Intelligent video fault analysis and early warning method and apparatus, and computer device

Abstract

The present application relates to an intelligent video fault analysis and early warning method and apparatus, and a computer device. The method comprises: acquiring an anomaly monitoring video of a power plant; performing fault analysis on the anomaly monitoring video, and identifying video anomaly data having an anomaly part in the anomaly monitoring video; on the basis of real-time environment change information of the power plant, performing scene perception analysis on the video anomaly data to obtain scene perception analysis data; on the basis of the scene perception analysis data and the video anomaly data, performing fault diffusion analysis on an abnormal condition of the power plant to obtain fault diffusion path information; and, on the basis of the scene perception analysis data and the fault diffusion path information, performing early warning analysis on a video monitoring system of the power plant to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system. The present method enables proactive management of video monitoring equipment maintenance, and improves the operation and maintenance efficiency of fault analysis for video monitoring equipment in power plants.

Description

Intelligent video fault analysis and early warning methods, devices and computer equipment Technical Field

[0001] This application relates to the field of power grid monitoring technology, and in particular to an intelligent video fault analysis and early warning method, device and computer equipment. Background Technology

[0002] With the development of digital monitoring technology, video fault analysis technology has emerged for power plant monitoring equipment. In traditional technology, when anomalies are detected in the images captured by the power grid monitoring equipment, technicians inspect the physical condition of the cameras, video transmission lines, and storage devices on-site; they also check the quality of the video footage, such as for signal loss, blurry images, or system crashes. However, traditional fault analysis methods require dispatching technical personnel to investigate after the power grid monitoring equipment malfunctions, which has certain deficiencies in terms of timeliness and comprehensiveness of anomaly analysis, resulting in insufficient operational efficiency for power plant monitoring equipment fault analysis. Summary of the Invention

[0003] Therefore, it is necessary to provide an intelligent video fault analysis and early warning method, device, computer equipment, computer-readable storage medium, and computer program product that can improve the operation and maintenance efficiency of power plant monitoring equipment fault analysis in response to the above-mentioned technical problems.

[0004] Firstly, this application provides an intelligent video fault analysis and early warning method. The method includes:

[0005] Acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant;

[0006] Fault analysis is performed on the anomaly monitoring video to identify abnormal video data with abnormal parts in the anomaly monitoring video;

[0007] Based on the real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information.

[0008] Based on the scene perception analysis data and the video anomaly data, a fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information.

[0009] Based on the scene perception analysis data and the fault propagation path information, the power plant's video monitoring system is subjected to early warning analysis to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0010] Secondly, this application also provides an intelligent video fault analysis and early warning device. The device includes:

[0011] The video data acquisition module is used to acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant.

[0012] The video data analysis module is used to perform fault analysis on the anomaly monitoring video and identify abnormal video data with abnormal parts in the anomaly monitoring video.

[0013] The video data analysis module is also used to perform scene perception analysis on the abnormal video data based on the real-time environmental change information of the power plant, and obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information.

[0014] The video data analysis module is also used to perform fault propagation analysis on the abnormal situation of the power plant based on the scene perception analysis data and the video anomaly data, and obtain fault propagation path information.

[0015] The system fault early warning module is used to perform early warning analysis on the power plant's video monitoring system based on the scene perception analysis data and the fault propagation path information, and to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0016] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the following steps:

[0017] Acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant;

[0018] Fault analysis is performed on the anomaly monitoring video to identify abnormal video data with abnormal parts in the anomaly monitoring video;

[0019] Based on the real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information.

[0020] Based on the scene perception analysis data and the video anomaly data, a fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information.

[0021] Based on the scene perception analysis data and the fault propagation path information, the power plant's video monitoring system is subjected to early warning analysis to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0022] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, performs the following steps:

[0023] Acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant;

[0024] Fault analysis is performed on the anomaly monitoring video to identify abnormal video data with abnormal parts in the anomaly monitoring video;

[0025] Based on the real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information.

[0026] Based on the scene perception analysis data and the video anomaly data, a fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information.

[0027] Based on the scene perception analysis data and the fault propagation path information, the power plant's video monitoring system is subjected to early warning analysis to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0028] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, performs the following steps:

[0029] Acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant;

[0030] Fault analysis is performed on the anomaly monitoring video to identify abnormal video data with abnormal parts in the anomaly monitoring video;

[0031] Based on the real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information.

[0032] Based on the scene perception analysis data and the video anomaly data, a fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information.

[0033] Based on the scene perception analysis data and the fault propagation path information, the power plant's video monitoring system is subjected to early warning analysis to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0034] The aforementioned intelligent video fault analysis and early warning method, device, computer equipment, storage medium, and computer program product acquire abnormal monitoring videos of a power plant; these abnormal monitoring videos are obtained by editing real-time monitoring videos from edge nodes of the power plant; fault analysis is performed on the abnormal monitoring videos to identify abnormal video data with abnormal components; based on real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of abnormal video data and real-time environmental change information; based on the scene perception analysis data and the abnormal video data, fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information; based on the scene perception analysis data and the fault propagation path information, early warning analysis is performed on the power plant's video monitoring system to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0035] By acquiring anomaly monitoring videos from power plants and utilizing real-time editing technology at edge nodes to efficiently process massive amounts of monitoring data, data transmission pressure and latency can be significantly reduced, improving the real-time performance and accuracy of anomaly detection. Simultaneously, combining real-time environmental change information from the power plant with scene-aware analysis allows for in-depth analysis of the potential correlation between anomalies in video monitoring equipment and changes in environmental conditions, providing more comprehensive background support for identifying the causes of failures. Fault propagation analysis further utilizes scene-aware data to predict the scope and propagation path of fault impacts, helping video monitoring equipment anticipate risks before subsequent failures occur and develop targeted maintenance and prevention strategies. Through this end-to-end analysis and prediction, power plants can achieve proactive management of video monitoring equipment maintenance, improve the operational efficiency of fault analysis for video monitoring equipment, significantly enhance the reliability and stability of video monitoring equipment operation, reduce unnecessary downtime, lower maintenance costs, and ensure the continuity and safety of power production. Attached Figure Description

