An intelligent security monitoring and response method for power generation enterprises

Abstract

The present invention relates to the technical field of industrial safety management, and particularly to an intelligent security monitoring and response method for power generation enterprises. The method includes the following steps: acquiring environmental data of a power generation enterprise; detecting enterprise abnormal event data from the environmental data of the power generation enterprise to obtain enterprise environmental abnormal event data; performing multi-source video linkage monitoring on the enterprise environmental abnormal event data to obtain multi-angle video data; analyzing abnormal events based on the enterprise environmental abnormal event data and the multi-angle video data to obtain abnormal event risk assessment data for assisting in intelligent security monitoring and response operations of the power generation enterprise. The present invention improves the detection and analysis capabilities of abnormal events and significantly enhances the safety and operation efficiency of enterprises.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial safety management, and in particular to an intelligent security monitoring and response method for power generation enterprises. Background Art

[0002] With the rapid development of the industrial field, power generation enterprises, as an important part of energy production, have a complex production environment and diverse risk factors. Traditional safety management methods mainly rely on regular inspections, single-point monitoring, and manual analysis, and have the following significant deficiencies. Traditional monitoring systems are only based on sensors or cameras at fixed positions, which are prone to form monitoring blind spots and lack comprehensive coverage of environmental changes. Currently, most systems only rely on a single data source (such as environmental sensor data or video data) for anomaly detection, resulting in low accuracy and reliability of the detection results. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention proposes an intelligent security monitoring and response method for power generation enterprises to solve at least one of the above technical problems.

[0004] The present application provides an intelligent security monitoring and response method for power generation enterprises, including the following steps:

[0005] Step S1: Obtain the environmental data of the power generation enterprise;

[0006] Step S2: Detect enterprise abnormal event data from the environmental data of the power generation enterprise to obtain enterprise environmental abnormal event data;

[0007] Step S3: Conduct multi-source video linkage monitoring on the enterprise environmental abnormal event data to obtain multi-angle video data;

[0008] Step S4: Analyze the abnormal event based on the enterprise environmental abnormal event data and the multi-angle video data to obtain abnormal event risk assessment data for assisting the intelligent security monitoring and response operation of the power generation enterprise.

[0009] In the present invention, environmental parameters (such as temperature and humidity, smoke, vibration, etc.) are collected in real time through a sensor network, covering all key areas of the power generation enterprise. Using data analysis methods, the occurrence location of abnormal events is quickly identified, providing a basis for monitoring and response. Based on the abnormal event data, cameras related to the event are automatically linked to conduct multi-angle real-time monitoring of the target area. The combination of multi-source video data avoids the blind spot problem of a single monitoring point and improves the comprehensiveness of monitoring coverage. Specific response suggestions or action plans (such as triggering an alarm, starting a fire extinguishing system) are generated according to the risk level, significantly shortening the time from discovery to response.

[0010] Preferably, step S1 is specifically:

[0011] Step S11: Deploy temperature and humidity sensors, smoke detectors, and vibration detection devices, initialize the device parameters, and complete the network connection with the central monitoring system to obtain the sensor metadata of the power generation enterprise;

[0012] Step S12: Collect sensing parameters based on the sensor metadata of the power generation enterprise to obtain preliminary sensing parameter data;

[0013] Step S13: Perform quality verification on the preliminary sensing parameter data to obtain verified environmental data;

[0014] Step S14: Conduct preliminary filtering on the verified environmental data to obtain filtered environmental data;

[0015] Step S15: Perform distributed caching and aggregation on the filtered environmental data to obtain local environmental data;

[0016] Step S16: Perform global sensor network mapping based on the local environmental data to obtain preliminary power generation enterprise environmental data;

[0017] Step S17: Perform hierarchical decoding on the preliminary power generation enterprise environmental data to obtain power generation enterprise environmental data.

[0018] In the present invention, multiple types of sensors are reasonably deployed according to the layout of enterprise equipment to ensure the comprehensiveness of environmental monitoring. Through network connection, unified management and real-time monitoring of sensors are achieved, reducing the complexity of equipment management. The deployed sensor network is used to collect environmental parameters in real time, such as temperature, humidity, vibration frequency, smoke concentration, etc. Error data is removed through verification to improve data quality and lay a foundation for accurate analysis. Low-pass filtering and smoothing algorithms are applied to preliminarily filter the verified data to extract effective environmental change characteristics. Distributed caching significantly reduces the data pressure on the central system and improves data transmission and processing efficiency. Local aggregation facilitates zonal environmental analysis and independent processing. The local environmental data is used to generate a global sensor network topology map for data spatial association. Through geometric projection or GIS tools, the monitoring range of sensors is mapped to the physical space of the enterprise. The preliminary power generation enterprise environmental data is decoded at multiple levels according to the environmental characteristics of the enterprise (such as stratification by time, space, and feature dimensions) for multi-dimensional analysis.

[0019] Preferably, step S2 is specifically as follows:

[0020] Step S21: Perform feature matching based on the power generation enterprise environmental data and a preset abnormal event recognition model to obtain preliminary abnormal event data;

[0021] Step S22: Perform complex event processing on the preliminary abnormal event data to obtain complex event data;

[0022] Step S23: Determine abnormal events for complex event data to obtain abnormal event determination data;

[0023] Step S24: Perform multi-stage verification and cross-check on the abnormal event determination events to obtain enterprise environment abnormal event data.

[0024] In the present invention, through feature matching, potential abnormal events are quickly screened from a large amount of environmental data. The preliminary abnormal event data is input into a complex event processing engine (such as Apache Flink) to analyze the time, space, and causal relationships of the events. Composite patterns (such as the superposition of multiple environmental features or multiple events within a time window) are extracted to generate complex event data. Based on abnormal criteria and models, the scope of abnormal events is further narrowed to improve the determination accuracy. Multi-stage verification is performed on the abnormal event determination data, including threshold verification (such as repeated events within a time window), historical data comparison (such as the occurrence frequency of similar events), and spatial consistency check (such as the event coverage area). The cross-check mechanism (Cross-check) is used to combine video data, sensor data, and log data to verify the authenticity of the events.

[0025] Preferably, step S3 is specifically as follows:

[0026] Step S31: Lock the event area according to the enterprise environment abnormal event data to obtain abnormal event area data;

[0027] Step S32: Activate the area monitoring device according to the abnormal event area data to obtain area monitoring data;

[0028] Step S33: Collect the original video according to the area monitoring data to obtain the original video data;

[0029] Step S34: Slice the original video according to the original video data to obtain video slice data, and synchronize the time axis of the video slice data to obtain synchronized video segment data;

[0030] Step S35: Extract video features from the synchronized video segment data to obtain video feature data, and identify abnormal frames from the video feature data to obtain abnormal frame segment data;

[0031] Step S36: Perform frame annotation and classification according to the abnormal frame segment data to obtain annotated video data;

[0032] Step S37: Construct a video network according to the monitoring location data corresponding to the area monitoring data and the annotated video data to obtain multi-angle video data.

[0033] In the present invention, based on enterprise environment abnormal event data, the abnormal event area is locked through mapping by a Geographic Information System (GIS) or a sensor network. The dynamic area modeling technology is used to optimize the accuracy of the locked area. According to the abnormal event area data, the monitoring cameras within the coverage area are activated, and parameters such as the camera angle and focal length are adjusted. The real-time video stream is extracted from the activated area monitoring devices. The efficient video coding technology (such as H.265) is used to optimize the storage and transmission of video data. The original video data is sliced according to time periods. The time synchronization protocol (NTP) is used to correct the time axis of multi-source video data to ensure a unified timeline. Feature extraction is performed on the synchronized video segment data to extract the target area features. Abnormal pictures (such as fires, illegal intrusions, etc.) are automatically detected to generate abnormal picture segment data. The abnormal picture segment data is manually or automatically annotated, including the objects, actions, and event types in the pictures. The content of the pictures is classified to annotate the video event types. The annotated video data is associated with the geographical location data of the area monitoring devices. A multi-angle video network is constructed to display the panoramic view and different-angle pictures of the abnormal event.

