Coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making method and system for cross-basin water resources system

Abstract

The present invention belongs to the technical field of intelligent hydrological and water resource management, and discloses a method and system for coordinated scheduling and dynamic decision-making for positive and negative bidirectional coupling across river basin water resource systems. The method comprises: S1: constructing a multimodal water conservancy infrastructure topological relationship map and performing bidirectional transmission optimization; S2: based on the multimodal water conservancy infrastructure topological relationship map and optimized bidirectional transmission, sensing the hydrological gap risk situation and analyzing the coupling of multi-source water supply potential; S3: based on the coupling of the hydrological gap risk situation and multi-source water supply potential, generating a flexible graded water allocation strategy and a cross-project replacement agreement; S4: based on the flexible graded water allocation strategy and the cross-project replacement agreement, constructing and optimizing a "positive and negative" bidirectional pre-rehearsal decision-making engine. The present invention solves the technical difficulties faced by traditional one-way scheduling models in coping with complex topological relationships, sudden water shortages, and cross-administrative region coordination.

Description

Technical Field

[0001] The present invention belongs to the field of intelligent management technology of hydrology and water resources, and is particularly suitable for intelligent decision-making support in scenarios of coordinated scheduling of multiple water source projects, dynamic supply and demand matching, and emergency reverse water flow conduction. It specifically relates to a method and system for coordinated scheduling of cross-basin water resource systems and positive-negative bidirectional coupling dynamic decision-making. Background Art

[0002] Traditional scheduling technologies primarily rely on one-way forward simulations and one-way water flow conduction models, simulating the one-way water transfer process from a water source project to users. For example, these technologies optimize water allocation through static topology networks. However, these technologies lack the ability to simulate reverse scheduling. Their static topology networks cannot support reverse water flow back-calculation, making it impossible to reverse-calculate project operating boundaries and scheduling plans based on water shortages.

[0003] The limitations of the existing forward scheduling model are further reflected in the inefficiency of the cross-regional coordination mechanism. Water transfer across administrative regions relies on manual negotiation and experience-based decision-making, and lacks a quantitative contribution model, leading to frequent disputes. The lack of coupling between this manual intervention and one-way simulation makes it difficult for scheduling plans to dynamically balance ecological and economic needs. The physical boundaries and equity distribution issues of water conservancy facilities in reverse water flow transmission have long been unresolved. The technical bottlenecks of traditional simulation systems are also significant. Although rigid water allocation strategies (such as fixed-priority step-by-step reductions) can guarantee basic needs, the lack of a flexible mechanism to fill water demand gaps leads to increased economic losses. More importantly, existing technologies are unable to embed physical constraints (such as cross-section design flow and reservoir characteristic water level) into the reverse simulation process, often generating scheduling plans that are theoretically feasible but unusable in engineering.

[0004] Overall, the existing technology system focuses on forward scheduling rehearsals, while reverse scheduling rehearsals suffer from systemic shortcomings: First, the ability to model reverse water transfer paths is insufficient, leading to delayed emergency responses; second, the inverse coupling between cross-regional coordination rules and physical constraints is insufficient, limiting the feasibility of the solution; and third, the rigid allocation strategy lacks a dynamic compensation mechanism, making losses difficult to control. These issues hinder the intelligent level of water resource scheduling, and there is an urgent need to overcome the bottleneck of reverse scheduling rehearsals and establish a "forward-reverse" two-way collaborative decision-making system. Summary of the Invention

[0005] The present invention is directly applied to digital systems of major water diversion projects such as river basin water resources management platforms, solving technical difficulties that traditional one-way scheduling models are difficult to cope with complex topological relationships, sudden water shortages and cross-administrative region collaboration. The present invention provides a cross-basin water resources system collaborative scheduling and positive-negative bidirectional coupling dynamic decision-making method and system. By constructing a two-way topological network model to support positive and negative water volume interaction, combined with a smart contract-driven cross-regional collaboration mechanism and a flexible priority dynamic allocation algorithm, the present invention uses a positive-negative bidirectional rehearsal engine and a water rights multi-criteria non-dominated sorting technology to generate a non-dominated water rights configuration set under physical constraints, and through dynamic weight adjustment and a multi-agent water rights Nash equilibrium mechanism, it realizes intelligent decision-making and dynamic optimization of water resources scheduling, significantly improving emergency response capabilities and cross-regional collaboration efficiency.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A cross-basin water resources system coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making method, the method comprising:

[0008] S1: Construct a multimodal water infrastructure topology map and perform bidirectional transmission optimization;

[0009] S2: Based on the multimodal water infrastructure topology relationship map and optimized bidirectional transmission, the risk situation of hydrological gap is perceived and coupled with the analysis of multi-source water supply potential;

[0010] S3: Generate flexible graded water allocation strategies and cross-project replacement agreements based on the coupling of hydrological deficit risk situation and multi-source water supply potential;

[0011] S4: Based on the flexible graded water allocation strategy and cross-project replacement agreement, build and optimize the "positive-negative" two-way pre-rehearsal decision engine.

[0012] Preferably, in S1, constructing a multimodal water conservancy infrastructure topology relationship map and performing bidirectional transmission optimization includes:

[0013] S1.1: Multi-dimensional correlation tensor modeling and adaptive weight evolution of water conservancy entities;

[0014] S1.2: Dynamic weight adjustment enabled by smart contracts and the construction of a multi-agent water rights Nash equilibrium mechanism.

[0015] Preferably, in said S1.1, the multi-dimensional correlation tensor modeling and adaptive weight evolution of water conservancy entities include:

[0016] Construction of source-sink relationship matrix and bidirectional water transfer rules;

[0017] Construction of supply and demand relationship matrix and intelligent adjustment of hierarchical water supply;

[0018] Regression relationship matrix construction and intelligent dynamic adjustment of recycled water replenishment;

[0019] Water resource partition matrix construction and federated learning collaborative scheduling;

[0020] Construction of administrative division matrix and cross-regional coordination mechanism;

[0021] The source-sink relationship matrix is constructed, including:

[0022] Determine the maximum water delivery capacity based on engineering design parameters and historical operating peak data;

[0023] Based on the maximum water delivery capacity, the source-sink relationship matrix is constructed by real-time monitoring of equipment health and combining machine learning algorithms to predict availability trends.

[0024] Construction of two-way water transfer rules, including:

[0025] Generate reverse water transfer paths based on the reverse water transfer capacity of the pumping station and the bidirectional overflow threshold of the pipeline;

[0026] Based on the reverse water transfer path, the weight of reverse water transfer is dynamically adjusted according to the water storage capacity of the water source project, equipment health, and emergency response priority parameters to establish a two-way water transfer rule;

[0027] Intelligent adjustment of graded water supply, including:

[0028] ;

[0029] in, is the water supply ratio of user i at time t, is the initial water supply ratio, is the current water storage capacity, To warn of water levels;

[0030] Intelligent dynamic adjustment of recycled water replenishment, including:

[0031]

[0032] in, is the recycled water replenishment ratio, is the maximum redemption ratio, is the amount of recycled water, is the water demand;

[0033] Water resource partition matrix construction and federated learning collaborative scheduling, including:

[0034]

[0035]

[0036]

[0037] in, is the model parameter, which represents the weight of the water resources scheduling model. Represents a partition The loss function is used to measure the error between the model prediction value and the actual value. For partition Local model parameters of ; is the optimal model parameter, which means that under the partition data, the model can minimize the parameter value of the scheduling error; is the global model parameter, which represents the model parameter after collaborative training of all partitions. Represents a partition The amount of data, that is, the scale of water resources data, is the total data volume, which represents the sum of the data volumes of all partitions. and Represents partitions respectively In the Second and The model parameters in the iteration, is the learning rate, which controls the step size of model parameter update. is the loss function The gradient of , which indicates the direction of change of the model parameters;

[0038] The construction of administrative division matrix and cross-regional coordination mechanism includes;

[0039] ;

[0040] in, For administrative districts The contribution of is the set of all administrative districts, is a subset, For subset benefits.

[0041] Preferably, S1.2: Dynamic weight adjustment enabled by smart contracts and construction of a multi-agent water rights Nash equilibrium mechanism, including:

[0042] Weight calculation and intelligent dynamic correction model construction;

[0043] Based on the constructed comprehensive weights, an improved multi-agent water rights Nash equilibrium rule is constructed;

[0044] Among them, the basic weight calculation includes:

[0045] Basic weight = physical transmission capacity × proportion of project water storage capacity;

[0046] Among them, physical transmission capacity refers to the maximum water transmission capacity of the water transmission path, and the proportion of project water storage capacity refers to the proportion of the current project water storage capacity to the total water storage capacity;

[0047] Intelligent dynamic correction model construction, including:

[0048] Dynamic correction factor = administrative coordination coefficient × emergency response priority;

[0049] Among them, the administrative coordination coefficient indicates the degree of coordination between administrative regions;

[0050] Comprehensive weight = basic weight × dynamic correction factor;

[0051] Constructing an improved multi-agent Nash equilibrium rule for water rights, including: constructing an improved Lemmings optimization algorithm (ALA); and arbitrating an atomic settlement protocol for water rights transactions based on the improved Lemmings optimization algorithm (ALA) and distributed ledgers.

