Water conservancy project data acquisition device and method based on internet of things

Abstract

The application discloses a water conservancy project data acquisition device and method based on Internet of Things, and belongs to the technical field of water conservancy project intelligent monitoring and risk management. The device comprises a sensor node module, a virtual center node module, a cloud management module, a token allocation module and a decision optimization module, wherein: the sensor node module is used for collecting multiple types of monitoring parameters in real time at a water conservancy project site, and realizing preliminary anomaly detection; the virtual center node module is responsible for gathering and fusing multi-source data, completing regional risk assessment and generating early warning information; the cloud management module is used for centrally storing and analyzing global data, providing trend prediction and decision support; the token allocation module dynamically allocates emergency, coordination and communication tokens according to event levels, so as to guarantee the priority processing of key events and efficient information transmission; and the decision optimization module dynamically adjusts event identification threshold values, token allocation strategies and permission boundary parameters based on execution feedback.

Description

Technical Field

[0001] This application belongs to the field of intelligent monitoring and risk management technology for water conservancy projects, and relates to a water conservancy project data acquisition device and method based on the Internet of Things. Through multi-parameter sensing, heterogeneous data fusion, dynamic token allocation and adaptive decision optimization, it realizes high-precision real-time monitoring, hierarchical early warning and intelligent global management of water conservancy projects. Background Technology

[0002] With the continuous expansion of water conservancy projects, infrastructure such as reservoirs, dams, rivers, and drainage systems play a vital role in ensuring flood control, irrigation, and urban water supply. However, the operating environment of water conservancy projects is complex and variable, affected by factors such as rainfall, water level fluctuations, flow changes, water quality, and meteorological conditions, posing numerous challenges to their safety and operational efficiency. Traditional water conservancy project monitoring methods mainly rely on manual inspections and fixed-point instrument measurements, which suffer from problems such as long data collection cycles, information lag, limited coverage, and weak anomaly identification capabilities, making it difficult to meet the demands of modern water conservancy projects for high-precision, real-time, and continuous monitoring.

[0003] In recent years, the application of IoT technology in environmental monitoring and industrial control has made significant progress, providing new solutions for intelligent monitoring of water conservancy projects. By deploying various types of sensor nodes at the project site, real-time acquisition of water level, flow rate, water quality, and environmental parameters can be achieved, and data transmission can be carried out in conjunction with wireless communication technology, which can improve monitoring efficiency and information acquisition capabilities to a certain extent. However, existing IoT monitoring systems still have several limitations: First, sensor nodes are scattered, lacking efficient multi-source data fusion and anomaly detection capabilities, making it difficult to accurately identify regional risk events; second, monitoring data is easily affected by delays and interference during transmission between different nodes, resulting in insufficient timeliness and reliability of early warning information; third, existing systems lack a unified decision-making optimization mechanism, limiting their emergency handling and resource scheduling capabilities for sudden events, and making it difficult to achieve global collaboration and dynamic adaptive optimization.

[0004] In summary, how to achieve real-time monitoring of multiple parameters, regional data fusion, risk classification and early warning, and intelligent decision-making across the entire network at water conservancy project sites, while ensuring priority handling of critical events and adaptive optimization of the system, has become an urgent technical problem to be solved. Summary of the Invention

[0005] To overcome a series of shortcomings in existing technologies, this application aims to provide an IoT-based water conservancy engineering data acquisition device, comprising a sensor node module, a virtual central node module, a cloud management module, a token allocation module, and a decision optimization module. Specifically: the sensor node module is used to collect multiple monitoring parameters in real time at the water conservancy engineering site and perform preliminary anomaly detection; the virtual central node module is responsible for aggregating and fusing multi-source data, completing regional risk assessment, and generating early warning information; the cloud management module is used for centralized storage and analysis of global data, providing trend prediction and decision support; the token allocation module dynamically allocates emergency, coordination, and communication tokens according to event levels to ensure priority handling of critical events and efficient information transmission; and the decision optimization module dynamically adjusts event identification thresholds, token allocation strategies, and permission boundary parameters based on execution feedback.

[0006] Furthermore, the sensor node module includes a water level detection unit, a flow rate detection unit, a water quality detection unit, and an environmental detection unit. Specifically, the water level detection unit uses piezoresistive or ultrasonic ranging principles to achieve millimeter-level accuracy in water level monitoring; the flow rate detection unit measures water flow velocity and flow rate changes through the Doppler effect or electromagnetic induction principle; the water quality detection unit integrates a pH detector, a dissolved oxygen detector, a turbidity detector, and a heavy metal ion detector to achieve real-time multi-parameter water quality analysis; and the environmental detection unit is equipped with temperature and humidity sensors, wind speed and direction sensors, rainfall sensors, and air pressure sensors to monitor meteorological conditions affecting the operation of water conservancy projects.

[0007] Furthermore, the virtual central node module adopts a hierarchical data processing architecture, including a data receiving layer, a fusion analysis layer, and a decision output layer. The data receiving layer is configured with a multi-protocol communication interface to ensure data transmission stability in complex electromagnetic environments. The fusion analysis layer uses a weighted average method, Kalman filtering algorithm, and Bayesian inference method to perform spatiotemporal fusion processing on heterogeneous data from different sensor nodes, thereby eliminating the influence of sensor measurement errors and environmental interference and generating high-precision comprehensive monitoring results. The decision output layer, based on fuzzy control theory and expert system rule base, realizes intelligent assessment of the operating status of water conservancy projects and generates early warning information.

[0008] Furthermore, the cloud management module adopts a microservice architecture design, including a data storage service unit, a computing and analysis service unit, a visualization service unit, and a system management service unit. The data storage service unit uses a hybrid storage architecture combining a distributed database cluster and a time-series database, enabling efficient storage and management of massive amounts of time-series monitoring data, and supporting rapid data querying and historical backtracking analysis. The computing and analysis service unit provides real-time stream computing, batch data analysis, and predictive modeling functions. The visualization service unit uses responsive web technology to achieve multi-dimensional data chart display, geographic information system integration, and mobile adaptation. The system management service unit is responsible for user permission management, system configuration maintenance, and operational status monitoring to ensure the stable operation of the entire cloud platform.

[0009] The purpose of this application is also to provide a method for acquiring water conservancy project data based on the Internet of Things, implemented using the aforementioned water conservancy project data acquisition equipment, including the following steps: Acquire the distribution information of monitoring points and the initial network topology at the site of the water conservancy project, and obtain the initial state and communication capability parameters of each sensor node; Based on the initial state of nodes, the distribution information of monitoring points, and the network topology, a three-layer intelligent decision-making network architecture is constructed, consisting of sensor node modules, virtual center node modules, and cloud management modules, and corresponding monitoring permissions and decision-making scopes are assigned to each layer. The sensor node module collects real-time monitoring data of the water conservancy project and transmits it to the virtual center node module for preliminary processing and risk assessment. Whenever monitoring data is received, the urgency of the current monitoring event is classified based on the magnitude of data change and preset risk thresholds. Time windows are dynamically divided according to the event level, and corresponding decision priorities and processing permissions are assigned to events of different levels. Based on the event level and time window configuration, the token allocation module dynamically generates and allocates emergency tokens, coordination tokens, and communication tokens to ensure that emergencies receive the highest processing priority, regional decisions receive collaborative permissions, and key data receives priority transmission channels. During the token allocation process, the token usage status and decision consistency of each decision node are continuously monitored. When a permission conflict or decision inconsistency is detected, the conflict is automatically resolved based on priority principles, timeliness requirements, and energy consumption balance strategies. If a decision conflict is detected to be unresolved, a global coordination mechanism is activated to reassess and reassign decision permissions through the cloud management module until network-wide decision consistency is achieved. If all decision conflicts have been resolved and the network is operating stably, the decision optimization module continuously collects decision execution feedback data from each layer of nodes, dynamically adjusts the emergency event identification threshold, token allocation strategy, and permission boundary parameters, and controls the water conservancy project data acquisition equipment to continue executing the next round of monitoring and decision-making according to the optimized parameter configuration.

