A community intelligent monitoring and emergency linkage method and system fusing BIM space semantics

Abstract

The application discloses a community intelligent monitoring and emergency linkage method and system fusing BIM space semantics, comprising: constructing a BIM scene atlas, obtaining the attributes and mutual relations of components and space regions in a BIM model; mapping a dynamic target detected in video monitoring to the BIM model to obtain its space semantic information; based on the BIM space semantic information of the dynamic target, constructing a target-environment interaction graph, training and reasoning by using a graph neural network model, and identifying a specific complex event related to a space context; taking the BIM model as a space-time reference, fusing multi-source heterogeneous data, and adopting a graph-based event correlation algorithm to reconstruct a complete event chain containing an atomic event sequence and a correlation relation; when an emergency event or an event chain indicates an emergency state, combining BIM preset information and real-time sensor data, dynamically generating an optimal emergency plan, and visualizing command and dispatching through a BIM three-dimensional scene and augmented reality.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of intelligent security technology, in particular to a community intelligent monitoring and emergency linkage method and system fusing BIM spatial semantics. BACKGROUND

[0002] With the acceleration of urbanization and the promotion of smart community construction, community security monitoring and emergency management are facing increasingly severe challenges. Traditional video monitoring systems mainly rely on two-dimensional image features for event detection, lacking understanding of three-dimensional spatial structure and deep semantic understanding of scene objects. For example, the system has difficulty in distinguishing between entering a normal office or a critical power distribution room, and it is also difficult to accurately determine whether the long-time lingering of personnel at the evacuation passage is a potential risk. Such limitations lead to insufficient recognition ability of the monitoring system for complex behaviors, with high false and missed alarm rates.

[0003] Secondly, community monitoring usually involves multiple cameras and various heterogeneous data sources such as access control, fire control, environmental sensors, etc. How to effectively fuse these multi-modal data to realize cross-camera target continuous tracking, event correlation analysis and comprehensive situation judgment is a technical problem to be solved. Existing systems often process these data in isolation, making it difficult to form a unified and global understanding.

[0004] In addition, when emergency events such as fire and illegal intrusion occur, it is crucial to quickly generate targeted emergency plans and efficiently visualize command and dispatch. Traditional emergency plans are mostly static texts or fixed processes, making it difficult to adapt to dynamic changes in situations. Although BIM (Building Information Modeling) technology provides rich three-dimensional geometric information and semantic information for buildings, how to fully utilize these information to realize intelligent plan generation and dynamic dispatch is still in the exploratory stage. For example, how to dynamically plan the optimal evacuation path according to the real-time location of the fire point and the distribution of personnel, and intuitively convey this information to the on-site personnel and command center, is the key to improving emergency response efficiency.

[0005] Therefore, how to fuse the rich spatial semantics provided by BIM to improve the recognition accuracy of video monitoring for complex behaviors and potential risks, realize effective fusion and event correlation of multi-modal information, and on this basis build intelligent emergency plan generation and visualization dispatch capabilities, is a technical problem to be solved in the field of community intelligent monitoring and emergency linkage. SUMMARY

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a community intelligent monitoring and emergency response method and system that integrates BIM spatial semantics. It aims to use BIM spatial semantic information to guide video event detection and understanding, especially to identify context-related complex events through graph neural network models, and to achieve deep fusion of multi-source heterogeneous data and comprehensive situational assessment through refined event chain reconstruction algorithms. This will support the dynamic generation of emergency plans and AR / VR visual command, thereby improving the intelligence level of community monitoring and the efficiency and accuracy of emergency response.

[0007] To achieve the above objectives, in a first aspect, the present invention provides a community intelligent monitoring and emergency response method integrating BIM spatial semantics, comprising the following steps:

[0008] S1: Construct a community building information model (BIM) scene map, which includes BIM component information, spatial area information and their respective semantic attribute information, as well as the topological relationship information between the BIM components and the spatial areas;

[0009] S2: Acquire video surveillance data and detect dynamic targets from it. Map the dynamic targets to the three-dimensional space corresponding to the BIM scene map through coordinate transformation and geometric projection, and obtain the semantic information of the BIM space where the dynamic targets are located.

[0010] S3: Based on the BIM spatial semantic information of the dynamic target, a local interaction graph containing the dynamic target node and its neighboring BIM element nodes is constructed, and a pre-trained graph neural network (GNN) event recognition model is used to analyze the local interaction graph, or a preset event rule base is combined to analyze the behavior of the dynamic target and identify specific events related to the spatial context. The node features of the GNN model include the kinematic features of the dynamic target and the functional and state features of the BIM spatial elements, and the edge features represent the interaction relationship between nodes.

[0011] S4: Using the BIM as a unified spatiotemporal benchmark, integrate multi-source heterogeneous data acquired by at least two different types of sensors, perform spatiotemporal alignment and correlation analysis, and form an atomic event set;

[0012] S5: Based on the topological relationship information in the BIM scene map, the motion trajectory of the dynamic target in the BIM space, and the set of atomic events formed in S4, an event chain construction algorithm is adopted. By evaluating the temporal sequence relationship, spatial proximity relationship, object consistency, and predefined logical association rules between atomic events, related atomic events are filtered and connected to reconstruct a complete event chain with temporal sequence and logical association.

[0013] S6: When an emergency event is identified in S3 or the event chain reconstructed in S5 indicates an emergency state, the optimal emergency plan for the current emergency event is dynamically calculated and generated based on the emergency resource information, evacuation route information, hazard source information preset in the BIM scene map and the real-time sensor data fused in S4.

[0014] S7: Visualize the optimal emergency plan in the BIM 3D scene, and choose to provide navigation and instructions to on-site personnel through augmented reality (AR) devices, or provide immersive situational awareness and command and dispatch support to command center personnel through virtual reality (VR) devices.

[0015] Secondly, this invention provides a community intelligent monitoring and emergency response system that integrates BIM spatial semantics, including:

[0016] The BIM semantic parsing module is used to construct a community building information model (BIM) scene map. The BIM scene map includes BIM component information, spatial area information and their respective semantic attribute information, as well as the topological relationship information between the BIM components and the spatial areas.

[0017] The video analysis and mapping module is used to acquire video surveillance data, detect dynamic targets, and map the dynamic targets to the three-dimensional space corresponding to the BIM scene map to obtain their BIM spatial semantic information.

