56 results about "Catchment hydrology" patented technology

The invention discloses a reservoir group-lake system flood control risk adaptive analysis method and system. The method comprises the following steps: acquiring a basin hydrological sequence and a pre-configured physical causal constraint matrix, and constructing a VineCopula joint distribution and initial Bayesian network conforming to physical topology according to the basin hydrological sequence and the pre-configured physical causal constraint matrix; risk sensitive weights of the sample points are calculated based on the Bayesian network, and weighted sampling density inclining towards the high-risk area is constructed; generating a self-adaptive scene set based on the density, solving an optimal scheduling scheme, feeding back a scheduling result to the Bayesian network to update parameters, carrying out closed-loop iteration of perception-sampling-scheduling-learning until risk assessment is converged, and determining a comprehensive flood control risk assessment result of the system based on the converged risk factor Bayesian network. According to the method, the problems of poor physical interpretability and scarcity of extreme risk samples of a traditional model are solved, and accurate capture and quantification of extreme risks of a complex flood control system are realized.

The invention relates to a reservoir dispatching management method and system based on digital twinning, and the method comprises the following steps: carrying out the real-time parameter collection of a target reservoir, and obtaining a reservoir characteristic parameter matrix; performing three-dimensional space modeling based on the reservoir characteristic parameter matrix and the reservoir hydrogeological environment in the target reservoir to obtain a reservoir digital twinborn model; performing drainage basin rainfall-runoff dynamic response analysis on the reservoir digital twin model to obtain a drainage basin hydrological response characteristic curve group; performing prospective water balance calculation on the watershed hydrological response characteristic curve group to obtain a multi-period reservoir flow prediction sequence; and formulating an optimal scheduling instruction set of the target reservoir based on the multi-period reservoir inflow prediction sequence, and regulating and controlling a reservoir gate control system in the target reservoir based on the optimal scheduling instruction set. And the method is difficult to adapt to hydro meteorological conditions and multi-target scheduling requirements which change in real time.

The invention discloses a hydrological forecasting method based on multi-feature combination and a Transform model. The method comprises the following steps: acquiring measured data of a target watershed hydrological station, building a physical hydrological model, acquiring output data and derivative feature data of the physical hydrological model, and integrating to obtain basic hydrological data; creating enhanced hydrological physical features, and forming a multi-dimensional original feature pool; constructing a plurality of combination strategies based on the basic hydrological data and the multi-dimensional original feature pool; capturing a long-term dependency relationship of the hydrological time sequence by using an improved Transform model; and the model performance is improved through automatic hyper-parameter optimization. According to the method, the influence of different input feature combinations on the flood forecasting precision is highlighted, and effective technical support is provided for water resource management and flood control and disaster reduction.

The invention relates to a flood and drought disaster risk assessment and deduction system and method based on slope multi-level regulation and storage, and belongs to the technical field of flood and drought disaster risk prevention and control and watershed hydrological regulation and control. The method comprises the following steps: acquiring multi-source data of a target drainage basin, performing water source refined analysis through coupling of water transfer coefficients (K1 and K2) based on a hierarchical synergistic effect of'upper closure-middle retention-lower dissipation 'of a slope, and calculating a controllable, effective and renewable water source amount; in combination with the regulation and storage capacity difference of each level, carrying out flood and drought disaster chain risk collaborative assessment, and outputting hierarchical risks and 0-10 level set risk indexes; a physical mechanism and data driving coupled scene deduction model is driven, and historical extreme event redisk and future climate scene prediction are completed; and after rendering by a visual engine, through hierarchical slope regulation and storage threshold comparison, outputting a regulation and storage measure layout optimization scheme. The system adopts a micro-service architecture and comprises a data management module, a multi-level regulation and storage collaborative analysis module, a water source analysis module, a risk evaluation module, a scene deduction module, a decision support module and a threshold comparison module. According to the method, high-resolution accurate positioning, low-error simulation and efficiency improvement of flood and drought disaster risks are realized, a decision closed loop from risk identification to optimization regulation is formed, and direct and quantitative technical support can be provided for flood and drought prevention and control engineering practice of drainage basins.

