimal reservoir capacity for each period, and replace the initial reservoir capacity sequence with the obtained updated reservoir capacity sequence for multiple rounds of iteration until the preset convergence condition is met, and output the reservoir scheduling scheme generated by the optimized reservoir capacity sequence.

[0153] Thirdly, this embodiment also provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute the reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization described above.

[0154] Fourthly, this embodiment also provides a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are executed by a processor, they implement the reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization described above.

[0155] In summary, this invention provides a reservoir flood control and peak shaving scheduling method, system, equipment, and medium based on time-series dynamic target water level and analytical POA optimization. Firstly, by integrating a multi-criteria automatic identification strategy that combines flood-valley flow ratio and inter-peak duration thresholds, automated and precise analysis of the peak-valley structure of complex flood processes is achieved. Peak, valley, and terminal intervals can be accurately defined without manual intervention, providing a reliable basis for subsequent dynamic scheduling, effectively reducing reliance on dispatcher experience, and improving the scientific rigor and consistency of scheduling decisions.

[0156] Next, a piecewise linear interpolation strategy based on flood peak and valley identification was used to construct a target water level process line that dynamically changes over time. This enables the reservoir to automatically guide peak interception and reduction during the flood peak period, automatically guide pre-release to reduce water levels during the flood valley period, and automatically return to the scheduled final water level during the final period. This dynamic target water level mechanism fundamentally solves the technical defects of poor adaptability of traditional fixed target water level strategies, significantly improves the utilization efficiency and peak reduction effect of reservoir flood control capacity, and effectively avoids the problem of insufficient response capacity to subsequent flood peaks due to untimely water release in multi-peak flood scenarios.

[0157] Then, by transforming the subproblems of the successive optimization algorithm into univariate quadratic functions and directly solving for the optimal reservoir capacity using analytical methods, the drawbacks of high computational overhead and slow convergence speed caused by traditional numerical iterative solutions are completely eliminated. This analytical solution method reduces the computational complexity of single-step optimization to the constant level, compressing the generation time of the entire flood control and dispatch plan to the second level, fully meeting the response speed requirements for real-time emergency decision-making in sudden flood scenarios.

[0158] Finally, by introducing a smoothing term into the objective function, the variation range of outflow between adjacent time periods is effectively constrained, avoiding violent fluctuations in the discharge process. The generated scheduling scheme has good engineering operability and is conducive to the stable operation of the gate equipment and the safe flood discharge of the downstream river channel.

[0159] Since the systems / devices described in the above embodiments of the present invention are systems / devices used to implement the methods of the above embodiments of the present invention, those skilled in the art can understand the specific structure and modifications of the systems / devices based on the methods described in the above embodiments of the present invention, and therefore will not be repeated here. All systems / devices used in the methods of the above embodiments of the present invention fall within the scope of protection of the present invention.

[0160] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0162] It should be noted that in the description of this invention, the word "a" or "an" preceding a component does not exclude the existence of multiple such components. This invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. The use of terms such as first, second, third, etc., is merely for convenience and does not indicate any order. These terms can be understood as part of the component names.

[0163] Furthermore, it should be noted that in the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0164] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning of the basic inventive concept, can make other changes and modifications to these embodiments.

[0165] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from the spirit and scope of the invention.

Claims

1. A reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization, characterized in that, include: The baseflow is segmented from the acquired forecast inflow data of the reservoir basin to obtain the start and end times of flood events; Based on the flood valley flow and interpeak duration of flood events, flood peaks and valleys are identified and divided within the start and end time, generating interval sequences that include flood peak intervals, flood valley intervals, and end intervals. Based on the interval sequence, a piecewise linearly varying time-series dynamic target water level process line is constructed, and storage compensation calculation is performed in conjunction with the predicted inflow process data to generate an initial reservoir capacity sequence; Under the constraint of the boundary capacity of adjacent time periods, the capacity relationship within the boundary capacity range is locally linearized. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal capacity of the current time period. The optimal reservoir capacity for each time period is solved by traversing all scheduling periods. The initial reservoir capacity sequence is replaced with the updated reservoir capacity sequence and multiple iterations are performed until the preset convergence condition is met. The reservoir scheduling scheme generated by the optimized reservoir capacity sequence is then output.