[0010] Furthermore, in the three-layer intelligent decision-making network architecture, the sensor node module is responsible for local data acquisition and primary anomaly detection, and is equipped with lightweight edge computing capabilities to identify basic anomalies such as data mutations and sensor failures; the virtual center node module undertakes regional data aggregation and intermediate risk analysis functions, integrates multi-source information fusion algorithms and engineering safety assessment models, and can identify regional risk events and cross-sensor correlation anomalies; the cloud management module provides global data management and advanced decision support, deploys deep learning models and expert knowledge bases, and can perform long-term trend prediction, complex fault diagnosis, and emergency plan recommendation.

[0011] Furthermore, the real-time monitoring data includes water level changes, flow velocity fluctuations, pressure anomalies, temperature gradients, and structural displacements. Each parameter is configured with an independent acquisition frequency and accuracy requirement: water level and pressure parameters are acquired at millisecond-level high-frequency acquisition, flow velocity and temperature parameters are acquired at second-level medium-frequency acquisition, and structural displacements are acquired at minute-level low-frequency acquisition. During data transmission, a hierarchical compression and priority queuing strategy is adopted: urgent data is losslessly compressed and given the highest transmission priority, while regular data is lossily compressed to save bandwidth resources. When the network is interrupted or transmission fails, lost data can be automatically retransmitted, and missing time-series data can be compensated through interpolation algorithms to ensure the continuity and availability of monitoring data.

[0012] Furthermore, the calculation of the magnitude of the data change includes the following steps: Obtain the current value and historical baseline value of the monitoring data as the basic input for the analysis of the magnitude of change; The absolute change is calculated by the difference between the current value and the benchmark value, and the relative rate of change is further obtained to reflect the magnitude and relative intensity of the data change. Based on time series methods, the rate of change of data is calculated using the first derivative or difference method to reveal the speed of change and inflection point characteristics. By using moving average, exponential smoothing, or regression analysis methods, we can predict future data trends and provide forward-looking information for early warning judgments. Based on the distribution characteristics, seasonal patterns, and operating cycles of historical data, a threshold system covering four categories—normal, alert, early warning, and emergency—is established. Combined with real-time monitoring characteristics and expert experience, a self-learning algorithm is used to dynamically adjust the threshold parameters. By comparing the absolute change, relative change rate, change speed, and trend results with the threshold system, the risk level is comprehensively determined and early warning information is output.

[0013] Furthermore, the urgency classification includes the following steps: Evaluation data were collected from five dimensions: degree of deviation, rate of change, scope of impact, historical similar events, and importance of the project. The indicators of each dimension were standardized from 0 to 1. The weight coefficients of each dimension were determined by the analytic hierarchy process (AHP), and the standardized scores of each dimension were weighted and summed to obtain a comprehensive urgency score. Based on the comprehensive urgency score, the urgency level is divided into four levels: Normal Level I, Attention Level II, Warning Level III, and Emergency Level IV, so as to achieve accurate identification of different risk levels; Based on the changing trends of continuous monitoring data, the risk level is automatically upgraded when it increases and automatically downgraded when the abnormal situation is controlled or eliminated, ensuring the dynamic real-time nature of the classification status. The classification results are matched and compared with the historical event database, and the experience of handling similar cases is referenced to provide auxiliary basis for current risk management and emergency decision-making. By verifying the results after the fact and providing continuous feedback, the correctness of the classification results is checked, and the classification algorithm and parameter settings are optimized accordingly to improve the stability and adaptability of the classification.

[0014] Furthermore, the decision priority allocation includes the following steps: A five-level priority system is established to classify and manage different types of events: ultra-high priority is used for emergency events related to life safety, high priority is used for important events related to engineering safety, medium priority is used for equipment failures and operational anomalies, low priority is used for routine maintenance and data statistics, and lowest priority is used for historical data queries and system optimization. An independent processing queue is set up for each priority level, and corresponding computing resources and communication bandwidth are configured. High-priority tasks can preempt the resources of low-priority tasks to ensure the timely processing of critical tasks. Dynamically modify task priorities based on the development trend of the event and the effectiveness of the handling to prevent resource waste caused by rigid priorities, while maintaining continuous and efficient response to critical tasks; Set a maximum waiting time for low-priority tasks and trigger priority promotion rules based on task waiting status or importance to ensure that long-term low-priority tasks are not ignored and achieve a balance in resource allocation. When multiple tasks of equal priority arrive at the same time, they are sorted by timestamp, task importance and resource requirements to avoid processing conflicts. During operation, the processing efficiency and resource utilization of tasks at all levels are continuously monitored, and priority allocation strategies and queue management parameters are adjusted based on feedback to optimize scheduling performance.

[0015] Compared with existing technologies, this application has the following beneficial effects: by real-time acquisition of multiple parameters from sensor nodes, fusion of multi-source data from virtual central nodes and regional risk assessment, global management and trend prediction in the cloud, and dynamic token allocation and decision optimization based on event levels, it realizes real-time closed-loop, precise classification and adaptive optimization of water conservancy project monitoring, early warning and emergency response. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the structure of a water conservancy engineering data acquisition device based on the Internet of Things disclosed in an embodiment of this application.

[0017] Figure 2 This is a flowchart illustrating a water conservancy project data acquisition method based on the Internet of Things disclosed in an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be described in more detail below with reference to the accompanying drawings. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some embodiments of this invention, but not all embodiments.

[0019] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] The embodiments and directional terms described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0021] like Figure 1As shown, an IoT-based water conservancy engineering data acquisition device includes a sensor node module, a virtual central node module, a cloud management module, a token allocation module, and a decision optimization module. Specifically: the sensor node module collects various monitoring parameters in real time at the water conservancy engineering site and performs preliminary anomaly detection; the virtual central node module aggregates and merges multi-source data to complete regional risk assessment and generate early warning information; the cloud management module centrally stores and analyzes global data, providing trend prediction and decision support; the token allocation module dynamically allocates emergency, coordination, and communication tokens based on event levels to ensure priority handling of critical events and efficient information transmission; and the decision optimization module dynamically adjusts event identification thresholds, token allocation strategies, and permission boundary parameters based on execution feedback.