[0018] The semantic event detection module is used to construct a local interaction graph containing the dynamic target node and its neighboring BIM element nodes based on the BIM spatial semantic information of the dynamic target, and to analyze the local interaction graph using a pre-trained graph neural network (GNN) event recognition model, or to identify specific events related to the spatial context by combining a preset event rule base.

[0019] The multi-source data fusion module is used to fuse multi-source heterogeneous data acquired by at least two different types of sensors, using the BIM as a unified spatiotemporal reference, to perform spatiotemporal alignment and correlation analysis, and form an atomic event set.

[0020] The event reasoning and early warning module is used to reconstruct a complete event chain with temporal and logical correlation based on the topological relationship information in the BIM scene map, the motion trajectory of dynamic targets in the BIM space and the set of atomic events formed, and to issue an early warning when an emergency event is identified or the event chain indicates an emergency state.

[0021] The emergency response plan generation module is used to dynamically calculate and generate the optimal emergency response plan for the current emergency event based on the emergency resource information, evacuation route information, hazard source information, and fused real-time sensor data preset in the BIM scene map after receiving an early warning.

[0022] The 3D visualization and interaction module is used to visualize the optimal emergency plan in the BIM 3D scene. It can also provide navigation and instructions to on-site personnel through augmented reality (AR) devices, or provide immersive situational awareness and command and dispatch support to command center personnel through virtual reality (VR) devices.

[0023] The technical solution provided by this invention may include the following beneficial effects:

[0024] This invention discloses a community intelligent monitoring and emergency response method and system that integrates BIM spatial semantics, aiming to solve the problems of traditional monitoring lacking spatial semantic understanding, difficulty in multi-source data fusion, and low intelligence level of emergency plans. The method includes: constructing a BIM scene map to obtain the attributes and interrelationships of components and spatial areas in the BIM model; mapping dynamic targets detected in video surveillance to the BIM model to obtain their spatial semantic information; constructing a target-environment interaction graph based on the BIM spatial semantic information of the dynamic targets, and using graph neural network (GNN) models such as graph attention network (GAT) for training and inference to identify specific complex events related to the spatial context; using the BIM model as a spatiotemporal reference, fusing multi-source heterogeneous data, and reconstructing a complete event chain containing atomic event sequences and relationships based on spatiotemporal proximity, object consistency, and predefined logical rules using a graph-based event association algorithm; when an emergency event or event chain indicates an emergency state, dynamically generating an optimal emergency plan by combining BIM preset information and real-time sensor data, and visually directing and dispatching through BIM 3D scenes, augmented reality (AR), or virtual reality (VR). This invention can significantly improve the intelligence level of community monitoring and the efficiency of emergency response.

[0025] This invention possesses at least the following prominent substantive features and significant advancements: First, by introducing a GNN model to analyze the dynamic interaction diagram between the target and the BIM environment, it can more accurately and robustly identify complex context-related events that are difficult to handle with traditional methods, such as subtle behavioral anomalies or multi-target collaborative anomalies. Second, the refined event chain reconstruction algorithm not only connects spatiotemporally related atomic events but also endows the event chain with advanced semantics through logical rules and scene templates, thereby gaining a deeper understanding of the nature and development trend of the situation and providing a more comprehensive basis for early warning and decision-making. Third, using BIM as a unified spatiotemporal benchmark, it effectively integrates heterogeneous data such as video, access control, and sensors, breaking down information silos and providing a more comprehensive and dynamic community security situation map through event chain reconstruction. Fourth, it can dynamically generate optimal emergency plans based on real-time situations (such as fire points and personnel distribution) and BIM model information, such as intelligently planning evacuation routes and allocating emergency resources, significantly improving the pertinence and effectiveness of emergency response. Fifth, by combining BIM 3D scenes with AR / VR technology, it provides on-site personnel and the command center with intuitive and immersive situational information and instruction guidance, improving collaborative efficiency and handling capabilities. Attached Figure Description

[0026] Figure 1 This is a functional module architecture diagram of a community intelligent monitoring and emergency response system that integrates BIM spatial semantics, provided by an embodiment of the present invention.

[0027] Figure 2 This is a flowchart of a community intelligent monitoring and emergency response method that integrates BIM spatial semantics, provided by an embodiment of the present invention.

[0028] Figure 3 This is a schematic diagram of the BIM scene map in an embodiment of the present invention.

[0029] Figure 4 This is a schematic diagram of video target mapping to BIM space and semantic acquisition in an embodiment of the present invention.

[0030] Figure 5 This is a schematic diagram of constructing a local interaction graph based on the GNN-based event recognition model in an embodiment of the present invention.

[0031] Figure 6 This is a schematic diagram of the event chain construction process in an embodiment of the present invention. Detailed Implementation

[0032] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention. It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0033] Example 1

[0034] Reference Figure 1 This embodiment provides a community intelligent monitoring and emergency response system 100 that integrates BIM spatial semantics. This system 100 aims to achieve intelligent monitoring of the community environment and rapid emergency response to sudden events. The system 100 mainly includes:

[0035] 1. BIM Semantic Analysis Module: This module is responsible for the parsing, semantic processing, and map construction of the original BIM model (usually in IFC format or the native format of software such as Revit). It includes:

[0036] BIM Data Interface Submodule 111: Used to read and parse BIM model files, extracting geometric information (location, dimensions, shape), physical properties (material), and functional properties (purpose) of building components (such as walls, doors, windows, columns, beams, and floor slabs), equipment (such as fire hydrants, fire extinguishers, cameras, smoke detectors, and access control card readers), and spaces (such as rooms, corridors, staircases, power distribution rooms, and pump rooms).

[0037] Semantic Enhancement and Ontology Mapping Submodule 112: Addressing potential semantic gaps or non-standardization issues in BIM models, this submodule enriches and standardizes extracted information semantically using a pre-built building domain ontology library (containing space function classifications, equipment type specifications, safety level definitions, etc.). For example, a space simply labeled "Room" can be further precisely labeled as "Fire Control Room" or "High-Voltage Power Distribution Room" based on its internal equipment and connections, and assigned corresponding safety levels and access permissions.

