Single reservoir dispatching method and device based on adaptive grid and micro-step simulation

Abstract

This invention provides a single reservoir scheduling method and device based on adaptive grid and micro-step simulation, relating to the field of flood control scheduling technology in water conservancy engineering. By using micro-step simulation, the integral error of traditional long-step calculation is overcome, making the simulation results closer to the actual hydraulic processes of the reservoir. Based on the simulation results, the effective reservoir capacity range is dynamically locked, and a search grid is adaptively constructed in conjunction with the inflow intensity, significantly compressing the state space while ensuring accuracy and solving the curse of dimensionality. Through physically driven grid density calculation, the difference in water volume between adjacent discrete states is ensured to be less than the reservoir's single-step regulation capacity, eliminating "grid deadlock" at its root. Dual capacity constraint verification is introduced into the dynamic programming to prevent the mathematically optimal solution from failing due to insufficient discharge capacity. Multiple engineering constraints can be flexibly incorporated, maximizing the peak reduction rate while ensuring safety, achieving synergistic optimization of safety and efficiency, and significantly improving the flood control scheduling efficiency of the reservoir.

Description

Technical Field

[0001] This invention relates to the field of flood control scheduling technology in water conservancy projects, and in particular to a single reservoir scheduling method and device based on adaptive grid and micro-step simulation. Background Technology

[0002] Reservoir flood control is a core component of flood prevention and disaster reduction. Existing control methods are mainly divided into two categories: rule-based control and optimization control. Rule-based control is based on fixed logic (such as water level-triggered discharge), which is simple to operate but lacks flexibility. When facing complex flood processes, its peak-shaving effect is limited, and it is difficult to achieve global optimization.

[0003] Dynamic programming (DP) in optimized scheduling can theoretically obtain the global optimal solution, but in practical applications, it faces the following technical bottlenecks: First, discretization of the entire reservoir capacity state easily leads to the "curse of dimensionality," resulting in low computational efficiency; second, coarse-grained discretization may cause the difference in water volume between adjacent states to exceed the reservoir's single-step regulation capacity, resulting in "grid deadlock," making the algorithm unsolvable or reducing the quality of the solution; third, traditional DP only verifies the discharge capacity at the initial moment, ignoring the impact of the water level drop at the final moment on subsequent discharge, leading to the physical infeasibility of the optimized scheme; fourth, conventional simulations use hourly step sizes, making it difficult to accurately characterize the nonlinearity of the reservoir capacity curve and the abrupt changes in discharge rules, resulting in insufficient simulation accuracy.

[0004] In summary, there is an urgent need for a single-reservoir flood optimization scheduling method that balances simulation accuracy, computational efficiency, and physical feasibility. Summary of the Invention

[0005] The purpose of this invention is to provide a single reservoir scheduling method and device based on adaptive grid and micro-step simulation. Micro-step simulation overcomes the integral error of traditional long-step calculations, making the simulation results closer to the actual hydraulic processes of the reservoir and providing a high-fidelity initial sequence for optimization. Based on the simulation results, the effective reservoir capacity range is dynamically locked, and a search grid is adaptively constructed in conjunction with the inflow intensity, significantly compressing the state space while ensuring accuracy and solving the curse of dimensionality. Through physically driven grid density calculation, the difference in water volume between adjacent discrete states is ensured to be less than the reservoir's single-step regulation capacity, eliminating "grid deadlock" at its root. Dual capacity constraint verification is introduced into the dynamic programming, ensuring that the optimized solution meets engineering physical limits throughout the entire time period, preventing the mathematically optimal solution from failing due to insufficient discharge capacity. Multiple engineering constraints can be flexibly incorporated, maximizing the peak reduction rate while ensuring safety, achieving synergistic optimization of safety and efficiency, and significantly improving the reservoir's flood control scheduling efficiency.

[0006] In a first aspect, the present invention provides a single reservoir scheduling method based on adaptive grid and micro-step size simulation, comprising:

[0007] Construct a digital physical model of the target reservoir; the digital physical model includes: a two-way mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity;

[0008] Based on preset scheduling rules, a micro-step simulation of the target reservoir inflow flood process is performed to obtain the initial reservoir capacity change sequence that reflects the basic response characteristics of the target reservoir.

[0009] Based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and the preset engineering constraint boundary, a dynamic programming search grid is adaptively constructed. The dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points.

[0010] On the dynamic programming search grid, with the goal of minimizing the maximum outbound flow during the entire scheduling process, the dynamic programming algorithm is used for global optimization to find the optimal storage capacity change path that satisfies the preset physical constraints, and the corresponding optimal outbound flow sequence is calculated accordingly.

[0011] The output includes an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence, which serves as the final scheduling instruction for the target reservoir.

[0012] In some preferred embodiments of the present invention, the step of performing micro-step simulation of the target inflow flood process based on preset scheduling rules to obtain the initial reservoir capacity change sequence reflecting the basic response characteristics of the target reservoir includes:

[0013] The preset standard hydrological time period step is subdivided into multiple micro-steps based on a preset segmentation constant; where the segmentation constant is an integer greater than or equal to 1.

[0014] Within each microstep, the following steps are executed sequentially: calculating the water level based on the current reservoir capacity, calculating the target discharge volume according to the preset scheduling rules, limiting the target discharge volume according to the maximum discharge capacity corresponding to the current water level, and updating the reservoir capacity using the water balance equation.

[0015] All simulation results with microsteps are aggregated according to standard hydrological time periods to output the initial reservoir capacity change sequence.

[0016] In some preferred embodiments of the present invention, the step of determining the effective search interval includes:

[0017] Extract the minimum and maximum values ​​of the initial storage capacity change sequence;

[0018] Based on the ratio of the current inflow intensity to the total capacity of the target reservoir, the adaptive buffer amount is dynamically calculated, and an initial search interval is constructed based on the minimum and maximum values ​​of the initial capacity change sequence and the adaptive buffer amount.

[0019] By introducing engineering constraint boundaries to truncate and correct the initial search interval, the final effective search interval is obtained.

[0020] In some preferred embodiments of the present invention, the step of calculating the discretization step size that guarantees state reachability includes:

[0021] Based on the mapping relationship between water level and maximum discharge capacity, determine the maximum water volume change capacity of the target reservoir within a single time period;

[0022] Determine the state discretization step size; wherein the state discretization step size is less than or equal to the ratio of the maximum water volume change capacity to the preset safe connectivity coefficient.

[0023] In some preferred embodiments of the present invention, the step of generating discrete storage capacity state points and pre-calculating the physical properties corresponding to the discrete storage capacity state points includes:

[0024] For each generated discrete reservoir capacity state point, the corresponding water level is determined through the two-way mapping relationship between water level and reservoir capacity;

[0025] The maximum physical discharge capacity corresponding to the water level is calculated by the mapping relationship between water level and maximum discharge capacity;

[0026] The mapping results are used to construct a lookup table.

[0027] In some preferred embodiments of the present invention, the step of using a dynamic programming algorithm for global optimization on the dynamic programming search grid, with the objective of minimizing the maximum outbound flow during the entire scheduling process, includes:

[0028] Define a state transition equation and back-calculate the required average outflow from the reservoir capacity state at adjacent times based on the water balance principle. In the state transition process, a preset dual capacity constraint check is introduced to ensure that the back-calculated outflow simultaneously meets the maximum discharge capacity limit corresponding to the start and end states of the transition.