[0022] As can be seen from the above, the device achieves real-time acquisition and anomaly detection of on-site data of water conservancy projects through multi-level collaboration, integrates multi-source information for regional risk assessment and early warning generation, centrally manages global data to support trend analysis and decision optimization, dynamically allocates tokens to ensure priority handling of critical events, and adjusts thresholds and strategies in real time based on execution feedback, thereby improving monitoring accuracy, response speed and overall management efficiency.

[0023] Furthermore, the sensor node module includes a water level detection unit, a flow rate detection unit, a water quality detection unit, and an environmental detection unit. Specifically, the water level detection unit uses piezoresistive or ultrasonic ranging principles to achieve millimeter-level accuracy in water level monitoring; the flow rate detection unit measures water flow velocity and flow rate changes through the Doppler effect or electromagnetic induction principle; the water quality detection unit integrates a pH detector, a dissolved oxygen detector, a turbidity detector, and a heavy metal ion detector to achieve real-time multi-parameter water quality analysis; and the environmental detection unit is equipped with temperature and humidity sensors, wind speed and direction sensors, rainfall sensors, and air pressure sensors to monitor meteorological conditions affecting the operation of water conservancy projects.

[0024] As can be seen from the above, this sensor node module, by integrating high-precision water level, flow rate, multi-parameter water quality and environmental monitoring units, realizes comprehensive real-time perception of the operation status of water conservancy projects and surrounding meteorological conditions. It can provide millimeter-level water level monitoring, accurate flow rate measurement, multi-index water quality analysis and environmental factor monitoring, providing reliable and accurate basic data support for subsequent data fusion, risk assessment and early warning.

[0025] Furthermore, the virtual central node module adopts a hierarchical data processing architecture, including a data receiving layer, a fusion analysis layer, and a decision output layer. The data receiving layer is configured with a multi-protocol communication interface to ensure data transmission stability in complex electromagnetic environments. The fusion analysis layer uses a weighted average method, Kalman filtering algorithm, and Bayesian inference method to perform spatiotemporal fusion processing on heterogeneous data from different sensor nodes, thereby eliminating the influence of sensor measurement errors and environmental interference and generating high-precision comprehensive monitoring results. The decision output layer, based on fuzzy control theory and expert system rule base, realizes intelligent assessment of the operating status of water conservancy projects and generates early warning information.

[0026] As can be seen from the above, the virtual central node module achieves stable reception, spatiotemporal fusion, and accurate analysis of multi-source heterogeneous data through a hierarchical architecture. It uses weighted averaging, Kalman filtering, and Bayesian inference to eliminate measurement errors and environmental interference, generate high-precision comprehensive monitoring results, and intelligently evaluate the operating status of water conservancy projects based on fuzzy control and expert rules. It can output reliable early warning information in a timely manner, thereby improving the overall monitoring accuracy and response capability.

[0027] Furthermore, the cloud management module adopts a microservice architecture design, including a data storage service unit, a computing and analysis service unit, a visualization service unit, and a system management service unit. The data storage service unit uses a hybrid storage architecture combining a distributed database cluster and a time-series database, enabling efficient storage and management of massive amounts of time-series monitoring data, and supporting rapid data querying and historical backtracking analysis. The computing and analysis service unit provides real-time stream computing, batch data analysis, and predictive modeling functions. The visualization service unit uses responsive web technology to achieve multi-dimensional data chart display, geographic information system integration, and mobile adaptation. The system management service unit is responsible for user permission management, system configuration maintenance, and operational status monitoring to ensure the stable operation of the entire cloud platform.

[0028] As can be seen from the above, this cloud management module achieves efficient storage, rapid querying, and historical backtracking of massive water conservancy monitoring data through a microservice architecture. It provides real-time calculation, batch analysis, and trend prediction functions, and intuitively presents monitoring results through multi-dimensional visualization and geographic information integration. At the same time, it realizes user permission management and system status monitoring, ensuring the stability of platform operation and the real-time, visualization, and intelligent level of data application.

[0029] like Figure 2 As shown, the purpose of this application is also to provide a data acquisition method for water conservancy projects based on the Internet of Things, implemented using the aforementioned water conservancy project data acquisition equipment, including the following steps: Acquire the distribution information of monitoring points and the initial network topology at the site of the water conservancy project, and obtain the initial state and communication capability parameters of each sensor node; Based on the initial state of nodes, the distribution information of monitoring points, and the network topology, a three-layer intelligent decision-making network architecture is constructed, consisting of sensor node modules, virtual center node modules, and cloud management modules, and corresponding monitoring permissions and decision-making scopes are assigned to each layer. The sensor node module collects real-time monitoring data of the water conservancy project and transmits it to the virtual center node module for preliminary processing and risk assessment. Whenever monitoring data is received, the urgency of the current monitoring event is classified based on the magnitude of data change and preset risk thresholds. Time windows are dynamically divided according to the event level, and corresponding decision priorities and processing permissions are assigned to events of different levels. Based on the event level and time window configuration, the token allocation module dynamically generates and allocates emergency tokens, coordination tokens, and communication tokens to ensure that emergencies receive the highest processing priority, regional decisions receive collaborative permissions, and key data receives priority transmission channels. During the token allocation process, the token usage status and decision consistency of each decision node are continuously monitored. When a permission conflict or decision inconsistency is detected, the conflict is automatically resolved based on priority principles, timeliness requirements, and energy consumption balance strategies. If a decision conflict is detected to be unresolved, a global coordination mechanism is activated to reassess and reassign decision permissions through the cloud management module until network-wide decision consistency is achieved. If all decision conflicts have been resolved and the network is operating stably, the decision optimization module continuously collects decision execution feedback data from each layer of nodes, dynamically adjusts the emergency event identification threshold, token allocation strategy, and permission boundary parameters, and controls the water conservancy project data acquisition equipment to continue executing the next round of monitoring and decision-making according to the optimized parameter configuration.

[0030] This IoT-based data acquisition method for water conservancy projects achieves comprehensive, dynamic monitoring and intelligent management of the project's operational status by constructing a three-layer intelligent decision-making network. First, by acquiring information on the distribution of monitoring points, the initial network topology, and the status and communication capabilities of each sensor node, a precise initial deployment foundation is provided. This ensures that the sensor nodes, virtual central node, and cloud management module are reasonably matched in terms of network structure, monitoring permissions, and decision-making scope, thereby constructing a fully functional and hierarchically distinct intelligent monitoring and decision-making architecture. Under this architecture, the sensor node module achieves real-time acquisition of various parameters of the water conservancy project, including water level, flow rate, water quality, and environmental conditions, ensuring high accuracy and timeliness of data acquisition. It also enables timely identification of preliminary anomalies, providing a reliable foundation for subsequent risk assessment and decision-making. The virtual central node module performs preliminary processing, fusion, and analysis of the data uploaded by the sensor nodes. It classifies monitoring events into urgency levels based on data change amplitude and preset risk thresholds, and dynamically divides time windows based on event levels, assigning corresponding decision priorities and processing permissions to different levels of events. This enables rapid response to emergencies and reasonable scheduling of routine events.

