Intelligent water affair dynamic monitoring system and monitoring method based on digital twinning

Abstract

The invention belongs to the technical field of water affair management, and discloses an intelligent water affair dynamic monitoring system and monitoring method based on digital twinning. Comprising the steps of collecting multi-modal water affair data in a preset water affair system and performing data cleaning to obtain standard multi-modal water affair data; adaptive weighted fusion is carried out on the standard multi-modal water affair data, and a multi-modal water affair data matrix is constructed; constructing a digital twin water affair architecture and constructing a simulated water affair management scene; constructing an ecological evaluation model to perform water quality monitoring on the simulated water affair management scene, and outputting water quality evaluation data; constructing a water affair management partition topological structure to perform water resource cross-region scheduling on the simulated water affair management scene, and outputting a water affair partition management strategy; constructing a water affair system assembly operation scene to carry out system fault monitoring, and outputting a system maintenance log; collecting real-time water affair data and combining with a system maintenance log to perform parameter updating on the digital twin water affair architecture to obtain an optimized digital twin water affair architecture; and efficient water affair management is realized.

Description

Technical Field

[0001] The present invention relates to the field of water management technology, and more specifically, to a smart water dynamic monitoring system and monitoring method based on digital twins. Background Art

[0002] With the continuous expansion of urban scale and the continuous increase in the complexity of water resource management, traditional water systems are faced with problems such as lagging water quality monitoring, inefficient water resource allocation, and difficulty in real-time monitoring of facility operating status. With the development of the Internet of Things and artificial intelligence, smart water systems are gradually emerging, but certain challenges still exist in today's water management technology.

[0003] Traditional smart water systems have weak capabilities for integrating multimodal water data and are unable to accurately capture the relationships between multi-source data, which easily leads to the phenomenon of data silos. Furthermore, in terms of simulation modeling, traditional smart water systems mostly use static rule models for scenario simulation, which makes it difficult to express the evolution mechanism of various types of data in complex scenarios and lacks deductive simulation capabilities, resulting in poor simulation effects and low simulation accuracy. In terms of water quality testing, traditional water systems are insensitive to abnormal water quality data, resulting in low water quality testing accuracy. Furthermore, in terms of water resource scheduling, existing smart water systems mostly rely on manual experience in the database to guide scheduling strategies and are unable to dynamically optimize scheduling strategies based on regional topological relationships and parameters such as water quality. Furthermore, in terms of fault monitoring, traditional smart water systems rely on simple parameter detection and lack the ability to perceive and predict potential faults, resulting in low maintenance effects.

[0004] In view of this, the present invention proposes a smart water dynamic monitoring system and monitoring method based on digital twins to solve the above problems. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned objectives, the present invention provides the following technical solution: a smart water affairs dynamic monitoring method based on digital twins, comprising: S1. Collect multimodal water service data from a preset water service system, clean the multimodal water service data, and obtain standard multimodal water service data; S2. Perform adaptive weighted fusion on standard multimodal water service data to construct a multimodal water service data matrix; S3. Build a digital twin water architecture based on a multimodal water data matrix and use the digital twin water architecture to construct simulated water management scenarios. S4. Build an ecological assessment model and embed it into the digital twin water management architecture. Use the ecological assessment model to monitor water quality in simulated water management scenarios and output water quality assessment data. S5. Construct a water management zoning topology. Based on the zoning topology and water quality assessment data, perform cross-regional water resource scheduling for simulated water management scenarios. Output a water management strategy for the zoning area and feed it back into the digital twin water architecture. S6. Construct water system component operation scenarios based on the digital twin water architecture, perform system fault monitoring on these water system component operation scenarios, and output system maintenance logs; S7. Collect real-time water service data, update the parameters of the digital twin water service architecture based on the system maintenance log and real-time water service data to obtain an optimized digital twin water service architecture, and deploy the optimized digital twin water service architecture to the preset water service system.

[0006] Furthermore, the method of cleaning the multimodal water service data includes: Perform time alignment on multimodal water service data to obtain time-aligned water service data; construct an anomaly detection model to process outliers in the time-aligned water service data to generate normal water service data; use a normalization function to normalize the normal water service data to obtain standard multimodal water service data; Methods for adaptive weighted fusion of standard multimodal water service data include: Perform modal classification on standard multimodal water affairs data to obtain standard water affairs data of any modality; calculate the entropy value of each data point in the input standard water affairs data of each modality, calculate the mutual information of each data point in the standard water affairs data of each modality based on the entropy value, and map the mutual information to the feature space to form a mutual information feature vector; deploy a fusion unit, receive the mutual information feature vector of the standard water affairs data of each modality, and construct a mutual information feature matrix; set the initial feature weights, use the activation function to calculate the weights of the mutual information feature matrix, and output the corrected weights; weight the standard water affairs data of each modality based on the corrected weights to obtain a multimodal water affairs data matrix.

[0007] Furthermore, the method of constructing a digital twin water service architecture based on a multimodal water service data matrix includes: Build a communication protocol library, calculate the system network bandwidth in real time, set the network bandwidth threshold, and dynamically select the optimal communication protocol based on the system network bandwidth and the network bandwidth threshold, while receiving a multimodal water data matrix as input data; construct a high-precision three-dimensional water network model and fluid simulation model of the preset water system; build a scenario simulation engine, and simultaneously build a water knowledge database and associate it with the scenario simulation engine; integrate the high-precision three-dimensional water network model, fluid simulation model, and scenario simulation engine to generate a digital twin water architecture; Ways to use the digital twin water architecture to construct simulated water management scenarios include: Perform physical structural modeling of the preset water system to generate a high-precision three-dimensional water network model, and use a fluid simulation model to simulate water flow. Label the water facilities in the high-precision three-dimensional water network model, and simultaneously label the parameters of the high-precision three-dimensional water network model based on a multimodal water data matrix. Extract the water management model from the high-precision three-dimensional water network model, and encode the water management model to obtain a water management model code. A management model database is constructed based on all water management model codes, and the management model database is associated with a high-precision three-dimensional water network model; a scenario simulation engine is used to generate simulated water management scenarios based on the high-precision three-dimensional water network model, and constraints for the simulated water management scenarios are set based on the water knowledge database.

[0008] Furthermore, the method of using the ecological assessment model to monitor water quality in a simulated water management scenario includes: Based on the ecological assessment model, a pollutant diffusion path for a simulated water management scenario is constructed, and a pollutant diffusion path map is drawn based on the pollutant diffusion path. Water quality parameters for the simulated water management scenario are extracted, and a water quality thermodynamic map is drawn based on the water quality parameters. The water quality thermodynamic map is then superimposed on the pollutant diffusion path map to obtain visualized water quality assessment data. A four-layer water quality assessment system is designed based on the water affairs knowledge database, a water quality assessment function is constructed based on the four-layer water quality assessment system, and the water quality assessment score of the simulated water affairs management scenario is calculated using the water quality assessment function; a water quality assessment grading threshold is set, and the water quality assessment grade of the simulated water affairs management scenario is determined based on the water quality assessment grading threshold and the water quality assessment score; the ecological assessment model is associated with the preset weather forecast system, and a water quality change trend map is constructed based on the weather forecast data of the preset weather forecast system; water quality assessment visualization data, water quality assessment grades and water quality change trend maps are integrated to obtain water quality assessment data for the simulated water affairs management scenario.