[0034] Preferably, in step S31, the event area locking is specifically as follows:

[0035] The video monitoring coverage area is locked according to the enterprise environment abnormal event data to obtain the first abnormal event area data;

[0036] The abnormal mode propagation path area is locked according to the enterprise environment abnormal event data to obtain the second abnormal event area data;

[0037] The environmental feature difference area is locked according to the enterprise environment abnormal event data to obtain the third abnormal event area data;

[0038] The area fusion is performed according to the first abnormal event area data, the second abnormal event area data, and the third abnormal event area data to obtain the abnormal event area data.

[0039] In the present invention, based on enterprise environmental abnormal event data, by using the geographical distribution and coverage range of video surveillance cameras, the video surveillance devices covering the event area are locked. The occurrence area of the abnormal event is quickly locked through the coverage range of the video surveillance cameras, ensuring the relevance between the video data and the event. By using the abnormal pattern propagation path model, the time and space trajectories of the event spreading from the occurrence point to the surrounding are analyzed, and the affected propagation area is locked. Considering the dynamic characteristics of the abnormal event propagation, the locked range is adjusted in real time. The differences between the abnormal event area and the reference environmental characteristics (such as temperature and humidity, smoke concentration, etc.) are compared, and the abnormal areas with significant characteristic changes are locked. It is not limited to the event source, but other affected areas are locked through the changes in environmental characteristics. By integrating the data of the first, second, and third abnormal event areas, area superposition, optimization, and boundary refinement are carried out to generate the final abnormal event area, integrating the advantages of multiple locking methods to ensure the comprehensiveness and accuracy of area locking.

[0040] Preferably, the locking of the video surveillance coverage area is specifically as follows:

[0041] Obtain the basic data of the monitoring devices;

[0042] Construct a spatial distribution model of the monitoring devices according to the basic data of the monitoring devices to obtain the spatial distribution model of the monitoring devices;

[0043] Perform event location mapping according to the enterprise environmental abnormal event data to obtain abnormal event location data;

[0044] Perform camera mapping according to the abnormal event location data and the spatial distribution model of the monitoring devices to obtain camera list data;

[0045] Perform geometric reconstruction of the video scene according to the camera list data to obtain a three-dimensional model of the monitoring area;

[0046] Extract the visible area of the abnormal event according to the three-dimensional model of the monitoring area to obtain key frame data of the visible area;

[0047] Generate the boundary of the monitoring area for the key frame data of the visible area to obtain the first abnormal event area data.

[0048] In the present invention, the basic data of all monitoring devices are collected, including parameters such as geographical location, viewing angle, focal length, resolution, coverage range, etc. A spatial distribution model is constructed using the basic data of the monitoring devices, including the geographical coordinates, field of view angle, coverage range, and blind spot location of the devices. Using GIS tools and triangular geometry calculations, a spatial polygon of the coverage area of the monitoring devices is generated. The location attributes in the enterprise environmental anomaly event data are geographically mapped to the spatial distribution model of the monitoring devices, and the anomaly events are quickly associated with specific locations in the enterprise spatial layout. According to the anomaly event location data and the spatial distribution model of the monitoring devices, the cameras covering that location are identified. A list of cameras associated with the event is generated, including information such as camera ID, location, and coverage area. Through three-dimensional reconstruction technology, a three-dimensional scene analysis is provided for the anomaly events. The visible areas related to the anomaly event location are extracted from the three-dimensional model of the monitoring area (SA3DM), and the key frame data within the visible areas are generated. Boundary analysis is performed on the key frame data of the visible areas (VRK) to generate the physical boundaries of the anomaly event area.

[0049] Preferably, the locking of the abnormal mode propagation path area is specifically as follows:

[0050] Based on the enterprise environmental anomaly event data, a Bayesian network model is constructed to obtain an event dynamic propagation model;

[0051] The event dynamic propagation model is used to generate a spatio-temporal node network to obtain event spatio-temporal node network data;

[0052] Based on the event spatio-temporal node network data, the propagation path probability is calculated to obtain event propagation path probability data;

[0053] The event propagation path probability data is used for regional impact allocation to obtain event regional impact weight data;

[0054] Based on the event regional impact weight data, multi-region propagation fusion is performed to obtain multi-region propagation path data;

[0055] Based on the multi-region propagation path data, propagation path prediction is performed to obtain predicted propagation path data;

[0056] Based on the predicted propagation path data, propagation path area reconstruction is performed to obtain propagation path coverage area data;

[0057] Based on the enterprise environmental anomaly event data and the propagation path coverage area data, region generation is performed to obtain second anomaly event area data.

[0058] In the present invention, based on the enterprise environmental anomaly event data, a Bayesian network is constructed to analyze the causal relationships and propagation laws of the anomaly events. The model nodes represent the influencing factors of the events, the edges represent the causal relationships, and the edge weights represent the event propagation probabilities. Based on the event dynamic propagation model, a node network containing spatio-temporal information is constructed, where the nodes represent spatial locations and the edges represent the propagation paths. In the event spatio-temporal node network data, the Markov chain is used to calculate the probabilities of the event propagation paths. Considering the influencing factors (such as time delay, environmental resistance) on each path to adjust the propagation probabilities, the propagation probabilities of the events on different paths are accurately calculated to help identify the high-risk propagation paths. According to the event propagation path probability data, the impacts of the path propagation are allocated to the covered areas, weights are assigned to each area, and the event influence scope and intensity are intuitively reflected. The event area influence weight data is fused with the propagation characteristics of multiple regions to generate the global multi-region propagation paths, and the multi-region propagation fusion improves the adaptability to complex propagation scenarios. The future propagation paths of the events are accurately predicted to help enterprises take prevention and control measures before the events expand. According to the predicted propagation path data, combined with the regional spatial information, the propagation path coverage range is reconstructed, the propagation path coverage area is accurately reconstructed, and the boundary errors in the analysis are reduced. By integrating the enterprise environmental anomaly event data and the propagation path coverage area data, a second anomaly event area is generated.

[0059] Preferably, the locking of the environmental feature difference region is specifically as follows:

[0060] Perform spatio-temporal mapping of environmental features according to the enterprise environmental anomaly event data to obtain the feature spatio-temporal distribution map data;

[0061] Generate a regional feature benchmark model according to the event environmental feature spatio-temporal distribution map data;

[0062] Perform real-time feature deviation analysis according to the regional feature benchmark model and the feature spatio-temporal distribution map data to obtain the feature deviation matrix data;

[0063] Perform difference region clustering according to the feature deviation matrix data to obtain the clustering region data;

[0064] Perform regional similarity analysis according to the clustering region data to obtain the difference region data;

[0065] Refine the difference boundaries of the difference region data to obtain the optimized region boundary data;

[0066] Generate the third anomaly event area data according to the feature deviation matrix data and the optimized region boundary data.

[0067] In the present invention, based on enterprise environmental anomaly event data, environmental features (such as temperature and humidity, vibration, smoke concentration) are extracted and spatio-temporal distribution modeling is carried out to provide an intuitive distribution of environmental features in time and space, revealing the potential change rules of events. Based on historical environmental data and feature spatio-temporal distribution map data, a feature benchmark model is constructed, and the benchmark model is generated by region to adapt to the environmental characteristics of different locations in the enterprise. The real-time environmental feature data is compared with the regional feature benchmark model to calculate the deviation of each feature point. Based on the feature deviation matrix data, clustering algorithms (such as DBSCAN or K-Means) are used to identify the difference regions. The clustering process highlights the regions with significant feature deviations and reduces the interference of invalid regions. The feature vectors of the clustering region data are extracted, and the similarity between regions (such as cosine similarity or dynamic time warping) is calculated. The overlapping regions are merged through similarity analysis to improve the efficiency of region analysis. The boundary of the difference region data is optimized, abnormal boundary points are removed, and the boundary curve is smoothed to remove redundant boundary points to ensure the accuracy of the region boundary. The deviation data and the boundary optimization results are combined to ensure the accuracy and reliability of the abnormal region.