[0031] The token allocation module dynamically generates and distributes emergency tokens, coordination tokens, and communication tokens based on event level and time window configuration. This ensures that high-risk emergencies receive the highest processing priority, regional decisions gain collaborative permissions, and critical data is efficiently transmitted through priority transmission channels, thereby guaranteeing the smooth and reliable flow of information and instructions. During token usage, the token status and decision consistency of each decision-making node are continuously monitored. When permission conflicts or decision inconsistencies are detected, the module automatically adjusts and resolves conflicts based on priority principles, timeliness requirements, and energy consumption balance strategies, ensuring collaborative work among nodes and avoiding decision delays caused by permission conflicts or information lag. For incompletely resolved decision conflicts, a global coordination mechanism can be activated. The cloud management module reassesses and redistributes the decision permissions of each node, ensuring network-wide decision consistency and enabling the entire water conservancy project's data acquisition equipment to maintain stable operation and efficient collaboration even when multiple levels, nodes, and events occur simultaneously.

[0032] The decision optimization module, by collecting decision execution feedback data from nodes at each level, can dynamically adjust emergency event identification thresholds, token allocation strategies, and permission boundary parameters. It continuously optimizes its operational logic based on historical data and real-time feedback, forming a closed-loop adaptive management model. This optimization not only improves the response speed and accuracy of handling emergencies but also enhances the scientific nature of regional collaborative decision-making, while reducing overall energy consumption and resource waste. The entire method achieves a fully intelligent closed-loop process from on-site data acquisition, transmission, fusion, analysis, event classification, priority allocation, conflict resolution to global optimization. It significantly improves the real-time performance, accuracy, robustness, and adaptability of water conservancy project data acquisition equipment, providing a scientific and reliable technical guarantee for risk management, operation scheduling, and maintenance decision-making in complex water conservancy environments. This method enables stable, efficient, and intelligent operation in complex water conservancy engineering scenarios involving multiple events, nodes, and variables. It achieves efficient transmission and intelligent processing of monitoring data and decision-making instructions, significantly improving engineering safety, management efficiency, and resource utilization. Simultaneously, it provides a solid data foundation and decision support for long-term trend prediction and strategy optimization. This method significantly improves upon the problems of information lag, slow response, fragmented decision-making, and difficulties in overall coordination inherent in traditional water conservancy engineering monitoring methods, fully demonstrating the application value and technical effectiveness of IoT technology in intelligent monitoring and management of water conservancy projects.

[0033] The entire approach achieves a high degree of integration, automation, and intelligence in water conservancy project data acquisition equipment through refined monitoring data collection, intelligent event classification and priority management, dynamic token allocation and conflict adjustment, and adaptive decision optimization. This enables the equipment to operate continuously, stably, and efficiently in complex environments, significantly improving risk perception capabilities, decision response speed, resource utilization efficiency, and reliability. It provides a complete and feasible technical solution for the safe operation, scientific management, and maintenance of water conservancy projects.

[0034] Furthermore, the process of acquiring the monitoring point distribution information includes: GPS is used to locate each sensor node, obtain precise geographic coordinates, and establish a spatial distribution database of monitoring points; Based on the structural characteristics of water conservancy projects and hydrogeological conditions, the analytic hierarchy process (AHP) is used to calculate the critical weights of each region in order to distinguish between key monitoring areas and general areas. By conducting on-site surveys and analyzing historical monitoring data, we can identify vulnerable areas prone to accidents and key locations requiring close monitoring. Based on the technical parameters, measurement accuracy, and coverage of the monitoring equipment, calculate the optimal spacing between sensor nodes and evaluate the monitoring coverage. An optimization model for monitoring point distribution is constructed, and the optimal sensor layout scheme is solved using the particle swarm optimization algorithm to maximize the monitoring effect while minimizing equipment cost and maintenance difficulty.

[0035] As can be seen from the above, by establishing a precise spatial database through GPS positioning, analyzing the key weights of the region in conjunction with engineering structures and hydrogeological conditions, identifying vulnerable links and key parts prone to danger, calculating the optimal sensor spacing to assess coverage, and using particle swarm optimization to solve for the optimal layout scheme, we can achieve efficient monitoring of key areas, maximize monitoring coverage, and reduce equipment investment and maintenance costs.

[0036] Furthermore, the process of constructing the initial network topology includes the following steps: Based on the geographic location information and communication capability parameters of sensor nodes, the adjacency graph between nodes is established using the Deloitte triangulation algorithm. The communication quality evaluation indicators between nodes are determined by signal strength testing and path loss calculation. The minimum spanning tree algorithm is used to construct the basic network backbone topology, minimizing the overall communication overhead while ensuring network connectivity. Based on the basic network backbone topology, a multi-path network structure is formed by adding redundant links to improve the reliability and fault tolerance of the network. Establish a dynamic topology maintenance mechanism that can automatically adjust the network topology when node failures, the addition of new nodes, or link quality degradation caused by environmental changes are detected. Network simulation tools were used to verify the performance indicators of the topology, including network connectivity, average path length, and load balancing, to ensure that the network architecture meets the real-time and reliability requirements of water conservancy project monitoring.

[0037] As can be seen above, by establishing adjacency relationships based on the geographical location and communication capabilities of sensor nodes, and evaluating communication quality using signal strength and path loss, a minimum spanning tree backbone network is constructed and redundant links are added to form a multi-path structure, achieving a balance between network connectivity and reliability. At the same time, a dynamic topology maintenance mechanism is used to cope with node failures, environmental changes, and the addition of new nodes, ensuring that the network adapts and adjusts itself. Combined with simulation verification of network connectivity, path length, and load balancing, the stability, real-time performance, and overall fault tolerance of data transmission are improved.

[0038] Furthermore, obtaining the initial state of the node includes the following steps: Using the node's built-in self-testing circuit, the sensor accuracy, battery voltage level, storage space utilization, and communication module operating status are periodically checked, and a hardware health assessment report is generated. By querying software version information, verifying configuration parameters, and analyzing operation logs, the integrity and normal functionality of the node software can be confirmed. Measure the node's transmit power, calibrate the receiver sensitivity, and test the data transmission rate. Analyze channel interference and establish a communication performance profile for the node. Based on the collected hardware, software and communication status information, the analytic hierarchy process and fuzzy comprehensive evaluation method are used to calculate the comprehensive performance index and reliability level of each node. The initial configuration parameters, performance benchmark values, and historical operating data of each node are recorded into the node status database.

[0039] As can be seen from the above, by performing self-checks on sensor hardware health, battery voltage, storage usage, and communication module status, combined with software integrity verification and operation log analysis, measuring transmit power, receive sensitivity, and data transmission rate, assessing channel interference, and calculating the node's comprehensive performance indicators and reliability level based on hierarchical analysis and fuzzy comprehensive evaluation, and storing the node's initial configuration, performance benchmarks, and historical data in the database, a comprehensive understanding of the node's status and reliability assessment are achieved.

[0040] Furthermore, in the three-layer intelligent decision-making network architecture, the sensor node module is responsible for local data acquisition and primary anomaly detection, and is equipped with lightweight edge computing capabilities to identify basic anomalies such as data mutations and sensor failures; the virtual center node module undertakes regional data aggregation and intermediate risk analysis functions, integrates multi-source information fusion algorithms and engineering safety assessment models, and can identify regional risk events and cross-sensor correlation anomalies; the cloud management module provides global data management and advanced decision support, deploys deep learning models and expert knowledge bases, and can perform long-term trend prediction, complex fault diagnosis, and emergency plan recommendation.

