Water conservancy information automation and security control method and system

Abstract

The invention belongs to the technical field of water conservancy informatization, and provides a water conservancy information automation and safety control method and system, and the method comprises the following steps: S1, collecting the water level, flow, water quality, weather and engineering structure safety data in real time through an Internet of Things sensing device disposed at a water conservancy project site; s2, on the basis of historical disaster data, real-time monitoring data and hydraulic engineering characteristic parameters, constructing a digital twinborn model containing four-dimensional linkage of'forecasting-early warning-rehearsing-plan '; s3, according to a risk assessment result, a control instruction is sent to an execution mechanism through the Internet of Things platform; by integrating advanced technologies such as the Internet of Things, edge calculation, deep learning, digital twinning and reinforcement learning, automatic acquisition, processing, analysis and intelligent decision making of water conservancy information are realized, the real-time monitoring, early warning and emergency response capabilities of a water conservancy project are remarkably improved, safe operation of the project is ensured, resource allocation is optimized, and the working efficiency is improved. And the management efficiency is improved.

Description

Technical Field

[0001] This invention belongs to the field of water conservancy information technology, specifically a method and system for automated and secure control of water conservancy information. Background Technology

[0002] In the field of water conservancy informatization, traditional water conservancy project management methods lack comprehensive analysis models based on historical disaster data and real-time monitoring data. Traditional methods suffer from delays in disaster early warning, unclear warning signal classifications, and consequently, untimely or inaccurate response measures. Especially in the event of sudden emergencies, they cannot achieve remote automatic control and intelligent coordinated response of water conservancy facilities, failing to meet the high requirements of modern water conservancy project management for real-time performance, accuracy, and intelligence.

[0003] To address these issues, those skilled in the art have proposed an automated and secure control method and system for water conservancy information. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides an automated and safe control method and system for water conservancy information, resolving the issues of lag in disaster early warning, unclear warning signal classification, and untimely or inaccurate response measures in existing technologies. Particularly in the event of sudden emergencies, it is impossible to achieve remote automatic control and intelligent coordinated response of water conservancy facilities, making it difficult to meet the high requirements of modern water conservancy project management for real-time performance, accuracy, and intelligence.

[0005] In a first aspect, the present invention provides a method for automated and secure control of water conservancy information, comprising the following steps:

[0006] S1. First, by deploying IoT sensing devices (including water level gauges, rain gauges, flow meters, multi-parameter water quality sensors, video surveillance cameras, and satellite remote sensing receivers) at the water conservancy project site, real-time data on water level, flow rate, water quality, meteorology, and engineering structure safety are collected. Then, edge computing nodes are used to clean, calibrate, and perform preliminary analysis on the raw data.

[0007] S2. Based on historical disaster data, real-time monitoring data, and water conservancy project characteristic parameters, a digital twin model is constructed that integrates four dimensions: forecasting, early warning, rehearsal, and contingency plan.

[0008] The forecasting module uses deep learning algorithms to make rolling predictions of rainfall and flood events;

[0009] The early warning module triggers tiered early warning signals by setting multiple threshold levels (including water level exceeding limits and abnormal seepage pressure);

[0010] The simulation module uses 3D simulation technology to simulate the engineering safety status and disaster evolution path under different working conditions;

[0011] The contingency plan module automatically generates an emergency dispatch plan based on the rehearsal results and pushes it to the decision-making terminal;

[0012] S3. Based on the risk assessment results, control commands are sent to the actuators (including gate controllers, water pump frequency converters, and drone inspection systems) through the IoT platform to achieve:

[0013] Remote automatic control of water conservancy engineering facilities;

[0014] Intelligent and coordinated response to emergencies (including activating backup power and closing valves in dangerous areas);

[0015] Dynamic optimization and allocation of inspection tasks (based on AI path planning algorithms to reduce blind spots in manual inspections).

[0016] Specific operating steps:

[0017] S1: Data Acquisition and Preliminary Processing

[0018] 1) Deploy IoT sensing devices:

[0019] IoT sensing devices such as water level gauges, rain gauges, flow meters, multi-parameter water quality sensors, video surveillance cameras, and satellite remote sensing receivers are deployed at water conservancy project sites.

[0020] Ensure that the equipment covers key monitoring points and can comprehensively collect data on water level, flow rate, water quality, meteorology, and engineering structure safety.

[0021] 2) Real-time data acquisition:

[0022] The IoT sensing devices are activated to begin collecting various types of water conservancy data in real time.

[0023] The data includes, but is not limited to, water level, flow rate, water quality indicators (including pH value, dissolved oxygen, etc.), meteorological conditions (including rainfall, wind speed, etc.), and the safety status of the engineering structure (including stress, deformation, etc.).

[0024] 3) Edge computing node processing:

[0025] The raw data is cleaned by using edge computing nodes deployed on-site to remove outliers and noise.

[0026] The data is calibrated to ensure its accuracy and consistency.