[0038] The Scene Graph Construction and Storage Submodule 113: This module uses semantically encoded BIM components, equipment, and spaces as nodes, and their physical connections (e.g., a door connecting two rooms), spatial topology (e.g., room A adjacent to room B, corridor C containing camera D), and functional dependencies (e.g., a fire hydrant belonging to a fire compartment) as edges, to construct a BIM scene graph. This graph is preferably stored and managed using a graph database (e.g., Neo4j) for efficient querying of complex spatial relationships and semantic information. Figure 3 A simplified BIM scene diagram is shown, where nodes represent BIM elements (such as room 201, door D1, camera C1), and edges represent the relationships between them (such as "located" or "connected").

[0039] 2. Video Analysis and Mapping Module: This module is responsible for processing real-time video streams from various surveillance cameras in the community, detecting dynamic targets, and locating them in the BIM 3D space. It includes:

[0040] Video Stream Access and Decoding Submodule 121: Accesses and decodes the video stream from the network camera via protocols such as RTSP and ONVIF.

[0041] The dynamic target detection and tracking submodule 122 employs deep learning target detection algorithms (such as YOLOv5 and Faster R-CNN) to perform real-time detection and classification of dynamic targets (mainly people and vehicles) in video frames. It also combines target tracking algorithms (such as DeepSORT and FairMOT) to continuously track the trajectory of the same target within a single camera.

[0042] Camera Calibration and Coordinate Transformation Submodule 123: The system pre-calibrates the intrinsic and extrinsic parameters (focal length, principal point, distortion coefficient, and position and orientation in the BIM world coordinate system) of each surveillance camera. This submodule uses these calibration parameters to map the target pixel coordinates detected in the 2D video image (e.g., the center point or foot point of the target bounding box) to rays in the BIM 3D space or locate them to specific 3D coordinates through projection transformation (such as back projection of a pinhole camera model).

[0043] BIM spatial semantic association submodule 124: associates the three-dimensional coordinates of the target in BIM with the scene map constructed by BIM semantic parsing module 110, and queries the BIM space area where the target is currently located (such as "corridor A01", "power distribution room P02"), nearby BIM components (such as "door M03", "fire hydrant F04") and their related semantic information (such as area function "prohibited entry", equipment status "available"). Figure 4 This illustrates the process by which a target (person) is detected from a video image and mapped to a specific room in a BIM model.

[0044] 3. Semantic Event Detection Module: This module identifies specific events based on the spatial semantic information of the target in the BIM. It includes:

[0045] The local interactive graph construction submodule 131a is responsible for constructing the input graph of the GNN model in real time. After the video analysis and mapping module detects a dynamic target and maps it to the BIM space, this submodule, centered on the target and combining it with the BIM scene map 110, extracts BIM elements (such as doors, windows, walls, cameras, fire hydrants, and specific area boundaries) within a certain spatial neighborhood (e.g., within a radius of R meters, or topologically, 1-hop / 2-hop neighbors). These dynamic targets and BIM elements together constitute the nodes of the local interactive graph. Node feature generation includes: calculating a feature vector for each dynamic target node, such as `[type (one-hot), normalized velocity_x, normalized velocity_y, relative BIM area position_x, relative BIM area position_y, orientation angle, cumulative dwell time]`. It also calculates a feature vector for each BIM element node, such as `[type (one-hot), function ID (one-hot), security level, real-time status (one-hot, e.g., door open / closed, equipment normal / damaged)]`.

[0046] Edge definition and feature generation: Define different types of edges, such as "target-target" (features: relative distance, direction), "target-BIM element" (features: distance, whether they are in contact, interaction direction), and "BIM element-BIM element" (inherited from the BIM scene map, feature: connectivity).

[0047] GNN event recognition model submodule 132: Employs a pre-trained graph neural network model, such as a heterogeneous graph neural network containing multiple graph attention layers (GATv2Conv). Each GATv2Conv layer aggregates information from neighboring nodes through an attention mechanism to update the representation of the central node. The model may also include a graph pooling layer (such as TopKPooling or MeanPooling) to obtain a global representation of the entire local interaction graph. Finally, it outputs predefined event category probabilities (e.g., normal passage, illegal intrusion into a designated area, tailgating, abnormal loitering, items left in a danger zone, fighting, falls, etc.) through one or more fully connected layers and a Softmax function. This graph neural network model is trained using a large amount of labeled data. The training data includes video clips of normal behavior and various abnormal events, as well as corresponding BIM model interaction data (target trajectory, BIM element state changes, etc.). The loss function typically uses cross-entropy loss.

[0048] Event Rule Engine Submodule 131: For simple events that are not adequately covered by the GNN model or have low confidence, or as pre- / post-processing for GNN, an IF-THEN-based rule engine can be used.

[0049] 4. Multi-source data fusion module: This module is responsible for integrating data from different sensors and fusing them based on BIM. It includes:

[0050] Multi-source data access submodule 141: Receives data from access control systems (card swipe records, door opening status), fire protection systems (smoke detectors, temperature detectors), environmental sensors (gas leaks, water immersion), security sensors (infrared beams, window and door magnetic sensors), etc.

[0051] Spatiotemporal Alignment and Association Submodule 142: The physical installation locations of all sensors are pre-marked in the BIM model. When sensor data is received, its BIM location and event timestamp are combined with the target information output by the video analysis and mapping module 120 for spatiotemporal alignment. For example, if the access control system reports that a door has been opened by swiping a card, and a camera near that door captures a person passing through, the two can be correlated to confirm the person's identity (if the card swipe information includes their identity). If a smoke detector alarms in a certain area, and a camera in that area captures smoke, the authenticity of the fire alarm can be cross-verified.

[0052] 5. Event Reasoning and Early Warning Module: This module performs in-depth analysis on the fused atomic events (GNN output from the semantic event detection module 130, sensor events from the multi-source data fusion module 140, etc.), reconstructs the event chain, and issues early warnings. This includes:

[0053] Cross-camera target re-identification submodule 151: Utilizes the spatial connectivity in the BIM scene map (such as corridors connecting multiple rooms, stairs connecting different floors) and the appearance features and movement patterns of pedestrians to realize the association of targets between different camera views, i.e., target re-identification (Re-ID). By using BIM path constraints to reduce the Re-ID search space, the complete movement trajectory of the target in the entire community can be constructed.