[0009] Furthermore, the method of constructing the water management zone topology structure includes: Extract the topology information of the preset water system in the simulated water management scenario, and construct a water topology map of the simulated water management scenario based on the topology information; Extract node information from each node in the water management topology map and construct a node information feature matrix based on the node information; construct an adaptive range window and dynamically adjust the window radius, while numbering each adaptive range window; construct a regional division structure to extract the water state parameters in each adaptive range window and group all water state parameters based on the number; divide the simulated water management scenario into regions based on the node information feature matrix and a set of water state parameters corresponding to the node to obtain an initial set of divided regions; Construct a balance constraint function and judge the rationality of the initial partition area set based on the balance constraint function; if it is unreasonable, readjust the boundaries of each initial partition area to obtain the optimized partition area set; integrate the water management topology map, regional division structure and balance constraint function to construct the water management zoning topology structure; The methods for dividing the simulated water management scenario into regions based on the node information feature matrix and a set of water state parameters corresponding to the node include: Node feature vectors are extracted from the node information feature matrix, and the node feature vectors are feature-concatenated with a set of water body state parameters corresponding to the node to obtain a partition input vector; all partition input vectors are mapped to the constructed input parameter data space to obtain a unified fusion vector; all unified fusion vectors are classified and processed to obtain a preliminary grouping vector set; a regional partitioning model is constructed, and the topological relationship in the water service topology map is extracted and associated with the regional partitioning model; the preliminary grouping vector set is partitioned using the regional partitioning model to output an initial partitioning region set; a regional topology factor is constructed, and the partitioning processing degree of the regional partitioning model is adjusted based on the regional topology factor.

[0010] Furthermore, the method of performing cross-regional water resource scheduling for the simulated water management scenario based on the water management zoning topology and water quality assessment data includes: Two types of inter-regional water resource dispatching conditions are set; the first type of inter-regional water resource dispatching condition is: detecting the water quality assessment data of each optimized division area in the simulated water management scenario, comparing the water quality assessment visualization data with the preset water body visualization database, and triggering inter-regional water resource dispatching if it meets the pollution water quality chart in the water body visualization database, and the water quality assessment level is lower than the preset water body assessment threshold or the water quality change amplitude in the water quality change trend map is greater than the preset change amplitude threshold; the second type of inter-regional water resource dispatching condition is: collecting the water body state parameters of each optimized division area in the simulated water management scenario, constructing a supply and demand load function based on the water body state parameters, and using the supply and demand load function to calculate the supply and demand load of each optimized division area, and triggering water resource inter-regional dispatching if it is greater than or equal to the preset load threshold; Based on the topological structure of water management zoning, water sub-regions with associated paths in the optimized divided areas that require cross-regional water resource scheduling are screened. At the same time, the connection relationship between the optimized divided areas and the water sub-regions with associated paths is extracted to construct an initial scheduling path dataset; a cost function is constructed, and the cost consumption score of each initial scheduling path in the initial scheduling path dataset is calculated using the cost function; the initial scheduling path with the smallest cost consumption score is used as the regional scheduling path, and the digital twin water architecture is used to predict the changes in water state parameters of the regional scheduling path, and output the water zoning management strategy.

[0011] Furthermore, the method of constructing the water system component operation scenario based on the digital twin water architecture includes: Extract the equipment status parameters of each water facility in each optimized partitioned area in the simulated water management scenario, use the digital twin water architecture to create a digital twin object for each water facility based on the equipment status parameters, and perform partition mapping on the digital twin objects of all water facilities based on the topological relationships between water facilities; Call the management mode database and extract the operating status of each water facility in any optimized divided area in the simulated water management scenario; encode the operating status of each water facility, and construct a facility operation event sequence based on the event code and timestamp; build an event-driven operation scenario based on the facility operation event sequence of all water facilities, and associate the event-driven operation scenario with the fluid simulation model and water topology map in the digital twin water architecture to construct the water system component operation scenario.

[0012] Furthermore, the method of performing system fault monitoring on the operation scenario of the water system components includes: Extract the equipment status parameters of water facilities in the water system component operation scenario; collect historical facility operation data to construct a system fault map; extract the water body state parameters in the optimized division area of ​​each water facility in the water system component operation scenario, and combine the water body state parameters and equipment state parameters to obtain fault operation data; map the fault operation data of the water facility to the vector space to form a fault operation vector, and integrate all fault operation vectors to construct an equipment fault matrix; build a fault monitoring model, and embed the system fault map into the fault monitoring model, use the fault monitoring model to perform fault detection on the equipment fault matrix, and output the fault vector and fault risk index; set the fault classification threshold, determine the fault level of each water facility based on the fault classification threshold and fault risk index, and determine the fault type based on the fault vector; Build a prediction model, design facility operation and maintenance rules based on the water affairs knowledge database, and combine the facility operation and maintenance rules with the prediction model; use the prediction model to predict faults in the equipment fault matrix and output the fault prediction results; draw a fault propagation path diagram based on the fault prediction results and the topological relationship between water affairs facilities; The fault level, fault type and fault propagation path diagram are integrated to obtain a fault information set; a system maintenance plan is generated based on the fault information set and the system fault map, and the system maintenance plan is encoded to obtain a system maintenance log.

[0013] Furthermore, the method of updating parameters of the digital twin water service architecture based on system maintenance logs and real-time water service data includes: A data buffer is constructed within the digital twin water utility architecture, and is used to store real-time water utility data and system maintenance logs. A log-water utility parameter matrix is ​​constructed based on the system maintenance logs and real-time water utility data, and simulation scenarios are generated based on the log-water utility parameter matrix using the digital twin water utility architecture. Simulation parameters corresponding to the parameters in the log-water utility parameter matrix are extracted from the simulation scenario, and the simulation parameters are compared with the log-water utility parameter matrix. Based on the comparison results, an adjustable parameter dataset is constructed. Extract the architectural parameters of the digital twin water architecture, establish a mapping relationship between the adjustable parameter data set and the architectural parameters, and construct a parameter space based on the adjustable parameter data set and the architectural parameters; construct an optimization objective function, and use a preset optimization algorithm to search for optimal parameters in the parameter space based on the optimization objective function, and output the optimal parameter combination; apply the optimal parameter combination to the digital twin water architecture, use the digital twin water architecture to re-simulate the scene, and perform a scene accuracy comparison. If the scene accuracy is higher than the preset scene accuracy threshold, the digital twin water architecture at this time will be used as the optimized digital twin water architecture, and the optimal parameter combination of the optimized digital twin water architecture will be stored in the preset parameter database.

[0014] The digital twin-based smart water affairs dynamic monitoring system is used to implement a digital twin-based smart water affairs dynamic monitoring method, including: A data acquisition module is used to collect multimodal water service data from a preset water service system, clean the multimodal water service data, and obtain standard multimodal water service data; Data fusion module: builds a data fusion framework to perform adaptive weighted fusion on standard multimodal water service data and construct a multimodal water service data matrix; The digital twin simulation module builds a digital twin water architecture based on a multimodal water data matrix and uses the digital twin water architecture to construct simulated water management scenarios. The water quality monitoring module builds an ecological assessment model and embeds it into the digital twin water management architecture. It uses the ecological assessment model to monitor water quality in simulated water management scenarios and outputs water quality assessment data. The water management scheduling module builds a water management zoning topology, performs cross-regional water resource scheduling for simulated water management scenarios based on the zoning topology and water quality assessment data, outputs water management zoning strategies, and feeds these strategies back to the digital twin water architecture. The fault monitoring module constructs water system component operation scenarios based on the digital twin water architecture, monitors system faults in the water system component operation scenarios, and outputs system maintenance logs; The parameter optimization module is used to collect real-time water service data, update the parameters of the digital twin water service architecture based on the system maintenance log and real-time water service data to obtain an optimized digital twin water service architecture, and deploy the optimized digital twin water service architecture to the preset water service system; each module is connected by wired and / or wireless means.