[0027] A preliminary analysis is conducted to extract key features and information, providing a foundation for subsequent processing.

[0028] S2: Building a digital twin model

[0029] 4) Collect historical and real-time data:

[0030] Integrate historical disaster data, real-time monitoring data, and water conservancy project characteristic parameters (including project scale, design standards, etc.).

[0031] 5) Construct a four-dimensional linkage model of "forecast-early warning-drill-contingency plan":

[0032] Forecasting module: Employs deep learning algorithms (including a hybrid LSTM-Transformer architecture) to perform rolling forecasts of rainfall and flood events, achieving a rainfall forecast error of ≤5% for the next 72 hours.

[0033] Early warning module: Set multiple threshold levels (including water level exceeding limits, abnormal seepage pressure, etc.), and trigger a graded early warning signal when the monitored data exceeds the threshold.

[0034] Pre-simulation module: Utilizes 3D simulation technology to simulate the engineering safety status and disaster evolution path under different working conditions, and evaluates the engineering response and stability under different conditions.

[0035] Contingency plan module: Automatically generates emergency dispatch plans based on the results of the rehearsal, including operations such as opening / closing gates and starting / stopping water pumps, and pushes the plans to the decision-making terminal for management personnel to review.

[0036] S3: Intelligent Decision-Making and Remote Control

[0037] 6) Risk assessment and decision-making:

[0038] Based on the forecast, early warning, and simulation results of the digital twin model, a risk assessment is conducted to determine the main risks and potential impacts currently facing the project.

[0039] Develop corresponding countermeasures and scheduling strategies to ensure project safety and operational efficiency.

[0040] 7) Send control commands:

[0041] Control commands are sent to actuators (including gate controllers, water pump frequency converters, drone inspection systems, etc.) through an IoT platform.

[0042] To achieve remote automatic control of water conservancy facilities, including automatically adjusting the gate opening and starting drainage pumps to lower the reservoir water level based on water level changes.

[0043] 8) Intelligent and coordinated response to sudden emergencies:

[0044] When a sudden emergency is detected (including when the seepage pressure of a reservoir dam exceeds the safety threshold), the intelligent linkage response process is automatically triggered.

[0045] This includes, but is not limited to, closing the inlet gate, starting the drainage pump, dispatching drones to capture high-definition videos of the leak points and uploading them to the cloud platform, and sending early warning information to management personnel through multiple channels.

[0046] 9) Dynamic optimization and allocation of inspection tasks:

[0047] Based on AI path planning algorithms, the allocation of inspection tasks is dynamically optimized, reducing blind spots in manual inspections, improving inspection efficiency, and expanding coverage.

[0048] Ensure that all parts of the project are monitored and maintained in a timely and effective manner.

[0049] Preferably, in step S1, the edge computing node is configured with a dynamic power consumption strategy:

[0050] During low-risk periods, the system enters a dormant state, with the edge gateway simulating sensor values ​​based on historical data. When the federated learning platform predicts that the probability of an anomaly exceeds the threshold, it wakes up the corresponding regional nodes to perform high-precision sampling.

[0051] Specifically, in step S1, when the edge computing node is configured with a dynamic power consumption strategy, the following sequential operation steps can be designed to optimize energy consumption and improve the efficiency and accuracy of data acquisition:

[0052] Operation steps of dynamic power consumption strategy for edge computing nodes

[0053] 1) Initialization and Configuration

[0054] Deploy edge computing nodes: Deploy edge computing nodes with dynamic power management capabilities at the water conservancy project site to ensure that they are correctly connected to IoT sensing devices (including water level gauges, rain gauges, flow meters, etc.).

[0055] Configure dynamic power consumption strategy parameters: set the criteria for determining low-risk periods, the threshold conditions for sleep and wake-up, the rules for historical data simulation, and the anomaly prediction threshold of the federated learning platform.

[0056] 2) Operations during low-risk periods

[0057] Enter hibernation mode:

[0058] Edge computing nodes continuously monitor the current risk level and automatically enter a sleep state to reduce energy consumption when the risk level is determined to be low.

[0059] In sleep mode, edge computing nodes suspend high-precision data sampling, and instead, edge gateways simulate sensor values ​​based on historical data to maintain basic data monitoring functions.

[0060] Historical data simulation:

[0061] The edge gateway uses stored historical data to simulate the current sensor values ​​according to preset algorithms and rules.

[0062] Simulated data is used to maintain basic system operational status monitoring, but not for advanced analysis and decision-making.

[0063] 3) Anomaly prediction and node wake-up

[0064] Federated learning platform predicts:

[0065] The federated learning platform continuously analyzes data from multiple edge computing nodes and IoT sensing devices, using machine learning algorithms to predict the probability of anomalies in the near future.

[0066] When the predicted probability of an anomaly in a certain area exceeds a preset threshold, a wake-up signal is immediately sent to the edge computing node corresponding to that area.

[0067] Wake up edge computing nodes:

[0068] The edge computing node that receives the wake-up signal resumes from its dormant state and begins high-precision data sampling.

