Water resources coordinated allocation method and system for drought risk response

Abstract

The present invention discloses a method and system for coordinating and allocating water resources in response to drought risk. Based on the monitoring data of the spatial distribution of the study area, a plurality of downscaling methods are used to downscale the climate model, and the data are input into a pre-constructed grid hydrological model to simulate and obtain the hydrological data of the study area; the drought risk assessment factors are screened, and the gridded data are input into the drought risk assessment model to obtain the drought risk zoning of the study area; the hydrological data of the study area, the available water volume of the study area and the drought risk zoning of the study area are input into a pre-constructed multi-objective optimization model of water resources to obtain the optimal allocation scheme of water resources; the optimal allocation scheme of water resources is input into a pre-constructed water engineering cluster optimization scheduling model, and the model is solved cyclically to obtain the optimal scheduling scheme of the water engineering cluster. The present invention is helpful to enhance the flexibility and adaptability of water resources management, reduce the impact of drought, maintain ecological balance and improve social and economic benefits.

Description

Technical Field

[0001] The invention relates to a method for coordinating and allocating water resources in response to drought risks. Background Art

[0002] Drought is one of the most common natural disasters in the world. It poses a serious threat to the ecological environment, economic development and social stability. Effective water resource management and allocation have become the key to alleviating the impact of drought. As climate change intensifies, water resources are no longer a static resource. How to optimize the allocation and use of water resources has become a pressing problem.

[0003] In the field of water resources management, a variety of technical means have been applied this year, including: water resources simulation models, water resources allocation models, intelligent water supply systems and data analysis and decision support systems. These technologies have played an important role in water resources management and provided technical support for coordinating and coordinating water resources in spatial and temporal imbalances, multi-objective conflicts and other difficult problems. However, existing water resources allocation methods often focus on the balance between water supply and demand, and due to the difficulty of drought prediction, existing model methods cannot make adaptive allocation of water resources according to future drought risks.

[0004] The present invention proposes a method and system for coordinating and allocating water resources to cope with drought risks, which solves the above-mentioned problems that currently exist, optimizes resource allocation, improves utilization efficiency, enhances flexibility and adaptability, reduces the impact of drought, maintains ecological balance, and improves social and economic benefits. Summary of the invention

[0005] Purpose of the invention: To provide a method for coordinating and allocating water resources in response to drought risks, so as to solve the above problems existing in the prior art. On the other hand, to provide a system for coordinating and allocating water resources in response to drought risks.

[0006] Technical solution: A method for coordinating and allocating water resources to cope with drought risks, including the following steps:

[0007] Step S1, collecting the climate model and the monitoring data of each meteorological station in the study area, using the weighted optimized Thiessen polygon method to spatially distribute the monitoring data to obtain distribution data; reducing the scale of the climate model to the scale of the distribution data, obtaining the meteorological grid data of the study area, inputting the meteorological grid data of the study area into a pre-built grid-type hydrological model for simulation, obtaining the hydrological data of the study area, and calculating the available water volume;

[0008] Step S2, collecting drought risk assessment literature, screening drought risk assessment factors, using the constructed drought risk assessment model to optimize the weights of drought risk assessment factors, rasterizing the drought risk assessment factors and inputting them into the drought risk assessment model to obtain the drought risk zoning of the study area;

[0009] Step S3, constructing a multi-objective optimization model for water resources and taking the hydrological data, available water volume and drought risk zoning of the study area as input, solving the model using an optimization algorithm to obtain the optimal allocation plan for water resources in the study area;

[0010] Step S4: construct a water project cluster optimization scheduling model and take the optimal water resource allocation plan of the study area as input, perform a cyclic solution, and obtain the water project cluster optimization scheduling plan of the study area.

[0011] According to one aspect of the present application, step S1 further comprises:

[0012] Step S11, collecting monitoring data of each meteorological station in the study area, and using weighted optimized Thiessen polygon method to spatially distribute the monitoring data to obtain the spatial distribution of the monitoring data, that is, to obtain distribution data;

[0013] Step S12: based on the spread data, downscaling the climate model using at least two downscaling methods, reducing the scale of the climate model to the scale of the spread data, so that the meteorological data of the study area matches the scale of the climate model, and obtaining the meteorological raster data of the study area based on the corresponding relationship between the meteorological data of the study area and the climate model, including: historical meteorological raster data of the study area and future meteorological raster data of the study area;

[0014] Step S13: using the meteorological grid data of the study area as input data, calling a pre-built grid-type hydrological model for solution, obtaining the hydrological data of the study area, and calculating the available water based on the hydrological data.

[0015] According to one aspect of the present application, the step S11 is further:

[0016] Step S11a, collecting monitoring data of each meteorological station in the study area, and determining the weight of each meteorological station based on the size of each meteorological station in the study area, the surrounding population density and the number of observations;

[0017] Step S11b, using GIS software to draw Thiessen polygons, each polygon corresponding to a meteorological station;

[0018] Step S11c, construct a weighted optimized Thiessen polygon method based on the weights of each meteorological station, adjust each Thiessen polygon according to the weight of each point in the Thiessen polygon method, that is, the boundary of each area, to obtain the spatial division, visualization results and data interpolation of the monitoring data.

