A water conservancy project operation and maintenance management system based on digital twin

Abstract

The present application provides a water conservancy project operation and maintenance management system based on digital twins, which relates to the field of operation and maintenance management technology, including: a hydrological sensitive point screening module to screen finite element stress hotspots and hydrological sensitive points of target water conservancy projects; a sensor deployment module to deploy multimodal sensors; a twin model construction module to push real-time data to the digital twin synchronization engine; a comparison loss value set determination module to perform a comparison loss analysis between the real-time digital twin model and the historical digital twin model constructed in the last synchronization; a reverse mapping module to reversely map the multimodal sensor network; an asynchronous sampling module to asynchronously sample the multimodal sensor network and identify the operation and maintenance management solution. This application can solve the technical problem in the prior art that the sampling frequency cannot be adaptively adjusted according to real-time changes, realize the technical goal of intelligent sampling of multimodal sensors, and achieve the technical effect of improving the accuracy of key data acquisition.

Description

Technical Field

[0001] The present application relates to the field of operation and maintenance management technology, and in particular to a water conservancy project operation and maintenance management system based on digital twins. Background Art

[0002] In the current intelligent operation and maintenance management of water conservancy projects, the integration of sensor networks and digital twin technology has gradually become a mainstream application method, especially in large-scale hydraulic structures such as dams, sluices or water diversion tunnels. Multimodal sensors are used to realize real-time monitoring of multi-dimensional information such as structural stress, water flow status, temperature, power operation, etc., combined with three-dimensional modeling and simulation analysis, which effectively improves the ability to understand the status of the project. However, existing technologies still have significant defects in sensor deployment strategies, data processing efficiency and dynamic response capabilities. At present, in the data acquisition and processing links, the data types generated by multimodal sensors are diverse and the frequencies are different. Existing solutions usually adopt fixed-frequency acquisition and unified data channel processing, and fail to dynamically adjust the sampling strategy according to the importance or change rate of the data, resulting in waste of network bandwidth and delayed response of some key data.

[0003] In summary, the existing technology has technical problems such as the lack of dynamic identification and differentiated processing mechanism for multimodal sensor data features, which leads to the inability to adaptively adjust the sampling frequency and data transmission strategy according to real-time changes, further affecting the data utilization efficiency, response timeliness and abnormal warning capabilities of the water conservancy project monitoring system. Summary of the Invention

[0004] The purpose of this application is to provide a water conservancy project operation and maintenance management system based on digital twins to solve the technical problems in the existing technology that the sampling frequency and data transmission strategy cannot be adaptively adjusted according to real-time changes due to the lack of dynamic recognition and differentiated processing mechanism of multimodal sensor data characteristics, further affecting the data utilization efficiency, response timeliness and abnormal warning capabilities of the water conservancy project monitoring system.

[0005] In view of the above problems, the present application provides a water conservancy project operation and maintenance management system based on digital twin, including: a hydrological sensitive point screening module, which is used to screen the finite element stress hotspots and hydrological sensitive points of the target water conservancy project, and obtain a finite element stress hotspot set and a hydrological sensitive point set; a sensor deployment module, which is used to traverse the finite element stress hotspot set and the hydrological sensitive point set to deploy multimodal sensors, and obtain a deployed multimodal sensor network; a twin model construction module, which is used to use the industrial edge gateway to push the real-time data collected in the multimodal sensor network to the digital twin synchronization engine, and build A real-time digital twin model; a comparison loss value set determination module, used to perform a comparison loss analysis on the real-time digital twin model and the historical digital twin model constructed synchronously last time, and determine the comparison loss value set; a reverse mapping module, used to perform reverse mapping on the multimodal sensor network based on the comparison loss value set, and determine the multimodal sensor anchor point set; an asynchronous sampling module, used to perform asynchronous sampling on the multimodal sensor network according to the multimodal sensor anchor point set and the comparison loss value set, and identify the operation and maintenance management plan based on the asynchronous sampling results to determine the target operation and maintenance management plan.

[0006] Preferably, the digital twin-based water conservancy project operation and maintenance management system also includes: the multimodal sensor is selected from at least one of the following sensors: fiber grating strain gauge, piezoelectric accelerometer, ultrasonic water level meter, RTD temperature probe, smart meter.

[0007] Preferably, the water conservancy project operation and maintenance management system based on digital twins also includes: an association tracing analysis unit, which is used to perform association tracing analysis on the multimodal sensor network starting from the multimodal sensor anchor point set, and determine the associated multimodal sensor group set; an asynchronous sampling unit, which is used to determine the operation and maintenance management sampling scale set based on the comparison loss value set, and distribute the operation and maintenance management sampling scale set to the multimodal sensor anchor point set and the associated multimodal sensor group set for asynchronous sampling, and determine the target operation and maintenance management plan according to the asynchronous sampling results.

[0008] Preferably, the water conservancy project operation and maintenance management system based on digital twins also includes: a network retrieval unit, which is used to search the multimodal sensor network according to a preset association range based on the position of the multimodal sensor anchor point set and the water flow direction of the target water conservancy project to determine the associated multimodal sensor group set.

[0009] Preferably, the water conservancy project operation and maintenance management system based on digital twins also includes: a normalization processing unit, used to perform normalization processing on the comparison loss value set respectively to obtain a comparison loss feature normalized value set; an operation and maintenance management coefficient set acquisition unit, used to respectively divide each comparison loss feature normalized value in the comparison loss feature normalized value set by the sum of the comparison loss feature normalized value set to obtain an operation and maintenance management coefficient set; an operation and maintenance management sampling scale set acquisition unit, used to traverse and calculate the difference between the operation and maintenance management coefficient set and 1, and multiply the calculation result by the preset standard sampling scale to obtain the operation and maintenance management sampling scale set.

