Scenario recognition-based dynamic weight scheduling method and system for basin water resources

Abstract

This invention discloses a dynamic weighted scheduling method and system for watershed water resources based on scenario recognition. By introducing coupled oscillation mode analysis of multi-source hydrological and meteorological elements, it achieves dynamic and accurate identification of watershed scenarios, effectively overcoming the limitations of traditional methods that rely on fixed rules or historical experience. The construction of a multi-dimensional constraint system ensures the organic combination of short-term scheduling objectives and long-term planning needs. By analyzing the changes in the feasible solution space and setting critical thresholds, the system can identify scheduling conflict states in a timely manner, thereby avoiding the potential impact of short-term behavior on long-term strategy in resource allocation decisions. This makes water resource scheduling no longer limited to immediate optimization, but also takes into account the temporal coordination and sustainability of the watershed system.

Description

Technical Field

[0001] This invention relates to the field of watershed water resources scheduling technology, and is particularly applicable to a dynamic weighted scheduling method and system for watershed water resources based on scenario recognition. Background Technology

[0002] Watershed water resource allocation is a core means of ensuring regional water security and achieving efficient water resource utilization. Traditional allocation methods often rely on fixed rules or historical experience, making it difficult to adapt to dynamic changes in watershed conditions. With the popularization of Internet of Things (IoT) technology, real-time monitoring systems based on sensor networks have been widely used in watershed management. These systems continuously collect hydrological and meteorological data through various sensors deployed in rivers, reservoirs, and meteorological stations, providing a data foundation for allocation decisions. Building on this, existing technologies are beginning to explore the introduction of dynamic weighting mechanisms. This involves assigning variable priorities to different water use objectives, such as domestic, industrial, agricultural, and ecological water use, based on real-time watershed conditions, aiming to achieve more refined resource allocation.

[0003] However, existing dynamic weight scheduling methods based on sensor networks have shortcomings in addressing multi-timescale decision-making needs. These methods typically rely on current or recent sensor data for scenario assessment and weight calculation, focusing their optimization on short-term resource allocation efficiency. This often leads to scheduling schemes conflicting with long-term water resource strategic planning goals based on monthly or quarterly cycles. For example, in short-term high-water scenarios, the system may tend to increase the weight of ecological or agricultural water supply, but this may excessively deplete water storage, thereby compromising the water supply security of urban and rural areas during subsequent dry seasons. Existing technologies lack consideration for the coordination mechanism between short-term scheduling and long-term planning, which means that while dynamic weight adjustments may improve the accuracy of scheduling in local time periods, they may also jeopardize the overall safety and sustainability of the watershed water resource system. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic weighted scheduling method and system for watershed water resources based on scenario recognition, which addresses the problem of the lack of a coordination mechanism between short-term scheduling and long-term planning, thereby endangering the overall safety and sustainability of the watershed water resources system.

[0005] To achieve the above objectives, the dynamic weighted scheduling method for watershed water resources based on scenario recognition described in this invention includes the following steps: S1. Acquire real-time hydrological and meteorological data of the watershed through sensor networks; S2. Based on real-time hydrological and meteorological data, analyze the quantitative characteristics of multi-source hydrological and meteorological elements and the stability of the coupled oscillation modes between elements to determine the current watershed situation. S3. Based on the current watershed situation, establish a multi-dimensional constraint system that includes hydrological conditions, engineering capabilities, ecological needs, and socio-economic goals; S4. Based on a multi-dimensional constraint system, analyze the changes in the feasible solution space that simultaneously satisfies short-term scheduling constraints and long-term planning constraints. When the feasible solution space is lower than the critical threshold, it is identified as a conflict state. S5. Simulate the operational status of the watershed water resources system under various predefined future adverse scenario evolution paths based on conflict state, and evaluate the resource supply vulnerability index of the system under each evolution path. S6. Adjust dynamic weights based on conflict status and resource supply vulnerability indicators to balance short-term and long-term objectives.

[0006] Furthermore, in step S1, real-time hydrological and meteorological data are collected through sensors deployed in rivers, reservoirs, and meteorological stations, including river water level data, river flow data, reservoir water storage data, rainfall data, temperature data, and humidity data.

[0007] Furthermore, step S2 specifically includes extracting time series data of multiple hydrological and meteorological elements from real-time hydrological and meteorological data; calculating the mean value of the time series data as a quantitative indicator; calculating the cross-correlation function between the time series data, using the absolute value of the correlation coefficient as a coupling strength indicator; calculating the variance of the time series data, using the reciprocal of the variance or the normalized variance value as an oscillation stability indicator; and comprehensively judging the current watershed situation based on the quantitative indicator, coupling strength indicator, and oscillation stability indicator.

[0008] Furthermore, the coupling strength index is compared with a first preset threshold, the oscillation stability index is compared with a second preset threshold, and the magnitude index is compared with a third preset threshold to determine the hydrological intensity. Based on the comparison results, predefined decision rules are applied to determine whether the current watershed scenario is a short-term high-water scenario, a normal-water scenario, or a low-water scenario, and the stable or unstable state under the scenario is identified.

[0009] Furthermore, the multi-dimensional constraint system in step S3 includes hydrological constraints based on river water level data and preset flood control limit water levels; engineering capacity constraints based on reservoir water storage data and reservoir design capacity; ecological demand constraints based on the minimum ecological flow determined by historical ecological monitoring data; and socio-economic objective constraints based on the calculation of basic water supply demand based on regional population distribution and industrial layout data.

[0010] Furthermore, step S4 specifically includes constructing a linear programming model based on a multi-dimensional constraint system; determining the range of solution sets that satisfy all constraints by solving the linear programming model; calculating the volume of the solution set range as a feasible solution space volume index, and comparing the feasible solution space volume index with a preset critical threshold; identifying a conflict state when the feasible solution space volume index is lower than the preset critical threshold.

[0011] Furthermore, step S4 specifically involves constructing a linear programming model with water resource allocation efficiency as the objective and a multi-dimensional constraint system as the constraint, and using a linear programming solution algorithm to calculate the range of values ​​for the decision variables as the range of the solution set.

[0012] Furthermore, step S5 includes selecting a predefined future adverse scenario evolution path based on the conflict state; simulating the dynamic changes in the supply and demand of the watershed water resources system under each evolution path through the water balance equation; and calculating the gap ratio between water resource supply and demand under each evolution path as an indicator of resource supply vulnerability.

[0013] Furthermore, step S6 includes mapping the conflict state to a conflict severity level and mapping the resource supply vulnerability index to a vulnerability level; querying a predefined weight adjustment table based on the combination of the conflict severity level and the vulnerability level to obtain the adjustment values ​​of the short-term target weight and the long-term target weight, and applying the adjustment values ​​to update the dynamic weights.