[0036] Figure 1 shows the application environment of an intelligent video fault analysis and early warning method in one embodiment;

[0037] Figure 2 is a flowchart illustrating an intelligent video fault analysis and early warning method in one embodiment;

[0038] Figure 3 is a flowchart illustrating a method for obtaining scene perception analysis data in one embodiment;

[0039] Figure 4 is a flowchart illustrating a data association identification method in one embodiment;

[0040] Figure 5 is a flowchart illustrating a method for obtaining fault propagation path information in one embodiment;

[0041] Figure 6 is a flowchart illustrating a fault condition information setting method in one embodiment;

[0042] Figure 7 is a flowchart illustrating the method for obtaining fault propagation path information in another embodiment;

[0043] Figure 8 is a flowchart illustrating a method for obtaining equipment fault early warning data in one embodiment;

[0044] Figure 9 is a structural block diagram of an intelligent video fault analysis and early warning device in one embodiment;

[0045] Figure 10 is an internal structure diagram of a computer device in one embodiment. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0047] This application provides an intelligent video fault analysis and early warning method, which can be applied to the application environment shown in Figure 1. The terminal 102 communicates with the server 104 via a network. A data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104 or placed on the cloud or other network servers. The server 104 acquires abnormal monitoring videos of the power plant through the terminal 102; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the power plant's edge nodes; fault analysis is performed on the abnormal monitoring videos to identify abnormal video data with abnormal parts; based on the power plant's real-time environmental change information, scene perception analysis is performed on the video abnormal data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of video abnormal data and real-time environmental change information; based on the scene perception analysis data and the video abnormal data, fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information; based on the scene perception analysis data and the fault propagation path information, early warning analysis is performed on the power plant's video monitoring system to obtain equipment fault early warning data and equipment fault repair data of the video monitoring system. The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, and smart in-vehicle systems. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted devices. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0048] In one embodiment, as shown in Figure 2, an intelligent video fault analysis and early warning method is provided. Taking the application of this method to the server in Figure 1 as an example, the method includes the following steps:

[0049] Step 202: Obtain the abnormal monitoring video of the power plant.

[0050] Among them, the abnormal monitoring video can be a segment obtained by filtering or editing the real-time monitoring video through the power plant's edge node equipment. These segments usually contain suspected abnormal situations.

[0051] Among them, real-time monitoring video can be on-site video data recorded 24 hours a day by the power plant monitoring system through cameras and other equipment, reflecting the operating status of equipment and environmental conditions in the power plant in real time.

[0052] Specifically, the power plant's monitoring system collects real-time monitoring video from cameras distributed across various key nodes. To reduce the burden of transmitting large amounts of raw real-time monitoring video to the central server, edge nodes use edge computing technology to preprocess the real-time monitoring video locally, including editing and filtering out potentially abnormal segments. The edited abnormal monitoring video is automatically generated based on preset anomaly judgment criteria (such as blurriness, color cast, noise, video signal loss, abnormal brightness, recording integrity, or pan-tilt-zoom (PTZ) malfunction, equipment online status, network interruption, etc. mapped to the abnormal monitoring video).

[0053] Step 204: Perform fault analysis on the anomaly monitoring video to identify abnormal video data with abnormal parts in the anomaly monitoring video.

[0054] Among them, video anomaly data can be specific data extracted from anomaly monitoring videos, containing specific information about the anomaly event, such as the anomaly type, the time of occurrence, and the anomaly area in the video.

[0055] Specifically, the anomaly monitoring video is input into the fault analysis module for processing. The system uses computer vision and deep learning algorithms (such as convolutional neural networks, CNN) to detect and identify abnormal parts in the anomaly monitoring video. This includes matching specific image features in the anomaly monitoring video with existing video monitoring equipment fault or abnormal situation templates, or identifying new fault features by learning known anomaly patterns in historical data through algorithms. The identified abnormal parts are extracted as video anomaly data, which contains information such as the specific abnormal event type, timestamp, and abnormal region.

[0056] Step 206: Based on the real-time environmental change information of the power plant, perform scene perception analysis on the abnormal video data to obtain scene perception analysis data.

[0057] Real-time environmental change information can be data related to the power plant's external environment collected in real time by sensors, such as temperature, humidity, wind speed, and air pressure. This information is closely related to the operation of the video monitoring equipment and affects its stability and failure rate.

[0058] Scene perception analysis can combine and comprehensively analyze abnormal video data with real-time environmental change information to explore the correlation between abnormal equipment operation and changes in environmental conditions, thereby more accurately determining the cause of equipment failure.

[0059] Among them, scene perception analysis data can be the result obtained through scene perception analysis, reflecting the dynamic combination of abnormal situations in the video and real-time environmental changes in the power plant, helping to determine whether equipment abnormalities are affected by environmental factors.