[0068] Preferably, the region fusion specifically includes:

[0069] Performing multi-region intersection calculation according to the first abnormal event region data, the second abnormal event region data, and the third abnormal event region data to obtain intersection region data;

[0070] Performing region weight assignment according to the intersection region data to obtain weighted region data;

[0071] Performing region feature fusion according to the weighted region data to obtain fused region feature data;

[0072] Performing region boundary optimization according to the fused region feature data to obtain first region data;

[0073] Performing feature extraction according to the first abnormal event region data, the second abnormal event region data, and the third abnormal event region data to respectively obtain first abnormal event region feature data, second abnormal event region feature data, and third abnormal event region feature data;

[0074] Performing Gaussian clustering calculation on the first abnormal event region feature data, the second abnormal event region feature data, and the third abnormal event region feature data to obtain region feature Gaussian clustering data;

[0075] Performing similarity and difference degree calculation according to the region feature Gaussian clustering data to obtain region feature similarity and difference degree data;

[0076] Performing weighted calculation on the intersection region data according to the region feature similarity and difference degree data to obtain second region data;

[0077] Calculate the regional differential distribution based on the first-region data and the second-region data to obtain the regional differential distribution data;

[0078] Perform regional compensation calculation on the first-region data and the second-region data according to the regional differential distribution data to obtain the abnormal event region data.

[0079] In the present invention, perform geometric intersection calculation on the first, second, and third abnormal event region data to identify the common coverage part of the three-region data. According to the source and importance of the intersection region data, assign a weight value to each region (such as according to the reliability or detection accuracy of the abnormal event). Integrate the environmental characteristics of each region in the weighted region data to generate comprehensive feature data. Optimize the abnormal region boundary according to the integrated region feature data, and eliminate redundant or fuzzy boundaries. Perform feature extraction on the first, second, and third abnormal event region data respectively, and extract the spatial, temporal, and environmental characteristics of each region. Use the Gaussian mixture model to cluster the extracted region feature data, and output the clustering centers and region feature distributions with high confidence. Gaussian clustering identifies the commonalities and differences between regions, providing a basis for the analysis of similarities and differences. Calculate the similarity and difference degrees between regions for the Gaussian clustering results to quantify the differences in regional characteristics. Perform weighted calculation on the intersection region data using the data on the degree of similarities and differences of region features. Calculate the regional differential distribution based on the first-region data and the second-region data, generate a regional risk map, provide the difference information between regions, and provide data support for the optimization of resource scheduling. Solve the problems of omission and redundancy through regional compensation, and generate the complete range of the abnormal region.

[0080] Preferably, step S4 is specifically as follows:

[0081] Step S41: Perform multimodal feature fusion on the enterprise environmental abnormal event data and the multi-angle video data to obtain the enterprise abnormal event multimodal data;

[0082] Step S42: Perform Bayesian risk assessment on the enterprise abnormal event multimodal data to obtain the preliminary enterprise abnormal event risk assessment data;

[0083] Step S43: Perform event cascade analysis on the enterprise abnormal event multimodal data according to the preliminary enterprise abnormal event risk assessment data to obtain the event cascade diagram data;

[0084] Step S44: Perform potential risk assessment according to the event cascade diagram data and the preset enterprise environmental abnormal event assessment model to obtain the secondary enterprise abnormal event risk assessment data;

[0085] Step S45: Integrate the preliminary enterprise abnormal event risk assessment data and the secondary enterprise abnormal event risk assessment data to obtain the abnormal event risk assessment data for the intelligent security monitoring and response assistance operation of the power generation enterprise.

[0086] In the present invention, enterprise environment abnormal event data (such as temperature and humidity, vibration, smoke concentration) and multi-angle video data (such as the picture features captured by cameras) are fused. A Bayesian network model is used to conduct a preliminary risk assessment on the multi-modal data of enterprise abnormal events, and the risk probability of abnormal events is quantified. According to the preliminary risk assessment data of enterprise abnormal events, causal relationship and time series analysis are carried out on the abnormal events in the multi-modal data to reveal the dynamic relationship between abnormal events and provide insights into complex event scenarios. Using a preset enterprise environment abnormal event assessment model, secondary risk analysis is carried out on the event cascade graph data to evaluate the risk probability and influence range of potential abnormal events and identify secondary risks that have not yet emerged. The preliminary and secondary assessment results are integrated to form a complete risk assessment system.

[0087] The beneficial effects of the present invention are as follows: Deploying multiple types of sensors (such as temperature and humidity, vibration, smoke, etc.) to cover the core areas of power generation enterprises, realizing the comprehensive collection of environmental data, forming high-quality enterprise environmental data, and ensuring the real-time, accuracy, and stability of the data. It realizes the full-link real-time response of abnormal events from occurrence to detection, shortening the event discovery time. The abnormal event detection process is deeply combined with enterprise environmental data, and abnormal features in complex scenarios (such as the environmental abnormal superposition effect) can be identified. It realizes the dynamic monitoring of abnormal events, captures accurate video data from multiple dimensions and perspectives, and provides visual support for event analysis. It effectively integrates different types of abnormal area data, reduces the redundant processing volume of video data, and improves system efficiency. The multi-source linkage mechanism ensures the complete recording of events in complex scenarios (such as secondary problems caused by equipment failures). The system can dynamically evaluate the current risk and secondary risks of abnormal events and early warn the disaster diffusion range in advance. The evaluation results accurately quantify the severity and impact of events, providing clear action priority suggestions for enterprise decision-makers. The risk assessment integrates static data and dynamic monitoring, significantly improving the comprehensiveness and depth of risk management. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] By reading the detailed description of the non-limiting embodiments with reference to the following drawings, other features, purposes, and advantages of the present application will become more obvious:

[0089] Figure 1 The flowchart of the steps of a method for intelligent security monitoring and response of a power generation enterprise according to an embodiment is shown;

[0090] Figure 2 The flowchart of the steps of a method for obtaining environmental data of a power generation enterprise according to an embodiment is shown;

[0091] Figure 3 The flowchart of the steps of a method for detecting enterprise abnormal events according to an embodiment is shown;

[0092] Figure 4 The flowchart of the steps of a multi-source video linkage monitoring method according to an embodiment is shown;

[0093] Figure 5 The flowchart of the steps of an abnormal event risk assessment method according to an embodiment is shown. Detailed implementation manners

[0094] The technical method of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0095] In addition, the accompanying drawings are only schematic diagrams of the present invention and are not necessarily drawn to scale. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor methods and / or microcontroller methods.

[0096] It should be understood that although terms such as "first" and "second" may be used here to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, the first unit can be called the second unit, and similarly the second unit can be called the first unit. The term "and / or" used here includes any and all combinations of one or more of the listed related items.

[0097] Please refer to Figures 1 to 5 , the present application provides an intelligent security monitoring and response method for power generation enterprises, including the following steps:

[0098] Step S1: Obtain the environmental data of the power generation enterprise;

[0099] Specifically, temperature and humidity sensors, smoke detectors, and vibration detection devices are deployed in each key area (such as boiler rooms, fuel depots, and transmission lines) within the power generation enterprise. These devices are connected to the central monitoring system through pre-configured wireless or wired networks. The data collected by the devices includes temperature, humidity, gas concentration, and vibration frequency, and the data is uploaded to the central monitoring system at set time intervals. The monitoring system performs real-time verification and formatting on the transmitted data, filters out invalid or abnormal data, and ensures the consistency and reliability of the collected data. Subsequently, the environmental data is divided into local data for different regions, and global data representing the overall environmental characteristics of the enterprise is generated in the system.

[0100] Step S2: Detect enterprise abnormal event data for the environmental data of power generation enterprises to obtain enterprise environmental abnormal event data;

[0101] Specifically, the central monitoring system performs real-time analysis on the received environmental data through the built-in abnormal event detection module. The system has preset multiple abnormal detection rules. For example, too high temperature indicates equipment overload, humidity change reflects leakage risk, and excessive gas concentration indicates dangerous gas leakage, etc. By matching the environmental data with the preset rules, preliminary abnormal event data is identified. Subsequently, the system performs complex event processing on the abnormal events. By analyzing the multiple variable relationships in the abnormal area, the severity and spread trend of the event are further confirmed. The system records the confirmed abnormal event information (including event type, occurrence time, location, and influence range) to generate enterprise environmental abnormal event data.

[0102] Step S3: Conduct multi-source video linkage monitoring on the enterprise environmental abnormal event data to obtain multi-angle video data;

[0103] Specifically, when the monitoring system confirms an abnormal event, it immediately activates the monitoring cameras covering the area according to the occurrence location of the abnormal event. The multi-angle video data collected by the cameras in real-time is uploaded to the monitoring center. The system will process the video images, including improving the image clarity, eliminating background noise, and detecting moving targets, etc. The images collected by multiple cameras will be integrated into a complete regional monitoring video through image stitching technology. Subsequently, the system extracts the key images related to the abnormal event in the video images, such as the segments with smoke, fire, or equipment abnormalities, and generates multi-angle video data for further analysis of the abnormal events.