[0015] The technical effects and advantages of the digital twin-based smart water affairs dynamic monitoring system and monitoring method of the present invention are as follows: By integrating adaptive weighted processing of multimodal water data, building a digital twin water architecture and 3D scenario simulation, embedding a water quality ecological assessment model, and designing a cross-regional water resource scheduling mechanism, coupled with system component operation monitoring and intelligent fault prediction models, this approach achieves intelligent, precise, and dynamic management of all elements, all time periods, and all processes of the water system. This digital twin-based smart water dynamic monitoring method not only enhances the real-time perception of key information such as water quality changes, equipment failures, and water pressure fluctuations, but also provides intelligent prediction and scheduling response capabilities for water management strategies. This effectively improves the operational safety, control efficiency, and maintenance autonomy of the overall water system, supporting the transformation of water management from traditional passive response to active prediction and self-optimization, significantly enhancing the intelligence level of water management and emergency decision-making capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 Schematic diagram of the digital twin-based smart water affairs dynamic monitoring method of the present invention; Figure 2 Schematic diagram of the digital twin-based smart water dynamic monitoring system of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0018] Example 1 See also Figure 1 As shown, the digital twin-based smart water affairs dynamic monitoring method described in this embodiment includes: S1. Collect multimodal water service data from a preset water service system, clean the multimodal water service data, and obtain standard multimodal water service data; S2. Perform adaptive weighted fusion on standard multimodal water service data to construct a multimodal water service data matrix; S3. Build a digital twin water architecture based on a multimodal water data matrix and use the digital twin water architecture to construct simulated water management scenarios. S4. Build an ecological assessment model and embed it into the digital twin water management architecture. Use the ecological assessment model to monitor water quality in simulated water management scenarios and output water quality assessment data. S5. Construct a water management zoning topology. Based on the zoning topology and water quality assessment data, perform cross-regional water resource scheduling for simulated water management scenarios. Output a water management strategy for the zoning area and feed it back into the digital twin water architecture. S6. Construct water system component operation scenarios based on the digital twin water architecture, perform system fault monitoring on these water system component operation scenarios, and output system maintenance logs; S7. Collect real-time water service data, update the parameters of the digital twin water service architecture based on the system maintenance log and real-time water service data to obtain an optimized digital twin water service architecture, and deploy the optimized digital twin water service architecture to the preset water service system.

[0019] The existing water system is used to manage and optimize the development, utilization and protection of water resources. Water sources, pipelines, pumping stations and sewage treatment sites are all part of the water system. IoT sensors deployed in the preset water system collect data such as water level, flow, water quality, chemical composition and equipment status, and integrate them to obtain multimodal water data.

[0020] Methods for cleaning multimodal water data include: The multimodal water service data is time aligned to obtain the time aligned water service data. The dynamic time warping algorithm is used to align the non-uniformly sampled data in the multimodal water service data. In order to ensure the efficiency of data collection and the accuracy of the collected multimodal water service data, the sampling rates of water service data of different modes may be different. For example, water quality data changes slowly, so it is set to every Sampling is performed once every minute; time alignment is used to solve the problem of time sequence misalignment of multi-source data caused by sampling frequency or transmission delay, and the time dimension of multimodal water service data is unified; an anomaly detection model is constructed to process the outliers of time-aligned water service data to generate normal water service data. In this embodiment, an unsupervised anomaly detection model is constructed based on an autoencoder to realize the recognition of abnormal data of time-aligned water service data, and the abnormal data is eliminated to obtain normal water service data; in the field of anomaly detection, traditional methods judge abnormal data based on fixed thresholds or adaptive thresholds, and have low sensitivity to nonlinear relationships. Therefore, an autoencoder is used to capture anomalies in complex relationship data, thereby improving the accuracy of anomaly detection; a normalization function is used to normalize the normal water service data to obtain standard multimodal water service data. In this embodiment, the Min-Max function is used to normalize the normal water service data. The Min-Max function is a commonly used normalization method that maps different modal data to , eliminating the interference of dimensional differences and improving the efficiency of subsequent data processing.

[0021] Methods for adaptive weighted fusion of standard multimodal water service data include: Perform modal classification on standard multimodal water service data to obtain standard water service data of any modality, and perform modal classification based on the physical meaning of each modal data in the standard multimodal water service data and the type of acquisition sensor, including water quality modality, equipment modality and fluid modality; wherein water quality modality includes data related to chemical elements in water bodies such as pH value and dissolved oxygen, equipment modality includes data related to water facilities such as pump current or vibration frequency; fluid modality includes data related to the physical properties of water flow such as water flow velocity or flow; calculate the entropy value of each data point in the input standard water service data of each modality, calculate the mutual information of each data point in the standard water service data of each modality based on the entropy value, and map the mutual information to the feature space to form a mutual information feature vector. In this embodiment, the entropy value of each data point in the standard water service data is calculated using the Shannon entropy formula. This formula is a basic concept in information theory and is common knowledge. Related field technology Personnel can obtain the entropy value through mathematical calculation; based on the entropy value, the mutual information between the mode to which the entropy value belongs and other modes is calculated using the k-nearest neighbor mutual information estimation algorithm, and the mutual information is mapped to the feature space to form a mutual information feature representation, which is the mutual information feature vector; deploy a fusion unit, receive the mutual information feature vector of the standard water service data of each mode to construct a mutual information feature matrix, wherein the fusion unit uses the cross-attention mechanism as the algorithm basis, and splices the received mutual information feature vectors through the cross-attention mechanism to form a cross-modal correlation matrix, which is the mutual information feature matrix, to ensure that the information complementarity between different modes is maximized; set the initial feature weight, use the activation function to calculate the weight of the mutual information feature matrix, and output the corrected weight, wherein the initial feature weight is generated based on the Xavier initialization method, which is a weight initialization method commonly used in deep learning models and can avoid gradient disappearance; in this embodiment, the activation function is Function, using this function to normalize each row of the mutual information feature matrix to obtain the modified weight, where each element is the weight of the corresponding dimension; based on the modified weight, the standard water service data of each mode is weighted to obtain a multimodal water service data matrix; according to each weight in the modified weight, the standard water service data of each mode is weighted and spliced ​​to obtain a multimodal water service data matrix.

[0022] Ways to build a digital twin water architecture based on a multimodal water data matrix include: Build a communication protocol library, calculate the system network bandwidth in real time, set the network bandwidth threshold, and dynamically select the optimal communication protocol based on the system network bandwidth and the network bandwidth threshold. At the same time, receive the multimodal water service data matrix as input data. In this embodiment, the communication protocols that can be selected include Communication protocols, communication protocols and Communication protocol: When implementing, you can select a communication protocol that better suits the working conditions from the preset communication protocol database; the network bandwidth thresholds include 100kb and 500kb, and it is enabled when the system network bandwidth is less than 100kb Communication protocol; enabled when the system network bandwidth is greater than or equal to 100kb and less than or equal to 500kb Communication protocol; enabled when the system network bandwidth is greater than 500kb communication protocol; making data transmission more flexible and efficient by dynamically switching communication protocols; constructing a high-precision three-dimensional water network model and a fluid simulation model of a preset water system. In this embodiment, a combination of BIM and GIS modeling methods is used to construct the high-precision three-dimensional water network model and the fluid simulation model; wherein in the existing modeling field, the BIM modeling method is suitable for modeling buildings, and therefore the method is used to model water facilities in the preset water system, such as sewage treatment plants, pumping stations, units, and valves; the GIS modeling method is suitable for environmental modeling based on geographic information, and therefore the method is used to model the surrounding environment and natural water bodies of the preset water system, and at the same time, the method is used to simulate water flow and construct a fluid simulation model; constructing a scenario simulation engine, and at the same time constructing a water knowledge database associated with the scenario simulation engine, wherein the scenario simulation engine is used to realize three-dimensional model visualization, interaction between equipment and environment, and scenario deduction, and the associating water knowledge database is to ensure that the simulated scenario conforms to real rules; wherein the water knowledge database includes normal operation data of the water system, which is used to restore the actual water system operation process; integrating the high-precision three-dimensional water network model, the fluid simulation model, and the scenario simulation engine to generate a digital twin water architecture.

