Emergency escape path planning system and method based on BIM and Internet of Things dynamic identification

Abstract

The invention discloses an emergency escape path planning system and method based on BIM and Internet of Things dynamic identification, and the method comprises the steps: constructing a three-dimensional building information model, forming a digraph structure for the constructed three-dimensional building information model, and extracting the geometric information of each node; constructing a space mapping table of corresponding space units in the three-dimensional building information model and networking sensors deployed in the building; then multi-dimensional information including environmental parameters, personnel positioning and structural states in the building is collected in real time, so that a real-time dynamic data source in the building is formed; mapping the collected dynamic real-time data to a corresponding space unit in the constructed three-dimensional building information model, and carrying out dynamic identification on a dangerous area and an obstacle; and then on the basis of the real-time space state diagram, an optimal escape path avoiding a high-risk area is automatically generated, and dynamic re-planning can be carried out. According to the scheme, the space expression ability of the BIM and the real-time sensing ability of the Internet of Things are deeply fused, dangerous factors and traffic obstacles in the building can be dynamically recognized, the real-time escape path is generated in combination with the path algorithm, and the problems existing in the prior art can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the field of integrated application of building information modeling (BIM) and Internet of Things technology, and in particular to an emergency escape path planning technology based on BIM and Internet of Things dynamic perception information. Background Art

[0002] In modern urban buildings, the ability to respond to sudden incidents like fires, explosions, and earthquakes is becoming an increasingly important indicator of a building's intelligence and public safety capabilities. Traditional emergency evacuation mechanisms rely heavily on pre-set routes and static signage, unable to dynamically adjust to changing on-site conditions. This can easily lead to people straying into dangerous areas or becoming stuck in bottlenecks during disasters, severely impacting evacuation efficiency and safety.

[0003] At the same time, building information modeling technology has been widely used in the design, construction, and operation and maintenance stages in recent years. Its powerful spatial representation, data integration, and information visualization capabilities have become a crucial foundation for building digital twins. Furthermore, the development of the Internet of Things (IoT) has enabled sensors deployed within buildings to collect real-time, multi-dimensional sensory data, including temperature, smoke, gas concentration, occupancy, and traffic status. This has made it possible to monitor building operations in real time.

[0004] However, the integration of BIM and the Internet of Things (IoT) in emergency response is still in its infancy. Existing solutions typically rely solely on embedding IoT monitoring data dashboards within BIM platforms or displaying sensor values ​​within model component properties through one-way data interfaces. These solutions lack mechanisms for dynamically binding sensor data in time and space and automatically identifying it. Most solutions remain at the "visualization level," mapping building sensor data as icons or heat maps and overlaying them within the BIM interface. This makes it difficult to achieve data-driven intelligent responses. For example, in the event of a building fire, while the system can display the alarm location, it cannot automatically identify smoke diffusion paths, determine obstructions, or generate optimal evacuation routes based on real-time conditions. Furthermore, there is a lack of mechanisms for deeply integrating data status with spatial topology, failing to support regularized dynamic identification of hazardous areas and coordinated route replanning responses. Consequently, emergency decision-making relies on manual judgment and fails to meet the needs for real-time risk avoidance and multi-terminal intelligent guidance during emergencies. These issues highlight that current BIM and IoT integration has yet to form a closed-loop technical chain: "spatial modeling - data fusion - intelligent decision-making - visual feedback."

[0005] As can be seen, existing solutions commonly suffer from technical bottlenecks such as a disconnect between models and sensory data, a single hazard identification mechanism, a lack of dynamic path planning updates, and untimely visual feedback. Particularly in emergencies, there is a lack of systematic, engineered solutions for identifying fire sources, high-temperature areas, toxic gas zones, structural damage points, and temporary obstacles within a building in real time, and for planning and displaying intelligent evacuation routes based on these indicators.