[0054] Event Chain Construction and Management Submodule 152: This submodule connects events identified by a single camera, events associated with multiple sensors, and target behaviors tracked across cameras, according to chronological order and logical relationships (such as causality and accompaniment), forming a complete event chain. For example: "Personnel A enters through the east gate at 08:00 (access control + video confirmation) -> is captured by camera Cam01 in the east corridor at 08:01 -> enters the power distribution room at 08:03 (video semantic event detection) -> smoke alarm in the power distribution room at 08:05 (sensor data)". Specifically, this includes:

[0055] Atomic event library management: Receives and stores atomic events from various modules. Each atomic event has a standard format: `{event_id, timestamp, bim_location_id, event_type, source_module, involved_objects_ids: [], confidence, description}`.

[0056] Association rule engine: Includes a built-in set of configurable event association rules, including:

[0057] Time-related: `AFTER(e1, e2, min_dt, max_dt)`, `CONCURRENT(e1, e2, window_t)`.

[0058] Spatial association: `ADJACENT_BIM(e1, e2, hops)` (BIM topological adjacency), `WITHIN_DISTANCE_BIM(e1, e2, dist)`.

[0059] Object association: `SAME_TARGET(e1, e2, target_id)`, `INTERACTS_WITH_SAME_BIM_ELEMENT(e1, e2, bim_element_id)`.

[0060] Logical / causal relationship: `IF (type(e1) == T1 AND condition1) THEN LIKELY_LEADS_TO (type(e2) == T2)` (e.g., "illegal entry" may occur after "broken window sound").

[0061] Dynamic event chain graph generation: New atomic events are added as nodes to a dynamic "potential event chain graph" in real time. If a new event is related to an existing event in the graph according to the association rule engine, a directed edge is added between them, and the edge can be weighted (indicating the strength or type of association).

[0062] Event Chain Extraction and Scoring: In this diagram, path search algorithms (such as heuristic search) are run periodically or triggered by high-priority atomic events. The search objective is to find paths that satisfy a specific pattern (such as a "template event chain": a predefined sequence of atomic event types and their expected associations), or paths whose weights (combining factors such as atomic event confidence, association strength, and path length) exceed a threshold. These paths are the identified event chains.

[0063] Event chain semantic annotation and early warning: The extracted event chains are matched with a higher-level "situation template library" (for example, the "theft attempt" template may contain an event chain pattern of "nighttime loitering in a restricted area -> attempting to unlock -> triggering the door magnetic alarm -> escaping"). Successfully matched event chains will be assigned an overall semantic label, and corresponding early warning levels will be triggered based on their severity.

[0064] Situation assessment and early warning submodule 153: Based on the severity and development trend of the input structured event chain and the preset early warning threshold, a comprehensive situation assessment is performed. When the assessment result reaches the early warning level (such as "illegal intrusion and risk of causing fire"), an early warning signal and related event information are sent to the emergency plan generation module 160 and the 3D visualization and interaction module 170.

[0065] 6. Emergency Response Plan Generation Module: Upon receiving an early warning, this module dynamically generates an emergency response plan. This includes:

[0066] Emergency Resource Matching Submodule 161: Based on the event type (fire, intrusion, medical assistance, etc.) and the location of the incident (obtained from the BIM scene map), query and match the most suitable emergency resources in the BIM model, such as the nearest and available fire hydrant, fire extinguisher, first aid kit, security personnel, etc.

[0067] Dynamic Path Planning Submodule 162: For scenarios requiring evacuation (such as fire), based on the pre-set evacuation routes and safety exit information in the BIM, and combined with real-time sensor data (such as fire location, smoke spread area, personnel density, and damaged or blocked routes), a path planning algorithm (such as the improved A* algorithm, where path cost considers distance, congestion, and hazard) is used to dynamically calculate the optimal evacuation route for personnel at different locations. For rescue forces, the optimal route to the incident site is planned.

[0068] Contingency Plan Instruction Generation Submodule 163: Integrates matching resources, planned routes, and necessary action instructions (such as blocking specific passages, starting smoke exhaust fans, and notifying personnel in specific positions) into a structured emergency plan.

[0069] 7. 3D Visualization and Interaction Module: This module is responsible for visualizing and interactively presenting monitoring information, event status, and emergency plans in a 3D environment. It includes:

[0070] BIM 3D Scene Rendering Engine 171: Renders BIM models in real time and overlays dynamic targets (from module 120), event information (from modules 130 and 150), sensor status (from module 140), and emergency plans (from module 160) on the model.

[0071] AR / VR Interface and Rendering Submodule 172: AR Augmented Reality: Projects pre-planned instructions (such as evacuation arrows, danger zone highlights, and target equipment indicators) into the real field of vision of on-site personnel through AR glasses (such as Microsoft HoloLens), providing intuitive navigation. Figure 5 This illustrates how evacuation guidance arrows displayed in AR glasses are superimposed on a real-world hallway.

[0072] VR Virtual Reality: Provide command center personnel with access to VR headsets (such as HTC Vive, Oculus Quest), enabling them to immerse themselves in BIM 3D scenes from a first-person perspective, observe real-time dynamics, simulate the effects of different handling plans, and conduct remote command and dispatch.

[0073] Interactive control submodule 173: Allows operators to interact on the BIM visualization interface, such as selecting and querying component information, manually adjusting the camera PTZ, simulating the triggering of specific equipment (such as remotely starting a fire pump), and manually intervening in contingency plans.

[0074] The modules communicate and collaborate with each other via an internal bus or network. For example, the output of the video analysis and mapping module 120 serves as the input of the semantic event detection module 130; the results of the multi-source data fusion module 140 support the analysis of the event reasoning and early warning module 150; and the results of the emergency plan generation module 160 are finally presented by the 3D visualization and interaction module 170.

[0075] Example 2

[0076] Reference Figure 2 This embodiment provides a detailed process for a community intelligent monitoring and emergency response method that integrates BIM spatial semantics:

[0077] Step S201: Construct a BIM scene map and initialize the system. The BIM scene map includes BIM component information, spatial area information and their respective semantic attribute information, as well as the topological relationship information between the BIM components and the spatial areas. This step corresponds to the operation of the BIM semantic parsing module 110 in the system.

[0078] First, load the community's BIM model (e.g., an IFC file). Then, use a parsing engine to extract the geometric information, ID, name, type, and other basic attributes of all building components (walls, floors, doors, windows, etc.), spatial objects (rooms, corridors, lobbies, stairwells, etc.), and equipment in the MEP (Mechanical, Electrical, Plumbing) system (fire hydrants, smoke detectors, cameras, access control card readers, etc.).