[0023] Ways to use the digital twin water architecture to construct simulated water management scenarios include: The preset water system is physically modeled to generate a high-precision three-dimensional water network model, and a fluid simulation model is used to simulate the flow of water bodies. The fluid simulation model is used to simulate the flow of water bodies so that the high-precision three-dimensional water network model conforms to the laws of fluid dynamics. At the same time, the high-precision three-dimensional water network model includes the topological relationship between various water facilities, such as the connection structure of the pipes in the pipe network and the layout of the pump room. The water facilities in the high-precision three-dimensional water network model are labeled, and the parameters of the high-precision three-dimensional water network model are marked based on the multimodal water data matrix. After each water facility is marked, it is easy to identify the specific water facility. The data in the multimodal water data matrix is ​​used to mark it in the high-precision three-dimensional water network model, and the parameters of the corresponding facilities or water bodies are reflected in real time. At the same time, the parameters are matched with the corresponding parts in the model to prevent later The problem of parameter confusion continues to occur; the water management mode of the high-precision three-dimensional water network model is extracted, and the water management mode is encoded to obtain the water management mode code, wherein the water management mode refers to the linkage operation path mode composed of multiple subsystems and water body changes in the water system preset at certain specific times. For example, the water management mode from six to nine in the morning every day is the peak water supply mode. By starting a water pump and opening a valve associated with the water pump, the pressurizing facility enters high-frequency operation, wherein the equipment parameter changes of each water facility and the water flow rate, water volume and water level and other water body physical information are all part of the peak water supply mode; wherein the water management mode code reflects the mode ID, mode type, time range, trigger conditions, water facility labels that need to be operated and water facility startup sequence of the water management mode, which is convenient for terminal identification.

[0024] A management model database is constructed based on the codes of all water management models, and the management model database is associated with a high-precision three-dimensional water network model. The associated management model database and the high-precision three-dimensional water network model are used to quickly switch the water management mode in the simulated water management scenario, thereby improving the authenticity of the simulated water management scenario. A scenario simulation engine is used to generate a simulated water management scenario based on the high-precision three-dimensional water network model, and the constraints of the simulated water management scenario are set based on the water knowledge database. The water knowledge database ensures that the simulated water management scenario conforms to real rules, thereby further ensuring the rationality of the simulated water management scenario.

[0025] Ways to use ecological assessment models to monitor water quality in simulated water management scenarios include: Based on the ecological assessment model, a pollutant diffusion path of a simulated water management scenario is constructed, and a pollutant diffusion path map is drawn based on the pollutant diffusion path. The ecological assessment model in this embodiment simulates the transmission and diffusion of pollutants in water as the water flow changes based on the convection-diffusion equation to obtain the diffusion path of pollutants. The convection-diffusion equation is a basic equation commonly used in fluid mechanics and is widely used to simulate pollutant propagation phenomena. The pollutant diffusion path map is drawn using Python's drawing tools, and pollutant attribute information such as pollutant concentration, type and diffusion rate is marked on the path map. The prediction accuracy of the pollutant diffusion path is improved through fluid mechanics simulation. The water quality parameters of the simulated water management scenario are extracted, and a water quality thermodynamic map is drawn based on the water quality parameters. The water quality thermodynamic map is superimposed with the pollutant diffusion path map to obtain water quality assessment visualization data. In this embodiment, the Kriging interpolation algorithm is used to spatially interpolate the water quality parameters of discrete monitoring points in the simulated water management scenario to capture the spatial distribution law of water quality changes and draw a water quality thermodynamic map, which intuitively reflects the spatial distribution of water quality changes in the simulated water management scenario. The rendering engine superimposes the water quality heat map with the pollutant diffusion path map to construct a visualization interface of the dynamic pollution evolution, which is the water quality assessment visualization data.

[0026] Based on the water affairs knowledge database, a four-layer water quality assessment system is designed, including physical indicator assessment, chemical indicator assessment, biological indicator assessment, and functional indicator assessment. Physical indicators represent water body physical attribute indicators such as temperature, turbidity, and flow rate; chemical indicators represent water body chemical attribute indicators such as pH value, dissolved oxygen content, and heavy metal content; biological indicators represent the complexity of microbial species; and functional indicators represent the self-purification capacity coefficient of water bodies. A water quality assessment function is constructed based on the four-layer water quality assessment system, and the water quality assessment score of the simulated water affairs management scenario is calculated using the water quality assessment function. The calculation formula of the water quality assessment function is as follows: ;in, represents the water quality assessment score of the simulated water management scenario; represents the physical indicator assessment score; represents the chemical index evaluation score; represents the bioindicator assessment score; represents the functional index assessment score; 、 、 and They are 、 、 and The weight, in this embodiment , , , ; The calculation formula for the physical index evaluation score is: ;in, represents the number of physical indicators in the simulated water management scenario; Indicates the first The value of a physical indicator; Indicates the first The ideal value of a physical indicator is the ideal value of the corresponding physical indicator obtained by querying the water affairs knowledge database; Indicates the allowable deviation. In this embodiment, the allowable deviation is ; The calculation formula for chemical index evaluation score is: ;in, represents the number of chemical indicators in the simulated water management scenario; Indicates the first The value of a chemical indicator; Indicates the first The ideal value of a chemical indicator is the ideal value of the corresponding chemical indicator obtained by querying the water affairs knowledge database; Indicates the decrease control factor, which is used to reflect the change in the content of chemical elements. In this embodiment, the value is ; The calculation formula for the biological indicator evaluation score is: ;in, Indicates the theoretical maximum value of microbial diversity, which is taken as ; The number of microbial species in simulated water management scenarios; Indicates the first The calculation formula of functional index evaluation score is: ;in, Represents the water body self-purification capacity coefficient for the simulated water management scenario; It represents the theoretical maximum water body self-purification capacity coefficient, which is obtained by querying the water affairs knowledge database; sets a water quality assessment grading threshold, determines the water quality assessment grade of the simulated water affairs management scenario based on the water quality assessment grading threshold and the water quality assessment score, divides the water quality assessment grades into five categories based on the water quality assessment grading threshold, and determines the water quality assessment grade to which the simulated water affairs management scenario belongs based on the water quality assessment score; associates the ecological assessment model with the preset weather forecast system, and constructs a water quality change trend map based on the weather forecast data of the preset weather forecast system, embeds the weather forecast data into the convection-diffusion equation, simulates the changes in pollutant concentration caused by changes in weather forecast data based on the convection-diffusion equation, and constructs a water quality change trend map; integrates water quality assessment visualization data, water quality assessment grades and water quality change trend maps to obtain water quality assessment data for the simulated water affairs management scenario.

[0027] Ways to construct a water management zoning topology include: The topological information of the preset water system in the simulated water management scenario is extracted, and a water topological map of the simulated water management scenario is constructed based on the topological information. In this embodiment, the topological information of the preset water system is extracted using a graph convolutional neural network model, and a water topological map is constructed; the water topological map is constructed to facilitate the subsequent regional partitioning architecture to perform regional division based on the topological structure therein, while improving the accuracy of regional division; the water facility entities in the preset water system are used as nodes of the water topological map, and the connection relationship of the water facilities in the preset water system is modeled as an edge of the water topological map, wherein the node attribute information in each node includes the water facility type, spatial coordinates and real-time status parameters, such as pressure, flow and energy consumption; the attribute information of each edge includes the length and direction of the connecting pipe section, and fluid parameters such as water flow velocity and flow direction.