[0006] The Chinese invention patent application with publication number CN111832811A specifically discloses an intelligent fire evacuation method based on digital twins. This solution primarily relies on video surveillance and historical evacuation data to predict personnel distribution, and uses a deep learning model based on a spatiotemporal graph convolutional network to predict crowd flow. However, its perception data source is relatively single, mainly concentrated in camera monitoring, and lacks the integration of multi-source sensor data such as temperature, smoke, and access control status, resulting in an incomplete perception of environmental changes. In addition, in terms of path planning, the solution mainly generates and evaluates evacuation plans based on historical data and preset models, lacking the ability to quickly respond to real-time environmental changes. When the actual evacuation path changes, the evacuation simulation analysis needs to be repeated, which is a relatively complex process and difficult to meet the needs of rapid response in emergency situations. Summary of the Invention

[0007] In response to the problems existing in the existing emergency response solutions based on the integration of BIM and the Internet of Things, the purpose of the present invention is to provide an emergency escape path planning solution based on dynamic identification of BIM and the Internet of Things. This solution deeply integrates the spatial expression capabilities of BIM with the real-time perception capabilities of the Internet of Things, can dynamically identify dangerous factors and traffic obstacles inside the building, and generate real-time escape paths in combination with path algorithms, which can effectively overcome the problems existing in the existing technology.

[0008] In order to achieve the above objectives, the present invention provides a BIM and IoT-based dynamic identification emergency escape path planning system, the system comprising:

[0009] A BIM three-dimensional spatial modeling unit, wherein the BIM three-dimensional spatial modeling unit is configured to construct a three-dimensional building information model that reflects the building's geometric structure, topological relationship, and semantic information. The three-dimensional building information model performs spatial segmentation and topological extraction on corresponding spatial units to form a directed graph structure. The nodes in the directed graph structure are configured as functional areas or spatial units, and the edges are configured as traffic paths. The geometric information between each node is extracted. The BIM three-dimensional spatial modeling unit also constructs a spatial mapping table between the spatial units in the three-dimensional building information model and the networked sensors in the Internet of Things perception unit.

[0010] An IoT sensing unit, which collects multi-dimensional information about the building's interior, including environmental parameters, occupant location, and structural status, in real time through a number of networked sensors deployed inside the building, thereby forming a real-time dynamic data source within the building;

[0011] A data fusion and hazard identification unit configured to interact with the BIM three-dimensional spatial modeling unit and the Internet of Things perception unit to map the dynamic real-time data collected by the Internet of Things perception unit to the corresponding spatial unit constructed by the BIM three-dimensional spatial modeling unit to generate a real-time spatial state diagram, and to dynamically identify dangerous areas and obstacles based on rules or models;

[0012] A path calculation unit is configured to interact with the data fusion and hazard identification unit to construct a multidimensional path evaluation function based on the real-time spatial state diagram dynamically generated in the data fusion and hazard identification unit, thereby automatically generating an optimal escape path that avoids high-risk areas and enabling dynamic replanning.

[0013] Furthermore, the IoT perception unit establishes a mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the three-dimensional building information model for the deployed networked sensors, and maps the perception data of the networked sensors to the corresponding spatial units in the three-dimensional building information model through three-dimensional coordinate matching or topological alignment.

[0014] Furthermore, the data fusion and hazard identification unit first locates the spatial units in the three-dimensional building information model corresponding to the dynamic real-time data collected by the Internet of Things sensing unit according to the mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the three-dimensional building information model; then, weighted averaging, maximum value extraction and / or time trend analysis are performed on the values ​​collected by the networked sensors corresponding to each spatial unit to obtain a comprehensive status evaluation value of each spatial unit, thereby generating a real-time spatial status diagram.

[0015] Furthermore, when calculating the optimal escape path, the path calculation unit first calls the fused BIM topology map and the real-time spatial status map to construct a weighted access map; then the constructed multi-dimensional path evaluation function calculates the optimal emergency escape path based on the weighted access map.

[0016] Furthermore, the emergency escape path planning system also includes a path display and interaction unit, which is configured to interact with the BIM three-dimensional space modeling unit and the path calculation unit data, and is used to display the path calculation results of the path calculation unit in a visual form in the three-dimensional building information model, and can also push personalized evacuation guidance information through multiple terminals.