[0079] Secondly, semantic enhancement is performed. By combining a predefined community function ontology library (e.g., defining "power distribution room" as a high-risk restricted area and "evacuation staircase" as a life-saving passage), more precise semantic labels are assigned to BIM objects, such as room function (office, meeting room, power distribution room), safety level, and management responsibility area. For example, a room simply named "RM-101" in BIM can be labeled as a "high-voltage power distribution room" by analyzing its internal equipment (such as high-voltage switchgear) or design document associations.

[0080] Next, a BIM scene graph is constructed. All BIM objects and their semantic attributes are used as nodes in the graph, and spatial relationships (e.g., room A contains equipment B, door C connects room D and corridor E, staircase F connects floors G and H) and logical relationships (e.g., fire hydrant I serves area J) between objects are used as edges in the graph. This graph is stored in a graph database.

[0081] At the same time, other modules in the system are initialized, such as loading the video analysis model, calibrating camera parameters, and establishing connections with various sensors.

[0082] Step S202: Real-time monitoring and multi-source data acquisition.

[0083] The video analysis and mapping module 120 continuously acquires real-time video streams from various surveillance cameras within the community. Simultaneously, the multi-source data fusion module 140 receives data in real-time from access control systems, fire alarm systems, environmental sensors, and other sources.

[0084] Step S203: Dynamic target detection, tracking and BIM space mapping, the dynamic target is mapped to the three-dimensional space corresponding to the BIM scene map through coordinate transformation and geometric projection, and the semantic information of the BIM space where the dynamic target is located is obtained.

[0085] The video analysis and mapping module 120 processes the received video stream:

[0086] 1. Object detection: Using deep learning models such as the YOLO series, detect dynamic targets of interest in each frame, mainly people or vehicles, and obtain their 2D bounding boxes.

[0087] 2. Single-camera tracking: For detected targets, tracking algorithms such as DeepSORT are used to assign a unique ID within the field of view of a single camera and continuously track their movement trajectory.

[0088] 3. BIM Spatial Mapping: Obtain the pre-calibration parameters (intrinsic and extrinsic parameters, with extrinsic parameters referring to the camera's position and orientation in the BIM world coordinate system) for each camera. Transform the pixel coordinates of the target in the 2D image (usually taking the midpoint of the bounding box's bottom edge as the foot point) into the BIM 3D world coordinate system using a back projection algorithm (such as based on homography matrix or direct linear transformation). By querying the BIM scene map, determine which BIM spatial object (e.g., room, corridor) the 3D coordinates fall within, or which BIM component (e.g., door, window) it is close to. Thus, the dynamic target acquires rich BIM spatial semantic context, for example, "Zhang San (target ID: P001) is located in room 305 (function: office) on the third floor, near window W003."

[0089] Step S204: Event recognition based on BIM semantics.

[0090] The semantic event detection module 130 performs event judgment based on the BIM spatial semantic information obtained from the target in step S203:

[0091] 1. Construct a Local Interaction Graph: For each (or selected key) dynamic target, using itself as the central node, query the BIM scene graph for BIM elements (such as the room entity, door, window, corner, fire hydrant, prohibited area boundary, etc.) within a certain radius (e.g., 5 meters) or topologically adjacent (e.g., 1-hop or 2-hop) of its current location as environment nodes, collectively forming a local interaction graph centered on that target. Assign feature vectors to each node (target node, BIM element node), such as target type, speed, orientation, BIM element type, function, status, etc. Assign features to the edges between nodes (target-target, target-BIM element, BIM element-BIM element), such as relative distance, direction, interaction type (e.g., the target is moving towards the door, the target has crossed the area boundary line).

[0092] 2. GNN Model Inference: The constructed local interaction graph (which may be a sequence of graphs within a time window to capture dynamic changes) is input into a pre-trained GNN model (such as the GAT model mentioned above). The GNN model learns the complex spatiotemporal interaction patterns between the target and the environment through graph convolution and attention mechanisms, and outputs the event classification results for the current scene (such as "normal passage", "illegal intrusion into the finance office", "loitering at the evacuation stairwell for more than 2 minutes", "item left in the fire escape") and their confidence scores.

[0093] 3. Rule Engine Supplement: For scenarios that the GNN model fails to cover or have low classification confidence, pre-defined rules based on BIM semantics (such as "personnel enter the 'prohibited' 'power distribution room' without authorized access records") can be used for supplementary judgment.

[0094] The output of this step is a video event with semantic labels and confidence levels, which is then used as an atomic event input to the next step.

[0095] Step S205: Spatiotemporal fusion and correlation analysis of multi-source data.

[0096] The multi-source data fusion module 140 performs the following steps: 1. Data alignment: Unify the timestamps of various sensor data collected in step S202 (such as access control card swipe records: card number, time, door ID; smoke alarm: sensor ID, time, status). Associate the physical ID of the sensor with its pre-marked location in the BIM model.

[0097] 2. Data Association: Using the BIM space as a link, data from different sources are associated. For example, if a card swipe event occurs at a certain access control point (e.g., D001), the system checks whether there is a camera near door D001 in the BIM scene map, retrieves video clips from that camera before and after the card swipe time, and combines this with the target detection results of S203 to perform facial comparison or behavior confirmation to determine whether it was the cardholder who operated the device or whether there were any anomalies such as tailgating. As another example, if a smoke sensor in a certain area (e.g., R101) alarms, the system checks the video from cameras in or near R101 to confirm whether there is visible smoke or flame.

[0098] Step S206: Event chain reconstruction and comprehensive situation assessment. Event reasoning and early warning module 150 executes:

[0099] 1. Atomic Event Collection and Standardization: This involves aggregating atomic events from all information sources, including video semantic events from S204, sensor alarms (such as smoke detectors, door sensors, and infrared beam detectors) fused from S205 (formerly S205, now understood as the result of multi-source data fusion), and access control records (success, failure, and duress codes). The data format is standardized, including timestamps, BIM spatial location IDs, event types, sources, a list of participating object IDs, and confidence levels.

[0100] 2. Construct a candidate event association graph: Maintain a dynamic graph structure where nodes are atomic events. When a new atomic event occurs, it is added to the graph. Then, based on a predefined set of association rules (temporal association, spatial association, object consistency, logical / causal association, with configurable weights in the rule base), evaluate the potential associations between the new event and existing events in the graph. If one or more rules are satisfied, create one or more directed edges between the corresponding atomic event nodes. The edge attributes can include association type and association strength (calculated from rule weights and atomic event confidence).

