Reservoir optimal operation method based on inter-period sensitivity sparse repair and related products

Abstract

This invention relates to the field of water resource optimization scheduling technology in water conservancy and hydropower engineering, specifically to a reservoir optimization scheduling method and related products based on intertemporal sensitivity sparse repair. The method involves constructing and solving implicit equations for power generation and water volume, establishing an evaluation system and calculating intertemporal sensitivity to obtain a set of key time periods. Then, sparse repair is implemented on the initial power generation plan within this set, and the repair operation is embedded in an iterative optimization framework. This invention avoids the use of empirical estimation or coarse trial calculations in traditional reservoir scheduling, enhances the water resource safety net of reservoirs, suppresses the oscillation phenomenon of traditional intelligent evolutionary algorithms, and maintains the scheduling intent and temporal consistency of the initial predetermined plan on the power grid side.

Description

Technical Field

[0001] This invention relates to the field of water resource optimization scheduling technology in water conservancy and hydropower projects, specifically to a reservoir optimization scheduling method and related products based on intertemporal sensitivity sparse repair, and particularly to a reservoir optimization scheduling method for coordinating monthly power generation plans and water supply plans under the principle that power dispatch follows water dispatch. Background Technology

[0002] As a core project for the rational allocation of water resources and the development and utilization of hydropower resources, reservoirs typically undertake multiple comprehensive utilization tasks such as flood control, water supply (including multi-dimensional water needs such as ecology, urban areas, and agriculture), and power generation.

[0003] In actual dispatch management, power generation plans are usually issued by the power grid side, while water supply plans are proposed by the water administration department or water users. There is often a mismatch between the two in terms of total amount and process: within the same period, power generation plans tend to operate at high water levels to improve efficiency, while water supply plans have certain guarantee targets and emphasize rigid demand during the period.

[0004] To reconcile the aforementioned contradictions, current practical dispatching generally follows the basic principle of "power dispatching subordinate to water dispatching." Existing methods mostly employ heuristic evolutionary algorithms to search for monthly power generation plans, and then calculate the power generation water volume and water supply allocation for each candidate plan. These methods still face three key engineering challenges: the intertemporal coupling problem, the minimum perturbation problem, and the implicit solution problem. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a reservoir optimization scheduling method and related products based on intertemporal sensitivity sparse repair. Under the constraints of reservoir capacity and power generation water intake, water supply priority, and the principle that power dispatch follows water dispatch, this method automatically identifies the key periods and directions for adjustment to address the mismatch between monthly power generation plans and water supply plans. It eliminates or significantly reduces ecological / urban / agricultural water shortage damage with minimal disturbance and outputs an optimal power generation plan that meets the executable boundaries.

[0006] This invention is achieved through the following technical solution:

[0007] A reservoir optimization scheduling method based on intertemporal sensitivity sparse repair includes:

[0008] Acquire basic operational data, initial power generation plans, and water supply plans of the reservoir during the scheduling period;

[0009] Based on the aforementioned basic operational data, an implicit equation is constructed between power generation and water volume for power generation, and the water volume for power generation in each time period is obtained by solving the equation, and the reservoir capacity in each time period is obtained by recursion.

[0010] The available water volume is determined based on the reservoir capacity and power generation water volume in each time period, and the water supply is allocated according to the preset priority order to calculate the corresponding water shortage.

[0011] Construct a target vector for evaluating candidate power generation plans, calculate the intertemporal sensitivity of the water shortage to the planned power generation, and identify a set of critical repair periods accordingly; perform sparse repair operations on the candidate power generation plans within the set of critical repair periods, and the repair process satisfies preset perturbation constraints;

[0012] The sparse repair operation is embedded in the iterative optimization framework to update the power generation plan until the termination condition is met, and the optimal power generation plan and the corresponding reservoir scheduling process are output.

[0013] Optionally, the basic operational data includes: the inflow sequence, initial reservoir capacity, dead reservoir capacity, operational reservoir capacity boundary, minimum operating water level of the intake structure, leakage parameters, water level-reservoir capacity curve data, and tailwater level-flow curve data during the scheduling period;

[0014] The initial power generation plan includes: a monthly power generation plan;

[0015] The water supply plan includes: ecological water supply plan, urban water supply plan and agricultural water supply plan;

[0016] Specifically, the power generation water intake limit reservoir capacity is determined by reverse calculation based on the minimum working water level of the water intake structure and the water level-reservoir capacity curve data.

[0017] Alternatively, methods for solving for the amount of water used for power generation in each time period using implicit equations include:

[0018] Construct the first The implicit equations for power generation and hydropower generation over a given period are established, and their residual functions are determined. ,in, For power generation coefficient, For the first Water volume for power generation during the period For the first Duration of time This is a water level-reservoir capacity curve function. This is a tailwater level-discharge curve function. For the first Time-related head loss, For the candidate project's power generation, For the first Initial storage capacity for the period For the first Storage capacity at the end of the period;

[0019] The solution is obtained using a higher-order univariate iteration, and the iteration formula is as follows: ,in, Number the iteration steps. For the first Time period The amount of water generated in the next iteration For the first Time period The amount of water generated in the next iteration For the first Time period Implicit equation residuals at the next iteration and These are the implicit equation residual functions. The first and second derivatives;

[0020] When the implicit equation is satisfied, the absolute value of the residual Or reach the preset maximum number of iterations The iteration stops at a certain point, and the power generation water volume for each time period is obtained. This is the tolerance for residual convergence.