[0017] In order to achieve the above object, the present invention provides a method for planning emergency escape paths based on dynamic identification of BIM and the Internet of Things, the method comprising:

[0018] A three-dimensional building information model is constructed that can reflect the building's geometric structure, topological relationship, and semantic information. Spatial segmentation and topological extraction are performed on corresponding spatial units in the constructed three-dimensional building information model to form a directed graph structure. Nodes in the directed graph structure are configured as functional areas or spatial units, and edges are configured as passage paths. Geometric information between each node is extracted. A spatial mapping table between the spatial units and networked sensors deployed in the building is also constructed for the corresponding spatial units in the constructed three-dimensional building information model.

[0019] Deploy several networked sensors inside the building to collect real-time multi-dimensional information including environmental parameters, occupant location, and structural status, forming a real-time dynamic data source inside the building;

[0020] Mapping the collected dynamic real-time data to the real-time spatial status diagram generated in the corresponding spatial unit in the constructed 3D building information model, and dynamically identifying dangerous areas and obstacles based on rules or models;

[0021] A multi-dimensional path evaluation function is constructed based on the dynamically generated real-time spatial state diagram, which automatically generates the optimal escape path avoiding high-risk areas and can perform dynamic replanning.

[0022] Furthermore, the method establishes a mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the three-dimensional building information model for the deployed networked sensors, and maps the perception data of the networked sensors to the corresponding spatial units in the three-dimensional building information model through three-dimensional coordinate matching or topological alignment.

[0023] Furthermore, the method, for dynamic real-time data collected by IoT sensing units, first locates the spatial units in the 3D building information model corresponding to the collected dynamic real-time data based on a mapping table between the networked sensor deployment coordinates and component IDs in the 3D building information model. Next, weighted averaging, maximum value extraction, and / or temporal trend analysis are performed on the values ​​collected by the networked sensors corresponding to each spatial unit to obtain a comprehensive status assessment value for each spatial unit, thereby generating a real-time spatial status diagram.

[0024] Furthermore, when calculating the optimal escape path, the method first calls the fused BIM topology map and the real-time spatial status map to construct a weighted access map; then the constructed multi-dimensional path evaluation function calculates the optimal emergency escape path based on the weighted access map.

[0025] Furthermore, the method also includes a path display and interaction step, in which the path calculation results of the path calculation unit can be displayed in a visual form in a three-dimensional building information model, and personalized evacuation guidance information can be pushed through multiple terminals.

[0026] The emergency escape path planning scheme provided by the present invention is based on BIM and Internet of Things dynamic recognition technology, deeply integrating the spatial expression ability of BIM with the real-time perception ability of the Internet of Things, combining multiple hazard recognition mechanisms to perform real-time dynamic recognition of dangerous factors and passage obstacles inside the building, and combining the path algorithm to generate a real-time escape path, thereby realizing dynamic update of the escape path, which can effectively overcome the problems existing in the existing technology.

[0027] The solution of the present invention deeply integrates various types of IoT sensors (such as temperature, smoke, access control, positioning, etc.) with the BIM model to construct a real-time spatial status diagram, thereby achieving comprehensive perception and dynamic update of the internal environmental status of the building, and improving the system's response speed and accuracy to emergencies; at the same time, an improved path search algorithm is adopted, combined with the real-time updated spatial status diagram, which can immediately trigger a local replanning mechanism when it detects that a certain side of the path becomes blocked or high-risk, and quickly generate the optimal escape route avoiding dangerous areas, ensuring the safety and efficiency of personnel evacuation.

[0028] On this basis, the solution of the present invention can also render the optimal escape route in real time in the BIM model, and can simultaneously push it to mobile terminals and evacuation guidance devices, so as to display evacuation guidance information to personnel intuitively and in real time through three-dimensional visualization means, and realize the closed-loop response mechanism of "perception-recognition-decision-making-guidance".

[0029] The emergency escape route planning scheme provided by the present invention can not only be used for emergency evacuation of building fires in practical applications, but also has the potential to be promoted and applied in complex spaces such as hospitals, schools, subways, and complexes, and has broad engineering value and practical significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The present invention is further described below with reference to the accompanying drawings and specific embodiments.