[0041] As can be seen from the above, the sensor node module realizes local data acquisition and basic anomaly detection, and has edge computing capabilities to quickly identify data mutations and sensor failures; the virtual center node module completes regional data aggregation and intermediate risk analysis, and uses multi-source information fusion and safety assessment models to identify regional risks and cross-node anomalies; the cloud management module is responsible for global data management and advanced decision support, and relies on deep learning models and expert knowledge bases to perform trend prediction, complex fault diagnosis and emergency plan recommendation, realizing hierarchical intelligent management from local monitoring to regional assessment and then to global decision-making, improving the accuracy, response speed and overall decision-making capabilities of water conservancy project monitoring.

[0042] Furthermore, the allocation of monitoring permissions and decision-making scope includes the following steps: A role-based access control mechanism is adopted. The sensor node module is granted data acquisition permissions, local storage permissions, and basic alarm permissions, enabling it to autonomously perform data collection, anomaly detection, and emergency alarms. The virtual center node module is granted regional coordination permissions, data fusion permissions, and intermediate decision-making permissions to schedule sensor nodes within the region, perform multi-source data fusion analysis, and make regional control decisions. The cloud management module has global management permissions, system configuration permissions, and the highest decision-making permissions, and is responsible for the unified management, parameter configuration, and major decision-making of the entire monitoring network. Establish a permission inheritance and delegation mechanism, whereby superior nodes can temporarily authorize or revoke permissions for subordinate nodes, and subordinate nodes can apply to superior nodes for permission upgrades. Digital signatures and timestamps are used to ensure the security and immutability of permission allocation, preventing unauthorized access and abuse of permissions.

[0043] As can be seen from the above, through the role-based access control mechanism, sensor nodes are equipped with data acquisition, local storage, and basic alarm capabilities, enabling them to autonomously perform data collection and preliminary anomaly handling. The virtual central node has regional coordination, data fusion, and intermediate decision-making capabilities, and can schedule regional sensors and complete multi-source analysis and control decisions. The cloud management module has global management, system configuration, and the highest decision-making authority, and is responsible for unified management and major decision-making. Combined with the permission inheritance and delegation mechanism, upper and lower level nodes can dynamically authorize or apply for permission upgrades. At the same time, digital signatures and timestamps ensure the security and immutability of permission allocation, realizing secure, efficient, and hierarchical collaborative management of the monitoring network.

[0044] Furthermore, the real-time monitoring data includes water level changes, flow velocity fluctuations, pressure anomalies, temperature gradients, and structural displacements. Each parameter is configured with an independent acquisition frequency and accuracy requirement: water level and pressure parameters are acquired at millisecond-level high-frequency acquisition, flow velocity and temperature parameters are acquired at second-level medium-frequency acquisition, and structural displacements are acquired at minute-level low-frequency acquisition. During data transmission, a hierarchical compression and priority queuing strategy is adopted: urgent data is losslessly compressed and given the highest transmission priority, while regular data is lossily compressed to save bandwidth resources. When the network is interrupted or transmission fails, lost data can be automatically retransmitted, and missing time-series data can be compensated through interpolation algorithms to ensure the continuity and availability of monitoring data.

[0045] As can be seen from the above, by setting different acquisition frequencies and accuracies for parameters such as water level, pressure, flow velocity, temperature, and structural displacement, layered and precise monitoring of key indicators of water conservancy projects is achieved. During data transmission, hierarchical compression and priority queuing are adopted to ensure that urgent data is transmitted with high priority and without loss, while saving bandwidth for regular data. At the same time, automatic retransmission and interpolation compensation mechanisms are provided to effectively ensure data continuity and integrity, thereby improving the accuracy, reliability, and real-time response capability of monitoring information.

[0046] Furthermore, the preliminary processing includes the following steps: Statistical methods are used to identify and process outliers, including outlier detection based on the 3σ criterion, outlier screening based on interquartile range, and mutation point identification based on time series analysis. Heterogeneous data from different sensors are converted into standard JSON format to unify numerical precision, timestamp format and unit identifier; By combining network time protocol and local clock correction, the consistency of time base of multi-source data is ensured, and the time synchronization accuracy reaches the millisecond level. The evaluation is conducted from five dimensions: completeness, accuracy, consistency, timeliness, and credibility. A weighted scoring method is used to calculate the overall quality score. Low-quality data is marked and corresponding data repair or re-collection is triggered.

[0047] As can be seen from the above, outliers are identified through statistical methods, and data is cleaned using outlier detection, interquartile range screening, and time series mutation point analysis; heterogeneous sensor data is uniformly converted into a standard format, and numerical precision, timestamps, and unit identifiers are standardized; millisecond-level time synchronization is achieved by combining network time protocols and local clock correction; and data quality is weighted and evaluated from five dimensions: completeness, accuracy, consistency, timeliness, and reliability. Low-quality data is marked, repaired, or re-acquired, thereby improving the reliability, comparability, and analytical accuracy of monitoring data.

[0048] Furthermore, the risk assessment includes the following steps: Based on monitoring data from a single sensor, threshold comparison and trend analysis methods are used to identify local anomalies and determine the preliminary risk level for each monitoring point. By integrating data from multiple sensors within the region, fuzzy inference and neural networks are used to assess the overall security status of the region and generate regional risk indicators. Based on the risk status and interrelationships of various regions, Bayesian network and Markov chain models are used to predict the global risk situation. Based on the assessment results, corresponding response strategies and measures are triggered to achieve risk visualization and hierarchical management.

[0049] As can be seen from the above, by comparing thresholds of single sensors and analyzing trends, local anomalies can be identified and preliminary risk levels can be determined. By integrating multi-sensor data within the region, fuzzy inference and neural networks can be used to assess the overall safety status of the region and form risk indicators. Furthermore, by comprehensively considering the risks and mutual influences of each region, Bayesian networks and Markov chains can be used to predict the global risk situation. Based on the assessment results, corresponding response strategies can be triggered to achieve the visualization, hierarchical management and precise prevention of risks, significantly improving the overall safety monitoring capabilities and decision-making response efficiency of water conservancy projects.

[0050] Furthermore, the calculation of the magnitude of the data change includes the following steps: Obtain the current value and historical baseline value of the monitoring data as the basic input for the analysis of the magnitude of change; The absolute change is calculated by the difference between the current value and the benchmark value, and the relative rate of change is further obtained to reflect the magnitude and relative intensity of the data change. Based on time series methods, the rate of change of data is calculated using the first derivative or difference method to reveal the speed of change and inflection point characteristics. By using moving average, exponential smoothing, or regression analysis methods, we can predict future data trends and provide forward-looking information for early warning judgments. Based on the distribution characteristics, seasonal patterns, and operating cycles of historical data, a threshold system covering four categories—normal, alert, early warning, and emergency—is established. Combined with real-time monitoring characteristics and expert experience, a self-learning algorithm is used to dynamically adjust the threshold parameters. By comparing the absolute change, relative change rate, change speed, and trend results with the threshold system, the risk level is comprehensively determined and early warning information is output.