2. The method as described in claim 1, characterized in that, The baseflow is segmented from the acquired forecast inflow data for the reservoir basin to obtain the start and end times of flood events, including: Acquire forecast inflow data for the reservoir basin and conduct basin size analysis to obtain scale characteristic parameters of the reservoir basin; Based on the scale and characteristic parameters of the reservoir basin, the corresponding baseflow segmentation method is selected from the preset baseflow segmentation method library to calculate the baseflow segmentation threshold and obtain the segmentation flow threshold of the reservoir basin. The baseflow segmentation methods include the average flow method, the flood peak percentage method, and the change point detection method. The forecast inflow data is compared with the segmented flow threshold. Based on the comparison results, the forecast inflow process of the reservoir basin is divided into baseflow and flood. The period when the flow rate is continuously higher than the segmented flow threshold is taken as the start and end time of the flood event.

3. The method as described in claim 1, characterized in that, Based on flood valley flow and interpeak duration of flood events, flood peaks and valleys are identified and segmented within the start and end times, generating interval sequences containing peak intervals, valley intervals, and end intervals, including: Based on the flood valley flow and interpeak duration of flood events, flood peaks and valleys within the start and end times are identified and divided using flood merging and flood division criteria, generating an initial interval sequence containing flood peak intervals, flood valley intervals, and end intervals. Based on the set constraints including minimum flood peak interval, minimum flood peak amplitude, and minimum flood peak duration, the flood peak intervals in the initial interval sequence are filtered out, and flood peak intervals that do not meet the constraints are removed to obtain the optimized interval sequence. The flood merging criterion is as follows: if the valley flow between adjacent flood peaks is greater than a preset proportional threshold multiplied by the flow of the preceding flood peak, and the duration between peaks is less than a preset time threshold, then the adjacent flood peaks and the intermediate process will be merged into the same flood peak interval. The flood segmentation criterion is as follows: if the flood valley flow drops to near the baseflow level or the interpeak duration exceeds the preset independent flood event time threshold, the flood process is segmented at the flood valley to define the flood valley interval.

4. The method as described in claim 1, characterized in that, Based on the interval sequence, a piecewise linearly varying time-series dynamic target water level process line is constructed, and storage compensation calculations are performed in conjunction with the predicted inflow process data to generate an initial reservoir capacity sequence including: Based on the flood control level information of the reservoir, obtain the target water level range of the reservoir basin at each flood stage; Based on the temporal characteristics of each flood peak interval, flood valley interval, and terminal interval in the interval sequence, the target water level of each interval is linearly adjusted to generate a time-series dynamic target water level process line, so as to guide the reservoir to perform peak interception and reduction in the flood peak interval, pre-release and empty the reservoir in the flood valley interval, and restore the reservoir level in the terminal interval. Based on the forecasted inflow data and the time-series dynamic target water level process line, the water balance equation is used to calculate the storage compensation and generate the initial reservoir capacity sequence for all scheduling periods.

5. The method as described in claim 4, characterized in that, Based on the temporal characteristics of each flood peak interval, flood trough interval, and terminal interval in the interval sequence, the target water level of each interval is linearly adjusted to generate a time-series dynamic target water level process line, including: Based on the start and end times t of the interval in which the current time t is located. start and t end Obtain the time progress factor ; Based on the interval type and the time progress factor, the target water level is linearly adjusted: Within the flood peak range, the target water level Z will be... t The water level rises linearly from the initial moment of the interval to the flood control high water level Z. max Dynamic target water level The calculation formula is: ； Within the flood valley area, the target water level Z will be... t The water level drops linearly from the initial moment of the interval to the flood control level Z. min Dynamic target water level The calculation formula is: ； Within the final interval, the target water level Z will be... t Linear regression from the water level at the beginning of the interval to the user-specified final water level Z of the scheduling target. tar Dynamic target water level The calculation formula is: 。 6. The method as described in claim 1, characterized in that, Under the constraint of the boundary storage capacity of adjacent time periods, the storage capacity relationship within the boundary storage capacity range is locally linearized. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the storage capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved analytically to obtain the optimal storage capacity for the current time period, including: Obtain the upstream and downstream boundary storage capacities of the current time period as boundary storage capacity constraints for adjacent time periods; Under the constraint of reservoir capacity at the boundary of adjacent time periods, the objective function for stepwise optimization of the reservoir capacity in the current time period is constructed with the goal of minimizing the weighted sum of the water level deviation term and the outflow smoothing term. The water level deviation term is the squared difference between the water level corresponding to the reservoir capacity at the end of the current time period and the target water level in the time-series dynamic target water level process line in the current time period, and the outflow smoothing term is the squared difference between the outflow in the current time period and the outflow in the previous time period. Substituting the water balance equation of the outflow smoothing term and the linear equation of the reservoir capacity curve for the current period into the objective function, we obtain a quadratic function of the reservoir capacity for the current period. By utilizing the convexity of a quadratic function, the optimal storage capacity for the current time period can be obtained analytically by taking the derivative and setting it to zero. Based on the closed interval formed by the dead storage capacity and flood control capacity of the reservoir, the optimal storage capacity is boundary-trimmed to obtain the feasible optimal storage capacity for the current period.