[0104] Step S4: Analyze the abnormal events based on the enterprise environmental abnormal event data and the multi-angle video data to obtain abnormal event risk assessment data for intelligent security monitoring and response assistance operations of power generation enterprises.

[0105] Specifically, the monitoring center combines the enterprise environmental abnormal event data with the multi-angle video data for multi-modal analysis. By matching the dynamic changes of the environmental data and the features in the video images, the system conducts a preliminary assessment of the severity of the abnormal event. Subsequently, using historical data and the preset risk assessment model, a more in-depth analysis of the potential influence range and development trend of the current abnormal event is carried out. For example, the system can predict the path of fire spread or whether the abnormal equipment will trigger other chain reactions. The system generates a complete abnormal event risk assessment report, including risk level, recommended emergency response measures, and subsequent risk points, and provides it to the operators for auxiliary decision-making.

[0106] Preferably, step S1 is specifically:

[0107] Step S11: Deploy temperature and humidity sensors, smoke detectors, and vibration detection devices, initialize the device parameters, and complete the network connection with the central monitoring system to obtain the sensor metadata of the power generation enterprise;

[0108] Specifically, install temperature and humidity sensors, smoke detectors, and vibration detection devices in the key areas of the power generation enterprise (such as boiler rooms, fuel storage areas, transmission equipment areas, etc.). Each sensor adjusts its installation position according to the specific requirements of the area. For example, temperature and humidity sensors are arranged in equipment-intensive areas and storage environments, smoke detectors are arranged in flammable areas, and vibration detection devices are installed near key mechanical equipment. The parameters of the sensor devices (such as sampling frequency, sensitivity range, and network address) are initialized after installation. The devices complete the docking of the data transmission network with the central monitoring system through a wireless network module (such as Wi-Fi or LoRa) or a wired connection method. After initialization, the system records the metadata information of each sensor, including device number, installation location, working range, and network connection status.

[0109] Step S12: Collect sensing parameters based on the sensor metadata of the power generation enterprise to obtain preliminary sensing parameter data;

[0110] Specifically, collect environmental data regularly through the deployed sensors according to the set sampling frequency. The data content collected by each device includes: temperature and humidity sensors record temperature and humidity changes; smoke detectors detect the smoke concentration in the area; vibration detection devices monitor the vibration frequency of equipment operation. The collected preliminary sensing parameter data is uploaded to the central monitoring system through the transmission module of the device and marked and archived according to the timestamp and location.

[0111] Step S13: Perform quality verification on the preliminary sensing parameter data to obtain verified environmental data;

[0112] Specifically, after the central monitoring system receives the preliminary sensing parameter data, it verifies the data quality, including the following steps: Check whether there are missing or abnormal fields in the data packet, such as whether the temperature, humidity, and smoke concentration values are empty. Judge whether each parameter value is within the normal working range of the device. For example, whether the temperature value exceeds the detection limit of the device. Process obviously abnormal data (such as sudden extreme values) by replacing them with the average value of the surrounding time period or the data of adjacent sensors. After completing the quality verification, output the verified environmental data.

[0113] Step S14: Perform preliminary filtering on the verified environmental data to obtain filtered environmental data;

[0114] Specifically, the verification environment data is preliminarily filtered to eliminate redundant information and low-correlation data, such as invalid data or duplicate sampling points when there is no abnormal state. The data after preliminary filtering retains records with strong correlation to the device monitoring range and time series, providing more refined input data for subsequent analysis.

[0115] Step S15: Perform distributed caching and aggregation on the filtered environment data to obtain local environment data;

[0116] Specifically, the preliminarily filtered data is cached and summarized on distributed nodes at the regional level. For example, the sensor data within each monitoring area is transmitted to the local cache server for short-term storage, and then the environmental data features of the area (such as average temperature, maximum vibration value, etc.) are generated through aggregation calculation. This not only reduces the central storage pressure of the system but also provides basic data for regional environmental analysis.

[0117] Step S16: Perform global sensor network mapping based on the local environment data to obtain preliminary power generation enterprise environmental data;

[0118] Specifically, through the local environment data uploaded by the distributed nodes, the central monitoring system performs global mapping on the sensor network of the entire power generation enterprise. In the mapping process, based on the geographical location of the sensors, the monitoring area coverage, and the current data status, preliminary environmental data within the enterprise is generated, including equipment working status, abnormal area positioning, and environmental variable distribution.

[0119] Step S17: Perform hierarchical decoding on the preliminary power generation enterprise environmental data to obtain power generation enterprise environmental data.

[0120] Specifically, the central monitoring system performs hierarchical decoding on the global data, deconstructing the complex environmental data according to the logical levels into: real-time monitoring layer: high-frequency data reflecting the current environmental status; historical trend layer: low-frequency data describing the long-term data change trend; abnormal state layer: data marking and correlating abnormal events. The environmental data after hierarchical decoding is stored as complete power generation enterprise environmental data to support abnormal event detection and response work.

[0121] Preferably, step S2 is specifically as follows:

[0122] Step S21: Perform feature matching based on the power generation enterprise environmental data and a preset abnormal event recognition model to obtain preliminary abnormal event data;

[0123] Specifically, the central monitoring system performs feature matching based on the environmental data of power generation enterprises in combination with a preset abnormal event recognition model. The abnormal event recognition model contains multiple abnormal feature templates. For example, too high temperature indicates equipment overload; a sharp increase in humidity reflects equipment leakage; and an excessive smoke concentration indicates a combustion or leakage accident. The monitoring system compares variables such as temperature, humidity, and smoke concentration in the environmental data with these templates one by one, and filters out the qualified data. For example, when the temperature in a certain area continuously exceeds the safety threshold and the humidity changes significantly, the system marks it as a potential abnormal event and records relevant data (such as time, location, and scope of influence). The filtered data is defined as preliminary abnormal event data.

[0124] Step S22: Perform complex event processing on the preliminary abnormal event data to obtain complex event data;

[0125] Specifically, for the preliminary abnormal event data, the central monitoring system further analyzes the potential complexity of the abnormality. For example: Check whether multiple abnormal features occur simultaneously, such as an excessive temperature accompanied by an increase in smoke concentration; evaluate the time series change pattern of the abnormal event, such as whether consecutive abnormal readings are periodic. The monitoring system determines whether these abnormalities have the possibility of causal relationship or chain reaction by analyzing the synchronous data of multiple sensors. For example, when a high-temperature event and abnormal vibration data in an adjacent area are highly consistent in time, the system combines them into a complex event. The complex event data includes event type, time series characteristics, and spatial distribution characteristics.

[0126] Step S23: Determine abnormal events for the complex event data to obtain abnormal event determination data;

[0127] Specifically, the complex event data is further analyzed to determine whether it constitutes an abnormal event. For example: Combine historical data to evaluate the severity of the current event, such as whether it exceeds the previous change range. Judge whether it poses a direct threat to equipment or the environment according to the scope of event influence. The system generates an abnormal event determination result for each complex event, including event type (such as fire, equipment failure), risk level (high, medium, low), and priority response suggestions. For example, when the system determines that a high-temperature event accompanied by abnormal vibration causes equipment overheating and failure, it marks it as a high-priority abnormal event.

[0128] Step S24: Perform multi-stage verification and cross-check on the abnormal event determination events to obtain enterprise environmental abnormal event data.

[0129] Specifically, to improve the accuracy of abnormal event determination, the system conducts multi-stage verification and cross-check on the determination results: Data resampling verification: Re-collect real-time data from relevant sensors to check whether the abnormal features persist. Historical pattern verification: Compare the current event with historical data patterns to determine whether it conforms to known abnormal patterns. Multi-sensor cross-check: Check whether the data of other sensors in the same area is consistent with the determination result. For example, whether the readings of surrounding sensors support the occurrence of an abnormal event. The abnormal events passing the multi-stage verification are further confirmed to generate enterprise environment abnormal event data, including detailed event information (type, location, time, and risk level).

