Hydraulic power plant safety risk dynamic evaluation system and method based on multi-source data fusion

Abstract

The invention relates to the technical field of hydraulic power plant safety management and risk assessment, and discloses a hydraulic power plant safety risk dynamic evaluation system and method based on multi-source data fusion. The system comprises a central processing unit, a storage unit, a video monitoring module, an environment sensing module, a personnel positioning module, a task management system, an emergency response database, a data integration unit, a risk assessment engine and an output module. The multi-source heterogeneous data is cleaned, aligned and fused through the data integration unit, the fused data is dynamically scored and graded by combining a risk assessment engine with a machine learning model, and a visual risk map and an automatic report are generated. According to the invention, real-time monitoring and dynamic risk assessment of the operation state of the hydraulic power plant can be realized, the safety management efficiency and accuracy are improved, and technical support is provided for stable power supply.

Description

Technical Field

[0001] This invention relates to the field of hydropower plant safety management and risk assessment technology, and in particular to a dynamic evaluation system and method for hydropower plant safety risks based on multi-source data fusion. Background Technology

[0002] As hydropower plants continue to expand in scale, their operational safety and risk management capabilities have become crucial factors in ensuring stable power supply. Hydropower plant safety management involves multiple factors, including equipment status, environmental conditions, and personnel behavior. Traditional safety management methods primarily rely on manual inspections and single-sensor monitoring. While these methods can reflect local conditions to some extent, they lack the ability to comprehensively analyze multi-source data.

[0003] Currently, some hydropower plants have introduced independent systems such as video surveillance and environmental sensors for auxiliary management. However, the data from these systems is isolated, making it difficult to form a unified risk assessment framework. Furthermore, traditional risk assessment methods are typically based on static models, which cannot reflect the dynamic risk status of hydropower plants in real time.

[0004] Therefore, existing technologies still have room for improvement in multi-source data fusion, dynamic risk assessment, and intelligent early warning response. There is an urgent need for a technical solution that can integrate multimodal data and provide real-time assessment to improve the safety management level of hydropower plants. Summary of the Invention

[0005] The present invention aims to at least partially solve one of the technical problems in the related art.

[0006] Therefore, the first objective of this invention is to propose a dynamic evaluation system for the safety risks of hydropower plants based on multi-source data fusion.

[0007] The second objective of this invention is to propose a dynamic evaluation method for the safety risks of hydropower plants based on multi-source data fusion.

[0008] The third objective of this invention is to provide an electronic device.

[0009] The fourth objective of this invention is to provide a computer-readable storage medium.

[0010] The fifth objective of this invention is to provide a computer program product.

[0011] To achieve the above objectives, a first aspect of the present invention proposes a dynamic safety risk assessment system for hydropower plants based on multi-source data fusion, comprising: The system includes: a central processing unit (CPU); a storage unit connected to the CPU for storing collected data and processing results; a video monitoring module connected to the storage unit for real-time acquisition of image information inside and outside the hydropower plant; an environmental sensing module connected to the storage unit for monitoring environmental parameters such as temperature, humidity, gas concentration, and water level; a personnel positioning module connected to the storage unit for obtaining the location information of on-site workers; a task management system connected to the storage unit for recording the status information of electronic work orders; an emergency response database connected to the storage unit for storing emergency plans and historical handling records; a data integration unit connected to the storage unit via an IoT gateway for integrating and preprocessing multi-source heterogeneous data; a risk assessment engine connected to the data integration unit for dynamically scoring and classifying the fused data using a machine learning model; and an output module connected to the risk assessment engine for generating visualized risk maps, early warning notifications, and emergency response suggestions, and supporting the automatic generation of reports in industry standard formats.

[0012] In one embodiment, the environmental sensing module includes: a temperature and humidity sensor connected to the storage unit for real-time monitoring of temperature and humidity changes inside the hydropower plant; a gas detector connected to the storage unit for measuring the concentration of harmful gases; and a level gauge connected to the storage unit for detecting water level in key areas.

[0013] In one embodiment, the hydropower plant safety risk dynamic assessment system based on multi-source data fusion further includes a vibration detection module, which is connected to the storage unit and is used to measure the vibration frequency and amplitude of key equipment.

[0014] In one embodiment, the hydropower plant safety risk dynamic assessment system based on multi-source data fusion further includes an image recognition unit connected to the storage unit, used for target detection and behavior analysis of the collected image data.

[0015] In one embodiment, the hydropower plant safety risk dynamic assessment system based on multi-source data fusion further includes an alarm module connected to the central processing unit, used to trigger an alarm when the risk reaches a set threshold.