[0029] A recursive formula that minimizes the maximum outbound flow rate is used for global optimization;

[0030] By backtracking from the termination point, the globally optimal storage capacity path is reconstructed.

[0031] In some preferred embodiments of the present invention, the dual capability constraint verification includes:

[0032] Verify whether the maximum discharge capacity corresponding to the storage capacity status at the start of the transfer is not less than the deduced outflow rate;

[0033] Verify whether the water level corresponding to the reservoir capacity at the time of transfer termination can support the inferred outflow rate.

[0034] In some preferred embodiments of the present invention, the recursive formula for minimizing the maximum outbound flow includes:

[0035] ;

[0036] in, To arrive at time from the initial time status The minimum of the maximum outbound flow experienced across all paths; Let be the storage capacity at time t; j is the storage capacity grid point at time t; Let be the storage capacity value at grid point j at time t; Let be the average outflow that the reservoir needs to discharge from its state at time t-1 to its state at time t.

[0037] In some preferred embodiments of the present invention, the engineering constraint boundary includes at least one of the following: starting water level, dead water level, warning water level, flood control high water level, and check flood level;

[0038] The method also includes:

[0039] The optimized scheduling scheme will be compared and analyzed with the benchmark scheme obtained by micro-step simulation based on preset scheduling rules. The comparison indicators include at least one of the following: highest water level, maximum outflow, and peak shaving rate.

[0040] Secondly, the present invention provides a single reservoir scheduling device based on adaptive grid and micro-step size simulation, comprising:

[0041] The physical model building module is used to construct a digital physical model of the target reservoir; the digital physical model includes: a two-way mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity;

[0042] The basic scheduling simulation module is used to perform micro-step simulation of the target inflow flood process based on preset scheduling rules, and obtain the initial reservoir capacity change sequence that reflects the basic response characteristics of the target reservoir.

[0043] The adaptive grid processing module is used to adaptively construct a dynamic programming search grid based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and preset engineering constraint boundaries. The dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points.

[0044] The global optimization module for scheduling paths is used to perform global optimization on the dynamic programming search grid with the goal of minimizing the maximum outbound flow during the entire scheduling process. It uses a dynamic programming algorithm to solve for the optimal storage capacity change path that satisfies the preset physical constraints, and calculates the corresponding optimal outbound flow sequence accordingly.

[0045] The optimization scheme processing module is used to output an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence, which serves as the final scheduling instruction for the target reservoir.

[0046] This invention brings the following beneficial effects:

[0047] This invention provides a single reservoir scheduling method and apparatus based on adaptive grid and micro-step simulation. The method includes: constructing a digital physical model of the target reservoir; wherein the digital physical model includes: a bidirectional mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity; performing micro-step simulation of the target inflow flood process based on preset scheduling rules to obtain an initial reservoir capacity change sequence reflecting the basic response characteristics of the target reservoir; adaptively constructing a dynamic programming search grid based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and preset engineering constraint boundaries; wherein the dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points; on the dynamic programming search grid, with the objective of minimizing the maximum outflow during the entire scheduling process, using a dynamic programming algorithm for global optimization to solve for the optimal reservoir capacity change path that satisfies the preset physical constraints. The system calculates the corresponding optimal outflow sequence based on this; it outputs an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence as the final scheduling instruction for the target reservoir; through micro-step simulation, it overcomes the integral error of traditional long-step calculation, making the rule simulation results closer to the actual hydraulic process of the reservoir and providing a high-fidelity initial sequence for optimization; based on the simulation results, it dynamically locks the effective reservoir capacity range and adaptively constructs a search grid in combination with the inflow intensity, significantly compressing the state space while ensuring accuracy and solving the problem of dimensionality curse; through physical-driven grid density calculation, it ensures that the difference in water volume between adjacent discrete states is less than the single-step regulation capacity of the reservoir, eliminating "grid deadlock" from the root; it introduces dual capacity constraint verification in dynamic programming, ensuring that the optimized scheme meets the engineering physical limits throughout the time period, avoiding the failure of the mathematically optimal solution due to insufficient discharge capacity; it supports the flexible implantation of multiple engineering constraints, maximizing the peak reduction rate under the premise of ensuring safety, achieving synergistic optimization of safety and efficiency, and significantly improving the flood control scheduling efficiency of the reservoir. Attached Figure Description

[0048] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0049] Figure 1 A flowchart of a single reservoir scheduling method based on adaptive grid and micro-step size simulation provided in this embodiment of the invention;

[0050] Figure 2 A schematic diagram of single reservoir scheduling provided in an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of a single reservoir scheduling device based on adaptive grid and micro-step simulation provided in an embodiment of the present invention;

[0052] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0053] Icons: 310 - Physical model building module; 320 - Basic scheduling simulation module; 330 - Adaptive mesh processing module; 340 - Global optimization of scheduling path module; 350 - Optimization scheme processing module; 400 - Memory; 401 - Processor; 402 - Bus; 403 - Communication interface. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0055] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0056] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0057] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. In addition, the terms "first," "second," "third," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0058] Furthermore, terms such as "horizontal," "vertical," and "sag" do not imply that components must be absolutely horizontal or suspended, but rather that they can be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0059] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0060] Reservoir flood control is a core component of the flood control and disaster reduction system. For single reservoir control, the optimization goal should be to improve the peak reduction rate and minimize the risk of downstream floods, while ensuring the safety of the reservoir project.

[0061] Current mainstream scheduling methods are mainly divided into two categories: rule-based scheduling and optimization scheduling. Rule-based scheduling is based on the practical experience of engineers to formulate fixed scheduling logic (such as using high flood control water levels as flood discharge trigger conditions). Its advantage lies in its simple operation process and ease of execution, but it lacks flexibility and has limited peak-shaving effectiveness when facing complex flood processes, making it difficult to achieve global optimization of the scheduling scheme. Optimization scheduling includes two major branches: heuristic algorithms and planning algorithms. Among them, dynamic programming (DP), as a classic planning algorithm, can theoretically obtain the global optimal solution, but it has the following technical bottlenecks in practical engineering applications:

[0062] The challenge of state space discretization: Traditional dynamic programming requires state discretization over the entire storage capacity. When the storage capacity is large or the step size is small, it is easy to cause the "curse of dimensionality", which leads to a surge in computation and makes it difficult to meet the timeliness requirements.

[0063] Grid deadlock risk: If coarse-grained discretization is used to improve computational efficiency, the water volume difference between adjacent discrete states may exceed the maximum water volume change capacity of the reservoir in a single time period, resulting in the state transition path being unreachable, i.e., "grid deadlock" occurs, making the algorithm unsolvable or the quality of the solution severely degraded;

[0064] Physical constraint conflict: Traditional dynamic programming often only considers the discharge capacity constraint at the beginning of the state transition process, ignoring the impact of the water level drop at the end on the subsequent discharge capacity, which leads to the failure of the optimized scheduling scheme in actual engineering due to physical infeasibility.