[0060] Specifically, since the operating status of the power plant's video monitoring equipment is closely related to environmental factors, the system integrates multiple sensors (such as temperature, humidity, wind speed, and pressure) to collect real-time information on environmental changes around the power plant. The scene perception and analysis module combines this environmental data with video anomaly data, using a dynamic weighted model to analyze how environmental conditions affect the equipment's operating status. For example, when an abnormally high equipment temperature is detected in the video, the system compares it with changes in ambient temperature to determine whether the equipment malfunction is caused by external overheating or by an internal fault within the equipment itself, ultimately obtaining scene perception and analysis data.

[0061] Step 208: Based on the scene perception analysis data and video anomaly data, perform fault propagation analysis on the abnormal situation of the power plant to obtain fault propagation path information.

[0062] Among them, fault propagation analysis can predict and analyze the potential impact of equipment failures on other parts of the monitoring system based on abnormal equipment conditions and scene perception analysis data. This analysis helps determine the scope of fault propagation and the equipment that may be affected.

[0063] Among them, the fault propagation path information can be the result of fault propagation analysis, showing the specific path and scope of impact of the fault spreading from the initial video monitoring device to other video monitoring devices, helping operators understand potential system risks.

[0064] Specifically, after identifying the video anomaly data and scene perception data, the fault propagation analysis module simulates and analyzes how anomalies in video monitoring equipment affect the overall operation of the monitoring system. This module establishes a fault propagation model, utilizing historical data, real-time device connectivity, and environmental changes to predict the propagation path and scope of the fault after an anomaly occurs. For example, if a video monitoring device malfunctions, the system can analyze its connectivity with other devices, predict its impact chain, and generate detailed fault propagation path information.

[0065] Step 210: Based on the scene perception analysis data and fault propagation path information, perform early warning analysis on the power plant's video monitoring system to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0066] Among them, early warning analysis can be based on comprehensive equipment status, fault propagation information and scene perception data to predict and assess the possible future failures of video monitoring equipment in the monitoring system.

[0067] Among them, equipment failure early warning data can be the result of early warning analysis, including information such as a list of equipment that is about to fail, the probability of failure, and the possible time range, to help operators intervene in advance and prevent the failure from escalating.

[0068] Among them, equipment fault repair data can be detailed information generated through early warning analysis for the maintenance of video monitoring equipment, including repair suggestions for faulty equipment, parts that may need to be replaced, and estimated repair time.

[0069] Specifically, combining scene perception analysis and fault propagation path information, the early warning analysis module continuously monitors the power plant's video monitoring system to assess potential future fault risks. Through learning from historical fault data and dynamically analyzing real-time monitoring data, the system can predict which equipment has potential fault hazards and generate equipment fault early warning data. This data can include a list of equipment about to fail, the probability of failure, and the timeframe. Simultaneously, the system generates specific equipment fault repair data, providing targeted maintenance recommendations, such as when to recommend shutdown for repair or replacement of equipment components, to ensure timely fault repair, prevent the fault from spreading throughout the entire system, and ultimately improve the stability and reliability of power plant equipment.

[0070] In the aforementioned intelligent video fault analysis and early warning method, abnormal monitoring videos of the power plant are acquired. These abnormal monitoring videos are obtained by editing real-time monitoring videos from the power plant's edge nodes. Fault analysis is performed on the abnormal monitoring videos to identify abnormal video data with abnormal components. Based on real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data. The scene perception analysis data represents the dynamic combination of abnormal video data and real-time environmental change information. Based on the scene perception analysis data and the abnormal video data, fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information. Based on the scene perception analysis data and the fault propagation path information, early warning analysis is performed on the power plant's video monitoring system to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0071] By acquiring anomaly monitoring videos from power plants and utilizing real-time editing technology at edge nodes to efficiently process massive amounts of monitoring data, data transmission pressure and latency can be significantly reduced, improving the real-time performance and accuracy of anomaly detection. Simultaneously, combining real-time environmental change information from the power plant with scene-aware analysis allows for in-depth analysis of the potential correlation between anomalies in video monitoring equipment and changes in environmental conditions, providing more comprehensive background support for identifying the causes of failures. Fault propagation analysis further utilizes scene-aware data to predict the scope and propagation path of fault impacts, helping video monitoring equipment anticipate risks before subsequent failures occur and develop targeted maintenance and prevention strategies. Through this end-to-end analysis and prediction, power plants can achieve proactive management of video monitoring equipment maintenance, improve the operational efficiency of fault analysis for video monitoring equipment, significantly enhance the reliability and stability of video monitoring equipment operation, reduce unnecessary downtime, lower maintenance costs, and ensure the continuity and safety of power production.

[0072] In one embodiment, as shown in Figure 3, based on real-time environmental change information of the power plant, scene perception analysis is performed on abnormal video data to obtain scene perception analysis data, including:

[0073] Step 302: Perform correlation analysis on real-time environmental change information and video anomaly data to identify the data correlation relationship between real-time environmental change information and video anomaly data.

[0074] Association analysis, in particular, is the process of identifying potential associations or correlations between different datasets using statistical and algorithmic techniques. In the application of video surveillance equipment, association analysis compares abnormal data from the video surveillance equipment with real-time environmental change information to determine whether a significant relationship exists between the two, thereby revealing the impact of environmental factors on the equipment's operating status.

[0075] Data correlation can be a statistical or causal relationship between two or more types of data discovered through correlation analysis. In a video monitoring equipment system, data correlation refers to the interaction or dependence between abnormal operation data of the video monitoring equipment and environmental change information, revealing how environmental changes (such as temperature, humidity, wind speed, etc.) affect the operating status of the video monitoring equipment or cause malfunctions.