[0016] To achieve the above objectives, a second aspect of the present invention proposes a method for dynamic risk assessment of hydropower plant safety risks using the multi-source data fusion-based dynamic risk assessment system for hydropower plants as described above, comprising: The video monitoring module collects real-time image information of the hydropower plant's interior and exterior and stores it in the storage unit; the environmental sensing module monitors environmental parameters such as temperature, humidity, gas concentration, and water level and stores them in the storage unit; the personnel positioning module obtains the location information of on-site workers and stores it in the storage unit; the task management system records the status information of electronic work orders and stores it in the storage unit; the data integration unit cleans, aligns, and merges the collected multi-source data to construct a unified risk data pool; the risk assessment engine dynamically scores and classifies the merged data; and the output module generates a visualized risk map, early warning notifications, and emergency response suggestions, and supports the automatic generation of reports in industry standard formats.

[0017] In one embodiment, dynamically scoring and classifying the fused data through the risk assessment engine includes: acquiring the fused data through the risk assessment engine, the fused data including image data, environmental parameter data, personnel location data, and task status data; using the risk assessment engine to call a trained machine learning model to perform feature extraction and pattern recognition on the fused data; calculating a risk score based on the extracted features through the risk assessment engine, and classifying the risk level based on the score result.

[0018] In one embodiment, the process of using the risk assessment engine to call a trained machine learning model to perform feature extraction and pattern recognition on the fused data includes: using the risk assessment engine to call a convolutional neural network (CNN) to extract features from the image data and identify potential security risks; using the risk assessment engine to call a long short-term memory network (LSTM) to model the time series data and predict risk development trends; and using the risk assessment engine to call an ensemble learning model to perform comprehensive analysis of multimodal data and improve the accuracy of risk assessment.

[0019] In one embodiment, after the risk assessment engine calculates a risk score based on extracted features and classifies risk levels according to the score results, the method further includes: if the risk score exceeds a preset threshold, triggering an alarm through the alarm module; pushing a warning notification to relevant personnel through the output module; matching the best handling plan for the current risk scenario through the emergency response database, and generating emergency response suggestions.

[0020] In one embodiment, after generating a visualized risk map, early warning notification, and emergency response recommendations through the output module, the method further includes: recording the results of each risk assessment through the output module to form a risk evolution map; and analyzing risk change trends through the risk evolution map to optimize the response strategies in the emergency plan library.

[0021] The technical solutions provided by the embodiments of the present invention bring at least the following beneficial effects: The risk dynamic assessment system and method provided by the present invention acquires real-time image information of the inside and outside of the hydropower plant through the video monitoring module, monitors environmental parameters such as temperature, humidity, gas concentration, and water level through the environmental perception module, obtains the location information of on-site workers through the personnel positioning module, and records the status information of electronic work orders through the task management system. The data integration unit cleans, aligns, and merges multi-source heterogeneous data to form a unified risk data pool. The risk assessment engine uses a trained machine learning model to dynamically score and classify the merged data, generating a visualized risk map and automated reports. The output module pushes early warning notifications and emergency response suggestions to relevant personnel and records the results of each risk assessment to form a risk evolution map. Through comprehensive analysis of multi-source data, the present invention realizes real-time monitoring and dynamic risk assessment of the hydropower plant's operating status, improves the efficiency and accuracy of safety management, and provides technical support for ensuring stable power supply.

[0022] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0023] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic diagram of the overall architecture of the system of the present invention; Figure 2 This is a schematic flowchart of the method of the present invention; Figure 3 This is a schematic diagram illustrating the composition of the environmental sensing module of the present invention; Figure 4 This is a schematic diagram illustrating the working principle of the risk assessment engine of this invention.

[0024] The system includes: 1. Central Processing Unit; 2. Storage Unit; 3. Video Surveillance Module; 4. Environmental Sensing Module; 5. Personnel Positioning Module; 6. Task Management System; 7. Emergency Response Database; 8. Data Integration Unit; 9. Risk Assessment Engine; 10. Output Module; 11. Temperature and Humidity Sensor; 12. Gas Detector; and 13. Liquid Level Gauge. Detailed Implementation

[0025] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments 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.