[0130] Preferably, step S3 is specifically as follows:

[0131] Step S31: Lock the event area according to the enterprise environment abnormal event data to obtain abnormal event area data;

[0132] Specifically, the monitoring system locates the occurrence area of the abnormal event according to the enterprise environment abnormal event data. For example, if high temperature and smoke abnormalities are detected by sensors in a certain area, the system will take this area as the center and determine the event area in combination with the influence range of the event. The event area data includes the area boundary, the location of the center point, and the influence radius, which are used to activate the monitoring equipment.

[0133] Step S32: Activate the area monitoring equipment according to the abnormal event area data to obtain area monitoring data;

[0134] Specifically, the monitoring system searches the device database for a list of monitoring devices covering the event area. For example, the camera devices in the area are activated and enter the real-time video acquisition mode. At the same time, the system checks the status of the devices to ensure their normal operation. The activated monitoring devices start to transmit video data to the central system to form area monitoring data.

[0135] Step S33: Collect the original video according to the area monitoring data to obtain the original video data;

[0136] Specifically, the activated monitoring devices collect the video data of the event area in real time and upload it to the central monitoring system. The original video data includes the video frame sequence and the corresponding meta-information such as time stamps and device numbers. These data are centrally stored for processing.

[0137] Step S34: Slice the original video according to the original video data to obtain video slice data, and synchronize the video slice data on the time axis to obtain synchronized video segment data;

[0138] Specifically, the central monitoring system slices the original video data, dividing the long video stream into multiple short video segments. The slicing is based on time intervals and the moments when events occur, ensuring that each video segment contains key frames. Subsequently, the system synchronizes the time axes of the video segments collected by different cameras to ensure that all video segments are aligned within the same time range.

[0139] Step S35: Extract video features from the synchronized video segment data to obtain video feature data, and identify abnormal frame segments from the video feature data to obtain abnormal frame segment data;

[0140] Specifically, the system extracts features from the synchronized video segments and analyzes the key elements in the frames: identifies moving targets in the frames, such as people, equipment, or smoke; extracts information such as the trajectories and color distributions of the targets; detects special visual features in the frames (such as flames, smoke filling). Through the above feature extraction, the system generates key frame information related to abnormal events.

[0141] Step S36: Perform frame annotation and classification based on the abnormal frame segment data to obtain annotated video data;

[0142] Specifically, the system combines the results of video feature extraction through deep learning algorithms to identify abnormal frames. For example, when features such as smoke diffusion or flame flickering are detected, the system marks these frames as abnormal frames. The abnormal frame segment data includes the time range, abnormal types (such as smoke, flame), and positions. The system annotates and classifies the abnormal frames. For example, the frames are marked as categories such as "smoke abnormality", "abnormality of high-temperature equipment", etc., and the corresponding time and space information are recorded. The annotated data is stored as annotated video data for quick retrieval, facilitating further analysis.

[0143] Step S37: Construct a video network based on the monitoring position data corresponding to the area monitoring data and the annotated video data to obtain multi-angle video data.

[0144] Specifically, the system integrates the annotated video data according to the geographical locations of the monitoring devices to generate a multi-angle video network. For example, by the overlapping relationship between the camera positions and the frame contents, the frames from different cameras are spliced to form a panoramic video of the event area. The multi-angle video data generated by the system includes multi-perspective frames with time synchronization, providing complete visual support for subsequent abnormal event analysis.

[0145] Preferably, the event area locking in step S31 is specifically:

[0146] Lock the video monitoring coverage area according to the enterprise environment abnormal event data to obtain the first abnormal event area data;

[0147] Specifically, the system determines the first abnormal event area based on the enterprise environment abnormal event data in combination with the coverage of video surveillance devices. For example, when an abnormal event occurs in the boiler room area, the system retrieves the surveillance devices covering this area through the camera database and generates a surveillance coverage area according to the field of view angle and surveillance range of the cameras. The coverage area is represented as polygon area data, marking the surveillance range containing the abnormal event. These data are defined as the first abnormal event area data.

[0148] Lock the propagation path area of the abnormal mode according to the enterprise environment abnormal event data to obtain the second abnormal event area data;

[0149] Specifically, the system combines the dynamic propagation model of the abnormal event to analyze the event diffusion path. For example, when abnormal temperature and increased smoke are detected in a certain area, the system predicts the direction and range of smoke diffusion through historical data and propagation rules. Based on the sensor network and environmental data, the system generates the path area of event propagation. The area includes the initial event point and the potential diffusion range, marked as the second abnormal event area data.

[0150] Lock the environmental feature difference area according to the enterprise environment abnormal event data to obtain the third abnormal event area data;

[0151] Specifically, the system identifies the abnormal area according to the environmental feature differences inside and outside the abnormal event area. For example, near the area where the abnormal event occurs, the humidity, temperature or vibration characteristics deviate significantly from the normal range. The system marks the area with a large difference from the normal environment by comparing the changes in these characteristic values. The area is defined as the third abnormal event area data.

[0152] Perform area fusion according to the first abnormal event area data, the second abnormal event area data and the third abnormal event area data to obtain the abnormal event area data.

[0153] Specifically, the system fuses the first, second and third abnormal event area data to generate the final abnormal event area. First, the system calculates the intersection part of these areas to ensure that the overlapping area is preferentially marked as a high-risk area. Subsequently, the non-overlapping areas are sorted according to the importance by the weight assignment method. The fused abnormal event area data includes the main range where the event occurs and the affected diffusion area, which is used for monitoring device activation and abnormal event analysis.

[0154] Preferably, the area locking of the video surveillance coverage range is specifically as follows:

[0155] Obtain the basic data of the surveillance device;

[0156] Specifically, the system extracts the basic information of the monitoring devices from the monitoring device database, including device numbers, location coordinates, camera orientations, field of view angles, maximum monitoring distances, and the current working status of the devices. The data records the spatial distribution and coverage of each device, which is used for subsequent area modeling and analysis.

[0157] Construct a spatial distribution model of the monitoring devices based on the basic data of the monitoring devices to obtain the spatial distribution model of the monitoring devices;

[0158] Specifically, the system constructs a spatial distribution model of the monitoring devices according to the obtained basic data of the monitoring devices. The monitoring range of each camera is defined as a frustum model with the camera position as the vertex, extending according to the field of view angle and monitoring distance. Subsequently, the frustums of all devices are integrated into a unified enterprise coordinate system to form a spatial distribution model of the monitoring devices covering the entire power generation enterprise, which is used to quickly query whether a certain location is within the monitoring range.

[0159] Perform event location mapping based on the enterprise environmental anomaly event data to obtain the anomaly event location data;

[0160] Specifically, according to the enterprise environmental anomaly event data, the system locates the specific location where the anomaly event occurs. By combining the geographical coordinates or affected areas of the sensors where the event occurs, event location data is generated. The data describes the central location of the anomaly event and its affected range, marking the areas that need to be monitored key.

[0161] Perform camera mapping based on the anomaly event location data and the spatial distribution model of the monitoring devices to obtain the camera list data;

[0162] Specifically, the system generates the camera list data according to the matching of the event location data and the spatial distribution model of the monitoring devices. The system filters out the camera devices that cover the anomaly event location; through the field of view angle and positional relationship of the cameras, the cameras with a relatively high overlap with the event area are further confirmed and added to the camera list. The camera list data includes the device numbers and their attributes of all devices used to monitor the event area.

[0163] Perform geometric reconstruction of the video scene based on the camera list data to obtain the three-dimensional model of the monitoring area;

[0164] Specifically, according to the camera list data, the system extracts real-time video frames from these devices and performs geometric reconstruction of the video scene using the camera position and orientation information. The process of geometric reconstruction includes mapping the two-dimensional images captured by the cameras into three-dimensional space to restore the spatial structure of the monitoring area. The reconstructed three-dimensional model of the monitoring area describes the terrain, buildings, and equipment layout within the event area, providing an accurate spatial reference for visual area analysis.

[0165] Extract the visible area of abnormal events based on the 3D model of the monitored area to obtain the key frame data of the visible area;

[0166] Specifically, based on the 3D model of the monitored area, the system extracts the visible area of abnormal events. By simulating the line-of-sight range of each camera, the system calculates which monitoring devices can directly observe the area of abnormal events. The data of the visible area includes the relationship between the event occurrence location and surrounding obstacles, and records the best viewing angle information of each camera. Subsequently, the system extracts the key frames of the video from these perspectives and marks the pictures related to abnormal events.