[0065] Insufficient simulation accuracy: Conventional scheduling rule simulations often use hourly steps for water balance calculations, which makes it difficult to accurately depict the hydraulic processes caused by the nonlinearity of the reservoir capacity curve and the abrupt changes in the discharge rules. As a result, there are discrepancies between the simulation results and the actual response.

[0066] This invention provides a single reservoir scheduling method and device based on adaptive grid and micro-step simulation, which can adaptively lock the optimal search interval, dynamically adapt the state grid, and strictly meet engineering physical constraints while ensuring high-fidelity simulation of flood physical processes.

[0067] The following detailed description of some embodiments of the present invention is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0068] Example 1

[0069] Firstly, this invention provides a single reservoir scheduling method based on adaptive grid and micro-step size simulation, see [link to relevant documentation]. Figure 1 The flowchart shown in this embodiment of the invention provides a single reservoir scheduling method based on adaptive grid and micro-step size simulation, including:

[0070] Step S102: Construct a digital physical model of the target reservoir; wherein, the digital physical model includes: a two-way mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity.

[0071] Specifically, based on measured or design data of the target reservoir, a digital mapping of the reservoir's basic physical characteristics is constructed. The bidirectional mapping relationship between water level and reservoir capacity is established using a linear interpolation algorithm to construct a transformation function between water level Z and reservoir capacity V, i.e. and This is used to enable mutual querying of water level and reservoir capacity. The mapping relationship between water level and maximum discharge capacity is constructed based on the characteristics of the spillway, flood discharge tunnel, and gates, establishing a relationship between water level Z and maximum physical discharge capacity. Functional relationship This is used to determine the upper limit of discharge at any water level. By establishing an accurate digital physical model, a unified data foundation is provided for subsequent simulation and optimization, ensuring the accuracy of the conversion between water level, reservoir capacity, and discharge capacity, as well as the efficiency of real-time querying, and avoiding decision-making biases caused by ambiguity in physical relationships.

[0072] Step S104: Based on the preset scheduling rules, perform micro-step simulation of the target inflow flood process to obtain the initial reservoir capacity change sequence that reflects the basic response characteristics of the target reservoir.

[0073] Specifically, within a preset micro-step size Δτ, the following operations are performed: Based on the instantaneous reservoir capacity at the current micro-step time, the instantaneous water level is inferred through a water level-capacity mapping; preset scheduling rules (such as tiered control rules) are invoked, and the target discharge is calculated by combining the current instantaneous water level and inflow rate; it is verified whether the target discharge exceeds the maximum discharge capacity corresponding to the current water level, and the smaller value is taken as the actual micro-step discharge; the reservoir capacity is updated using the water balance equation. All simulation results for micro-step sizes are aggregated according to standard hydrological time periods, outputting the initial reservoir capacity change sequence and outflow rate sequence. Through micro-step simulation, the integral error caused by nonlinear reservoir capacity curves and complex discharge rules in long-step calculations is effectively overcome, making the simulation results closer to the actual hydraulic processes of the reservoir. This provides a high-fidelity initial reservoir capacity change sequence for subsequent optimization, while also improving the accuracy of the rule-based scheduling itself.

[0074] Furthermore, in some preferred embodiments of the present invention, the step of performing micro-step simulation of the target inflow flood process based on preset scheduling rules to obtain the initial reservoir capacity change sequence reflecting the basic response characteristics of the target reservoir includes: subdividing the preset standard hydrological time period step into multiple micro-steps based on a preset segmentation constant; wherein, the segmentation constant is an integer greater than or equal to 1; within each micro-step, sequentially executing: back-calculating the water level based on the current reservoir capacity, calculating the target discharge according to the preset scheduling rules, limiting the target discharge according to the maximum discharge capacity corresponding to the current water level, and updating the reservoir capacity using the water balance equation; aggregating the simulation results of all micro-steps according to the standard hydrological time period, and outputting the initial reservoir capacity change sequence.

[0075] Specifically, first, the hydrological time step ΔT (outer layer step, in seconds) is set, and then it is subdivided into N micro-steps Δτ (inner layer step), i.e. In some preferred embodiments of the present invention, N=60, that is, the 1-hour period is divided into 60 1-minute micro-steps. Within each micro-step, the instantaneous water level is first calculated by back-calculating the instantaneous reservoir capacity based on the water level-reservoir capacity mapping; then, a preset scheduling rule (such as a tiered control rule) is invoked, and the target discharge is calculated by combining the current instantaneous water level and the inflow rate; finally, the maximum discharge capacity corresponding to the current water level is obtained from the water level-maximum discharge capacity relationship. The target leakage will be limited to no more than This serves as the actual micro-step discharge volume; utilizing the water balance equation. Update the database capacity and handle overflow at the top. The water level is either below the dead water level or below engineering constraints. After a hydrological period ends, the simulation results of all micro-steps within that period (such as the reservoir capacity at the end of the period and the average outflow during the period) are aggregated to obtain the reservoir capacity change and outflow during that period. The above process is repeated for all periods to finally obtain the complete initial reservoir capacity change sequence. This micro-step simulation mechanism simulates the reservoir regulation process at a minute-level resolution, accurately depicting the instantaneous interaction between water level changes and discharge decisions, avoiding the cumulative errors caused by traditional hourly step sizes, making the rule-based scheduling results more valuable, and providing a reliable dynamic range basis for subsequent adaptive grid construction.

[0076] Step S106: Based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and the preset engineering constraint boundary, an adaptive dynamic programming search grid is constructed; wherein, the dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points.

[0077] Specifically, the minimum value of the initial storage capacity change sequence is first extracted. and maximum value As a baseline; then, a preliminary search interval is constructed using a preset adaptive buffer amount δ. Next, preset engineering constraint boundaries (such as reservoir capacity corresponding to the warning water level) are introduced to truncate and correct the initial search interval, resulting in the final effective search interval. Based on this, to ensure connectivity during state transitions, the maximum possible water volume change capacity is preset. And based on the maximum possible water volume change capacity The state discretization step size Δv is determined. Within the effective search interval, M discrete reservoir capacity state points are generated according to the step size Δv, and the corresponding water level and maximum discharge capacity are pre-calculated for each state point, constructing a lookup table for subsequent use. The effective reservoir capacity interval is dynamically locked through rule-based simulation results, significantly compressing the state space size and avoiding the curse of dimensionality in traditional dynamic programming when traversing the entire reservoir capacity. At the same time, a physics-driven adaptive grid density algorithm is introduced to ensure that the water volume difference between adjacent discrete states is less than the maximum adjustment capacity in a single step, fundamentally eliminating grid deadlock and guaranteeing the robustness and feasibility of the algorithm.

[0078] Furthermore, in some preferred embodiments of the present invention, the step of determining the effective search interval includes: extracting the minimum and maximum values ​​of the initial reservoir capacity change sequence; dynamically calculating the adaptive buffer amount based on the ratio of the current inflow intensity to the total reservoir capacity of the target reservoir; constructing a preliminary search interval based on the minimum and maximum values ​​of the initial reservoir capacity change sequence and the adaptive buffer amount; and introducing engineering constraint boundaries to truncate and correct the preliminary search interval to obtain the final effective search interval.