[0014] The present invention discloses a dynamic weighted scheduling system for watershed water resources based on scenario recognition, and a dynamic weighted scheduling method for watershed water resources based on scenario recognition, comprising the following modules: The data acquisition module is used to acquire real-time hydrological and meteorological data of the watershed through a sensor network; The scenario assessment module is used to analyze the stability of coupled oscillation modes among multiple hydrological and meteorological elements based on real-time hydrological and meteorological data in order to determine the current watershed scenario. The system establishment module is used to establish a multi-dimensional constraint system that includes hydrological conditions, engineering capabilities, ecological needs, and socio-economic goals when the current watershed scenario is determined to be a short-term high-water scenario. The state recognition module is used to analyze the changes in the feasible solution space that simultaneously satisfies short-term scheduling constraints and long-term planning constraints based on a multi-dimensional constraint system. When the feasible solution space is lower than the critical threshold, it is identified as a conflict state. The indicator evaluation module is used to simulate the operation of the watershed water resources system under various predefined future adverse scenario evolution paths based on conflict state, and to evaluate the system's resource supply vulnerability indicators under each evolution path. The weight adjustment module is used to adjust dynamic weights based on conflict status and resource supply vulnerability indicators to balance short-term and long-term objectives.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By introducing coupled oscillation mode analysis of multi-source hydrological and meteorological elements, dynamic and accurate identification of watershed scenarios is achieved, effectively overcoming the limitations of traditional methods that rely on fixed rules or historical experience. The construction of a multi-dimensional constraint system ensures the organic combination of short-term scheduling objectives and long-term planning needs. By analyzing the changes in the feasible solution space and setting critical thresholds, the system can identify scheduling conflict states in a timely manner, thereby avoiding the potential impact of short-term behavior on long-term strategy in resource allocation decisions. This makes water resource scheduling no longer limited to immediate optimization, but also takes into account the temporal coordination and sustainability of the watershed system.

[0016] 2. By simulating the system's operational state under predefined adverse future scenarios and quantifying resource supply vulnerability indicators, a scientific basis for dynamic weight adjustment is provided. By combining conflict states with vulnerability assessment, the weight allocation can respond to real-time scenario changes and proactively adapt to risks and challenges under different evolution paths. This two-way adjustment mechanism significantly improves the adaptability and robustness of the water resource scheduling system, ensuring an effective balance between short-term efficiency and long-term security in complex and ever-changing environments, and ultimately achieving overall optimization of watershed water resource management. Attached Figure Description

[0017] Figure 1 This is a flowchart of a dynamic weighted scheduling method for watershed water resources based on scenario recognition, according to the present invention.

[0018] Figure 2 This is a schematic diagram of the structure of a dynamic weighted scheduling system for water resources in a basin based on scenario recognition, according to the present invention. Detailed Implementation

[0019] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] Example 1 The present invention provides a dynamic weighted scheduling method for watershed water resources based on scenario recognition, such as... Figure 1 As shown, it includes the following steps: S1. Acquire real-time hydrological and meteorological data of the watershed through a sensor network; specifically implemented as follows: Real-time hydrological and meteorological data are collected through a sensor network deployed in rivers, reservoirs, and meteorological stations. Sensor types include water level sensors, flow sensors, rainfall sensors, temperature sensors, and humidity sensors. Installation locations cover key areas of the watershed; for example, in rivers, the confluence of main streams and tributaries is selected; in reservoirs, the area in front of the dam and in the center of the reservoir area is covered; and meteorological stations are distributed in mountainous, plain, and valley areas. Data acquisition frequency is set according to monitoring requirements, and the sensors transmit data to a central processing platform in real time via wireless communication technologies such as 4G or LoRa.

[0021] The collected data mainly includes river water level data, river flow data, reservoir storage data, rainfall data, temperature data, and humidity data, as well as engineering design parameters and historical statistical data retrieved from the database. The engineering design parameters include preset flood control limit water level, reservoir capacity curve, and minimum ecological flow standard. The historical statistical data is used to calculate quantile thresholds in subsequent steps.

[0022] River water level data is acquired through pressure or ultrasonic water level sensors. The sensors are installed on fixed supports on both banks of the river to measure the vertical distance between the water surface and the reference plane. Wave interference is taken into account during the measurement, and a stable value is output by taking the average of multiple measurements, such as calculating the arithmetic mean through 10 samples.

[0023] River flow data is acquired using Doppler current meters or weir / flute systems, with sensors installed on the riverbed or sidewalls. Instantaneous flow is calculated based on the velocity-area method. Velocity measurements are combined with cross-sectional geometry data, and total flow is calculated through integration, validated using methods such as Manning's formula or measured calibration curves.

[0024] Reservoir water storage data is obtained through water level sensors and a reservoir capacity curve model. The water level sensors measure the reservoir water level. The reservoir capacity curve is constructed using historical topographic surveys and design data, for example, describing the relationship between water level and reservoir capacity using a polynomial function, and calculating the current reservoir capacity through linear interpolation.

[0025] Rainfall data is acquired using tipping bucket or radar-type rain gauges, which are installed in open areas 1.5 meters above the ground. The sensors record the depth of precipitation per unit time. The data is smoothed using a moving average method during data acquisition, and outliers such as bird interference are removed.

[0026] Temperature data is acquired using platinum resistance or thermocouple temperature sensors, which are placed inside a Stevenson screen to avoid direct sunlight. The air temperature is measured and calibrated to an accuracy within ±0.5 degrees Celsius.

[0027] Humidity data is acquired through capacitive or resistive humidity sensors installed in ventilated environments to measure relative humidity. The sensors are periodically calibrated using a standard saturated salt solution, with accuracy controlled within ±3%.

[0028] Data transmission adopts a multi-protocol compatible approach. Data packets are encapsulated with timestamps and location information and sent to the server via an encryption protocol. The server performs format standardization processing, and the data is stored in a distributed database, supporting real-time querying and historical backtracking.

[0029] All data collection and processing methods follow the general technical specifications for hydrological and meteorological monitoring to ensure consistency and comparability. The data is ultimately used for subsequent scenario assessment and scheduling decisions, providing basic support for watershed water resources management.

[0030] S2. Based on real-time hydrological and meteorological data, analyze the stability of coupled oscillation modes among multi-source hydrological and meteorological elements to determine the current watershed situation. Specific implementation details are as follows: First, time series data of multiple hydrological and meteorological elements are extracted from real-time hydrological and meteorological data. For example, river water level data, river flow data, rainfall data, temperature data, and humidity data are selected as key elements. The time series data are organized at fixed time intervals, such as using 1 hour or 1 day as the time step. The data length covers recent periods, such as 30 days. Missing values ​​are handled during the extraction process, and the continuity of the time series is ensured by linear interpolation or fill-in methods. The time series data is stored in arrays or lists, with each element corresponding to an independent sequence for subsequent calculations.

[0031] The cross-correlation function between time series data is calculated to obtain the coupling strength index. The cross-correlation function measures the correlation between two time series at different time lags, such as calculating the Pearson correlation coefficient between water level data and rainfall data at zero lag. The coupling strength index is defined as the absolute value of the correlation coefficient, ranging from 0 to 1, where 0 indicates no coupling and 1 indicates strong coupling. The stationarity of the time series is considered during the calculation, and differencing or detrending processing is performed when necessary, such as eliminating linear trends through first-order differencing, to ensure that the calculation results reflect the true coupling relationship. The coupling strength index is output in numerical form for subsequent stability assessment.

[0032] The variance of time series data is calculated to obtain an oscillation stability index. Variance measures the degree of fluctuation in a time series; for example, calculating the variance of water level data within a specified time window. The oscillation stability index is defined as the reciprocal of the variance or normalized by the maximum variance, making the index between 0 and 1, where 0 indicates high instability and 1 indicates high stability. The calculation uses the sample variance formula, involving the sum of squared deviations of data points from the mean divided by the number of data points minus 1, ensuring consistent dimensions; for example, water level data is in meters, while variance is in square meters. The oscillation stability index output is a dimensionless numerical value used for comprehensive evaluation. The statistical mean of time series data is calculated to obtain a quantitative index, reflecting the hydrological intensity of the current period; for example, calculating the arithmetic mean of water level or flow data within the current time window. The quantitative index output is a numerical form used to identify the abundance or scarcity of a watershed.