[0010] Preferably, the water conservancy project operation and maintenance management system based on digital twins also includes: a data acquisition unit, which is used for the multimodal sensor anchor point set and the associated multimodal sensor group set to collect data according to the corresponding operation and maintenance management sampling scale set, and obtain a multimodal sensor anchor point sampling data sequence set and an associated multimodal sensor group sampling data sequence set; an analysis unit, which is used to pre-build an operation and maintenance management scheme identifier, and use the operation and maintenance management scheme identifier to analyze the multimodal sensor anchor point sampling data sequence set and the associated multimodal sensor group sampling data sequence set to obtain the target operation and maintenance management scheme.

[0011] Preferably, the water conservancy project operation and maintenance management system based on digital twin also includes: a preprocessing unit, which is used to use Kalman filtering to preprocess the real-time data collected in the multimodal sensor network in the industrial edge gateway to obtain a preprocessed multimodal sensor preprocessing real-time data set; a transmission unit, which is used to transmit the multimodal sensor preprocessing real-time data set to the digital twin synchronization engine according to the HTTP protocol; a digital twin model construction unit, which is used for the digital twin synchronization engine to construct a digital twin model according to the multimodal sensor preprocessing real-time data set to obtain the real-time digital twin model.

[0012] Preferably, the water conservancy project operation and maintenance management system based on digital twin also includes: a simulation calculation unit, which is used for the digital twin synchronization engine to call the finite element model and the fluid mechanics model, perform simulation calculations on the multimodal sensor preprocessing real-time data set, and obtain the real-time digital twin model.

[0013] Preferably, the water conservancy project operation and maintenance management system based on digital twin also includes: a three-dimensional model construction unit, which is used to use BIM technology to construct a three-dimensional model of the target water conservancy project and set the external load boundary; a force simulation unit, which is used to use ANSYS to mesh the three-dimensional model, and perform force simulation based on the external load boundary to determine the structural stress distribution diagram; a finite element stress hotspot set acquisition unit, which is used to take the points in the structural stress distribution diagram that exceed the preset stress threshold as the finite element stress hotspot set; a water flow simulation unit, which is used to perform water flow simulation based on the three-dimensional model and determine the hydrological sensitive point set.

[0014] Preferably, the water conservancy project operation and maintenance management system based on digital twin also includes: a historical data acquisition unit, used to acquire historical hydrological data and historical meteorological data of the area where the target water conservancy project is located; a flow velocity field distribution map generation unit, used to use the historical hydrological data and historical meteorological data as initial conditions, import the three-dimensional model into water flow simulation software for water flow simulation, and generate a flow velocity field distribution map; a hydrological sensitive point set determination unit, used to determine the hydrological sensitive point set based on the flow velocity field distribution map.

[0015] The technical solution provided in this application has at least the following technical effects or advantages: by realizing the technical goal of linking multimodal sensor intelligent sampling and digital twin synchronous modeling driven by comparison loss, the technical effects of improving the accuracy of key data collection, optimizing system resource allocation, and enhancing the ability to perceive the operating status of water conservancy projects and respond to abnormalities are achieved.

[0016] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, which can be implemented in accordance with the contents of the description, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically listed below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in this application or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely exemplary, and a person of ordinary skill in the art can obtain other drawings based on the provided drawings without creative work.

[0018] Figure 1 This application presents a structural diagram of a water conservancy project operation and maintenance management system based on digital twins.

[0019] Figure 2 This application presents a structural diagram of an asynchronous sampling module in a water conservancy project operation and maintenance management system based on digital twins.

[0020] Explanation of the accompanying drawings: hydrological sensitive point screening module 1, sensor deployment module 2, twin model construction module 3, comparison loss value set determination module 4, reverse mapping module 5, asynchronous sampling module 6, association tracing analysis unit 61, asynchronous sampling unit 62. DETAILED DESCRIPTION

[0021] This application provides a water conservancy project operation and maintenance management system based on digital twins, which solves the technical problem in the existing technology that the sampling frequency and data transmission strategy cannot be adaptively adjusted according to real-time changes due to the lack of dynamic recognition and differentiated processing mechanism for multimodal sensor data features, further affecting the data utilization efficiency, response timeliness and abnormal warning capabilities of the water conservancy project monitoring system. The technical goal of linking multimodal sensor intelligent sampling driven by comparison loss with digital twin synchronous modeling is achieved, achieving the technical effects of improving the accuracy of key data collection, optimizing system resource allocation, and enhancing the ability to perceive the operating status of water conservancy projects and respond to abnormalities.

[0022] Below, the technical solutions in this application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited to the example embodiments described herein. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. It should also be noted that, for the convenience of description, only the parts related to this application, rather than all of them, are shown in the accompanying drawings.

[0023] Please see the attached Figure 1 , this application provides a water conservancy project operation and maintenance management system based on digital twins, specifically including:

[0024] The hydrological sensitive point screening module 1 is used to screen the finite element stress hotspots and hydrological sensitive points of the target water conservancy project to obtain the finite element stress hotspot set and the hydrological sensitive point set.