[0052] Among them, the construction of the improved Lemming Optimization Algorithm ALA includes: based on the Lemming Optimization Algorithm ALA, introducing three improvement strategies: optical inverse mapping, Gaussian perturbation strategy and adaptive mutation;

[0053] Among them, optical inverse mapping: for each individual , generating a symmetric solution , and calculate the fitness value corresponding to the symmetric solution , and finally select the solution whose fitness value meets the preset requirements. The calculation formula is:

[0054] ; ;

[0055] in, and are the lower and upper bounds of the search space, Is the fitness value corresponding to each individual;

[0056] Gaussian perturbation strategy: for each individual , generate new candidate solutions , and calculate the fitness value corresponding to the new solution , and finally select the solution whose fitness value meets the preset requirements. The calculation formula is:

[0057] ; ;

[0058] in, The mean is 0 and the variance is Gaussian distributed random numbers, is the step size of Gaussian random walk, and the dynamic adjustment formula is:

[0059] ;

[0060] in, and are the maximum and minimum values of the step length, and are the maximum number of iterations and the current number of iterations respectively;

[0061] Adaptive mutation: Calculate the mutation probability of the current iteration , for each individual Generate its mutation vector , and generate new candidate solutions , calculate the fitness value corresponding to the new solution , and finally select the solution whose fitness value meets the preset requirements. The calculation formula is:

[0062] ; ; ; ;

[0063] in, and are the minimum and maximum values of the mutation probability respectively;

[0064] Among them, the atomic settlement agreement arbitration of water rights transactions based on distributed ledgers includes:

[0065] Calculate the contribution of each administrative district based on the Shapley value method ;

[0066] According to contribution Automatically allocate water transfer rights, the calculation formula is:

[0067] ;

[0068] in, For administrative districts water transfer rights, is the total water transfer volume;

[0069] When a cross-administrative district dispute occurs, the smart contract automatically calculates the contribution of each administrative district based on the preset Shapley value method and generates a rights and interests distribution plan.

[0070] Preferably, in S2, based on the multimodal water conservancy infrastructure topological relationship map and optimized bidirectional transmission, sensing the hydrological gap risk situation and analyzing the multi-source water supply potential are coupled, including:

[0071] S2.1: Dynamic assessment and early warning of water shortage in sections based on real-time hydrological data;

[0072] S2.2: Calculate the system water supply potential under engineering physics constraints. The system water supply potential includes: project water storage, external water transfer, and recycled water.

[0073] S2.3: Generate a multi-source water supply collaborative decision-making interval; wherein, the multi-source water supply collaborative decision-making interval is: adjustable water volume = total water supply potential - total rigid water demand.

[0074] Preferably, in S3, based on the coupling of the hydrological deficit risk situation and the multi-source water supply potential, a flexible graded water allocation strategy and a cross-project replacement agreement are generated, including:

[0075] S3.1: Construct a three-level dynamic allocation mode switching mechanism;

[0076] S3.2: Based on the three-level dynamic allocation mode switching mechanism, the balanced allocation mode and the time-space decoupling allocation;

[0077] S3.3: Optimize compensation patterns and cross-project replacement protocols based on spatiotemporal decoupling of allocations.

[0078] Preferably, in S3.2, based on the three-level dynamic allocation mode switching mechanism, the balanced allocation mode and the time-space decoupling allocation include:

[0079] Through time series analysis, the fluctuation components of different time scales are decoupled to give priority to meeting the peak water shortage period;

[0080] Dynamically adjust the water supply ratio in each period to achieve balanced distribution of water resources over time;

[0081] Based on the sum of the project's water storage capacity and incoming water, calculate the dynamic balance and allocation weight of water resources in spatial dimensions;

[0082] Among them, the water supply ratio of each time period is dynamically adjusted to achieve a balanced distribution of water resources over time, including:

[0083] Assume the time window size is , the data in the window is ,in For the Water demand during the time period;

[0084] Calculate the average water demand within the window and standard deviation :

[0085] ; ;

[0086] Use the data within the window to predict water demand for the next period , the calculation formula is:

[0087] ;

[0088] in, is the adjustment coefficient, which is used to control the fluctuation range of the predicted value;

[0089] Based on predicted water demand and adjustable water volume , calculate the water supply ratio for the next period ,if , then take , indicating that the water supply is sufficient; otherwise, it is distributed proportionally, and the calculation formula is:

[0090] ;

[0091] Based on the sum of the project's water storage capacity and incoming water, the dynamic balance and allocation weights of water resources in spatial dimensions are calculated, including:

[0092] ;

[0093] in, 、 、 Respectively The allocated water volume, water storage volume and water inflow volume of each project;

[0094] In S3.3, based on the temporal and spatial decoupling of allocation, the compensation model and cross-project replacement protocol are optimized, including:

[0095] Two-stage allocation strategy and elastic filling of water demand gap;

[0096] Cross-project replacement protocol and blockchain verification;

[0097] Among them, the two-stage allocation strategy includes:

[0098] ;

[0099] ;

[0100] Flexible filling of water gaps, including:

[0101] ;

[0102] ;

[0103] Cross-project replacement protocol and blockchain verification, including:

[0104] Establish the substitution rules between external water and local water to obtain the substitution ratio between different water sources; the substitution rules are:

[0105] ;

[0106] in, 、 are the replacement amounts of externally transferred water and local water, respectively;

[0107] Verification of atomic settlement protocol for water rights transactions based on distributed ledgers:

[0108] ;

[0109] The upper limit of the replacement ratio is the maximum value in the preset replacement rule library.

[0110] Preferably, in said S4, based on the flexible graded water allocation strategy and the cross-project replacement protocol, a "positive-negative" two-way pre-play decision engine is constructed and optimized, including:

[0111] S4.1: Forward risk deduction and target quantification;

[0112] S4.2: Reverse scheduling preview and multi-criteria non-dominated ranking of water rights;

[0113] Among them, the forward risk deduction and target quantification in S4.1 include:

[0114] Risk situation simulation and project water storage capacity assessment;

[0115] Entropy evaluation of user water demand gap and quantification of scheduling targets;

[0116] In S4.2, the reverse scheduling preview and water rights multi-criteria non-dominated sorting include;

[0117] Constraint inversion and physical constraint mapping;

[0118] Multi-criteria non-dominated sorting algorithm for water rights and generation of non-dominated water rights configuration sets.

[0119] Preferably, risk situation simulation and project water storage capacity assessment, including:

[0120] Through hydrological models and scheduling models, key indicators such as reservoir water storage, river flow, and user water demand in the future scheduling period are simulated;

[0121] Based on the simulation results, a thermal map of the project's water storage exceeding the limit is generated, marking the areas where the water storage exceeds the designed storage capacity. The calculation formula is:

[0122] ;

[0123] in, For the Time period position The water storage capacity exceeds the limit. For the Time period position The water storage capacity, For location Design storage capacity;

[0124] According to the simulation results, the spatial distribution map of the user water demand gap entropy value is drawn, and the high-risk areas are marked. The calculation formula is:

[0125] ;

[0126] in, 、 、 Respectively Time period position User water shortage rate, user water demand, and actual water supply;

[0127] User water demand gap entropy assessment and scheduling target quantification, including:

[0128] ;

[0129] in, 、 are the actual and target water supply or discharge flow respectively;

[0130] Constraint inversion and physical constraint mapping, including:

[0131] Dynamically adjust the minimum operating water level of the project based on real-time data and forecast data and maximum water lifting capacity , the calculation formula is:

[0132] ;

[0133] ;

[0134] in, For the The minimum operating water level of the time period, For the The maximum water extraction capacity during the period, and For dynamic adjustment function;

[0135] Through projection technology, the physical constraint, namely the minimum operating water level, and maximum water lifting capacity Mapped to the feasible solution space, the calculation formula is:

[0136] ;

[0137] in, 、 are feasible solutions and primal solutions respectively, is the set of physical constraints;

[0138] Water rights multi-criteria non-dominated sorting algorithm and non-dominated water rights configuration set generation, including:

[0139] Through the water rights multi-criteria non-dominated sorting algorithm, the three objectives of water supply stability, discharge compliance rate and scheduling cost are optimized simultaneously; among them, water supply stability :Minimize water supply fluctuations and maintain a long dispatch period The calculation formula is:

[0140] ;

[0141] in, 、 They are Actual and target water supply for the time period;

[0142] Discharge compliance rate :Maximize the rate of compliance of downstream flow rate:

[0143] ;

[0144] in, for Discharge flow rate during the period;

[0145] Scheduling costs : Minimize scheduling cost:

[0146] ;

[0147] in, 、 They are Physical constraints and total flow during the time period;

[0148] An improved multi-objective dream optimization algorithm, IMODOA, is used to output the non-dominated water rights configuration set. Based on the dream optimization algorithm DOA, dynamic forgetting dimension adjustment, adaptive weight adjustment, hybrid search strategy, and constraint processing mechanism improvement strategy are integrated to obtain the improved multi-objective dream optimization algorithm, IMODOA:

[0149] In the exploration phase, the number of iterations ranges from 0 to :

[0150] Dynamic forgetting dimension adjustment: the number of forgotten dimensions Dynamically adjusted with the number of iterations:

[0151] ;

[0152] in, It is The number of forgotten dimensions in iterations, is the number of initial forgotten dimensions, is the maximum number of iterations;

[0153] Memory strategy: Individuals remember the optimal solution of their group:

[0154] ;

[0155] in, For the In the iteration group The optimal solution is The individual location;

[0156] Forgetting and replenishing strategy: The individual randomly updates the position on the forgetting dimension:

[0157] ;

[0158] in, For the Individuals in the iteration In dimension location, For the In the iteration group The optimal solution is in the dimension location, and Dimensions The lower and upper limits of is the current iteration number, is the maximum number of iterations in the exploration phase, Respectively indicate the first, second, ... At the iteration, the selected dimension index in the forget dimension is forgotten;

[0159] Dream sharing strategy: Individuals randomly obtain the location information of other individuals in the forgotten dimension:

[0160] ;

[0161] in, 、 Respectively sequence Individuals in the iteration In dimension location; For range A natural number randomly selected from the source represents the individual index;

[0162] During the development phase, the number of iterations starts from arrive :

[0163] Dynamic forgetting dimension adjustment: the number of forgetting dimensions Dynamically adjusted with the number of iterations:

[0164] ;

[0165] in, It is The number of forgotten dimensions in the iteration, is the number of initial forgotten dimensions;

[0166] Memory strategy: Individuals remember the optimal solution of the entire population:

[0167] ;

[0168] in, For the The optimal solution of the entire population in the iteration becomes the The individual location;

[0169] Forgetting and replenishing strategy: The individual randomly updates the position on the forgetting dimension:

[0170] ;

[0171] in, For the The optimal solution of the entire population in the iteration is in dimension location;

[0172] Adaptive weight adjustment: Dynamically adjust the weight of the objective function according to the convergence of the objective function:

[0173] ;

[0174] in, For the The objective function in the iteration The weight of For the The optimal solution in the iteration The objective function value of

[0175] Constraint handling mechanism:

[0176] Constraint correction: Correct solutions that violate constraints:

[0177] ;

[0178] Constraint Penalty: Penalize solutions that violate constraints:

[0179] ;

[0180] in, is the penalty coefficient, For population The optimal solution of For population The penalty function when the optimal solution is reached.

