Drainage basin water quality simulation and scene deduction method and system based on digital twinning

Abstract

The invention discloses a drainage basin water quality simulation and scene deduction method and system based on digital twinning, and relates to the technical field of drainage basin water quality management. Constructing an initial digital twinborn body framework; continuously injecting real-time data into the initial digital twinborn body, and carrying out model training and parameter optimization operation to obtain a digital twinborn model; an interactive scene deduction mechanism is introduced, and the future change trend of the drainage basin water quality under different intervention scenes is simulated based on a digital twinborn model; calculating water quality conditions of each control section of the watershed under different intervention scenes; key abnormal parameter data are found out through a comparison judgment module in combination with the future change trend of water quality, and a water quality regulation scheme is formulated. According to the method, the consistency of the physical drainage basin and the virtual drainage basin is realized through the multi-source data, pollution diffusion is presented in a visual form through the digital twinborn model, multi-scene comparison is supported, and the accuracy of water quality simulation and decision scientificity are further improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of watershed water quality management, in particular to a watershed water quality simulation and scenario deduction method and system based on digital twinning. BACKGROUND

[0002] Watershed water quality refers to the overall condition of the physical, chemical and biological properties of water bodies within a specific watershed, which determines the ability of water bodies to support specific uses and the level of ecological health. Therefore, it is necessary to accurately manage and protect watershed water quality. However, the existing watershed water quality management faces some deficiencies, making it difficult to achieve accurate simulation and scientific deduction: 1. Traditional water quality models rely on static parameters and historical data, and cannot effectively integrate multi-source real-time monitoring data, making it difficult to accurately reflect the dynamic changes of water quality in complex watersheds; 2. Existing methods lack scenario deduction mechanisms and cannot quickly deduce water quality evolution trends under different intervention scenarios; 3. The application of digital twinning technology in watershed water quality management is still in its early stages, and multi-source data integration is not smooth, model coupling is poor, and cannot meet the needs of watershed water quality simulation and scenario deduction.

[0003] Therefore, the present application provides a watershed water quality simulation and scenario deduction method and system based on digital twinning, which can eliminate the drawbacks of existing technical solutions. SUMMARY

[0004] The present application aims to provide a watershed water quality simulation and scenario deduction method and system based on digital twinning to solve the problems of existing watershed water quality management.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions: The watershed water quality simulation and scenario deduction method based on digital twinning specifically includes the following steps: Step S1, responsible for watershed multi-source data acquisition and preprocessing operation, collecting and investigating historical statistical data, real-time monitoring data, remote sensing data of the target water area from the aspects of water area, land area and air area, and performing basic analysis, hydrological analysis and pollution analysis; Step S2, constructing an initial digital twin framework based on the basic analysis results, formulating hydrological setting conditions based on the hydrological condition results, and determining key pollutants and pollutant trends based on the pollution analysis results; Step S3, continuously injecting the real-time data into the initial digital twin to perform model training and parameter optimization operations, obtaining a digital twin model, and predicting the future trend of the water quality of the corresponding target water area; Step S4, an interactive scenario deduction mechanism is introduced to support user input of different intervention scenario parameters, including pollution source emission parameters, extreme weather parameters and treatment engineering parameters, and to simulate future change trends of the water quality of the basin under different intervention scenarios based on the digital twin model; Step S5, based on the future change trend data of the water quality of the basin under different intervention scenarios output by the digital twin model, the future change trend data is analyzed, and the water quality of each control section of the basin under different intervention scenarios is calculated; Step S6, combining the future change trend of the corresponding target water area water quality and the future change trend of the water quality of the basin under different intervention scenarios, the key abnormal parameter data is found out through the comparison judgment module, and the water quality improvement scheme is formulated according to the key abnormal parameter data.

[0006] Preferably, the basin multi-source data collection operation in step S1 specifically includes: The water quality data, hydrological data and meteorological data of the water quality monitoring station, the hydrological monitoring station and the meteorological monitoring station are obtained in real time through the monitoring network, the water quality monitoring station is used to collect pH value, dissolved oxygen, chemical oxygen demand and heavy metal concentration data, the hydrological monitoring station is used to collect water level, flow rate and runoff data, and the meteorological monitoring station is used to collect rainfall, air temperature and wind speed data; The basin terrain elevation data, vegetation coverage data and water area change data are obtained through satellite remote sensing and unmanned aerial vehicle aerial photography, and unified geographic coordinate conversion operation is performed on the data; The industrial pollution source emission data, agricultural non-point source pollution load data, urban sewage treatment plant operation data and land use data of the surrounding area of the target water area for several years are obtained through the database.