[0025] Specifically, target water conservancy projects refer to specific engineering entities, such as dams, canals, pumping stations, or gates, which are the objects of monitoring and maintenance by this system. Finite element stress hotspots and hydrologically sensitive points are screened for target water conservancy projects. Finite element stress hotspots are identified by calculating and identifying critical locations within a three-dimensional model where stress concentrations are high and structural fatigue or failure is likely to occur. Finite element analysis discretizes a continuous structure into a finite number of small elements and numerically calculates mechanical responses such as stress and displacement. For example, ANSYS software can be used to simulate the stress state of a dam under varying water pressure and temperature. Next, hydrologically sensitive points must be screened. Hydrologically sensitive points are locations that are sensitive to hydrological changes and are located in areas where water velocity fluctuates dramatically, vortices are prone to form, or are significantly affected by climate. Hydrologically sensitive points are identified through statistical analysis of historical hydrological and meteorological data, followed by dynamic flow simulation using simulation software such as MIKE or Fluent. Finally, we obtain a set of finite element stress hotspots and a set of hydrological sensitive points, that is, a list of all high-risk, key-monitored structural parts and hydrological locations, which determines the key areas for subsequent modeling and data synchronization.

[0026] The sensor deployment module 2 is used to traverse the finite element stress hotspot set and the hydrological sensitive point set to perform multimodal sensor deployment and obtain a deployed multimodal sensor network.

[0027] Specifically, the finite element stress hotspot set and the hydrologically sensitive point set are traversed, meaning each identified key monitoring point is examined and processed in turn, and the environmental conditions and monitoring requirements at each point are determined one by one, which is then used to develop a sensor deployment plan. Finite element stress hotspots are locations where stress values ​​are significantly higher than other areas found in structural analysis, concentrated at structural corners, boundary connections, or locations with sudden changes in stress. Hydrologically sensitive point sets are areas that are particularly sensitive to changes in water flow and are prone to eddies, violent water level fluctuations, or scours.

[0028] Next, multimodal sensor deployment is carried out. Different types of sensors are installed at corresponding locations according to the monitoring objectives of different points. Multimodal sensors refer to a group of sensors with multiple monitoring functions, such as fiber Bragg grating strain gauges for detecting structural strain, piezoelectric accelerometers for recording vibration and shock, ultrasonic water level gauges for measuring water level, RTD temperature probes for measuring temperature, and smart meters for collecting equipment electricity usage data. The deployment of different points is determined by their characteristics. For example, the downstream bottom of the dam body may be equipped with strain gauges and temperature probes, while the upstream spillway may only require a water level gauge.

[0029] With the completion of sensor deployment at all key points, a complete multimodal sensor network is eventually formed. It is a collaborative monitoring system composed of multiple sensor nodes. It has the characteristics of wide distribution, multiple types, and complementary functions. It can collect multi-angle and real-time data on the structural status and operating environment of water conservancy projects.

[0030] The twin model construction module 3 is used to use the industrial edge gateway to push the real-time data collected in the multimodal sensor network to the digital twin synchronization engine to build a real-time digital twin model.

[0031] Specifically, industrial edge gateways are used to process and transmit data collected by multimodal sensor networks. Industrial edge gateways are computing devices deployed on-site that perform pre-processing, protocol conversion, format encapsulation, and other operations on sensor data, while also possessing local computing capabilities. They can perform preliminary data screening and compression locally, reducing communication latency and improving real-time performance and bandwidth efficiency.

[0032] The processed data is then pushed to the digital twin synchronization engine. This module is responsible for building and updating the digital twin model. It receives real-time data from the edge gateway and uses it to synchronize the virtual model with the actual project. The digital twin synchronization engine not only receives data but also dynamically adjusts the model state based on pre-set physical rules, simulation logic, and model parameters. For example, if the vibration data of the dam body uploaded by the edge gateway increases abnormally, the digital twin engine will adjust the structural response of the virtual dam model in real time to reflect potential fatigue risks.

[0033] Ultimately, a real-time digital twin model is constructed. A real-time digital twin model is a digital replica of the project that is continuously updated based on real-time sensor data. It can synchronously reflect the state changes of the project entity at different time points. For example, during flood season, if real-time data indicates a continuous rise in water levels, the water flow simulation in the real-time digital twin model will change accordingly, predicting overflow points and stress spike locations, thereby providing accurate early warning.

[0034] The comparison loss value set determination module 4 is used to perform a comparison loss analysis on the real-time digital twin model and the historical digital twin model constructed synchronously last time to determine the comparison loss value set.

[0035] Specifically, a comparative loss analysis is conducted between the real-time digital twin model and the historical digital twin model. The current digital twin model, generated by sensor data, is compared point by point and parameter by parameter with the model saved at a previous moment. The real-time digital twin model is a virtual replica of the current state of the project, reflecting real-time changes in the structure, environment, and equipment; the historical digital twin model, on the other hand, represents the state at a past moment. The core of the comparative loss analysis is to identify differences between the two models in key physical parameters, such as stress, deformation, water level, and flow rate, to assess whether there are any abnormal changes in project operation.

[0036] Next, all differences found during the comparison are categorized and quantified, generating a data set for analysis and decision-making, and determining a set of comparison loss values. A comparison loss value is the difference between the same position and the same physical quantity at two different times. For example, if the displacement of a measurement point was 4 mm last time and 9 mm now, the comparison loss value is 5 mm.

[0037] The reverse mapping module 5 is configured to perform reverse mapping on the multimodal sensor network based on the comparison loss value set to determine a multimodal sensor anchor point set.