[0033] The current watershed scenario is determined based on a comprehensive evaluation of coupling strength, oscillation stability, and magnitude indices. The coupling strength index is compared to a first preset threshold, which is set based on the statistical quantiles of historical coupling strength data and determined by sorting historical data and calculating quantiles. The oscillation stability index is compared to a second preset threshold, which is also set based on the statistical quantiles of historical stability data. Predefined decision rules are applied; for example, if the coupling strength index is higher than the first preset threshold and the oscillation stability index is lower than the second preset threshold, the current watershed scenario is determined to be unstable; otherwise, it is determined to be stable.

[0034] Extracting time series data involves data preprocessing, including cleaning the raw real-time hydrological and meteorological data, removing outliers or noise, and using a sliding window method to segment the data. The window size is set to 7 days or 30 days, adjusted according to the watershed characteristics. At least two elements are extracted, such as simultaneously analyzing combinations of water level and rainfall, or flow and temperature, to cover multi-source coupling relationships. Data standardization is also performed, for example, using the Z-score method to convert the data into a series with a mean of 0 and a standard deviation of 1, reducing the influence of dimensions.

[0035] The cross-correlation function is calculated using the sliding window method. The correlation coefficients under different lags are calculated, and the maximum absolute value is selected as the coupling strength index. A high value of the coupling strength index indicates a strong dynamic correlation between elements, while a low value indicates a weak correlation.

[0036] Variance calculation is based on the volatility of time series. For example, to calculate the variance of each element within the window, the formula is used to sum the squared differences between each data point and the mean, and then divided by the number of data points minus 1. The oscillation stability index is obtained by taking the reciprocal of the variance or by linear transformation. For example, the stability index is equal to 1 minus the normalized variance. Normalization uses the historical maximum variance as a benchmark to ensure that the index range is consistent. The data distribution is considered during the calculation, such as performing a logarithmic transformation on skewed distributions.

[0037] The first and second preset thresholds are set based on historical data statistical analysis. For example, data on coupling strength index and oscillation stability index from the past 5 years are collected and their distribution characteristics are calculated. The first preset threshold is taken as the upper quartile of the coupling strength index, and the second preset threshold is taken as the lower quartile of the oscillation stability index. The threshold update cycle is set to 1 year and dynamically adjusted according to the latest data to ensure that the thresholds reflect the trend of watershed changes.

[0038] The third preset threshold is set based on historical data statistical analysis, such as collecting quantitative index data (including flow, water level, or rainfall) from the past 5 years, calculating its distribution characteristics, and taking the upper quartile (P75) of the quantitative index as the third preset threshold. The threshold update cycle is set to 1 year, and it is dynamically adjusted according to the latest data to ensure that the threshold reflects the changing trend of the watershed's wet and dry seasons. Through the combination mapping of the first, second, and third preset thresholds, a precise characterization of the "wet-dry" and "stable-chaotic" multi-dimensional scenarios can be achieved.

[0039] The predefined decision rules are based on expert knowledge or historical cases. For example, historical scenario data is trained through machine learning classifiers to generate rule sets. The rule form is a conditional statement. For example, when the coupling strength index is greater than 0.8 and the oscillation stability index is less than 0.3, it is determined to be an unstable scenario; otherwise, it is a stable scenario. When the rules are applied, logical evaluation is performed to output the scenario type for subsequent scheduling decisions.

[0040] The selection of multi-source hydrological and meteorological elements is based on watershed characteristics. For example, in arid areas, rainfall and water level are given priority, while in humid areas, humidity and flow factors are added. The number of elements can be expanded, for example, from 2 to 5. The time series length can be adjusted according to the analysis needs. For example, 7 days of data can be used for short-term analysis, and 30 days of data can be used for long-term analysis. The extraction method is compatible with different data formats, such as CSV or database records.

[0041] The calculation of coupling strength indices supports various statistical methods, including Spearman rank correlation in addition to Pearson correlation coefficient. The calculation of oscillation stability indices can incorporate coefficients of variation, such as the ratio of variance to mean, to enhance the robustness of stability characterization.

[0042] The threshold comparison process introduces a buffer mechanism, such as performing multiple verifications when the indicator is close to the threshold to reduce false judgments. The decision rules are configurable, such as modifying the threshold and logical conditions through configuration files to adapt to different watershed scenarios. The comprehensive evaluation results not only output the scenario type but can also be supplemented with confidence scores, such as calculating the probability value based on the distance between the indicator and the threshold.

[0043] When calculating the cross-correlation function, non-stationary sequences are processed, for example, by decomposing the time series through wavelet transform and then calculating the correlation of the subsequences. The coupling strength index can aggregate multiple factor pairs, for example, by calculating the average correlation of all pairs as the comprehensive coupling strength. The oscillation stability index can be calculated by weighting, for example, by assigning weights according to the importance of the factors. The variance calculation takes into account seasonal effects, for example, by using deseasonalization methods to eliminate the periodic effects.

[0044] Threshold settings reference historical events in the watershed, such as using past flood or drought periods as threshold benchmarks. Decision rules integrate opinions from multiple experts, such as using the Delphi method to determine rule parameters. The comprehensive evaluation process is iterative, such as reviewing indicator trends after the initial assessment to improve accuracy. Output scenarios are used to trigger subsequent scheduling actions, such as initiating emergency response in unstable scenarios.

[0045] S3. Establish a multi-dimensional constraint system based on the current watershed situation, including hydrological conditions, engineering capabilities, ecological needs, and socio-economic objectives; the specific implementation is as follows: For example, if the current watershed scenario is determined to be a short-term high-water unstable scenario (i.e., a scenario where the quantitative index exceeds the third preset threshold and the oscillation stability index is below the second preset threshold), a multi-dimensional constraint system is established, encompassing hydrological conditions, engineering capabilities, ecological needs, and socio-economic objectives. The dimensional composition of this constraint system remains consistent across different scenarios, but the specific constraint limits dynamically switch based on the scenario identification results. First, river level data from real-time hydrological and meteorological data are combined with preset flood control limits to form hydrological condition constraints. The preset flood control limits are set based on historical flood data and flood control engineering design standards. For example, by statistically analyzing the highest water level records of the past 50 years, the flood level value with a frequency of 1% is calculated as the preset flood control limit, expressed in meters. Real-time river level data comes from sensor acquisition in step S1. The hydrological condition constraint is defined as the river level not exceeding the preset flood control limit, and the constraint form is an inequality, such as the water level being less than or equal to the preset flood control limit. This constraint is used to control flood risk and ensure watershed flood control safety.

[0046] The reservoir's water storage data and its design capacity constitute an engineering capacity constraint. The design capacity is obtained from the reservoir's engineering design documents or planning data. Real-time reservoir water storage data comes from sensor data collected in step S1. The engineering capacity constraint is defined as the reservoir's water storage not exceeding its design capacity. The constraint takes the form of an inequality, such as water storage being less than or equal to the design capacity. This constraint is used to maintain the safety of the reservoir structure and prevent dam overflow or dam failure events.