[0031] Figure 1 This is a flow chart of the method for dynamic identification of emergency escape routes based on BIM and the Internet of Things in the present invention;

[0032] Figure 2 This is a block diagram of the emergency escape path planning system based on BIM and the Internet of Things dynamic identification in the present invention;

[0033] Figure 3 This is an example diagram of the composition of the BIM three-dimensional space modeling unit in the present invention;

[0034] Figure 4 This is an example diagram of the structure of the Internet of Things sensing unit in the present invention;

[0035] Figure 5 This is an example diagram of the structure of the data fusion and hazard identification unit in the present invention;

[0036] Figure 6 This is a diagram illustrating an example of the structure of a path calculation unit in the present invention;

[0037] Figure 7 This is an example diagram of the structure of the path display and interaction unit in the present invention. DETAILED DESCRIPTION

[0038] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below with reference to specific illustrations.

[0039] Based on BIM and IoT dynamic recognition technology, the present invention deeply integrates the spatial expression capability of BIM with the real-time perception capability of IoT, thereby realizing dynamic recognition of obstacles, fire sources, toxic gases, etc. in buildings under disaster environments, automatically avoiding obstacles and planning feasible escape routes in real time.

[0040] Based on this, the present invention provides a method for planning emergency escape paths based on dynamic identification of BIM and the Internet of Things. Figure 1 As shown in FIG, the emergency escape path planning method mainly includes the following stages:

[0041] S1: Construction of three-dimensional building information model (abbreviated as: BIM model).

[0042] This phase is used to construct a 3D building information model (BIM model) that reflects the building's geometric structure, topological relationships, and semantic information. This BIM model will provide a digital foundation for subsequent data fusion mapping, path calculation, and visualization.

[0043] In this stage, spatial segmentation and topology extraction are performed on the corresponding spatial units in the constructed BIM model to form a corresponding directed graph structure. The nodes in this directed graph structure are configured as functional areas or spatial units, and the edges are configured as traffic paths. At the same time, the geometric spatial information between each node is extracted as the basic parameters for subsequent escape path calculations.

[0044] In this stage, a spatial mapping table between the corresponding spatial units in the constructed three-dimensional building information model and the networked sensors deployed in the building is also constructed.

[0045] S2: IoT perceives the internal environment of buildings.

[0046] In this stage, several networked sensors are deployed inside the building, and a spatial mapping relationship is formed between the deployed networked sensors and the corresponding spatial units in the BIM model constructed in stage S1.

[0047] On this basis, a number of networked sensors are deployed to collect multi-dimensional information inside the building, including environmental parameters, personnel positioning and structural status, in real time, to form a real-time dynamic data source inside the building.

[0048] The environmental parameters mentioned here include temperature, humidity, smoke, carbon monoxide concentration, light intensity and other environmental parameters.

[0049] The structural status mentioned here includes temporary obstacles or passage blockages within the building.

[0050] S3: Data fusion and hazard identification.

[0051] This stage maps the dynamic real-time data collected in stage S2 to the corresponding spatial units in the BIM model constructed in stage S1 to form a real-time spatial status diagram, and dynamically identifies dangerous areas and obstacles based on rules or models.

[0052] Furthermore, this stage coordinately maps and semantically binds the real-time data from various IoT sensors such as smoke, temperature, access control, and positioning deployed in stage S2 to the spatial units in the BIM model, and assigns real-time status information to each spatial unit, including labels such as risk level, passability, and congestion. All information is written into the structured graph data to construct a real-time spatial status diagram. This real-time spatial status diagram not only contains static geometric topology, but also dynamically expresses the real-time status of each traffic path and node, which will serve as the basis for obtaining effective path information in the subsequent emergency escape path calculation stage.

[0053] S4: Emergency escape route calculation.

[0054] Based on the real-time spatial state diagram generated in stage S3, this stage combines the improved path search algorithm to automatically generate the optimal escape path that avoids high-risk areas and can perform dynamic replanning.

[0055] In this stage, the generated real-time spatial status diagram is integrated with the BIM spatial topology to form a weighted graph model. An improved A* path search algorithm that introduces risk factors, congestion coefficients, and traffic probabilities is used, and a multi-factor path evaluation mechanism is introduced to generate the optimal escape path that avoids high-risk areas.

[0056] Specifically, in the weighted graph model formed in this stage, the weight of each path edge not only includes distance, but also superimposes risk factors, personnel density factors and obstacle penalty coefficients, and prioritizes the search for low-risk traversable paths by setting a heuristic function; at the same time, when multiple feasible paths exist, the optimal escape path is dynamically selected by comprehensively evaluating factors such as path continuity, replanning cost and travel time.