The application relates to the technical field of hydrology and water resources simulation, and discloses a SWAT watershed hydrology model task planning system, method and product, which comprises a knowledge graph construction module, an embedded index construction module, a query coding and two-stage retrieval module, a context construction module and a large language model task planning module. Through the complete planning process of knowledge graph modeling, vector index construction, two-stage retrieval, context generation and structured task output, the whole-process intelligentization from a user's natural language demand to SWAT watershed hydrology model executable task planning is realized, the SWAT file / parameter structure can be accurately matched, the professionalism, accuracy and executability of task planning are improved, and the problems of high operation threshold, large artificial workload, unstable and unexecutable output of a general large model of a traditional SWAT watershed hydrology model are solved.

The application discloses a large-scale basin deep learning flood forecasting method based on runoff lag information and relates to the technical field of hydrological simulation and flood forecasting of machine learning, and comprises the following steps: collecting basin hydrological observation site information data, basin characteristic data, basin historical hydrological data and meteorological forecasting characteristic data; dividing a large-scale basin into a plurality of sub-basins; preprocessing data, extracting runoff lag information obtained by using a convolutional neural network model, and constructing a runoff lag information database; adopting a deep learning long short-term memory model to couple the runoff lag information database, obtaining a plurality of sub-basin runoff lag information models; obtaining optimal runoff lag information parameters of the plurality of sub-basins according to a model precision evaluation index, obtaining an optimal hydrological forecasting model, and realizing accurate flood forecasting of the large-scale basin. The application makes up for the deficiency of flood forecasting under the condition of lack of hydrological data of a large-scale basin and has a strong application prospect in areas with little or no data.

The invention relates to the field of hydrological data analysis and model correction, in particular to a digital twin basin correction method based on non-stationary reservoir hydrology, which comprises the following steps: acquiring hydrological data of a basin in a predetermined season period to extract corresponding hydrological motion features; in combination with hydrological motion characteristics and the number of branch runoff paths, analyzing characteristic parameters representing hydrological connectivity stability of the watershed to divide hydrological fluctuation categories of the watershed; adaptively evaluating and correcting twinborn parameters of the drainage basin based on hydrological fluctuation categories; and determining a corresponding prediction deviation level based on the corrected prediction deviation value of the simulation prediction result and the actual result. According to the method, the adaptability and robustness of the digital twin basin model are improved while the accuracy of hydrological simulation and prediction is improved, and then the practicability and efficiency of engineering practical application are improved.

The application introduces a kind of aquatic organism protection threshold identification method based on four water coupling models, comprising: S1, constructing a watershed hydrological model, then inputting rainfall data, obtaining runoff output data;S2, constructing a river water power and water quality model, then inputting water quantity data, water level data and water quality concentration data, obtaining water quantity output data, flow velocity output data, water level output data, flow direction output data and water quality concentration output data;S3, determine the aquatic organism sampling point, and obtain the hydrology, hydrodynamic, water quality data of the corresponding aquatic organism sampling point through the watershed hydrological model and the river water power and water quality model;S4, extract the hydrology, hydrodynamic, water quality data matched with the aquatic organism sampling point from the model, construct a multi-factor coupling model based on random forest, and identify the aquatic organism protection threshold. Through the above design, the technical effect of solving the data matching in the aquatic organism protection threshold identification process is realized.

The application provides a hydrological simulation method and system based on one-way driving of a nested model in the Yellow River Basin, and belongs to the technical field of hydrological simulation; the method comprises the following steps: taking a SHUD model as a core layer for simulating the hydrological cycle of the basin, taking a MIKE hydrodynamic model as a river channel hydrodynamic simulation layer, coupling the two models to construct a nested model, and performing one-way driving based on a spatial nesting technology; a two-way data interface is constructed, data exchange between the SHUD model and the MIKE hydrodynamic model is realized, a data assimilation algorithm is introduced, historical observation data are used to calibrate the parameters of the nested model, and key parameters are optimized; and then, the hydrological process of the basin is simulated based on the optimized nested model. The application can realize high-precision simulation of the hydrological process of the basin, and further provides scientific guidance for water resource management in the Yellow River Basin.