[0019] According to one aspect of the present application, step S12 further comprises:

[0020] Step S12a, collecting data of at least two climate models, each climate model including: historical climate data and future climate data;

[0021] Step S12b, downscaling the data of each climate mode by using at least two statistical downscaling methods and at least two dynamic downscaling methods to obtain meteorological grid data of each climate mode, and fusing the meteorological grid data of each climate mode by using Bayesian averaging method to obtain fused meteorological grid data of each climate mode, wherein the fused meteorological grid data includes: historical meteorological grid data and future meteorological grid data;

[0022] Step S12c, based on expert opinions, weight each climate mode, calculate the meteorological grid data of the comprehensive climate mode based on the weight of each climate mode and the meteorological grid data, extract the historical meteorological grid data and compare it with the meteorological data of the study area and correct the weight of each climate mode, calculate the comprehensive meteorological grid data of the final climate mode, the comprehensive meteorological grid data of the final climate mode is the meteorological grid data of the study area, including the historical meteorological grid data of the study area and the future meteorological grid data of the study area. According to one aspect of the present application, the step S13 is further:

[0023] Step S13a, collect digital elevation data, land use and soil type of the study area and input them into the ArcSWAT module to obtain hydrological response units, and use the boundary of the meteorological grid data as a constraint to combine the various hydrological response units to obtain the calculation unit of the grid SWAT model, that is, to obtain the grid hydrological model;

[0024] Step S13b, collecting typical drought events in the study area, identifying and calculating the sensitivity of each parameter of the grid hydrological model, screening the grid hydrological model parameters based on the sensitivity, and obtaining calibrated grid hydrological model parameters;

[0025] Step S13c, extract the meteorological grid data of the study area and input it into the calibrated grid-type hydrological model to simulate and obtain the hydrological data of the study area, and calculate the available water volume based on the hydrological data.

[0026] According to one aspect of the present application, step S13b further comprises:

[0027] Step S13b1, collecting typical drought events in the study area and identifying grid-type hydrological model parameters;

[0028] Step S13b2, changing the value of a certain parameter and keeping other parameters unchanged, calculating the degree of change of flow and low water level, and obtaining the sensitivity of this parameter;

[0029] Step S13b3, calculating the sensitivity of all grid-type hydrological model parameters in turn, setting a threshold, and extracting all grid-type hydrological model parameters with a sensitivity greater than the threshold for parameter calibration;

[0030] Step S13b4, sorting the collected typical drought events in the study area from small to large according to the severity of drought, and the first typical drought event in the study area has the smallest severity of drought;

[0031] Step S13b5: based on expert opinions, select the value with the largest probability of the parameter as the prior distribution of the parameter in the first typical drought event scenario in the study area, and perform Bayesian inference to obtain the posterior distribution as the prior distribution of the parameter in the next typical drought event scenario in the study area;

[0032] Step S13b6, sequentially calculating the prior distribution and posterior distribution of the parameters under each typical drought event scenario in the study area;

[0033] Step S13b7: Use the posterior distribution of the parameters under the last typical drought event scenario in the study area as the grid hydrological model parameters.

[0034] According to one aspect of the present application, step S2 further comprises:

[0035] Step S21, collecting drought risk assessment literature, using multiple meta-analyses to identify drought risk assessment factors, obtaining an initial drought risk assessment factor set, counting the number of occurrences of each initial drought risk assessment factor to calculate the occurrence frequency and average frequency, removing initial drought risk assessment factors with an occurrence frequency lower than the average frequency, and obtaining a drought risk assessment factor set;

[0036] Step S22: Optimizing drought risk assessment factor weights by using the constructed drought risk assessment model for the drought risk assessment factor set;

[0037] Step S23, classifying drought risk assessment factors to obtain five types of drought risk assessment factors, namely: meteorological factors, underlying surface factors, hydrological factors, engineering factors and socio-economic factors; extracting meteorological factor raster data from the meteorological raster data of the study area, extracting hydrological data raster data of the study area from the hydrological data of the study area, and extracting underlying surface factor raster data;

[0038] A multi-objective optimization model for the project group was constructed, with the objective function being the most uniform water shortage and the least damage caused by water shortage. The multi-objective optimization model was solved to obtain the size of the engineering capacity of each project allocated to the downstream grid of the project, and the engineering factor grid data was obtained. The night light brightness data of each spatial grid was generated using the NPP-VIIRS satellite night light remote sensing image data. The weight of each spatial grid was determined based on the night light brightness data of each spatial grid, and the socioeconomic factors were rasterized based on the weight of each spatial grid.

[0039] Step S24, input the meteorological factor raster data, underlying surface factor raster data, hydrological factor raster data, engineering factor raster data and socio-economic factor raster data of the study area, as well as the drought risk assessment factor weights into the drought risk assessment model to obtain the drought risk zoning of the study area.

[0040] According to one aspect of the present application, step S3 further comprises:

[0041] Step S31, constructing a multi-objective optimization model for water resources, wherein the objective function is to minimize the area of ​​high-risk areas in drought risk zoning, minimize the area of ​​relatively high-risk areas, minimize the area of ​​medium-risk areas, and minimize the total water shortage, and the constraints are the regulation and storage and water delivery capacity of rivers and projects;

[0042] Step S32: input the hydrological data, available water volume and drought risk zoning of the study area into the multi-objective optimization model of water resources, and solve the model using the multi-objective particle swarm optimization algorithm optimized by an adaptive multi-agent system to obtain the optimal configuration plan for water resources in the study area.

[0043] According to one aspect of the present application, in step S32, the process of solving the model using a multi-objective particle swarm optimization algorithm optimized by an adaptive multi-agent system to obtain the optimal configuration plan for water resources in the study area is further as follows:

[0044] Step S32a, construct multiple proxy representative particles, classify the optimization objectives, and correspond each proxy to an optimization objective;

[0045] Step S32b, randomly generate the initial position and speed of each agent in the solution space, and evaluate their respective fitness;

[0046] Step S32c: The agents share the local optimal solution and the global optimal solution, and adjust their respective optimization strategies based on the local optimal solution and the global optimal solution, and update the speed and position;

[0047] Step S32d, record the current Pareto frontier, determine whether the new solution is better than the existing solution and update the Pareto frontier;

[0048] Step S32e, iterate repeatedly until complete convergence, and obtain the optimal configuration plan for water resources in the study area.