[0057] The optimal escape route generated in this way takes into account path safety, real-time performance and flexibility, significantly improving emergency escape response capabilities.

[0058] The following is a detailed description of the specific implementation scheme of each stage in the emergency escape path planning method based on BIM and Internet of Things dynamic identification given in the present invention.

[0059] When constructing the BIM model in stage S1, this method preferably establishes a complete building structure model based on a BIM modeling platform such as Revit and ArchiCAD, and exports it in IFC or a custom format for system integration in actual application.

[0060] On this basis, the constructed BIM model includes components such as walls, floor slabs, doors and windows, evacuation exits, stairs, elevators, and evacuation signs, and each component is given a unique code (i.e. ID) and semantic attributes. The BIM model constructed in this way accurately reflects the building's geometric structure, topological relationship, and semantic information.

[0061] The semantic attributes here include component usage (such as walls, evacuation doors), access type (passable / closed), safety level, material, floor, space unit, and other information.

[0062] In this solution, corresponding semantic attributes are assigned to each component to support perception data binding, evacuation path constraint judgment, and spatial accessibility calculation, thereby building data-driven intelligent decision-making logic.

[0063] Topological relationships are embodied through the connectivity and flow of traffic between building spaces, such as doors connecting two rooms, stairs connecting upper and lower floors, and corridors connecting rooms to evacuation exits. By extracting these relationships and constructing a directed graph structure, the path planning module can identify accessible paths based on the BIM model and calculate the optimal route.

[0064] Furthermore, in order to facilitate subsequent efficient data analysis and spatial calculation, when constructing a BIM model, this method performs spatial segmentation and topology extraction on spatial areas such as rooms, passages, floors, etc. (i.e., spatial units) in the constructed BIM model to form a directed graph structure. The nodes in the directed graph structure are configured as functional areas or other corresponding spatial units in the BIM model, and the edges in the directed graph structure are configured as passage paths in the BIM model.

[0065] On this basis, during the construction of the BIM model, geometric space information such as the geometric distance between nodes, floor height difference, and passage width are also extracted synchronously to serve as the basic parameters for subsequent escape path calculations.

[0066] Furthermore, this method pre-sets a binding relationship between each spatial unit in the constructed BIM model and the networked sensors deployed on-site in the building, thereby forming a spatial mapping table between the spatial units in the BIM model and the networked sensors deployed on-site in the building. For example, a spatial unit in the BIM model includes a corresponding spatial area, and each spatial area includes a corresponding component. Thus, a "room ID-component ID-sensor ID" mapping table can be pre-set, thereby enabling spatial mapping between the BIM model and IoT sensors.

[0067] Furthermore, in stage S1, the method can import the constructed BIM model into Cesium, Unity or WebGL engine, and realize lightweight three-dimensional display through glTF or custom format, which facilitates subsequent path rendering and human-computer interaction.

[0068] Furthermore, to support multi-source data access and semantic query, this method encapsulates the BIM model interface into a RESTful API for the constructed BIM model, providing services such as component attribute query, channel connectivity retrieval, and evacuation path query.

[0069] At this stage, this method can realize the standardized construction, spatial topology abstraction and three-dimensional lightweight visualization of the BIM model, providing structural support for the operation of the overall plan.

[0070] When this method perceives the internal environment of a building based on the Internet of Things in stage S2, multiple types of networked sensors are deployed for each spatial unit inside the building to collect multi-dimensional information such as temperature, humidity, smoke concentration, carbon monoxide concentration, light intensity, occupant location, and structural status in real time within the corresponding spatial unit, thus forming a dynamic data input source.

[0071] For example, in specific deployments, temperature and humidity sensors and smoke sensors should be installed in high-risk areas such as corridors, computer rooms, and storage areas within a building;

[0072] UWB or Bluetooth AOA positioning base stations are deployed in public areas, entrances and exits, and evacuation passages within the building to achieve accurate positioning of personnel;

[0073] LiDAR and infrared sensors are used to identify temporary obstacles or channel blockages.