The invention discloses a stage-by-stage dynamic flood control water level regulation and control method suitable for a hydropower station, and the method comprises the steps: obtaining watershed hydro meteorology, engineering operation and topographic geology multi-source basic data, and carrying out the Spark distributed cleaning, wavelet denoising and GAN missing value complementation preprocessing, thereby forming a standardized data set; an optimal flood season staging scheme is screened through coupling of qualitative analysis and a quantitative method; then, the safe pre-discharge amount and the effective pre-discharge time serve as sensitive factors, a flood control water level dynamic control model is built in combination with fuzzy mathematics and Copula joint distribution, and flood control water level process lines and regulation and control domain values of all flood season stages are generated; the safety of a regulation and control threshold value is verified through finite element simulation and a two-dimensional hydrodynamic model, a regulation and control scheme meeting flood control safety and efficient utilization of water resources is output, dynamic and accurate regulation and control of the flood control water level of the hydropower station are achieved, and the problems that in traditional static flood control water level dispatching, the contradiction between flood control and benefit making is prominent, the staging precision is low, and safety verification is insufficient are solved.

The application relates to a two-dimensional flood risk prediction method, device, electronic equipment, medium and product, which comprises the following steps: acquiring multi-modal data in a target river basin; obtaining simulated runoff results based on a plurality of hydrological models pre-established according to the multi-modal data in the target river basin, and establishing a hydrological-machine learning coupled model based on the simulated runoff results; simulating the river basin hydrological process under a future scenario through the hydrological-machine learning coupled model to obtain a river basin hydrological process simulation result under the future scenario, and extracting flood duration and flood volume characteristic values under a plurality of warming level time windows by using a preset optimal model based on the river basin hydrological process simulation result, obtaining a two-dimensional flood risk prediction result based on the flood duration and flood volume characteristic values under the plurality of warming level time windows, thereby solving the problems of low flood risk prediction accuracy and poor applicability in the related art, and improving the flood risk prediction accuracy and applicability.

The invention relates to the technical field of watershed hydrological simulation, and discloses a hydrological model parameter calibration method and device, equipment, a medium and a product, and the method comprises the steps: obtaining the spatial distribution characteristics of a first model parameter; determining a reasonable value range of the first model parameter according to the actual value range and the theoretical value range of the first model parameter; determining a search interval of the first model parameters according to the reasonable value range and the spatial distribution characteristics; the first model parameter is calibrated according to the search interval and the calibration database by taking the first index optimization as the target to obtain a target hydrological model, the first index represents the simulation precision of the hydrological model through the predicted hydrological data and the actual hydrological data, the predicted hydrological data is the output value of the hydrological model, and the actual hydrological data is the output value of the hydrological model. The actual hydrological data is hydrological data corresponding to the hydrological model input in the calibration database. According to the method, the spatial difference and the physical consistency of the calibration parameters can be ensured, and the stability of the finally obtained target hydrological model is improved.

The present application relates to the technical field of hydrological simulation, and more particularly to a rainfall-runoff-concentration-evolution whole-process dynamic simulation system and method. The method comprises the following steps: obtaining basin topographic information and extracting basin boundaries; dividing grid cells according to the basin boundaries and setting river nodes using the grid cells; collecting rainfall data using a rain measurement radar; calculating runoff using the rainfall data; inputting the runoff into a pre-trained distributed concentration model for concentration analysis and recording the concentration flow; inputting the concentration flow and the river nodes into a river flood evolution model to perform river evolution simulation and generate river concentration data; determining the amount of water entering the main river according to the river concentration data; generating a flow time series according to the amount of water entering the main river; and predicting the flood peak flow based on the flow time series. The present application realizes accurate simulation of the hydrological process of a basin based on hydrological simulation technology, and improves the accuracy and timeliness of flood prediction and early warning.