[0049] According to one aspect of the present application, step S4 is further:

[0050] Step S41, constructing a water project cluster optimization scheduling model, including: an optimization scheduling module and a process simulation module, and the objective function is that the deviation between each area and the optimization configuration plan is the minimum;

[0051] Step S42, inputting the optimal allocation plan of water resources in the study area into the optimization scheduling module of the water engineering cluster optimization scheduling model to obtain the initial scheduling plan of the water engineering cluster in the study area;

[0052] Step S43, input the initial scheduling plan of the water engineering cluster in the study area into the process simulation module, obtain the scheduling simulation process of the water engineering cluster in the study area, and calculate the deviation degree of each area;

[0053] Step S44: set a deviation degree threshold, extract all areas with deviation degrees greater than the threshold, adjust the corresponding gate openings based on their respective deviation flows, and obtain a revised scheduling plan;

[0054] Step S45, repeat step S43-step S44 until the deviation degree of all areas is less than the threshold, and obtain the optimal scheduling plan for the water project cluster in the study area.

[0055] According to another aspect of the present application, a water resource allocation system for coping with drought risks is provided, comprising:

[0056] at least one processor; and

[0057] a memory communicatively connected to at least one of the processors; wherein,

[0058] The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the method for coordinated allocation of water resources to cope with drought risks described in any of the above technical solutions.

[0059] Beneficial effects: A coordinated allocation of water resources to cope with drought risks is adopted to provide scientific and effective response strategies for areas with high incidence of drought risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flow chart of the present invention.

[0061] Figure 2 It is a flow chart of step S1 of the present invention.

[0062] Figure 3 It is a flow chart of step S2 of the present invention.

[0063] Figure 4 It is a flow chart of step S3 of the present invention.

[0064] Figure 5 It is a flow chart of step S4 of the present invention. DETAILED DESCRIPTION

[0065] like Figure 1According to one aspect of the present application, a method for coordinating and allocating water resources to cope with drought risks is provided, comprising the following steps:

[0066] Step S1, collecting the climate model and the monitoring data of each meteorological station in the study area, using the weighted optimized Thiessen polygon method to spatially distribute the monitoring data to obtain distribution data; reducing the scale of the climate model to the scale of the distribution data, obtaining the meteorological grid data of the study area, inputting the meteorological grid data of the study area into a pre-built grid-type hydrological model for simulation, obtaining the hydrological data of the study area, and calculating the available water volume;

[0067] Step S2, collecting drought risk assessment literature, screening drought risk assessment factors, using the constructed drought risk assessment model to optimize the weights of drought risk assessment factors, rasterizing the drought risk assessment factors and inputting them into the drought risk assessment model to obtain the drought risk zoning of the study area;

[0068] Step S3, constructing a multi-objective optimization model for water resources and taking the hydrological data, available water volume and drought risk zoning of the study area as input, solving the model using an optimization algorithm to obtain the optimal allocation plan for water resources in the study area;

[0069] Step S4: construct a water project cluster optimization scheduling model and take the optimal water resource allocation plan of the study area as input, perform a cyclic solution, and obtain the water project cluster optimization scheduling plan of the study area.

[0070] like Figure 2 As shown, according to one aspect of the present application, the step S1 is further:

[0071] Step S11, collecting monitoring data from each meteorological station in the study area, and using the weighted optimized Thiessen polygon method to spatially distribute the monitoring data to obtain the spatial distribution of the monitoring data, that is, to obtain distribution data; since there are significant differences between meteorological data in different geographical locations, spatial distribution can help us understand these differences and influencing factors, and through spatial distribution, the spatial analysis ability of data can be improved, providing more accurate information support for subsequent steps;

[0072] Therefore, in this embodiment, Thiessen polygons are used to spatially distribute the monitoring data. However, since Thiessen polygons are very sensitive to extreme values ​​in the data, some areas may be disproportionately enlarged or reduced. At the same time, when processing a large number of points, the time complexity of calculating the Thiessen polygon method is high, especially in high-dimensional space, the amount of calculation will increase sharply. Therefore, the weighted Thiessen polygon method is used to spatially distribute the monitoring data, and the spatial division, visualization results and data interpolation of the obtained monitoring data are more accurate.

[0073] Step S12: based on the spread data, downscaling the climate model using at least two downscaling methods, reducing the scale of the climate model to the scale of the spread data, so that the meteorological data of the study area matches the scale of the climate model, and obtaining the meteorological raster data of the study area based on the corresponding relationship between the meteorological data of the study area and the climate model, including: historical meteorological raster data of the study area and future meteorological raster data of the study area;

[0074] Step S13: using the meteorological grid data of the study area as input data, calling a pre-built grid-type hydrological model for solution, obtaining the hydrological data of the study area, and calculating the available water based on the hydrological data.

[0075] According to one aspect of the present application, the step S11 is further:

[0076] Step S11a, collecting monitoring data of each meteorological station in the study area, and determining the weight of each meteorological station based on the size of each meteorological station in the study area, the surrounding population density and the number of observations;

[0077] Step S11b, using GIS software to draw Thiessen polygons, each polygon corresponding to a meteorological station;

[0078] Step S11c: construct a weighted optimized Thiessen polygon method based on the weights of each meteorological station, adjust each Thiessen polygon according to the weight of each point in the Thiessen polygon method, that is, the boundary of each area, and obtain the spatial division, visualization results and data interpolation of the monitoring data. In a certain embodiment, specifically:

[0079] Collect data on average annual temperature, average seasonal temperature and precipitation over the past 20 years, and collect regional geographic information data, including: location of monitoring stations, size of monitoring stations, surrounding population density, number of observations, land use and topography;

[0080] The weight of each meteorological station is determined based on the size of each meteorological station in the study area, the surrounding population density and the number of observations.