[0181] The present invention also provides a cross-basin water resources system coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making system, which is used to implement the above-mentioned method. The system includes: a graph construction module, a perception and analysis module, a generation module, and a strategy optimization module;

[0182] The graph construction module is used to construct a multi-modal water conservancy infrastructure topology relationship graph and perform bidirectional transmission optimization;

[0183] The perception and analysis module is used to perceive the hydrological gap risk situation and analyze the coupling of multi-source water supply potential based on the multi-modal water conservancy infrastructure topological relationship map and optimized bidirectional transmission;

[0184] The generation module is used to generate a flexible hierarchical water allocation strategy and a cross-project replacement agreement based on the coupling of the hydrological gap risk situation and the multi-source water supply potential;

[0185] The strategy optimization module is used to build and optimize a "positive-negative" two-way pre-rehearsal decision engine based on a flexible graded water allocation strategy and a cross-project replacement protocol.

[0186] Compared with the prior art, the present invention has the following beneficial effects:

[0187] This invention systematically solves the core problem of weak reverse scheduling rehearsal capability in the existing technology by constructing a bidirectional reversible topology network model, introducing a cross-regional collaboration mechanism driven by smart contracts, designing a flexible priority dynamic allocation algorithm, and developing a forward-reverse bidirectional rehearsal engine. Its beneficial effects are mainly reflected in the following aspects:

[0188] 1. Omnidirectional topology network improves emergency reverse flow conduction response capabilities:

[0189] The topological network based on dynamic weighted bidirectional graph modeling breaks through the limitations of traditional one-way water transmission paths and realizes real-time analysis of the interaction between two-way water volume between nodes of water source projects. By embedding parameters such as the reverse water transmission capacity of the pump station and the two-way overflow threshold of the pipeline into the network weight calculation, the system can automatically generate an emergency scheduling plan that includes the reverse water transmission path. Compared with the existing technology that only supports the forward rehearsal mode, the present invention can significantly shorten the generation time of the reverse scheduling plan in the emergency scenario, and the engineering feasibility verification pass rate of the reverse water transmission path in the plan is significantly improved, solving the problem of delayed emergency response from the root.

[0190] 2. Smart contracts ensure the efficiency and fairness of cross-regional coordinated scheduling:

[0191] A cross-administrative water transfer collaboration platform is established through an atomic settlement protocol for water rights transactions based on a distributed ledger. Parameters such as water contribution, ecological compensation coefficient, and historical water use rights are encoded into automatically executable contractual rules. When reverse water flow triggers a cross-regional dispute, the system generates a real-time equity allocation plan based on a pre-set quantitative model (such as the Shapley value method for allocating contribution), replacing traditional manual negotiation mechanisms. This technical solution not only reduces dispute resolution time to a negligible level but also, through a dynamic mapping model of contribution and compensation value, ensures that water-exporting regions receive reasonable ecological compensation for reverse water flow, thus achieving a virtuous cycle of "equal rights and responsibilities, and immediate settlement" for cross-regional water transfer.

[0192] 3. A forward-reverse bidirectional rehearsal engine ensures the engineering feasibility of the scheduling plan:

[0193] By bidirectionally coupling a forward water propagation model with a backward constraint inversion algorithm, a simulation and verification system driven by engineering physics constraints was constructed. The forward simulation simulates the water delivery process from the water source project to the user end. The backward simulation, starting from the user's water shortage, reversely derives the water dispatch path that meets the project's capacity boundaries (such as the maximum head of the pumping station and the pipeline's pressure limit). The two-stage simulation results are iterated through alternating iterations to search for the global optimal solution, ensuring that the generated dispatch plan meets water allocation requirements while strictly adhering to the project's safe operation constraints.

[0194] 4. Systematically improve the level of intelligent decision-making in water resource scheduling:

[0195] The synergistic effect of these technical solutions propels inter-basin water transfer systems from "unidirectional static scheduling" to "positive and negative bidirectional dynamic optimization." Through the deep integration of bidirectional topological network modeling, smart contract collaboration, elastic allocation algorithms, and a rehearsal engine, this invention improves emergency response speed, cross-regional collaboration efficiency, socioeconomic loss control, and the coupling degree of physical boundaries of water conservancy facilities, providing core technical support for addressing the challenges of water resource management under climate change. BRIEF DESCRIPTION OF THE DRAWINGS

[0196] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0197] Figure 1A flow chart of a cross-basin water resources system coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making method according to an embodiment of the present invention;

[0198] Figure 2 This is a schematic diagram of the recycled water reuse path according to an embodiment of the present invention;

[0199] Figure 3 Schematic diagram of the cross-region coordination mechanism according to an embodiment of the present invention. DETAILED DESCRIPTION

[0200] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0201] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0202] Example 1

[0203] As can be seen from the background technology,

[0204] This invention addresses the core defects of the reverse scheduling preview technology in the existing inter-basin water diversion system and aims to solve the following key problems:

[0205] Delayed response to emergency reverse water flow conduction: The existing forward scheduling model lacks the ability to model reverse water transmission paths, resulting in the inability to quickly generate emergency reverse water flow conduction plans in sudden water shortage events, seriously threatening water supply safety and project stability.

[0206] Inefficient cross-regional coordinated scheduling: The cross-administrative water allocation mechanism, which relies on manual negotiation, is lengthy and lacks a quantitative equity distribution model, making it difficult to ensure fairness and execution efficiency in reverse water flow transmission scenarios.

[0207] Socioeconomic losses of the rigid water allocation strategy: The fixed-priority step-by-step water supply reduction method leads to intensified water supply conflicts between high-economic-value users and low-priority users, resulting in unnecessary waste of resources and economic losses.

[0208] The scheduling plan is out of line with engineering physical constraints: The scheduling plan generated by reverse simulation often ignores key engineering constraints such as pump station head and reservoir dead water level, resulting in the inability to actually implement the theoretical deduction results.

[0209] By overcoming the above-mentioned technical bottlenecks, the present invention significantly improves the intelligent decision-making capabilities of the cross-basin water diversion system, achieving rapid response to emergency reverse water flow conduction, efficient cross-regional collaborative execution, dynamic and flexible allocation optimization, and a high degree of matching between the scheduling plan and the physical boundaries of water conservancy facilities.

[0210] The present invention provides a cross-basin water resource system coordinated scheduling and forward-reverse bidirectional coupling dynamic decision-making method, which supports forward and reverse water volume interaction by constructing a bidirectional topological network model, combines the cross-regional collaborative mechanism driven by smart contracts and the elastic priority dynamic allocation algorithm, and uses the forward-reverse bidirectional preview engine and water rights multi-criteria non-dominated sorting technology to generate a non-dominated water rights configuration set under physical constraints, and realizes intelligent decision-making and dynamic optimization of water resource scheduling through dynamic weight adjustment and multi-agent water rights Nash equilibrium mechanism, significantly improving emergency response capabilities and cross-regional collaboration efficiency. The basic flow chart of the present invention is as follows: Figure 1 As shown:

[0211] S1 Multimodal water infrastructure topology relationship map construction and bidirectional transmission optimization

[0212] Through a multi-dimensional relationship matrix, we systematically model the physical connections and scheduling rules between water source projects, establish a dynamic weight adjustment mechanism, and achieve accurate characterization and adaptive optimization of complex water resource systems. We focus on building a bidirectional water transfer topology between water source projects, breaking through the limitations of traditional unidirectional water flow models and supporting emergency reverse water diversion and cross-regional coordinated scheduling.

[0213] S1.1 Multi-dimensional correlation tensor modeling and adaptive weight evolution of water conservancy entities

[0214] 1. Construction of source-sink relationship matrix and bidirectional water transfer rules

[0215] (1) Physical connection modeling

[0216] Through topological network modeling technology, the topological structure of water transmission paths between water resource projects such as reservoirs, pumping stations, and dams is depicted, including various connection methods such as natural river channels, artificial channels, and pressure pipes. Specific modeling methods include:

[0217] ① Natural river channel: The water transfer path is constructed based on the hydrological characteristics of the river channel (such as flow, velocity, and cross-sectional shape), and the water transfer capacity of the river channel is calculated using the Manning formula:

[0218]

[0219] in, For traffic, is the Manning roughness coefficient, is the cross-sectional area of water flow, is the hydraulic radius, The slope of the water surface.

[0220] ② Artificial channel: Model the channel based on its design parameters (such as cross-sectional dimensions, slope, and lining materials) and use the open channel uniform flow formula to calculate the water transfer capacity, such as the Xie Cai formula:

[0221]

[0222] in, is the Xie Cai coefficient, which reflects the roughness of the channel and the water flow characteristics.

[0223] ③ Pressure pipeline: Based on the modeling of parameters such as pipeline diameter, material, and pressure level, it supports the evaluation of bidirectional water transmission capacity and uses the Darcy-Weisbach formula to calculate the pipeline head loss:

[0224]

[0225] in, is the head loss, is the friction coefficient, is the pipe length, is the pipe diameter, is the flow rate, is the acceleration due to gravity.