[0021] Optionally, the recursive method for calculating the storage capacity for each time period includes:

[0022] Based on the principle of water balance, the reservoir capacity is calculated recursively: ,in, For the first Initial storage capacity for the time period For the first The final storage capacity of the time period For the first The amount of water entering the reservoir during a given period For the first Water volume for power generation during the period For the first Non-power generation water discharge or supplementary water supply during the period, For the first Leakage volume over a period of time;

[0023] The recursive storage capacity for each time period must meet the following constraints: and ,in, and These are the lower and upper limits of the reservoir's operating capacity, respectively. Dead storage capacity of the reservoir;

[0024] The convergence and feasibility verification conditions are that the obtained power generation water volume and the corresponding reservoir capacity satisfy the boundary constraints of reservoir capacity and water volume.

[0025] Optionally, methods for determining the available water supply and calculating the water shortage include:

[0026] Based on the priority water supply mode for power generation tailwater, the first Water availability during the period : ,in, For the first Water volume for power generation during the period For the first The final storage capacity of the time period Reservoir capacity is limited due to water intake for power generation;

[0027] The preset priority order is as follows: ecology, urban, agriculture; and the first priority is calculated according to the aforementioned priority order. Ecological water supply during different time periods Urban water supply and agricultural water supply : ,in, , and The first Ecological water demand plan, urban water demand plan, and agricultural water demand plan for each time period;

[0028] Based on the calculated water supply, the corresponding ecological water shortage is calculated respectively. Urban water shortage and agricultural water shortage : .

[0029] Alternatively, methods for constructing the target vector include:

[0030] Based on the water shortage in each time period, the total number of time periods is calculated separately. Summary of ecological water shortage targets Urban water shortage summary target and summary targets for agricultural water shortage : ;

[0031] Based on the candidate power generation plan and the initial power generation plan, the target deviation from the plan is calculated. : ,in, For the first Candidate planned power generation for the specified time period For the first Initial baseline planned power generation for the time period;

[0032] Construct a lexicographically ordered target vector for evaluating candidate power generation plans. : ;

[0033] The optimal power generation plan is obtained by screening candidate power generation plans according to the principle that power dispatch follows water dispatch.

[0034] Optionally, methods for identifying the set of critical repair periods based on intertemporal sensitivity include:

[0035] Introducing a water-scarce weighted aggregation agent target Calculate the first Intertemporal sensitivity : ,in, , This represents the total number of time periods. , , The weights for ecological water shortage, urban water shortage, and agricultural water shortage are respectively, and they satisfy the following conditions: ; For the first Candidate power generation for the specified time period;

[0036] Based on the calculated inter-period sensitivity of each time period, a set of key repair time periods was identified. : ,in, To select the one with the largest absolute value An operator that assigns each element a time period number. The number of key time periods is preset.

[0037] Optionally, methods for performing sparse repair operations include:

[0038] The set of critical repair periods Within the process, a sparsity repair operation is performed on the candidate planned power generation to obtain the repaired planned power generation. : ,in, To repair the step size, This is the interval projection operator.

[0039] The repair process must simultaneously satisfy the following total disturbance constraint and single-month disturbance constraint: ,in, The total disturbance threshold, This is the monthly disturbance threshold.

[0040] A reservoir optimization scheduling method terminal based on intertemporal sensitivity sparse repair includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the reservoir optimization scheduling method based on intertemporal sensitivity sparse repair as described above.

[0041] A computer program product includes a computer program / instructions that, when executed by a processor, implement the reservoir optimization scheduling method based on intertemporal sensitivity sparse repair as described above.

[0042] Compared with the prior art, the present invention has the following features and beneficial effects:

[0043] This invention constructs and solves implicit equations for power generation and water generation, then establishes an evaluation system and calculates intertemporal sensitivity to obtain a set of key time periods. Subsequently, sparse repair is implemented on the initial power generation plan within the set, and the repair operation is embedded in an iterative optimization framework.

[0044] This invention avoids the use of empirical estimation or rough trial calculation in traditional reservoir scheduling by constructing and solving a nonlinear implicit equation between power generation and water volume, ensuring water balance and engineering boundary feasibility while guaranteeing algorithm execution efficiency. By constructing a target vector through water supply priority, it enhances the reservoir's ability to guarantee water resources. By calculating the intertemporal sensitivity of multi-target water shortage to planned power generation in each period to extract key repair periods and performing sparse repair operations, it suppresses the oscillation phenomenon of traditional intelligent evolutionary algorithms and maintains the scheduling intent and temporal consistency of the initial plan on the power grid side. Attached Figure Description

[0045] The accompanying drawings illustrate exemplary embodiments of the present invention and, together with the description thereof, serve to explain the principles of the invention. These drawings are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, but do not constitute a limitation on the embodiments of the present invention.

[0046] Figure 1 This is a detailed schematic diagram of a reservoir optimization scheduling method based on intertemporal sensitivity sparse repair according to the present invention.

[0047] Figure 2 This is a flowchart illustrating step S3 in Embodiment 2 of the present invention.

[0048] Figure 3 This is a flowchart illustrating steps S6 to S8 in Embodiment 2 of the present invention.

[0049] Figure 4 This is a schematic diagram of the overall process of a reservoir optimization scheduling method based on intertemporal sensitivity sparse repair according to the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0051] It should also be noted that, for ease of description, only the parts relevant to the present invention are shown in the accompanying drawings.