[0047] The minimum ecological flow, determined based on historical ecological monitoring data, constitutes an ecological demand constraint. Historical ecological monitoring data includes long-term recorded river flow, water quality parameters, and biological indicators. For example, it can be obtained by collecting average daily flow data from ecological monitoring stations over the past 10 years. The minimum ecological flow is set based on river health assessment methods, such as using the Tennant method to calculate the minimum flow required to maintain aquatic habitats. The unit is cubic meters per second. The ecological demand constraint is defined as the river flow not being lower than the minimum ecological flow. The constraint takes the form of an inequality, such as the flow being greater than or equal to the minimum ecological flow. This constraint is used to protect the watershed's ecosystem function and prevent ecological degradation.

[0048] The basic water supply demand, calculated based on regional population distribution and industrial layout data, constitutes a socio-economic objective constraint. Regional population distribution data comes from population statistics reports or census data, such as population figures by administrative division. Industrial layout data includes demand information for agricultural irrigation, industrial production, and domestic water use, such as water quotas for each industry obtained from economic development plans. Basic water supply demand is calculated by multiplying the average daily water consumption per capita by the total population, and then adding the sum of water consumption for each industry. For example, the average daily water consumption per capita is taken as 100 liters. Agricultural water consumption is determined based on irrigated area and crop water requirement coefficients, while industrial water consumption is calculated based on output value and water use efficiency. The unit is expressed in cubic meters per day. The socio-economic objective constraint is defined as the water supply system must meet the basic water supply demand. The constraint takes the form of an inequality, such as water supply exceeding or equaling the basic water supply demand. This constraint is used to ensure stable water use for people's lives and economic activities.

[0049] The process of setting the preset flood control limit water level involves historical data analysis, such as collecting the annual maximum water level sequence of multiple stations in the basin, fitting the data with extreme value distribution models such as the Gumbel distribution, and calculating the water level value with a return period of 100 years as the preset flood control limit water level. When setting the water level, the changes in river topography are taken into account, such as updating the river cross-section data through digital elevation models to ensure that the threshold reflects the current basin conditions.

[0050] The design capacity of a reservoir is obtained based on reservoir engineering design specifications, such as by consulting the capacity curve in the reservoir construction drawings or operation manual to determine the maximum safe water storage capacity. The capacity data is verified through on-site measurement or remote sensing technology, such as using a sonar depth sounder to measure the bottom topography of the reservoir and updating the capacity calculation model to ensure the accuracy of the engineering capacity constraints.

[0051] Methods for determining the minimum ecological flow include eco-hydrological analysis, such as analyzing ecological response data from historical flow sequences to identify the flow threshold for maintaining the reproduction of key species, or calculating the relationship between the wetted perimeter of a river cross section and the flow through the wetted perimeter method, selecting the inflection point flow as the minimum ecological flow, and taking into account seasonal changes when setting it, such as increasing the threshold during the dry season to protect the ecosystem.

[0052] The calculation of basic water supply demand takes into account population growth and industrial transformation. For example, population forecasting models are used to estimate the population size in the next 5 years. Combined with the industrial water quota adjustment coefficient, dynamic water supply demand is calculated. Data sources include statistical yearbooks and industry reports. For example, agricultural water quotas are obtained from irrigation test data, and industrial water quotas are extracted from enterprise water use audit reports to ensure that socio-economic goals are consistent with the actual development of the region.

[0053] The construction process of the multi-dimensional constraint system integrates the above constraints. For example, it combines hydrological condition constraints, engineering capacity constraints, ecological demand constraints, and socio-economic goal constraints into a constraint set. Each constraint is expressed in the form of a mathematical inequality and stored in a database or configuration file for subsequent analysis. The constraint parameters are updated regularly, such as reviewing the preset flood control limit water level and minimum ecological flow every year and adjusting the thresholds according to the latest data.

[0054] When river level data from real-time hydrological and meteorological data is used for hydrological condition constraints, data quality control is performed, such as removing outliers caused by sensor errors, using the moving average method to smooth the data sequence, and ensuring the reliability of constraint comparisons.

[0055] When reservoir storage data is used to constrain engineering capacity, the constraints are dynamically adjusted in conjunction with reservoir operation rules, such as considering different limit water levels during flood season and non-flood season, to prevent constraint conflicts.

[0056] When historical ecological monitoring data is used to constrain ecological demand, the problem of missing data should be addressed, for example by supplementing missing flow records through interpolation or using data from neighboring sites to ensure the continuity of minimum ecological flow.

[0057] When regional population distribution and industrial layout data are used to constrain socio-economic goals, spatial aggregation is performed. For example, township-level population data is aggregated to the watershed scale, and industrial water use data is aggregated by industry category to calculate total water supply demand. When constraining applications, the priority of water source allocation is considered, such as prioritizing domestic water use over industrial water use.

[0058] The mechanism for updating the preset flood control limit water level is based on real-time monitoring and early warning. For example, when the water level approaches the preset value, an alarm is triggered, the reasonableness of the threshold is manually reviewed, and the threshold is adjusted according to the weather forecast if necessary.

[0059] Verification of reservoir design capacity is achieved through regular safety assessments, such as conducting reservoir structure inspections every 5 years and adjusting the capacity value based on the inspection results to ensure the safety margin of engineering capacity constraints.

[0060] The setting of minimum ecological flow takes into account the opinions of multiple experts, such as collecting suggestions from ecologists and hydrologists through the Delphi method, and comprehensively determining the threshold to avoid subjective bias.

[0061] The calculation of basic water supply demand incorporates water-saving factors, such as considering improvements in water resource utilization efficiency and adjusting per capita water consumption or industrial water quotas, to make socio-economic goals more sustainable.

[0062] All constraints are constructed based on publicly available data and methods to ensure that those skilled in the art can implement them. The constraint system is implemented in programming on a computing platform, for example, by using optimization software to integrate constraints and outputting them as executable constraint files for subsequent feasible solution space analysis.

[0063] The preset flood control limit water level constrained by hydrological conditions takes into account the local characteristics of the watershed when setting it. For example, a higher threshold is used in mountainous rivers and a lower threshold is used in plain rivers. The threshold unit is uniformly set in meters, and the data source is noted as the historical flood database.

[0064] When obtaining the design capacity of a reservoir under engineering capacity constraints, different types of reservoirs are distinguished, such as gravity dams and arch dams, and different safety factors are used. The capacity value is verified by geometric measurement or fluid dynamics calculation.

[0065] When determining the minimum ecological flow under ecological demand constraints, biodiversity data is incorporated, for example, higher flow requirements are set for the spawning season of specific fish species, and the threshold is obtained through ecological model simulation.

[0066] When calculating the basic water supply demand constrained by socio-economic goals, climate factors are taken into account, such as reducing the water supply quota in drought years, and the constraint value is verified by a water resource supply and demand balance model.

[0067] The coordination of the multi-dimensional constraint system is achieved through weight allocation. For example, ecological needs are prioritized in water-abundant scenarios, while socio-economic goals are prioritized in water-scarce scenarios. However, weight setting is not part of this step and is only used as a reference for constraint integration.

[0068] Real-time data is used for time synchronization during constraint construction, such as unifying all data to the same timestamp to avoid constraint failure due to inconsistencies in time sequence.

[0069] When using historical data for threshold setting, perform normality tests, such as using the Shapiro-Wilk test to verify the data distribution and ensure the applicability of statistical methods.