[0076] Specifically, the system first acquires real-time environmental change information (including temperature, humidity, wind speed, pressure, etc.) from multiple sensors and extracts abnormal data from the video anomaly data of the video monitoring equipment. Subsequently, the system aligns this data using time synchronization technology to ensure a precise temporal correspondence between environmental change information and equipment anomaly data. The system uses association analysis algorithms (such as correlation analysis, regression analysis, or causal reasoning based on Bayesian networks) to calculate the degree of correlation between environmental variables and abnormal events in the video, identifying the data correlation between real-time environmental change information and video anomaly data. For example, when the temperature of a device abnormally rises, the system analyzes whether this change is strongly correlated with an increase in the external ambient temperature, or whether the device's vibration increases synchronously when the wind speed increases dramatically.

[0077] Step 304: Based on the data correlation, perform scene fault perception on real-time environmental change information and video anomaly data to obtain scene perception analysis data.

[0078] Scene fault perception involves analyzing and determining the root cause of a video surveillance equipment malfunction and the context in which it occurred by combining abnormal equipment data and environmental change information. It considers not only the internal state of the video surveillance equipment itself but also the impact of external environmental conditions, helping the system accurately identify the source of the fault—whether it's an anomaly caused by the external environment or a problem with the video surveillance equipment itself.

[0079] Specifically, based on the identified data correlations, the system further performs scene fault perception analysis on real-time environmental change information and video anomaly data. The system analyzes the impact of the environment on video monitoring equipment faults by introducing scene perception models (such as a multimodal model integrating the state of the environment and the video monitoring equipment). The specific implementation process includes: First, the system constructs a model of environmental changes and video monitoring equipment faults based on historical and real-time data to predict the potential impact of environmental changes on the video monitoring equipment; then, it uses this model to determine whether the anomaly of the video monitoring equipment is caused by external factors resulting from environmental changes or by an internal fault of the video monitoring equipment itself, ultimately obtaining scene perception analysis data. For example, when the video monitoring equipment malfunctions in a high-temperature environment, the system can use scene perception analysis to confirm whether this is due to increased load on the video monitoring equipment caused by excessively high ambient temperature or internal overheating caused by a failure in the cooling system of the video monitoring equipment.

[0080] In this embodiment, by performing correlation analysis on real-time environmental change information and video anomaly data, the system can accurately identify the potential correlation between environmental factors (such as temperature and humidity) and abnormal operation of video monitoring equipment. Based on these correlations, scene fault perception analysis allows the system to more comprehensively understand the causes of video monitoring equipment failures and the impact of the external environment. This significantly improves the accuracy and real-time performance of fault warnings, enabling timely detection of potential video monitoring equipment failures caused by changes in the external environment. This facilitates the implementation of effective preventative and maintenance measures in advance, reducing the probability of failures and indirectly improving the operational stability and safety of the power plant.

[0081] In one embodiment, as shown in Figure 4, correlation analysis is performed on real-time environmental change information and video anomaly data to identify the data correlation relationship between them, including:

[0082] Step 402: For any target association dimension, use the association rule algorithm to perform association analysis on real-time environmental change information and video anomaly data to identify the dimensional association relationship between environmental change information and video anomaly data.

[0083] The target correlation dimension can refer to a specific data dimension selected in data analysis to explore the correlation between different variables under that dimension.

[0084] Association rule algorithms are algorithms used to mine potential associations in a dataset, typically to discover co-occurrence patterns between different events or variables. Common algorithms include Apriori and FP-growth, which identify the frequency and strength of the co-occurrence of two types of data (such as environmental changes and equipment malfunctions) under a specific condition by calculating indicators such as support and confidence, in order to reveal potential causal relationships or association patterns.

[0085] In this context, dimensional correlation refers to the correlation or dependency between two or more datasets discovered through correlation analysis under a specific target correlation dimension. It indicates the correlation pattern between certain changes in that dimension (such as an increase in ambient temperature) and another type of data (such as equipment malfunctions or failures).

[0086] Specifically, the system first extracts real-time environmental change information and corresponding time-period video anomaly data from video monitoring equipment for any selected target association dimension, such as temperature, humidity, or wind speed. Association rule algorithms (such as Apriori or FP-growth) are then applied to perform frequent itemset mining and association analysis on this data. The specific implementation process includes calculating the frequency of co-occurrence of certain environmental changes and abnormal events of video monitoring equipment under the selected target dimension, as well as the strength or confidence of this co-occurrence, to identify the dimensional association between environmental change information and video anomaly data. For example, the incidence of video monitoring equipment failure under high temperature conditions, or abnormal current in video monitoring equipment when humidity increases.

[0087] Step 404: Stack the dimension relationships corresponding to each target's associated dimension to obtain the data association relationship.

[0088] Specifically, the system stacks and comprehensively analyzes the dimensional relationships of multiple target-related dimensions (such as temperature, humidity, wind speed, and pressure). The system organizes the relationships under each dimension and comprehensively processes the correlations across different dimensions through weighting and multi-dimensional fusion. Then, based on the degree and frequency of the impact of different environmental factors on equipment anomalies, the system constructs a multi-dimensional comprehensive model. For example, by stacking the dimensional relationships of temperature, humidity, and wind speed, the system can discover that increased ambient temperature and wind speed may simultaneously lead to equipment overheating and abnormal vibration. Finally, the system merges these dimensional relationships into a global data correlation graph, demonstrating the complex interactions between various environmental factors and equipment anomalies, thus revealing the data correlation relationships.