[0081] Using the location information of the weather stations, GIS tools are used to calculate the Thiessen polygons of temperature and precipitation, where each polygon represents the influence area of ​​a weather station, showing the impact of different weather stations on the climate characteristics of the surrounding areas;

[0082] Generate a Thiessen polygon map of temperature and precipitation, and use different colors to represent the high and low temperatures and precipitation in different polygon areas.

[0083] In this embodiment, the weight-optimized Thiessen polygon method is used to spatially distribute the monitoring data, which improves the scientific nature of data processing and provides accurate data for subsequent calculations.

[0084] According to one aspect of the present application, step S12 further comprises:

[0085] Step S12a, collecting data of at least two climate models, each climate model including: historical climate data and future climate data;

[0086] Step S12b, downscaling the data of each climate mode by using at least two statistical downscaling methods and at least two dynamic downscaling methods to obtain meteorological grid data of each climate mode, and fusing the meteorological grid data of each climate mode by using Bayesian averaging method to obtain fused meteorological grid data of each climate mode, wherein the fused meteorological grid data includes: historical meteorological grid data and future meteorological grid data;

[0087] In this embodiment, the statistical downscaling methods used include: transfer function method, weather situation method and weather generator; the dynamic downscaling methods used include: proportional scaling, nested operation and physical process parameterization.

[0088] Step S12c, assign weights to each climate model based on expert opinions, calculate the meteorological raster data of the comprehensive climate model based on the weight of each climate model and the meteorological raster data, extract historical meteorological raster data and compare with the meteorological data of the study area and correct the weight of each climate model, calculate the comprehensive meteorological raster data of the final climate model, the comprehensive meteorological raster data of the final climate model is the meteorological raster data of the study area, including the historical meteorological raster data of the study area and the future meteorological raster data of the study area.

[0089] The resolution of the climate model is low, and there is often a scale mismatch problem when coupling it with the hydrological model. In order to solve the scale mismatch problem between the climate model and the hydrological model, the monitoring data needs to be downscaled. At present, the commonly used downscaling methods are mainly divided into two categories: statistical downscaling and dynamic downscaling. The statistical downscaling method downscales the monitoring data on the basis of establishing an empirical relationship between large-scale prediction factors and local small-scale prediction quantities; the dynamic downscaling method performs dynamic downscaling by constructing a high-resolution regional climate model. In this embodiment, multiple statistical downscaling and multiple dynamic downscaling methods are used simultaneously to downscale the monitoring data. After obtaining multiple groups of results, the Bayesian averaging method is used to fuse multiple groups of meteorological raster data. The obtained meteorological raster data is more accurate and has a higher degree of matching with the study area.

[0090] In this embodiment, firstly, an initial weight is given to each climate model according to expert opinions, and based on the initial weight, the meteorological grid data of all climate models are fused to obtain the initial meteorological grid data of a comprehensive climate model, including: initial historical meteorological grid data and initial future meteorological grid data, and the initial historical meteorological grid data of the comprehensive climate model corresponding to the meteorological data of the study area in the initial historical meteorological grid data is extracted and compared with the meteorological data of the study area, and the degree of deviation is calculated, and the initial weight of each climate model is optimized and adjusted based on the degree of deviation, and the weighted fusion, extraction, comparison and optimization adjustment steps are repeated until the historical meteorological grid data of the comprehensive climate model has the highest matching degree with the meteorological data of the study area in the same time period, and the corresponding climate model weight is the optimal weight, and the optimal weight of each climate model is fused with the meteorological grid data to calculate the final meteorological grid data of the climate model, which is the meteorological grid data of the study area, including the historical meteorological grid data of the study area and the future meteorological grid data of the study area.

[0091] In a certain embodiment, specifically:

[0092] Collect 100 types of climate model data, including historical meteorological raster data and future meteorological raster data;

[0093] The 100 climate model data were downscaled using the transfer function method, weather situation method, weather generator, same-proportion scaling, nested operation and physical process parameterization, and 6 sets of downscaled meteorological data were obtained for each of the 100 climate model data.

[0094] For the six groups of downscaled meteorological data corresponding to each climate model coupling, the corresponding comprehensive downscaled meteorological data are obtained;

[0095] The 100 climate models were weighted separately, and the product of each comprehensive downscaled meteorological data and its corresponding weight was calculated and summed. The obtained result was then compared with the meteorological data of the area to be studied for matching degree. The weight of each climate model was adjusted according to the matching degree, and it was iterated repeatedly until the matching degree was greater than the threshold.

[0096] The meteorological grid data finally coupled with 100 climate models is the meteorological grid data of the research area, including: historical meteorological grid data and future meteorological grid data.

[0097] Bayesian averaging is a method for improving prediction and decision-making by considering uncertainty in decision analysis and model combination. It is based on Bayesian theorem and combines different data by weighted averaging the probability of future results. In a certain embodiment, it is specifically as follows:

[0098] Identify multiple models that need to be combined;

[0099] Based on historical data, the Bayesian method is used to calculate the parameters of each model and evaluate the adaptability of the model;

[0100] A Bayesian approach was used to calculate the posterior probability of each model given the data;

[0101] The prediction results of each model are weighted based on the posterior probability to obtain a comprehensive prediction;

[0102] Check the prediction performance of the Bayesian averaging method through cross-validation, and adjust the model and weighting strategy as needed.

[0103] According to one aspect of the present application, step S13 is further:

[0104] Step S13a, collect digital elevation data, land use and soil type of the study area and input them into the ArcSWAT module to obtain hydrological response units, and use the boundary of the meteorological grid data as a constraint to combine the various hydrological response units to obtain the calculation unit of the grid SWAT model, that is, to obtain the grid hydrological model;

[0105] The hydrological response unit is the innovation of the SWAT model that is different from other hydrological models. In this embodiment, due to the existence of spatial variation, in order to reflect the characteristics of the hydrological cycle under different underlying surface conditions, the hydrological response unit is introduced. The hydrological response unit is an area with the same land use type, soil type and slope type, and is the smallest plot unit in the model operation.