[0052] Where there is no conflict, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0053] Example 1

[0054] like Figure 4 As shown in the figure, this embodiment provides a reservoir optimization scheduling method based on intertemporal sensitivity sparse repair. This method is a closed-loop iterative process for diagnosing and controlling the repair of multi-objective (power generation and water supply) conflicts in reservoirs. The overall operation logic can be summarized as follows: data acquisition → physical process simulation (solving equations and reservoir capacity extrapolation) → water supply effect evaluation (allocation and calculation of water shortage) → diagnosis and fine-tuning (intertemporal sensitivity analysis and sparse repair) → closed-loop optimization (iteration until the optimal solution is output).

[0055] Sparse repair refers to the process of performing localized adjustments to a small number of critical periods with high intertemporal sensitivity after completing intertemporal sensitivity calculations and identifying the set of critical repair periods.

[0056] The operation steps of this embodiment are described below:

[0057] Step 1: Obtain the basic operational data, initial power generation plan, and water supply plan of the reservoir during the scheduling period.

[0058] Basic operational data refers to the objective physical parameters and historical / predictive data (such as inflow, water level-storage ratio, etc.) required to maintain reservoir hydraulics and hydrological calculations.

[0059] The initial power generation plan and the water supply plan refer to the power generation targets set in advance by the power grid department and the rigid water demand proposed by water users or management departments, respectively.

[0060] Step 2: Based on the basic operating data, construct an implicit equation between power generation and water volume for power generation, solve for the water volume for power generation in each time period, and recursively obtain the reservoir capacity for each time period.

[0061] In mathematics, implicit equations refer to equations where the independent and dependent variables are intertwined and cannot be directly written as an expression of "dependent variable = independent variable". In reservoir management, power generation, water volume for power generation, and water level (determined by reservoir capacity) are mutually causal, forming an implicit coupling.

[0062] Recursion refers to a calculation method that uses the state value of the previous period (such as the initial storage capacity) and the decision quantity of the current period (such as the inflow and outflow of water) to calculate the initial state value (end storage capacity) of the next period.

[0063] This step reconstructs the actual nonlinear hydrodynamic process of the reservoir by constructing implicit equations; and calculates the actual water consumption required to maintain the target power generation through mathematical solutions. Then, relying on the principle of water balance, the dynamic change process of the reservoir water level is calculated time by time.

[0064] Step 3: Determine the available water volume based on the reservoir capacity and power generation water volume for each time period, and allocate the water supply according to the preset priority order, and calculate the corresponding water shortage.

[0065] Given limited water resources, different water users (such as ecological, urban, and agricultural users) have different priorities in obtaining water resources. After obtaining the available water from the reservoir, water resources are allocated according to the established priority levels, and the degree of water shortage for each water user is quantified.

[0066] Step 4: Construct a target vector for evaluating candidate power generation plans, calculate the intertemporal sensitivity of the water shortage to the planned power generation, and identify the set of key repair periods accordingly.

[0067] Intertemporal sensitivity is used to measure the degree to which changes in power generation plans during a specific period affect the overall water shortage situation throughout the entire dispatch period.

[0068] Sparse repair operations are performed on the candidate power generation plans within the set of critical repair periods, and the repair process satisfies preset perturbation constraints.

[0069] Sensitivity analysis was used to identify a few key months (sets) that resulted in the most severe water shortages, and power generation plans were only modified for these specific months (sparseness), rather than blindly modifying all months. At the same time, strict upper and lower limits were set for the magnitude of modifications to the power generation plans to prevent over-modification.

[0070] Step 5: Embed the sparse repair operation into the iterative optimization framework to update the power generation plan until the termination condition is met, and output the optimal power generation plan and the corresponding reservoir scheduling process.

[0071] By repeating the process of "evaluation-modification-re-evaluation", the optimal solution is gradually approached, and the iteration stops when the termination condition is met (such as reaching the maximum number of calculations or the water shortage has been reduced to an allowable range).

[0072] This embodiment starts by collecting boundary data, first deriving the physical trajectory of water volume and reservoir capacity through implicit equations, and then allocating water resources according to priority. Subsequently, it uses intertemporal sensitivity indicators to pinpoint time points of water scarcity and performs sparse repair under perturbation constraints. The process is then iterated and looped within an optimization framework until the optimal scheduling scheme is obtained. This effectively resolves the spatiotemporal conflicts in hydropower planning, ensuring the minimum water supply demand for various purposes while maintaining the initial power generation plan to the greatest extent possible.

[0073] Example 2

[0074] This embodiment further refines and implements the technical solution in Embodiment 1, demonstrating the nine core operational steps (S1-S9) included in the optimized scheduling method, such as... Figure 1 As shown, the specific steps of this embodiment will be described in detail.

[0075] S1. Obtain data about the reservoir during the scheduling period.

[0076] Basic operational data includes: inflow sequence, initial reservoir capacity, dead reservoir capacity, operating reservoir capacity boundary, minimum working water level of intake structures, leakage parameters, water level-reservoir capacity curve data, and tailwater level-flow curve data during the scheduling period;

[0077] The initial power generation plan includes: a monthly power generation plan;

[0078] Water supply plans include: ecological water supply plans, urban water supply plans, and agricultural water supply plans;

[0079] Specifically, the power generation water intake limit reservoir capacity is determined by reverse calculation based on the minimum working water level of the water intake structure and the water level-reservoir capacity curve data.

[0080] In addition, to ensure the output of this embodiment is feasible in engineering and management, the minimum working water level of the water intake structure (used to inversely calculate the reservoir capacity for power generation water intake) and the grid-side planned boundary are given as external inputs to the algorithm and have a clear source:

[0081] Reservoir capacity for power generation water intake restrictions This is a restriction used to characterize the prohibition (or inappropriateness) of water intake for power generation when the upstream water level is lower than the minimum operating water level of the intake (or water intake structure). Let the minimum operating water level of the water intake structure be... The water level-reservoir capacity curve is ,but (Equation 1), where Given by engineering design data, equipment operating conditions, or operating procedures; when When the points are discrete curve points, It can be obtained through inverse interpolation or curve lookup.