[0038] Specifically, the multimodal sensor network is reverse mapped based on the comparison loss value set. The parameter anomaly information contained in the comparison loss value set is used to trace back to the data source that caused the abnormal change, and identify the sensor that played a key role in the model difference. The comparison loss value set is the difference data between the digital twin models at two moments in time in terms of structural stress, water level, temperature, etc., while the multimodal sensor network is a sensor system deployed at various key points in the water conservancy project, including various types of sensors such as strain gauges, water level gauges, and accelerometers. Reverse mapping refers to the process of inferring the cause from the result. That is, by analyzing the changes in the comparison loss values, the sensor whose measured physical quantity has changed significantly is checked to find the source of the abnormal change.

[0039] Next, the key sensor points identified during the reverse mapping process are aggregated to determine a multimodal sensor anchor set, forming a sensor set for focused observation or response. This multimodal sensor anchor set serves as an anchor for monitoring structural state changes, representing the source of fault initiation or the location of the most active changes.

[0040] The asynchronous sampling module 6 is used to perform asynchronous sampling on the multimodal sensor network according to the multimodal sensor anchor point set and the comparison loss value set, and identify the operation and maintenance management plan according to the asynchronous sampling result to determine the target operation and maintenance management plan.

[0041] Specifically, asynchronous sampling is performed on the multimodal sensor network based on a set of multimodal sensor anchor points and a set of comparison loss values. Based on the identified key sensor points and their corresponding loss levels, the sampling frequency and timing of each sensor are determined, enabling targeted data collection at non-uniform time intervals. Asynchronous sampling means that each sensor no longer collects data simultaneously or at a fixed frequency, but instead dynamically adjusts the sampling strategy based on its importance and rate of change.

[0042] Next, the operation and maintenance management plan is identified based on the asynchronous sampling results. The collected asynchronous data is input into a pre-built operation and maintenance management identification model to analyze and determine the project's operating status. Relying on machine learning or rule-matching techniques, the model can classify the current structural status as normal, slightly abnormal, or severely abnormal. The identification model comprehensively considers multiple parameters such as stress, water level, and vibration, and combines historical experience and expert knowledge to output a set of recommendations or decision paths. Finally, the target operation and maintenance management plan is determined, and a set of targeted project operation and maintenance strategies are recommended, including measures such as on-site manual inspections, activation of backup pumping stations, slope reinforcement, and adjustment of sluice openings. The goal is to intervene promptly in potential risks and prevent further deterioration.

[0043] Furthermore, the present application also includes: the multimodal sensor is selected from at least one of the following sensors: a fiber grating strain gauge, a piezoelectric accelerometer, an ultrasonic water level meter, an RTD temperature probe, and a smart meter.

[0044] Specifically, the sensors deployed in water conservancy project monitoring are not of a single type, but include a variety of devices that can sense different physical quantities. Multimodal sensors are selected from at least one of the following sensors, and then one or more of them are flexibly selected and combined according to actual monitoring needs. This helps to obtain information on structural and environmental changes from multiple dimensions, thereby building a more comprehensive and detailed digital twin model.

[0045] A fiber Bragg grating (FBG) strain gauge is a sensor that uses wavelength variations in light transmitted through an optical fiber to measure strain. Due to the sensitivity of the grating's reflected wavelength to stress changes, it is used to monitor the deformation of concrete or steel structures, making it suitable for high-precision and long-distance monitoring. For example, a fiber Bragg grating (FBG) strain gauge can be embedded within a dam structure for continuous stress observation. A piezoelectric accelerometer is a device that measures acceleration by generating an electric charge when subjected to force. It is used for vibration detection and can monitor the dynamic response of engineering structures to external disturbances such as earthquakes and water hammer. It is installed in locations susceptible to impact, such as bridge piers and gates. An ultrasonic water level gauge calculates water level by measuring the time it takes for an ultrasonic wave to be emitted from the sensor and reflected back from the water surface. It is unaffected by water quality and is used to monitor water level changes in rivers, reservoirs, and gates. An RTD temperature probe, or resistance temperature detector, senses temperature by measuring the change in metal resistance as temperature changes. It is used to monitor the temperature distribution of water bodies, structural materials, or equipment during operation. For example, an RTD probe can be deployed in a pump room to track motor heating. A smart meter is an electric power monitoring device with data recording, communication and remote control functions. It can not only record electricity consumption in real time, but also report abnormal power changes. It is used to determine whether electrical equipment such as water pumps and gates are in normal operation, and plays a role in energy consumption analysis and equipment health assessment.

[0046] Further, if Figure 2 As shown, the present application also includes: an association tracing analysis unit 61, which is used to perform association tracing analysis on the multimodal sensor network starting from the multimodal sensor anchor point set, and determine an associated multimodal sensor group set; an asynchronous sampling unit 62, which is used to determine an operation and maintenance management sampling scale set based on the comparison loss value set, and distribute the operation and maintenance management sampling scale set to the multimodal sensor anchor point set and the associated multimodal sensor group set for asynchronous sampling, and determine a target operation and maintenance management plan according to the asynchronous sampling result.

[0047] Specifically, starting with the multimodal sensor anchor point set and starting from the key sensor points identified in the previous analysis, the multimodal sensor network is subjected to correlation tracing analysis. In the entire sensor network, combined with the physical location of the anchor point, water flow direction, structural topology and other information, other sensor nodes with correlation, similar trends or causal relationships in the monitoring data with the multimodal sensor anchor point are found, and the associated multimodal sensor group set is determined to facilitate the subsequent expansion of the monitoring area in space and achieve more comprehensive data analysis.