[0074] On this basis, edge computing gateways are further deployed to perform edge data processing on networked sensors deployed on site.

[0075] Among them, all sensor networks deployed on site are connected to the edge computing gateway through wired RS485 or wireless LoRa, Wi-Fi, NB-IoT, etc.

[0076] The edge computing gateway can perform preliminary data filtering and anomaly identification on the collected data uploaded by networked sensors, and cache the processed data; on this basis, the edge computing gateway further encapsulates the processed collected data through the MQTT protocol and uploads it to the back-end system for subsequent processing.

[0077] Here, the edge computing gateway encapsulates the perception data collected by networked sensors at the front end according to a unified structure, including fields such as device number, data type, collection time, spatial coordinates, numerical value and status code, thereby ensuring the accuracy and reliability of data transmission and improving the efficiency of subsequent data fusion processing.

[0078] Furthermore, in order to ensure the spatial matching between the front-end perception data and the BIM model, this method establishes a mapping relationship table between the deployment coordinates of the networked sensors deployed in the building and the component ID in the BIM model, and adopts three-dimensional coordinate matching or topological alignment technology to ensure that the perception data is accurately mapped to the corresponding spatial unit.

[0079] Specifically, to achieve spatial matching between the perception data of IoT sensors deployed on-site and the BIM model, and to ensure that the perception data of IoT sensors can be accurately mapped to the corresponding spatial units in the BIM model, this solution first establishes a "networked sensor deployment coordinates-component ID" mapping relationship table based on the unique geometric boundaries and component IDs of all spatial units in the BIM model (such as rooms, corridors, and stairwells);

[0080] Next, when deploying sensors, the 3D installation coordinates of each sensor are recorded and uniformly converted into a building coordinate system consistent with the BIM model;

[0081] Subsequently, based on the three-dimensional coordinate matching method, the system performs a spatial containment judgment on the spatial position of the sensor and the geometric boundaries of the components or spatial units in the BIM model. Using the three-dimensional spatial containment judgment algorithm, the spatial position of each sensor is compared with the room boundary, component volume or channel geometry in the BIM model. Through spatial envelope operations (such as the point-in-polyhedron algorithm), it is determined whether the sensor is inside a specific spatial unit, and then the spatial unit ID or component ID to which it belongs is determined; if multiple spaces are adjacent or overlapping, the closest distance priority or weight rule is used to determine the unique mapping relationship.

[0082] For sensors deployed in continuous spaces such as passages and stairs, topological alignment technology is used to map their positions to adjacent nodes or edges in the spatial topology map generated by the BIM model, ensuring that the data is bound to the key structural units of the spatial passage path to ensure that dynamic state changes can be accurately reflected during path calculation.

[0083] Finally, the system establishes an association relationship table of "sensor ID-coordinate-component ID-spatial unit ID" to achieve accurate one-to-one mapping between perception data and spatial semantic units, providing a data basis for subsequent spatial state assignment, hazard identification and path planning, and ensuring the spatial accuracy and real-time performance of subsequent hazard identification and path generation.

[0084] Furthermore, at this stage, the method further sets corresponding data sampling and uploading frequencies for the deployed networked sensors to ensure real-time data. The specific setting scheme is not limited here and can be determined according to actual needs.

[0085] At this stage, this method provides a real-time and reliable data foundation for subsequent hazard identification and path planning through multi-source and multi-dimensional environment and status perception.

[0086] When performing data fusion and hazard identification in stage S3, this method first receives dynamic real-time perception data streams with spatial encoding from the networked sensors and edge computing gateways deployed in stage S2; on this basis, the spatial units in the BIM model corresponding to the collected dynamic real-time data are located according to the mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the BIM model, that is, the spatial units to which the dynamic real-time perception data streams belong, such as rooms, channels, nodes, etc., are located. In this way, on the basis of the BIM spatial topology diagram, the real-time environmental perception data from the Internet of Things sensors can be fused (such as coordinate mapping and semantic binding of the real-time data of multiple types of Internet of Things sensors with the spatial units in the BIM model) to generate the corresponding spatial status diagram.

[0087] The spatial encoding here is a logical identifier used to uniquely identify the spatial unit or component to which the perception data belongs in the building BIM model, which can be generated when the sensor is deployed in stage S2.