[0082] Permissible monthly generation planning boundaries on the grid side This serves as the scheduling and management boundary, used to limit the scope of planned adjustments during the repair and search processes. Let the baseline plan be... The allowable upward / downward adjustment range is... , ,but (Equation 2), where and The power grid dispatching procedures, transaction / contract constraints, or assessment criteria are given as external inputs to ensure that the output plan of this invention is executable.

[0083] S2. Based on the inflow of water and the initial reservoir capacity, the initial reservoir capacity for each period is recursively obtained and the available water volume for power generation in each period is calculated. The feasible range of water volume for power generation is determined and used as the boundary for S3 root finding and feasibility verification. That is, under the premise of meeting the physical capacity of the reservoir (not exceeding the maximum operating capacity, nor lower than the restricted capacity), how much water can be released at most and at least in the current period.

[0084] Based on the principle of water balance, the reservoir capacity is calculated recursively: (Equation 3), where, For the first Initial storage capacity for the time period For the first The final storage capacity of the time period For the first The amount of water entering the reservoir during a given period For the first Water volume for power generation during the period For the first Non-power generation water discharge or supplementary water supply during the period, For the first Leakage volume over a period of time;

[0085] The recursive storage capacity for each time period must meet the following constraints: and ,in, and These are the lower and upper limits of the reservoir's operating capacity, respectively. Dead storage capacity of the reservoir;

[0086] The preferred method is to calculate the total available water supply under a coupled mode that prioritizes the supply of tailwater from power generation. Supplementary water supply is allowed when the final reservoir capacity exceeds the water intake limit for power generation. This determines the... Water availability during the period : (Equation 4), where, For the first Water volume for power generation during the period For the first The final storage capacity of the time period Reservoir capacity is limited due to water intake for power generation.

[0087] S3. Based on the water level-storage capacity curve function, tailwater level-flow curve function and head loss, an implicit equation for power generation-power generation water volume is constructed, and the power generation water volume for each time period is obtained by solving it through a high-order single-variable root-finding algorithm.

[0088] Construct the first The implicit equations for power generation and hydropower generation over a given period are established, and their residual functions are determined. (Equation 5), where, For power generation coefficient, For the first Water volume for power generation during the period For the first Duration of time This is a water level-reservoir capacity curve function. This is a tailwater level-discharge curve function. For the first Time-related head loss, For the candidate project's power generation, For the first Initial storage capacity for the period For the first Storage capacity at the end of the period; and It can be constructed from curve data through interpolation (such as Newton interpolation or spline interpolation), which is a common engineering implementation method.

[0089] The solution is obtained using a higher-order univariate iteration (such as Halley iteration), and the iterative formula is as follows: (Equation 6), where, Number the iteration steps. For the first Time period The amount of water generated in the next iteration For the first Time period The amount of water generated in the next iteration For the first Time period Implicit equation residuals at the next iteration and These are the implicit equation residual functions. The first and second derivatives;

[0090] The obtained power generation water volume and corresponding reservoir capacity satisfy the reservoir capacity and water volume boundary constraints (see Equation 3) as the convergence and feasibility verification conditions.

[0091] When the implicit equation is satisfied, the absolute value of the residual Or reach the preset maximum number of iterations The iteration stops at a certain point, and the power generation water volume for each time period is obtained. This is the tolerance for residual convergence.

[0092] like Figure 2 The execution process of this step is explained in detail as shown below.

[0093] S31, Obtain the first candidate (or pending repair) plan. Basic data include the planned power generation for the time period, the initial reservoir capacity for this time period from step S2, the water level-storage capacity curve, the tailwater level-flow curve, head loss, power generation coefficient, and time period length.

[0094] Subsequently, the feasible range of the current power generation water volume is determined by the available power generation water volume calculated in step S2, and this range is used as the boundary for subsequent root finding and the feasibility check boundary.

[0095] S32. Based on the above water level-storage capacity curve, tailwater level-flow curve and head loss, construct the first equation strictly according to the aforementioned formula (5). Implicit equation for power generation versus water generation during a given period.

[0096] S33. Select initial values ​​for the iteration. To improve the convergence speed, the initial values ​​are preferably the midpoint of the feasible interval, the feasible value generated by the previous generation algorithm, or the solution result of the previous time period; at the same time, initialize the iteration step number to... .

[0097] S34, Calculate the first The residual function value of the implicit equation at the current iteration point is calculated, and its first and second derivative values ​​are calculated simultaneously.

[0098] S35. Substitute the calculated residual function value and its first and second derivatives into the aforementioned equation (6), perform Halley iteration update, and obtain the next step of power generation water volume.

[0099] S36. Determine whether the absolute value of the residual of the current implicit equation satisfies the convergence condition (i.e., ).

[0100] If the convergence condition is not met: further determine the number of iterations. Has the preset maximum number of iterations been exceeded? .

[0101] If the time limit is not exceeded, return to step S34 to continue the next iteration calculation; if the time limit is exceeded, trigger the failure handling mechanism (e.g., retry by adjusting the initial iteration value, reducing the solution step size, or resetting the iteration process), and return to step S33.

[0102] If the convergence condition is met: proceed to the feasibility verification stage.

[0103] S37. The power generation water volume obtained by convergence is calculated and the boundary is checked according to the aforementioned formula (3) to determine whether it meets the preset reservoir capacity and water volume boundary.