[0048] Based on the comparison loss value set, the O&M sampling scale set is determined. The deviation between the real-time and historical states of the digital twin model is quantified as a comparison loss value. This value is then used to calculate the required sampling frequency, data accuracy, or sensor activation strategy for each monitoring area. A larger loss value indicates more dramatic or abnormal state changes in the area or device, requiring a finer sampling scale, and thus requiring more frequent or accurate monitoring. The O&M sampling scale set refers to the combination of sampling scales required for different locations or sensor points in the sensor network. This can be achieved by varying time intervals or data resolution. The O&M sampling scale set is distributed to the multimodal sensor anchor point set and the associated multimodal sensor group set for asynchronous sampling. Instead of collecting data at a uniform frequency, each point updates data according to its own dynamic sampling scale, saving energy while focusing on abnormal areas. Subsequently, pattern analysis based on the asynchronous sampling results determines the targeted O&M management plan, such as whether to implement structural reinforcement, flood discharge regulation, or equipment maintenance. Table 1 shows a partial record of the most recent associated retrospective analysis and asynchronous sampling management.

[0049] Table 1: Partial records of the most recent correlation tracing analysis and asynchronous sampling management

[0050] Sensor number Is it an anchor sensor? Associated traceability sensor group Comparison loss value Operation and maintenance management coefficient Summary of Asynchronous Sampled Data Characteristics Whether to trigger an operation and maintenance response S001 yes S002,S003 0.75 5 Large strain fluctuation (>5%) yes S002 no - 0.42 10 Stable fluctuation (<1%) no S003 no - 0.66 6 Small changes (2%-3%) no S004 yes S005 0.88 4 Vibration signal peak is significant yes S005 no - 0.39 12 Normal current fluctuations no .

[0051] Furthermore, the present application also includes: a network retrieval unit, which is used to search the multimodal sensor network according to a preset association range based on the position of the multimodal sensor anchor point set and the water flow direction of the target water conservancy project to determine the associated multimodal sensor group set.

[0052] Specifically, the relationships between sensors are determined based on the locations of the multimodal sensor anchor points and the flow direction of the target hydraulic project. This means that not only spatial locations but also the direction and path of water flow in the actual project are considered when determining the relationships between related sensors. For example, in a dam or spillway, where water flows from upstream to downstream, changes in water levels sensed by upstream anchor points may affect stress in downstream structures.

[0053] Searching a multimodal sensor network according to a preset association range means limiting the search scope to a pre-set spatial or physical influence range when identifying associated sensor groups. The preset association range is a specific distance threshold, such as searching for other sensors within a 500-meter radius around an anchor point, or it can be used to determine the influence transmission chain based on the logical order of the water flow path. This allows efficient screening of nodes directly or indirectly associated with key anchor points throughout the sensor network. Ultimately, a set of associated multimodal sensor groups is determined: a set of sensor points near the anchor point that are relevant to the analysis, selected based on water flow direction and structural connectivity, and that can collaboratively provide more comprehensive operational status information.

[0054] Furthermore, the present application also includes: a normalization processing unit, which is used to perform normalization processing on the comparison loss value set respectively to obtain a comparison loss feature normalized value set; an operation and maintenance management coefficient set acquisition unit, which is used to divide each comparison loss feature normalized value in the comparison loss feature normalized value set by the sum of the comparison loss feature normalized value set to obtain an operation and maintenance management coefficient set; an operation and maintenance management sampling scale set acquisition unit, which is used to traverse and calculate the difference between the operation and maintenance management coefficient set and 1, and multiply the calculation result by a preset standard sampling scale to obtain the operation and maintenance management sampling scale set.

[0055] Specifically, the comparison loss value set is normalized, standardizing the comparison loss values ​​generated by each sensor monitoring point to a range between 0 and 1, facilitating comparison and subsequent calculations. This normalization process uses a maximum and minimum normalization method, which helps to quantify and analyze loss values ​​of different orders of magnitude on the same scale.

[0056] Next, each normalized value in the normalized value set of the comparison loss features is divided by the sum of the normalized values ​​of the comparison loss features to obtain the set of operation and maintenance management coefficients. The ratio operation is a weighted distribution strategy used to measure the relative contribution of each sensor point to the overall anomaly. For example, if the normalized value set contains three values: 0.2, 0.3, and 0.5, and their sum is 1, then the operation and maintenance management coefficients are 0.2, 0.3, and 0.5, respectively. The set of operation and maintenance management coefficients reflects the locations with the most significant deviations, facilitating subsequent precise sampling and resource allocation.

[0057] Finally, the difference between the operation and maintenance management coefficient set and 1 is traversed and calculated, and the calculation result is multiplied by the preset standard sampling scale to obtain the operation and maintenance management sampling scale set, reflecting the importance of each monitoring point in the actual sampling frequency or accuracy. The larger the difference, the smaller the deviation of the point and the lower the required sampling density; otherwise, encrypted sampling is required.

[0058] Furthermore, the present application also includes: a data acquisition unit, configured to perform data acquisition on the multimodal sensor anchor point set and the associated multimodal sensor group set according to the corresponding operation and maintenance management sampling scale set, to obtain a multimodal sensor anchor point sampling data sequence set and an associated multimodal sensor group sampling data sequence set; and an analysis unit, configured to pre-build an operation and maintenance management solution identifier, and use the operation and maintenance management solution identifier to analyze the multimodal sensor anchor point sampling data sequence set and the associated multimodal sensor group sampling data sequence set to obtain the target operation and maintenance management solution.