[0226] (2) Dynamic transmission capability

[0227] Dynamic transmission capability is the core feature of the source-sink relationship matrix, specifically including:

[0228] ① Maximum water transfer capacity: determined based on engineering design parameters (such as reservoir capacity and pump station rated flow) and historical operating peak data (such as historical maximum water transfer capacity) to ensure the engineering feasibility of the water transfer route.

[0229] ② Real-time availability: Dynamically adjust actual water delivery capacity based on factors such as equipment aging, siltation, and maintenance status. Real-time monitoring of equipment health (such as pump station efficiency and pipeline wear rate) is combined with machine learning algorithms to predict availability trends. The availability calculation formula is:

[0230]

[0231] in, is the real-time availability rate, is the initial availability rate, is the attenuation coefficient, For time.

[0232] (3) Two-way water transfer rules

[0233] Under emergency conditions, reverse water transfer is allowed (for example, reservoir A supplies reservoir B through a pumping station). Specific rules include:

[0234] ① Reverse water transfer path generation: Based on the reverse water transfer capacity of the pump station (such as reverse head, flow limit) and the bidirectional overflow threshold of the pipeline, a feasible reverse water transfer path is generated.

[0235] ② Emergency priority assessment: Dynamically adjust the weight of reverse water transfer based on parameters such as water storage capacity of water source projects, equipment health, and emergency response priority. The weight calculation formula is:

[0236]

[0237] in, is the path weight, The water storage capacity for water source projects, For device health, For emergency priority, is the weight coefficient.

[0238] (4) Intelligent dynamic adjustment of weights

[0239] The intelligent dynamic weight adjustment mechanism is the key innovation of the source-sink relationship matrix, specifically including:

[0240] ① Weight calculation model: Based on real-time data of parameters such as project water storage capacity, equipment health, and emergency priority, the connection weight is dynamically updated in real time to ensure dynamic optimization of the water transmission path.

[0241] Basic weight = physical transmission capacity × proportion of project water storage capacity

[0242] Dynamic correction factor = administrative coordination coefficient × emergency response priority

[0243] ② Alternative path search algorithm: When a water transmission channel interruption is detected, an alternative path search algorithm (such as Dijkstra shortest path optimization) is automatically triggered to generate an alternative water transmission path and update the weight. The path search formula of the Dijkstra algorithm is:

[0244]

[0245] in, For nodes The shortest path, For nodes The shortest path, For nodes To Node The weight of .

[0246] 2. Construction of supply and demand relationship matrix and intelligent adjustment of hierarchical water supply

[0247] (1) User category binding:

[0248] Establish a water supply priority list based on four categories of users: domestic, industrial, agricultural, and ecological, to ensure that the water needs of different user groups are met in a differentiated manner. Specific classification standards include:

[0249] ① Domestic water: highest priority, 100% guaranteed.

[0250] ②Industrial water use: medium priority, with flexible water supply agreement.

[0251] ③Agricultural water use: lower priority, with a certain degree of reduction allowed.

[0252] ④ Ecological water use: the lowest priority, but at least 80% is guaranteed during the flood season.

[0253] (2) Grading water supply threshold:

[0254] According to user priority and water supply capacity, set graded water supply thresholds and dynamically adjust water supply ratios:

[0255] ① Domestic water: 100% guaranteed, no reduction allowed.

[0256] ② Ecological water use: A minimum of 80% is guaranteed during the flood season, and dynamic adjustments are made based on actual conditions during the dry season.

[0257] ③Industrial water use: A flexible water supply agreement is adopted, allowing the actual water supply to fluctuate within ±15% of the contract value to balance the contradiction between supply and demand.

[0258] ④ Agricultural water use: Dynamically adjusted according to the water storage capacity of water source projects, with a minimum guarantee of 50%.

[0259] (3) Intelligent dynamic adjustment rules:

[0260] When the water storage capacity of the water source project is lower than the warning water level, the intelligent dynamic adjustment rules are automatically triggered:

[0261] ① Priority adjustment: reduce the water supply ratio for low-priority users (such as agriculture and ecology).

[0262] ② Flexible reduction mechanism: Dynamically adjust the water supply ratio for industrial users based on real-time supply and demand conditions to ensure water supply security for high-priority users (such as domestic water).

[0263] Relevant formula for water supply ratio adjustment:

[0264]

[0265] in, is the water supply ratio of user i at time t, is the initial water supply ratio, is the current water storage capacity, To warn of water levels.

[0266] 3. Construction of regression relationship matrix and intelligent dynamic adjustment of recycled water replenishment

[0267] (1) Water reuse path: Track the path of recycled water such as industrial wastewater and agricultural irrigation water back to the water source project, and dynamically adjust the replenishment ratio, including:

[0268] ① Industrial wastewater reuse: Model the reuse path based on the effluent quality and quantity of the industrial wastewater treatment plant.

[0269] ② Agricultural irrigation water withdrawal: Based on the seasonal changes in irrigation water withdrawal, model the water withdrawal reuse path.

[0270] (2) Recycled water replenishment rules: The replenishment ratio is dynamically adjusted based on the water storage capacity of the water source project and the quality of recycled water.

[0271] The formula for the recycled water replenishment ratio is:

[0272]

[0273] in, is the recycled water replenishment ratio, is the maximum redemption ratio, is the amount of recycled water, is the water demand. Figure 2 shown.

[0274] 4. Water Resource Partition Matrix Construction and Federated Learning Collaborative Scheduling

[0275] (1) Membership quantification:

[0276] Define the affiliation between users and the three-level water resource divisions, using a 0-1 value to represent the strength of the association. Specific methods include:

[0277] ① Membership calculation: Calculate the membership between the user and the partition based on the user's geographical location, water consumption, historical water rights and other parameters. The membership calculation formula is:

[0278]

[0279] in, For users The membership degree of is the weight coefficient, For the characteristic functions (such as geographical location, water consumption, etc.).

[0280] ② Partition scheduling strategy: Prioritize users with membership > 0.8. Cross-partition water transfer requires collaborative computing through federated learning. The core principles of federated learning include local model training, global model aggregation, and model update. The mathematical formulas are:

[0281]

[0282]

[0283]

[0284] in, is a model parameter, which represents the weight of the water resources scheduling model, such as the water release volume of the reservoir and the water delivery capacity of the channel. Represents a partition The loss function is used to measure the error between the model prediction value and the actual value. In water resource scheduling, the loss function can be expressed as the difference between the predicted water consumption and the actual water consumption. For partition Local model parameters of ; is the optimal model parameter, which means that under the partition data, the model can minimize the parameter value of the scheduling error; are global model parameters, which represent the model parameters after collaborative training of all partitions. Represents a partition The amount of data, that is, the scale of water resources data (such as the number of reservoirs and the number of users). is the total data volume, which represents the sum of the data volumes of all partitions. and Represents partitions respectively In the Second and The model parameters in iterations. is the learning rate, which controls the step size of model parameter updates. is the loss function The gradient of , which indicates the direction of change of the model parameters.

[0285] (2) Sensitive data protection:

[0286] Homomorphic encryption technology is used for sensitive data (such as industrial user locations) to ensure data privacy and security. Homomorphic encryption technology includes four core steps: encryption, homomorphic addition, homomorphic multiplication, and decryption. The corresponding mathematical formulas are:

[0287]

[0288]

[0289]

[0290]

[0291] in, It is plain text data, representing sensitive data in water resource scheduling (such as reservoir water storage and user water demand). is the public key, used to encrypt data. 、 、 All are ciphertext, indicating encrypted data. and Represent the ciphertext after addition and multiplication operations respectively. is the private key, used to decrypt data. The decrypted plaintext data.

[0292] 5. Construction of administrative division matrix and cross-regional coordination mechanism

[0293] (1) Cross-regional coordination mechanism:

[0294] Establish a benefit distribution model based on the Shapley value method to quantify the contribution of each administrative district. Specifically include:

[0295] ① Contribution calculation: Calculate the contribution based on the water consumption, water transfer volume, ecological compensation and other parameters of each administrative area. The Shapley value calculation formula is:

[0296]

[0297] in, For administrative districts The contribution of is the set of all administrative districts, is a subset, For subset benefits.

[0298] ② Equity distribution rules: The equity distribution plan is automatically executed through the atomic settlement protocol of water rights transactions based on distributed ledgers to ensure the fairness and efficiency of cross-regional water transfer. Figure 3 shown.

[0299] S1.2 Dynamic weight adjustment and multi-agent water rights Nash equilibrium mechanism enabled by smart contracts

[0300] The dynamic weight adjustment mechanism is a core module for optimizing water resource scheduling within the inter-basin water transfer system. By dynamically adjusting the weights of water transfer routes, the system can respond in real time to changes in project water storage, equipment health, and emergency response priorities, ensuring the flexibility and efficiency of scheduling solutions.

[0301] 1. Weight calculation and intelligent dynamic correction model

[0302] (1) Basic weight calculation: The basic weight reflects the relative importance of the physical transmission capacity of the water transmission path and the water storage capacity of the project. The calculation formula is:

[0303] Basic weight = physical transmission capacity × proportion of project water storage capacity

[0304] Variable explanation:

[0305] ① Physical transmission capacity: refers to the maximum water transmission capacity of the water transmission route, which is determined based on engineering design parameters (such as pipeline diameter, pump station flow) and historical operating peak data.

[0306] ② Project water storage capacity ratio: represents the ratio of the current project water storage capacity to the total water storage capacity. The calculation formula is:

[0307]

[0308] in, For engineering The water storage capacity, is the total water storage capacity of all projects in the system.