[0104] If the boundary conditions are not met, it indicates that the obtained mathematical solution is physically infeasible, triggering the failure handling mechanism described above, and returning to step S33 for retry.

[0105] If the boundary conditions are met, the iterative process is successful, and the power generation and residual verification results for each time period are finally output. These results are then substituted back into the reservoir capacity recursion step S2 to support the rolling calculation of the state throughout the subsequent time period.

[0106] S4. Based on the available water volume for a given time period, calculate the ecological water supply, urban water supply, and agricultural water supply in the order of ecological priority, urban water supply, and agricultural water supply, and calculate the corresponding water shortage.

[0107] Calculate the first according to the priority order. Ecological water supply during different time periods Urban water supply and agricultural water supply : (Equation 7), where, , and The first Ecological water demand plan, urban water demand plan, and agricultural water demand plan for each time period;

[0108] Based on the calculated water supply, the corresponding ecological water shortage is calculated respectively. Urban water shortage and agricultural water shortage : (Equation 8).

[0109] S5. Construct a target vector for lexicographical comparison of power dispatch according to water dispatch, and evaluate candidate power generation plans. Lexicographical comparison is a multi-objective evaluation mechanism, like looking up words in a dictionary, with an absolute order (e.g., compare 'a' first, then compare 'b' if 'a' is the same). That is, by constructing a target vector and performing lexicographical comparison, the algorithm will absolutely prioritize assessing the water supply gap when evaluating the merits of the plans. Only when the water supply security level is the same will the completion rate of the power generation plan be compared.

[0110] Based on the water shortage in each time period, the total number of time periods is calculated separately. Summary of ecological water shortage targets Urban water shortage summary target and summary targets for agricultural water shortage : (Equation 9);

[0111] Based on the candidate power generation plan and the initial power generation plan, the target deviation from the plan is calculated. : ,in, For the first Candidate planned power generation for the specified time period For the first Initial baseline planned power generation for the time period;

[0112] Construct a lexicographically ordered target vector for evaluating candidate power generation plans. : (Equation 10);

[0113] The candidate power generation plan is selected to obtain the optimal plan according to the principle that power dispatch follows water dispatch. That is, the optimal plan is selected by first minimizing the power generation plan, and then comparing the plans if they are equal (or the difference is less than the tolerance). The optimal plan is selected by comparing the plans when all three types of water shortage are equal.

[0114] The specific principle of ESC conforming to SC can be:

[0115] Prioritize minimizing the total ecological water shortage target ;

[0116] like If the values ​​are equal or the difference is less than the preset tolerance, then the aggregated target for urban water shortage is compared and minimized. ;

[0117] like If the values ​​are equal or the difference is less than the preset tolerance, then the aggregated agricultural water shortage target is compared and minimized. ;

[0118] When the combined targets for the three types of water shortage are all equal or the differences are all less than the preset tolerance, the final comparison and minimization of the planned deviation target will be performed. .

[0119] S6. Calculate the intertemporal sensitivity of the water shortage target to the monthly planned power generation, and identify the set of critical repair periods accordingly. If the water supply is insufficient in a certain month, it may be due to excessive water release in the previous months. Therefore, by using intertemporal sensitivity, it is possible to determine which months' planned power generation led to the water shortage.

[0120] To facilitate sensitivity calculation, a water-scarce weighted aggregation proxy target is introduced. , (Equation 11) , , The weights for ecological water shortage, urban water shortage, and agricultural water shortage are respectively, and they satisfy the following conditions: .

[0121] Calculate the first Intertemporal sensitivity : (Equation 12), where, , This represents the total number of time periods. For the first Candidate power generation for the specified time period; The calculation can be performed by recalculating S2 to S5 of the candidate plan under a small disturbance in a single month to obtain a difference approximation, or by combining the implicit solution process of equation (5) to obtain analytical / semi-analytical direction information.

[0122] Based on the calculated inter-period sensitivity of each time period, a set of key repair time periods was identified. (Source of sparsity): (Equation 13), where, To select the one with the largest absolute value An operator that assigns each element a time period number. The number of key time periods is preset.

[0123] like Figure 3 As shown, the underlying implementation includes an "inner loop" structure for differential calculation.

[0124] S61. Based on the various water shortage indicators obtained in step S5, construct the water shortage summary proxy target according to the aforementioned formula (11);

[0125] S62. For each month within the scheduling period A small perturbation (preferably a differential perturbation) is applied to the current planned power generation. After applying the perturbation, the evaluation process of steps S2 to S5 is repeated once for the perturbation plan (i.e., the reservoir capacity recursion, implicit root finding, and water supply allocation are re-performed), thereby calculating the first... Monthly inter-period sensitivity (according to Equation 12); iterate through all months to obtain the sensitivity sequence for the entire time period;

[0126] S63. Take the absolute value of the sensitivity sequence calculated above, and use the TopK operator to truncate the sequence with the highest absolute sensitivity. These months constitute a set of key repair periods. (According to formula 13).