[0070] Once constraints are formalized into mathematical expressions, they are stored in a structured database, such as using SQL tables to record constraint types, thresholds, and conditions, facilitating subsequent queries and modifications.

[0071] S4. Based on a multi-dimensional constraint system, analyze the changes in the feasible solution space that simultaneously satisfies short-term scheduling constraints and long-term planning constraints. When the feasible solution space is lower than a critical threshold, it is identified as a conflict state. The specific implementation is as follows: First, a linear programming model is constructed. The multi-dimensional constraint system includes hydrological constraints, engineering capacity constraints, ecological demand constraints, and socio-economic goal constraints. These constraints are transformed into a system of linear inequalities. For example, the hydrological constraint is expressed as river water level less than or equal to the preset flood control limit; the engineering capacity constraint is expressed as reservoir storage less than or equal to the reservoir's design capacity; the ecological demand constraint is expressed as river flow greater than or equal to the minimum ecological flow; and the socio-economic goal constraint is expressed as water supply greater than or equal to basic water supply demand. The objective function is set to maximize water resource allocation efficiency, for example, defined as the ratio of total water supply to total demand multiplied by a weighting coefficient. Decision variables include the allocation of each water source and the reservoir's discharge. The variable values ​​are set according to the actual water supply capacity; for example, the lower limit for water source allocation is 0, and the upper limit is the maximum available water volume. Model parameters are obtained from the watershed's basic database; for example, water supply capacity comes from engineering records, and demand comes from population and industry data. After model construction, the model is stored in the computing system in matrix form. The coefficient matrix contains constraint coefficients, and the objective function vector contains efficiency weights.

[0072] The range of solutions satisfying all constraints is determined by solving a linear programming model, using linear programming algorithms such as the simplex method or the interior-point method. The simplex method constructs an initial feasible basis and iteratively improves the solution until the optimal solution is found. The interior-point method searches for the optimal solution within the feasible region. Convergence conditions are set during the calculation, such as terminating when the change in the objective function is less than 0.001 or the number of iterations exceeds 1000. The range of solutions is determined by analyzing the optimal range of the decision variables. For example, the range of values ​​for each decision variable is calculated near the optimal solution. Considering the boundary effects of the constraints, the range of solutions is represented as a set of intervals for the decision variables. For example, the reservoir discharge is between 100,000 and 500,000 cubic meters, and the river diversion is between 50,000 and 300,000 cubic meters. The range of values ​​is verified through sensitivity analysis, such as observing changes in the right-hand side of the constraints to ensure the stability of the solution set.

[0073] The volume of the solution set's domain is used as the volume index of the feasible solution space. High-dimensional volume calculation methods are employed, such as treating the feasible region comprised of decision variables as a polyhedron and calculating its volume through triangulation or Monte Carlo methods. Triangulation decomposes the polyhedron into simplexes, calculates the volume of each simplex, and sums the volumes. The Monte Carlo method estimates the volume by statistically analyzing the proportion of randomly sampled points falling within the feasible region and multiplying this by the total volume of the sampled space. The feasible solution space volume index is expressed in cubic units; for example, if the decision variables are in units of 10,000 cubic meters, the volume index is in cubic meters. Normalization is performed during calculation, such as dividing by the theoretical maximum volume to obtain a relative volume value, ensuring the index ranges between 0 and 1 for easier subsequent comparisons.

[0074] The feasible solution space volume index is compared with a preset critical threshold. The preset critical threshold is set based on the statistical characteristics of historical feasible solution space volume data. For example, the feasible solution space volume index sequence of the past 10 years is collected, and its average value or a specific quantile is calculated as the threshold. For example, the 10th percentile value of historical data is taken as the preset critical threshold. The threshold setting takes into account the watershed characteristics and management needs. For example, a lower threshold is used in arid watersheds and a higher threshold is used in water-rich watersheds. The threshold unit is consistent with the feasible solution space volume index. For example, both are dimensionless relative values. The preset critical threshold is stored in the configuration file and is updated once a year according to the latest data to ensure the adaptability of the threshold.

[0075] When the volume index of the feasible solution space is lower than a preset critical threshold, it is identified as a conflict state. The conflict state is represented by a Boolean value or a level flag. For example, when the volume index is less than the threshold, the conflict state is output as true, otherwise it is false. Conflict state identification is achieved through conditional judgment, such as writing a comparison statement in the calculation program and recording the result in the state variable. Conflict states are used to trigger subsequent scheduling adjustments, such as marking abnormal situations that need to be prioritized in a water resource management system.

[0076] The construction of a linear programming model specifically includes defining the dimensions of decision variables. For example, if there are 3 water sources and 5 users, the number of decision variables is 15. The number of constraints is determined according to a multi-dimensional constraint system. For example, there are 2 hydrological constraints, 3 engineering capacity constraints, 2 ecological demand constraints, and 4 socio-economic goal constraints, for a total of 11 constraints. The coefficients of the objective function are set according to the priority of water resource allocation. For example, the weight of domestic water use is 0.5, the weight of industrial water use is 0.3, and the weight of agricultural water use is 0.2. The coefficients are determined through expert consultation or historical data analysis. The model input parameters are updated in real time, for example, by obtaining the latest constraint values ​​from the monitoring system through a data interface.

[0077] When solving linear programming models, it handles cases where there is no solution or the solution is unbounded. For example, it returns a no-solution flag when constraints are contradictory, and checks the integrity of constraints when the objective function is unbounded. The calculation of the solution set range considers the basic solution and the boundary solution. For example, it finds the coordinates of the vertices of the feasible region by enumerating vertices, and then calculates the volume of the space enclosed by the vertices. The solution set range output is a list of the minimum and maximum values ​​of each decision variable. For example, the allocation of water source 1 is between 0 and 1 million cubic meters, and the allocation of water source 2 is between 0 and 800,000 cubic meters.

[0078] The volume index of the feasible solution space is calculated using a numerical integration method. For example, uniform grid points are generated within the decision variable range, the proportion of grid points that satisfy all constraints is statistically analyzed, and the volume index is obtained by multiplying it by the total volume of the decision space. The calculation accuracy is controlled by the grid density, for example, 100 grid points are divided in each dimension, and the volume index is retained to 4 decimal places to ensure the reliability of the calculation results.

[0079] The preset critical threshold setting introduces a dynamic adjustment mechanism. For example, different thresholds are set according to seasonal changes, with the threshold set to 0.6 during the rainy season and 0.4 during the dry season. Threshold verification is achieved by backtracking through historical conflict events, such as checking whether the volume index was below the threshold when conflicts occurred in the past. If necessary, the threshold selection is optimized through ROC curves.

[0080] The conflict status identification adds a buffer zone. For example, when the volume index is below the threshold but exceeds 90% of the threshold, it is marked as a potential conflict. The status output includes the conflict level, such as mild conflict, moderate conflict and severe conflict. The level is divided according to the difference between the index and the threshold. For example, a difference of less than 10% is mild, 10% to 30% is moderate, and more than 30% is severe.

[0081] The integration of a multi-dimensional constraint system ensures constraint linearization. For example, nonlinear ecological demand constraints can be approximated as piecewise linear constraints. Constraint consistency is checked during model construction, and redundant constraints can be eliminated through linear programming preprocessing to improve solution efficiency.