[0106] Step S13b, collecting typical drought events in the study area, identifying and calculating the sensitivity of each parameter of the grid hydrological model, screening the grid hydrological model parameters based on the sensitivity, and obtaining calibrated grid hydrological model parameters;

[0107] Step S13c, extract the meteorological grid data of the study area and input it into the calibrated grid-type hydrological model to simulate and obtain the hydrological data of the study area, and calculate the available water volume based on the hydrological data.

[0108] The available water volume is the simulated runoff process minus the most basic ecological flow demand to maintain the river.

[0109] According to one aspect of the present application, step S13b further comprises:

[0110] Step S13b1, collecting typical drought events in the study area and identifying grid-type hydrological model parameters;

[0111] Step S13b2, changing the value of a certain parameter and keeping other parameters unchanged, calculating the degree of change of flow and low water level, and obtaining the sensitivity of this parameter;

[0112] Step S13b3, calculating the sensitivity of all grid-type hydrological model parameters in turn, setting a threshold, and extracting all grid-type hydrological model parameters with a sensitivity greater than the threshold for parameter calibration;

[0113] Step S13b4, sorting the collected typical drought events in the study area from small to large according to the severity of drought, and the first typical drought event in the study area has the smallest severity of drought;

[0114] Step S13b5: based on expert opinions, select the value with the largest probability of the parameter as the prior distribution of the parameter in the first typical drought event scenario in the study area, and perform Bayesian inference to obtain the posterior distribution as the prior distribution of the parameter in the next typical drought event scenario in the study area;

[0115] Step S13b6, sequentially calculating the prior distribution and posterior distribution of the parameters under each typical drought event scenario in the study area;

[0116] Step S13b7: Use the posterior distribution of the parameters under the last typical drought event scenario in the study area as the grid hydrological model parameters.

[0117] In this embodiment, drought event scenarios are sorted from small to large according to the degree of drought, and then the grid hydrological model parameters are calibrated in order. The hydrological model parameters that are sensitive to drought are inferred and corrected in sequence from small to large drought conditions. The parameters finally obtained are closer to the actual situation and avoid the situation where the parameters suddenly increase in data and cause excessive changes due to sensitivity, thereby causing the final result to have too large an error.

[0118] like Figure 3 As shown, according to one aspect of the present application, the step S2 is further:

[0119] Step S21, collecting drought risk assessment literature, using multiple meta-analyses to identify drought risk assessment factors, obtaining an initial drought risk assessment factor set, counting the number of occurrences of each initial drought risk assessment factor to calculate the occurrence frequency and average frequency, removing initial drought risk assessment factors with an occurrence frequency lower than the average frequency, and obtaining a drought risk assessment factor set;

[0120] In this embodiment, multiple meta-analysis methods are used simultaneously to identify drought risk assessment factors, which can complement each other and optimize indicator screening for different research designs and data characteristics. This comprehensiveness can reduce the bias caused by a single method and provide more comprehensive analysis results; at the same time, it lubricates the handling of heterogeneity problems, and through random effects, the results from different studies can be better integrated, the statistical power can be improved, and the scientificity and effectiveness of subsequent decision-making can be improved.

[0121] Step S22: Optimizing drought risk assessment factor weights by using the constructed drought risk assessment model for the drought risk assessment factor set;

[0122] Step S23, classifying drought risk assessment factors to obtain five types of drought risk assessment factors, namely: meteorological factors, underlying surface factors, hydrological factors, engineering factors and socio-economic factors; extracting meteorological factor raster data from the meteorological raster data of the study area, extracting hydrological data raster data of the study area from the hydrological data of the study area, and extracting underlying surface factor raster data;

[0123] A multi-objective optimization model for a project group is constructed, and the objective function is: the most uniform water shortage and the least water shortage damage. The multi-objective optimization model is solved to obtain the size of the engineering capacity of each project allocated to the downstream grid of the project, and the engineering factor raster data is obtained; the NPP-VIIRS satellite night light remote sensing image data is used to generate night light brightness data, and the weight of each spatial grid is determined based on the night light brightness data of each spatial grid, and the socio-economic factors are rasterized based on the weight of each spatial grid; the meteorological factor, underlying surface factor and hydrological factor data are obtained from step S1 and are already rasterized data. Therefore, in this embodiment, the engineering factor and socio-economic factor are rasterized.

[0124] The contrast between bright and dark areas in light images makes it a powerful tool for studying intensive human activities and their impacts. Therefore, there is a positive correlation between light intensity and socioeconomic factors. Based on the nighttime light brightness data of each spatial grid, the weight of the socioeconomic factors of each spatial grid can be determined.

[0125] Step S24, input the meteorological factor raster data, underlying surface factor raster data, hydrological factor raster data, engineering factor raster data and socio-economic factor raster data of the study area, as well as the drought risk assessment factor weights into the drought risk assessment model to obtain the drought risk zoning of the study area.

[0126] like Figure 4 As shown, according to one aspect of the present application, step S3 is further:

[0127] Step S31, constructing a multi-objective optimization model for water resources, wherein the objective function is to minimize the area of ​​high-risk areas in drought risk zoning, minimize the area of ​​relatively high-risk areas, minimize the area of ​​medium-risk areas, and minimize the total water shortage, and the constraints are the regulation and storage and water delivery capacity of rivers and projects;

[0128] Step S32: input the hydrological data, available water volume and drought risk zoning of the study area into the multi-objective optimization model of water resources, and solve the model using the multi-objective particle swarm optimization algorithm optimized by an adaptive multi-agent system to obtain the optimal configuration plan for water resources in the study area.