The application discloses a basin hydrology simulation method and system based on multi-source remote sensing, relates to the technical field of data processing, and comprises the following steps: acquiring multi-source remote sensing data by interacting with a satellite remote sensing platform and a ground measurement station; constructing a basin simulation model comprising a physical simulation layer and a digital simulation layer by analyzing the digital elevation structure and physical elements of a target basin; performing simulation solving on the digital simulation layer of the basin simulation model, mapping to the physical simulation layer, and generating a water area simulation state; and performing optional display of the water area simulation state. The application solves the technical problems of strong dependence on ground measurement stations, insufficient spatial coverage, difficulty in fusing multi-source heterogeneous remote sensing data, and low accuracy and efficiency of hydrology control equation solving under complex terrain conditions in the prior art, and achieves the technical effects of improving the spatial resolution, model accuracy, calculation efficiency and multi-source data fusion capability of basin hydrology simulation.

The invention relates to a watershed hydrological prediction method and system based on meteorological driving and causal reconstruction, and belongs to the technical field of watershed hydrological prediction. The method comprises the following steps: constructing a causal structure diagram between meteorological driving variables and between meteorological driving variables and rainfall to form an interpretable meteorological and rainfall causal diagram, explicitly converting a functional relationship in a sub-module into a series of structural equations by taking a traditional hydrological model as a construction basis, and uniformly embedding the structural equations into a causal model framework to form a meteorological and rainfall causal diagram; and a unified structural causal graph is constructed through a graph structure merging technology, and the whole process description of a causal path from external climate factors to drainage basin runoff and runoff response is realized. According to the method, a physical mechanism, statistical learning and causal reasoning are fused, the defect that a traditional model lacks explanatory ability and generalization ability when coping with non-stationary meteorological conditions is overcome, and a system modeling method with causal support is provided for flood forecasting, water resource management and extreme climate response simulation.

The invention provides a drainage basin hydrological toughness evaluation and regulation method based on a Budyko framework, and belongs to the technical field of hydrological water resource management and environment monitoring. Firstly, an improved Budyko framework and a standardized rainfall index (SPI) are adopted to carry out hydrological toughness analysis calculation and mutation analysis research on the Weihe river basin; then, the driving force influencing the hydrological toughness of the watershed and the regulation and control mechanism of the driving force are analyzed by adopting CCM-Geoderator; and finally, constructing a natural scene without human activities through a VIC model, and illuminating the influence of human climate change on hydrological toughness. The new hydrological toughness evaluation method provided by the invention can comprehensively explore the space-time evolution characteristics and driving force of the hydrological toughness in the Weihe River basin, and provides effective support for improving the ecological restorability of the basin and promoting sustainable development.

The invention provides a reservoir discharge flow continuous prediction method and device, and belongs to the field of basin hydrological simulation. The method comprises the steps of obtaining daily reservoir storage flow, total reservoir storage flow, water taking amount and timestamps corresponding to each day of a to-be-predicted time period, respectively normalizing the daily reservoir storage flow, the total reservoir storage flow of one month, six months, one year, two years and five years in the past, constructing a first input feature data set, and then inputting a reservoir water storage model to obtain a second input feature data set; obtaining a prediction result of the reservoir water storage and the reservoir water level at the beginning of each day; and constructing a second input feature data set after standardizing the reservoir inflow rates of the to-be-predicted time period every day and the first two days, the total reservoir inflow rates of the past one month, six months, one year, two years and five years, the water taking amount, the timestamps and an output result of the reservoir water storage model, and inputting the second input feature data set into the reservoir outflow model to obtain a reservoir outflow rate sequence of the to-be-predicted time period. The method obviously improves the accuracy of continuously predicting the reservoir outlet flow, and is suitable for the condition of lacking continuous water storage observation data.