[0028] Extract the node information of each node in the water management topology map, build a node information feature matrix based on the node information, convert the node information into a vector representation, and stack it to obtain the node information feature matrix, which provides data support for subsequent area division; build an adaptive range window based on the spatial position of each node in the simulated water management scenario and dynamically adjust the window radius, wherein the spatial coordinates of any node are used as the center point of the adaptive range window, and each adaptive range window is numbered, and the size of the adaptive range window is adjusted based on the specific situation. For example, if there are a large number of water facilities, the size of the adaptive range window is reduced to improve the partitioning precision; in this embodiment, the initial size of the adaptive range window is set to the radius Meters, this radius is the length of the real scene, and it needs to be proportionally reduced in the simulated water management scene; a regional division structure is constructed to extract the water state parameters in each adaptive range window, and all water state parameters are grouped based on the number, where the water state parameters include real-time flow, flow velocity, water quality indicators, pressure and other water-related parameters; grouping is used to ensure the consistency of water state data in each area during the regional division process, thereby improving the regional division effect; the simulated water management scene is divided into regions based on the node information feature matrix and a set of water state parameters corresponding to the node, and an initial divided region set is obtained. A hybrid clustering algorithm is used for regional division to realize sub-region division based on physical location, topological structure and water state parameters.

[0029] A balance constraint function is constructed, and the rationality of the division of the initial divided area set is judged based on the balance constraint function. The calculation formula of the balance constraint function is: ;in, represents modularity; Indicates the number of edges of the initial partition area; Representation node With node The edge weight between them is obtained by querying the water affairs knowledge database to obtain the weight of the connection relationship between water affairs facilities, which is the edge weight; Representation and Node The sum of the weights of connected edges; Representation and Node The sum of the weights of connected edges; Is a judgment function, if the node With node Belong to the same initial divided area, then the function value is , otherwise ; If it is unreasonable, the boundaries of each initial divided area are readjusted to obtain an optimized divided area set. In this embodiment, the boundary node reallocation mechanism is used to adjust the boundary. This mechanism is an optimization method for adjusting the attribution of partition boundary nodes that is widely used in the existing technical field. The partition boundary is adjusted by evaluating the correlation between the node and different partitions; the water management topology map, regional division structure and balance constraint function are integrated to construct a water management partition topology structure.

[0030] The methods for dividing the simulated water management scenario into regions based on the node information feature matrix and a set of water state parameters corresponding to the node include: Extract node feature vectors from the node information feature matrix, perform feature splicing on the node feature vector and a set of water body state parameters corresponding to the node to obtain a partition input vector, normalize the node information feature matrix to obtain a node feature vector, convert the water body state parameter into a vector representation, and perform vector splicing on the node feature vector to obtain a partition input vector; map all partition input vectors to the constructed input parameter data space to obtain a unified fusion vector, in this embodiment, use the principal component analysis algorithm to compress the dimension of all partition input vectors, and map them to the same input parameter data space to generate a unified fusion vector, unify the data dimension, and eliminate the dimensional difference; classify all unified fusion vectors to obtain a preliminary grouping vector set, wherein the DBSCAN density clustering algorithm is used for preliminary clustering processing, which retains the topological connectivity while ensuring the efficiency of clustering processing; construct regional division The model extracts the topological relationship in the water service topology map and associates it with the regional partitioning model. The regional partitioning model is based on the GCN graph convolutional network model. The graph convolution layer extracts the topological relationship in the water service topology map, captures the spatial connection characteristics, and strengthens the connection dependency in the same area. The regional partitioning model is used to partition the preliminary grouping vector set to output the initial partitioning area set. The preliminary grouping vector set is used as the input of the regional partitioning model. The node information of each node is extracted through the regional partitioning model. The regional partitioning label is generated using the semi-supervised label propagation mechanism. The initial partitioning area set after division is output, where each area is composed of multiple nodes and their associated paths, ensuring the topological connectivity and integrity of each initial partitioning area. The regional topology factor is constructed, and the partitioning processing degree of the regional partitioning model is adjusted based on the regional topology factor. The calculation formula of the regional topology factor is: ;in, Indicates initialization of nodes in the sub-area The sum of the edge weights of all connected nodes; Indicates the initialization of the edge of the sub-region Edge betweenness centrality, where edge betweenness centrality is used to quantify the role of an edge in different parts of a region and is a method used in the prior art to measure the importance of an edge; and Respectively and The weight is adjusted based on the specific division situation. In this embodiment, and The initial values ​​of and .

[0031] Methods for cross-regional water resource scheduling in simulated water management scenarios based on the water management zoning topology and water quality assessment data include: Two types of inter-regional water resource dispatching conditions are set, which represent two situations: abnormal water quality and regional supply and demand imbalance. One type of inter-regional water resource dispatching condition is: detecting the water quality assessment data of each optimized divided area in the simulated water management scenario, and comparing the water quality assessment visualization data with the preset water body visualization database. If it meets the pollution water quality chart in the water body visualization database, and the water quality assessment level is lower than the preset water body assessment threshold or the water quality change amplitude in the water quality change trend map is greater than the preset change amplitude threshold, then the inter-regional water resource dispatching is triggered. The water body visualization database includes pollution water quality charts corresponding to typical pollution, and each pollution water quality chart is a pollution. Distribution template; when the water quality assessment visualization data is consistent with the polluted water quality chart and the water quality assessment level is low or the water quality change amplitude suddenly changes, it can be considered that the water quality is abnormal at this time and a pollution incident may have occurred. In order to prevent the spread of pollution, clean water needs to be introduced into the area to dilute the pollution, so cross-regional scheduling of water resources is required; the second-class cross-regional scheduling conditions of water resources are: collect the water state parameters of each optimized division area in the simulated water management scenario, construct the supply and demand load function based on the water state parameters, and use the supply and demand load function to calculate the supply and demand load of each optimized division area. If it is greater than or equal to the preset load threshold, the cross-regional scheduling of water resources is triggered. The calculation formula of the supply and demand load function is: ;in Represents the supply and demand load of any optimally divided area; Indicates the water demand of the optimized divided area; Indicates the available water volume in the optimized divided area; if the supply and demand load is greater than or equal to the preset load threshold, it is determined that the water supply is insufficient at this time and a certain amount of water needs to be dispatched to supplement the supply.

[0032] Based on the topological structure of water management zoning, water sub-regions with associated paths are screened with the optimized divided areas that require inter-regional water resource scheduling. At the same time, the connection relationship between the optimized divided areas and the water sub-regions with associated paths is extracted to construct an initial scheduling path dataset. The areas that can be connected to the optimized divided areas that require inter-regional water resource scheduling are regarded as water sub-regions with associated paths, and the topological relationships therein are extracted to construct an initial scheduling path dataset. A cost function is constructed and the cost consumption score of each initial scheduling path in the initial scheduling path dataset is calculated using the cost function. The calculation formula of the cost function is: ;in Represents the cost consumption fraction of any initial scheduling path; represents the energy loss cost; It represents the water loss cost. Water loss is likely to occur during water resource scheduling, which is the water loss cost. It represents the facility disturbance cost. When the facility is started and stopped, it is easy to cause equipment friction and cause disturbance to the system. 、 and They are 、 and The weight, in this embodiment 、 、 ; The initial scheduling path with the smallest cost consumption score is used as the regional scheduling path, and the digital twin water architecture is used to predict the changes in water state parameters of the regional scheduling path, and the water zoning management strategy is output. The fluid simulation model in the digital twin water architecture is used to simulate the changes in water state parameters in the regional scheduling path, and the water zoning management strategy is output based on the changes in water state parameters and the topological relationship in the regional scheduling path; the water zoning management strategy reflects the regional topological relationship for cross-regional water resource scheduling, the control instructions of all water facilities on the path, and the corresponding changes in water state parameters.