[0082] The choice of algorithm is based on the problem size. For example, the simplex method is used when there are fewer than 100 variables, and the interior point method is used when there are more than 100 variables. The algorithm parameters are set to default values. For example, the maximum number of iterations for the simplex method is 5000, and the tolerance for the interior point method is 0.0001.

[0083] The calculation of the solution set range is combined with parametric programming. For example, by changing the coefficients of the objective function and observing the changes in the solution, the range output is an interval matrix. For example, a 3x2 matrix represents the range of values ​​of the three decision variables in two scenarios.

[0084] Optimize sampling strategies for calculating volume indices in feasible solution spaces, such as using important sampling to improve efficiency in Monte Carlo methods, and add uncertainty estimation to volume indices, such as obtaining the standard deviation through repeated calculations.

[0085] The preset critical threshold is updated using a sliding window method, such as using data from the most recent 5 years to calculate the threshold, to avoid historical data becoming outdated. Threshold adjustments are verified through management strategies, such as testing the impact of the threshold on conflict identification in a simulated environment.

[0086] The output of conflict state identification is integrated into the decision support system. The conflict state is recorded with a timestamp and a cause code. For example, cause code 1 indicates a hydrological constraint conflict, and code 2 indicates an ecological constraint conflict.

[0087] The objective function of a linear programming model can be expanded, for example, by adding cost minimization or fairness maximization as multiple objectives, which can be transformed into a single objective through weighted summation. Model validation is achieved through test cases, such as constructing problems with known solutions to check the correctness of the solution.

[0088] Visualization of the solution set range is achieved through dimensionality reduction, for example, by selecting key decision variables to draw a two-dimensional feasible region, which helps to understand the solution space structure. The range data is stored in JSON format for easy access by other systems.

[0089] Trend analysis of volume indicators in the feasible solution space is achieved through time series analysis, such as calculating weekly volume indicators to observe the changing trends. Anomaly detection is achieved through control chart methods, such as triggering checks when the volume exceeds 3 times the standard deviation.

[0090] The regional differences in the preset critical thresholds take into account the characteristics of the sub-basins. For example, the upstream threshold is set to 0.7 and the downstream threshold is set to 0.5. The threshold setting is based on the watershed planning objectives. For example, a stricter threshold is used in ecological protection areas.

[0091] The response strategy for conflict states is predefined, such as activating an emergency dispatch plan when the conflict state is true, and the status is continuously monitored through real-time data streams, such as updating volume indicators and conflict states every 6 hours.

[0092] S5. Simulate the operational status of the watershed water resources system under various predefined adverse future scenario evolution paths based on conflict states, and evaluate the system's resource supply vulnerability index under each evolution path; the specific implementation is as follows: First, predefined future adverse scenario evolution paths, such as drought exacerbation scenarios and water demand growth scenarios, are selected based on the conflict state. The conflict state is derived from the output of step S4; for example, the scenario selection process is automatically triggered when the conflict state is true. The drought exacerbation scenario is defined as an unfavorable condition of continuously decreasing rainfall and rising temperatures in the future period. For example, rainfall is set to decrease by 20% compared to the historical average for the same period, and temperatures rise by 2 degrees Celsius, with a scenario duration of 3 months. The water demand growth scenario is defined as an unfavorable condition of significantly increased water demand due to population growth or economic development. For example, the annual population growth rate is set to 1.5%, the annual industrial output growth rate to 5%, and the agricultural irrigation area to expand by 10%, with a scenario duration of 1 year. Scenario parameters are set based on historical climate data and socio-economic forecast data. For example, rainfall trend data is obtained from meteorological departments, and population and industry forecast data are obtained from statistical departments. The predefined future adverse scenario evolution paths are stored in the scenario parameter database. Applicable scenarios are selected through query logic; for example, scenario intensity is matched according to the severity of the conflict state, with a parameter change of 10% for mild conflict and 30% for severe conflict.

[0093] The dynamic changes in water supply and demand of the basin's water resources system under various evolution paths are simulated using water balance equations. The water balance equations are constructed based on the principle of mass conservation, and the expression is that the total input water volume minus the total output water volume equals the change in water storage. Input water volume includes rainfall, river inflow, and inter-basin water transfer; output water volume includes evaporation, river outflow, water intake, and seepage. The simulation of dynamic changes in supply and demand employs a time-stepping method, for example, using days as the time step to calculate the supply and demand balance state at each step. The supply calculation considers the available surface water and the exploitable groundwater. For example, the available surface water volume is determined based on river flow and reservoir storage, while the exploitable groundwater volume is determined based on aquifer reserves and recharge rate. Water demand calculations include domestic water demand, industrial water demand, and agricultural water demand. For example, domestic water demand is obtained by multiplying the average daily water consumption per capita by the population; industrial water demand is obtained by multiplying the water consumption per 10,000 yuan of output value by the industrial output value; and agricultural water demand is obtained by multiplying the irrigation water consumption per unit area by the irrigated area. The simulation process begins with initial hydrological conditions, such as using the current river level and reservoir storage as initial values, and iteratively calculates the changes in storage and supply and demand at each time step until the entire scenario period is covered. The simulation output includes the supply and demand sequences for each time step, as well as the storage change curve, used to analyze the dynamic behavior of the system.

[0094] The ratio of water supply to demand gaps under each evolutionary path is calculated as a resource supply vulnerability indicator. The gap ratio is defined as the difference between demand and supply, divided by the demand, and then multiplied by 100% to obtain a percentage value. The resource supply vulnerability indicator is the maximum gap ratio at each time step within the simulation period; for example, after calculating the daily gap ratio, the maximum value is selected as the vulnerability indicator for that scenario. The indicator value ranges from 0% to 100%, where 0% indicates that supply fully meets demand, and 100% indicates a complete lack of supply. Cases where supply exceeds demand are handled during calculation; for example, when supply exceeds demand, the gap ratio is set to 0% to avoid the impact of negative values. The resource supply vulnerability indicator is output as a single value for each scenario; for example, the vulnerability indicator is 15% under the drought exacerbation scenario and 25% under the water demand increase scenario. The indicator is stored in the assessment results table for subsequent comparison and decision-making.

[0095] The scenario selection process specifically includes parameterizing scenario characteristics. For example, in a drought exacerbation scenario, a monthly rainfall reduction gradient is set: a 10% reduction in the first month, a 20% reduction in the second month, and a 30% reduction in the third month. In a water demand growth scenario, quarterly water demand growth rates are set: a 3% increase in the first quarter, a 5% increase in the second quarter, a 7% increase in the third quarter, and a 10% increase in the fourth quarter. Scenario parameters are determined through expert consultation or analysis of historical extreme events. For example, drought exacerbation scenario parameters are set by referencing rainfall data from past drought years, and water demand growth scenario parameters are set by referencing water consumption data from periods of rapid economic development. The selection logic is based on conflict state attributes; for example, when the conflict state includes conflict levels, higher-level conflicts correspond to more severe scenario parameters.

[0096] The water balance equation simulation involves data preprocessing, such as smoothing the input hydrological data, using moving averages to eliminate random fluctuations, and seasonally adjusting water demand data to account for annual variations in water use habits. The equations are solved using numerical methods, such as the Euler method or the Runge-Kutta method for iterative calculations. Step size accuracy is ensured through error control, for example, setting a relative error tolerance of 0.001. Simulation input parameters are updated in real time; for example, rainfall data is obtained from meteorological forecasting systems, and water demand data is obtained from economic monitoring platforms to ensure the simulation reflects the latest conditions.