[0089] In this embodiment, by using association rule algorithms to perform correlation analysis on real-time environmental change information and video anomaly data for different target correlation dimensions, the system can identify specific correlations between environmental factors and anomalies in video monitoring equipment under each dimension. Furthermore, by stacking these dimensional correlations, the system derives overall data correlations, comprehensively capturing the complex interactions between anomalies in video monitoring equipment and multiple environmental factors. This multi-dimensional correlation analysis significantly improves the accuracy of fault prediction, helps the system better identify potential fault risks, optimizes early warning strategies, and thus enhances the operational efficiency and reliability of the monitoring system.

[0090] In one embodiment, as shown in Figure 5, based on scene perception analysis data and video anomaly data, fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information, including:

[0091] Step 502: Based on the scene perception analysis data and video anomaly data, set the fault condition information for fault propagation analysis.

[0092] In this context, fault condition information refers to the triggering conditions set in fault analysis to determine whether a device or system is in an abnormal state. These conditions typically include key parameters of the device (such as temperature, power consumption, vibration frequency, etc.) exceeding set safety thresholds, or the possibility that environmental factors (such as temperature, humidity, etc.) may cause equipment malfunctions.

[0093] Specifically, the system combines scene perception analysis data and video anomaly data to extract meaningful features from the operating status of video monitoring equipment, changes in the external environment, and historical fault records. Scene perception analysis data reflects abnormal performance of video monitoring equipment in specific environments, such as reduced efficiency or frequent malfunctions in high-temperature environments. Video anomaly data provides information on abnormal situations of video monitoring equipment at different times, locations, and states. The system analyzes this data to determine which conditions can serve as potential fault triggers for fault propagation analysis, and accordingly sets fault thresholds (such as video monitoring equipment temperature exceeding the normal range, sudden increase in power consumption, abnormal vibration frequency, etc.). It also defines corresponding fault types and impact levels for different video monitoring equipment and environmental variables, thus obtaining fault condition information.

[0094] Step 504: Based on the fault condition information, set the fault propagation mechanism for fault propagation analysis.

[0095] Specifically, the system, based on the operating topology of the power plant's video monitoring equipment, the physical and logical connections between the equipment, and the fault condition information established in the previous step, combines the analysis of the dependencies between the video monitoring equipment, energy transmission paths, and the speed and scope of fault propagation to establish a mathematical model for simulating and calculating fault propagation. Furthermore, the system considers the impact of environmental changes in scene perception data on fault propagation; for example, extreme weather conditions may accelerate fault propagation or expand its impact range, thus constructing a fault propagation mechanism. This mechanism defines how a fault in one video monitoring device propagates to other devices through the power plant's network structure. For instance, if a generator shuts down due to overheating, it may affect the load distribution of the transmission video monitoring equipment connected to it, leading to overload faults in downstream video monitoring equipment.

[0096] Step 506: Based on fault condition information and fault propagation mechanism, calculate the propagation path of the power plant's abnormal situation according to scene perception analysis data and video anomaly data, and obtain fault propagation path information.

[0097] Specifically, the system combines fault condition information, fault propagation mechanisms, scene perception analysis data, and video anomaly data to identify the video monitoring equipment most likely to malfunction, designating it as the fault source. Then, based on the established propagation mechanism, the system sequentially simulates the fault's spread path, such as through physical connections or logical dependencies in the power network, determining which video monitoring equipment the fault might extend to and the extent of its impact. The system calculates the time sequence of fault propagation, its propagation speed, and the specific impact on each affected video monitoring equipment. Finally, the system generates detailed fault propagation path information, demonstrating the path of the fault from its source to other video monitoring equipment, its impact range, and its severity.

[0098] In this embodiment, by combining scene perception analysis data and video anomaly data to set fault propagation analysis fault condition information, the system can accurately determine when video monitoring equipment is in a potential fault state. Based on these fault conditions, the system sets a fault propagation mechanism, clarifying the way and path of fault propagation from one video monitoring device to other video monitoring devices. Subsequently, the system calculates the propagation path of the anomaly in the monitoring system according to these conditions and mechanisms, deriving fault propagation path information. This method can not only identify and predict the propagation path and impact range of faults in advance, but also help operators take more effective prevention and emergency measures, reduce the impact of faults on the operation of the monitoring system, and improve the system's safety and stability.

[0099] In one embodiment, as shown in Figure 6, fault condition information for fault propagation analysis is set based on scene perception analysis data and video anomaly data, including:

[0100] Step 602: Apply the scene perception analysis data and video anomaly data to the power plant's monitoring equipment network topology to obtain the equipment anomaly network topology.

[0101] The network topology of the monitoring equipment can be a mapping of the physical and logical connections between various video monitoring devices within the monitoring system. It shows the location of each video monitoring device in the system, as well as the data flow, transmission, or dependencies between them.

[0102] The device anomaly network topology can be generated based on the monitoring device network topology and combined with the device's anomaly data. It not only displays the connectivity of the video monitoring devices but also identifies the video monitoring devices currently experiencing anomalies and their anomaly types.

[0103] Specifically, the system combines scene perception analysis data and video anomaly data with the power plant's monitoring equipment network topology. This network topology illustrates the physical connections and logical dependencies of the video monitoring equipment within the power plant system. By mapping the abnormal states of the video monitoring equipment to the equipment topology, the system generates an anomaly network topology. This structure not only describes the operational anomalies of each individual video monitoring device but also demonstrates how these devices influence each other through their topologies.

[0104] Step 604: Set fault condition information from the abnormal network topology of the device.