[0033] Methods for constructing water system component operation scenarios based on the digital twin water architecture include: The equipment status parameters of each water facility in each optimized divided area in the simulated water management scenario are extracted, and the digital twin water architecture is called. Based on the equipment status parameters, a digital twin object of each water facility is created in the water network construction layer, and the digital twin objects of all water facilities are partitioned and mapped based on the topological relationship between water facilities. The equipment status parameters include pump speed, valve switch status, facility pressure, facility temperature difference and other parameters related to water facilities; the attribute information of the created digital twin object includes spatial location, topological connection relationship, current equipment status parameters and historical operation records; the digital twin object is assigned to the corresponding optimized divided area according to the topological relationship, and the topological relationship within and between partitions is unified modeling, thereby improving the calculation efficiency and local reconstruction capability of the constructed water system component operation scenario.

[0034] The management mode database is called to extract the operating status of each water facility in the management mode database belonging to any optimized divided area in the simulated water management scenario. The operating status is the water management mode being implemented by the water facility during that period, reflecting the changes in the equipment status parameters of the water facility. The operating status of each water facility is event-coded, and a facility operation event sequence is constructed based on the event code and timestamp. The event code is expressed in the form of (facility label, management mode ID, specific equipment status parameter change). All time codes are sorted based on the timestamp to obtain the facility operation event sequence. An event-driven operation scenario is constructed based on the facility operation event sequence of all water facilities. The event-driven operation scenario is associated with the fluid simulation model and water topology map in the digital twin water architecture to construct the water system component operation scenario. The digital twin entity of the corresponding water facility is constructed based on the facility operation event sequence. The equipment status parameters of the digital twin entity are adjusted based on the facility operation event sequence to form an event-driven operation scenario. The event sequence change path is updated through the water topology map, and water flow simulation is performed based on the fluid simulation model to finally construct the water system component operation scenario.

[0035] Methods for monitoring system faults in water system component operation scenarios include: Extract the equipment status parameters of water facilities in the water system component operation scenario, and use the equipment status parameters as the input data of the fault monitoring architecture; collect historical facility operation data, and build a system fault map based on the historical facility operation data, where the system fault map includes information such as changes in equipment status parameters and fault types before and after the fault occurs in the historical data, and use water facilities as map nodes and fault events as edges; the edges can reflect the potential correlation between faults; extract the water body status parameters of each water facility in the optimized division area in the water system component operation scenario, and combine the water body status parameters and equipment status parameters to obtain fault operation data. Since there may be certain components in the water body that are prone to corrosion and other faults in the water facility, it is necessary to collect water body status parameters; map the fault operation data of the water facility to the vector space to form a fault operation vector, integrate all fault operation vectors to construct an equipment fault matrix, where each fault operation vector corresponds to a water facility, and all fault operation vectors are mapped to the vector space to form a fault operation vector. Sorting is performed according to facility labels, and vector splicing is performed to obtain an equipment fault matrix; a fault monitoring model is constructed, and the system fault map is embedded in the fault monitoring model, and the fault monitoring model is used to perform fault detection on the equipment fault matrix, and a fault vector and a fault risk index are output. In this embodiment, the fault monitoring model uses a graph neural network model based on a graph attention mechanism as the basic structure, and extracts fault correlation and fault information in the system fault map as judgment conditions for fault detection; the fault vector includes the category probability of each fault type, and the fault risk index represents a scalar vector output by the fault monitoring model; a fault classification threshold is set, and the fault level of each water facility is determined based on the fault classification threshold and the fault risk index, and the fault type is determined based on the fault vector. In this embodiment, the fault level is divided into five levels based on the fault classification threshold, and the fault level to which the water facility belongs is determined based on the value of the fault risk index; if a fault is determined to exist, the fault category with the highest probability in the fault vector is output to determine the fault type.

[0036] Construct a prediction model, design facility operation and maintenance rules based on the water conservancy knowledge database, and combine the facility operation and maintenance rules with the prediction model, where the prediction model is an LSTM long short-term memory network model trained using historical facility operation data; the facility operation and maintenance rules include operating constraints such as the monthly average startup frequency upper limit, continuous operation time limit and voltage upper limit; the facility operation and maintenance rules are combined with the prediction model to realize rule-driven and data-driven collaborative fault prediction; use the prediction model to predict faults on the equipment fault matrix and output the fault prediction results, where the prediction model performs sequence prediction on the equipment fault matrix and outputs several prediction vectors; each prediction vector represents the change in fault level and fault type of the corresponding water conservancy facility in the future; draw a fault propagation path diagram based on the fault prediction results and the topological relationship between the water conservancy facilities, and visualize the fault propagation path by drawing the fault propagation path diagram, so that technical personnel can maintain the corresponding water conservancy facilities in advance to prevent faults.

[0037] Integrate the fault level, fault type and fault propagation path diagram to obtain a fault information set; generate a system maintenance plan based on the fault information set and the system fault map, encode the system maintenance plan to obtain a system maintenance log, match the fault corresponding to the fault information set based on the system fault map, and obtain the system maintenance plan corresponding to the fault type.

[0038] Methods for updating parameters of the digital twin water utility architecture based on system maintenance logs and real-time water utility data include: A data buffer is constructed in the digital twin water service architecture, and the data buffer is used to store real-time water service data and system maintenance logs. A log-water service parameter matrix is ​​constructed based on the system maintenance log and real-time water service data. The simulation scenario generation layer in the digital twin water service architecture is used to generate simulation scenarios based on the log-water service parameter matrix. Based on the specific maintenance operations, water service facility tags, and operating time in the system maintenance log, the real-time water service data is associated and mapped to obtain the log-water service parameter matrix. The simulation scenario generation layer is called to simulate the specific operation scenarios of the facilities corresponding to the log-water service parameter matrix. The simulation scenarios that are related to the log-water service parameter matrix are extracted. The simulation parameters corresponding to the parameters in the digital twin water service parameter matrix are compared with the log-water service parameter matrix, and an adjustable parameter data set is constructed based on the comparison results. Since there are certain differences between the simulation scenario and the actual scenario, the parameters of the actual scenario are extracted and compared with the simulation parameters of the simulation scenario, and all the parameters with differences are constructed as an adjustable parameter data set; the architectural parameters of the digital twin water service architecture are extracted, and a mapping relationship between the adjustable parameter data set and the architectural parameters is established. The parameter space is constructed based on the adjustable parameter data set and the architectural parameters, where the architectural parameters include the viscosity coefficient, water quality diffusion coefficient and Architectural parameters such as valve response delay are mapped to architectural parameters and adjustable parameter data sets. For example, a mapping relationship is generated between the flow velocity deviation in the pipe network and the pipe network friction coefficient in the architectural parameters, indicating that the facility may be aging. An optimization objective function is constructed, and a preset optimization algorithm is used to search for optimal parameters in the parameter space based on the optimization objective function, and the optimal parameter combination is output. The optimization algorithm is the particle swarm optimization algorithm, and the optimization objective function is the fitness function of the particle swarm optimization algorithm. The particle swarm optimization algorithm is used to search the parameter space to obtain the optimal parameter combination. The optimal parameter combination is applied to the digital twin water architecture, and the digital twin is used to The digital twin water architecture is re-simulated and the scene accuracy is compared. If the scene accuracy is higher than the preset scene accuracy threshold, the current digital twin water architecture is used as the optimized digital twin water architecture, and the optimal parameter combination of the optimized digital twin water architecture is stored in the preset parameter database to determine whether the digital twin water architecture with the optimal parameter combination can construct a higher-precision simulation scene. If it meets the expected effect, the digital twin water architecture is used as the optimized digital twin water architecture. The optimal parameter combination is stored in the preset parameter database to be able to quickly switch the parameter combination to ensure the performance of the digital twin water architecture.