[0079] Specifically, let the minimum value of the initial storage capacity change sequence be... The maximum value is The adaptive buffer size δ is based on the inbound flow rate during the current time period. With the total reservoir capacity The ratio is dynamically calculated, for example, using a linear function or a lookup table, so that the buffer capacity increases appropriately with the increase of inflow intensity to cope with flood uncertainty. The initial search interval is [ Subsequently, if engineering constraints exist, such as the warning water level Z... alert Then convert it to the corresponding storage capacity V. alert If V alert If the value is less than the upper limit of the initial search interval, the upper limit of the search interval will be forcibly adjusted to V. alert Similarly, the reservoir capacity V corresponding to the dead water level dead This can be used as a lower bound constraint. The final effective search interval is then obtained. This method not only narrows the search range using rule-based simulation results, but also flexibly responds to flood fluctuations through adaptive buffering. Furthermore, it directly incorporates engineering safety constraints into the search space construction, ensuring that the optimization process always operates within safe boundaries, thus achieving a dual guarantee of efficiency and safety.

[0080] Furthermore, in some preferred embodiments of the present invention, the step of calculating the discretization step size to ensure state reachability includes: determining the maximum water volume change capacity of the target reservoir within a single time period based on the mapping relationship between water level and maximum discharge capacity; determining the state discretization step size; wherein the state discretization step size is less than or equal to the ratio of the maximum water volume change capacity to a preset safe connectivity coefficient.

[0081] Specifically, firstly, based on the relationship between water level and maximum discharge capacity... Furthermore, the water level-reservoir capacity relationship allows calculation of the maximum reduction in water volume that can be achieved by releasing water from a reservoir at its maximum discharge capacity within a given time period Δt, starting from a certain water level. And the maximum amount of water that can be added by storing water at maximum inflow capacity. The larger of the two values ​​is taken as the maximum water volume change capacity. To ensure the feasibility of transitions between adjacent discrete states, the state discretization step size Δv must satisfy... Where K is the safety connectivity coefficient (K≥1), a larger K results in a denser grid and better connectivity, but also increases the computational load. In practical applications, the value of K can be selected according to flood characteristics and accuracy requirements (e.g., K=2). This condition ensures that starting from any discrete state, at least the adjacent discrete state can be reached within a single time period, thus avoiding deadlock. The maximum water volume change capability is calculated through physical mechanisms, and the safety connectivity coefficient is introduced to constrain the discretization step size, making the grid density match the actual regulation capacity of the reservoir. This fundamentally eliminates the state unreachability problem caused by excessively coarse grids in traditional dynamic programming, ensuring that the algorithm can find feasible solutions under any flood process, significantly improving the robustness and adaptability of the method.

[0082] Furthermore, in some preferred embodiments of the present invention, the step of generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points includes: for each generated discrete reservoir capacity state point, determining the corresponding water level through the bidirectional mapping relationship between water level and reservoir capacity; calculating the maximum physical discharge capacity corresponding to the water level through the mapping relationship between water level and maximum discharge capacity; and constructing a lookup table from the mapping results.

[0083] Specifically, within the valid search range [ Within the [inventory], M discrete storage capacity state points {v1, v2, ..., v3} are obtained by uniformly dividing the storage area according to the step size Δv. M For each state point v k Calling the water level-reservoir capacity relationship Obtain the corresponding water level Z k Then call the water level-maximum discharge capacity relationship. Obtain the maximum discharge capacity under this state. (v) k Z k Q max_k The triples are stored in a lookup table for direct use during subsequent dynamic programming solutions, eliminating the need for repeated curve interpolation calculations. Pre-computation embeds frequently used physical properties in the lookup table during dynamic programming recursion, avoiding repeated interpolation calculations at each state transition. This significantly improves the algorithm's solution speed, making it particularly suitable for long-duration flood events or real-time scheduling scenarios requiring rapid response, thus saving valuable decision-making time for engineering applications.

[0084] Step S108: On the dynamic programming search grid, with the goal of minimizing the maximum outbound flow during the entire scheduling process, the dynamic programming algorithm is used for global optimization to find the optimal storage capacity change path that satisfies the preset physical constraints, and the corresponding optimal outbound flow sequence is calculated accordingly.

[0085] Specifically, on the adaptive grid, the state variable is defined as the storage capacity V at time t. t And V tOnly values ​​from discrete state points can be taken. Starting from the known initial storage capacity V0 at the initial time t=0, the calculation proceeds recursively for each subsequent time period. For each time period, the path from all feasible states at the previous time to all feasible states at the current time is considered, and the required average outflow Q is calculated. out The process recursively aims to minimize the maximum outbound flow. After reaching the endpoint T, all feasible states are traversed, and the state with the minimum maximum outbound flow is selected as the optimal endpoint. Then, based on the recorded optimal preceding state index, the process backtracks from the endpoint to the starting point to reconstruct the globally optimal storage capacity sequence, and calculates the corresponding optimal outbound flow sequence accordingly.

[0086] Furthermore, the core of dynamic programming's optimal backtracking lies in the fact that at each step of the forward recursive calculation, the current optimal predecessor node must be recorded synchronously. That is, when establishing an optimal state for the current time period, the source state from the previous time period must be saved as the corresponding index, thus forming a continuous chain of pointers at the endpoint for subsequent reverse lookups. This dynamic programming solution process uses peak shaving rate as the core optimization objective, employs minimization-maximization recursion, and identifies the discharge strategy that minimizes downstream flood risk from a global perspective, significantly improving the reservoir's flood control benefits.

[0087] For example, consider a flood control scheduling system that includes three points in time: t=0 (starting point), t=1 (intermediate point), and t=2 (ending point).

[0088] Forward derivation and process recording: When calculating the storage capacity A reaching t=1, if the transfer from the initial storage capacity at t=0 is optimal, then record on the storage capacity A node: "Optimal predecessor is the initial storage capacity at t=0". When calculating the storage capacity C reaching t=2, if the transfer from storage capacity A at t=1 has the best peak-shaving effect, then record on the storage capacity C node: "Optimal predecessor is storage capacity A at t=1", and save the current minimum flood peak value. Endpoint optimization: After recursively reaching the end point t=2, compare all end point states and find that "storage capacity C" experiences the smallest global maximum outflow, thus determining it as the globally optimal end point.

[0089] Furthermore, in some preferred embodiments of the present invention, the steps of using a dynamic programming algorithm for global optimization on the dynamic programming search grid, with the objective of minimizing the maximum outflow during the entire scheduling process, include: defining a state transition equation, and back-calculating the required average outflow based on the reservoir capacity state at adjacent times according to the water balance principle; wherein, a preset dual capacity constraint check is introduced during the state transition process to ensure that the back-calculated outflow simultaneously satisfies the maximum discharge capacity limit corresponding to the start and end states of the transition; global optimization is performed using a recursive formula that minimizes the maximum outflow; and backtracking from the end time to reconstruct the globally optimal reservoir capacity path.