[0069] High-precision sampling data is used for more detailed analysis and decision support, ensuring a rapid response in abnormal situations.

[0070] 4) High-precision sampling and data processing

[0071] High-precision data sampling:

[0072] Once awakened, the edge computing node performs high-precision sampling of the raw data collected by IoT sensing devices, including but not limited to key indicators such as water level, flow rate, and water quality.

[0073] After being cleaned, calibrated, and preliminarily analyzed, the sampled data provides accurate input for subsequent digital twin model construction and risk assessment.

[0074] Data processing and analysis:

[0075] Edge computing nodes further process and analyze high-precision sampled data to extract key features and information.

[0076] The analysis results are uploaded to a cloud platform or local server for use by the digital twin model and other advanced analytics functions.

[0077] 5) Continuous monitoring and strategy adjustment

[0078] Continuous monitoring of risk levels:

[0079] After being woken up, the edge computing node continuously monitors the current risk level and dynamically adjusts its working status based on real-time data.

[0080] When the risk level drops to low risk again, edge computing nodes can re-enter a dormant state to save energy.

[0081] Strategy Adjustment and Optimization:

[0082] Based on system operation and actual results, regularly evaluate and adjust the parameters and rules of the dynamic power consumption strategy.

[0083] Optimize key parameters such as historical data simulation algorithms, anomaly prediction models, and wake-up thresholds to improve the overall performance and energy efficiency of the system.

[0084] Preferably, the deep learning algorithm adopts an LSTM-Transformer hybrid architecture, combining meteorological radar data and ground monitoring station data, to achieve a rainfall prediction error of ≤5% for the next 72 hours.

[0085] Specific operating steps:

[0086] 1) Data collection and integration

[0087] Weather radar data collection:

[0088] Obtain high-resolution weather radar data, including parameters such as reflectivity factor and radial velocity, from meteorological departments or self-built weather radar stations.

[0089] Radar data is preprocessed, including noise reduction, calibration, and format conversion, to ensure data quality.

[0090] Ground monitoring station data collection:

[0091] Real-time meteorological data such as rainfall, temperature, and humidity are obtained from ground meteorological monitoring stations located in various locations.

[0092] The data from ground monitoring stations are cleaned and verified to remove outliers and erroneous data.

[0093] Data integration:

[0094] The meteorological radar data and the ground monitoring station data are synchronized in time and matched spatially to form a unified dataset.

[0095] Feature extraction is performed on the integrated data, including time series features and spatial distribution features, to provide input for subsequent model training.

[0096] 2) Constructing an LSTM-Transformer hybrid architecture model

[0097] LSTM layer construction:

[0098] Design an LSTM network structure to capture long-term dependencies and dynamic changes in time series data.

[0099] Determine the hyperparameters of the LSTM layer, such as the number of hidden units, the number of layers, and the activation function.

[0100] Transformer layer construction:

[0101] A Transformer encoder is connected after the LSTM layer to analyze the spatial correlation and long-term dependencies of the data using a self-attention mechanism.

[0102] Configure key parameters of the Transformer, such as the number of multi-head attention points and the size of the feedforward neural network.

[0103] Hybrid architecture integration:

[0104] By concatenating the LSTM layer and the Transformer layer, a hybrid LSTM-Transformer architecture is formed.

[0105] A fully connected layer is added at the end of the hybrid architecture to output the rainfall forecast for the next 72 hours.

[0106] 3) Model training and optimization

[0107] Divide the dataset into training and test sets:

[0108] The integrated dataset is divided into a training set and a test set to ensure that the data distributions of the training set and the test set are similar and do not overlap.

[0109] Model training:

[0110] The LSTM-Transformer hybrid architecture model is trained using the training set, and the model parameters are updated using the backpropagation algorithm and optimizers (including Adam).

[0111] During training, monitor the changes in the model's loss function to ensure that the model gradually converges.

[0112] Model optimization:

[0113] The trained model is evaluated using a test set, and the prediction error (including mean squared error, MSE) is calculated.

[0114] Adjust the model hyperparameters (including the number of LSTM hidden units, the number of Transformer multi-head attention units, etc.) based on the evaluation results, and retrain the model to optimize performance.

[0115] Repeat the training and optimization process until the model's error in predicting rainfall in the next 72 hours is controlled within ≤5%.

[0116] 4) Rainfall forecast and results analysis

[0117] Real-time data input:

[0118] The latest weather radar data and ground monitoring station data are input into the trained LSTM-Transformer hybrid architecture model.

[0119] Rainfall forecast:

[0120] The model outputs rainfall forecasts for the next 72 hours, including hourly rainfall forecasts and total rainfall forecasts.

[0121] Results analysis:

[0122] The forecast results are visualized, including a curve showing the change in rainfall over time and a spatial distribution map.

[0123] Analyzing the sources and distribution of prediction errors provides a basis for further model optimization.