[0105] Specifically, the system first identifies key video monitoring devices or nodes in the topology. These devices are crucial to the overall operation of the monitoring system, and malfunctions in these nodes could trigger widespread system failures. Based on this identification, the system sets specific fault conditions for these key nodes, such as equipment temperature exceeding a certain threshold, voltage instability, or interruption of connection signals with other devices. Furthermore, the system analyzes the interdependencies between video monitoring devices to set trigger conditions for chain reactions. For example, when a generator fails, it determines whether it will affect connected transmission equipment and further impact the entire power grid, ultimately obtaining the set fault condition information.

[0106] In this embodiment, by applying scene-aware analysis data and video anomaly data to the network topology of the monitoring system's video surveillance devices, the system can generate an abnormal network topology for the video surveillance devices, comprehensively displaying the physical and logical connections between the devices and their abnormal states. Based on this structure, the system can accurately set fault condition information, ensuring more targeted and comprehensive fault detection. This allows the system to more accurately identify potential fault risks, improve the accuracy of fault early warning and propagation analysis, thereby optimizing the fault management process, reducing sudden failures during system operation, and ensuring system stability and security.

[0107] In one embodiment, as shown in Figure 7, based on fault condition information and fault propagation mechanism, and according to scene perception analysis data and video anomaly data, the propagation path of the power plant's abnormal situation is calculated to obtain fault propagation path information, including:

[0108] Step 702: Based on fault condition information and fault propagation mechanism, determine the fault propagation scenario according to scene perception analysis data and video anomaly data.

[0109] Among them, the fault propagation scenario can be a fault propagation situation or pattern inferred under specific environmental conditions (such as temperature, humidity, wind speed, etc.) and abnormal equipment conditions.

[0110] Specifically, based on fault condition information and fault propagation mechanisms, the system analyzes the impact of environmental conditions (such as temperature, humidity, and wind speed) in scene perception data on the operation of video monitoring equipment. This is combined with anomalies recorded in video anomaly data. The determination of the fault propagation scenario depends on the combined effect of the cause of the video monitoring equipment failure and environmental factors, as well as the dependencies between video monitoring equipment. Through this data, the system infers how a fault propagates from one video monitoring device to other related devices in a specific scenario. For example, if a critical video monitoring device fails in a high-temperature environment, the system will determine whether other video monitoring devices with strong dependencies on that device are also at risk and include them in the fault propagation scenario.

[0111] Step 704: Using the fault propagation scenario as a constraint, calculate the fault propagation path and the fault range information corresponding to the fault propagation path based on the scenario perception analysis data and video anomaly data.

[0112] Among them, the fault propagation path can refer to the specific path by which a fault spreads from one device to other devices within the power plant system.

[0113] Among them, the fault range information can be the number, location and range of devices that the fault may affect, calculated by the system during the occurrence and propagation of the fault.

[0114] Specifically, the system uses fault propagation scenarios as constraints, inputting the physical connections, logical dependencies, and fault condition information between video monitoring devices in the scenario into the fault propagation model to simulate the process of a fault spreading from the initial video monitoring device to other video monitoring devices. For example, if a video monitoring device stops operating due to overheating, the system will sequentially analyze its impact on upstream and downstream video monitoring devices, predict possible chain reactions, and obtain the fault propagation path. Next, the system performs quantitative analysis on each propagation path, calculating which video monitoring devices the fault spreads to, the duration of the path, the propagation range, and the specific status of the affected video monitoring devices, thus obtaining fault range information.

[0115] Step 706: Obtain fault propagation path information based on the fault propagation path and fault range information of the power plant.

[0116] Specifically, based on the fault propagation path and fault range information generated in the previous step, the system generates the final fault propagation path information. The fault propagation path information includes not only the specific fault propagation path (i.e., the chain reaction from the initial video monitoring device to other affected video monitoring devices), but also the propagation time series (the speed and time of fault propagation), as well as the fault range information (the number, location, and degree of impact of affected video monitoring devices).

[0117] In this embodiment, by combining fault condition information and fault propagation mechanisms with scene perception analysis data and video anomaly data, the system can accurately identify fault propagation scenarios and use these as constraints to calculate the fault propagation path and its corresponding range information. Through this multi-level analysis, the system can generate fault propagation path information, comprehensively demonstrating the possibility and impact range of a fault propagating from the initial video monitoring device to other video monitoring devices. This process significantly improves the accuracy of fault propagation prediction, helping operators identify fault risks and impact ranges in advance, thereby effectively reducing the risk of systemic downtime and ensuring the continuity and stability of monitoring equipment operation.

[0118] In one embodiment, as shown in Figure 8, based on scene perception analysis data and fault propagation path information, an early warning analysis is performed on the power plant's video monitoring system to obtain equipment fault early warning data for the video monitoring system, including:

[0119] Step 802: Construct a multi-domain fault map based on scene perception analysis data and fault propagation path information.

[0120] Among them, the multi-domain fault map can be a comprehensive diagram that combines various data such as equipment status, environmental factors, and fault propagation paths to show the potential relationships and propagation patterns of equipment faults in different domains (such as equipment domain, environmental domain, and fault propagation domain).

[0121] Specifically, scene perception analysis data provides status information of video monitoring equipment under different environmental conditions, such as the impact of environmental factors like temperature and humidity on equipment anomalies. Fault propagation path information shows the specific path along which a fault spreads from one video monitoring device to others. The system integrates this data to build a multi-domain fault map, including a video monitoring equipment status domain, an environmental domain, and a fault propagation domain. This map interconnects the data in each domain, demonstrating the propagation of faults under different environmental conditions and the mutual influence between video monitoring devices.