[0090] Specifically, define the state transition equation: for the state storage capacity at time t The state storage capacity is transferred to time t+1. According to the principle of water balance, the required average outflow is: When performing a state transition, it is necessary to verify that... Does it satisfy the dual capability constraint? First, It should not be greater than the initial state. Corresponding maximum discharge capacity Secondly, It should also not be greater than the termination state. Corresponding maximum discharge capacity This is to prevent insufficient discharge capacity in subsequent periods due to a rapid drop in water level. Only transfer paths that simultaneously meet both of these conditions are permitted. Then, the arrival state is calculated using a preset recursive formula. Find the optimal value of F and record the predecessor state. After the recursion is completed, select F from the final time T. T The state with the minimum value is taken as the optimal endpoint. By backtracking along the predecessor chain of records, the optimal reservoir capacity path from the starting point to the endpoint is obtained. By embedding dual capability constraints into the state transition process, it is ensured that the optimized scheduling scheme is not only mathematically optimal but also physically fully executable. At the same time, the recursive and backtracking mechanisms guarantee a rigorous mathematical proof of global optimality, providing a theoretically optimal solution for reservoir flood control scheduling.

[0091] Furthermore, in some preferred embodiments of the present invention, the dual capability constraint verification includes: verifying whether the maximum discharge capacity corresponding to the reservoir capacity status at the start of the transfer is not less than the deduced outflow rate; and verifying whether the water level corresponding to the reservoir capacity status at the end of the transfer can support the deduced outflow rate.

[0092] Specifically, regarding the state arrive The transition first obtains the initial state from the pre-computed lookup table. The corresponding maximum discharge capacity Q max_start and termination status The corresponding maximum discharge capacity Q max_end The average outbound flow rate derived from this is... Q must be satisfied out ≤Q max_start This is to ensure that the flow rate can be released at the current water level; at the same time, it is also necessary to meet Q. out ≤Q max_endThis ensures that the reservoir can maintain the outflow rate at the next water level (or at least prevent further discharge due to a sudden drop in water level). If either condition is not met, the transfer path is deemed infeasible and excluded during the recursive process. This dual verification ensures the coordination of the scheduling process from a temporal perspective, avoiding blind spots in gate operation or subsequent attenuation of discharge capacity caused by rapid water level drops. This ensures the continuous operability of the optimized scheme throughout the entire time period, improving the stability and safety of scheduling execution.

[0093] Furthermore, in some preferred embodiments of the present invention, the recursive formula for minimizing the maximum outbound flow includes: ;in, To arrive at time from the initial time status The minimum of the maximum outbound flow experienced across all paths; Let be the storage capacity at time t; j is the storage capacity grid point at time t; Let be the storage capacity value at grid point j at time t; Let be the average outflow that the reservoir needs to discharge from its state at time t-1 to its state at time t.

[0094] Specifically, this recursive formula embodies the Min-Max optimization concept. For each current state... Iterate through all feasible states from the previous time step. For each preceding state, calculate the maximum outbound flow accumulated on the current path, i.e. Then, among all the paths corresponding to the preceding states, the minimum value of the cumulative maximum flow is selected as... The preceding state that achieves the minimum value is recorded. It should be noted that t and t-1 both represent time steps (scheduling periods); where t is the current time (e.g., the current hour), and t-1 is the previous time (e.g., the past hour); i and j are both indices (numbers) of the discrete storage capacity states. In this embodiment, the storage capacity range is divided into many grid points (discrete states), where i represents a specific storage capacity grid point at the current time t; j represents a specific storage capacity grid point at the previous time t-1; v i This indicates the specific storage capacity of the reservoir at the current moment, which is also the target point. j This represents the reservoir's specific capacity at the previous moment, i.e., the starting point. Qout represents the average outflow that the reservoir needs to discharge to change its capacity from the previous moment's state to the current moment's state. F t-1 This represents the maximum outflow rate experienced along that route when moving from the initial time step to a certain storage capacity state in the previous time step. F tThis represents the minimum maximum outflow rate experienced among all possible routes from the initial time to a certain storage capacity state at the current time.

[0095] In this way, the recursive process ensures the optimal substructure at each state, ultimately obtaining the optimal path that minimizes the maximum outflow at the endpoint. This recursive formula, with minimizing the maximum value as its core, perfectly aligns with the essential requirements of peak-shaving scheduling. It can identify the optimal discharge strategy that minimizes downstream flood risk from a global perspective. Its mathematical form guarantees the optimality principle of dynamic programming, ensuring that the obtained solution is globally optimal, demonstrating a significant peak-shaving advantage compared to local greedy strategies or rule-based scheduling.

[0096] Step S110: Output an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence, as the final scheduling instruction for the target reservoir.

[0097] Specifically, the optimal storage capacity change path V obtained through backtracking will be... opt (t) and the corresponding optimal outflow sequence Q opt (t) Output the final scheduling plan for the target reservoir during this flood event in tabular or graphical form. This plan clarifies the reservoir capacity target and gate discharge for each scheduling period, and can be directly used to guide actual flood control scheduling operations. Outputting a clear optimized scheduling plan provides reservoir managers with scientific and executable scheduling instructions, realizing a closed loop from theoretical optimization to engineering practice, and improving the intelligence level and decision-making efficiency of reservoir flood control scheduling.

[0098] For example, let's take flood control scheduling that includes three moments: t=0 (starting point), t=1 (intermediate point), and t=2 (ending point).

[0099] Backtracking: Starting from the optimal endpoint "capacity C", read its records and trace back to "capacity A" at t=1; then, starting from "capacity A", read the records and trace back to "initial capacity" at t=0. Finally, connect them in ascending time order to reconstruct the complete globally optimal scheduling path (initial capacity -> capacity A -> capacity C).

[0100] Furthermore, in some preferred embodiments of the present invention, the engineering constraint boundary includes at least one of the following: starting water level, dead water level, warning water level, flood control high water level, and check flood level; the method also includes: comparing and analyzing the optimized scheduling scheme with the benchmark scheme obtained by micro-step simulation based on preset scheduling rules, and the comparison index includes at least one of the following: highest water level, maximum outflow, and peak reduction rate.

[0101] Specifically, the engineering constraint boundaries can be set according to the actual safe operation procedures of the target reservoir. For example, the initial adjustment water level is the initial water level for scheduling, the dead water level is the minimum allowable water level, the warning water level is the water level that needs to be closely monitored, and the flood control high water level and the check flood level are the ultimate safe water levels. These constraints are used for truncation correction of the search interval in step S106. The flexible implementation of engineering constraints enables the method to adapt to the safety procedures of different reservoirs, and has strong versatility.

[0102] After outputting the optimized solution, the system automatically compares it with the baseline solution for rule-based scheduling, calculating key indicators such as the rise and fall of the highest water level, the reduction in the maximum outflow, and the peak reduction rate ((maximum outflow under rule - maximum outflow under optimization) / peak inflow), and presenting the optimization benefits in chart form. The comparative analysis with the baseline solution provides dispatchers with intuitive quantitative data on benefits, helping to evaluate the practical value of the optimization method and promoting its application and decision-making in engineering practice.

[0103] For example, taking a reservoir as an example, two core physical characteristic curves of the reservoir are constructed using an interpolation algorithm (such as linear interpolation): the water level-storage capacity curve (ZV curve) and the water level-discharge capacity curve (ZQ curve). The ZV curve data points include a water level of 100m corresponding to a reservoir capacity of 10 million cubic meters. A water level of 115m corresponds to a reservoir capacity of 150 million cubic meters. A water level of 125m corresponds to a reservoir capacity of 300 million cubic meters. .