[0124] 5) Model Deployment and Application

[0125] Model Deployment:

[0126] The trained LSTM-Transformer hybrid architecture model is deployed to the cloud platform or edge computing node of the water conservancy information automation and safety control system.

[0127] Ensure that the model can receive weather radar data and ground monitoring station data in real time and make online predictions.

[0128] System Integration and Application:

[0129] Integrating rainfall forecast results into the automated and safe control system for water conservancy information provides data support for water conservancy project scheduling.

[0130] Based on the forecast results, corresponding water conservancy project scheduling plans are formulated, including adjusting reservoir water levels and opening or closing gates, to ensure the safe operation of water conservancy projects.

[0131] Preferably, in step S3, the intelligent linkage processing includes:

[0132] When the seepage pressure of the reservoir dam exceeds the safety threshold, the following actions are automatically triggered:

[0133] 1) Close the inlet gate;

[0134] 2) Start the drainage pump to lower the reservoir water level;

[0135] 3) Dispatch drones to capture high-definition video of the leak points and upload it to the cloud platform;

[0136] 4) Send Level 3 early warning information to management personnel via SMS, APP push and sound and light alarm.

[0137] Specific operating steps:

[0138] 1) Real-time monitoring and threshold triggering

[0139] Data collection:

[0140] Seepage pressure sensors deployed at key parts of the reservoir dam continuously collect seepage pressure data and upload it to the cloud platform in real time via IoT communication modules (including LoRa, 4G / 5G).

[0141] Threshold determination:

[0142] The cloud platform analyzes the received seepage pressure data in real time. When the monitored value exceeds the preset safety threshold (including the maximum seepage pressure value allowed by the design), the system automatically triggers the emergency response process.

[0143] 2) Automated equipment linkage control

[0144] Close the inlet gate:

[0145] The cloud platform sends a closing command to the intake gate control system via industrial control protocols (including Modbus and OPC UA).

[0146] After receiving the instruction, the gate actuator (including electric valves and hydraulic hoists) completes the gate closing action within 10 seconds to prevent external water from continuing to flow into the reservoir area.

[0147] Start the drainage pumps to lower the reservoir water level:

[0148] The cloud platform synchronously sends start commands to the drainage pump control system and dynamically adjusts the operating frequency of the drainage pump (including full power operation or stepped adjustment) according to the degree of seepage pressure exceeding the limit.

[0149] The drainage pumps started working, draining water from the reservoir through the drainage pipes and gradually lowering the water level to a safe range.

[0150] 3) Drone inspection and data transmission

[0151] Task scheduling:

[0152] The cloud platform sends task instructions to the drone dispatch system, specifying the inspection area (including the back slope of the dam and high-risk seepage areas) and shooting requirements (including high-definition video and infrared thermal imaging).

[0153] Drones perform inspections:

[0154] The drone automatically takes off from the hangar and flies to the target area along a preset route, using high-definition cameras and infrared sensors to take pictures of the leak point from all angles.

[0155] During filming, the drone transmits footage back to the cloud platform via real-time video stream, while also storing the raw data for subsequent analysis.

[0156] Data Upload and Processing:

[0157] After receiving video data transmitted back by the drone, the cloud platform automatically performs image recognition (including a deep learning-based leak detection algorithm), marks suspicious leak locations, and generates an inspection report.

[0158] 4) Multi-channel early warning information dissemination

[0159] Early warning classification and content generation:

[0160] Based on the degree of seepage pressure exceeding the limit (including mild, moderate, and severe) and the potential risk level, the system automatically generates three levels of early warning information (including blue, yellow, and orange warnings).

[0161] The early warning information includes key information such as the event type (pressure exceeding the limit), the time of occurrence, the current pressure value, and recommended treatment measures.

[0162] Multi-channel push:

[0163] SMS notification: Send early warning SMS messages to key personnel such as reservoir management personnel and technicians through SMS gateway.

[0164] APP push: Early warning information is pushed to relevant personnel through a dedicated water conservancy management APP, along with links to inspection videos and handling suggestions.

[0165] Audible and visual alarms: Audible and visual alarms are deployed at sites such as reservoir control rooms and dam management stations. When triggered, they emit a high-decibel alarm and flashing lights to remind on-site personnel to take immediate action.

[0166] 5) Closed-loop management of emergency response

[0167] Follow-up on the handling process:

[0168] The cloud platform continuously monitors the gate status, drainage pump operating parameters, and reservoir water level changes, and updates the emergency response progress in real time.

[0169] Managers can view the handling logs via the APP or Web, including equipment operation time, drone inspection results, and early warning information transmission records.

[0170] Results Feedback and Review:

[0171] After the emergency response is completed, the system automatically generates a response report, summarizing key information such as seepage pressure change curves, equipment operation data, and inspection images.

[0172] Experts were organized to review and analyze the handling process, optimize threshold settings, linkage strategies, and early warning information templates, and improve the efficiency of subsequent emergency response.