[0051] As can be seen from the above, by comparing the current value of monitoring data with historical benchmark values, the absolute change and relative rate of change are calculated to reveal the magnitude and intensity of data changes; the rate of change is identified by using time series derivatives or difference analysis; future trends are predicted by combining moving averages, exponential smoothing, and regression analysis, providing forward-looking information for early warning; a four-category threshold system is established based on historical data patterns and operating cycles, and the thresholds are dynamically adjusted through a self-learning algorithm; finally, the risk level is determined by comprehensively considering the change amount, rate of change, rate of change, and trend results, achieving real-time, scientific, and dynamic early warning output, and improving the accuracy and response capability of water conservancy project monitoring.

[0052] Furthermore, the urgency classification includes the following steps: Evaluation data were collected from five dimensions: degree of deviation, rate of change, scope of impact, historical similar events, and importance of the project. The indicators of each dimension were standardized from 0 to 1. The weight coefficients of each dimension were determined by the analytic hierarchy process (AHP), and the standardized scores of each dimension were weighted and summed to obtain a comprehensive urgency score. Based on the comprehensive urgency score, the urgency level is divided into four levels: Normal Level I, Attention Level II, Warning Level III, and Emergency Level IV, so as to achieve accurate identification of different risk levels; Based on the changing trends of continuous monitoring data, the risk level is automatically upgraded when it increases and automatically downgraded when the abnormal situation is controlled or eliminated, ensuring the dynamic real-time nature of the classification status. The classification results are matched and compared with the historical event database, and the experience of handling similar cases is referenced to provide auxiliary basis for current risk management and emergency decision-making. By verifying the results after the fact and providing continuous feedback, the correctness of the classification results is checked, and the classification algorithm and parameter settings are optimized accordingly to improve the stability and adaptability of the classification.

[0053] As can be seen from the above, by collecting information from five dimensions—data deviation, rate of change, scope of impact, historical similar events, and project importance—and standardizing the data, the Analytic Hierarchy Process (AHP) is used to determine the weights and calculate a comprehensive urgency score. Based on the score, the risk is divided into four levels: normal, attention, warning, and emergency, achieving accurate identification of different risk levels. The grading status is dynamically adjusted by combining continuous monitoring data, and the historical event database is used to provide handling references. At the same time, the algorithm and parameter settings are optimized through post-event verification and feedback, thereby improving the accuracy, real-time performance, and adaptability of urgency determination.

[0054] Furthermore, the dynamic division of the time window includes the following steps: Based on the nature of the monitoring task, the frequency of data collection, and the characteristics of the project operation, a comprehensive judgment is made on whether to adopt an event-driven, period-driven, or a combination of both strategies. In event-driven mode, the granularity of the time window is dynamically adjusted according to the urgency level of the monitored event: the normal level I uses an hourly time window, the attention level II uses a minute-level time window, the warning level III uses a second-level time window, and the emergency level IV uses a millisecond-level time window, so as to achieve rapid response and key data capture for different risk levels. In the cycle-driven mode, a fixed time window is set according to the operation cycle of water conservancy projects and natural laws for long-term trend and regularity analysis; Within the defined time window, the sliding window technique is used for data processing, and overlapping or non-overlapping modes are selected according to the monitoring objectives and analysis needs, thereby achieving a balance between data accuracy and computational load. During data processing, the frequency and magnitude of data changes are monitored in real time, and the window size, overlap ratio and update cycle are automatically adjusted to cope with changes in environmental conditions or data characteristics. Data analysis is conducted in parallel at different time scales, combining short-term responses with long-term trends to generate multi-level, panoramic monitoring results.

[0055] As can be seen from the above, by combining the nature of the monitoring task, the frequency of data acquisition, and the characteristics of engineering operation, and flexibly adopting event-driven, period-driven, or hybrid strategies, a refined response to different monitoring needs can be achieved. In event-driven mode, the window granularity is dynamically adjusted according to the risk level, from hourly to millisecond-level, to quickly capture key data and improve response speed. In period-driven mode, a fixed window is set for long-term trend analysis. By using sliding window technology and adjusting the overlap ratio, a balance between data accuracy and computational load can be achieved. By automatically adjusting window parameters through real-time detection of data changes, and simultaneously analyzing short-term events and long-term trends in parallel at different time scales, multi-level, panoramic monitoring results are generated, significantly improving monitoring accuracy, response efficiency, and adaptability.

[0056] Furthermore, the decision priority allocation includes the following steps: A five-level priority system is established to classify and manage different types of events: ultra-high priority is used for emergency events related to life safety, high priority is used for important events related to engineering safety, medium priority is used for equipment failures and operational anomalies, low priority is used for routine maintenance and data statistics, and lowest priority is used for historical data queries and system optimization. An independent processing queue is set up for each priority level, and corresponding computing resources and communication bandwidth are configured. High-priority tasks can preempt the resources of low-priority tasks to ensure the timely processing of critical tasks. Dynamically modify task priorities based on the development trend of the event and the effectiveness of the handling to prevent resource waste caused by rigid priorities, while maintaining continuous and efficient response to critical tasks; Set a maximum waiting time for low-priority tasks and trigger priority promotion rules based on task waiting status or importance to ensure that long-term low-priority tasks are not ignored and achieve a balance in resource allocation. When multiple tasks of equal priority arrive at the same time, they are sorted by timestamp, task importance and resource requirements to avoid processing conflicts. During operation, the processing efficiency and resource utilization of tasks at all levels are continuously monitored, and priority allocation strategies and queue management parameters are adjusted based on feedback to optimize scheduling performance.

[0057] As can be seen from the above, by constructing a five-level event priority system, different types of events such as life safety, engineering safety, equipment failure, routine maintenance, and historical data query are classified and managed. Independent queues, computing resources, and communication bandwidth are allocated to each level of task, and high-priority tasks can preempt low-priority resources to ensure timely handling of critical events. Priorities are dynamically adjusted based on event development trends and handling effects to avoid resource waste. Waiting times are set for low-priority tasks and escalation rules are triggered to ensure that long-term tasks are not ignored. Conflicts between tasks of the same level are handled through comprehensive sorting, and task processing efficiency and resource utilization are continuously monitored to dynamically optimize scheduling strategies, achieving high efficiency in task processing, timely response, and balanced resource allocation.

[0058] Furthermore, the processing permissions adopt a distributed permission management architecture, with each decision level configured with a corresponding permission management node responsible for the allocation, monitoring, and revoke of permissions at its level; permission types include four categories: data access permissions, device control permissions, system configuration permissions, and emergency response permissions; permissions of higher-level nodes can be partially or fully inherited by lower-level nodes, and temporary authorization and timed revoke of permissions are supported.

[0059] As can be seen from the above, by configuring permission management nodes at each decision-making level through a distributed architecture, the allocation, monitoring, and revocation of permissions for data access, equipment control, system configuration, and emergency response are realized. Permissions of higher-level nodes can be partially or fully delegated to lower-level nodes, and temporary authorization and timed revocation mechanisms are supported, thereby ensuring flexible control, security, reliability, and efficient execution of permissions at each level, and improving the operation management and emergency response capabilities of the water conservancy project monitoring and decision-making system.