[0129] According to one aspect of the present application, in step S32, the process of solving the model using a multi-objective particle swarm optimization algorithm optimized by an adaptive multi-agent system to obtain the optimal configuration plan for water resources in the study area is further as follows:

[0130] Step S32a, construct multiple proxy representative particles, classify the optimization objectives, and correspond each proxy to an optimization objective;

[0131] Step S32b, randomly generate the initial position and speed of each agent in the solution space, and evaluate their respective fitness;

[0132] Step S32c: The agents share the local optimal solution and the global optimal solution, and adjust their respective optimization strategies based on the local optimal solution and the global optimal solution, and update the speed and position;

[0133] Step S32d, record the current Pareto frontier, determine whether the new solution is better than the existing solution and update the Pareto frontier;

[0134] Step S32e, iterate repeatedly until complete convergence, and obtain the optimal configuration plan for water resources in the study area.

[0135] After obtaining the hydrological data, water supply and drought risk zoning of the study area, the multi-objective optimization model of water resources constructed in advance was input and the multi-objective particle swarm optimization algorithm optimized by adaptive multi-agent system was used to solve the multi-objective optimization model of water resources, and the optimal configuration plan of water resources in the study area was obtained.

[0136] The multi-objective particle swarm optimization algorithm for adaptive multi-agent system optimization is an improved particle swarm optimization algorithm that combines adaptive mechanism and multi-agent system. This algorithm dynamically adjusts parameters through adaptive mechanism to adapt to the optimization requirements of different stages, making the algorithm more flexible in convergence speed and accuracy. When solving complex goals, the adaptive strategy can effectively balance the process of exploration and development and improve the efficiency of finding the optimal solution.

[0137] The multi-agent system decomposes the optimization process into multiple intelligent agents, each of which can independently explore the solution space, thus promoting diversity and cooperation, and effectively avoiding falling into the local optimal solution. At the same time, each agent can collaborate based on local and global information to improve the overall search efficiency. In this embodiment, a multi-objective particle swarm optimization algorithm optimized by an adaptive multi-agent system is used to solve the multi-objective optimization model of water resources. It can simultaneously optimize multiple objectives involved in water resources management, and can dynamically adjust the optimization strategy according to real-time data to cope with changes in water resources demand and availability due to meteorological factors, thereby providing decision makers with a more comprehensive information basis and finding better solutions.

[0138] In a certain embodiment, specifically:

[0139] A multi-objective optimization model was constructed, and the objective functions included: maximizing irrigation efficiency, minimizing water supply costs, and maintaining ecological flow;

[0140] Adopt the multi-objective particle swarm optimization algorithm of adaptive multi-agent system optimization to determine the parameters of the particle swarm and set the adaptive strategy and the cooperation mode of the agents;

[0141] Use multiple agents to explore different water sources separately, and quickly converge to a better solution by communicating and sharing information. At the same time, monitor the status of each water source in real time and adjust the agent strategy;

[0142] The multi-objective particle swarm optimization algorithm was used to calculate and obtain the optimal water resource allocation plan, which not only met the agricultural irrigation needs, but also effectively reduced the water supply cost while keeping the ecological flow within a safe range.

[0143] like Figure 5 As shown, according to one aspect of the present application, step S4 is further:

[0144] Step S41, constructing a water project cluster optimization scheduling model, including: an optimization scheduling module and a process simulation module, and the objective function is that the deviation between each area and the optimization configuration plan is the minimum;

[0145] The water project cluster optimization scheduling model obtains the detailed scheduling operation plan of the water project cluster through the coupling method of process simulation and optimization scheduling under the optimal configuration plan of known water resources;

[0146] The optimization scheduling module comprehensively considers the actual water release process at the head of the canal, the water inflow process in the head of the canal interval, and the flood inflow process in the canal calculated by the canal flood control forecasting model. It takes the highest matching degree of system water supply as the optimization scheduling target, scientifically considers the utilization of rainwater resources, coordinates the joint adjustment and control of the head of the canal, the water inlet gate, and the control gate, and generates the scheduling plan for the head of the canal hub / gate, each water inlet gate, and the control gate;

[0147] The process simulation module, on the one hand, provides channel flow evolution simulation for the optimization scheduling module; on the other hand, it pushes the scheduling plan to the simulation model for deduction, provides display, and pushes relevant business data to users. Users can modify the scheduling plan according to the model output results until the results meet the requirements.

[0148] Step S42, inputting the optimal allocation plan of water resources in the study area into the optimization scheduling module of the water engineering cluster optimization scheduling model to obtain the initial scheduling plan of the water engineering cluster in the study area;

[0149] In this embodiment, the actual water discharge process at the head of the canal, the water inflow process in the head of the canal interval, and the flood inflow process in the canal calculated by the canal flood control forecasting model are comprehensively considered, and the optimization scheduling target is the highest matching degree between the water supply in each area and the optimization configuration plan. The utilization of rainwater resources is scientifically considered, and the head of the canal, the water inlet gate, and the control gate are coordinated and coordinated to generate the scheduling plan of the head of the canal hub / gate, each water inlet gate, and the control gate;

[0150] In this embodiment, the deviation between the water supply of each area and the optimal configuration scheme is selected as the decision variable:

[0151]

[0152] in, is the degree of deviation under the optimal configuration scheme of the i-th water receiving sub-area in the t-th period, t=1,2,…,T, T is the scheduling period; is the actual water supply obtained by the i-th water receiving sub-area in the t-th period under the current scheduling plan;

[0153] The objective function is: the deviation between each area and the optimal configuration scheme is the lowest:

[0154]

[0155] in: is the weight of the ith water receiving sub-area in the tth period, which is used to distinguish the importance and water security priority of different areas in different periods, and N is the total number of water receiving sub-areas;

[0156] The constraints are: upper and lower limits of water level, water level constraints at the beginning of the dispatching period, water level constraints at the end of the dispatching period, water supply constraints in the receiving area, and water discharge capacity constraints;

[0157] The daily water distribution flow of each area in the water resources optimization allocation plan is mapped to the corresponding gate station project, the daily gate opening is calculated, the initial opening calculation of the irrigation area joint optimization scheduling gate station project is completed, and the initial scheduling plan is obtained.