[0167] Generate the boundary of the monitored area for the key frame data of the visible area to obtain the first abnormal event area data.

[0168] Specifically, according to the extracted key frame data of the visible area, the system generates the boundary of the first abnormal event area. By analyzing the visible ranges of different cameras, the system fits the boundary of the abnormal event area to generate an accurate spatial boundary. This boundary represents the monitoring range affected by the abnormal event and marks the abnormal features and important monitoring points in the area in combination with the key frame data.

[0169] Preferably, the locking of the abnormal mode propagation path area is specifically as follows:

[0170] Perform Bayesian network modeling based on the enterprise environment abnormal event data to obtain the event dynamic propagation model;

[0171] Specifically, the system uses the enterprise environment abnormal event data to construct a Bayesian network model describing the event propagation dynamics. The propagation dynamics of the event are described by the relationships between nodes and edges, where nodes represent specific environmental states (such as rising temperature, increasing smoke concentration), and edges represent the causal relationships between these states. The system analyzes the influence probability of each state on subsequent states based on historical abnormal event data. For example, a high-temperature area causes abnormal equipment vibration, and the vibration abnormality leads to a fire. The output of the model is the event dynamic propagation model, indicating how environmental abnormalities spread in time and space.

[0172] Generate a spatio-temporal node network for the event dynamic propagation model to obtain the event spatio-temporal node network data;

[0173] Specifically, the system generates a spatio-temporal node network based on the event dynamic propagation model. Each node represents an abnormal state at a specific time and location (for example, the high-temperature state at a certain time point). The connections between nodes represent the propagation relationships between states, for example, the abnormality of one node affects the surrounding nodes. The system organizes these nodes in chronological order and spatial distance by analyzing the spatial location and time series characteristics of abnormal events to generate the event spatio-temporal node network data.

[0174] Calculate the propagation path probability based on the event spatio-temporal node network data to obtain the event propagation path probability data;

[0175] Specifically, the system analyzes the propagation probability of each connection in the spatio-temporal node network. Use historical data to evaluate the likelihood of an abnormal state spreading to neighboring nodes. For example, the probability of a high-temperature node spreading to a neighboring location depends on the range and time interval of temperature change. By accumulating the propagation probabilities of all paths, event propagation path probability data is generated for identifying the propagation direction and the affected area.

[0176] Perform regional impact allocation on the event propagation path probability data to obtain the event regional impact weight data;

[0177] Specifically, the system performs impact allocation on the enterprise area according to the probability data of the propagation path. The area is divided into multiple units (such as grids), and the system assigns impact weights to each regional unit through location-based linear calculation based on the probability values of the propagation paths. The level of the weight reflects the likelihood of this area being affected by the abnormal event. For example, the regional unit closer to the high-temperature node is assigned a higher weight.

[0178] Perform multi-region propagation fusion based on the event regional impact weight data to obtain multi-region propagation path data;

[0179] Specifically, the system fuses the impact weights of each region to generate multi-region propagation path data. The fusion process synthesizes the impacts of different propagation paths on the region, superimposes the weights of all paths, and forms a unified regional impact distribution. The result is multi-region propagation path data covering the entire enterprise scope, showing the main areas where the abnormal event spreads.

[0180] Perform propagation path prediction based on the multi-region propagation path data to obtain predicted propagation path data;

[0181] Specifically, the system predicts the future propagation path based on the multi-region propagation path data. The system analyzes the direction and speed of the current propagation path, and through the time-shift average algorithm or the long-short time series convolutional network prediction algorithm, speculates the scope of the abnormal event spreading in the future period. The predicted propagation path data includes the regions affected in the future and the expected occurrence time of the abnormal state in these regions.

[0182] Perform propagation path area reconstruction based on the predicted propagation path data to obtain propagation path coverage area data;

[0183] Specifically, the system utilizes the predicted propagation path data to reconstruct the propagation path coverage area. Based on the direction and influence range of the propagation path, the system generates a set of polygon regions representing the affected spatial scope. Each coverage area is labeled with a time node and a predicted abnormal status for monitoring and response purposes.

[0184] Based on the enterprise environmental abnormal event data and the propagation path coverage area data, region generation is performed to obtain the second abnormal event region data.

[0185] Specifically, the system combines the initial abnormal event data and the propagation path coverage area to generate the second abnormal event region data. The system fuses the initial abnormal event region with the propagation path coverage area to generate a comprehensive region, marking the key areas that require the most attention. The generated second abnormal event region data includes the central location, influence range, time distribution, and propagation direction of the abnormal event.

[0186] Preferably, the locking of the environmental feature difference region is specifically as follows:

[0187] Perform environmental feature spatio-temporal mapping based on the enterprise environmental abnormal event data to obtain the feature spatio-temporal distribution map data;

[0188] Specifically, the system performs spatio-temporal mapping of the environmental features according to the enterprise environmental abnormal event data to generate a feature spatio-temporal distribution map. First, extract environmental feature data such as temperature, humidity, smoke concentration, and vibration frequency within the abnormal event region. Organize these feature data into a three-dimensional spatio-temporal distribution map according to the time series and spatial distribution. For example, the feature data at each time point corresponds to a specific grid area in the enterprise map, reflecting the changes in the feature values over time and location. The feature spatio-temporal distribution map data completely describes the dynamic changes of the environmental features within the abnormal event region.

[0189] Generate a regional feature benchmark model based on the event environmental feature spatio-temporal distribution map data;

[0190] Specifically, the system generates a regional feature benchmark model by analyzing the feature spatio-temporal distribution map data. Based on historical data, calculate the normal feature value range for each region. For example, the average value and fluctuation range of the normal temperature, humidity, and smoke concentration in a specific region. The generated feature benchmark model provides reference data for each grid area in the normal state for deviation analysis.

[0191] Perform real-time feature deviation analysis based on the regional feature benchmark model and the feature spatio-temporal distribution map data to obtain the feature deviation matrix data;

[0192] Specifically, the system compares the current feature data with the feature benchmark model to analyze the real-time deviation. For each grid area, the deviation between the current feature value and the benchmark value is calculated, such as whether the temperature is higher than the normal range or whether the humidity shows an abnormal decrease. The deviation analysis result is represented in the form of a feature deviation matrix, where each matrix cell corresponds to an area and its feature deviation data, describing the environmental changes within the abnormal event area.

[0193] Cluster the difference areas based on the feature deviation matrix data to obtain the clustered area data;

[0194] Specifically, based on the feature deviation matrix data, the system performs a clustering analysis on the areas to identify areas with similar feature changes. The system divides the grids with similar feature deviations into a clustered area. For example, the areas with both high temperature and high humidity are clustered into one unit. The data of each clustered area includes the list of grids it belongs to and the overall description of the feature deviation, and the output result is the clustered area data.

[0195] Perform a regional similarity analysis based on the clustered area data to obtain the difference area data;

[0196] Specifically, the system performs a similarity analysis on different clustered areas to identify areas with significant differences. By comparing the feature deviations of each clustered area, it identifies which areas have significant differences in environmental features. For example, the humidity in a certain area is significantly lower than that in other areas. The output difference area data includes the locations of these areas with significant differences and their main feature deviations.

[0197] Refine the difference boundaries of the difference area data to obtain the optimized area boundary data;

[0198] Specifically, based on the difference area data, the system refines its boundaries. First, the preliminary boundaries of the difference area are extracted. For example, a boundary polygon is generated based on the spatial distribution of the feature deviations in the grid. Then, by smoothing the boundaries or filling the gaps in the boundaries, the spatial boundaries of the difference area are optimized to make them more conform to the actual situation. The output optimized area boundary data describes the exact scope of the difference area.

[0199] Generate the third abnormal event area data based on the feature deviation matrix data and the optimized area boundary data.

[0200] Specifically, the system combines the feature deviation matrix data and the optimized area boundary data to generate the third abnormal event area data. The difference areas are further integrated and labeled to form a complete abnormal event area, which describes the spatial distribution and feature changes of the event. The third abnormal event area data includes the overall outline of the difference area, the internal feature distribution, and the deviation from the environmental benchmark.