This invention discloses a flood simulation method and system that couples the dynamic characteristics of the underlying surface and the river channel, relating to the fields of watershed hydrological simulation and disaster prevention and mitigation. The method includes: multi-source data acquisition and preprocessing; prediction of underlying surface evolution and hydrophysical parameterization; inversion of river channel dynamic roughness based on water-sediment coupling simulation and inverse Manning's formula; coupling the future runoff generation parameter field and dynamic roughness sequence to a distributed hydrological model for simulation; and quantitatively decoupling the contributions of underlying surface and river channel changes to flood characteristics through multi-scenario comparative experiments. This invention overcomes the shortcomings of existing technologies, such as the disconnect between underlying surface and river channel processes and the static nature of roughness parameters, achieving full-chain dynamic coupling simulation. It significantly improves the physical realism and prediction accuracy of flood simulation under complex environments, and is particularly suitable for data-scarce high-altitude watersheds.

The invention belongs to the technical field of water conservancy projects, and particularly relates to a data-free area runoff sediment transport simulation method and system based on an SWAT model. The method comprises the steps that a data-free small watershed is selected as a target watershed, and a system collects comprehensive geographic information data of the target watershed; based on a hydrological similarity principle, selecting data basins with similar hydrological attributes as reference basins; constructing, calibrating and verifying a reference basin SWAT hydrological model; and transplanting the verified model parameters to the target drainage basin to realize long-term sequence simulation of annual runoff volume and sediment transport volume of the target drainage basin. Compared with the prior art, the parameter transplantation method based on the SWAT model realizes runoff-sediment transport synchronous simulation under the condition of lack of actual measurement data, data requirements and calculation cost are reduced, and simulation results can provide scientific basis for regional water resource total amount analysis, drainage basin treatment measure evaluation and hydraulic engineering design; the method is of great significance in improving the scientific management capability and strategic planning level of water resources in areas without data.

The application provides a flow prediction method based on a Transformer-mixed expert module, acquires multi-source hydrological feature data and timestamp data of a target area and performs preprocessing, inputs the preprocessed multi-source hydrological feature data into independent feature encoders to obtain feature vectors in a unified hidden dimension, converts the timestamp into a time index and performs time embedding, splices each feature vector and the time embedding vector and inputs the spliced feature vector into an attention fusion module to obtain a fusion feature vector, and then performs dynamic calculation through a mixed expert module containing a gating network and an expert network set to output a flow prediction value. The application can effectively fuse multi-source heterogeneous features, dynamically adapt to different hydrological scenes, balance prediction accuracy, calculation efficiency and cross-basin generalization ability, and is suitable for the scenes of basin hydrological monitoring and real-time flow prediction.

The application provides a kind of based on multiple model's watershed flow prediction method, terminal and storage medium, the method includes: obtaining and integrating the hydrological data of each monitoring point in the upstream of watershed, obtain the watershed hydrological data;Pretreatment watershed hydrological data and input to Xinanjiang model to obtain runoff flow sequence, then by maskingin model to obtain prior flow sequence;Collect and pretreatment watershed actual time flow, obtain measured flow sequence;With the residual value between prior flow sequence and measured flow sequence as the learning goal of long short-term memory neural network, output residual flow sequence;Based on prior flow sequence and residual flow sequence, use hybrid density network to generate time-varying probability distribution parameters of the flow of watershed in future period, generate flow prediction interval that meets pre-set reliability based on time-varying probability distribution parameters, can realize the watershed flow prediction of explainable, high precision and with uncertainty quantification, to improve the accuracy and reliability of flood forecast.