[0039] This embodiment integrates adaptive weighted processing of multimodal water data, the construction of a digital twin water architecture and three-dimensional scenario simulation, the embedding of a water quality ecological assessment model, the design of a cross-regional water resource scheduling mechanism, and the coordination of system component operation monitoring and intelligent fault prediction models to achieve intelligent, precise, and dynamic management of all elements, all time periods, and all processes of the water system. The digital twin-based smart water dynamic monitoring method not only improves the real-time perception of key information such as water quality changes, equipment failures, and water pressure fluctuations, but also has the ability to intelligently predict and dispatch water management strategies. It effectively improves the operational safety, control efficiency, and maintenance autonomy of the overall water system, supports the transformation of water management from traditional passive response to active prediction and self-optimization, and significantly enhances the intelligence level of water management and emergency decision-making capabilities.

[0040] Example 2 See also Figure 2 As shown, for the parts not described in detail in this embodiment, please refer to the description of Example 1. A smart water affairs dynamic monitoring system based on digital twins is provided, including: A data acquisition module is used to collect multimodal water service data from a preset water service system, clean the multimodal water service data, and obtain standard multimodal water service data; The data fusion module performs adaptive weighted fusion on standard multimodal water service data to construct a multimodal water service data matrix; The digital twin simulation module builds a digital twin water architecture based on a multimodal water data matrix and uses the digital twin water architecture to construct simulated water management scenarios. The water quality monitoring module builds an ecological assessment model and embeds it into the digital twin water management architecture. It uses the ecological assessment model to monitor water quality in simulated water management scenarios and outputs water quality assessment data. The water management scheduling module builds a water management zoning topology, performs cross-regional water resource scheduling for simulated water management scenarios based on the zoning topology and water quality assessment data, outputs water management zoning strategies, and feeds these strategies back to the digital twin water architecture. The fault monitoring module constructs water system component operation scenarios based on the digital twin water architecture, monitors system faults in the water system component operation scenarios, and outputs system maintenance logs; The parameter optimization module is used to collect real-time water service data, update the parameters of the digital twin water service architecture based on the system maintenance log and real-time water service data to obtain an optimized digital twin water service architecture, and deploy the optimized digital twin water service architecture to the preset water service system; each module is connected by wired and / or wireless means.

[0041] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art will be able to modify the technical solutions described in the foregoing embodiments or to substitute equivalents for some of the technical features. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

[0042] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0043] In the description of the present invention, it should be understood that the terms "first", "second", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0044] In the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0045] In the description of the present invention, “several” means one or more, and “a large number” means two or more.

[0046] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0047] The formulas in this manual are all dimensionless and calculated using numerical values. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field based on actual conditions.

[0048] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

Claims

1. The digital twin-based smart water affairs dynamic monitoring method is characterized by: include: S1. Collect multimodal water service data from a preset water service system, clean the multimodal water service data, and obtain standard multimodal water service data; S2. Perform adaptive weighted fusion on standard multimodal water service data to construct a multimodal water service data matrix; S3. Build a digital twin water architecture based on a multimodal water data matrix and use the digital twin water architecture to construct simulated water management scenarios. S4. Build an ecological assessment model and embed it into the digital twin water management architecture. Use the ecological assessment model to monitor water quality in simulated water management scenarios and output water quality assessment data. S5. Construct a water management zoning topology. Based on the zoning topology and water quality assessment data, perform cross-regional water resource scheduling for simulated water management scenarios. Output a water management strategy for the zoning area and feed it back into the digital twin water architecture. S6. Construct water system component operation scenarios based on the digital twin water architecture, perform system fault monitoring on these water system component operation scenarios, and output system maintenance logs; S7. Collect real-time water service data, update the parameters of the digital twin water service architecture based on the system maintenance log and real-time water service data to obtain an optimized digital twin water service architecture, and deploy the optimized digital twin water service architecture to the preset water service system.

2. The digital twin-based smart water affairs dynamic monitoring method according to claim 1 is characterized in that: The method of performing data cleaning on multimodal water service data includes: Perform time alignment on multimodal water service data to obtain time-aligned water service data; construct an anomaly detection model to process outliers in the time-aligned water service data to generate normal water service data; use a normalization function to normalize the normal water service data to obtain standard multimodal water service data; Methods for adaptive weighted fusion of standard multimodal water service data include: Perform modal classification on standard multimodal water affairs data to obtain standard water affairs data of any modality; calculate the entropy value of each data point in the input standard water affairs data of each modality, calculate the mutual information of each data point in the standard water affairs data of each modality based on the entropy value, and map the mutual information to the feature space to form a mutual information feature vector; deploy a fusion unit, receive the mutual information feature vector of the standard water affairs data of each modality, and construct a mutual information feature matrix; set the initial feature weights, use the activation function to calculate the weights of the mutual information feature matrix, and output the corrected weights; weight the standard water affairs data of each modality based on the corrected weights to obtain a multimodal water affairs data matrix.

3. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 2 is characterized in that: The method of constructing a digital twin water service architecture based on a multimodal water service data matrix includes: Build a communication protocol library, calculate the system network bandwidth in real time, set the network bandwidth threshold, and dynamically select the optimal communication protocol based on the system network bandwidth and the network bandwidth threshold, while receiving a multimodal water data matrix as input data; construct a high-precision three-dimensional water network model and fluid simulation model of the preset water system; build a scenario simulation engine, and simultaneously build a water knowledge database and associate it with the scenario simulation engine; integrate the high-precision three-dimensional water network model, fluid simulation model, and scenario simulation engine to generate a digital twin water architecture; Ways to use the digital twin water architecture to construct simulated water management scenarios include: Perform physical structural modeling of the preset water system to generate a high-precision three-dimensional water network model, and use a fluid simulation model to simulate water flow. Label the water facilities in the high-precision three-dimensional water network model, and simultaneously label the parameters of the high-precision three-dimensional water network model based on a multimodal water data matrix. Extract the water management model from the high-precision three-dimensional water network model, and encode the water management model to obtain a water management model code. A management model database is constructed based on all water management model codes, and the management model database is associated with a high-precision three-dimensional water network model; a scenario simulation engine is used to generate simulated water management scenarios based on the high-precision three-dimensional water network model, and constraints for the simulated water management scenarios are set based on the water knowledge database.

4. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 3 is characterized in that: The method of using the ecological assessment model to monitor water quality in a simulated water management scenario includes: Based on the ecological assessment model, a pollutant diffusion path for a simulated water management scenario is constructed, and a pollutant diffusion path map is drawn based on the pollutant diffusion path. Water quality parameters for the simulated water management scenario are extracted, and a water quality thermodynamic map is drawn based on the water quality parameters. The water quality thermodynamic map is then superimposed on the pollutant diffusion path map to obtain visualized water quality assessment data. A four-layer water quality assessment system is designed based on the water affairs knowledge database, a water quality assessment function is constructed based on the four-layer water quality assessment system, and the water quality assessment score of the simulated water affairs management scenario is calculated using the water quality assessment function; a water quality assessment grading threshold is set, and the water quality assessment grade of the simulated water affairs management scenario is determined based on the water quality assessment grading threshold and the water quality assessment score; the ecological assessment model is associated with the preset weather forecast system, and a water quality change trend map is constructed based on the weather forecast data of the preset weather forecast system; water quality assessment visualization data, water quality assessment grades and water quality change trend maps are integrated to obtain water quality assessment data for the simulated water affairs management scenario.

5. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 4 is characterized in that: The method of constructing the water management zoning topology structure includes: Extract the topology information of the preset water system in the simulated water management scenario, and construct a water topology map of the simulated water management scenario based on the topology information; Extract node information from each node in the water management topology map and construct a node information feature matrix based on the node information; construct an adaptive range window and dynamically adjust the window radius, while numbering each adaptive range window; construct a regional division structure to extract the water state parameters in each adaptive range window and group all water state parameters based on the number; divide the simulated water management scenario into regions based on the node information feature matrix and a set of water state parameters corresponding to the node to obtain an initial set of divided regions; Construct a balance constraint function and judge the rationality of the initial partition area set based on the balance constraint function; if it is unreasonable, readjust the boundaries of each initial partition area to obtain the optimized partition area set; integrate the water management topology map, regional division structure and balance constraint function to construct the water management zoning topology structure; The methods for dividing the simulated water management scenario into regions based on the node information feature matrix and a set of water state parameters corresponding to the node include: Node feature vectors are extracted from the node information feature matrix, and the node feature vectors are feature-concatenated with a set of water body state parameters corresponding to the node to obtain a partition input vector; all partition input vectors are mapped to the constructed input parameter data space to obtain a unified fusion vector; all unified fusion vectors are classified and processed to obtain a preliminary grouping vector set; a regional partitioning model is constructed, and the topological relationship in the water service topology map is extracted and associated with the regional partitioning model; the preliminary grouping vector set is partitioned using the regional partitioning model to output an initial partitioning region set; a regional topology factor is constructed, and the partitioning processing degree of the regional partitioning model is adjusted based on the regional topology factor.

6. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 5 is characterized in that: The method of performing cross-regional water resource scheduling for a simulated water management scenario based on the water management zoning topology and water quality assessment data includes: Two types of inter-regional water resource dispatching conditions are set; the first type of inter-regional water resource dispatching condition is: detecting the water quality assessment data of each optimized division area in the simulated water management scenario, comparing the water quality assessment visualization data with the preset water body visualization database, and triggering inter-regional water resource dispatching if it meets the pollution water quality chart in the water body visualization database, and the water quality assessment level is lower than the preset water body assessment threshold or the water quality change amplitude in the water quality change trend map is greater than the preset change amplitude threshold; the second type of inter-regional water resource dispatching condition is: collecting the water body state parameters of each optimized division area in the simulated water management scenario, constructing a supply and demand load function based on the water body state parameters, and using the supply and demand load function to calculate the supply and demand load of each optimized division area, and triggering water resource inter-regional dispatching if it is greater than or equal to the preset load threshold; Based on the topological structure of water management zoning, water sub-regions with associated paths in the optimized divided areas that require cross-regional water resource scheduling are screened. At the same time, the connection relationship between the optimized divided areas and the water sub-regions with associated paths is extracted to construct an initial scheduling path dataset; a cost function is constructed, and the cost consumption score of each initial scheduling path in the initial scheduling path dataset is calculated using the cost function; the initial scheduling path with the smallest cost consumption score is used as the regional scheduling path, and the digital twin water architecture is used to predict the changes in water state parameters of the regional scheduling path, and output the water zoning management strategy.

7. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 6 is characterized in that: The method of constructing the water system component operation scenario based on the digital twin water architecture includes: Extract the equipment status parameters of each water facility in each optimized partitioned area in the simulated water management scenario, use the digital twin water architecture to create a digital twin object for each water facility based on the equipment status parameters, and perform partition mapping on the digital twin objects of all water facilities based on the topological relationships between water facilities; Call the management mode database and extract the operating status of each water facility in any optimized divided area in the simulated water management scenario; encode the operating status of each water facility, and construct a facility operation event sequence based on the event code and timestamp; build an event-driven operation scenario based on the facility operation event sequence of all water facilities, and associate the event-driven operation scenario with the fluid simulation model and water topology map in the digital twin water architecture to construct the water system component operation scenario.

8. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 7 is characterized in that: The method of performing system fault monitoring on the operation scenario of the water system components includes: Extract the equipment status parameters of water facilities in the water system component operation scenario; collect historical facility operation data to construct a system fault map; extract the water body state parameters in the optimized division area of ​​each water facility in the water system component operation scenario, and combine the water body state parameters and equipment state parameters to obtain fault operation data; map the fault operation data of the water facility to the vector space to form a fault operation vector, and integrate all fault operation vectors to construct an equipment fault matrix; build a fault monitoring model, and embed the system fault map into the fault monitoring model, use the fault monitoring model to perform fault detection on the equipment fault matrix, and output the fault vector and fault risk index; set the fault classification threshold, determine the fault level of each water facility based on the fault classification threshold and fault risk index, and determine the fault type based on the fault vector; Build a prediction model, design facility operation and maintenance rules based on the water affairs knowledge database, and combine the facility operation and maintenance rules with the prediction model; use the prediction model to predict faults in the equipment fault matrix and output the fault prediction results; draw a fault propagation path diagram based on the fault prediction results and the topological relationship between water affairs facilities; The fault level, fault type and fault propagation path diagram are integrated to obtain a fault information set; a system maintenance plan is generated based on the fault information set and the system fault map, and the system maintenance plan is encoded to obtain a system maintenance log.

9. The method for dynamic monitoring of smart water affairs based on digital twins according to claim 8, characterized in that: The method of updating parameters of the digital twin water service architecture based on system maintenance logs and real-time water service data includes: A data buffer is constructed within the digital twin water utility architecture, and is used to store real-time water utility data and system maintenance logs. A log-water utility parameter matrix is ​​constructed based on the system maintenance logs and real-time water utility data, and simulation scenarios are generated based on the log-water utility parameter matrix using the digital twin water utility architecture. Simulation parameters corresponding to the parameters in the log-water utility parameter matrix are extracted from the simulation scenario, and the simulation parameters are compared with the log-water utility parameter matrix. Based on the comparison results, an adjustable parameter dataset is constructed. Extract the architectural parameters of the digital twin water architecture, establish a mapping relationship between the adjustable parameter data set and the architectural parameters, and construct a parameter space based on the adjustable parameter data set and the architectural parameters; construct an optimization objective function, and use a preset optimization algorithm to search for optimal parameters in the parameter space based on the optimization objective function, and output the optimal parameter combination; apply the optimal parameter combination to the digital twin water architecture, use the digital twin water architecture to re-simulate the scene, and perform a scene accuracy comparison. If the scene accuracy is higher than the preset scene accuracy threshold, the digital twin water architecture at this time will be used as the optimized digital twin water architecture, and the optimal parameter combination of the optimized digital twin water architecture will be stored in the preset parameter database.

10. A digital twin-based smart water affairs dynamic monitoring system, which is used to implement the digital twin-based smart water affairs dynamic monitoring method according to any one of claims 1 to 9, characterized in that: include: A data acquisition module is used to collect multimodal water service data from a preset water service system, clean the multimodal water service data, and obtain standard multimodal water service data; The data fusion module performs adaptive weighted fusion on standard multimodal water service data to construct a multimodal water service data matrix; The digital twin simulation module builds a digital twin water architecture based on a multimodal water data matrix and uses the digital twin water architecture to construct simulated water management scenarios. The water quality monitoring module builds an ecological assessment model and embeds it into the digital twin water management architecture. It uses the ecological assessment model to monitor water quality in simulated water management scenarios and outputs water quality assessment data. The water management scheduling module builds a water management zoning topology, performs cross-regional water resource scheduling for simulated water management scenarios based on the zoning topology and water quality assessment data, outputs water management zoning strategies, and feeds these strategies back to the digital twin water architecture. The fault monitoring module constructs water system component operation scenarios based on the digital twin water architecture, monitors system faults in the water system component operation scenarios, and outputs system maintenance logs; The parameter optimization module is used to collect real-time water service data, update the parameters of the digital twin water service architecture based on the system maintenance log and real-time water service data to obtain an optimized digital twin water service architecture, and deploy the optimized digital twin water service architecture to the preset water service system; each module is connected by wired and / or wireless means.

Images