[0026] This invention provides a dynamic assessment system and method for the safety risks of hydropower plants based on multi-source data fusion, as detailed below. Figure 1 To be continued Figure 4 The specific embodiments of the present invention will be described in detail below. Figure 1 This is a schematic diagram of the overall system architecture of the present invention, showing the connection relationship between the central processing unit 1, storage unit 2, video monitoring module 3, environmental perception module 4, personnel positioning module 5, task management system 6, emergency response database 7, data integration unit 8, risk assessment engine 9, and output module 10. Figure 2 This is a flowchart of the method of the present invention, which describes the complete process from multi-source data collection to risk assessment and output. Figure 3 This is a schematic diagram of the composition of the environmental sensing module 4, showing the connection method and functional distribution of the temperature and humidity sensor 11, gas detector 12, and liquid level gauge 13 with the storage unit 2. Figure 4 This diagram illustrates the working principle of Risk Assessment Engine 9, showcasing the process by which machine learning models extract features, recognize patterns, and score risks from image data, time-series data, and multimodal data.

[0027] In this embodiment, the central processing unit 1 is the core control unit, connected to the storage unit 2 via a bus, and is used to coordinate data processing and instruction transmission among various modules. The storage unit 2 is internally divided into multiple partitions, used to store image information acquired by the video surveillance module 3, environmental parameters monitored by the environmental perception module 4, location information obtained by the personnel positioning module 5, electronic work order status information recorded by the task management system 6, and emergency plans and historical handling records from the emergency response database 7. The storage unit 2 also provides temporary cache space for the data integration unit 8 to support the cleaning, alignment, and fusion of multi-source heterogeneous data.

[0028] The video surveillance module 3 connects to the storage unit 2 via a network interface, acquiring real-time image information from inside and outside the hydropower plant, and storing the acquired image data in a designated partition of the storage unit 2 according to a preset format. The cameras in the video surveillance module 3 are deployed in key areas of the hydropower plant, including generator units, control rooms, and dam spillways, ensuring comprehensive coverage without blind spots. The image acquisition frequency of the video surveillance module 3 can be adjusted according to actual needs; for example, the acquisition frequency can be increased during high-risk periods to enhance monitoring effectiveness.

[0029] The environmental sensing module 4 consists of a temperature and humidity sensor 11, a gas detector 12, and a level gauge 13, such as... Figure 3As shown. Temperature and humidity sensors 11 are installed in densely equipped areas within the hydropower plant to monitor temperature and humidity changes in real time and transmit the data to storage unit 2 via wired or wireless means. Gas detectors 12 are deployed in areas where hazardous gas leaks may occur, such as generator rooms and chemical reagent storage areas, to measure the concentration of hazardous gases and store the data in storage unit 2. Level gauges 13 are installed at water level monitoring points in critical areas, such as reservoirs and drainage pipes, to detect water level height and upload the data to storage unit 2. Each sensor in the environmental sensing module 4 has a self-calibration function, enabling it to automatically adjust the measurement accuracy according to environmental conditions.

[0030] Personnel positioning module 5 connects to storage unit 2 via wireless communication technology to acquire the location information of on-site workers. The positioning terminal of personnel positioning module 5 is worn by the worker, calculates location coordinates by receiving base station signals, and uploads the location information to storage unit 2 in real time. The positioning accuracy of personnel positioning module 5 can reach the meter level, accurately reflecting the activity trajectory and distribution of workers. In addition, personnel positioning module 5 also supports setting up electronic fences, triggering an alarm signal when a worker enters a hazardous area.

[0031] The task management system 6 is connected to storage unit 2 via a local area network to record the status information of electronic work orders. Each electronic work order includes the task name, executor, start time, end time, and task progress. The task management system 6 updates this information to storage unit 2 in real time. The task management system 6 also supports task priority settings and task reminders to ensure that critical tasks are completed in a timely manner. The emergency response database 7 is connected to storage unit 2 to store emergency plans and historical handling records. Emergency plans include handling procedures and resource allocation schemes for different risk scenarios, while historical handling records document the handling process and results of past risk events, providing a reference for subsequent risk assessments.

[0032] Data integration unit 8 connects to storage unit 2 via an IoT gateway and is responsible for integrating and preprocessing multi-source heterogeneous data. First, data integration unit 8 cleans the collected data to remove noise and outliers. Then, it performs time and spatial alignment to ensure that data from different sources can be compared within the same time and spatial range. Data integration unit 8 also employs data fusion algorithms to comprehensively process image data, environmental parameter data, personnel location data, and task status data, forming a unified risk data pool. The data in the risk data pool is stored in a structured format for easy subsequent analysis and processing.