[0309] (2) Calculation of dynamic correction factor:

[0310] The dynamic correction factor is used to adjust the basic weight to reflect the priority of administrative coordination and emergency response. The calculation formula is:

[0311] Dynamic correction factor = administrative coordination coefficient × emergency response priority

[0312] Variable explanation:

[0313] ① Administrative coordination coefficient: It indicates the degree of coordination between administrative regions, with a value range of 0 to 1. The higher the value, the smoother the coordination.

[0314] ② Emergency response priority: Indicates the urgency of the current water diversion task, with a value range of 1 to 10. The higher the value, the higher the priority.

[0315] (3) Calculation of comprehensive weight:

[0316] The comprehensive weight is the product of the basic weight and the dynamic correction factor, which is used to ultimately determine the priority of the water transfer path. The calculation formula is:

[0317] Comprehensive weight = basic weight × dynamic correction factor

[0318] 2. Multi-agent water rights Nash equilibrium rules and optimization algorithms

[0319] (1) Improved Lemming Optimization Algorithm (IALA):

[0320] The Arctic Lemming Algorithm (ALA) is an intelligent optimization algorithm based on the behavior of lemming colonies. It simulates the migration, foraging, and obstacle avoidance of lemmings in polar environments. The ALA algorithm boasts strong global search capabilities and rapid convergence, making it suitable for solving the multi-criteria non-dominated ranking problem of water rights in inter-basin water transfers.

[0321] Based on the original Lemming Optimization Algorithm (ALA), the Improved Arctic Lemming Algorithm (IALA) is proposed by introducing three improved strategies: optical inverse mapping, Gaussian perturbation strategy, and adaptive mutation. Lens imaging reverse learning is a strategy that expands the search range by generating reverse solutions, enhancing global search capabilities and improving solution diversity. Gaussian random walk is a random search strategy based on a Gaussian distribution that performs local search near the current position, enhancing local search capabilities and improving optimization accuracy. Adaptive mutation is a strategy that dynamically adjusts the mutation intensity. Its core concept is to dynamically adjust the mutation probability or mutation mode according to the algorithm's iterative process to balance global and local search capabilities and improve the algorithm's robustness. As a result, the improved IALA algorithm significantly improves the performance and adaptability of dynamic weight adjustment tasks.

[0322] When applied to inter-basin water transfer, the objective function is the same as the multi-criteria non-dominated sorting algorithm for water rights, optimizing the ecological water demand gap, water supply benefit loss, and scheduling costs. Constraints include project water storage capacity constraints and water transfer capacity constraints.

[0323] The core steps of the IALA algorithm are:

[0324] ① Initialize the population: randomly generate a set of initial populations ,in is the population size, It is The location of the lemmings. Each solution represents a possible water resource scheduling scheme. And calculate the fitness value of each individual .

[0325] ② Migration behavior: Simulating the migration process of lemmings, generating new locations for lemmings through random walks and groups, and collaboratively exploring the solution space. Introducing lens imaging reverse learning to generate new solutions.

[0326]

[0327] in, 、 Lemmings In the Second and The position in the iteration. is the optimal solution in the current population. The position of a randomly selected lemming. The coefficient for dynamically adjusting the step size. is a Gaussian distributed random number used to introduce random perturbations.

[0328] ③ Digging behavior (local search): simulates the digging process of lemmings, performs local search near the current position, and optimizes the current solution. The calculation formula is:

[0329]

[0330] in, is the step size coefficient of the local search. is a Gaussian random number for local search, is the step size of the local search.

[0331] ④ Foraging behavior (random exploration): simulates the foraging process of lemmings, performs random searches near the current location, and explores new solution spaces. The calculation formula is:

[0332]

[0333] in, is the step size coefficient of the random search. is a uniformly distributed random number.

[0334] ⑤ Obstacle avoidance behavior (avoiding local optimality): Simulate the obstacle avoidance process of lemmings and avoid falling into the local optimal solution through random perturbations. The calculation formula is:

[0335]

[0336] in: is the step size coefficient and escape coefficient of the obstacle avoidance behavior. is a Gaussian distributed random number for obstacle avoidance behavior, is the step size of the obstacle avoidance behavior.

[0337] ⑥Optical inverse mapping (enhanced global search): for each individual , generating its symmetric solution , and calculate the fitness value corresponding to the symmetric solution , and finally select the solution with better fitness value. The specific calculation formula is:

[0338]

[0339]

[0340] in, and are the lower and upper bounds of the search space.

[0341] ⑦Gaussian perturbation strategy (enhanced local search): for each individual , generate new candidate solutions , and calculate the fitness value corresponding to the new solution , and finally select the solution with better fitness value. The specific calculation formula is:

[0342]

[0343]

[0344] in, The mean is 0 and the variance is Gaussian distributed random numbers. is the step size of Gaussian random walk, and the dynamic adjustment formula is:

[0345]

[0346] in, and are the maximum and minimum step lengths respectively. and are the maximum number of iterations and the current number of iterations respectively.

[0347] ⑧Adaptive mutation (enhanced population diversity): calculate the mutation probability of the current iteration , for each individual Generate its mutation vector . And generate new candidate solutions , calculate the fitness value corresponding to the new solution , and finally select the solution with better fitness value. The specific calculation formula is:

[0348]

[0349]

[0350]

[0351]

[0352] in, and are the minimum and maximum values of the mutation probability, is the fitness value corresponding to each individual.

[0353] ⑨ Population update: Update the population according to the fitness value and retain the best solutions. The mathematical formula is:

[0354]

[0355] ⑩Iterative optimization: Repeat the above process until the termination condition is met (such as reaching the maximum number of iterations or the fitness value converges).

[0356] (2) Arbitration of atomic settlement agreements for water rights transactions based on distributed ledgers

[0357] Cross-district disputes are resolved through predefined arbitration rules automatically executed by the distributed ledger-based atomic settlement protocol for water rights transactions. In this invention, smart contracts are used to resolve disputes over the distribution of rights and interests in cross-district water transfers, ensuring the fairness and efficiency of the water transfer process. The specific arbitration rules are as follows:

[0358] ① Contribution calculation: Calculate the contribution of each administrative district based on the Shapley value method , the specific formula refers to the above formula (17).

[0359] ② Equity distribution: Water transfer equity is automatically distributed based on contribution. The calculation formula is:

[0360]

[0361] in, For administrative districts water transfer rights, is the total water transfer volume.

[0362] ③Automatic execution of arbitration:

[0363] When a cross-district dispute arises, the smart contract automatically calculates each district's contribution based on a pre-set Shapley value model and generates a plan for allocating equity. The smart contract automatically enforces arbitration decisions through an atomic settlement protocol for water rights transactions based on a distributed ledger, ensuring transparency and immutability in dispute resolution.

[0364] S2 Hydrological gap risk situation awareness and multi-source water supply potential coupling analysis

[0365] Based on the dynamic deviation assessment of real-time downstream flow and ecological flow targets, a multi-level hydrological risk situation awareness matrix is triggered, and the water supply potential of multiple sources such as engineering water storage, external water transfer, and recycled water is jointly analyzed to generate a credible and adjustable water allocation decision range to achieve real-time water shortage risk response.

[0366] S2.1 Dynamic assessment and early warning of water shortage in sections based on real-time hydrological data

[0367] 1. Gap calculation: The water shortage of a section is the difference between the real-time discharge flow and the target discharge flow, which is used to assess the current water supply and demand gap. The calculation formula is:

[0368]

[0369] in, 、 、 Respectively The section water shortage, target discharge flow and actual discharge flow in the time period.

[0370] 2. Gap evolution trend surface: Generate a gap evolution trend surface using historical and real-time data to predict future gap changes. This can be done using time series analysis or machine learning models (such as ARIMA and LSTM). The calculation formula is:

[0371]

[0372] in, For the The predicted water shortage for the time period. is the prediction model function. This is historical water shortage data.

[0373] 3. Dynamic triggering algorithm and graded response of the three-level hydrological risk situation awareness matrix: trigger different levels of warnings based on the size of the gap:

[0374] (1) Red Alert:

[0375] (2) Orange alert:

[0376] (3) Yellow Warning:

[0377] in, These are the thresholds for red, orange, and yellow warnings, respectively.

[0378] S2.2 Calculation of system water supply potential under engineering physics constraints

[0379] 1. Project water storage capacity: Project water storage capacity is the amount of water available from water source projects such as reservoirs and dams, taking into account the constraints of effective storage capacity and dead storage capacity. The calculation formula is:

[0380]

[0381] in, is the total project water storage capacity. 、 Respectively The current water storage capacity and dead storage capacity of the project.

[0382] 2. Externally transferred water volume: Externally transferred water volume is the committed water volume dynamically revised according to the inter-basin water transfer agreement. The calculation formula is:

[0383]

[0384] in, is the total external water volume. 、 Respectively The committed water volume and correction factor (between 0 and 1) of each water transfer agreement.

[0385] 3. Recycled water volume: The recycled water volume is the available volume calculated based on the real-time production capacity of the water plant. The calculation formula is:

[0386]

[0387] in, is the total recycled water volume. 、 For the The real-time capacity and utilization rate of each water plant (between 0 and 1).

[0388] S2.3 Generation of collaborative decision intervals for multi-source water supply

[0389] 1. Rigid water demand deduction: Rigid water demand refers to the minimum water demand that must be guaranteed, including domestic water, ecological water, industrial water and agricultural water. The calculation formula is:

[0390]

[0391] in, is the total rigid water demand. For domestic water consumption, 100% guarantee is required. For ecological water use, it is dynamically adjusted according to the threshold of sensitive period (such as ensuring 50% during the breeding period). For industrial water consumption, 70% of the contract value is retained. 50% of the water used is reserved for agriculture.

[0392] 2. Decision on available water allocation: Available water allocation is the total water supply potential minus the total rigid water demand. If the available water allocation is negative, a red alert is triggered and cross-regional water transfer negotiations are initiated.