[0101] 3. Event Chain Extraction: Periodically or when a high-priority atomic event (such as a fire alarm or intrusion alarm) occurs, a path search algorithm is executed in the candidate event association graph. For example, a template-based search can be designed to find subgraphs or paths that match a predefined "scenario template" (such as which atomic event types a typical theft process would contain and the expected associations between them). Alternatively, a score-based search can be used to find meaningful event sequences in the graph whose total weight (representing the credibility or importance of the entire chain) exceeds a certain threshold. The extracted paths constitute a reconstructed event chain. An event chain can be represented as an ordered list of atomic events and the explicit relationships between them.

[0102] 4. Event Chain Semantic Annotation and Situation Assessment: For the extracted event chain, a comprehensive semantic description and risk level are assigned to the entire event chain by matching it with a higher-level "situation ontology library" (for example, a "fire initiation" situation may consist of an event chain of "smoke alarm -> heat alarm -> smoke appearing in the corresponding area video -> personnel evacuation"). If the event chain indicates an urgent or potentially serious risk, an early warning is triggered, and the process proceeds to step S207.

[0103] Step S207: Determine whether it is an emergency.

[0104] If S204 identifies a high-risk event (such as a fire alarm or forced entry) or the situation assessment result of S206 reaches the preset emergency threshold, it is determined to be an emergency event, and step S208 is executed; otherwise, if it is a general event, the log is recorded, relevant management personnel are notified, and then the process returns to step S202 to continue monitoring.

[0105] Step S208: Dynamically generate emergency response plans. The emergency response plan generation module 160 is activated:

[0106] Information Acquisition: Retrieve static information related to the incident location from the BIM scene map (such as evacuation routes, safety exit locations, distribution of fire-fighting facilities, and hazard source information), and obtain real-time dynamic information from the multi-source data fusion module (such as the precise location of the fire point / intruder, the range of smoke spread, personnel distribution density, and the status of available / damaged facilities).

[0107] Resource matching and route planning: Emergency resources are automatically matched based on event type. For example, in the event of a fire, the nearest available fire hydrants and fire extinguishers are located; in the event of injuries, the nearest first-aid kits and AEDs are located. Using algorithms such as A* or Dijkstra, the optimal evacuation route to the nearest safe exit is planned for trapped personnel on the BIM road network model (considering real-time obstacles such as fire zones, dense smoke zones, and closed passages), and the fastest access route to the incident point is planned for rescue personnel.

[0108] 3. Contingency Plan Integration: Integrate evacuation orders, rescue orders, resource allocation plans, and control measures to be taken (such as initiating smoke extraction and cutting off power to specific areas) into a complete dynamic emergency plan tailored to the current situation.

[0109] Step S209: Visualized command and AR / VR coordinated dispatch. 3D visualization and interaction module 170 execution:

[0110] BIM Visualization: On the BIM 3D visualization platform in the command center, the location of the incident, the scope of the incident's impact, dynamically generated evacuation routes, the location of emergency resources, and the movement routes of rescue teams are highlighted.

[0111] AR navigation and instructions transmit information such as evacuation routes, hazard warnings, and operation guidelines to AR glasses worn by on-site personnel via wireless network. The information is then overlaid in their real-time field of vision in the form of arrows, icons, and text, providing accurate navigation.

[0112] VR immersive perception: Commanders can enter a virtual BIM accident scene through VR headsets, observe the overall situation from a first-person or God's-eye view, and conduct simulations and remote command.

[0113] Command issuance and feedback: Commanders issue commands through the platform, and on-site personnel receive and provide feedback on the execution status through mobile terminals or AR devices, forming a closed-loop command system.

[0114] Step S210: The incident handling is completed, and normal monitoring is restored.

[0115] Once the emergency has been resolved, the system will deactivate the warning status, clear the temporary emergency information, restore the relevant modules to normal monitoring, and return to step S202.

[0116] Example 3

[0117] Suppose a fire breaks out in a residential building in a smart community.

[0118] 1. Fire detection and preliminary location (S202-S204):

[0119] The smoke sensor (Sensor_Smoke_502) located in room 502 on the 5th floor was triggered first, sending an alarm signal to the system's multi-source data fusion module 140 via the IoT gateway. Almost simultaneously, the video analysis and mapping module 120 of the smart camera (Cam_Corridor_5F) installed in the 5th-floor corridor detected smoke emerging from the door crack of room 502 through its video analysis algorithm. The video analysis module mapped the location of the smoke (2D image coordinates) onto the BIM model, confirming that the smoke source pointed to room 502. The semantic event detection module 130, combining the function of room 502 ("bedroom") in the BIM scene map, the smoke sensor alarm, and the smoke confirmed by the video, determined that a "fire occurred in room 502" event. Before the fire officially broke out, the GNN model may have detected an event by analyzing the video from the corridor cameras near room 502: "A person (possibly a resident) was pacing anxiously at the door of room 502 and repeatedly trying to open the door (through the interaction analysis between the target and the BIM door model)." This video semantic event will be linked with the later "smoke alarm in room 502" and "internal opening record of the access control system (if it is a smart lock) in room 502 (if the resident escapes)" in the event chain reconstruction step, forming a more complete picture of the initial development of the fire, which helps to determine the cause of the fire or whether anyone is trapped.

[0120] 2. Multi-source confirmation and situation escalation (S205-S206):

[0121] The event reasoning and early warning module 150 received a smoke alarm and video confirmation of a fire. The system queried the BIM model and found that there was no thermal imaging camera in room 502. However, Cam_Corridor_5F, located in the 5th-floor corridor, began detecting residents (targets P_Res01, P_Res02) running out of their respective rooms (e.g., 501, 503) and moving along the corridor towards the stairwell. Through cross-camera tracking, the system found that some residents chose the east stairwell, while others chose the west stairwell. The BIM scene map showed that room 502 was near the east stairwell. An initial event chain was formed: "Sensor_Smoke_502 alarm -> Cam_Corridor_5F detects smoke at the door of 502 -> P_Res01, P_Res02 ​​evacuate from the 5th floor towards the stairwell." Based on the severity of the fire, the system immediately classified it as an emergency.