[0097] The gap ratio calculation incorporates weighting factors, such as assigning different weights to the gap ratio at different time steps, with a higher weight for the dry season than the rainy season, to highlight the impact of critical periods. The resource supply vulnerability index can be expanded into a comprehensive index, for example, by calculating a weighted average based on the gap duration. The index is also standardized, for example, by dividing it by the historical maximum gap ratio to obtain a relative value, making the index comparable across scenarios.

[0098] Simulation validation is achieved through historical backtesting, such as running simulations using data from the past 5 years, comparing the simulation results with actual events, and adjusting model parameters as necessary. Resource supply vulnerability indicators are assessed considering uncertainty, for example, by generating multiple parameter combinations using Monte Carlo methods, calculating the probability distribution of the indicators, and outputting expected values ​​and confidence intervals.

[0099] The entire simulation and evaluation process is automated on the computing platform. For example, scripts are written to invoke the water balance model, inputting conflict state and scenario parameters, outputting resource supply vulnerability indicators, and logging simulation steps and intermediate results for auditing and analysis. Integrated decision tree logic automatically selects a more severe scenario if the conflict lasts longer than 7 days. The water balance equation parameterizes watershed characteristics, such as setting infiltration coefficients based on watershed area and soil type, and setting evaporation coefficients based on vegetation cover.

[0100] The simulation outputs visualizations of dynamic supply and demand changes, such as plotting curves of supply and demand over time, to help identify points of supply-demand imbalance. Resource supply vulnerability indicators are used for prioritization; for example, high-vulnerability scenarios are marked as requiring urgent response. The simulated resource allocation considers multiple water sources, such as surface water, groundwater, and reclaimed water, and refines water demand classifications, such as domestic water demand for urban and rural residents, and industrial water demand for high-water-consuming and low-water-consuming industries.

[0101] The scenario evolution path can be expanded to include more types, such as scenarios of increased pollution or infrastructure failure, but this step focuses on drought and increased water demand. The water balance equation is extended to account for the impacts of climate change, for example, by setting up scenarios where rising temperatures lead to increased evaporation, and the evaporation rate is dynamically updated during simulation. The deficit ratio calculation addresses missing data, for example, by using interpolation estimation when some water demand data is unavailable.

[0102] Resource supply vulnerability indicators are stored in a database and correlated with scenario parameters for easy historical querying and trend analysis. The indicator calculation code is either open-source or documented to ensure reproducibility by those skilled in the art.

[0103] S6. Adjust dynamic weights based on conflict status and resource supply vulnerability indicators to balance short-term and long-term objectives. Specific implementation details are as follows: First, conflict states are mapped to conflict severity levels. Conflict states are derived from the output of step S4, such as a Boolean value (true or false) or a numerical flag (0 or 1). The mapping process uses predefined conflict severity levels, for example, setting three levels: low, medium, and high. If a conflict state is false, it is mapped to a low level; if a conflict state is true, it is further subdivided based on the conflict duration or associated feasible solution space volume index value. For example, a conflict lasting less than 3 days is mapped to a medium level, and a conflict lasting 3 days or more is mapped to a high level. The mapping rules are set based on historical conflict event analysis, for example, by determining threshold points through statistical analysis of the duration distribution of conflict data over the past 5 years. The level output is a string code such as low, medium, or high, stored in a state variable for later use.

[0104] Resource supply vulnerability indicators are mapped to vulnerability levels. These indicators are derived from the output of step S5, for example, as a percentage value between 0% and 100%. The mapping process uses predefined vulnerability level classifications, such as setting three levels: low, medium, and high. Threshold ranges are set; for example, a resource supply vulnerability indicator between 0% and 30% maps to a low level, between 30% and 60% to a medium level, and between 60% and 100% to a high level. These thresholds are set based on watershed management objectives and historical vulnerability data; for example, by analyzing the distribution of vulnerability indicators in drought events over the past 10 years, 30% and 60% are used as dividing points. The mapping results are output as level identifiers such as L, M, and H, ensuring consistency with the conflict severity level format for easy subsequent combined queries.

[0105] The predefined weight adjustment table retrieves the adjusted values ​​for short-term and long-term target weights based on a combination of conflict severity and vulnerability levels. The predefined weight adjustment table is a two-dimensional lookup table, with row indices representing conflict severity levels and column indices representing vulnerability levels. Each cell contains two values: a short-term target weight adjustment value and a long-term target weight adjustment value. For example, in the table structure, when both the conflict severity and vulnerability levels are high, the short-term target weight adjustment value is +0.2, and the long-term target weight adjustment value is -0.2; when both the conflict severity and vulnerability levels are low, the short-term target weight adjustment value is -0.1, and the long-term target weight adjustment value is +0.1. The weight adjustment table is optimized based on expert experience or historical scheduling results, for example, by using regression analysis to determine the adjustment value range based on the relationship between weights and performance in past scheduling cases. The query process retrieves the corresponding adjustment value pairs from the table by matching level combinations, outputting the short-term and long-term adjustment values. For example, +0.1 and -0.1 represent a short-term weight increase of 0.1 and a long-term weight decrease of 0.1.

[0106] The dynamic weights are updated using adjusted values. Initial dynamic weight values ​​are obtained from the system configuration library; for example, the initial short-term target weight is 0.6, and the initial long-term target weight is 0.4. The update process uses additive adjustment; for example, the new short-term target weight equals the original short-term target weight plus the adjusted short-term target weight, and the new long-term target weight equals the original long-term target weight plus the adjusted long-term target weight. After adjustment, the total weight sum is checked to ensure it equals 1. If not, normalization is performed; for example, the new short-term and long-term target weight sums are divided by their sum to ensure the adjusted total weight sum equals 1. The updated dynamic weights are stored in the system parameter database for subsequent water resource scheduling models, such as as objective function weights in optimization algorithms. The entire adjustment process is automated; for example, mapping, querying, and updating logic can be implemented using programming scripts, taking conflict status and resource supply vulnerability indicators as input, and outputting updated dynamic weight values.

[0107] The mapping of conflict severity levels can be refined to more levels, such as adding an "emergency" level, corresponding to situations where the conflict lasts for more than 10 days and the feasible solution space volume index is below 10% of the historical low. The mapping rules consider a combination of multiple factors; for example, if the conflict state is true and the associated feasible solution space volume index is below a preset critical threshold of 50%, it is directly mapped to a high level. The threshold setting is dynamically updated; for example, the duration threshold is recalculated annually based on the latest conflict data, and a moving average method is used to smooth historical data fluctuations.

[0108] The vulnerability level mapping incorporates weighting factors, such as assigning different weights to resource supply vulnerability indicators under different scenarios. For example, the weight for an exacerbated drought scenario is 0.7, and the weight for a water demand increase scenario is 0.3. The mapping level is then calculated comprehensively. The threshold range is adjustable; for instance, during the dry season, the lower limit of the vulnerability level threshold can be lowered from 30% to 20% to enhance sensitivity. The mapping results are cached in memory to improve query efficiency.

[0109] The predefined weight adjustment table construction process includes data collection and validation, such as collecting conflict levels, vulnerability levels, and corresponding weight adjustment values ​​from historical scheduling records, and generating table content using machine learning algorithms such as decision trees. The table supports dynamic loading, such as reading table data from configuration files, allowing administrators to modify adjustment values ​​based on watershed changes. Query logic is optimized, for example, using hash tables to accelerate level combination lookups and ensure real-time performance.