[0007] Preferably, the basic analysis in step S1 includes data integrity evaluation report, catchment range vector diagram, meteorological characteristic curve, land type and rainfall condition acquisition operation, the hydrological analysis includes basin annual average runoff, different water period flow threshold value, key section water level about flow relationship curve acquisition operation, and the pollution analysis includes basin main pollution source, each pollution source contribution rate, key pollutant and concentration change trend acquisition operation.

[0008] Preferably, in the step S2, the construction of the initial digital twin framework specifically includes: Based on the catchment range vector diagram of the basic analysis in step S1, a three-dimensional geographic space model of the basin is built, and the water body terrain and land use type of the land are embedded; The basin annual average runoff and different water period flow threshold value of the hydrological analysis are used as the initial boundary conditions of the three-dimensional geographic space model of the basin, the river channel hydraulic conduction coefficient and roughness parameter are set, so that the initial hydrological state of the three-dimensional geographic space model of the basin matches the real basin; Based on the pollution analysis result of the pollution analysis, the core parameters of the pollution migration and transformation in the three-dimensional geographical space model of the river basin are configured to form the basic framework of the initial digital twin; The initial digital twin is constructed by verifying the initial digital twin framework through historical statistical data, calculating the root mean square error between the simulation value and the measured value, and completing the construction operation when the root mean square error is less than a preset threshold.

[0009] Preferably, the step S3 specifically comprises: The real-time monitoring data is denoised by using a time series decomposition algorithm to extract the trend item, periodic item and random item of the data; The processed real-time data and historical statistical data are proportionally divided into a training set and a verification set, the initial digital twin is iteratively trained using the training set, and the model parameters are adjusted using a gradient descent algorithm; The trained model is verified for accuracy using the verification set, the root mean square error between the simulation value and the actual monitoring value is calculated, if the root mean square error exceeds a preset threshold, the parameters are adjusted and the training is continued until the error meets the requirements, and an optimized digital twin model is obtained.

[0010] Preferably, the pollution source emission parameters in the step S4 include industrial enterprise wastewater discharge amount and discharge concentration, agricultural non-point source pollution fertilizer and pesticide application intensity and livestock and poultry breeding scale, urban sewage treatment plant influent concentration and fault overflow time, the extreme weather parameters include extreme rainfall, rainfall duration, high temperature duration, wind speed and duration, and the treatment engineering parameters include the treatment scale and process type of newly built sewage treatment plant, the repair length and vegetation configuration of river ecological restoration engineering, and the area and hydraulic retention time of artificial wetland.

[0011] Preferably, the specific steps of introducing an interactive scenario deduction mechanism in the step S4 include: An intervention scenario parameter input interactive interface is built to support user selection of a preset scenario template and custom input parameters, and the preset scenario template includes sewage treatment plant fault overflow, extreme rainfall and river ecological restoration engineering; The intervention scenario parameters input by the user are received, and the parameters are verified for legality to determine whether the parameters meet the actual physical constraints of the river basin, and if not, the user is prompted to correct; The intervention scenario parameters that pass the verification are converted into an input format recognizable by the digital twin model; The scenario simulation calculation of the digital twin model is started, the model is run at a preset time step, the future change trend of the water quality of the river basin under different intervention scenarios is simulated, and the water quality simulation data of different time nodes and different waters is stored in real time.

[0012] Preferably, the workflow of the comparison and judgment module in the step S6 specifically comprises: Setting a normal water quality parameter range threshold, comparing the target water area water quality future change trend data with the normal water quality parameter range threshold, and screening out abnormal parameters that do not meet the threshold requirements; Comparing the watershed water quality future change trend data under different intervention scenarios, analyzing the differences of the same parameter under different intervention scenarios, and finding out the key sensitive parameters affecting water quality; Comprehensive abnormal parameters and key sensitive parameters, determine the key abnormal parameter data, and establish the correlation between the key abnormal parameters and water quality problems.

[0013] The watershed water quality simulation and scenario deduction system based on digital twinning is used to realize the watershed water quality simulation and scenario deduction method based on digital twinning, and the watershed water quality simulation and scenario deduction system comprises: The data acquisition module is used to collect multi-source data of the target watershed from the water area, land area and air area, including historical statistical data, real-time monitoring data and remote sensing data, and complete data preprocessing operation through abnormal value filtering and space-time alignment; The digital twinning construction module is used to build an initial digital twinning body framework according to the preprocessed data, and determine hydrological setting conditions in combination with hydrological analysis, and lock key pollutants and change trends in combination with pollution analysis; The model optimization module is used to receive the initial digital twinning body framework and real-time monitoring data, optimize model parameters through iterative training and precision verification, and obtain the final digital twinning model; The scenario deduction module is used to provide an interactive interface to support user input of intervention scenario parameters, and drive the digital twinning model to run according to a preset time step, and output watershed water quality change trend data under different intervention scenarios; The water quality analysis module is used to receive water quality future change trend data and multi-scenario simulation data, extract water quality index simulation values of each control section of the watershed, and calculate water quality under different scenarios based on water quality assessment standards; The comparison and judgment module is used to screen abnormal parameters by setting a normal water quality parameter range threshold, identify key sensitive parameters by comparing multi-scenario data, comprehensively determine key abnormal parameter data, and establish the correlation between the key abnormal parameters and water quality problems; The result output module is used to receive key abnormal parameter data, combine expert experience and historical data, and provide several water quality improvement suggestions.