[0122] 3. Dynamic Emergency Response Plan Generation (S208): The emergency response plan generation module 160 is activated.

[0123] Hazard assessment: Based on the location of room 502 and the ventilation design in the BIM, it is preliminarily determined that the fire and smoke may spread primarily to the east staircase.

[0124] Resource location: According to the BIM scene map, there is a fire hydrant (FH-5E, FH-5W) and a fire extinguisher box at each end of the east and west sides of the 5th floor corridor. The nearest fire control room is on the 1st floor.

[0125] Evacuation Routes: For residents on the 5th floor and above, the system dynamically calculates evacuation routes. Because the fire is near the east staircase, the system temporarily reduces the access weight of the east staircase (Stair_E) at the entrance of that floor or marks it as "high risk." Residents on the 5th floor and above are preferentially recommended to use the west staircase (Stair_W) for evacuation. For residents on the 4th floor and below, if there is no smoke impact, they can use the nearest staircase normally. The system also checks the status of normally closed fire doors marked in the BIM model (assuming there are sensors providing feedback).

[0126] Rescue route: Plan the route from the ground entrance to the 5th floor fire scene for the arriving firefighters, and prioritize fire hydrants that are close to the fire but upwind (such as FH-5W, if the wind direction is known or defaulted).

[0127] Command generation: The generated commands include: notifying the community mini fire station to be dispatched; activating the fire broadcast and audible and visual alarms on the 5th floor and above; advising residents on the 5th floor and above to prioritize evacuation via the west staircase; and notifying the property control center to remotely activate the smoke exhaust fan (if applicable and linked to the system).

[0128] 4. Visualized Command and AR / VR Integration (S209): 3D Visualization and Interaction Module 170 execution:

[0129] Command Center Large Screen: In the BIM 3D model, room 502 is marked as a red fire point, and the smoke spread simulation area (based on preliminary CFD estimates or empirical models) is displayed with a semi-transparent effect. Recommended evacuation routes are marked in the model with green dynamic arrows. Real-time footage from camera Cam_Corridor_5F is embedded in the display. Nearby fire hydrants and fire extinguishers are highlighted.

[0130] On-site AR guidance: Assuming a resident on the 5th floor (wearing AR glasses provided by the community or using a mobile AR app) is preparing to evacuate via the east staircase, the AR interface will immediately display a red warning: "Large fire ahead, please take a detour!" and overlay a green navigation arrow pointing to the west staircase in their field of vision. The first batch of community mini fire station members to arrive will wear AR glasses, which will highlight the location of the FH-5W fire hydrant and provide brief operation instructions.

[0131] VR Remote Command: Fire commanders wearing VR headsets in the command center can virtually view the fire situation on the 5th floor (a virtual scene synthesized from BIM models and real-time sensor data), assess the fire's development, and direct on-site firefighting and search and rescue operations. If cameras can see parts of the 502 apartment, these video feeds will also be projected onto the corresponding locations in the VR scene.

[0132] Information linkage: If a resident reports their location as trapped via a mobile app (e.g., via GPS or by manually selecting on a BIM map), the location will be displayed in real time on the command center's large screen and on the rescuers' AR devices.

[0133] 5. Incident Handling and Recovery: Firefighters arrive and conduct firefighting and rescue operations according to the emergency plan and on-site command. Once the fire is extinguished, all personnel have been evacuated, and there is no risk of reignition, the command center declares the emergency over. The system records all data from this emergency event for subsequent review, analysis, and plan optimization. The system resumes normal monitoring.

[0134] In summary, this invention discloses a community intelligent monitoring and emergency response method and system integrating BIM spatial semantics, comprising: constructing a BIM scene map to obtain the attributes and interrelationships of components and spatial areas in the BIM model; mapping dynamic targets detected in video surveillance to the BIM model to obtain their spatial semantic information; constructing a target-environment interaction graph based on the BIM spatial semantic information of the dynamic targets, using a graph neural network model for training and inference to identify specific complex events related to the spatial context; using the BIM model as a spatiotemporal reference, integrating multi-source heterogeneous data, and based on spatiotemporal proximity, object consistency, and predefined logical rules, using a graph-based event association algorithm to reconstruct a complete event chain containing atomic event sequences and related relationships; when an emergency event or event chain indicates an emergency state, dynamically generating an optimal emergency plan by combining BIM preset information and real-time sensor data, and performing visualized command and dispatch through BIM 3D scenes and augmented reality.

[0135] Through the above embodiments, it can be seen that the method and system of the present invention can significantly improve the community's monitoring and early warning capabilities, emergency response speed, and command and dispatch efficiency in the face of emergencies such as fires, thereby maximizing the protection of residents' lives and property.

[0136] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A community intelligent monitoring and emergency response method integrating BIM spatial semantics, characterized in that, Includes the following steps: S1: Construct a community building information model (BIM) scene map, which includes BIM component information, spatial area information and their respective semantic attribute information, as well as the topological relationship information between the BIM components and the spatial areas; S2: Acquire video surveillance data and detect dynamic targets from it. Map the dynamic targets to the three-dimensional space corresponding to the BIM scene map through coordinate transformation and geometric projection, and obtain the semantic information of the BIM space where the dynamic targets are located. S3: Based on the BIM spatial semantic information of the dynamic target, a local interaction graph containing the dynamic target node and its neighboring BIM element nodes is constructed, and the local interaction graph is analyzed using a pre-trained graph neural network (GNN) event recognition model, or the behavior of the dynamic target is analyzed in combination with a preset event rule base to identify specific events related to the spatial context. The node features of the GNN model include the kinematic features of the dynamic target and the functional and state features of the BIM spatial elements, and the edge features represent the interaction relationship between nodes. S4: Using the BIM as a unified spatiotemporal benchmark, integrate multi-source heterogeneous data acquired by at least two different types of sensors, perform spatiotemporal alignment and correlation analysis, and form an atomic event set; S5: Based on the topological relationship information in the BIM scene map, the motion trajectory of the dynamic target in the BIM space, and the set of atomic events formed in S4, an event chain construction algorithm is adopted. By evaluating the temporal sequence relationship, spatial proximity relationship, object consistency, and predefined logical association rules between atomic events, related atomic events are filtered and connected to reconstruct a complete event chain with temporal sequence and logical association. S6: When an emergency event is identified in S3 or the event chain reconstructed in S5 indicates an emergency state, the optimal emergency plan for the current emergency event is dynamically calculated and generated based on the emergency resource information, evacuation route information, hazard source information preset in the BIM scene map and the real-time sensor data fused in S4. S7: Visualize the optimal emergency plan in the BIM 3D scene, and choose to provide navigation and instructions to on-site personnel through augmented reality (AR) devices, or provide immersive situational awareness and command and dispatch support to command center personnel through virtual reality (VR) devices.