[0104] The ZQ curve defines the physical discharge limit of hydraulic structures (such as spillways and gates) at different water levels. Specifically, the discharge capacity is 0 at a water level of 100m. The water level is 120m and the value is 1500. The water level is 3000 when it is 125m. .

[0105] In setting the input boundary conditions, this embodiment constructs a typical single-peak flood process as the excitation input for the system. To ensure the standardization and controllability of the verification process, this embodiment specifically selects a Gaussian function to simulate the inflow flood process. The main reason for selecting the Gaussian function is that: the bell-shaped curve characteristic of the Gaussian function mathematically closely matches the "rising-peak-receding" process of a single-peak flood in nature; at the same time, the function has good continuity and smoothness, which facilitates the elimination of numerical oscillations caused by data noise in algorithm verification; in addition, by adjusting the function parameters, the occurrence time and peak size of the flood peak can be precisely controlled, which is convenient for constructing a standard test set. The specific formula for calculating the inflow is as follows:

[0106] ;

[0107] in for Inflow rate at any given time, expressed in cubic meters per second (m³ / s) ); Indexed by time (hours). This process simulates a field with a base current of 200. The peak flood volume was 800. (That is, the total flood peak is about 1000) The total scheduling duration is set to 48 hours, and the time step is (). The time limit is 3600 seconds, the initial water level is 110.0m, and the warning water level constraint for downstream flood control or dam operation is set at 115.0m.

[0108] Before executing the global optimization algorithm, this embodiment first runs a high-precision simulation based on conventional rules. The purpose of this step is to obtain a benchmark scheme and determine the effective solution space of the subsequent optimization algorithm through rule scheduling. In order to solve the problems of water balance calculation error and lag in judging discharge capacity constraints under large time steps (such as 1 hour), this embodiment introduces micro-step technology, which further subdivides the outer time step into 60 micro-steps (i.e., 1 minute each) for iterative calculation.

[0109] The specific rule logic for regular scheduling is as follows:

[0110] Set a constant base current (e.g., 10) ) and evaporation loss (set to 0 in this example).

[0111] Hierarchical control strategy:

[0112] Condition A (water level < 112.0m): At this time, the water level is low, the gate remains closed, and the discharge flow is only the base flow.

[0113] Operating Condition B (112.0m ≤ water level < 120.0m): Entering the normal dispatching range. The target discharge flow rate is the minimum of the following three values: ① current inflow rate, ② preset control flow rate (700... ), ③ The maximum physical discharge capacity corresponding to the current water level. That is: .

[0114] Condition C (Water level ≥ 120.0m): The water level is high, requiring increased discharge. The target discharge volume is limited to [700m]. Between 700 and 700. That is: if the inflow is less than 700, 700 will be released to accelerate the receding of water; if the inflow is greater than the maximum capacity, the flood discharge will be fully opened; otherwise, the discharge will be based on the inflow.

[0115] Based on simulations using this rule, the baseline scenario yields a maximum water level of 114.17m and a maximum outflow of 552.19m. The simulated sequence of changes in storage capacity ( This will serve as an important input for subsequent steps.

[0116] Based on the above simulation results, this embodiment abandons the traditional full-capacity discretization method and instead constructs an adaptive search grid. Specifically, the algorithm first extracts the extreme value range of the simulated reservoir capacity sequence, and then calculates the buffer space based on this. The size of the buffer space is dynamically determined according to the maximum possible inflow volume in a single step, and finally locks the search reservoir capacity range to approximately [5650, 15000] ten thousand. Because a warning water level of 115.0m was set, the algorithm forcibly limited the upper bound of the search to the reservoir capacity corresponding to the warning water level (150 million cubic meters). Within a certain range, infeasible solutions exceeding the warning water level are directly eliminated at the solution space construction level. Simultaneously, to prevent state transition "deadlock" caused by an excessively sparse mesh, the algorithm adaptively calculates the number of discrete states based on the time step precision (157 state points in this example).

[0117] Generate a set of discrete storage capacity states .

[0118] Within a defined grid space, a dynamic programming (DP) algorithm is used to solve the Min-Max problem of minimizing the maximum outbound flow. During the recursive process, the state transition equation is defined as:

[0119] ;

[0120] in, The state from the previous moment is obtained by reverse calculation using the water balance equation. Transition to the current state The required outbound flow rate. In each state transition calculation, the algorithm performs rigorous physical constraint checks to ensure the accuracy of the calculated results. The physical discharge capacity shall not exceed the corresponding water level at any given time. From the end time Among all feasible states, the state with the minimum peak flow is selected as the endpoint, and the optimal outflow flow sequence is obtained by backtracking in reverse. and optimal storage capacity sequence .

[0121] The optimized scheduling scheme is compared with the conventional rule-based scheduling scheme. Based on the calculation data of this embodiment (see Table 1 for details):

[0122] Table 1

[0123]

[0124] See Figure 2The diagram shown in this embodiment of the invention illustrates a single reservoir scheduling scheme, comparing and analyzing the optimized scheme proposed in this embodiment with the conventional rule-based scheduling scheme. Data indicates that under the same inflow flood (peak flood 1000...),... Under the given conditions and starting regulation, although the conventional rules ensured that the water level (114.17m) did not exceed the warning line, its passive regulation resulted in a maximum discharge flow of 552.19 m. In contrast, the optimized method in this embodiment can fully utilize the limited flood control capacity between 114.17m and the warning water level of 115.00m. Through the strategy of "pre-release and peak shaving," the maximum outflow is significantly reduced to 357.12 cubic meters per second, while adhering to the 115.00m warning water level limit. The results show that the method in this embodiment increases the peak shaving rate of the reservoir to 64.29%.

[0125] This invention provides a single reservoir scheduling method based on adaptive grid and micro-step simulation, comprising: constructing a digital physical model of the target reservoir; wherein the digital physical model includes: a bidirectional mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity; performing micro-step simulation of the target inflow flood process based on preset scheduling rules to obtain an initial reservoir capacity change sequence reflecting the basic response characteristics of the target reservoir; adaptively constructing a dynamic programming search grid based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and preset engineering constraint boundaries; wherein the dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical attributes corresponding to the discrete reservoir capacity state points; on the dynamic programming search grid, with the objective of minimizing the maximum outflow during the entire scheduling process, using a dynamic programming algorithm for global optimization to solve for the optimal reservoir capacity change path that satisfies the preset physical constraints, and... Based on this, the corresponding optimal outflow sequence is calculated; an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence is output as the final scheduling instruction for the target reservoir; through micro-step simulation, the integral error of traditional long-step calculation is overcome, making the rule simulation results closer to the actual hydraulic process of the reservoir, providing a high-fidelity initial sequence for optimization; based on the simulation results, the effective reservoir capacity range is dynamically locked, and the search grid is adaptively constructed in combination with the inflow intensity, which greatly compresses the state space while ensuring accuracy and solves the problem of dimensionality curse; through physical-driven grid density calculation, it is ensured that the difference in water volume between adjacent discrete states is less than the single-step regulation capacity of the reservoir, eliminating "grid deadlock" from the root; dual capacity constraint verification is introduced into dynamic programming, so that the optimization scheme meets the engineering physical limits throughout the time period, avoiding the failure of the mathematically optimal solution due to insufficient discharge capacity; it supports the flexible implantation of multiple engineering constraints, maximizing the peak reduction rate under the premise of ensuring safety, realizing the synergistic optimization of safety and efficiency, and significantly improving the flood control scheduling efficiency of the reservoir.