[0059] Specifically, the multimodal sensor anchor point set and the associated multimodal sensor group set each collect data according to the corresponding operation and maintenance management sampling scale set. Based on the calculated sampling frequency requirements (i.e., the operation and maintenance management sampling scale set), environmental or structural data is acquired in a non-fixed periodic manner. A multimodal sensor is a device capable of collecting different physical quantities, such as strain, acceleration, water level, and temperature simultaneously. A sampled data sequence set is a collection of multiple sets of data continuously acquired by a sensor within a specific time interval, forming the time series input for analysis. For example, an anchor point sensor may sample every two minutes, while some sensors in the associated group may sample every five minutes. This results in two sets of data sequences with different densities, reflecting the varying importance of each monitoring point in the project operation.

[0060] Next, a pre-built operation and maintenance management solution identifier is an intelligent model or algorithm module designed and trained before system deployment. Its task is to receive sensor data and determine whether an operation and maintenance action is currently required. This identifier can use technologies such as rule engines, support vector machines, decision trees, or neural networks. Based on historical data, it trains a set of recognition logic to determine whether anomalies have occurred or whether operation and maintenance thresholds have been reached.

[0061] The O&M solution identifier then analyzes the data sequences from the multimodal sensor anchor points and the associated multimodal sensor groups. Using previously collected data sequences as input, the O&M solution identifier automatically calculates and identifies features such as trends, mutations, and abnormal amplitudes. The analysis may include indicators such as stress growth rate, water level change slope, and vibration frequency fluctuations. Through cross-comparison, it can determine whether there are structural risks, hydrological risks, or equipment degradation.

[0062] Finally, the target operation and maintenance management plan is derived through analysis. The identifier outputs a specific response plan tailored to the current project status, such as recommendations for local structural reinforcement, drainage optimization, or increased monitoring point density. This target operation and maintenance management plan is not a static, standardized process; rather, it is a dynamically generated strategy based on real-time data characteristics and historical model deviations.

[0063] Furthermore, the present application also includes: a preprocessing unit, which is used to use Kalman filtering to preprocess the real-time data collected in the multimodal sensor network in the industrial edge gateway to obtain a preprocessed multimodal sensor preprocessing real-time data set; a transmission unit, which is used to transmit the multimodal sensor preprocessing real-time data set to the digital twin synchronization engine according to the HTTP protocol; a digital twin model construction unit, which is used for the digital twin synchronization engine to construct a digital twin model according to the multimodal sensor preprocessing real-time data set to obtain the real-time digital twin model.

[0064] Specifically, the industrial edge gateway uses Kalman filtering to preprocess the real-time data collected from the multimodal sensor network. This refers to using the Kalman filter algorithm on an edge computing device near the sensor deployment location to remove noise and perform estimation corrections on the raw data transmitted by the sensors. Kalman filtering is a recursive filtering algorithm based on a state-space model. It is suitable for real-time data estimation in dynamic systems and can effectively reduce data fluctuations caused by sensor errors, external disturbances, and other factors. Industrial edge gateways are edge computing devices used in industrial scenarios. They process data close to the data source, significantly reducing communication latency. A multimodal sensor network refers to a monitoring system composed of multiple types of sensors (such as fiber Bragg grating strain gauges, accelerometers, and water level gauges) that can simultaneously acquire information from multiple physical dimensions. The preprocessed real-time data set of multimodal sensors is a filtered, smoother, and more reliable data set for use by downstream modules.

[0065] Next, the industrial edge gateway transmits the pre-processed real-time data set from the multimodal sensors to the digital twin synchronization engine using the HTTP protocol. The edge gateway uses the Hypertext Transfer Protocol (HTTP) to package the data and send it to the synchronization module in the remote server. HTTP is an application-layer communication protocol based on a request-and-response model, used for data exchange between clients and servers. The digital twin synchronization engine is the core module used to build and update digital twin models. It receives real-time data collected by sensors and maps it to a digital simulation system corresponding to the real system, ensuring consistency in state and behavior between the virtual model and the physical system.

[0066] The digital twin synchronization engine then constructs a digital twin model based on the real-time data set pre-processed by multimodal sensors. This real-time digital twin model is then updated or recreated in real time using filtered sensor data from the field, enabling the model to dynamically reflect the operational status of the real project. A digital twin model is a virtual object that integrates structural and behavioral models with data-driven algorithms. It fully replicates the status, changing trends, and potential risks of the physical project in a digital space, facilitating remote monitoring, simulation prediction, and intelligent analysis.

[0067] Furthermore, the present application also includes: a simulation calculation unit, which is used for the digital twin synchronization engine to call the finite element model and the fluid mechanics model, perform simulation calculations on the multimodal sensor preprocessing real-time data set, and obtain the real-time digital twin model.

[0068] Specifically, in the process of building a real-time digital twin model, the digital twin synchronization engine accesses finite element models and fluid dynamics models. The finite element model is a mathematical model that discretizes complex structures into many small units for numerical analysis. It is used to simulate the stress, deformation, and heat conduction behaviors of physical structures. The fluid dynamics model is used to describe the flow characteristics of water or gas under different boundaries and conditions, including water level changes, flow velocity distribution, and water pressure changes. The digital twin synchronization engine is the system's core computing module, responsible for inputting real-time sensor data into the simulation model and generating a dynamically updated digital representation.