[0014] Preferably, it further comprises a processor and a memory, the memory is built-in with a database, and further stores program instructions executable by the processor, and the processor calls the program instructions to realize the watershed water quality simulation and scenario deduction method.

[0015] Compared with the prior art, the beneficial effects of the present application are as follows: The present application sets up a digital-twin-based watershed water quality simulation and scenario deduction method and system, by fusing multi-source data of water area, land area and airspace, a digital twin model highly consistent with the physical watershed is constructed, the accuracy and dynamic response capability of water quality simulation are improved, the present application also introduces an interactive scenario deduction mechanism, allows users to flexibly set various intervention scenario parameters such as pollution discharge, extreme weather, treatment engineering, supports multi-scenario comparison, enables to quickly deduce the future change trend of watershed water quality under different scenarios based on the optimized digital twin model, presents the analysis results in the form of visual graphics, curves, maps and the like, and identifies key abnormal parameters through a comparison judgment module, facilitating the formulation of subsequent measures. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 It is a schematic diagram of the steps of the watershed water quality simulation and scenario deduction method of the present application.

[0017] Figure 2 It is a structural schematic diagram of the watershed water quality simulation and scenario deduction system of the present application.

[0018] Legend of reference signs: data acquisition module 10, digital twin construction module 20, model optimization module 30, scenario deduction module 40, water quality analysis module 50, comparison judgment module 60, result output module 70, processor 80, memory 90. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical scheme and advantages of the present application more clear and explicit, the present application is further described in detail below in combination with the drawings and examples. Example 1