[0158] Step S43, input the initial scheduling plan of the water engineering cluster in the study area into the process simulation module, obtain the scheduling simulation process of the water engineering cluster in the study area, and calculate the deviation degree of each area;

[0159] Step S44: set a deviation degree threshold, extract all areas with deviation degrees greater than the threshold, adjust the corresponding gate openings based on their respective deviation flows, and obtain a revised scheduling plan;

[0160] Calculate the degree of deviation for each patch and irrigation district objective function , select the degree of deviation For the area, the corresponding gate opening is adjusted according to the deviation flow to obtain the corrected scheduling plan.

[0161] Step S45, repeat step S43-step S44 until the deviation degree of all areas is less than the threshold, and obtain the optimal scheduling plan for the water project cluster in the study area.

[0162] The previous round of plans are gradually revised and optimized. The calculation is completed and the final solution is output as the termination condition, which is the optimal scheduling solution for the water project cluster in the study area.

[0163] According to another aspect of the present application, a water resource allocation system for coping with drought risks is provided, comprising:

[0164] at least one processor; and

[0165] a memory communicatively connected to at least one of the processors; wherein,

[0166] The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement any of the above-mentioned methods for coordinated allocation of water resources to cope with drought risks.

[0167] The preferred embodiments of the present invention are described in detail above; however, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all belong to the protection scope of the present invention.