[0033] The risk assessment engine 9 is connected to the data integration unit 8, and uses a machine learning model to dynamically score and classify the fused data. For example... Figure 4As shown, Risk Assessment Engine 9 first uses a convolutional neural network to extract features from image data, identifying potential safety hazards such as signs of equipment malfunction or personnel misconduct. Next, it uses a long short-term memory network to model time-series data, predicting risk trends. Finally, it uses an ensemble learning model to comprehensively analyze multimodal data, improving the accuracy of risk assessment. Based on the extracted features, Risk Assessment Engine 9 calculates a risk score and classifies risks into levels such as low, medium, and high risk.

[0034] Output module 10 connects to risk assessment engine 9 to generate visualized risk maps, early warning notifications, and emergency response recommendations. The visualized risk map displays risk distribution in heatmap format, with color intensity indicating risk level. Early warning notifications are pushed to relevant personnel via SMS or email to alert them to potential risks. Emergency response recommendations match the best handling plan based on the current risk scenario and generate detailed execution steps. Output module 10 also supports automatically generating industry-standard format reports, including risk assessment results, early warning information, and emergency response recommendations.

[0035] In this embodiment, the system also includes a vibration detection module and an image recognition unit as extended functional modules. The vibration detection module is connected to the storage unit 2 via wired or wireless means and is used to measure the vibration frequency and amplitude of key equipment. The sensors of the vibration detection module are mounted on the equipment housing and can monitor the equipment's operating status in real time and upload vibration data to the storage unit 2. The image recognition unit is connected to the storage unit 2 and is used to perform target detection and behavior analysis on the collected image data. The image recognition unit can identify abnormal behavior or violations and transmit relevant information to the risk assessment engine 9 to adjust the risk score.

[0036] Based on the aforementioned hardware architecture, this invention also provides a method for dynamic risk assessment using this system. First, the video monitoring module 3 collects real-time image information of the hydropower plant's interior and exterior, and stores it in storage unit 2. Next, the environmental sensing module 4 monitors environmental parameters such as temperature, humidity, gas concentration, and water level, and stores them in storage unit 2. Simultaneously, the personnel positioning module 5 acquires the location information of on-site workers and stores it in storage unit 2. The task management system 6 records the status information of electronic work orders and stores it in storage unit 2. Subsequently, the data integration unit 8 cleans, aligns, and merges the collected multi-source data to construct a unified risk data pool. The risk assessment engine 9 dynamically scores and classifies the merged data. Finally, the output module 10 generates a visualized risk map, early warning notifications, and emergency response suggestions, and supports the automatic generation of reports in industry-standard formats.

[0037] During the risk assessment process, Risk Assessment Engine 9 first acquires fused data, including image data, environmental parameter data, personnel location data, and task status data. Next, Risk Assessment Engine 9 uses a pre-trained machine learning model to perform feature extraction and pattern recognition on the fused data. For image data, Risk Assessment Engine 9 uses a convolutional neural network for feature extraction to identify potential safety hazards. For time-series data, Risk Assessment Engine 9 uses a long short-term memory network for modeling to predict risk development trends. For multimodal data, Risk Assessment Engine 9 uses an ensemble learning model for comprehensive analysis to improve the accuracy of risk assessment. Risk Assessment Engine 9 calculates a risk score based on the extracted features and classifies the risk level according to the score results.

[0038] If the risk score exceeds a preset threshold, an alarm is triggered via the alarm module. The alarm module sends an alarm signal to relevant personnel via an audible and visual alarm or a mobile application. Simultaneously, the output module 10 pushes a warning notification to relevant personnel and matches the best handling plan for the current risk scenario using the emergency response database 7, generating emergency response suggestions. Furthermore, the risk assessment engine 9 assesses the equipment's health status using vibration data and incorporates this into the risk score calculation process. The image recognition unit 15 identifies abnormal behavior or violations and transmits the relevant information to the risk assessment engine 9 to adjust the risk score.

[0039] After generating a visual risk map, early warning notices, and emergency response recommendations through output module 10, output module 10 records the results of each risk assessment, forming a risk evolution map. The risk evolution map displays risk change trends in a timeline format, facilitating managers to analyze risk development patterns and optimize response strategies in the emergency plan library.