[0020] In this embodiment, as shown in Figure 1 the digital-twin-based watershed water quality simulation and scenario deduction method specifically includes the following steps: Step S1, responsible for watershed multi-source data acquisition and preprocessing operation, collects and investigates historical statistical data, real-time monitoring data and remote sensing data of the target watershed from water area, land area and airspace, and performs basic analysis, hydrological analysis and pollution analysis; Specifically, the watershed multi-source data acquisition operation specifically includes: Real-time acquisition of water quality data, hydrological data and meteorological data of water quality monitoring stations, hydrological monitoring stations and meteorological monitoring stations through a monitoring network, the water quality monitoring stations are used to collect pH value, dissolved oxygen, chemical oxygen demand and heavy metal concentration data, the hydrological monitoring stations are used to collect water level, flow rate and runoff data, and the meteorological monitoring stations are used to collect rainfall, air temperature and wind speed data; The basin terrain elevation data, vegetation coverage data and water area change data are obtained by satellite remote sensing and unmanned aerial vehicle aerial photography, and unified geographic coordinate conversion operation is performed on the data; The industrial pollution source emission data, agricultural non-point source pollution load data, urban sewage treatment plant operation data and land use data of the target water area surrounding area for several years are obtained through the database; Specifically, the basic analysis includes data integrity assessment report, catchment range vector diagram, meteorological characteristic curve, land type and rainfall condition acquisition operation, hydrological analysis includes acquisition operation of annual average runoff of the basin, different water period flow threshold, relationship curve of key section water level with flow, pollution analysis includes acquisition operation of main pollution sources of the basin, contribution rate of each pollution source, key pollutants and concentration change trend; In the embodiment, the water area level can obtain pH value, dissolved oxygen (DO, unit: mg / L), chemical oxygen demand (COD, unit: mg / L), total phosphorus (TP, unit: mg / L) data for nearly 5 years through several water quality monitoring stations arranged in the basin, real-time data and daily average historical data for nearly 5 years of water level, flow rate and runoff are obtained through several hydrological monitoring stations, and data of influent concentration, treatment efficiency and effluent discharge of urban sewage treatment plants in the basin for nearly 5 years are synchronously collected, the land level can obtain wastewater discharge data of industrial pollution sources for nearly 5 years through the database of the local environmental protection department, obtain farmland fertilizer and pesticide application data and livestock and poultry breeding scale through the local agricultural department, and obtain land use data and change data for nearly 5 years of the basin through the local natural resources department, and the air level can obtain terrain elevation, vegetation coverage and water area change data of the basin through satellite remote sensing data, and can obtain real-time data and historical statistical data for nearly 5 years of rainfall, air temperature and wind speed through several meteorological monitoring stations in the basin; For real-time monitoring data, based on hydrological physical law, data obviously violating physical common sense is removed, then Z-score detection is used to calculate data standardization residual, abnormal data is screened out, dynamic time warping algorithm is used in time dimension, unified time reference and space coordinate reference are used, all data are stored according to the preset pattern of time-space-index, for example: the water quality data format is {2025-10-21-10:00; monitoring station A; COD: 25 mg / L, DO: 6.8 mg / L}, based on the preprocessed data, basic analysis, hydrological analysis and pollution analysis are completed; Step S2, constructing an initial digital twin framework according to the basic analysis result, formulating hydrological setting conditions according to the hydrological condition result, and determining key pollutants and pollution change trend according to the pollution analysis result; Specifically, constructing the initial digital twin framework specifically includes: Based on the catchment range vector diagram analyzed in step S1, a three-dimensional geographic space model of the catchment is built, and water body terrain and land use type are embedded; The annual average runoff of the catchment analyzed by hydrology and the flow threshold value in different water periods are used as the initial boundary conditions of the three-dimensional geographic space model of the catchment, the river water conveyance coefficient and the roughness parameter are set, so that the initial hydrological state of the three-dimensional geographic space model of the catchment matches the real catchment; Based on the pollution analysis result, the core parameters of pollutant migration and transformation are configured in the three-dimensional geographic space model of the catchment to form the basic framework of the initial digital twin; The initial digital twin framework is verified by historical statistical data, and the root mean square error of the simulation value and the measured value is calculated. When the root mean square error is less than the preset threshold, the construction of the initial digital twin is completed. In this embodiment, based on the catchment data and land use data obtained in step S1, a 1:10000 three-dimensional geographic space model is constructed, 10 catchment sub-units are divided, each catchment sub-unit is about 50km², the spatial boundary and topological relationship of each catchment sub-unit are determined, the vegetation coverage data and water body boundary data are superimposed and periodically updated to ensure that the model is consistent with the geographical features of the real catchment. According to the vegetation coverage, the area is marked as forest land, construction land and farmland, etc. Combined with the analysis result of step S1, the initial parameters such as the initial boundary conditions of the wet period, the flat water period and the dry period are set, the discharge position and discharge intensity of each pollution source are input, the migration path of pollutants from the source to the water body is constructed, and the movement process of water flow in the catchment is simulated; Step S3, continuously inject real-time data into the initial digital twin, perform model training and parameter optimization operation, obtain digital twin model, and predict future change trend of corresponding target water quality; Specifically, step S3 specifically includes: The time series decomposition algorithm is used to denoise the real-time monitoring data, and the trend item, periodic item and random item of the data are extracted. The processed real-time data and historical statistical data are proportionally divided into training set and verification set, the training set is used to iteratively train the initial digital twin, and the gradient descent algorithm is used to adjust the model parameters; The trained model is verified for accuracy using the verification set, the root mean square error of the simulation value and the actual monitoring value is calculated, if the root mean square error exceeds the preset threshold, the parameters are adjusted again and the training is continued until the error meets the requirements, and the optimized digital twin model is obtained; In this embodiment, the historical data of the past 5 years is selected from the pretreated data of step S1, 70% is taken as the training set, and 30% is taken as the validation set. The time series decomposition algorithm is used to denoise the real-time data, and the trend item, periodic item and random item are extracted. The trend item includes but is not limited to the long-term downward trend of COD concentration, the periodic item includes but is not limited to the seasonal fluctuation of TP concentration, and the random item includes but is not limited to the instantaneous concentration increase caused by heavy rain. The gradient descent algorithm is used to iteratively train the initial digital twin. First, the training set data is input into the initial digital twin, and the water quality simulation value is output. The root mean square error (RMSE) is calculated with the actual monitoring value. For the indicators with excessive root mean square error, the model parameters are adjusted, and the process of repeated training-error calculation-parameter adjustment is repeated until the root mean square error of the final validation set meets the preset threshold. The preset threshold data is determined according to expert experience and historical data, which can be adjusted according to actual environmental needs. Based on the optimized digital twin model, the future trend of water quality in the target water area is simulated. In this embodiment, the initial digital twin framework is a static three-dimensional structure model constructed based on static and historical data such as watershed geographical space, hydrology, and pollution sources, which has not been deeply integrated with real-time data. The digital twin model is a dynamic and predictable model that can evolve synchronously with the physical watershed and simulate its water quality dynamics by continuously injecting real-time monitoring data into the initial digital twin framework and optimizing the model training and parameters. Step S4, introduce an interactive scenario deduction mechanism to support user input of different intervention scenario parameters, including pollution source emission parameters, extreme weather parameters and treatment engineering parameters. Based on the digital twin model, the future trend of watershed water quality under different intervention scenarios is simulated. Specifically, the pollution source emission parameters include industrial wastewater discharge amount and discharge concentration, agricultural non-point source pollution fertilizer and pesticide application intensity and livestock and poultry breeding scale, urban sewage treatment plant influent concentration and fault overflow duration. Extreme weather parameters include extreme rainfall, rainfall duration, high temperature duration, wind speed and duration, treatment scale and process type of newly built sewage treatment plant, repair length and vegetation configuration of river ecological restoration project, and area and hydraulic retention time of constructed wetland. Specifically, the specific steps of introducing an interactive scenario deduction mechanism include: Build an intervention scenario parameter input interactive interface to support user selection of preset scenario templates and custom input parameters. The preset scenario templates include sewage treatment plant fault overflow, extreme rainfall and river ecological restoration project. Receive user input of intervention scenario parameters, and perform legality verification on the parameters to determine whether the parameters meet the actual physical constraints of the watershed. If not, prompt the user to correct. convert the intervention scenario parameters that pass the check into an input format recognizable by the digital twin model; start the scenario simulation calculation of the digital twin model, run the model at a preset time step, simulate the future change trend of the water quality in the basin under different intervention scenarios, and store the water quality simulation data of different time nodes and different waters in real time; In this embodiment, a visual interactive interface is built, which is divided into a scenario parameter input area, a template library, a result display area and the like. The scenario parameter input area supports manual input of parameters, the template library pre-stores multiple typical scenario templates including but not limited to industrial pollution overflow, extreme rainfall, ecological restoration engineering and the like. The basin water quality simulation and scenario deduction system based on digital twinning has a built-in physical constraint rule library, such as "wastewater treatment plant treatment efficiency ≤ 95%", "rainfall ≤ regional historical extreme value 150 mm / 24h" and the like. If the user inputs "treatment efficiency 100%", it will prompt "parameter exceeds the technical upper limit, please adjust it to ≤ 95%". Until the parameter is legal, the model can recognize the input format. An example can be that the "ecological restoration engineering" parameter is converted into "roughness coefficient of basin sub-unit 3 is adjusted to 0.035 and pollutant degradation coefficient is improved by 20%". The mapping relationship between the parameters and the hydrology and water quality sub-modules is established. The result display area dynamically displays the simulation results in real time and supports the user to pause and adjust during the simulation process. In this embodiment, the interactive scenario deduction mechanism is realized through a visual interactive interface, which can be in the form of a Web page application, a desktop client software or a mobile terminal. The user's interactive operation through the interface includes but is not limited to: in the scenario parameter input area, input or adjust the intervention scenario parameters, directly select a typical scenario from the preset scenario template library, load the corresponding parameter set and the like. The mechanism can drive the digital twin model to complete the scenario simulation calculation. Step S5, based on the future change trend data of the water quality in the basin under different intervention scenarios output by the digital twin model, analyze the future change trend data and calculate the water quality of each control section in the basin under different intervention scenarios. In the embodiment, key control sections in the basin are selected, including upstream inlet section, midstream industrial concentration area section, downstream outlet section, etc. The water quality assessment standards of each section are determined, for example: the upstream section executes the water quality standard: COD≤15 mg / L, TP≤0.1 mg / L, and the downstream section executes the water quality standard: COD≤20 mg / L, TP≤0.2 mg / L. The water quality of each control section in the basin under different intervention scenarios is calculated, including: determining the position and number of key control sections in the basin, determining the water quality assessment standards of each section, extracting the water quality index simulation values of each control section at different time nodes based on the water quality change trend data under different intervention scenarios output by the digital twin model, then comparing the water quality simulation values of each control section with the assessment standards, calculating the exceeding standard duration and exceeding standard multiple, and generating the water quality compliance rate and exceeding risk assessment report of each control section. Step S6, in combination with the future change trend of the water quality of the corresponding target water area and the future change trend of the water quality of the basin under different intervention scenarios, the key abnormal parameter data is found out by the comparison judgment module 60, and the water quality improvement scheme is formulated according to the key abnormal parameter data; Specifically, the working process of the comparison judgment module 60 specifically includes: The normal water quality parameter range threshold is set, the future change trend data of the target water area water quality is compared with the normal water quality parameter range threshold, and the abnormal parameters that do not meet the threshold requirements are screened out; The future change trend data of the basin water quality under different intervention scenarios are compared, the differences of the same parameter under different intervention scenarios are analyzed, and the key sensitive parameters affecting the water quality are found out; The key abnormal parameter data is determined by combining the abnormal parameters and the key sensitive parameters, and the correlation between the key abnormal parameters and the water quality problems is established; In the embodiment, the simulation results of different intervention scenarios can be combined to preferentially select the treatment project that can significantly improve the water quality compliance rate of each control section, and the water quality improvement scheme is formulated, specifically including: for the pollution source corresponding to the key abnormal parameter, measures such as limiting industrial pollution source emission, optimizing agricultural fertilization method, improving sewage treatment plant treatment efficiency, determining monitoring frequency, monitoring index and evaluation period, etc. are formulated in combination with the expert system or historical treatment cases; In this embodiment, the normal threshold range of each water quality indicator needs to be set first. This range is set according to national water quality standards or local standards. The simulated data is compared with the normal water quality parameter range threshold. Any parameter that exceeds the threshold range is marked as an "abnormal parameter" and its exceedance multiple and duration of continuous exceedance are recorded. Key sensitive parameters are identified by sensitivity coefficient. The sensitivity coefficient is the ratio of the change rate of the target indicator to the change rate of the intervention parameter. Parameters with sensitivity coefficients higher than the preset critical value are key sensitive parameters. The intersection and weight analysis of the above two types of parameters are performed. The key abnormal parameter determination includes: indicators that are both "abnormal parameters" and "key sensitive parameters", as well as single abnormal parameters that are not extremely sensitive but have a very large exceedance multiple or a very wide range of influence. Example 2