[0069] Next, multimodal sensor preprocessing real-time data sets refer to high-quality data after noise removal using methods such as Kalman filtering in edge devices. This includes data on multiple dimensions, such as strain, water level, flow rate, and temperature. Simulation calculations use these preprocessed real-time data sets as boundary conditions or loads and input them into finite element and fluid models. Numerical calculations simulate the stress state and hydrodynamic response of actual water conservancy projects, ensuring that large amounts of calculations can be completed in a short period of time.

[0070] Finally, a real-time digital twin model is obtained, creating a virtual model that is highly synchronized with the physical entity. It not only contains structural information and historical status, but also reflects the current operating status based on real-time data and has the ability to predict future status.

[0071] Furthermore, the present application also includes: a three-dimensional model construction unit, which is used to use BIM technology to construct a three-dimensional model of the target water conservancy project and set an external load boundary; a force simulation unit, which is used to use ANSYS to mesh the three-dimensional model, and perform force simulation based on the external load boundary to determine the structural stress distribution diagram; a finite element stress hotspot set acquisition unit, which is used to take the points in the structural stress distribution diagram that exceed the preset stress threshold as the finite element stress hotspot set; a water flow simulation unit, which is used to perform water flow simulation based on the three-dimensional model and determine the hydrological sensitive point set.

[0072] Specifically, BIM technology is used to construct a three-dimensional model of the target water conservancy project, and building information modeling technology is used to digitally model the water conservancy project, generating a three-dimensional visual model that includes geometric structure, physical properties, and construction details. BIM, or Building Information Modeling, is a technical means of integrating design and construction information. In water conservancy projects, it can not only represent the structure of components such as dams, gates, and waterways, but also simulate the external environment. Setting external load boundaries refers to defining external physical conditions that may be applied to the engineering structure, such as water flow pressure, soil support, wind force, and gravity, during the modeling process, providing a basis for subsequent simulation analysis.

[0073] Next, the 3D model was meshed using ANSYS, breaking the complex structure into many small cells. This allowed the finite element software ANSYS to accurately calculate the numerical solution for each small cell. ANSYS is an industrial-grade engineering simulation tool used for stress analysis, thermal analysis, and multi-physics field simulation. The level of meshing directly affects the accuracy of the simulation results. Subsequently, a stress simulation was performed based on the external load boundary. The load was input into the mesh model, and by calculating the stress response between each node, a structural stress distribution diagram was obtained, showing the distribution of forces experienced by various parts of the project under the action of the external load.

[0074] Then, points in the structural stress distribution map that exceed a preset stress threshold are identified as finite element stress hotspots. Areas or points exceeding the safety upper limit or critical value are identified from the stress distribution map to form a finite element stress hotspot set. The stress threshold is the maximum allowable stress value set in advance based on material strength, structural specifications, or safety factors. Areas exceeding this threshold may pose a risk of damage and require special attention and sensor deployment.

[0075] Finally, water flow simulation is performed based on the three-dimensional model to determine the set of hydrological sensitive points. The three-dimensional BIM model is imported into the hydrodynamic simulation software. By simulating the flow state of water inside and outside the structure, the locations most sensitive to hydrological changes are identified, such as flow velocity mutation points, water level sudden change areas or backflow convergence areas.

[0076] Furthermore, the present application also includes: a historical data acquisition unit, used to acquire historical hydrological data and historical meteorological data of the area where the target water conservancy project is located; a flow velocity field distribution map generation unit, used to use the historical hydrological data and historical meteorological data as initial conditions, import the three-dimensional model into water flow simulation software for water flow simulation, and generate a flow velocity field distribution map; a hydrological sensitive point set determination unit, used to determine the hydrological sensitive point set based on the flow velocity field distribution map.

[0077] Specifically, historical hydrological and meteorological data for the area where the target water conservancy project is located is obtained, collecting hydrological and meteorological information related to the project over a period of time. Hydrological data includes numerical values ​​such as water level changes, flow, and rainfall that reflect the characteristics and dynamics of the water body, while meteorological data includes climatic factors such as temperature, rainfall intensity, and wind speed that affect the water environment. Historical data typically comes from hydrological and meteorological stations and serves as the foundation for subsequent analysis.

[0078] Next, historical hydrological and meteorological data were used as initial conditions and fed into the water flow simulation software, ensuring that the simulation process reflects past environmental conditions. The software then combined this input data with the 3D model to generate a velocity field distribution map. The 3D model is a digital representation of the project's structure, while the velocity field distribution map displays the magnitude and direction of water flow at different locations within the water body, reflecting the hydrodynamic characteristics and patterns of change.

[0079] Then, based on the velocity field distribution map, a set of hydrologically sensitive points was identified. Areas with significant velocity variations or complex hydrodynamic characteristics were analyzed to identify key points most sensitive to hydrological changes. Hydrologically sensitive points are locations where velocity changes suddenly occur or where eddies are concentrated, making them prone to flow disturbances, scouring, or siltation. Therefore, they are of great significance to project safety and operational management.

[0080] To sum up, the water conservancy project operation and maintenance management system based on digital twins provided by this application has the following technical effects: by realizing the technical goal of linking multimodal sensor intelligent sampling driven by comparison loss with digital twin synchronous modeling, the technical effects of improving the accuracy of key data collection, optimizing system resource allocation, and enhancing the ability to perceive the operating status of water conservancy projects and respond to abnormalities are achieved.