[0110] When updating dynamic weights with adjusted values, boundary cases are handled. For example, if the adjusted weight exceeds a reasonable range (e.g., less than 0 or greater than 1), it is forcibly truncated to 0 or 1. The update strategy is configurable, allowing for options such as additive or multiplicative adjustment. In multiplicative adjustment, the new short-term target weight is equal to the original short-term target weight multiplied by the adjustment coefficient. A log is recorded after the weight update, including the adjustment time, original weight value, new weight value, and reason for adjustment, for auditing and backtracking purposes.

[0111] Example 2: like Figure 2 The diagram shows a schematic representation of a dynamic weighted scheduling system for watershed water resources based on scenario recognition, according to the present invention, comprising the following modules: The data acquisition module is used to acquire real-time hydrological and meteorological data of the watershed through a sensor network; The scenario assessment module is used to analyze the stability of coupled oscillation modes among multiple hydrological and meteorological elements based on real-time hydrological and meteorological data in order to determine the current watershed scenario. The system establishment module is used to establish a multi-dimensional constraint system that includes hydrological conditions, engineering capabilities, ecological needs, and socio-economic goals when the current watershed scenario is determined to be a short-term high-water scenario. The state recognition module is used to analyze the changes in the feasible solution space that simultaneously satisfies short-term scheduling constraints and long-term planning constraints based on a multi-dimensional constraint system. When the feasible solution space is lower than the critical threshold, it is identified as a conflict state. The indicator evaluation module is used to simulate the operation of the watershed water resources system under various predefined future adverse scenario evolution paths based on conflict state, and to evaluate the system's resource supply vulnerability indicators under each evolution path. The weight adjustment module is used to adjust dynamic weights based on conflict status and resource supply vulnerability indicators to balance short-term and long-term objectives.

[0112] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0113] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Whether these functions are implemented in hardware or software depends on the specific application of the technical solution and the constraints of the invention. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0114] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0115] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0116] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic weighted scheduling method for watershed water resources based on scenario recognition, characterized in that, Includes the following steps: S1. Acquire real-time hydrological and meteorological data of the watershed through sensor networks; S2. Based on real-time hydrological and meteorological data, analyze the quantitative characteristics of multi-source hydrological and meteorological elements and the stability of the coupled oscillation modes between the elements to judge the current watershed situation. S3. Based on the current watershed situation, establish a multi-dimensional constraint system that includes hydrological conditions, engineering capabilities, ecological needs, and socio-economic goals; S4. Based on a multi-dimensional constraint system, analyze the changes in the feasible solution space that simultaneously satisfies short-term scheduling constraints and long-term planning constraints. When the feasible solution space is lower than the critical threshold, it is identified as a conflict state. S5. Simulate the operational status of the watershed water resources system under various predefined future adverse scenario evolution paths based on conflict state, and evaluate the resource supply vulnerability index of the system under each evolution path. S6. Adjust dynamic weights based on conflict status and resource supply vulnerability indicators to balance short-term and long-term objectives.

2. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 1, characterized in that: In step S1, real-time hydrological and meteorological data are collected by sensors deployed in rivers, reservoirs, and meteorological stations, including river water level data, river flow data, reservoir water storage data, rainfall data, temperature data, and humidity data.

3. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 1, characterized in that: Step S2 specifically includes extracting time series data of multiple hydrological and meteorological elements from real-time hydrological and meteorological data, calculating the mean value of the time series data as a quantitative indicator, and calculating the cross-correlation function between the time series data, using the absolute value of the correlation coefficient as a coupling strength indicator. Calculate the variance of time series data and use the reciprocal of the variance or the normalized variance as an indicator of oscillation stability. The current watershed situation is judged by comprehensively considering quantitative indicators, coupling strength indicators, and oscillation stability indicators.

4. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 3, characterized in that: The coupling strength index is compared with a first preset threshold, the oscillation stability index is compared with a second preset threshold, the magnitude index is compared with a third preset threshold, and based on the comparison results, a predefined decision rule is applied to determine whether the current watershed scenario is a short-term high-water scenario, a normal-water scenario, or a low-water scenario, and the stable or unstable state under the scenario is identified.

5. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 1, characterized in that: The multi-dimensional constraint system in step S3 includes: hydrological constraints based on river water level data and preset flood control limit water level; engineering capacity constraints based on reservoir water storage data and reservoir design capacity; ecological demand constraints based on the minimum ecological flow determined by historical ecological monitoring data; and socio-economic objective constraints based on the calculation of basic water supply demand based on regional population distribution and industrial layout data.

6. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 1, characterized in that: Step S4 specifically includes: constructing a linear programming model based on a multi-dimensional constraint system; determining the range of solution sets that satisfy all constraints by solving the linear programming model; calculating the volume of the solution set range as a feasible solution space volume index, and comparing the feasible solution space volume index with a preset critical threshold; identifying a conflict state when the feasible solution space volume index is lower than the preset critical threshold.

7. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 6, characterized in that: Step S4 specifically involves constructing a linear programming model with water resource allocation efficiency as the objective and a multi-dimensional constraint system as the constraint, and using a linear programming solution algorithm to calculate the range of values ​​for the decision variables as the solution set range.

8. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 1, characterized in that: Step S5 includes selecting a predefined future adverse scenario evolution path based on the conflict state; simulating the dynamic changes in the supply and demand of the watershed water resources system under each evolution path using the water balance equation; and calculating the gap ratio between water resource supply and demand under each evolution path as an indicator of resource supply vulnerability.

9. The dynamic weighted scheduling method for watershed water resources based on scenario recognition according to claim 1, characterized in that: Step S6 includes mapping the conflict state to a conflict severity level and mapping the resource supply vulnerability index to a vulnerability level; querying a predefined weight adjustment table based on the combination of the conflict severity level and vulnerability level to obtain the adjustment values ​​of the short-term target weight and the long-term target weight, and applying the adjustment values ​​to update the dynamic weights.

10. A dynamic weighted scheduling system for watershed water resources based on scenario recognition, used to implement the dynamic weighted scheduling method for watershed water resources based on scenario recognition as described in any one of claims 1-9, characterized in that, Includes the following modules: The data acquisition module is used to acquire real-time hydrological and meteorological data of the watershed through a sensor network; The scenario assessment module is used to analyze the stability of coupled oscillation modes among multiple hydrological and meteorological elements based on real-time hydrological and meteorological data in order to determine the current watershed scenario. The system establishment module is used to establish a multi-dimensional constraint system that includes hydrological conditions, engineering capabilities, ecological needs, and socio-economic goals when the current watershed scenario is determined to be a short-term high-water scenario. The state recognition module is used to analyze the changes in the feasible solution space that simultaneously satisfies short-term scheduling constraints and long-term planning constraints based on a multi-dimensional constraint system. When the feasible solution space is lower than the critical threshold, it is identified as a conflict state. The indicator evaluation module is used to simulate the operation of the watershed water resources system under various predefined future adverse scenario evolution paths based on conflict state, and to evaluate the system's resource supply vulnerability indicators under each evolution path. The weight adjustment module is used to adjust dynamic weights based on conflict status and resource supply vulnerability indicators to balance short-term and long-term objectives.

Images