[0021] The difference from Example 1 is that, as in Example 1, Figure 2 As shown, a watershed water quality simulation and scenario extrapolation system based on digital twins is used to implement a watershed water quality simulation and scenario extrapolation method based on digital twins. The watershed water quality simulation and scenario extrapolation system includes: The data acquisition module 10 is used to collect multi-source data of the target watershed from various levels of water area, land area and air area, including historical statistical data, real-time monitoring data and remote sensing data, and to complete data preprocessing operations through outlier filtering and spatiotemporal alignment. This module can collect dynamic data from monitoring stations such as water quality monitoring stations, hydrological monitoring stations and meteorological monitoring stations in real time, and access satellite remote sensing and UAV aerial photography data. The digital twin construction module 20 is used to build an initial digital twin framework based on preprocessed data, determine hydrological setting conditions in conjunction with hydrological analysis, identify key pollutants and their changing trends in conjunction with pollution analysis, build a virtual model framework that is highly mapped to the real watershed using this module, construct a three-dimensional geospatial model of the watershed based on preprocessed data, divide the watershed into sub-units, set water flow boundary conditions and configure pollutant migration parameters. The model optimization module 30 is used to receive the initial digital twin framework and real-time monitoring data, optimize the model parameters through iterative training and accuracy verification, and obtain the final digital twin model. This module keeps the digital twin model consistent with the real watershed state, improves the accuracy of water quality simulation, and can update the model parameters regularly to adapt to changes in watershed topography, pollution source distribution, etc. The scenario simulation module 40 is used to provide an interactive interface to support users to input intervention scenario parameters and drive the digital twin model to run according to a preset time step, outputting watershed water quality change trend data under different intervention scenarios. This module realizes interactive scenario simulation, solves the problem of insufficient scenario simulation capability of traditional methods, and can pre-store typical scenario templates, which users can directly call or customize. This module can serve as the human-computer interaction interface between users and the digital twin model. The water quality analysis module 50 is used for receiving water quality future change trend data and multi-scenario simulation data, extracting water quality index simulation values of each control section of a basin, and calculating water quality conditions under different scenarios based on water quality evaluation standards; The comparison judgment module 60 is used for screening abnormal parameters by setting a normal water quality parameter range threshold, identifying key sensitive parameters by comparing multi-scenario data, comprehensively determining key abnormal parameter data, and establishing a correlation between the key abnormal parameter data and water quality problems. The result output module 70 is used for receiving key abnormal parameter data, providing several water quality improvement suggestions in combination with expert experience and historical data. As shown in the figure, Figure 2 The processor 80 and the memory 90 are further included, the memory 90 is built-in with a database and further stores program instructions executable by the processor 80, and the processor 80 calls the program instructions to realize the basin water quality simulation and scenario deduction method. To sum up, the present application solves the weak generalization problem of the traditional model by using multi-source data, realizes the consistency of the physical basin and the virtual basin, presents the pollution diffusion in a visual form through the digital twin model, supports multi-scenario comparison, and then obtains intervention measures according to the historical data and expert experience. In practical application, the method and system can improve the accuracy of water quality simulation and the scientific nature of decision-making, are suitable for sudden pollution emergency disposal, extreme precipitation and other scenarios, and have a good application prospect.