[0173] Secondly, the present invention provides an automated and safe water conservancy information control system, applicable to the aforementioned automated and safe water conservancy information control method, comprising:

[0174] The perception layer consists of a cluster of IoT sensing devices, including water level sensors, flow meters, water quality monitors, video surveillance terminals, and satellite remote sensing receivers.

[0175] Network layer: Adopts 5G+LoRa dual-mode communication protocol to achieve high-speed data transmission between sensing devices and cloud platform;

[0176] Platform layer: Deployed on a private cloud server, including:

[0177] Data governance module (supports petabyte-level data storage and real-time analysis);

[0178] Digital twin engine (integrating GIS maps, BIM models, and hydrological simulation algorithms);

[0179] Intelligent decision-making module (based on reinforcement learning algorithm to optimize scheduling scheme);

[0180] Application layer: Provides dual entry points for web and mobile devices, supporting:

[0181] Real-time monitoring data visualization;

[0182] Early warning information is disseminated through multiple channels;

[0183] Remote control commands are issued;

[0184] Digital management of emergency response plans.

[0185] Preferably, in the intelligent decision-making module, the reinforcement learning algorithm adopts the PPO (Proximal Policy Optimization) framework, with the optimization objective of "minimizing disaster losses + maximizing water resource utilization efficiency", and generates the optimal scheduling strategy through interactive training with the digital twin model.

[0186] Preferably, all the sensing layer devices integrate a self-diagnostic module, which can monitor the device's operating status in real time (including sensor accuracy drift and communication interruption), and store the device's health data on the blockchain to ensure that the data is tamper-proof and traceable.

[0187] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described method.

[0188] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0189] Compared with the prior art, the present invention has the following beneficial effects:

[0190] This invention integrates advanced technologies such as the Internet of Things, edge computing, deep learning, digital twins, and reinforcement learning to achieve automated collection, processing, analysis, and intelligent decision-making of water conservancy information. It significantly improves the real-time monitoring, early warning, and emergency response capabilities of water conservancy projects, ensuring the safe operation of the projects, while optimizing resource allocation and improving management efficiency. Attached Figure Description

[0191] Figure 1 This is a flowchart of the water conservancy information automation and safety control method of the present invention;

[0192] Figure 2 This is a framework diagram of the water conservancy information automation and safety control system of the present invention. Detailed Implementation

[0193] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0194] Example: This invention provides an automated and secure control method for water conservancy information, such as... Figure 1 As shown, it includes the following steps:

[0195] S1. First, by deploying IoT sensing devices (including water level gauges, rain gauges, flow meters, multi-parameter water quality sensors, video surveillance cameras, and satellite remote sensing receivers) at the water conservancy project site, real-time data on water level, flow rate, water quality, meteorology, and engineering structure safety are collected. Then, edge computing nodes are used to clean, calibrate, and perform preliminary analysis on the raw data.

[0196] The data cleaning algorithm uses Z-score normalization, and its algorithm formula is as follows:

[0197]

[0198] Where x is the original data point, μ is the mean of the data, and σ is the standard deviation of the data. Z-score standardization can eliminate the influence of different units of measurement, making the data easier to compare and analyze.

[0199] S2. Based on historical disaster data, real-time monitoring data, and water conservancy project characteristic parameters, a digital twin model is constructed that integrates four dimensions: forecasting, early warning, rehearsal, and contingency plan.

[0200] The forecasting module employs deep learning algorithms to perform rolling forecasts of rainfall and flood events. This involves applying the Transformer model to rainfall and flood predictions, improving forecast accuracy by capturing long-range dependencies in the data. Specifically, this includes:

[0201] 1) Self-attention mechanism:

[0202]

[0203] Where Q, K, and V are the query, key, and value matrices, respectively, and d k It is the dimension of the key.

[0204] 2) Bullish Attention:

[0205] The self-attention mechanism is extended to multi-head attention, which involves parallel computation of multiple different query, key, and value matrices, and then concatenating the results.

[0206] 3) Feedforward Neural Network:

[0207] Each location is independently applied to a feedforward neural network, which typically contains two linear transformations and a ReLU activation function.

[0208] The early warning module triggers tiered early warning signals by setting multiple threshold levels (including water level exceeding limits and abnormal seepage pressure);

[0209] The simulation module utilizes 3D simulation technology to model the safety status of the project and the evolution path of disasters under different working conditions. The specific steps are as follows: First, based on the design drawings, geological survey data, and historical operation records of the water conservancy project, a high-precision 3D digital model is constructed. This model must cover key elements such as the project structure, surrounding topography, and hydrological environment. Second, various working condition parameters are set in the 3D model, including different rainfall intensities, reservoir water level change rates, and structural damage scenarios. Simulation algorithms such as finite element analysis or computational fluid dynamics are used to simulate the stress distribution, deformation, and stability changes of the project structure under various working conditions. Simultaneously, combined with disaster evolution theory, the module simulates disaster processes such as flood overflow and levee breach, analyzing their impact paths and potential damage range on project safety. Finally, the simulation results are visually displayed using visualization technology, providing a scientific basis for project safety assessment, emergency plan formulation, and dispatching decisions.