[0126] Example 2

[0127] Based on the above embodiments, this invention provides a single reservoir scheduling device based on adaptive grid and micro-step size simulation, see [link to relevant documentation]. Figure 3 The diagram shown illustrates a single reservoir scheduling device based on adaptive grid and micro-step size simulation, according to an embodiment of the present invention. The device includes:

[0128] The physical model building module 310 is used to construct a digital physical model of the target reservoir; wherein, the digital physical model includes: a two-way mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity;

[0129] The basic scheduling simulation module 320 is used to perform micro-step simulation of the target inflow flood process based on preset scheduling rules, and obtain the initial reservoir capacity change sequence that reflects the basic response characteristics of the target reservoir.

[0130] The adaptive grid processing module 330 is used to adaptively construct a dynamic programming search grid based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and the preset engineering constraint boundary. The dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points.

[0131] The scheduling path global optimization module 340 is used to perform global optimization on the dynamic programming search grid with the goal of minimizing the maximum outbound flow in the entire scheduling process. It uses dynamic programming algorithm to solve for the optimal storage capacity change path that meets the preset physical constraints, and calculates the corresponding optimal outbound flow sequence accordingly.

[0132] The optimization scheme processing module 350 is used to output an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence, which serves as the final scheduling instruction for the target reservoir.

[0133] Furthermore, in some preferred embodiments of the present invention, the basic scheduling simulation module 320 is used to subdivide the preset standard hydrological time period step into multiple micro-steps based on a preset segmentation constant; wherein, the segmentation constant is an integer greater than or equal to 1; within each micro-step, the following are executed sequentially: back-calculating the water level based on the current reservoir capacity, calculating the target discharge according to the preset scheduling rules, limiting the target discharge according to the maximum discharge capacity corresponding to the current water level, and updating the reservoir capacity using the water balance equation; the simulation results of all micro-steps are aggregated according to the standard hydrological time period to output the initial reservoir capacity change sequence.

[0134] Furthermore, in some preferred embodiments of the present invention, the adaptive grid processing module 330 is used to extract the minimum and maximum values ​​of the initial reservoir capacity change sequence; dynamically calculate the adaptive buffer amount based on the ratio of the current inflow intensity to the total reservoir capacity of the target reservoir; construct a preliminary search interval based on the minimum and maximum values ​​of the initial reservoir capacity change sequence and the adaptive buffer amount; and introduce engineering constraint boundaries to truncate and correct the preliminary search interval to obtain the final effective search interval.

[0135] Furthermore, in some preferred embodiments of the present invention, the adaptive grid processing module 330 is used to determine the maximum water volume change capacity of the target reservoir within a single time period based on the mapping relationship between water level and maximum discharge capacity; and to determine the state discretization step size; wherein the state discretization step size is less than or equal to the ratio of the maximum water volume change capacity to the preset safety connectivity coefficient.

[0136] Furthermore, in some preferred embodiments of the present invention, the adaptive grid processing module 330 is used to determine the corresponding water level for each generated discrete reservoir capacity state point through the bidirectional mapping relationship between water level and reservoir capacity; calculate the maximum physical discharge capacity corresponding to the water level through the mapping relationship between water level and maximum discharge capacity; and construct a lookup table from the mapping results.

[0137] Furthermore, in some preferred embodiments of the present invention, the global optimization module 340 for scheduling paths is used to define state transition equations and back-calculate the required average outflow from the reservoir capacity state at adjacent times based on the water balance principle; wherein, a preset dual capacity constraint verification is introduced during the state transition process to ensure that the back-calculated outflow simultaneously meets the maximum discharge capacity limit corresponding to the start and end states of the transition; global optimization is performed using a recursive formula that minimizes the maximum outflow; and the global optimal reservoir capacity path is reconstructed by backtracking from the end time.

[0138] Furthermore, in some preferred embodiments of the present invention, the dual capability constraint verification includes: verifying whether the maximum discharge capacity corresponding to the reservoir capacity status at the start of the transfer is not less than the deduced outflow rate; and verifying whether the water level corresponding to the reservoir capacity status at the end of the transfer can support the deduced outflow rate.

[0139] Furthermore, in some preferred embodiments of the present invention, the global optimization module 340 for scheduling paths, used to minimize the recursive formula for the maximum outbound flow, includes: ;in, To arrive at time from the initial time status The minimum of the maximum outbound flow experienced across all paths; Let be the storage capacity at time t; j is the storage capacity grid point at time t; Let be the storage capacity value at grid point j at time t; Let be the average outflow that the reservoir needs to discharge from its state at time t-1 to its state at time t.

[0140] Furthermore, in some preferred embodiments of the present invention, the engineering constraint boundary includes at least one of the following: starting water level, dead water level, warning water level, flood control high water level, and check flood level; the device also includes: a comparison analysis module, used to compare and analyze the optimized scheduling scheme with the benchmark scheme obtained by micro-step simulation based on preset scheduling rules, and the comparison index includes at least one of the following: highest water level, maximum outflow, and peak reduction rate.

[0141] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the single reservoir scheduling device based on adaptive grid and micro-step simulation described above can be referred to the corresponding process in the aforementioned embodiments of the single reservoir scheduling method based on adaptive grid and micro-step simulation, and will not be repeated here.

[0142] Example 3

[0143] This invention also provides an electronic device for running a single reservoir scheduling method based on adaptive grid and micro-step simulation; see also Figure 4 The schematic diagram of an electronic device provided by an embodiment of the present invention is shown. The electronic device includes a memory 400 and a processor 401. The memory 400 is used to store one or more computer instructions, and the one or more computer instructions are executed by the processor 401 to realize the above-mentioned single reservoir scheduling method based on adaptive grid and micro-step simulation.

[0144] Furthermore, Figure 4 The electronic device shown also includes a bus 402 and a communication interface 403. The processor 401, the communication interface 403 and the memory 400 are connected via the bus 402.