[0022] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for watershed water quality simulation and scenario deduction based on digital twinning, characterized in that, Specifically comprising the following steps: Step S1, responsible for the collection and preprocessing of multi-source data of the river basin, collecting and investigating historical statistical data, real-time monitoring data, remote sensing data of the target water area from the aspects of water area, land area and air area, and performing basic analysis, hydrological analysis and pollution analysis; Step S2, constructing an initial digital twin framework according to the basic analysis results, formulating hydrological setting conditions according to the hydrological condition results, and determining key pollutants and pollution change trends according to the pollution analysis results; Step S3, continuously injecting the real-time data into the initial digital twin to perform model training and parameter optimization operations, obtaining a digital twin model, and predicting the future change trend of the water quality of the corresponding target water area; Step S4, introducing an interactive scenario deduction mechanism to support user input of different intervention scenario parameters, including pollution source emission parameters, extreme weather parameters and treatment engineering parameters, and simulating the future change trend of the water quality of the river basin under different intervention scenarios based on the digital twin model; Step S5, based on the future change trend data of the water quality of the river basin under different intervention scenarios output by the digital twin model, analyzing the future change trend data, and calculating the water quality of each control section of the river basin under different intervention scenarios; Step S6, combining the future change trend of the water quality of the corresponding target water area and the future change trend of the water quality of the river basin under different intervention scenarios, finding out key abnormal parameter data through a comparison judgment module (60), and formulating a water quality improvement plan according to the key abnormal parameter data.