7. The method as described in claim 1, characterized in that, The optimal reservoir capacity for each time period is calculated by traversing all scheduling periods. The updated reservoir capacity sequence is then used to replace the initial reservoir capacity sequence for multiple iterations until a preset convergence condition is met. The output reservoir scheduling scheme generated from the optimized reservoir capacity sequence includes: Using the initial storage capacity sequence as the initial solution for the current iteration, the optimal storage capacity for each time period is solved sequentially in chronological order to obtain the updated storage capacity sequence after one complete traversal. Get the maximum relative change between the updated storage capacity sequence and the storage capacity sequence before this traversal; If the maximum relative change is not less than the preset convergence threshold, the updated storage capacity sequence will be used as the initial solution for the next traversal, and the optimization process will be repeated for all time periods. If the maximum relative change is less than the preset convergence threshold, convergence is determined and iteration stops, and the current updated storage capacity sequence is output as the optimized storage capacity sequence; The optimized reservoir capacity sequence is converted into the corresponding reservoir water level process line and outflow process line, generating a reservoir scheduling scheme that includes time-period water level control values ​​and discharge operation instructions.

8. A reservoir flood control and peak shaving scheduling system based on time-series dynamic target water level and analytical POA optimization, characterized in that, include: The baseflow segmentation module is used to segment the acquired forecast inflow data of the reservoir basin to obtain the start and end times of flood events. The flood peak and valley identification module is used to identify and divide the flood peaks and valleys within the start and end time based on the flood valley flow and inter-peak duration of a flood event, and generate an interval sequence containing the flood peak interval, flood valley interval and end interval. The dynamic target water level and initial reservoir capacity generation module is used to construct a piecewise linearly changing time-series dynamic target water level process line based on the interval sequence, and to perform storage compensation calculations in conjunction with the forecast inflow process data to generate the initial reservoir capacity sequence. The POA optimization module is used to perform local linearization of the storage capacity relationship within the boundary storage capacity range under the constraint of the boundary storage capacity of adjacent time periods. Then, the preset stepwise optimization objective function is transformed into a univariate quadratic function with the storage capacity of the intermediate time period to be optimized as the variable. The univariate quadratic function is then solved by analytical differentiation to obtain the optimal storage capacity of the current time period. The iterative convergence and scheduling output module is used to traverse all scheduling periods to solve for the optimal reservoir capacity for each period, and replace the initial reservoir capacity sequence with the obtained updated reservoir capacity sequence for multiple rounds of iteration until the preset convergence condition is met, and output the reservoir scheduling scheme generated by the optimized reservoir capacity sequence.

9. An electronic device, characterized in that, include: At least one processor; and memory that is communicatively connected to at least one processor; The memory stores instructions that can be executed by at least one processor, which are executed by at least one processor to enable the at least one processor to perform a reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization as described in any one of claims 1-7.

10. A computer-readable storage medium storing computer-executable instructions thereon, characterized in that, When executed by a processor, the computer-executable instructions implement the reservoir flood control and peak shaving scheduling method based on time-series dynamic target water level and analytical POA optimization as described in any one of claims 1-7.

Images