[0145] The memory 400 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 403 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc. The bus 402 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0146] Processor 401 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 401 or by instructions in software form. Processor 401 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a readily available storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 400, and processor 401 reads information from memory 400 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0147] This invention also provides a computer-readable storage medium storing computer-executable instructions. When these computer-executable instructions are called and executed by a processor, they cause the processor to implement the above-described single reservoir scheduling method based on adaptive grid and micro-step simulation. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0148] The computer program product of the single reservoir scheduling method, device and electronic device based on adaptive grid and micro-step simulation provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0149] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and / or device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0150] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0151] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A single reservoir scheduling method based on adaptive grid and micro-step size simulation, characterized in that, include: Construct a digital physical model of the target reservoir; wherein, the digital physical model includes: a two-way mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity; Based on preset scheduling rules, a micro-step simulation of the target inflow flood process is performed to obtain the initial reservoir capacity change sequence that reflects the basic response characteristics of the target reservoir. Based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and the preset engineering constraint boundary, a dynamic programming search grid is adaptively constructed; wherein, the dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points; On the dynamic programming search grid, with the goal of minimizing the maximum outbound flow during the entire scheduling process, a dynamic programming algorithm is used for global optimization to find the optimal storage capacity change path that satisfies the preset physical constraints, and the corresponding optimal outbound flow sequence is calculated accordingly. The output includes an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence, which serves as the final scheduling instruction for the target reservoir. The steps for performing micro-step simulation of the target reservoir inflow process based on preset scheduling rules to obtain the initial reservoir capacity change sequence reflecting the basic response characteristics of the target reservoir include: The preset standard hydrological time period step is subdivided into multiple micro-steps based on a preset segmentation constant; wherein, the segmentation constant is an integer greater than or equal to 1; Within each microstep, the following steps are executed sequentially: calculating the water level based on the current reservoir capacity, calculating the target discharge volume according to the preset scheduling rules, limiting the target discharge volume according to the maximum discharge capacity corresponding to the current water level, and updating the reservoir capacity using the water balance equation. All simulation results with microsteps are aggregated according to the standard hydrological time period to output the initial reservoir capacity change sequence. On the dynamic programming search grid, the steps for global optimization using a dynamic programming algorithm, with the objective of minimizing the maximum outbound flow during the entire scheduling process, include: Define a state transition equation and back-calculate the required average outflow from the reservoir capacity state at adjacent times based on the water balance principle. In the state transition process, a preset dual capacity constraint check is introduced to ensure that the back-calculated outflow simultaneously meets the maximum discharge capacity limit corresponding to the start and end states of the transition. A recursive formula that minimizes the maximum outbound flow rate is used for global optimization; By backtracking from the termination point, the globally optimal storage capacity path is reconstructed.

2. The single reservoir scheduling method based on adaptive grid and micro-step simulation according to claim 1, characterized in that, The steps to determine the valid search range include: Extract the minimum and maximum values ​​of the initial storage capacity change sequence; The adaptive buffer amount is dynamically calculated based on the ratio of the current inflow intensity to the total capacity of the target reservoir, and an initial search interval is constructed based on the minimum and maximum values ​​of the initial capacity change sequence and the adaptive buffer amount. The initial search interval is truncated and corrected by introducing the engineering constraint boundary to obtain the final effective search interval.

3. The single reservoir scheduling method based on adaptive grid and micro-step simulation according to claim 1, characterized in that, The steps for calculating the discretization step size that guarantees state reachability include: Based on the mapping relationship between the water level and the maximum discharge capacity, the maximum water volume change capacity of the target reservoir within a single time period is determined; Determine the state discretization step size; wherein the state discretization step size is less than or equal to the ratio of the maximum water volume change capability to the preset safe connectivity coefficient.

4. The single reservoir scheduling method based on adaptive grid and micro-step simulation according to claim 1, characterized in that, The steps of generating discrete storage capacity state points and pre-calculating the physical properties corresponding to the discrete storage capacity state points include: For each of the generated discrete reservoir capacity state points, the corresponding water level is determined through the bidirectional mapping relationship between the water level and the reservoir capacity; The maximum physical discharge capacity corresponding to the water level is calculated by the mapping relationship between the water level and the maximum discharge capacity; The mapping results are used to construct a lookup table.

5. The single reservoir scheduling method based on adaptive grid and micro-step simulation according to claim 1, characterized in that, The dual capability constraint verification includes: Verify whether the maximum discharge capacity corresponding to the storage capacity status at the start of the transfer is not less than the deduced outflow rate; Verify whether the water level corresponding to the reservoir capacity status at the time of transfer termination can support the inferred outflow rate.

6. The single reservoir scheduling method based on adaptive grid and micro-step simulation according to claim 1, characterized in that, The recursive formula for minimizing the maximum outbound flow includes: ； in, To arrive at time from the initial time status The minimum maximum outflow rate experienced along all paths; j is the storage capacity grid point at time t; Let be the storage capacity value at grid point j at time t; Let be the average outflow that the reservoir needs to discharge from its state at time t-1 to its state at time t.

7. The single reservoir scheduling method based on adaptive grid and micro-step simulation according to claim 1, characterized in that, The engineering constraint boundaries include at least one of the following: starting water level, dead water level, warning water level, flood control high water level, and check flood level; The method further includes: The optimized scheduling scheme is compared and analyzed with the benchmark scheme obtained by micro-step simulation based on preset scheduling rules. The comparison indicators include at least one of the following: highest water level, maximum outflow, and peak shaving rate.

8. A single reservoir scheduling device based on adaptive grid and micro-step simulation, characterized in that, include: The physical model building module is used to construct a digital physical model of the target reservoir; wherein, the digital physical model includes: a two-way mapping relationship between water level and reservoir capacity, and a mapping relationship between water level and maximum discharge capacity; The basic scheduling simulation module is used to perform micro-step simulation of the target reservoir inflow flood process based on preset scheduling rules, and obtain the initial reservoir capacity change sequence reflecting the basic response characteristics of the target reservoir. An adaptive grid processing module is used to adaptively construct a dynamic programming search grid based on the dynamic range of the initial reservoir capacity change sequence, combined with the inflow flood intensity and preset engineering constraint boundaries; wherein, the dynamic programming search grid includes: determining the effective search interval, calculating the discretization step size to ensure state reachability, and generating discrete reservoir capacity state points and pre-calculating the physical properties corresponding to the discrete reservoir capacity state points; The global optimization module for scheduling paths is used to perform global optimization on the dynamic programming search grid with the goal of minimizing the maximum outbound flow during the entire scheduling process. It uses a dynamic programming algorithm to solve for the optimal storage capacity change path that satisfies the preset physical constraints, and calculates the corresponding optimal outbound flow sequence accordingly. The optimization scheme processing module is used to output an optimized scheduling scheme containing the optimal reservoir capacity sequence and the outflow sequence, which serves as the final scheduling instruction for the target reservoir. The basic scheduling simulation module is used to subdivide the preset standard hydrological time period step into multiple micro-steps based on a preset segmentation constant; wherein, the segmentation constant is an integer greater than or equal to 1; within each micro-step, the following are executed sequentially: back-calculating the water level based on the current reservoir capacity, calculating the target discharge according to the preset scheduling rules, limiting the target discharge according to the maximum discharge capacity corresponding to the current water level, and updating the reservoir capacity using the water balance equation; the simulation results of all micro-steps are aggregated according to the standard hydrological time period, and the initial reservoir capacity change sequence is output; The global optimization module for the scheduling path is used to define the state transition equation and, based on the principle of water balance, back-calculate the required average outflow from the reservoir capacity state at adjacent times. During the state transition process, a preset dual capacity constraint check is introduced to ensure that the back-calculated outflow simultaneously meets the maximum discharge capacity limits corresponding to the start and end states of the transition. A recursive formula minimizing the maximum outflow is used for global optimization. The module then backtracks from the end time to reconstruct the globally optimal reservoir capacity path.

Images