[0201] Preferably, the area fusion is specifically as follows:

[0202] Perform multi - region intersection calculation based on the first abnormal event area data, the second abnormal event area data, and the third abnormal event area data to obtain intersection area data;

[0203] Specifically, the system calculates the intersection of multiple regions according to the first abnormal event area data, the second abnormal event area data, and the third abnormal event area data. Identify the spatial overlapping parts of these three regions, such as grids or position points included in multiple regions. Define the overlapping part as the intersection area data. The intersection area reflects the common part of all abnormal event areas and is the core impact area of the event.

[0204] Perform area weight assignment based on the intersection area data to obtain weighted area data;

[0205] Specifically, according to the intersection area data, the system assigns weights to area units. The weight reflects the importance of the area. For example, the importance of each unit in the intersection area can be determined by the coverage degree of the three abnormal event areas on it. The system assigns values to each unit. For example, a unit covered by all three regions gets a higher weight, while a unit covered by only one region has a lower weight. After the weighted area data is output, it provides support for subsequent area feature analysis.

[0206] Perform area feature fusion based on the weighted area data to obtain fused area feature data;

[0207] Specifically, the system fuses the feature data within the area according to the weighted area data. The system performs weighted averaging on the feature values (such as temperature, humidity, smoke concentration, etc.) within the intersection area to ensure that the calculation result of each area feature comprehensively considers the contributions of multiple regions to this feature. The fused feature data not only contains the common features of the abnormal areas but also retains the weight differences of different regions in describing the event. The output fused area feature data can be used to accurately describe the environmental features of the comprehensive abnormal area.

[0208] Perform area boundary optimization based on the fused area feature data to obtain first area data;

[0209] Specifically, the system optimizes the boundary of the fused area feature data. First, extract the boundary points with significant abnormalities in the fused features and mark the initial boundary. Then, optimize the boundary by smoothing the boundary line, complementing the missing parts of the boundary, or removing the boundary noise points to make it more accurately reflect the spatial distribution of the event. Generate the first area data, which represents the optimized fused area.

[0210] Feature extraction is performed based on the first abnormal event area data, the second abnormal event area data, and the third abnormal event area data to obtain the first abnormal event area feature data, the second abnormal event area feature data, and the third abnormal event area feature data respectively;

[0211] Specifically, the system extracts feature data from the first, second, and third abnormal event areas respectively. The feature data of each area includes environmental variables such as the average temperature, humidity, vibration amplitude, and smoke concentration of the area. After extraction, the first abnormal event area feature data, the second abnormal event area feature data, and the third abnormal event area feature data are generated respectively for further analysis.

[0212] Perform Gaussian clustering calculation on the first abnormal event area feature data, the second abnormal event area feature data, and the third abnormal event area feature data to obtain area feature Gaussian clustering data;

[0213] Specifically, the system performs Gaussian clustering calculation on the feature data of the three areas to identify the feature distribution law. By analyzing the distribution pattern of the feature data, the system clusters the grids with similar features within the area into one category. The generated area feature Gaussian clustering data describes the similarity or difference between different area features.

[0214] Calculate the similarity and difference degree based on the area feature Gaussian clustering data to obtain area feature similarity and difference degree data;

[0215] Specifically, based on the Gaussian clustering data, the system calculates the similarity and difference degree of the area features. By comparing the central values and distribution ranges of different area clusters, the system identifies the feature similarities and differences. The system outputs the area feature similarity and difference degree data for measuring the differences in features between different abnormal areas.

[0216] Perform weighted calculation on the intersection area data according to the area feature similarity and difference degree data to obtain the second area data;

[0217] Specifically, the system re-weights the intersection area data according to the area feature similarity and difference degree data. The areas with greater similarity and difference degree of features are given higher weights, indicating that these areas play a more important role in the event impact. The system outputs the second area data, which describes the re-weighted area features.

[0218] Perform area difference distribution calculation based on the first area data and the second area data to obtain area difference distribution data;

[0219] Specifically, the system calculates the differential distribution of the first region data and the second region data. By comparing the spatial distribution and feature distribution of the two regions, the system identifies which positions have significant differences in these two regions. The regional difference distribution data includes the difference value and the spatial position of the difference, providing a basis for regional compensation.

[0220] Based on the regional difference distribution data, the system performs regional compensation calculations on the first region data and the second region data to obtain the abnormal event region data.

[0221] Specifically, the system performs compensation calculations on the first region data and the second region data according to the regional difference distribution data to generate the abnormal event region data. For regions with large differences, the system adjusts the feature values or weights to make these regions better reflect the true situation of the event. The generated abnormal event region data includes the core region, boundary information, and comprehensive feature description of the event, which can comprehensively characterize the influence range and nature of the abnormal event.

[0222] Preferably, step S4 is specifically as follows:

[0223] Step S41: Perform multimodal feature fusion on the enterprise environmental abnormal event data and the multi-angle video data to obtain the enterprise abnormal event multimodal data;

[0224] Specifically, the system performs feature fusion on the enterprise environmental abnormal event data and the multi-angle video data to generate the enterprise abnormal event multimodal data. The system extracts the environmental data related to the abnormal event (such as temperature, humidity, smoke concentration) and the video data from multiple monitoring devices (such as key frames, moving target information). According to the time stamp and spatial position, the environmental data and the video data are aligned to ensure that the two types of data can jointly describe the abnormal event within the same time and region. The system combines the environmental features (such as rising temperature) with the video features (such as smoke diffusion, appearance of flames) to form a multimodal feature vector to describe the comprehensive features of the abnormal event. The output multimodal data includes detailed descriptions of time, space, environmental state, and video features.

[0225] Step S42: Perform Bayesian risk assessment on the enterprise abnormal event multimodal data to obtain the preliminary enterprise abnormal event risk assessment data;

[0226] Specifically, the system performs Bayesian risk assessment on the multimodal data of enterprise abnormal events to generate preliminary risk assessment data for enterprise abnormal events. The characteristics of the multimodal data serve as the input for risk assessment, including environmental anomalies (such as high temperature, smoke) and visual anomalies (such as flames, personnel activities). By analyzing historical data and current characteristics, the risks caused by abnormal events are evaluated. For example, smoke diffusion and increased equipment vibration indicate a fire risk. According to the risk assessment results, the abnormal events are classified into three risk levels: low, medium, and high. The output preliminary risk assessment data includes the location, time, feature details, and risk level of the event.

[0227] Step S43: Perform event cascade analysis on the multimodal data of enterprise abnormal events according to the preliminary risk assessment data for enterprise abnormal events to obtain event cascade graph data;

[0228] Specifically, based on the preliminary risk assessment data, the system performs event cascade analysis on the multimodal data to generate event cascade graph data. The system takes each abnormal event as a node, and the node attributes include event time, location, risk level, and feature information. According to the time series and spatial distribution, the association between events is judged. For example, high temperature and smoke anomalies in adjacent areas indicate a chain reaction of events. The connection relationship between events is represented as a graph, where the weight of each edge reflects the association strength between the two events. The output event cascade graph data shows the association structure between abnormal events, providing a basis for potential risk assessment.

[0229] Step S44: Perform potential risk assessment according to the event cascade graph data and the preset enterprise environmental abnormal event assessment model to obtain secondary risk assessment data for enterprise abnormal events;

[0230] Specifically, the system evaluates the potential risks according to the event cascade graph data and the preset enterprise environmental abnormal event assessment model (constructed by deep learning algorithm through historical data, or constructed by regression calculation through historical data) to generate secondary risk assessment data for enterprise abnormal events. The system analyzes the propagation direction and scope of abnormal events according to the event connection relationship in the cascade graph. For example, high temperature spreads to adjacent areas, causing equipment failure. Calculate the combined impact of multiple events, such as smoke diffusion and vibration anomalies jointly causing a fire. Calculate the potential risk level for each event according to the propagation and cumulative effects of the events. The output secondary risk assessment data includes the location, time, and potential risk level of the secondary event.

[0231] Step S45: Integrate the preliminary risk assessment data for enterprise abnormal events and the secondary risk assessment data for enterprise abnormal events to obtain risk assessment data for abnormal events, so as to carry out intelligent security monitoring and response assistance operations for power generation enterprises.

[0232] Specifically, the system integrates the preliminary enterprise abnormal event risk assessment data and the secondary enterprise abnormal event risk assessment data to generate complete abnormal event risk assessment data. The system combines the direct risks of the preliminary risk assessment and the potential risks of the secondary assessment to form a comprehensive risk level. For each event, the system labels its direct risk, potential risk, and comprehensive risk level to form an all-round risk description. The generated risk assessment data includes the comprehensive description of the event (location, time, characteristics, risk level), which is used to guide the intelligent security monitoring and response of power generation enterprises.