[0060] Furthermore, the emergency token adopts a design scheme that combines timeliness and specialization. The token content includes a unique identifier, generation timestamp, validity period, scope of permissions, and digital signature information to ensure the security and non-forgeability of the token. The generation trigger conditions for the emergency token include monitoring data exceeding the danger threshold, sensor equipment failure, communication link interruption, and manual emergency call, supporting both automatic generation and manual application modes. Token allocation adopts the principles of proximity and capability matching, prioritizing allocation to nodes closest to the event and with the corresponding processing capabilities. By tracking the usage status, execution progress, and processing results of the token in real time, it is ensured that emergency events are handled promptly and effectively. When the emergency event is handled and the token expires, the token is automatically revoked and related permissions are cleared to prevent permission abuse and security vulnerabilities.

[0061] As can be seen from the above, by combining timeliness and specialization in its design, the token contains a unique identifier, timestamp, validity period, scope of permissions, and digital signature to ensure security and prevent forgery. When monitoring data exceeds the threshold, equipment malfunctions, communication is interrupted, or manual calls are made, tokens can be automatically or manually generated and allocated to the most suitable node according to the principle of proximity and capability matching. The token usage status and processing progress are tracked in real time to ensure timely and effective handling of emergencies. After the event is completed or the token expires, permissions are automatically revoked and token information is cleared, thereby improving emergency response efficiency, security, and management reliability.

[0062] Furthermore, the coordination token employs a distributed negotiation and consensus algorithm to ensure collaborative decision-making among multiple nodes. The token content includes a coordination region identifier, a list of participating nodes, a description of the coordination task, an expected completion time, and coordination rule information. The allocation of coordination tokens adopts a balancing strategy based on node capabilities and load conditions. A capability assessment algorithm is used to calculate the processing capability score of each node, and the optimal allocation is made in combination with the current load conditions. Point-to-point coordination, intra-regional coordination, or cross-regional coordination modes are selected according to the complexity of the task and the scope of impact. Distributed locks and token ring technology are used to prevent coordination conflicts and ensure that the same resource can only be occupied by one coordination task at the same time. When a coordination node failure, network partition, or coordination timeout anomaly is detected, a backup coordination scheme is automatically activated or the coordination task is reassigned.

[0063] As can be seen from the above, multi-node collaborative decision-making is achieved through distributed negotiation and consensus algorithms. The token contains information such as the coordination area, participating nodes, task description, expected completion time, and coordination rules. During allocation, the node processing capacity and load conditions are considered for balanced optimization, and point-to-point, intra-regional, or cross-regional coordination modes are selected according to the task complexity. Distributed locks and token ring technology are used to prevent resource conflicts and ensure that the same resource is occupied by only a single task at the same time. In the event of coordination node failure, network partitioning, or task timeout, backup plans can be automatically activated or tasks can be reallocated, thereby ensuring the reliability, continuity, and efficiency of multi-node collaboration.

[0064] Furthermore, the communication token employs a dynamic bandwidth allocation and quality of service (QoS) assurance mechanism. The token content includes communication priority, bandwidth quota, latency requirements, reliability level, and transmission security level. The priority transmission channel for critical data utilizes dedicated frequency bands and redundant paths, ensuring real-time reliable transmission of critical data through channel contention and path selection algorithms. Channel utilization, transmission delay, packet loss rate, and signal strength indicators are monitored in real time, and communication parameters and transmission strategies are dynamically adjusted based on the monitoring results. The optimal transmission rate and retransmission count are dynamically selected based on channel conditions and data importance, maximizing transmission efficiency while ensuring transmission reliability. A fairness guarantee mechanism for the communication token is established to prevent high-priority data from occupying communication resources for extended periods, thus preventing low-priority data from being transmitted. Minimum bandwidth guarantees and maximum occupation time limits are set.

[0065] As can be seen from the above, the dynamic bandwidth allocation and quality of service guarantee mechanism achieves priority transmission and high reliability of critical data; the token includes communication priority, bandwidth quota, latency requirements, reliability and security level; the critical data channel adopts dedicated frequency band and redundant path design, and combines channel contention and path selection algorithms to ensure real-time transmission; channel utilization, latency, packet loss rate and signal strength are monitored in real time, and transmission parameters and strategies are dynamically adjusted. The transmission rate and retransmission number are optimized according to channel conditions and data importance. At the same time, the fairness mechanism restricts the occupation of high-priority data, sets minimum bandwidth and maximum occupation time, and ensures the transmission of low-priority data, thus achieving a balance between communication efficiency, reliability and resource utilization.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A water conservancy engineering data acquisition device based on the Internet of Things, characterized in that, The system comprises a sensor node module, a virtual central node module, a cloud management module, a token allocation module, and a decision optimization module. Specifically: the sensor node module collects various monitoring parameters in real time at the water conservancy project site and performs preliminary anomaly detection; the virtual central node module aggregates and merges multi-source data to complete regional risk assessment and generate early warning information; the cloud management module centrally stores and analyzes global data, providing trend prediction and decision support; the token allocation module dynamically allocates emergency, coordination, and communication tokens based on event levels to ensure priority handling of critical events and efficient information transmission; and the decision optimization module dynamically adjusts event identification thresholds, token allocation strategies, and permission boundary parameters based on execution feedback.

2. The water conservancy engineering data acquisition device based on the Internet of Things according to claim 1, characterized in that, The sensor node module includes a water level detection unit, a flow rate detection unit, a water quality detection unit, and an environmental detection unit. Specifically, the water level detection unit uses piezoresistive or ultrasonic ranging principles to achieve millimeter-level accuracy in water level monitoring; the flow rate detection unit measures water flow velocity and flow rate changes through the Doppler effect or electromagnetic induction principle; the water quality detection unit integrates a pH detector, a dissolved oxygen detector, a turbidity detector, and a heavy metal ion detector to achieve real-time multi-parameter water quality analysis; and the environmental detection unit is equipped with temperature and humidity sensors, wind speed and direction sensors, rainfall sensors, and air pressure sensors to monitor meteorological conditions that affect the operation of water conservancy projects.

3. The water conservancy engineering data acquisition device based on the Internet of Things according to claim 2, characterized in that, The virtual central node module adopts a hierarchical data processing architecture, including a data receiving layer, a fusion analysis layer, and a decision output layer. The data receiving layer is configured with a multi-protocol communication interface to ensure data transmission stability in complex electromagnetic environments. The fusion analysis layer uses a weighted average method, Kalman filtering algorithm, and Bayesian inference method to perform spatiotemporal fusion processing on heterogeneous data from different sensor nodes, thereby eliminating the influence of sensor measurement errors and environmental interference and generating high-precision comprehensive monitoring results. The decision output layer, based on fuzzy control theory and expert system rule base, realizes intelligent assessment of the operating status of water conservancy projects and generates early warning information.

4. The water conservancy engineering data acquisition device based on the Internet of Things according to claim 3, characterized in that, The cloud management module adopts a microservice architecture design, including a data storage service unit, a computing and analysis service unit, a visualization service unit, and a system management service unit. The data storage service unit uses a hybrid storage architecture combining a distributed database cluster and a time-series database, enabling efficient storage and management of massive amounts of time-series monitoring data, and supporting rapid data querying and historical backtracking analysis. The computing and analysis service unit provides real-time stream computing, batch data analysis, and predictive modeling functions. The visualization service unit uses responsive web technology to achieve multi-dimensional data chart display, geographic information system integration, and mobile adaptation. The system management service unit is responsible for user permission management, system configuration maintenance, and operational status monitoring to ensure the stable operation of the entire cloud platform.