2. The digital-twin-based watershed water quality simulation and scenario deduction method according to claim 1, characterized in that, The river basin multi-source data collection operation in step S1 specifically includes: Real-time acquisition of water quality data, hydrological data and meteorological data of water quality monitoring stations, hydrological monitoring stations and meteorological monitoring stations through a monitoring network, the water quality monitoring stations are used to collect pH value, dissolved oxygen, chemical oxygen demand and heavy metal concentration data, the hydrological monitoring stations are used to collect water level, flow rate and runoff data, and the meteorological monitoring stations are used to collect rainfall, air temperature and wind speed data; Acquiring terrain elevation data, vegetation coverage data and water area change data of the river basin through satellite remote sensing and unmanned aerial vehicle aerial photography, and performing unified geographic coordinate conversion operation on the data; Acquiring industrial pollution source emission data, agricultural non-point source pollution load data, urban sewage treatment plant operation data and land use data of the surrounding area of the target water area through a database.

3. The digital-twin-based river basin water quality simulation and scenario deduction method according to claim 1, characterized in that, The basic analysis in step S1 includes data integrity evaluation report, catchment range vector diagram, meteorological feature curve, land type and rainfall condition acquisition operation, the hydrological analysis includes acquisition operation of river basin annual average runoff, different water period flow threshold value, and relationship curve of key section water level with flow, and the pollution analysis includes acquisition operation of river basin main pollution source, contribution rate of each pollution source, key pollutant and concentration change trend.

4. The digital-twin-based watershed water quality simulation and scenario deduction method according to claim 3, characterized in that, In step S2, constructing an initial digital twin framework specifically includes: Based on the catchment range vector diagram of the basic analysis in step S1, a three-dimensional geographic space model of the river basin is built, and water body terrain and land use type are embedded; The annual runoff of the hydrological analysis basin and the different water period flow threshold values are used as initial boundary conditions of the three-dimensional geographical space model of the basin, the river channel hydraulic conduction coefficient and the roughness parameter are set, and the initial hydrological state of the three-dimensional geographical space model of the basin is matched with the real basin; Based on the pollution analysis result of the pollution analysis, the core parameters of the pollutant migration and transformation are configured in the three-dimensional geographical space model of the basin to form a basic framework of the initial digital twin; The initial digital twin framework is verified through historical statistical data, the root mean square error of the simulation value and the measured value is calculated, and when the root mean square error is less than a preset threshold value, the construction operation of the initial digital twin is completed.

5. The digital-twin-based watershed water quality simulation and scenario deduction method according to claim 1, characterized in that, The step S3 specifically comprises: The time series decomposition algorithm is used to carry out noise reduction processing on the real-time monitoring data, and the trend item, periodic item and random item of the data are extracted; The processed real-time data and historical statistical data are proportionally divided into a training set and a verification set, the initial digital twin is iteratively trained using the training set, and the model parameters are adjusted using the gradient descent algorithm; The accuracy of the trained model is verified using the verification set, the root mean square error of the simulation value and the actual monitoring value is calculated, if the root mean square error exceeds a preset threshold value, the parameters are re-adjusted and the training is continued until the error meets the requirements, and an optimized digital twin model is obtained.