[0393] S3 Flexible graded water allocation strategy and cross-project replacement agreement

[0394] According to the adjustable water volume, the three modes of emergency supply, balanced distribution and optimized compensation are dynamically switched. Through cascade reduction, time-space decoupling distribution and cross-project replacement agreements, social benefits and water supply security are balanced.

[0395] S3.1 Three-level dynamic allocation mode switching mechanism

[0396] When the available water supply is insufficient, the system enters emergency supply mode, giving priority to domestic water use and reducing other water demands.

[0397] 1. Cascading reduction rules in emergency supply guarantee mode:

[0398] (1) Ecological water replenishment: Reduce to 0%, but retain the minimum ecological base flow.

[0399] (2) Agricultural water use: Reduce to 30% of rigid water demand.

[0400] (3) Industrial water use: Reduce to 50% of rigid water demand.

[0401] 2. Time-space decoupling optimization of balanced distribution mode:

[0402] (1) Allocation based on water demand ratio: The remaining water is allocated based on the user's water demand ratio to generate the path with the least social impact.

[0403] (2) Turn off non-essential landscape water: Automatically turn off non-essential landscape water to reduce water waste. The calculation formula is:

[0404]

[0405] in, 、 For users water allocation and water demand. The remaining adjustable water volume. The water demand of all users.

[0406] S3.2 Balanced Allocation Mode and Spatiotemporal Decoupling Allocation

[0407] When the adjustable water supply is insufficient to completely fill the gap, the system enters balanced distribution mode, giving priority to peak water shortage periods and key areas through dynamic adjustments in time and space dimensions.

[0408] 1. Dynamic allocation of time dimension and time-domain convolution algorithm of hydrological fluctuations:

[0409] (1) Decoupling daily, weekly and monthly fluctuation components: Through time series analysis, decouple the fluctuation components of different time scales and give priority to meeting the peak water shortage period.

[0410] (2) Hydrological Fluctuation Time Domain Convolution Algorithm: Dynamically adjust the water supply ratio in each time period to ensure the balanced distribution of water resources over time. The steps of the hydrological fluctuation time domain convolution algorithm are:

[0411] ①Define the time window: Set the time window size to (such as 7 days, 30 days, etc.), the data in the window is ,in For the Water demand during the period.

[0412] ②Calculate the statistical characteristics of water demand within the window: Calculate the mean of water demand within the window and standard deviation :

[0413]

[0414]

[0415] ③Predict water demand for the next period: Use the data in the window to predict water demand for the next period You can use a simple moving average method or a more complex forecasting model (such as ARIMA, LSTM):

[0416]

[0417] in, It is an adjustment coefficient used to control the fluctuation range of the predicted value.

[0418] ④Dynamically adjust the water supply ratio: according to the predicted water demand and adjustable water volume , calculate the water supply ratio for the next period ,if , then take , indicating that the water supply is sufficient; otherwise, it is distributed proportionally.

[0419] The specific formula is:

[0420]

[0421] ⑤ Sliding window update: Slide the window forward one period, update the data in the window, and repeat the above steps.

[0422] 2. Dynamic balance of spatial dimensions and calculation of apportioned weights:

[0423] (1) Calculation of apportionment weight: The apportionment weight is determined based on the sum of the project water storage capacity and the water inflow.

[0424] (2) Dynamic balancing algorithm: Ensure that all projects are adjusted synchronously to avoid local overload. The calculation formula is:

[0425]

[0426] in, 、 、 Respectively The allocated water volume, water storage volume and water inflow volume of each project.

[0427] S3.3 Optimizing compensation models and cross-project replacement agreements

[0428] When the available water volume is sufficient, the system enters the optimized compensation mode, maximizing social benefits through two-stage allocation and cross-project replacement agreements.

[0429] 1. Two-stage allocation strategy and elastic filling of water demand gap:

[0430] (1) Basic support stage:

[0431] ① Basic water volume guarantee: meet the minimum discharge demand of the section and reserve 10% safety redundancy.

[0432] ② Water volume guaranteed by the project: the remaining adjustable water volume is allocated according to the proportion of water inflow from the project.

[0433]

[0434]

[0435] (2) Water demand gap elastic filling stage:

[0436] ① Flexible agricultural water demand: agriculture will be compensated first, with the maximum compensation ratio being 50%.

[0437] ② Industrial compensation water demand: The industrial compensation amount is the smaller value of 60% of the remaining water volume and the order loss threshold.

[0438]

[0439]

[0440] 2. Cross-project replacement protocol and blockchain verification:

[0441] (1) Asymmetric substitution rule base: Establish substitution rules between external water and local water, and define the substitution ratio between different water sources. For example, 1 unit of external water can replace 1.2 units of local water. The rules in the substitution rule base can be expressed as:

[0442]

[0443] in, 、 are the replacement amounts of external water and local water respectively.

[0444] (2) Verification of the atomic settlement protocol for water rights transactions based on distributed ledgers: The replacement operation must be verified by the atomic settlement protocol for water rights transactions based on distributed ledgers to ensure fairness and transparency. When performing a replacement operation, the smart contract first checks whether the replacement ratio complies with the preset rule base. The specific process is as follows:

[0445]

[0446] The upper limit of the replacement ratio is the maximum value in the preset replacement rule library.

[0447] (3) Automatic replacement:

[0448] ① If the replacement operation passes verification, the smart contract automatically executes the replacement operation and records the replacement result in the water rights transaction atomic settlement protocol based on the distributed ledger, ensuring the transparency and non-tamperability of the operation.

[0449] ②If the swap operation fails verification, the smart contract refuses to execute and generates a corresponding error log for subsequent analysis and adjustment.

[0450] S4 Construction and optimization of "positive-negative" two-way preview decision engine

[0451] "Forward" risk simulation should preview the risk situation and impact, and propose the achievement of water supply and use and river discharge flow targets under different setting schemes such as the current scheduling mode; "reverse" simulation should deduce the safe operation constraints of water resources projects through water sources, engineering conditions and supply and demand relationship constraints, and formulate reasonable and feasible water allocation plans (i.e., multi-objective non-dominant water rights configuration sets under physical constraints) through simulation optimization, supporting three-dimensional visualization scheme comparison and traceability verification.

[0452] S4.1 Forward risk deduction and target quantification

[0453] Forward risk simulation simulates the operating status of the water resources system in the future scheduling period, evaluates the achievement of water supply and river discharge flow targets, and generates risk heat maps and water demand gap entropy distribution maps.

[0454] 1. Risk situation simulation and project water storage capacity assessment:

[0455] (1) Simulate the operating status of the water resources system during the future scheduling period: Through the hydrological model and scheduling model, simulate key indicators such as reservoir water storage, river flow, and user water demand during the future scheduling period.

[0456] (2) Generate a thermal map of project water storage exceeding the limit: Based on the simulation results, generate a thermal map of project water storage exceeding the limit, marking the areas where the water storage exceeds the designed storage capacity. The calculation formula is:

[0457]

[0458] in, For the Time period position The water storage capacity exceeds the limit (1 means exceeded, 0 means not exceeded). For the Time period position of water storage capacity. For location Design storage capacity.

[0459] (3) Calculate the spatial distribution map of user water demand gap entropy values: Based on the simulation results, draw a spatial distribution map of user water demand gap entropy values and mark high-risk areas. The calculation formula is:

[0460]

[0461] in, 、 、 Respectively Time period position User water shortage rate, user water demand, and actual water supply.

[0462] 2. Entropy evaluation of user water demand gap and quantification of scheduling targets:

[0463] The achievement of water supply and discharge targets under different scenarios, including the current dispatch mode, is proposed. By comparing the simulation results of different dispatch scenarios, the achievement of water supply and discharge targets is quantified. The calculation formula is:

[0464]

[0465] in, 、 They are the actual and target water supply (or discharge flow) respectively.

[0466] S4.2 Reverse Scheduling Preview and Water Rights Multi-Criteria Non-Dominated Sorting Algorithm

[0467] The reverse scheduling preview generates a non-dominated water rights configuration set under physical constraints through constraint inversion and multi-criteria non-dominated sorting of water rights, and formulates a reasonable and feasible reverse scheduling plan for water resources projects.

[0468] 1. Constraint inversion and physical constraint mapping

[0469] (1) Dynamically adjust the minimum operating water level, maximum water lifting capacity and other boundary conditions of water resources projects: Dynamically adjust the minimum operating water level of the project based on real-time data and forecast data and maximum water lifting capacity The calculation formula is:

[0470]

[0471]

[0472] in, For the The minimum operating water level for the time period. For the The maximum water extraction capacity during a period. and Dynamic adjustment function.

[0473] (2) Use physical constraint projection technology to map the physical constraint set to the feasible solution space: Through projection technology, physical constraints (such as minimum operating water level and maximum water lifting capacity) are mapped to the feasible solution space to ensure the feasibility of the scheduling plan. The calculation formula is:

[0474]

[0475] in, 、 are the feasible solution and the original solution respectively. is a set of physical constraints.

[0476] 2. Water rights multi-criteria non-dominated sorting algorithm and non-dominated water rights configuration set generation:

[0477] (1) Through the water rights multi-criteria non-dominated sorting algorithm, the three objectives of water supply stability, discharge compliance rate, and scheduling cost are optimized simultaneously:

[0478] ① Water supply stability :Minimize water supply fluctuations and maintain a long dispatch period Below, the calculation formula is:

[0479]

[0480] in, 、 They are Actual and target water supply for the time period.

[0481] ②Discharge compliance rate :Maximize the rate of compliance of downstream flow rate:

[0482]

[0483] in, for The discharge flow rate during the period.

[0484] ③ Scheduling costs : Minimize scheduling cost:

[0485]

[0486] in, 、 They are Physical constraints and total flow rate for the time period.