Claims

1. A method for coordinating and allocating water resources to cope with drought risks, characterized in that: The steps include: Step S1, collecting the climate model and the monitoring data of each meteorological station in the study area, using the weighted optimized Thiessen polygon method to spatially distribute the monitoring data to obtain distribution data; reducing the scale of the climate model to the scale of the distribution data, obtaining the meteorological grid data of the study area, inputting the meteorological grid data of the study area into a pre-built grid-type hydrological model for simulation, obtaining the hydrological data of the study area, and calculating the available water volume; Step S2, collecting drought risk assessment literature, screening drought risk assessment factors, using the constructed drought risk assessment model to optimize the weights of drought risk assessment factors, rasterizing the drought risk assessment factors and inputting them into the drought risk assessment model to obtain the drought risk zoning of the study area; Step S3, constructing a multi-objective optimization model for water resources and taking the hydrological data, available water volume and drought risk zoning of the study area as input, solving the model using an optimization algorithm to obtain the optimal allocation plan for water resources in the study area; Step S4, constructing a water project cluster optimization scheduling model and taking the optimal allocation plan of water resources in the study area as input, performing a cyclic solution to obtain an optimal scheduling plan for the water project cluster in the study area; The step S1 is further as follows: Step S11, collecting monitoring data of each meteorological station in the study area, and using weighted optimized Thiessen polygon method to spatially distribute the monitoring data to obtain the spatial distribution of the monitoring data, that is, to obtain distribution data; Step S12: based on the spread data, downscaling the climate model using at least two downscaling methods, reducing the scale of the climate model to the scale of the spread data, so that the meteorological data of the study area matches the scale of the climate model, and obtaining the meteorological raster data of the study area based on the corresponding relationship between the meteorological data of the study area and the climate model, including: historical meteorological raster data of the study area and future meteorological raster data of the study area; Step S13: using the meteorological grid data of the study area as input data, calling a pre-built grid-type hydrological model for solving, obtaining the hydrological data of the study area, and calculating the available water volume based on the hydrological data; The step S11 is further as follows: Step S11a, collecting monitoring data of each meteorological station in the study area, and determining the weight of each meteorological station based on the size of each meteorological station in the study area, the surrounding population density and the number of observations; Step S11b, using GIS software to draw Thiessen polygons, each polygon corresponding to a meteorological station; Step S11c, constructing a weighted optimized Thiessen polygon method based on the weights of each meteorological station, adjusting each Thiessen polygon, that is, the boundary of each area, according to the weight of each point in the Thiessen polygon method, to obtain the spatial division, visualization results and data interpolation of the monitoring data; The step S12 is further as follows: Step S12a, collecting data of at least two climate models, each climate model including: historical climate data and future climate data; Step S12b, downscaling the data of each climate mode by using at least two statistical downscaling methods and at least two dynamic downscaling methods to obtain meteorological grid data of each climate mode, and fusing the meteorological grid data of each climate mode by using Bayesian averaging method to obtain fused meteorological grid data of each climate mode, wherein the fused meteorological grid data includes: historical meteorological grid data and future meteorological grid data; Step S12c, weighting each climate mode based on expert opinions, calculating the meteorological grid data of the comprehensive climate mode based on the weight of each climate mode and the meteorological grid data, extracting the historical meteorological grid data and comparing it with the meteorological data of the study area and correcting the weight of each climate mode, calculating the comprehensive meteorological grid data of the final climate mode, the comprehensive meteorological grid data of the final climate mode is the meteorological grid data of the study area, including the historical meteorological grid data of the study area and the future meteorological grid data of the study area; The step S13 is further as follows: Step S13a, collect digital elevation data, land use and soil type of the study area and input them into the ArcSWAT module to obtain hydrological response units, and use the boundary of the meteorological grid data as a constraint to combine the various hydrological response units to obtain the calculation unit of the grid SWAT model, that is, to obtain the grid hydrological model; Step S13b, collecting typical drought events in the study area, identifying and calculating the sensitivity of each parameter of the grid hydrological model, screening the grid hydrological model parameters based on the sensitivity, and obtaining calibrated grid hydrological model parameters; Step S13c, extracting meteorological grid data of the study area and inputting it into a calibrated grid-type hydrological model, simulating the hydrological data of the study area, and calculating the available water based on the hydrological data; The step S13b is further as follows: Step S13b1, collecting typical drought events in the study area and identifying grid-type hydrological model parameters; Step S13b2, changing the value of a certain parameter and keeping other parameters unchanged, calculating the degree of change of flow and low water level, and obtaining the sensitivity of this parameter; Step S13b3, calculating the sensitivity of all grid-type hydrological model parameters in turn, setting a threshold, and extracting all grid-type hydrological model parameters with a sensitivity greater than the threshold for parameter calibration; Step S13b4, sorting the collected typical drought events in the study area from small to large according to the severity of drought, and the first typical drought event in the study area has the smallest severity of drought; Step S13b5: based on expert opinions, select the value with the largest probability of the parameter as the prior distribution of the parameter in the first typical drought event scenario in the study area, and perform Bayesian inference to obtain the posterior distribution as the prior distribution of the parameter in the next typical drought event scenario in the study area; Step S13b6, sequentially calculating the prior distribution and posterior distribution of the parameters under each typical drought event scenario in the study area; Step S13b7, using the posterior distribution of the parameters under the last typical drought event scenario in the study area as the grid hydrological model parameters; The step S2 is further as follows: Step S21, collecting drought risk assessment literature, using multiple meta-analyses to identify drought risk assessment factors, obtaining an initial drought risk assessment factor set, counting the number of occurrences of each initial drought risk assessment factor to calculate the occurrence frequency and average frequency, removing initial drought risk assessment factors with an occurrence frequency lower than the average frequency, and obtaining a drought risk assessment factor set; Step S22: Optimizing drought risk assessment factor weights by using the constructed drought risk assessment model for the drought risk assessment factor set; Step S23, classifying drought risk assessment factors to obtain five types of drought risk assessment factors, namely: meteorological factors, underlying surface factors, hydrological factors, engineering factors and socio-economic factors; extracting meteorological factor raster data from the meteorological raster data of the study area, extracting hydrological data raster data of the study area from the hydrological data of the study area, and extracting underlying surface factor raster data; A multi-objective optimization model for the project group was constructed, with the objective function being the most uniform water shortage and the least damage caused by water shortage. The multi-objective optimization model was solved to obtain the size of the engineering capacity of each project allocated to the downstream grid of the project, and the engineering factor grid data was obtained. The night light brightness data of each spatial grid was generated using the NPP-VIIRS satellite night light remote sensing image data. The weight of each spatial grid was determined based on the night light brightness data of each spatial grid, and the socioeconomic factors were rasterized based on the weight of each spatial grid. Step S24, inputting the meteorological factor raster data, underlying surface factor raster data, hydrological factor raster data, engineering factor raster data and socio-economic factor raster data of the study area, as well as the drought risk assessment factor weights into the drought risk assessment model to obtain the drought risk zoning of the study area; The step S3 is further as follows: Step S31, constructing a multi-objective optimization model for water resources, wherein the objective function is to minimize the area of ​​high-risk areas in drought risk zoning, minimize the area of ​​relatively high-risk areas, minimize the area of ​​medium-risk areas, and minimize the total water shortage, and the constraints are the regulation and storage and water delivery capacity of rivers and projects; Step S32: input the hydrological data, available water volume and drought risk zoning of the study area into the multi-objective optimization model of water resources, and solve the model using the multi-objective particle swarm optimization algorithm optimized by the adaptive multi-agent system to obtain the optimal configuration plan of water resources in the study area; In step S32, the multi-objective particle swarm optimization algorithm of adaptive multi-agent system optimization is used to solve the model and obtain the optimal configuration plan of water resources in the study area, which is further: Step S32a, construct multiple proxy representative particles, classify the optimization objectives, and correspond each proxy to an optimization objective; Step S32b, randomly generate the initial position and speed of each agent in the solution space, and evaluate their respective fitness; Step S32c: The agents share the local optimal solution and the global optimal solution, and adjust their respective optimization strategies based on the local optimal solution and the global optimal solution, and update the speed and position; Step S32d, record the current Pareto frontier, determine whether the new solution is better than the existing solution and update the Pareto frontier; Step S32e, iterate repeatedly until complete convergence, and obtain the optimal allocation plan for water resources in the study area; The step S4 is further as follows: Step S41, constructing a water project cluster optimization scheduling model, including: an optimization scheduling module and a process simulation module, and the objective function is that the deviation between each area and the optimization configuration plan is the minimum; Step S42, inputting the optimal allocation plan of water resources in the study area into the optimization scheduling module of the water engineering cluster optimization scheduling model to obtain the initial scheduling plan of the water engineering cluster in the study area; Step S43, input the initial scheduling plan of the water engineering cluster in the study area into the process simulation module, obtain the scheduling simulation process of the water engineering cluster in the study area, and calculate the deviation degree of each area; Step S44: set a deviation degree threshold, extract all areas with deviation degrees greater than the threshold, adjust the corresponding gate openings based on their respective deviation flows, and obtain a revised scheduling plan; Step S45, repeat step S43-step S44 until the deviation degree of all areas is less than the threshold, and obtain the optimal scheduling plan for the water project cluster in the study area.

2. A water resource allocation system to cope with drought risks, characterized by: include: at least one processor; as well as a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement the method for coordinated allocation of water resources to cope with drought risks as described in claim 1.

Images