6. The digital-twin-based watershed water quality simulation and scenario deduction method according to claim 1, characterized in that, The pollution source emission parameters in the step S4 include industrial enterprise wastewater discharge amount and discharge concentration, agricultural non-point source pollution fertilizer and pesticide application intensity and livestock and poultry breeding scale, urban sewage treatment plant influent concentration and fault overflow time length, the extreme weather parameters include extreme rainfall, rainfall duration, high temperature duration, wind speed and duration, and the treatment engineering parameters include treatment scale and process type of newly-built sewage treatment plant, repair length and vegetation configuration of river ecological restoration project, and area and hydraulic retention time of artificial wetland.

7. The digital-twin-based watershed water quality simulation and scenario deduction method according to claim 1, characterized in that, The specific steps of introducing the interactive scenario deduction mechanism in the step S4 comprise: An intervention scenario parameter input interactive interface is built to support user selection of a preset scenario template and custom input parameters, and the preset scenario template includes sewage treatment plant fault overflow, extreme rainfall and river ecological restoration project; The intervention scenario parameters input by the user are received, and the parameters are legally verified to determine whether the parameters meet the actual physical constraints of the basin, and if not, the user is prompted to correct; The intervention scenario parameters that pass the verification are converted into an input format recognizable by the digital twin model; The scenario simulation calculation of the digital twin model is started, the model is run at a preset time step, the future change trend of the water quality of the basin under different intervention scenarios is simulated, and the water quality simulation data of different time nodes and different water areas are stored in real time.

8. The digital-twin-based watershed water quality simulation and scenario deduction method according to claim 1, characterized in that, The working process of the comparison judgment module (60) in the step S6 specifically comprises: A normal water quality parameter range threshold value is set, the future change trend data of the water quality of the target water area is compared with the normal water quality parameter range threshold value, and abnormal parameters that do not meet the threshold value requirements are screened out; The future change trend data of the water quality of the basin under different intervention scenarios are compared, the differences of the same parameters under different intervention scenarios are analyzed, and the key sensitive parameters affecting the water quality are found out; The key abnormal parameter data is determined by combining the abnormal parameters and the key sensitive parameters, and the correlation between the key abnormal parameters and the water quality problems is established.

9. A digital-twin-based river basin water quality simulation and scenario deduction system, characterized in that, The watershed water quality simulation and scenario deduction method based on digital twinning comprises a data acquisition module (10), a digital twinning construction module (20), a model optimization module (30), a scenario deduction module (40), a water quality analysis module (50), a comparison and judgment module (60), and a result output module (70). The data acquisition module (10) is used for collecting multi-source data of the target watershed from the water area, land area, and air area, including historical statistical data, real-time monitoring data, and remote sensing data, and performing data preprocessing operations through abnormal value filtering and space-time alignment. The digital twinning construction module (20) is used for constructing an initial digital twinning framework according to the preprocessed data, determining hydrological setting conditions in combination with hydrological analysis, and locking key pollutants and change trends in combination with pollution analysis. The model optimization module (30) is used for receiving the initial digital twinning framework and real-time monitoring data, optimizing model parameters through iterative training and precision verification, and obtaining a final digital twinning model. The scenario deduction module (40) is used for providing an interactive interface to support user input of intervention scenario parameters, driving the digital twinning model to run according to a preset time step, and outputting change trend data of the watershed water quality under different intervention scenarios. The water quality analysis module (50) is used for receiving the future change trend data of the water quality and multi-scenario simulation data, extracting water quality index simulation values of each control section of the watershed, and calculating the water quality under different scenarios based on water quality assessment standards. The comparison and judgment module (60) is used for setting a normal water quality parameter range threshold to screen abnormal parameters, comparing multi-scenario data to identify key sensitive parameters, comprehensively determining key abnormal parameter data, and establishing a correlation between the key abnormal parameters and water quality problems. The result output module (70) is used for receiving the key abnormal parameter data, providing several water quality improvement suggestions in combination with expert experience and historical data.

10. The digital-twin-based river basin water quality simulation and scenario deduction system according to claim 9, characterized in that, The watershed water quality simulation and scenario deduction method based on digital twinning comprises a data acquisition module (10), a digital twinning construction module (20), a model optimization module (30), a scenario deduction module (40), a water quality analysis module (50), a comparison and judgment module (60), and a result output module (70). The watershed water quality simulation and scenario deduction method based on digital twinning comprises a data acquisition module (10), a digital twinning construction module (20), a model optimization module (30), a scenario deduction module (40), a water quality analysis module (50), a comparison and judgment module (60), and a result output module (70).

Images