[0081] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

[0082] Obviously, those skilled in the art may make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A water conservancy project operation and maintenance management system based on digital twin, characterized by: The system comprises: The hydrological sensitive point screening module is used to screen the finite element stress hotspots and hydrological sensitive points of the target water conservancy project to obtain the finite element stress hotspot set and the hydrological sensitive point set; A sensor deployment module, configured to traverse the finite element stress hotspot set and the hydrological sensitive point set to perform multimodal sensor deployment and obtain a deployed multimodal sensor network; A twin model construction module is used to push the real-time data collected in the multimodal sensor network to the digital twin synchronization engine using the industrial edge gateway to build a real-time digital twin model; A comparison loss value set determination module is used to perform a comparison loss analysis on the real-time digital twin model and the historical digital twin model constructed synchronously last time to determine a comparison loss value set; a reverse mapping module, configured to reverse map the multimodal sensor network based on the comparison loss value set to determine a multimodal sensor anchor point set; an asynchronous sampling module, configured to perform asynchronous sampling on the multimodal sensor network according to the multimodal sensor anchor point set and the comparison loss value set, and identify an operation and maintenance management solution based on the asynchronous sampling result to determine a target operation and maintenance management solution; an association tracing analysis unit, configured to perform an association tracing analysis on the multimodal sensor network starting from the multimodal sensor anchor point set, and determine an associated multimodal sensor group set; an asynchronous sampling unit, configured to determine an operation and maintenance management sampling scale set based on the comparison loss value set, distribute the operation and maintenance management sampling scale set to the multimodal sensor anchor point set and the associated multimodal sensor group set for asynchronous sampling, and determine a target operation and maintenance management solution based on the asynchronous sampling results; A normalization processing unit, configured to perform normalization processing on each of the comparison loss value sets to obtain a comparison loss feature normalized value set; an operation and maintenance management coefficient set obtaining unit, configured to obtain an operation and maintenance management coefficient set by dividing each comparison loss feature normalized value in the comparison loss feature normalized value set by the sum of the comparison loss feature normalized value set; The operation and maintenance management sampling scale set obtaining unit is configured to traverse and calculate the difference between the operation and maintenance management coefficient set and 1, and multiply the calculated result by a preset standard sampling scale to obtain the operation and maintenance management sampling scale set.

2. A water conservancy project operation and maintenance management system based on digital twins according to claim 1, characterized in that: The multimodal sensor is selected from at least one of the following sensors: a fiber grating strain gauge, a piezoelectric accelerometer, an ultrasonic water level gauge, an RTD temperature probe, and a smart meter.

3. A water conservancy project operation and maintenance management system based on digital twins according to claim 1, characterized in that: The network retrieval unit is used to search the multimodal sensor network according to a preset association range based on the position of the multimodal sensor anchor point set and the water flow direction of the target water conservancy project to determine the associated multimodal sensor group set.

4. A water conservancy project operation and maintenance management system based on digital twins according to claim 1, characterized in that: include: a data acquisition unit configured to acquire data from the multimodal sensor anchor point set and the associated multimodal sensor group set according to the corresponding operation and maintenance management sampling scale set, and obtain a multimodal sensor anchor point sampling data sequence set and an associated multimodal sensor group sampling data sequence set; An analysis unit is configured to pre-build an operation and maintenance management solution identifier, and use the operation and maintenance management solution identifier to analyze the multimodal sensor anchor point sampling data sequence set and the associated multimodal sensor group sampling data sequence set to obtain the target operation and maintenance management solution.

5. The water conservancy project operation and maintenance management system based on digital twin according to claim 1, characterized in that: include: A preprocessing unit, configured to preprocess the real-time data collected in the multimodal sensor network in the industrial edge gateway using a Kalman filter to obtain a preprocessed multimodal sensor preprocessed real-time data set; A transmission unit, configured to transmit the multimodal sensor pre-processed real-time data set to the digital twin synchronization engine by the industrial edge gateway according to the HTTP protocol; A digital twin model construction unit is used for the digital twin synchronization engine to construct a digital twin model according to the real-time data set preprocessed by the multimodal sensor to obtain the real-time digital twin model.

6. A water conservancy project operation and maintenance management system based on digital twins according to claim 5, characterized in that: A simulation calculation unit is used for the digital twin synchronization engine to call the finite element model and the fluid mechanics model, perform simulation calculations on the multimodal sensor preprocessing real-time data set, and obtain the real-time digital twin model.

7. The water conservancy project operation and maintenance management system based on digital twin according to claim 1, characterized in that: include: A 3D model construction unit is used to construct a 3D model of the target water conservancy project using BIM technology and set external load boundaries; A stress simulation unit, configured to mesh the three-dimensional model using ANSYS, perform stress simulation based on the external load boundary, and determine a structural stress distribution diagram; a finite element stress hotspot set obtaining unit, configured to take points in the structural stress distribution diagram that exceed a preset stress threshold as the finite element stress hotspot set; A water flow simulation unit is used to perform water flow simulation based on the three-dimensional model and determine a set of hydrological sensitive points.

8. A water conservancy project operation and maintenance management system based on digital twins according to claim 7, characterized in that: include: A historical data acquisition unit, configured to acquire historical hydrological data and historical meteorological data of the area where the target water conservancy project is located; a flow velocity field distribution diagram generating unit, configured to use the historical hydrological data and historical meteorological data as initial conditions, import the three-dimensional model into water flow simulation software to perform water flow simulation, and generate a flow velocity field distribution diagram; The hydrological sensitive point set determining unit is configured to determine the hydrological sensitive point set based on the velocity field distribution map.

Images