[0210] The contingency plan module automatically generates an emergency dispatch plan based on the rehearsal results and pushes it to the decision-making terminal;

[0211] S3. Based on the risk assessment results, control commands are sent to the actuators (including gate controllers, water pump frequency converters, and drone inspection systems) through the IoT platform to achieve:

[0212] Remote automatic control of water conservancy engineering facilities;

[0213] Intelligent and coordinated response to emergencies (including activating backup power and closing valves in dangerous areas);

[0214] Dynamic optimization and allocation of inspection tasks (based on AI path planning algorithms to reduce blind spots in manual inspections).

[0215] During operation, multi-dimensional data is first collected in real time through IoT sensing devices deployed at water conservancy project sites, and edge computing nodes are used for data cleaning, calibration, and preliminary analysis. Subsequently, a digital twin model containing "forecast-early warning-rehearsal-contingency plan" is constructed based on historical and real-time data. The rainfall and flood process is predicted through deep learning algorithms, multi-level thresholds are set to trigger graded early warnings, and 3D simulation is used to simulate the safety status of the project and disaster paths, and emergency dispatch plans are automatically generated. Finally, based on the risk assessment results, control commands are sent to the implementing agencies through the IoT platform to realize remote automatic control of water conservancy facilities, intelligent linkage response to sudden emergencies, and dynamic optimization allocation of inspection tasks.

[0216] As can be seen from the above, this method significantly improves the real-time monitoring, early warning and emergency response capabilities of water conservancy projects, ensures the safe operation of the projects, optimizes resource allocation and improves management efficiency.

[0217] In one exemplary embodiment, in step S1, the edge computing node configures a dynamic power consumption strategy:

[0218] During low-risk periods, the system enters a dormant state, with the edge gateway simulating sensor values ​​based on historical data. When the federated learning platform predicts that the probability of an anomaly exceeds the threshold, it wakes up the corresponding regional nodes to perform high-precision sampling.

[0219] During operation: In low-risk periods, edge computing nodes enter a dormant state to reduce energy consumption. At this time, the edge gateway uses historical data to simulate sensor values ​​and maintain basic data monitoring functions. When the federated learning platform predicts that the probability of an anomaly exceeds a preset threshold, it immediately wakes up the edge computing nodes in the corresponding area to perform high-precision data sampling to ensure the accuracy and real-time nature of the data.

[0220] As can be seen from the above, this step can effectively reduce the overall energy consumption of the system, extend the service life of the equipment, and respond quickly at critical moments, ensuring the reliability and accuracy of the water conservancy information automation and safety control system.

[0221] In one exemplary embodiment, the deep learning algorithm employs an LSTM-Transformer hybrid architecture, combining meteorological radar data with ground monitoring station data, to achieve a 72-hour rainfall prediction error of ≤5%.

[0222] During operation, meteorological radar data and ground monitoring station data are first fused together. Then, the LSTM layer is responsible for capturing the dynamic characteristics of rainfall in the time series, while the Transformer layer analyzes spatial correlation and long-term dependence through a self-attention mechanism. Finally, the rainfall prediction for the next 72 hours is achieved through joint optimization of the hybrid architecture, with the error controlled within ≤5%.

[0223] As can be seen from the above, this step significantly improves the accuracy and timeliness of rainfall forecasting, provides reliable data support for water conservancy project scheduling, and effectively reduces the engineering safety risks caused by inaccurate rainfall forecasting.

[0224] In one exemplary embodiment, step S3 includes:

[0225] When the seepage pressure of the reservoir dam exceeds the safety threshold, the following actions are automatically triggered:

[0226] 1) Close the inlet gate;

[0227] 2) Start the drainage pump to lower the reservoir water level;

[0228] 3) Dispatch drones to capture high-definition video of the leak points and upload it to the cloud platform;

[0229] 4) Send Level 3 early warning information to management personnel via SMS, APP push and sound and light alarm.

[0230] In the intelligent linkage response in step S3, when the monitoring system detects that the seepage pressure of the reservoir dam exceeds the preset safety threshold, the system immediately and automatically triggers a series of emergency operations: first, the inlet gate is closed to block the water source input, and at the same time, the drainage pump is started to quickly lower the reservoir water level to reduce the pressure on the dam; then, a drone is dispatched to fly to the seepage point to take high-definition video shots and upload the real-time images to the cloud platform for remote analysis; finally, three-level early warning information is sent to the management personnel through multiple channels such as SMS, APP push and sound and light alarms to ensure timely response and handling.

[0231] As can be seen from the above, this step can quickly control the development of the danger and prevent the disaster from spreading. At the same time, through multi-channel early warning, it ensures that managers can obtain danger information as soon as possible, thereby improving the efficiency and accuracy of emergency response.