[0122] Step 804: Based on the multi-domain fault map, perform early warning analysis on the power plant's video monitoring system to obtain the initial fault early warning data of the video monitoring system.

[0123] The initial fault warning data can be the initial warning results generated by analyzing the multi-domain fault map, reflecting the risk of equipment failure in the future.

[0124] Specifically, the system analyzes the status of video monitoring equipment, environmental impacts, and fault propagation paths in the data map to predict the likelihood of future faults and their propagation range. For example, if the temperature of a video monitoring device rises and environmental conditions become severe, the system can predict the risk of future faults in that device and the possible propagation paths of the faults, based on the dependencies between the video monitoring devices. Based on these analysis results, the system generates initial fault warning data. This initial warning data includes a list of video monitoring devices that may fail in the future, the warning level, the time window for the potential fault to occur, and the range of affected video monitoring devices.

[0125] Step 806: Detect the operating status information of the video monitoring system under fault conditions and obtain the early warning data correction amount.

[0126] Among them, the operating status information can be specific data about the health status and performance of the equipment or system during real-time operation. This information includes parameters such as temperature, vibration, current, and voltage.

[0127] Among them, the correction amount of the early warning data can be the difference value obtained by comparing the initial fault early warning data with the real-time operating status information of the equipment.

[0128] Specifically, the system analyzes the status data of the video monitoring equipment during actual operation, such as temperature, vibration, and voltage changes, to confirm whether the real-time behavior of the video monitoring equipment matches the early warning data. If a discrepancy is found between the actual status of the video monitoring equipment and the initial early warning analysis, the system will calculate the correction amount for the early warning data based on this new information.

[0129] Step 808: Optimize the initial fault warning data based on the warning data correction amount to obtain equipment fault warning data.

[0130] Specifically, the system applies the correction amount of the early warning data to the initial early warning model to adjust the risk assessment of video monitoring equipment failure, the prediction of failure occurrence time, and the prediction of failure propagation range, thereby obtaining equipment failure early warning data.

[0131] In this embodiment, by constructing a multi-domain fault map based on scene-aware analysis data and fault propagation path information, the system can comprehensively integrate multi-dimensional data on video monitoring equipment, environment, and fault propagation, providing accurate fault early warning analysis for the power plant's video monitoring system. Initial fault early warning data provides early warning of potential risks. Subsequently, by real-time monitoring of the video monitoring system's operating status under fault conditions, the system can calculate the correction amount for the early warning data and optimize the initial warning data. This process ensures the accuracy and real-time nature of fault early warnings, helping operators take timely preventative measures, reducing the occurrence of sudden faults, and improving the safety and operational efficiency of the monitoring system.

[0132] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0133] Based on the same inventive concept, this application also provides an intelligent video fault analysis and early warning device for implementing the intelligent video fault analysis and early warning method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations of one or more intelligent video fault analysis and early warning device embodiments provided below can be found in the limitations of the intelligent video fault analysis and early warning method described above, and will not be repeated here.

[0134] In one embodiment, as shown in FIG9, an intelligent video fault analysis and early warning device is provided, comprising: a video data acquisition module 902, a video data analysis module 904, and a system fault early warning module 906, wherein:

[0135] The video data acquisition module 902 is used to acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant.

[0136] The video data analysis module 904 is used to perform fault analysis on the anomaly monitoring video and identify abnormal video data with abnormal parts in the anomaly monitoring video.

[0137] The video data analysis module 904 is also used to perform scene perception analysis on abnormal video data based on real-time environmental change information of the power plant, and obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of abnormal video data and real-time environmental change information.

[0138] The video data analysis module 904 is also used to perform fault propagation analysis on abnormal situations in power plants based on scene perception analysis data and video anomaly data, and obtain fault propagation path information.

[0139] The system fault early warning module 906 is used to perform early warning analysis on the power plant's video monitoring system based on scene perception analysis data and fault propagation path information, and to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

[0140] In one embodiment, the video data analysis module 904 is further configured to perform correlation analysis on real-time environmental change information and video anomaly data, identify the data correlation relationship between real-time environmental change information and video anomaly data, and perform scene fault perception on real-time environmental change information and video anomaly data based on the data correlation relationship to obtain scene perception analysis data.

[0141] In one embodiment, the video data analysis module 904 is further configured to perform correlation analysis on real-time environmental change information and video anomaly data for any target correlation dimension using a correlation rule algorithm, identify the dimensional correlation relationship between environmental change information and video anomaly data, and stack the dimensional correlation relationships corresponding to each target correlation dimension to obtain the data correlation relationship.

[0142] In one embodiment, the video data analysis module 904 is further configured to: set fault condition information for fault propagation analysis based on scene perception analysis data and video anomaly data; set fault propagation mechanism for fault propagation analysis based on fault condition information; and calculate the propagation path of the power plant's abnormal situation based on fault condition information and fault propagation mechanism, according to scene perception analysis data and video anomaly data, to obtain fault propagation path information.

[0143] In one embodiment, the video data analysis module 904 is further configured to apply scene perception analysis data and video anomaly data to the monitoring equipment network topology of the power plant to obtain an equipment anomaly network topology; and to set fault condition information from the equipment anomaly network topology.

[0144] In one embodiment, the video data analysis module 904 is further configured to determine the fault propagation scenario based on fault condition information and fault propagation mechanism, according to scene perception analysis data and video anomaly data; calculate the fault propagation path and the fault range information corresponding to the fault propagation path based on the scene perception analysis data and video anomaly data, using the fault propagation scenario as a constraint; and obtain the fault propagation path information based on the power plant's fault propagation path and fault range information.