2. The method according to claim 1, characterized in that, The steps for constructing the BIM scene map in S1 specifically include: Parse the IFC file or native format file of the BIM model to extract the geometric information, physical properties and functional properties of the BIM components; Semantic enhancement is performed on the extracted BIM components and spaces to supplement or correct their semantic tags regarding function, safety level, and area of ​​origin. By using graph database technology, BIM components and spatial regions are used as nodes, and the inclusion, adjacency, or connectivity relationships between them are used as edges to construct the BIM scene graph.

3. The method according to claim 1, characterized in that, The graph neural network (GNN) event recognition model in S3 is either a graph attention network (GAT) or a relational graph convolutional network (R-GCN). The nodes of the input local interaction graph of the model include: dynamic target nodes, whose feature vectors contain target category, normalized velocity, direction of movement, and relative position in BIM space; and BIM element nodes, whose feature vectors contain element type, functional semantics, safety level, and current state. The edges of the local interaction graph are defined as: dynamic target-dynamic target edges, representing the relative distance and velocity between targets; dynamic target-BIM element edges, representing the distance between the target and the BIM element, whether contact or traversal behavior occurs, and the interaction type; and BIM element-BIM element edges, which inherit the topological relationship from the BIM scene graph.

4. The method according to claim 3, characterized in that, The training process of the GNN event recognition model includes: collecting labeled video data containing multiple predefined event scenarios and synchronous BIM interaction data; constructing a corresponding local interaction graph sequence for each labeled event scenario; using the local interaction graph sequence and its event labels as training samples, and using supervised learning to optimize the model parameters by minimizing the cross-entropy loss function between the predicted event category and the real event label.

5. The method according to claim 1, characterized in that, The event chain construction algorithm in S5 specifically includes: a. Define atomic events: Unify the video semantic events identified by S3, the sensor alarm events fused by S4, and the access control records into atomic events that include timestamps, BIM spatial location IDs, event types, a list of participating object IDs, and confidence scores; b. Construct a candidate event association graph: Using atomic events as nodes, add directed edges between pairs of atomic events that meet the conditions according to preset association rules. The association rules include: time order rules, spatial proximity rules, object consistency rules, and causal inference rules. c. Event chain extraction: In the candidate event association graph, a weight-based path search algorithm is used to extract one or more atomic event sequences as event chains. The weight of the path takes into account both the confidence of the atomic events and the strength of the associated edges. d. Event chain semantic annotation: The extracted event chain is matched with a predefined complex event scenario template library to give the event chain a high-level semantic description.

6. The method according to claim 1, characterized in that, The steps in S6 for dynamically generating the optimal emergency response plan specifically include: Based on the location of the fire point, the range of the hazardous gas leak, or the location of the intruder determined by real-time sensor data, the passage weight of relevant passages in the evacuation route is dynamically adjusted or marked as impassable in the BIM scene map. Based on the current personnel distribution information and the updated evacuation route information, the Dijkstra algorithm is used to calculate the optimal evacuation route from each trapped person to the safe exit. Based on the type and location of the emergency event, automatically find and recommend the nearest available emergency resources such as fire hydrants, fire extinguishers, and first aid kits marked in the BIM scene map.

7. A community intelligent monitoring and emergency response system integrating BIM spatial semantics, characterized in that, include: The BIM semantic parsing module is used to construct a BIM scene map of the community building information model. The BIM scene map includes BIM component information, spatial area information and their respective semantic attribute information, as well as the topological relationship information between the BIM components and the spatial areas. The video analysis and mapping module is used to acquire video surveillance data, detect dynamic targets, and map the dynamic targets to the three-dimensional space corresponding to the BIM scene map to obtain their BIM spatial semantic information. The semantic event detection module is used to construct a local interaction graph containing the dynamic target node and its neighboring BIM element nodes based on the BIM spatial semantic information of the dynamic target, and to analyze the local interaction graph using a pre-trained graph neural network (GNN) event recognition model, or to identify specific events related to the spatial context by combining a preset event rule base. The multi-source data fusion module is used to fuse multi-source heterogeneous data acquired by at least two different types of sensors, using the BIM as a unified spatiotemporal reference, to perform spatiotemporal alignment and correlation analysis, and form an atomic event set. The event reasoning and early warning module is used to reconstruct a complete event chain with temporal and logical correlation based on the topological relationship information in the BIM scene map, the motion trajectory of dynamic targets in the BIM space and the set of atomic events formed, and to issue an early warning when an emergency event is identified or the event chain indicates an emergency state. The emergency response plan generation module is used to dynamically calculate and generate the optimal emergency response plan for the current emergency event based on the emergency resource information, evacuation route information, hazard source information, and fused real-time sensor data preset in the BIM scene map after receiving an early warning. The 3D visualization and interaction module is used to visualize the optimal emergency plan in the BIM 3D scene. It can also provide navigation and instructions to on-site personnel through augmented reality (AR) devices, or provide immersive situational awareness and command and dispatch support to command center personnel through virtual reality (VR) devices.

8. The system according to claim 7, characterized in that, The BIM semantic parsing module is specifically configured as follows: it parses BIM model files through the built-in IFC parsing engine and semantic ontology library, extracts and enriches the semantics of BIM components and spaces, and uses a graph database to store and manage BIM scene maps.

9. The system according to claim 7, characterized in that, The graph neural network (GNN) event recognition model in the semantic event detection module has an architecture that includes multiple graph attention layers for learning weighted message passing between nodes, a graph pooling layer for generating graph-level representations, and a fully connected output layer for event classification. The model uses the Adam optimizer and cross-entropy loss function during training.

10. The system according to claim 7, characterized in that, The event reasoning and early warning module contains an event chain knowledge base for storing predefined atomic event types, association rules, and complex event scenario templates. It is also equipped with an event chain reasoning engine, which dynamically constructs and updates event chains based on real-time generated atomic events and association rules, and matches them with the template library for high-level semantic annotation and early warning judgment.

Images