[0127] S7. Only within the set of critical repair periods, execute the sparse repair operator on the monthly planned power generation to generate a repaired power generation plan, satisfying both total disturbance and single-month disturbance constraints. Sparse repair refers to, after completing the inter-period sensitivity calculation and identifying the set of critical repair periods, not making overall adjustments to the planned power generation for all periods throughout the entire scheduling period, but only making local corrections for a small number of critical periods with high inter-period sensitivity; for other periods not identified as critical repair periods, keep their candidate planned power generation unchanged. Specifically, let the set of critical repair periods be... When the t-th time period belongs to When the planned power generation for that period is adjusted according to inter-period sensitivity, repair step size, and planning boundary constraints; when the t-th period does not belong to No corrections are made at that time. "Sparse" means that the number of time periods actually adjusted is less than the total number of time periods in the scheduling period, and the total amount of repair and the amount of repair per time period are both limited by preset disturbance constraints; "repair" means that the water shortage damage caused by the mismatch between power generation plan and water supply demand is reduced or eliminated through the above local adjustments, while maintaining the temporal continuity of the original candidate power generation plan and the scheduling intention as much as possible.

[0128] After the specific month is determined in S6, the power generation plan is modified accordingly, and the extent of the modification is limited to the total disturbance and the monthly disturbance constraints.

[0129] The set of critical repair periods Within this process, a sparse repair operation is performed on the candidate planned power generation, adjusting only within the set of critical repair periods. The interval projection operator is used to ensure that the repaired planned power generation falls within the planned boundary interval given by equation (2), thus obtaining the repaired planned power generation. : (Equation 14), where, To repair the step size, For the interval projection operator, it is preferred to determine the baseline plan and allowable adjustment range according to Equation (2) to ensure that the management is executable.

[0130] The repair process must simultaneously satisfy the following total disturbance constraint and single-month disturbance constraint: (Equation 15), where, The total disturbance threshold, This is the monthly disturbance threshold. It is used to control the total number of repairs and the monthly disturbance, ensuring fewer modifications and more timely modifications.

[0131] Combination Figure 3 This step specifically includes:

[0132] S71, Only for sets in critical repair periods The monthly planned power generation is adjusted (i.e., the sparse repair operator of Equation 14 above is executed), and through interval projection operation, it is forcibly ensured that the adjusted plan strictly falls within the pre-set grid-side allowed plan boundary.

[0133] S72. Determine whether the above repair operation meets the preset minimum disturbance constraint (including the total disturbance threshold and the monthly disturbance threshold, i.e., the aforementioned Equation 15). If not, return to step S71 to adjust the repair step size or constraint conditions and re-execute the repair; if it meets the requirement, a valid post-repair power generation plan is obtained.

[0134] S73. For the legitimate repaired power generation plan, recalculate steps S2 to S5 to obtain the new target vector after repair. Then, compare it with the plan before repair according to the "lexicographical order rule" to update the current individual optimal / global optimal.

[0135] S8. Embed the sparse repair operator into the rolling evolution framework to iteratively update the power generation plan and repeat S2 to S7 until the termination condition is met. Continuously execute the cycle of "calculate water volume - allocate water supply - diagnose sensitivity - fine-tune plan" until the water shortage is reduced to the minimum or the set number of calculations is reached (termination condition).

[0136] Combination Figure 3 Specifically, this includes...

[0137] S81. At the beginning of the algorithm startup, initialize the rolling evolution population (i.e. generate an initial population containing multiple candidate monthly power generation plans) and set the parameters.

[0138] S82. The complete “evaluation-diagnosis-repair” process of S2 to S7 above is used as a reinforcement operator and applied to each candidate plan individual in the population.

[0139] S83. After all individuals have completed one repair and evaluation, perform regular crossover, mutation and position update operations under the rolling evolution framework to generate the next generation candidate plan;

[0140] S84. Determine if the termination condition is met (e.g., whether the set maximum number of iterations has been reached, or whether the water shortage target has been reduced to the preset minimum threshold). If the termination condition is not met, reject the next generation candidate plan and repeat the outer loop above (i.e., return to execute S2 to S7).

[0141] If the termination condition is met, the loop will exit and proceed to step S9 to output the final optimal result.

[0142] S9. Output the optimal power generation plan that satisfies the constraints and targets, along with its corresponding water supply and reservoir capacity processes. After sufficient iteration, the final comprehensive scheduling scheme is output, which includes both the corrected optimal monthly power generation plan and the matching water supply details and reservoir water level fluctuation trajectory.

[0143] If only S2 to S5 of this invention are executed to calculate the power generation water volume, water supply allocation and target vector, but S6 to S7 are not executed: the inter-period sensitivity is not calculated, the key set is not constructed, the sparse repair operator that is only updated within the key repair period set is not adopted, but the monthly planned power generation is updated in a regular evolutionary manner throughout the entire time period, and the plan falls within the planned boundary interval given by equation (2) is only guaranteed by boundary projection.

[0144] The following engineering phenomena will then occur:

[0145] Due to a lack of cross-period directional information, plan changes are often scattered across multiple non-critical months, making it difficult to meet the management requirements of making fewer changes and making changes in the right time periods.

[0146] Because the algorithm does not explicitly characterize the transmission effect of power generation water consumption in a certain month on subsequent reservoir capacity and water supply damage, it is difficult to identify the month with the root cause of water shortage, which can easily lead to repeated iterative adjustments, slow convergence, or oscillations.

[0147] When water supply segment constraints are frequently triggered or water shortage constraints are tight, the comparison method often requires a larger deviation from the plan to achieve the same level of water shortage improvement, which is not conducive to balancing feasibility and the goal of minimizing disturbance.

[0148] Therefore, this embodiment utilizes intertemporal sensitivity—TopK—only when Internal sparse repair focuses adjustments on a few key periods, with clear direction and strong interpretability. It can prioritize reducing water shortage damage under controlled disturbance conditions, and further optimize the deviation when the water shortage index is not worse than the comparison method, thus demonstrating stronger engineering feasibility and stability.