5. A method for acquiring water conservancy project data based on the Internet of Things (IoT), implemented based on the IoT-based water conservancy project data acquisition device described in claim 4, characterized in that... Includes the following steps: Acquire the distribution information of monitoring points and the initial network topology at the site of the water conservancy project, and obtain the initial state and communication capability parameters of each sensor node; Based on the initial state of nodes, the distribution information of monitoring points, and the network topology, a three-layer intelligent decision-making network architecture is constructed, consisting of sensor node modules, virtual center node modules, and cloud management modules, and corresponding monitoring permissions and decision-making scopes are assigned to each layer. The sensor node module collects real-time monitoring data of the water conservancy project and transmits it to the virtual center node module for preliminary processing and risk assessment. Whenever monitoring data is received, the urgency of the current monitoring event is classified based on the magnitude of data change and preset risk thresholds. Time windows are dynamically divided according to the event level, and corresponding decision priorities and processing permissions are assigned to events of different levels. Based on the event level and time window configuration, the token allocation module dynamically generates and allocates emergency tokens, coordination tokens, and communication tokens to ensure that emergencies receive the highest processing priority, regional decisions receive collaborative permissions, and key data receives priority transmission channels. During the token allocation process, the token usage status and decision consistency of each decision node are continuously monitored. When a permission conflict or decision inconsistency is detected, the conflict is automatically resolved based on priority principles, timeliness requirements, and energy consumption balance strategies. If a decision conflict is detected to be unresolved, a global coordination mechanism is activated to reassess and reassign decision permissions through the cloud management module until network-wide decision consistency is achieved. If all decision conflicts have been resolved and the network is operating stably, the decision optimization module continuously collects decision execution feedback data from each layer of nodes, dynamically adjusts the emergency event identification threshold, token allocation strategy, and permission boundary parameters, and controls the water conservancy project data acquisition equipment to continue executing the next round of monitoring and decision-making according to the optimized parameter configuration.

6. The method for acquiring water conservancy project data based on the Internet of Things according to claim 5, characterized in that, In the three-layer intelligent decision-making network architecture, the sensor node module is responsible for local data acquisition and primary anomaly detection, and is equipped with lightweight edge computing capabilities, which can identify basic anomalies such as data mutations and sensor failures; the virtual center node module undertakes regional data aggregation and intermediate risk analysis functions, integrates multi-source information fusion algorithms and engineering safety assessment models, and can identify regional risk events and cross-sensor correlation anomalies. The cloud management module provides global data management and advanced decision support, deploys deep learning models and expert knowledge bases, and is capable of long-term trend prediction, complex fault diagnosis, and emergency plan recommendation.

7. The method for acquiring water conservancy project data based on the Internet of Things according to claim 5, characterized in that, The real-time monitoring data includes water level changes, flow velocity fluctuations, pressure anomalies, temperature gradients, and structural displacements. Each parameter is configured with an independent acquisition frequency and accuracy requirement: water level and pressure parameters are acquired at millisecond-level high frequency, flow velocity and temperature parameters at second-level medium frequency, and structural displacements at minute-level low frequency. During data transmission, a hierarchical compression and priority queuing strategy is adopted: urgent data is losslessly compressed and given the highest transmission priority, while regular data is lossily compressed to save bandwidth resources. When the network is interrupted or transmission fails, lost data can be automatically retransmitted, and missing time-series data can be compensated for through interpolation algorithms to ensure the continuity and availability of monitoring data.

8. The method for acquiring water conservancy project data based on the Internet of Things according to claim 5, characterized in that, The calculation of the magnitude of the data change includes the following steps: Obtain the current value and historical baseline value of the monitoring data as the basic input for the analysis of the magnitude of change; The absolute change is calculated by the difference between the current value and the benchmark value, and the relative rate of change is further obtained to reflect the magnitude and relative intensity of the data change. Based on time series methods, the rate of change of data is calculated using the first derivative or difference method to reveal the speed of change and inflection point characteristics. By using moving average, exponential smoothing, or regression analysis methods, we can predict future data trends and provide forward-looking information for early warning judgments. Based on the distribution characteristics, seasonal patterns, and operating cycles of historical data, a threshold system covering four categories—normal, alert, early warning, and emergency—is established. Combined with real-time monitoring characteristics and expert experience, a self-learning algorithm is used to dynamically adjust the threshold parameters. By comparing the absolute change, relative change rate, change speed, and trend results with the threshold system, the risk level is comprehensively determined and early warning information is output.

9. A method for acquiring water conservancy project data based on the Internet of Things according to claim 5, characterized in that, The urgency level classification includes the following steps: Evaluation data were collected from five dimensions: degree of deviation, rate of change, scope of impact, historical similar events, and importance of the project. The indicators of each dimension were standardized from 0 to 1. The weight coefficients of each dimension were determined by the analytic hierarchy process (AHP), and the standardized scores of each dimension were weighted and summed to obtain a comprehensive urgency score. Based on the comprehensive urgency score, the urgency level is divided into four levels: Normal Level I, Attention Level II, Warning Level III, and Emergency Level IV, so as to achieve accurate identification of different risk levels; Based on the changing trends of continuous monitoring data, the risk level is automatically upgraded when it increases and automatically downgraded when the abnormal situation is controlled or eliminated, ensuring the dynamic real-time nature of the classification status. The classification results are matched and compared with the historical event database, and the experience of handling similar cases is referenced to provide auxiliary basis for current risk management and emergency decision-making. By verifying the results after the fact and providing continuous feedback, the correctness of the classification results is checked, and the classification algorithm and parameter settings are optimized accordingly to improve the stability and adaptability of the classification.

10. A method for acquiring water conservancy project data based on the Internet of Things according to claim 5, characterized in that, The decision priority allocation includes the following steps: A five-level priority system is established to classify and manage different types of events: ultra-high priority is used for emergency events related to life safety, high priority is used for important events related to engineering safety, medium priority is used for equipment failures and operational anomalies, low priority is used for routine maintenance and data statistics, and lowest priority is used for historical data queries and system optimization. An independent processing queue is set up for each priority level, and corresponding computing resources and communication bandwidth are configured. High-priority tasks can preempt the resources of low-priority tasks to ensure the timely processing of critical tasks. Dynamically modify task priorities based on the development trend of the event and the effectiveness of the handling to prevent resource waste caused by rigid priorities, while maintaining continuous and efficient response to critical tasks; Set a maximum waiting time for low-priority tasks and trigger priority promotion rules based on task waiting status or importance to ensure that long-term low-priority tasks are not ignored and achieve a balance in resource allocation. When multiple tasks of equal priority arrive at the same time, they are sorted by timestamp, task importance and resource requirements to avoid processing conflicts. During operation, the processing efficiency and resource utilization of tasks at all levels are continuously monitored, and priority allocation strategies and queue management parameters are adjusted based on feedback to optimize scheduling performance.

Images