[0040] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A dynamic safety risk assessment system for hydropower plants based on multi-source data fusion, characterized in that, include: The central processing unit (1) is used to coordinate the data processing and instruction transmission of each module; Storage unit (2), connected to the central processing unit (1), is used to store the collected data and processing results; The video monitoring module (3) is connected to the storage unit (2) and is used to collect image information inside and outside the hydropower plant in real time; An environmental sensing module (4) is connected to the storage unit (2) and is used to monitor temperature, humidity, gas concentration and water level parameters; The personnel positioning module (5) is connected to the storage unit (2) and is used to obtain the location information of the on-site workers; The task management system (6) is connected to the storage unit (2) and is used to record the status information of electronic work orders; An emergency response database (7) is connected to the storage unit (2) and is used to store emergency plans and historical handling records; The data integration unit (8) is connected to the storage unit (2) through an IoT gateway and is used to clean, align and fuse multi-source heterogeneous data; The risk assessment engine (9) is connected to the data integration unit (8) and is used to dynamically score and classify the fused data using a machine learning model; The output module (10) is connected to the risk assessment engine (9) to generate a visual risk map, early warning notices and emergency response suggestions, and supports the automatic generation of reports in industry standard formats.

2. The dynamic safety risk assessment system for hydropower plants based on multi-source data fusion according to claim 1, characterized in that, The environment sensing module (4) includes: A temperature and humidity sensor (11) is connected to the storage unit (2) and is used to monitor the temperature and humidity changes inside the hydropower plant in real time. A gas detector (12), connected to the storage unit (2), is used to measure the concentration of harmful gases; A level gauge (13) is connected to the storage unit (2) and is used to detect the water level in a critical area.

3. The dynamic safety risk assessment system for hydropower plants based on multi-source data fusion according to claim 1, characterized in that, It also includes a vibration detection module, which is connected to the storage unit (2) and is used to measure the vibration frequency and amplitude of the key equipment.

4. The dynamic safety risk assessment system for hydropower plants based on multi-source data fusion according to claim 1, characterized in that, It also includes an image recognition unit, which is connected to the storage unit (2) and is used to perform target detection and behavior analysis on the collected image data.

5. The dynamic safety risk assessment system for hydropower plants based on multi-source data fusion according to claim 1, characterized in that, It also includes an alarm module, which is connected to the central processing unit (1) and is used to trigger an alarm when the risk score exceeds a preset threshold.

6. A method for dynamic risk assessment of hydropower plant safety risks using the multi-source data fusion-based dynamic risk assessment system for any one of claims 1-5, characterized in that, include: The video monitoring module (3) collects image information of the inside and outside of the hydropower plant in real time and stores it in the storage unit (2). The environmental sensing module (4) monitors temperature, humidity, gas concentration and water level parameters and stores them in the storage unit (2). The location information of on-site workers is obtained through the personnel positioning module (5) and stored in the storage unit (2). The status information of the electronic work order is recorded by the task management system (6) and stored in the storage unit (2). The data integration unit (8) cleans, aligns and merges the collected multi-source data to build a unified risk data pool; The risk assessment engine (9) is used to dynamically score and classify the fused data; The output module (10) generates a visual risk map, early warning notices and emergency response suggestions, and supports the automatic generation of reports in industry standard formats.

7. The method for dynamic evaluation of hydropower plant safety risks based on multi-source data fusion according to claim 6, characterized in that, The risk assessment engine (9) performs dynamic scoring and grading of the fused data, including: The risk assessment engine (9) obtains the fused data, which includes image data, environmental parameter data, personnel location data, and task status data. The risk assessment engine (9) calls the trained machine learning model to perform feature extraction and pattern recognition on the fused data; The risk assessment engine (9) calculates a risk score based on the extracted features and classifies the risk level based on the score results.

8. The method for dynamic evaluation of hydropower plant safety risks based on multi-source data fusion according to claim 7, characterized in that, The risk assessment engine (9) calls the trained machine learning model to perform feature extraction and pattern recognition on the fused data, including: The risk assessment engine (9) calls a convolutional neural network to extract features from the image data and identify potential security risks. The risk assessment engine (9) calls the long short-term memory network to model time series data and predict risk development trends; The risk assessment engine (9) calls the ensemble learning model to perform comprehensive analysis of multimodal data.

9. The method for dynamic evaluation of hydropower plant safety risks based on multi-source data fusion according to claim 6, characterized in that, If the risk score exceeds the preset threshold, an alarm will be triggered through the alarm module, and a warning notification will be pushed to relevant personnel through the output module (10); The emergency response database (7) is used to match the best handling plan for the current risk scenario and generate emergency response recommendations.

10. The method for dynamic evaluation of hydropower plant safety risks based on multi-source data fusion according to claim 6, characterized in that, Also includes: The output module (10) records the results of each risk assessment to form a risk evolution map; By analyzing the risk evolution map, the risk change trend can be improved, and the response strategies in the emergency plan database can be optimized.

Images