[0149] Example 3

[0150] This embodiment combines Embodiment 1 and Embodiment 2 to provide the specific execution process of the reservoir optimization scheduling method.

[0151] Data preparation and boundary parameter setting stage (corresponding to step S1): Obtain the inflow water sequence, initial reservoir capacity, dead reservoir capacity, operating reservoir capacity boundary, leakage parameters, baseline planned power generation, power generation to be optimized / candidate planned power generation, and ecological / urban / agricultural water demand plan for each period of the scheduling period.

[0152] Reservoir capacity for power generation water intake restrictions The lowest working water level of the water intake structure The water level-reservoir capacity curve is obtained by inverse calculation using equation (1); the allowable planning boundary on the power grid side. Equation (2) is given by the baseline plan and the allowable adjustment range, which is used to constrain the executability of subsequent repair and search output.

[0153] Reservoir capacity estimation and power generation water quantity calculation stage (corresponding to steps S2 to S3): For any candidate plan The reservoir capacity is calculated using equation (3); within each time period, the reservoir capacity is calculated based on the water level-reservoir capacity curve. Tailwater level-discharge curve With head loss Construct the implicit equation of equation (5) and use the higher-order iteration of equation (6) (the number of iterations does not exceed 100). ) Calculate the amount of water used for power generation at each time period The process of obtaining storage capacity With the process of generating water volume .

[0154] Water supply allocation and target evaluation stage (corresponding steps S4 to S5): Calculate the available water volume for the time period according to formula (4). The ecological / urban / agricultural water supply is allocated according to formula (7), and the water shortage is obtained according to formula (8). The target vector is further calculated according to formulas (9) to (10). And the candidate plans for electric dispatching to obey water dispatching are selected using lexicographical order rules.

[0155] Intertemporal sensitivity and sparse repair stage (corresponding to steps S6-S7): Construct a water shortage summary proxy target according to equation (11), calculate the intertemporal sensitivity according to equation (12), preferably using differential perturbation method to facilitate project implementation. Obtain the critical set according to equation (13), and perform sparse repair within the critical repair period set according to equation (14), while satisfying the minimum perturbation constraint of equation (15). After repair, repeat the above reservoir capacity recursion and water supply allocation steps until the water shortage target meets the threshold. .

[0156] It should be noted that when using the differential perturbation method to calculate the inter-period sensitivity, the perturbation scale is preferably taken as 1% to 3% of the planned power generation for the month and not less than the minimum adjustment granularity on the grid side, and the perturbation plan is guaranteed to still fall within the planning boundary interval given by formula (2) for differential calculation; if the implicit root finding does not converge or the feasibility check fails, the perturbation scale is halved one by one until the solution is stable.

[0157] Rolling evolution embedding and output stage (corresponding to steps S8 to S9): In the rolling evolution framework such as particle swarm optimization and differential evolution, the above-mentioned sensitivity-TopK-sparse repair operation is first performed on each generation of candidate plans, and then the regular evolution update is completed; each generation uses lexicographical order rules to select the global optimum and individual optimum, and finally outputs the optimal power generation plan, water supply process and reservoir capacity process.

[0158] Example 4

[0159] A reservoir optimization scheduling method terminal based on intertemporal sensitivity sparse repair includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the reservoir optimization scheduling method based on intertemporal sensitivity sparse repair as described above.

[0160] Memory is used to store software programs and modules. The processor executes various terminal functions and data processing by running the software programs and modules stored in memory. Memory can mainly consist of a program storage area and a data storage area. The program storage area can store the operating system, at least one executable program required for a given function, etc.

[0161] The storage data area can store data created based on the use of the terminal. Furthermore, the memory can include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory, or other volatile solid-state storage devices.

[0162] A computer program product includes a computer program / instructions that, when executed by a processor, implement the reservoir optimization scheduling method based on intertemporal sensitivity sparse repair as described above.

[0163] Computer program products include computer programs or instruction sets used to perform specific tasks or achieve specific functions. These programs or instructions are designed to be executed by a processor to implement a series of predefined steps or operations. The program product may be stored in various forms of computer storage media, such as memory, hard disks, solid-state drives, optical discs, or other forms of digital storage devices. It may exist in the form of compiled binary code or in the form of scripts or bytecode that can be executed by an interpreter. Through carefully designed algorithms and logical instructions, the program product enables the processor to process data in a specific order and manner, performing various functions such as data analysis, user interaction, and device control.

[0164] In the description of this specification, the references to terms such as "one embodiment / mode," "some embodiments / modes," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment / mode or example is included in at least one embodiment / mode or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment / mode or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments / modes or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments / modes or examples described in this specification, as well as the features of different embodiments / modes or examples.

[0165] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0166] Those skilled in the art should understand that the above embodiments are merely for illustrating the present invention and are not intended to limit the scope of the invention. Those skilled in the art can make other changes or modifications based on the above invention, and these changes or modifications still fall within the scope of the present invention.