[0233] Therefore, from any perspective, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended application documents rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the application documents are intended to be encompassed within the present invention.

[0234] The above are only specific embodiments of the present invention, which enable those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent security monitoring and response method for a power generation enterprise, characterized in that: The following steps are involved: Step S1: Obtain environmental data of power generation enterprises; Step S2: Perform enterprise abnormal event detection on the environmental data of the power generation enterprise to obtain enterprise environmental abnormal event data; Step S3: locking the event area according to the enterprise environment abnormal event data to obtain abnormal event area data, and performing multi-source video linkage monitoring on the abnormal event area data to obtain multi-angle video data; Step S4: Analyze abnormal events based on abnormal event data of the enterprise environment and multi-angle video data to obtain abnormal event risk assessment data to perform intelligent security monitoring and response auxiliary operations for power generation enterprises; The event area locking in step S3 is specifically as follows: The video surveillance coverage area is locked according to the enterprise environment abnormal event data to obtain the first abnormal event area data; According to the enterprise environment abnormal event data, the abnormal mode propagation path area is locked to obtain the second abnormal event area data; According to the enterprise environment abnormal event data, the environment characteristic difference area is locked to obtain the third abnormal event area data; Performing regional fusion according to the first abnormal event region data, the second abnormal event region data, and the third abnormal event region data to obtain abnormal event region data; The specific locking of the abnormal mode propagation path area is as follows: Based on the abnormal event data of the enterprise environment, Bayesian network modeling is carried out to obtain the event dynamic propagation model; Generate a spatiotemporal node network for the event dynamic propagation model to obtain event spatiotemporal node network data; Calculate the probability of the propagation path based on the event space-time node network data to obtain the event propagation path probability data; Perform regional impact allocation on the event propagation path probability data to obtain event regional impact weight data; Multi-region propagation fusion is performed based on the event area impact weight data to obtain multi-region propagation path data; Predicting the propagation path based on the multi-region propagation path data to obtain predicted propagation path data; Reconstruct the propagation path area according to the predicted propagation path data to obtain the propagation path coverage area data; Region generation is performed based on the enterprise environment abnormal event data and the propagation path coverage area data to obtain second abnormal event area data.

2. The method according to claim 1, characterized in that Step S1 is specifically as follows: Step S11: deploy temperature and humidity sensors, smoke detectors, and vibration detection equipment, initialize equipment parameters, complete network connection with the central monitoring system, and obtain sensor metadata of the power generation enterprise; Step S12: collecting sensor parameters according to the sensor metadata of the power generation enterprise to obtain preliminary sensor parameter data; Step S13: Performing quality verification on the preliminary sensing parameter data to obtain verification environment data; Step S14: Preliminarily filter the verification environment data to obtain filtered environment data; Step S15: Distribute and aggregate the filtered environment data to obtain local environment data; Step S16: Perform global sensor network mapping based on local environmental data to obtain preliminary power generation enterprise environmental data; Step S17: Perform hierarchical decoding on the preliminary power generation enterprise environmental data to obtain the power generation enterprise environmental data.

3. The method according to claim 2, characterized in that Step S2 is specifically as follows: Step S21: performing feature matching according to the environmental data of the power generation enterprise and the preset abnormal event recognition model to obtain preliminary abnormal event data; Step S22: performing complex event processing on the preliminary abnormal event data to obtain complex event data; Step S23: performing abnormal event determination on the complex event data to obtain abnormal event determination data; Step S24: Perform multi-stage verification and cross-check on the abnormal event determination event to obtain enterprise environment abnormal event data.

4. The method according to claim 1, characterized in that Step S3 is specifically as follows: Step S31: locking the event area according to the enterprise environment abnormal event data to obtain abnormal event area data; Step S32: activating the regional monitoring device according to the abnormal event regional data to obtain regional monitoring data; Step S33: collecting original video according to the regional monitoring data to obtain original video data; Step S34: Slicing the video according to the original video data to obtain video slice data, and performing time axis synchronization on the video slice data to obtain synchronized video segment data; Step S35: extracting video features from the synchronized video segment data to obtain video feature data, and performing abnormal picture recognition on the video feature data to obtain abnormal picture segment data; Step S36: annotate and classify the images according to the abnormal image segment data to obtain annotated video data; Step S37: construct a video network according to the monitoring location data corresponding to the regional monitoring data and the annotated video data to obtain multi-angle video data.

5. The method according to claim 1, characterized in that The specific areas of video surveillance coverage are as follows: Obtain basic data of monitoring equipment; Constructing a monitoring equipment spatial distribution model based on the basic data of the monitoring equipment to obtain a monitoring equipment spatial distribution model; Mapping the event location according to the abnormal event data of the enterprise environment to obtain the abnormal event location data; Perform camera mapping based on abnormal event location data and monitoring equipment spatial distribution model to obtain camera list data; Reconstruct the video scene geometry based on the camera list data to obtain a three-dimensional model of the monitoring area; Extract the visible area of ​​abnormal events based on the three-dimensional model of the monitoring area to obtain key frame data of the visible area; The monitoring area boundary is generated for the key frame data of the visible area to obtain the first abnormal event area data.

6. The method according to claim 1, characterized in that The specific areas of environmental feature difference locking are: Performing spatiotemporal mapping of environmental characteristics based on enterprise environmental abnormal event data to obtain characteristic spatiotemporal distribution map data; Generate a feature benchmark model based on the event environment feature spatiotemporal distribution map data to obtain a regional feature benchmark model; Perform real-time feature deviation analysis based on the regional feature benchmark model and feature spatiotemporal distribution map data to obtain feature deviation matrix data; Perform difference region clustering according to feature deviation matrix data to obtain cluster region data; Conduct regional similarity analysis based on clustered regional data to obtain differential regional data; Refining the difference boundary of the difference area data to obtain optimized area boundary data; Region generation is performed based on the characteristic deviation matrix data and the optimized region boundary data to obtain the third abnormal event region data.

7. The method according to claim 1, characterized in that The specific regional integration includes: Perform multi-region intersection calculation according to the first abnormal event region data, the second abnormal event region data, and the third abnormal event region data to obtain intersection region data; Assign regional weights according to the intersection area data to obtain weighted regional data; Performing regional feature fusion according to weighted regional data to obtain fused regional feature data; Optimize the region boundary according to the fused region feature data to obtain first region data; Perform feature extraction based on the first abnormal event region data, the second abnormal event region data, and the third abnormal event region data to obtain first abnormal event region feature data, second abnormal event region feature data, and third abnormal event region feature data, respectively; Performing Gaussian clustering calculation on the first abnormal event regional feature data, the second abnormal event regional feature data, and the third abnormal event regional feature data to obtain regional feature Gaussian clustering data; Calculate the degree of similarity and difference based on the regional characteristic Gaussian clustering data to obtain the regional characteristic similarity and difference degree data; Perform weighted calculation on the intersection area data according to the regional feature similarities and differences to obtain the second area data; Performing regional difference distribution calculation based on the first regional data and the second regional data to obtain regional difference distribution data; Regional compensation calculation is performed on the first regional data and the second regional data according to the regional difference distribution data to obtain abnormal event regional data.

8. The method according to claim 1, characterized in that Step S4 is specifically as follows: Step S41: performing multimodal feature fusion on the enterprise environment abnormal event data and the multi-angle video data to obtain multimodal data of enterprise abnormal events; Step S42: performing Bayesian risk assessment on the multimodal data of enterprise abnormal events to obtain preliminary enterprise abnormal event risk assessment data; Step S43: performing event cascade analysis on the multimodal data of enterprise abnormal events according to the preliminary enterprise abnormal event risk assessment data to obtain event cascade graph data; Step S44: Perform potential risk assessment based on the event cascade graph data and the preset enterprise environment abnormal event assessment model to obtain secondary enterprise abnormal event risk assessment data; Step S45: Integrate the preliminary enterprise abnormal event risk assessment data and the secondary enterprise abnormal event risk assessment data to obtain abnormal event risk assessment data for performing intelligent security monitoring and response assistance operations for power generation enterprises.

Images