[0145] In one embodiment, the system fault early warning module 906 is further configured to construct a multi-domain fault map based on scene perception analysis data and fault propagation path information; perform early warning analysis on the power plant's video monitoring system based on the multi-domain fault map to obtain initial fault early warning data for the video monitoring system; detect the operating status information of the video monitoring system under fault conditions to obtain early warning data correction amount; and optimize the initial fault early warning data based on the early warning data correction amount to obtain equipment fault early warning data.

[0146] The modules in the aforementioned intelligent video fault analysis and early warning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0147] In one embodiment, a computer device, which may be a server, is provided, and its internal structure is shown in Figure 10. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores server data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements an intelligent video fault analysis and early warning method.

[0148] Those skilled in the art will understand that the structure shown in Figure 10 is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or may combine certain components, or may have different component arrangements.

[0149] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0150] In one embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0151] In one embodiment, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the steps in the above method embodiments.

[0152] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0153] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0154] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0155] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An intelligent video fault analysis and early warning method, characterized in that, The method includes: Acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant; Fault analysis is performed on the anomaly monitoring video to identify abnormal video data with abnormal parts in the anomaly monitoring video; Based on the real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information. Based on the scene perception analysis data and the video anomaly data, a fault propagation analysis is performed on the abnormal situation of the power plant to obtain fault propagation path information. Based on the scene perception analysis data and the fault propagation path information, the power plant's video monitoring system is subjected to early warning analysis to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

2. The method according to claim 1, characterized in that, Based on the real-time environmental change information of the power plant, scene perception analysis is performed on the abnormal video data to obtain scene perception analysis data, including: A correlation analysis is performed on the real-time environmental change information and the video anomaly data to identify the data correlation relationship between the real-time environmental change information and the video anomaly data; Based on the data correlation, scene fault perception is performed on the real-time environmental change information and the video anomaly data to obtain the scene perception analysis data.

3. The method of claim 2, wherein, The step of performing correlation analysis on the real-time environmental change information and the video anomaly data to identify the data correlation relationship between the real-time environmental change information and the video anomaly data includes: For any target association dimension, an association rule algorithm is used to perform association analysis on the real-time environmental change information and the video anomaly data to identify the dimensional association relationship between the environmental change information and the video anomaly data; The data association relationship is obtained by stacking the dimension association relationships corresponding to each of the target association dimensions.

4. The method of claim 1, wherein, The step of performing fault propagation analysis on the abnormal situation of the power plant based on the scene perception analysis data and the video anomaly data to obtain fault propagation path information includes: Based on the scene perception analysis data and the video anomaly data, the fault condition information for the fault propagation analysis is set; Based on the fault condition information, the fault propagation mechanism of the fault propagation analysis is set; Based on the fault condition information and the fault propagation mechanism, and according to the scene perception analysis data and the video anomaly data, the propagation path of the abnormal situation of the power plant is calculated to obtain the fault propagation path information.

5. The method of claim 4, wherein, The step of setting the fault condition information for the fault propagation analysis based on the scene perception analysis data and the video anomaly data includes: The scene perception analysis data and the video anomaly data are applied to the monitoring equipment network topology of the power plant to obtain the equipment anomaly network topology. The fault condition information is set from the abnormal network topology of the device.

6. The method of claim 4, wherein, Based on the fault condition information and the fault propagation mechanism, and according to the scene perception analysis data and the video anomaly data, the propagation path of the abnormal situation in the power plant is calculated to obtain the fault propagation path information, including: Based on the fault condition information and the fault propagation mechanism, the fault propagation scenario is determined according to the scene perception analysis data and the video anomaly data. Using the fault propagation scenario as a constraint, the fault propagation path is calculated based on the scenario perception analysis data and the video anomaly data, and the fault range information corresponding to the fault propagation path is also calculated. The fault propagation path information is obtained based on the fault propagation path of the power plant and the fault range information.

7. The method of claim 1, wherein, The step of performing early warning analysis on the power plant's video monitoring system based on the scene perception analysis data and the fault propagation path information to obtain equipment fault early warning data for the video monitoring system includes: Based on the scene perception analysis data and the fault propagation path information, a multi-domain fault map is constructed. Based on the multi-domain fault map, an early warning analysis is performed on the power plant's video monitoring system to obtain the initial fault early warning data of the video monitoring system; The operating status information of the video monitoring system under fault conditions is detected to obtain the early warning data correction amount; Based on the correction amount of the warning data, the initial fault warning data is optimized to obtain the equipment fault warning data.

8. An intelligent video fault analysis and early warning device, characterized in that, The device includes: The video data acquisition module is used to acquire abnormal monitoring videos of the power plant; the abnormal monitoring videos are obtained by editing real-time monitoring videos from the edge nodes of the power plant. The video data analysis module is used to perform fault analysis on the anomaly monitoring video and identify abnormal video data with abnormal parts in the anomaly monitoring video. The video data analysis module is also used to perform scene perception analysis on the abnormal video data based on the real-time environmental change information of the power plant, and obtain scene perception analysis data; the scene perception analysis data represents the dynamic combination of the abnormal video data and the real-time environmental change information. The video data analysis module is also used to perform fault propagation analysis on the abnormal situation of the power plant based on the scene perception analysis data and the video anomaly data, and obtain fault propagation path information. The system fault early warning module is used to perform early warning analysis on the power plant's video monitoring system based on the scene perception analysis data and the fault propagation path information, and to obtain equipment fault early warning data and equipment fault maintenance data of the video monitoring system.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images