Claims

1. A reservoir optimization scheduling method based on intertemporal sensitivity sparse repair, characterized in that, include: Acquire basic operational data, initial power generation plans, and water supply plans of the reservoir during the scheduling period; Based on the aforementioned basic operational data, an implicit equation is constructed between power generation and water volume for power generation, and the water volume for power generation in each time period is obtained by solving the equation, and the reservoir capacity in each time period is obtained by recursion. The available water volume is determined based on the reservoir capacity and power generation water volume in each time period, and the water supply is allocated according to the preset priority order to calculate the corresponding water shortage. Construct a target vector for evaluating candidate power generation plans, calculate the intertemporal sensitivity of the water shortage to the planned power generation, and identify a set of critical repair periods accordingly. Sparse repair operations are performed on the candidate power generation plans within the set of critical repair periods, and the repair process satisfies preset perturbation constraints. The sparse repair operation is embedded in the iterative optimization framework to update the power generation plan until the termination condition is met, and the optimal power generation plan and the corresponding reservoir scheduling process are output. Among them, the methods for solving the hydropower generation volume for each time period through implicit equations include: Construct the first The implicit equations for power generation and hydropower generation over a given period are established, and their residual functions are determined. ,in, For power generation coefficient, For the first Water volume for power generation during the period For the first Duration of time This is a water level-reservoir capacity curve function. This is a tailwater level-discharge curve function. For the first Time-related head loss, For the candidate project's power generation, For the first Initial storage capacity for the period For the first Storage capacity at the end of the period; The solution is obtained using a higher-order univariate iteration, and the iteration formula is as follows: ,in, Number the iteration steps. For the first Time period The amount of water generated in the next iteration For the first Time period The amount of water generated in the next iteration For the first Time period Implicit equation residuals at the next iteration and These are the implicit equation residual functions. The first and second derivatives; When the implicit equation is satisfied, the absolute value of the residual Or reach the preset maximum number of iterations The iteration stops at a certain point, and the power generation water volume for each time period is obtained. This is the tolerance for residual convergence. Among them, methods for identifying the set of critical repair periods based on intertemporal sensitivity include: Introducing a water-scarce weighted aggregation agent target Calculate the first Intertemporal sensitivity : ,in, , This represents the total number of time periods. , , The weights for ecological water shortage, urban water shortage, and agricultural water shortage are respectively, and they satisfy the following conditions: ; For the first Candidate power generation for the specified time period; Based on the calculated inter-period sensitivity of each time period, a set of key repair time periods was identified. : ,in, To select the one with the largest absolute value An operator that assigns each element a time period number. The number of key time periods is preset; The methods for performing sparse repair operations include: The set of critical repair periods Within the process, a sparsity repair operation is performed on the candidate planned power generation to obtain the repaired planned power generation. : ,in, To repair the step size, For interval projection operators; The repair process must simultaneously satisfy the following total disturbance constraint and single-month disturbance constraint: ,in, The total disturbance threshold, This is the monthly disturbance threshold.

2. The reservoir optimization scheduling method based on intertemporal sensitivity sparse repair according to claim 1, characterized in that, The basic operational data includes: the inflow sequence during the scheduling period, initial reservoir capacity, dead reservoir capacity, operating reservoir capacity boundary, minimum working water level of water intake structures, leakage parameters, water level-reservoir capacity curve data, and tailwater level-flow curve data; The initial power generation plan includes: a monthly power generation plan; The water supply plan includes: ecological water supply plan, urban water supply plan and agricultural water supply plan; Specifically, the power generation water intake limit reservoir capacity is determined by reverse calculation based on the minimum working water level of the water intake structure and the water level-reservoir capacity curve data.

3. The reservoir optimization scheduling method based on intertemporal sensitivity sparse repair according to claim 1, characterized in that, The recursive methods for estimating the storage capacity for each time period include: Based on the principle of water balance, the reservoir capacity is calculated recursively: ,in, For the first Initial storage capacity for the time period For the first The final storage capacity of the time period For the first The amount of water entering the reservoir during a given period For the first Water volume for power generation during the period For the first Non-power generation water discharge or supplementary water supply during the period, For the first Leakage volume over a period of time; The recursive storage capacity for each time period must meet the following constraints: and ,in, and These are the lower and upper limits of the reservoir's operating capacity, respectively. Dead storage capacity of the reservoir; The convergence and feasibility verification conditions are that the obtained power generation water volume and the corresponding reservoir capacity satisfy the boundary constraints of reservoir capacity and water volume.

4. The reservoir optimization scheduling method based on intertemporal sensitivity sparse repair according to claim 1, characterized in that, Methods for determining available water and calculating water shortage include: Based on the priority water supply mode for power generation tailwater, the first Water availability during the period : ,in, For the first Water volume for power generation during the period For the first The final storage capacity of the time period Reservoir capacity is limited due to water intake for power generation; The preset priority order is as follows: ecology, urban, agriculture; and the first priority is calculated according to the aforementioned priority order. Ecological water supply during different time periods Urban water supply and agricultural water supply : ,in, , and The first Ecological water demand plan, urban water demand plan, and agricultural water demand plan for each time period; Based on the calculated water supply, the corresponding ecological water shortage is calculated respectively. Urban water shortage and agricultural water shortage : .

5. The reservoir optimization scheduling method based on intertemporal sensitivity sparse repair according to claim 4, characterized in that, Methods for constructing target vectors include: Based on the water shortage in each time period, the total number of time periods is calculated separately. Summary of ecological water shortage targets Urban water shortage summary target and summary targets for agricultural water shortage : ; Based on the candidate power generation plan and the initial power generation plan, the target deviation from the plan is calculated. : ,in, For the first Candidate planned power generation for the specified time period For the first Initial baseline planned power generation for the time period; Construct a lexicographically ordered target vector for evaluating candidate power generation plans. : ; The optimal power generation plan is obtained by screening candidate power generation plans according to the principle that power dispatch follows water dispatch.

6. A reservoir optimization scheduling method terminal based on intertemporal sensitivity sparse repair, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the reservoir optimization scheduling method based on intertemporal sensitivity sparse repair as described in any one of claims 1-5.

7. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the reservoir optimization scheduling method based on intertemporal sensitivity sparse repair as described in any one of claims 1-5.

Images