[0232] Working principle: The system collects multi-data of water conservancy projects in real time through a cluster of IoT sensing devices, and performs data cleaning and preliminary analysis using edge computing nodes. Subsequently, a digital twin model is constructed based on historical disaster data, real-time monitoring data, and engineering characteristic parameters to achieve four-dimensional linkage of forecasting, early warning, rehearsal, and contingency planning. At the same time, the intelligent decision-making module uses the PPO reinforcement learning framework to interact and train with the digital twin model to generate the optimal scheduling strategy. Finally, the system achieves remote automatic control, intelligent linkage response, and inspection task optimization through the IoT platform to ensure the safe and efficient operation of water conservancy projects.

[0233] Furthermore, the automation and security control method for water conservancy information in this embodiment is compared with the existing traditional water conservancy project management methods, and the results are shown in the table below:

[0234] Comparison indicators Method of the present invention Traditional water conservancy project management methods Data acquisition delay ≤200ms ≥2s Early warning accuracy 98.7％ ≤60％ Scheduling scheme generation time ≤30 seconds ≥30 minutes

[0235] As shown in the table above, by comparing the performance of the system of this invention with existing traditional systems / manual decision-making in three key indicators—data acquisition latency, early warning accuracy, and scheduling scheme generation time—the table intuitively demonstrates the significant advantages of the system of this invention in terms of real-time performance (data acquisition latency ≤200ms), accuracy (early warning accuracy 98.7%), and decision-making efficiency (scheduling scheme generation time ≤30 seconds). This highlights the innovation and leading position of its technical solution in the field of water conservancy information automation and safety control.

[0236] An automated and safe water conservancy information control system, such as Figure 2 As shown, the above-mentioned method for automated and safe control of water conservancy information includes:

[0237] The perception layer consists of a cluster of IoT sensing devices, including water level sensors, flow meters, water quality monitors, video surveillance terminals, and satellite remote sensing receivers.

[0238] Network layer: Adopts 5G+LoRa dual-mode communication protocol to achieve high-speed data transmission between sensing devices and cloud platform;

[0239] Platform layer: Deployed on a private cloud server, including:

[0240] Data governance module (supports petabyte-level data storage and real-time analysis);

[0241] Digital twin engine (integrating GIS maps, BIM models, and hydrological simulation algorithms);

[0242] Intelligent decision-making module (based on reinforcement learning algorithm to optimize scheduling scheme);

[0243] Application layer: Provides dual entry points for web and mobile devices, supporting:

[0244] Real-time monitoring data visualization;

[0245] Early warning information is disseminated through multiple channels;

[0246] Remote control commands are issued;

[0247] Digital management of emergency response plans.

[0248] During operation, the perception layer collects water conservancy data in real time through a cluster of IoT sensing devices, and the network layer uses the 5G+LoRa dual-mode communication protocol to transmit the data to the cloud platform at high speed. The data governance module of the platform layer performs petabyte-level data storage and real-time analysis, the digital twin engine integrates GIS, BIM and hydrological algorithms to build a virtual model, and the intelligent decision-making module generates optimized scheduling schemes based on reinforcement learning algorithms. The application layer provides real-time data visualization, multi-channel early warning push, remote command issuance and digital management functions of emergency plans through web and mobile terminals.

[0249] As can be seen from the above, the system has achieved full-process automation and intelligence, significantly improving the monitoring accuracy, response speed and decision-making scientificity of water conservancy projects, effectively ensuring project safety and optimizing water resource management efficiency.

[0250] In one exemplary embodiment, the reinforcement learning algorithm in the intelligent decision-making module adopts the PPO (Proximal Policy Optimization) framework, with the optimization objective of "minimizing disaster losses and maximizing water resource utilization efficiency", and generates the optimal scheduling strategy through interactive training with the digital twin model.

[0251] When the PPO framework in this intelligent decision-making module is working, it takes "minimizing disaster losses and maximizing water resource utilization efficiency" as the optimization goal. First, it obtains the current status of water conservancy projects and future evolution predictions by interacting with digital twin models. Then, it tries various scheduling strategies in a simulation environment and evaluates their effects. It uses the gradient ascent characteristic of the PPO algorithm to continuously adjust the strategy parameters to approach the optimal solution, and finally generates a water conservancy project scheduling strategy that takes into account both safety and efficiency.

[0252] As can be seen from the above, this step can dynamically adapt to the complex and ever-changing water conservancy environment, maximize the efficiency of water resource utilization while ensuring the safety of the project, and significantly improve the scientific nature and accuracy of decision-making.

[0253] In one exemplary embodiment, the sensing layer devices are all integrated with a self-diagnostic module, which can monitor the device's operating status in real time (including sensor accuracy drift and communication interruption), and store the device's health data on the blockchain to ensure that the data is tamper-proof and traceable.

[0254] The self-diagnostic module integrated into this sensing layer device monitors the device's operating status in real time, including key indicators such as whether sensor accuracy has drifted and whether communication has been interrupted. This device health data is then stored on the blockchain using blockchain technology. In this process, the distributed ledger characteristics of blockchain ensure the immutability and traceability of the data.