The invention discloses a method for quantitatively evaluating the influence of the pattern change of a watershed ecosystem on a wading function. The method comprises the following steps: constructing an SWAT model; performing watershed discretization and HRU (hydrological response unit) division, parameter setting and model operation, and calibrating and verifying the model; replacing ecological system types of forests, grassland and wetlands in the land utilization data with bare lands, and operating the watershed hydrological model again; carrying out watershed hydrological process simulation under different land utilization scenes to obtain corresponding rainfall, evapotranspiration, runoff depth, water yield and river runoff; the basin water source conservation amount is calculated; the water yield is obtained; a standardized runoff index is calculated through the runoff volume of the river channel, and flood / drought occurrence months are counted. The method provided by the invention can quantitatively evaluate the influence of the pattern change of the main natural ecosystem on the watershed water source conservation effect, the water production effect, the flood regulation and storage and the drought relief effect, and provides theoretical support for scientific management, protection and recovery measures of the watershed ecosystem.

The invention provides a Yellow River basin hydrological simulation method and system based on nested model one-way driving, and belongs to the technical field of hydrological simulation. Comprising the steps that an SHUD model serves as a core layer of watershed hydrological cycle simulation, an MIKE hydrodynamic model serves as a river channel hydrodynamic simulation layer, a nested model is constructed through coupling between the two models, and one-way driving is conducted based on the space nested technology; a bidirectional data interface is constructed, data exchange between the SHUD model and the MIKE hydrodynamic model is achieved, meanwhile, a data assimilation algorithm is introduced, historical observation data is utilized to calibrate nested model parameters, and key parameters are optimized; and then simulating the watershed hydrological process based on the optimized nested model. According to the invention, high-precision simulation of the watershed hydrological process can be realized, and scientific guidance is provided for water resource management of the Yellow River watershed.

This invention relates to the field of dynamic simulation of river flow modulus, and discloses a method for dynamic simulation of flow modulus in mountainous rivers. First, it acquires data on hydrological and meteorological elements, topographic river networks, soil properties, land cover, and glacier characteristics. Based on this data, it divides the river into sub-basins and contour zones, constructs a watershed hydrological model, and calculates the flow of each sub-basin along the main river channel. Based on satellite remote sensing data and the distribution location of the sub-basins along the main river channel, it calculates the average gradient of the sub-basins. Based on the flow and average gradient of each sub-basin along the main river channel, it calculates the flow modulus of all sub-basins along the river channel in the hydrological model under different seasons using the flow modulus calculation formula. This invention solves the problem in existing technologies where it is impossible to conduct on-site measurements of flow and gradient in mountainous rivers, thus making it impossible to draw dynamic flow modulus data for mountainous rivers.

The application discloses a kind of quantitative calculation method, device, equipment, medium and product of watershed runoff variation attribution, it is related to runoff variation quantitative analysis field, the method includes: obtaining the daily runoff data of the watershed outlet in research period in the watershed to be analyzed;Year runoff of watershed is calculated based on daily runoff data, and the time series of year runoff of watershed is obtained;Based on the time series of year runoff of watershed, the mutation point of year runoff is obtained;According to the mutation point, the research period is divided into benchmark period and influence period;The runoff data of benchmark period is used to calibrate SWAT model, and the model calibrated in benchmark period is obtained;Based on the model calibrated in benchmark period, various data of benchmark period and influence period are used to calculate the contribution value of water conservancy facilities operation, human social and economic activity water intake, human social and economic activity discharge and land use type change to watershed runoff variation.The application improves the simulation accuracy of watershed hydrological model to watershed runoff variation.

The invention relates to the technical field of ecological value accounting, and discloses a water and soil conservation ecological product value accounting method and system.The water and soil conservation ecological product value accounting method comprises the following steps that S101, a normalized sand production intensity difference value is calculated; step S102, generating a slope irradiation coupling flux; step S103, calculating to obtain a propagation graph weight matrix; step S104, calculating mismatching potential energy; step S105, mapping the mismatch potential energy into an audit weight; and step S106, summarizing to obtain an accounting result. According to the method, on the basis of the influence rule of the slope direction on the earth surface process and the watershed hydrological connectivity characteristic, data is split according to the slope direction, the correlation difference of illumination vegetation sand production is captured, closed forward propagation is combined to quantify mismatch accumulative influence, linear offset of terrain illumination on vegetation indexes is eliminated, mismatch risk differences are adapted through auditing weights, and the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction of the slope direction; and the vegetation index better fits the real coverage state.