[0487] (2) Use the improved multi-objective dream optimization algorithm (IMODOA) to output the non-dominated water rights configuration set:

[0488] The Dream Optimization Algorithm (DOA) is a novel metaheuristic algorithm inspired by human dreams. The algorithm combines a basic memory strategy with a forgetting and replenishing strategy to balance exploration and exploitation.

[0489] Based on DOA, this paper integrates improved strategies such as dynamic forgetting dimension adjustment, adaptive weight adjustment, hybrid search strategy, and constraint processing mechanism to propose an improved multi-objective dream optimization algorithm (IMODOA). It can efficiently generate non-dominated water rights configuration sets, ensure the feasibility and efficiency of the scheduling scheme, and is suitable for complex optimization problems such as water resource scheduling. The core steps are:

[0490] ① Initialization phase: randomly generate the initial population , each individual represents a possible water resource scheduling solution:

[0491]

[0492] in: For the The location of an individual. and are the lower and upper bounds of the search space respectively. is a random number in the range [0,1]. Indicates the population size.

[0493] ② Exploration phase (iteration number from 0 to ):

[0494] Dynamic forgetting dimension adjustment: the number of forgetting dimensions Dynamically adjusted with the number of iterations:

[0495]

[0496] in, It is The number of forgotten dimensions in iterations. is the number of initial forgotten dimensions. is the maximum number of iterations.

[0497] Memory strategy: Individuals remember the optimal solution of their group:

[0498]

[0499] in, For the In the iteration group The optimal solution is The individual location.

[0500] Forget and replenish strategy: The individual randomly updates its position on the forgetting dimension.

[0501]

[0502] in, For the Individuals in the iteration In dimension location. For the In the iteration group The optimal solution is in the dimension location. and Dimensions lower and upper limits. is the current iteration number. is the maximum number of iterations in the exploration phase. Respectively indicate the first, second, ... At the iteration, the dimension index selected in the forget dimension is

[0503] Dream sharing strategy: Individuals randomly obtain the location information of other individuals in the forgotten dimension:

[0504]

[0505] in, 、 Respectively sequence Individuals in the iteration In dimension location. For range A natural number randomly selected from the , indicating the individual index.

[0506] ③ Development phase (iteration number from arrive ):

[0507] Dynamic forgetting dimension adjustment: the number of forgetting dimensions Dynamically adjusted with the number of iterations:

[0508]

[0509] in, It is The number of forgotten dimensions in iterations. is the number of initial forgotten dimensions.

[0510] Memory strategy: Individuals remember the optimal solution of the entire population:

[0511]

[0512] in, For the The optimal solution of the entire population in the iteration becomes the The individual location.

[0513] Forgetting and replenishing strategy: The individual randomly updates the position on the forgetting dimension:

[0514]

[0515] in, For the The optimal solution of the entire population in the iteration is in dimension location.

[0516] ④ Adaptive weight adjustment: Dynamically adjust the weight of the objective function according to the convergence of the objective function:

[0517]

[0518] in, For the The objective function in the iteration The weight of . For the The optimal solution in the iteration The objective function value of .

[0519] ⑤Constraint processing mechanism:

[0520] Constraint correction: Correct solutions that violate constraints:

[0521]

[0522] Constraint Penalty: Penalize solutions that violate constraints:

[0523]

[0524] in, is the penalty coefficient. For population The optimal solution of . For population The penalty function when the optimal solution is reached.

[0525] ⑥ Generation of non-dominated water rights configuration set: Select the best solution based on the fitness value and generate the non-dominated water rights configuration set:

[0526]

[0527] in, is the non-dominated water rights configuration set.

[0528] Example 2

[0529] The present invention also provides a cross-basin water resource system coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making system, which is used to implement the method of embodiment 1. The system includes: a graph construction module, a perception and analysis module, a generation module, and a strategy optimization module;

[0530] A graph construction module is used to construct a multimodal water infrastructure topology relationship graph and perform bidirectional transmission optimization;

[0531] A perception and analysis module, configured to perceive the hydrological gap risk situation and analyze the coupling of multi-source water supply potential based on the multi-modal water infrastructure topological relationship map and optimized bidirectional transmission;

[0532] A generation module is used to generate flexible graded water allocation strategies and cross-project replacement agreements based on the coupling of hydrological gap risk situation and multi-source water supply potential;

[0533] The strategy optimization module is used to build and optimize the "positive-negative" two-way pre-rehearsal decision engine based on the flexible graded water allocation strategy and cross-project replacement protocol.

[0534] The embodiments described above are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by persons skilled in the art should fall within the scope of protection defined by the claims of the present invention.

Claims

1. A cross-basin water resources system coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making method, characterized by: The method comprises: S1: Construct a multimodal water infrastructure topology map and perform bidirectional transmission optimization; S2: Based on the multimodal water infrastructure topology relationship map and optimized bidirectional transmission, the risk situation of hydrological gap is perceived and coupled with the analysis of multi-source water supply potential; S3: Generate flexible graded water allocation strategies and cross-project replacement agreements based on the coupling of hydrological deficit risk situation and multi-source water supply potential; S4: Build and optimize a "positive-negative" two-way pre-rehearsal decision engine based on a flexible hierarchical water allocation strategy and cross-project replacement protocols; In S4, based on the flexible graded water allocation strategy and cross-project replacement protocol, a "positive-negative" two-way pre-play decision engine is constructed and optimized, including: S4.1: Forward risk deduction and target quantification; S4.2: Reverse scheduling preview and multi-criteria non-dominated ranking of water rights; Among them, the forward risk deduction and target quantification in S4.1 include: Risk situation simulation and project water storage capacity assessment; Entropy evaluation of user water demand gap and quantification of scheduling targets; In S4.2, the reverse scheduling preview and water rights multi-criteria non-dominated sorting include; Constraint inversion and physical constraint mapping; Multi-criteria non-dominated sorting algorithm for water rights and generation of non-dominated water rights configuration sets; Risk situation simulation and project water storage capacity assessment, including: Through hydrological models and scheduling models, key indicators such as reservoir water storage, river flow, and user water demand in the future scheduling period are simulated; Based on the simulation results, a thermal map of the project's water storage exceeding the limit is generated, marking the areas where the water storage exceeds the designed storage capacity. The calculation formula is: ； in, For the Time period position The water storage capacity exceeds the limit. For the Time period position The water storage capacity, For location Design storage capacity; According to the simulation results, the spatial distribution map of the user water demand gap entropy value is drawn, and the high-risk areas are marked. The calculation formula is: ； in, 、 、 Respectively Time period position User water shortage rate, user water demand, and actual water supply; User water demand gap entropy assessment and scheduling target quantification, including: ； in, 、 are the actual and target water supply or discharge flow respectively; Constraint inversion and physical constraint mapping, including: Dynamically adjust the minimum operating water level of the project based on real-time data and forecast data and maximum water lifting capacity , the calculation formula is: ； ； in, For the The minimum operating water level of the time period, For the The maximum water extraction capacity during the period, and For dynamic adjustment function; Through projection technology, the physical constraint, namely the minimum operating water level, and maximum water lifting capacity Mapped to the feasible solution space, the calculation formula is: ； in, 、 are feasible solutions and primal solutions respectively, is the set of physical constraints; Water rights multi-criteria non-dominated sorting algorithm and non-dominated water rights configuration set generation, including: Through the water rights multi-criteria non-dominated sorting algorithm, the three objectives of water supply stability, discharge compliance rate and scheduling cost are optimized simultaneously; among them, water supply stability :Minimize water supply fluctuations and maintain a long dispatch period The calculation formula is: ； in, 、 They are Actual and target water supply for the time period; Discharge compliance rate :Maximize the rate of compliance of downstream flow rate: ； in, for Discharge flow rate during the period; Scheduling costs : Minimize scheduling cost: ； in, 、 They are Physical constraints and total flow during the time period; An improved multi-objective dream optimization algorithm, IMODOA, is used to output the non-dominated water rights configuration set. Based on the dream optimization algorithm DOA, dynamic forgetting dimension adjustment, adaptive weight adjustment, hybrid search strategy, and constraint processing mechanism improvement strategy are integrated to obtain the improved multi-objective dream optimization algorithm, IMODOA: In the exploration phase, the number of iterations ranges from 0 to : Dynamic forgetting dimension adjustment: the number of forgetting dimensions Dynamically adjusted with the number of iterations: ； in, It is The number of forgotten dimensions in the iteration, is the number of initial forgotten dimensions, is the maximum number of iterations; Memory strategy: Individuals remember the optimal solution of their group: ； in, For the In the iteration group The optimal solution is The individual location; Forgetting and replenishing strategy: The individual randomly updates the position on the forgetting dimension: ； in, For the Individuals in the iteration In dimension location, For the In the iteration group The optimal solution is in the dimension location, and Dimensions The lower and upper limits of is the current iteration number, is the maximum number of iterations in the exploration phase, Respectively indicate the first, second, ... At the iteration, the selected dimension index in the forget dimension is forgotten; Dream sharing strategy: Individuals randomly obtain the location information of other individuals in the forgotten dimension: ； in, 、 Respectively sequence Individuals in the iteration In dimension location; For range A natural number randomly selected from the source represents the individual index; During the development phase, the number of iterations starts from arrive : Dynamic forgetting dimension adjustment: the number of forgotten dimensions Dynamically adjusted with the number of iterations: ； in, It is The number of forgotten dimensions in iterations, is the number of initial forgotten dimensions; Memory strategy: Individuals remember the optimal solution of the entire population: ； in, For the The optimal solution of the entire population in the iteration becomes the The individual location; Forgetting and replenishing strategy: The individual randomly updates the position on the forgetting dimension: ； in, For the The optimal solution of the entire population in the iteration is in dimension location; Adaptive weight adjustment: Dynamically adjust the weight of the objective function according to the convergence of the objective function: ； in, For the The objective function in the iteration The weight of For the The optimal solution in the iteration The objective function value of Constraint handling mechanism: Constraint correction: Correct solutions that violate constraints: ； Constraint Penalty: Penalize solutions that violate constraints: ； in, is the penalty coefficient, For population The optimal solution of For population The penalty function when the optimal solution is reached.