[0255] As can be seen from the above, this step can promptly detect and warn of equipment failures, ensure the accuracy and continuity of data collection, and enhance the security and credibility of data through blockchain technology, providing reliable data support for the precise management and decision-making of water conservancy projects.

[0256] This application provides an electronic device applicable to the above-mentioned automated and safe control method for water conservancy information, including:

[0257] Memory is used to protect computer programs and data;

[0258] A processor is used to run system programs.

[0259] This application provides a computer storage medium applicable to the above-mentioned automated and secure control method for water conservancy information, and performs hierarchical confidentiality management of the above-mentioned system and data in accordance with confidentiality management requirements.

[0260] Those skilled in the art will understand that embodiments of this application can be provided as a system or a computer program product. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0261] This application is described with reference to flowchart illustrations and / or block diagrams of apparatus (systems) and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0262] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0263] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0264] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0265] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0266] Computer-readable media include both permanent and non-permanent, removable and non-removable media, which can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0267] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, article, or apparatus that includes that element.

[0268] The embodiments of the present invention are given for the purposes of illustration and description. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for automated and secure control of water conservancy information, characterized in that, Includes the following steps: S1. First, by deploying IoT sensing devices at the water conservancy project site, real-time data on water level, flow rate, water quality, weather, and engineering structure safety are collected, and edge computing nodes are used to clean, calibrate, and perform preliminary analysis on the raw data. S2. Based on historical disaster data, real-time monitoring data, and water conservancy project characteristic parameters, construct a digital twin model that includes four-dimensional linkage of "forecast-early warning-drill-contingency plan"; S3. Based on the risk assessment results, send control commands to the implementing agency through the IoT platform to achieve: Remote automatic control of water conservancy engineering facilities; Intelligent and coordinated response to sudden emergencies; Dynamic optimization allocation of inspection tasks.

2. The water conservancy information automation and security control method as described in claim 1, characterized in that: In step S1, the edge computing node configures a dynamic power consumption strategy: During low-risk periods, the system enters a dormant state, with the edge gateway simulating sensor values ​​based on historical data. When the federated learning platform predicts that the probability of an anomaly exceeds the threshold, it wakes up the corresponding regional nodes to perform high-precision sampling.

3. The method for automated and secure control of water conservancy information as described in claim 1, characterized in that: In step S2, the four-dimensional linkage digital twin model of "forecast-early warning-rehearsal-contingency plan" includes a forecast module, an early warning module, a rehearsal module, and a contingency plan module, wherein: The forecasting module uses deep learning algorithms to make rolling predictions of rainfall and flood events; The early warning module triggers tiered early warning signals by setting multiple threshold levels; The simulation module uses 3D simulation technology to simulate the engineering safety status and disaster evolution path under different working conditions; The contingency plan module automatically generates an emergency dispatch plan based on the rehearsal results and pushes it to the decision-making terminal.

4. The water conservancy information automation and security control method as described in claim 3, characterized in that: The deep learning algorithm adopts an LSTM-Transformer hybrid architecture, combining meteorological radar data and ground monitoring station data.

5. The method for automated and secure control of water conservancy information as described in claim 1, characterized in that: In step S3, the intelligent linkage processing includes: When the seepage pressure of the reservoir dam exceeds the safety threshold, the following actions are automatically triggered: 1) Close the inlet gate; 2) Start the drainage pump to lower the reservoir water level; 3) Dispatch drones to capture high-definition video of the leak points and upload it to the cloud platform; 4) Send Level 3 early warning information to management personnel via SMS, APP push and sound and light alarm.

6. A water conservancy information automation and safety control system, characterized in that: The water conservancy information automation and safety control method as described in any one of claims 1-5 includes: The perception layer consists of a cluster of IoT sensing devices, including water level sensors, flow meters, water quality monitors, video surveillance terminals, and satellite remote sensing receivers. Network layer: Adopts 5G+LoRa dual-mode communication protocol to achieve high-speed data transmission between sensing devices and cloud platform; Platform layer: Deployed on a private cloud server, including: Data governance module; Digital twin engine; Intelligent decision-making module; Application layer: Provides dual entry points for web and mobile devices, supporting: Real-time monitoring data visualization; Early warning information is disseminated through multiple channels; Remote control commands are issued; Digital management of emergency response plans.

7. The water conservancy information automation and safety control system as described in claim 6, characterized in that: In the intelligent decision-making module, the reinforcement learning algorithm adopts the PPO framework, with the optimization objective of "minimizing disaster losses and maximizing water resource utilization efficiency". It generates the optimal scheduling strategy through interactive training with the digital twin model.

8. The water conservancy information automation and safety control system as described in claim 6, characterized in that: All the sensing layer devices integrate a self-diagnostic module, which can monitor the device's operating status in real time and store the device's health data on the blockchain to ensure that the data is tamper-proof and traceable.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

Images