The invention discloses a series flow prediction method based on a Xinanjiang model and a long short-term memory network, and the method comprises the steps: firstly obtaining the hydrological data of a drainage basin, and employing the Xinanjiang model to calculate and obtain the hydrological intermediate parameters, such as soil water storage, net rainfall and simulated flow; as an input feature of an LSTM model, constructing an XAJ-LSTM series prediction architecture; further adopting an Optuna optimization algorithm to carry out adaptive optimization on hyper-parameters such as the number of neurons, the dropout rate and the learning rate of the LSTM model; and finally, evaluating the precision of the model through a plurality of indexes such as a Nash efficiency coefficient and a root-mean-square error, and verifying the effectiveness of the model in combination with a visualization means. According to the method, the physical mechanism advantages of the Xinanjiang model and the capability of the LSTM in the aspect of time series data fitting are fused, the precision and stability of flow prediction are remarkably improved, particularly, the performance in the aspect of flood extreme value capture is excellent, and the method can be widely applied to basin flood forecasting, water resource scheduling and other actual scenes.

The invention discloses a spatial heterogeneity time-varying mixed runoff production calculation method based on infiltration capacity dynamic evolution. The method comprises the following steps: step 1, partitioning spatial heterogeneity of a target watershed; step 2, constructing a runoff production characteristic curve of the small watershed unit; step 3, carrying out runoff production parameter functional representation based on underlying surface features; and 4, spatial heterogeneity time-varying mixed runoff production calculation is carried out. The method systematically solves the problems that according to an existing runoff production method, key parameter time-varying characteristics and spatial heterogeneity characteristics are insufficient, coupling of different runoff production mechanisms is weak and the like, and the physical mechanism rationality of runoff production calculation of the watershed under the complex hydro meteorology and underlying surface conditions is remarkably improved; and a more reliable technical support is provided for basin hydrological simulation.

The present application belongs to the technical field of hydrological forecasting, and in particular to a physical constraint basin hydrological real-time forecasting method based on semantic causal reasoning. In view of the semantic gap of multi-source heterogeneous data and the missing problem of physical mechanism existing in the prior art, the following scheme is proposed: first, a knowledge graph within a basin is constructed based on water conservancy text data; second, a semantic causal reasoning is performed by using a large language model combined with the knowledge graph to dynamically identify key disaster-causing factors and generate an input feature set; then, a semantic constraint prediction network is constructed, which realizes the spatio-temporal aggregation of upstream station features through a knowledge graph attention mechanism, and modifies the forget gate by introducing a rainfall attenuation coefficient and adds a water balance constraint term in the loss function to explicitly inject the hydrological physical law into the neural network; finally, the model is trained by using historical data and features are completed and dynamically optimized in the real-time data stream combined with the knowledge graph to output the water level prediction value at the future time.

The application discloses a regional surface water resource prediction simulation and optimization configuration method, which comprises the following steps: step one, collecting data for a research area, and constructing a watershed hydrological simulation model, namely, a SWAT model; step two, collecting future meteorological data, and inputting the data into the calibrated and verified SWAT model to predict and simulate the regional surface water resource quantity under the background of climate change in the research area; step three, integrating the regional surface water resource prediction simulation model under the background of climate change into an interval multi-objective optimization model; and step four, solving the interval multi-objective optimization configuration model of the regional surface water resource under the background of climate change; the beneficial effect is that, on the basis of predicting the future regional surface water resource quantity, the water resource configuration scheme of a reservoir under different climate change scenarios, which meets the power generation, water supply and water abandonment targets, can be obtained.