2. The method according to claim 1, characterized in that In S1, a multimodal water conservancy infrastructure topology relationship map is constructed and bidirectional transmission optimization is performed, including: S1.1: Multi-dimensional correlation tensor modeling and adaptive weight evolution of water conservancy entities; S1.2: Dynamic weight adjustment enabled by smart contracts and the construction of a multi-agent water rights Nash equilibrium mechanism.

3. The method according to claim 2, characterized in that In S1.1, the multi-dimensional correlation tensor modeling and adaptive weight evolution of water conservancy entities include: Construction of source-sink relationship matrix and bidirectional water transfer rules; Construction of supply and demand relationship matrix and intelligent adjustment of hierarchical water supply; Regression relationship matrix construction and intelligent dynamic adjustment of recycled water replenishment; Water resource partition matrix construction and federated learning collaborative scheduling; Construction of administrative division matrix and cross-regional coordination mechanism; The source-sink relationship matrix is constructed, including: Determine the maximum water delivery capacity based on engineering design parameters and historical operating peak data; Based on the maximum water delivery capacity, the source-sink relationship matrix is constructed by real-time monitoring of equipment health and combining machine learning algorithms to predict availability trends. Construction of two-way water transfer rules, including: Generate reverse water transfer paths based on the reverse water transfer capacity of the pumping station and the bidirectional overflow threshold of the pipeline; Based on the reverse water transfer path, the weight of reverse water transfer is dynamically adjusted according to the water storage capacity of the water source project, equipment health, and emergency response priority parameters to establish a two-way water transfer rule; Intelligent adjustment of graded water supply, including: ； in, is the water supply ratio of user i at time t, is the initial water supply ratio, is the current water storage capacity, To warn of water levels; Intelligent dynamic adjustment of recycled water replenishment, including: in, is the recycled water replenishment ratio, is the maximum redemption ratio, is the amount of recycled water, is the water demand; Water resource partition matrix construction and federated learning collaborative scheduling, including: in, is the model parameter, which represents the weight of the water resources scheduling model. Represents a partition The loss function is used to measure the error between the model prediction value and the actual value. For partition Local model parameters of ; is the optimal model parameter, which means that under the partition data, the model can minimize the parameter value of the scheduling error; is the global model parameter, which represents the model parameter after collaborative training of all partitions. Represents a partition The amount of data, that is, the scale of water resources data, is the total data volume, which represents the sum of the data volumes of all partitions. and Represents partitions respectively In the Second and The model parameters in the iteration, is the learning rate, which controls the step size of model parameter update. is the loss function The gradient of , which indicates the direction of change of the model parameters; The construction of administrative division matrix and cross-regional coordination mechanism includes; ； in, For administrative districts The contribution of is the set of all administrative districts, is a subset, For subset benefits.

4. The method according to claim 3, characterized in that S1.2: Smart contract-enabled dynamic weight adjustment and multi-agent water rights Nash equilibrium mechanism construction, including: Weight calculation and intelligent dynamic correction model construction; Based on the constructed comprehensive weights, an improved multi-agent water rights Nash equilibrium rule is constructed; Among them, the basic weight calculation includes: Basic weight = physical transmission capacity × proportion of project water storage capacity; Among them, physical transmission capacity refers to the maximum water transmission capacity of the water transmission path, and the proportion of project water storage capacity refers to the proportion of the current project water storage capacity to the total water storage capacity; Intelligent dynamic correction model construction, including: Dynamic correction factor = administrative coordination coefficient × emergency response priority; Among them, the administrative coordination coefficient indicates the degree of coordination between administrative regions; Comprehensive weight = basic weight × dynamic correction factor; Constructing an improved multi-agent Nash equilibrium rule for water rights, including: constructing an improved Lemmings optimization algorithm (ALA); and arbitrating an atomic settlement protocol for water rights transactions based on the improved Lemmings optimization algorithm (ALA) and distributed ledgers. Among them, the construction of the improved Lemming Optimization Algorithm ALA includes: based on the Lemming Optimization Algorithm ALA, introducing three improvement strategies: optical inverse mapping, Gaussian perturbation strategy and adaptive mutation; Among them, optical inverse mapping: for each individual , generating a symmetric solution , and calculate the fitness value corresponding to the symmetric solution , and finally select the solution whose fitness value meets the preset requirements. The calculation formula is: ； ； in, and are the lower and upper bounds of the search space, Is the fitness value corresponding to each individual; Gaussian perturbation strategy: for each individual , generate new candidate solutions , and calculate the fitness value corresponding to the new solution , and finally select the solution whose fitness value meets the preset requirements. The calculation formula is: ； ； in, The mean is 0 and the variance is Gaussian distributed random numbers, is the step size of Gaussian random walk, and the dynamic adjustment formula is: ； in, and are the maximum and minimum values of the step length, and are the maximum number of iterations and the current number of iterations respectively; Adaptive mutation: Calculate the mutation probability of the current iteration , for each individual Generate its mutation vector , and generate new candidate solutions , calculate the fitness value corresponding to the new solution , and finally select the solution whose fitness value meets the preset requirements. The calculation formula is: ； ； ； ； in, and are the minimum and maximum values of the mutation probability respectively; Among them, the atomic settlement agreement arbitration of water rights transactions based on distributed ledgers includes: Calculate the contribution of each administrative district based on the Shapley value method ; According to contribution Automatically allocate water transfer rights, the calculation formula is: ； in, For administrative districts water transfer rights, is the total water transfer volume; When a cross-administrative district dispute occurs, the smart contract automatically calculates the contribution of each administrative district based on the preset Shapley value method and generates a rights and interests distribution plan.

5. The method according to claim 1, wherein In S2, based on the multimodal water infrastructure topology relationship map and optimized bidirectional transmission, the perception of the hydrological gap risk situation and the analysis of the multi-source water supply potential are coupled, including: S2.1: Dynamic assessment and early warning of water shortage in sections based on real-time hydrological data; S2.2: Calculate the system water supply potential under engineering physics constraints. The system water supply potential includes: project water storage, external water transfer, and recycled water. S2.3: Generate a multi-source water supply collaborative decision-making interval; wherein, the multi-source water supply collaborative decision-making interval is: adjustable water volume = total water supply potential - total rigid water demand.

6. The method according to claim 1, characterized in that In S3, based on the coupling of hydrological deficit risk situation and multi-source water supply potential, a flexible hierarchical water allocation strategy and cross-project replacement agreement are generated, including: S3.1: Construct a three-level dynamic allocation mode switching mechanism; S3.2: Based on the three-level dynamic allocation mode switching mechanism, the balanced allocation mode and the time-space decoupling allocation; S3.3: Optimize compensation patterns and cross-project replacement protocols based on spatiotemporal decoupling of allocations.

7. The method according to claim 5, characterized in that In S3.2, based on the three-level dynamic allocation mode switching mechanism, the balanced allocation mode and the time-space decoupling allocation include: Through time series analysis, the fluctuation components of different time scales are decoupled to give priority to meeting the peak water shortage period; Dynamically adjust the water supply ratio in each period to achieve balanced distribution of water resources over time; Based on the sum of the project's water storage capacity and incoming water, calculate the dynamic balance and allocation weight of water resources in spatial dimensions; Among them, the water supply ratio of each time period is dynamically adjusted to achieve a balanced distribution of water resources over time, including: Assume the time window size is , the data in the window is ,in For the Water demand during the time period; Calculate the average water demand within the window and standard deviation : ； ； Use the data within the window to predict water demand for the next period , the calculation formula is: ； in, is the adjustment coefficient, which is used to control the fluctuation range of the predicted value; Based on predicted water demand and adjustable water volume , calculate the water supply ratio for the next period ,if , then take , indicating that the water supply is sufficient; otherwise, it is distributed proportionally, and the calculation formula is: ； Based on the sum of the project's water storage capacity and incoming water, the dynamic balance and allocation weights of water resources in spatial dimensions are calculated, including: ； in, 、 、 Respectively The allocated water volume, water storage volume and water inflow volume of each project; In S3.3, based on the temporal and spatial decoupling of allocation, the compensation model and cross-project replacement protocol are optimized, including: Two-stage allocation strategy and elastic filling of water demand gap; Cross-project replacement protocol and blockchain verification; Among them, the two-stage allocation strategy includes: ； ； Flexible filling of water gaps, including: ； ； Cross-project replacement protocol and blockchain verification, including: Establish the substitution rules between external water and local water to obtain the substitution ratio between different water sources; the substitution rules are: ； in, 、 are the replacement amounts of externally transferred water and local water, respectively; Verification of atomic settlement protocol for water rights transactions based on distributed ledgers: ； The upper limit of the replacement ratio is the maximum value in the preset replacement rule library.

8. A cross-basin water resources system coordinated scheduling and positive-negative bidirectional coupling dynamic decision-making system, the system being used to implement the method according to any one of claims 1 to 7, characterized in that: The system includes: a graph construction module, a perception and analysis module, a generation module, and a strategy optimization module; The graph construction module is used to construct a multi-modal water conservancy infrastructure topology relationship graph and perform bidirectional transmission optimization; The perception and analysis module is used to perceive the hydrological gap risk situation and analyze the coupling of multi-source water supply potential based on the multi-modal water conservancy infrastructure topological relationship map and optimized bidirectional transmission; The generation module is used to generate a flexible hierarchical water allocation strategy and a cross-project replacement agreement based on the coupling of the hydrological gap risk situation and the multi-source water supply potential; The strategy optimization module is used to build and optimize a "positive-negative" two-way pre-rehearsal decision engine based on a flexible graded water allocation strategy and a cross-project replacement protocol.

Images