[0088] Furthermore, the spatial code includes but is not limited to the sensor-bound component ID, spatial unit ID, floor number, installation location code, etc. For further explanation, there are two main ways to generate spatial codes: one is to automatically extract the ID of the component to which each sensor belongs as the spatial code by matching the 3D installation coordinates of each sensor with the component in the BIM model during the deployment phase; the other is to establish a "sensor ID-component ID-spatial unit ID" mapping relationship table during manual configuration or system initialization.

[0089] By automatically matching the data collected by the sensors with the corresponding spatial coding during the data collection process, real-time data fusion and positioning analysis can be carried out in stage S3, enabling the system to accurately identify the location of the data source space and its topological context; by introducing spatial coding, it can facilitate the subsequent deep integration of BIM and IoT data, spatial semantic binding and dynamic status tracking.

[0090] Next, the fusion calculation stage begins, and weighted averaging, maximum value extraction, or time trend analysis is performed on the values ​​collected by multiple networked sensors corresponding to each spatial unit to obtain a comprehensive status assessment value of the space. After calculating the comprehensive status assessment value of each spatial unit, dangerous areas and obstacles are dynamically identified through the preset risk judgment rule library and semantic threshold model, and the identification results are written into the spatial status diagram in real time.

[0091] In order to effectively identify hazards, this method configures multiple risk judgment rules in the risk judgment rule library. The risk judgment rules here are based on logical judgment conditions set by multiple real-time perception data sources (such as temperature, smoke concentration, carbon monoxide concentration, infrared occlusion status, personnel density, etc.), and are used to dynamically identify dangerous areas, obstacles, congestion status, etc. inside the building. Specifically, in stage S3, the comprehensive state evaluation value of each spatial unit is first calculated. This value integrates the data from multiple sensors in the unit through methods such as weighted average, maximum value extraction or time trend analysis to form a quantitative expression of the current state of the space; based on this, the risk judgment rule sets the threshold judgment logic based on the calculated evaluation value. In this way, the comprehensive state evaluation value can be used as an input condition, and the judgment rule can be used as a trigger mechanism. The two together constitute the judgment logic chain of hazard identification, ensuring that risk areas can be quickly and accurately identified and the spatial state map can be updated based on the multi-source perception results.

[0092] As a further explanation, the various risk judgment rules here include space unit risk judgment rules, space unit obstacle identification rules, and space unit personnel positioning rules.

[0093] The spatial unit risk assessment rules pre-set corresponding thresholds for various environmental parameters within each spatial unit, and set corresponding risk levels based on different environmental parameter threshold conditions. This is achieved by using environmental parameter sensors deployed in the corresponding spatial units within the building to obtain the corresponding environmental parameters of each spatial unit. Based on this, the environmental status assessment value (such as the smoke concentration status assessment value, smoke density status assessment value, etc.) of each spatial unit is obtained, and the risk level of each spatial unit is then determined according to the assessment rules.

[0094] For example, within a certain spatial unit, the on-site smoke concentration state assessment value within the spatial unit is calculated based on the smoke concentration value collected by the on-site sensor. If the on-site smoke concentration state assessment value exceeds the set threshold, the area corresponding to the spatial unit is marked as a "suspicious fire area";

[0095] For a certain spatial unit, the on-site temperature status assessment value within the spatial unit is calculated based on the temperature value collected by the on-site sensor. If the on-site temperature status assessment value exceeds the set threshold of 60°C, the area corresponding to the spatial unit is marked as a "high temperature danger zone";

[0096] For a certain spatial unit, the corresponding on-site smoke concentration state assessment value and on-site temperature state assessment value are calculated based on the smoke concentration value and temperature value collected by the on-site sensors. If the on-site smoke concentration state assessment value exceeds the set threshold, and the on-site temperature state assessment value also exceeds the set threshold of 60°C, and both continue to be abnormal within the sliding time window, that is, multiple judgment conditions are met at the same time, then the area corresponding to the spatial unit is marked as a "high-risk closed area"; its node is given a "dangerous" state label in the spatial state diagram and the pass cost is increased or it is set to be impassable.

[0097] Furthermore, for the spatial unit obstacle recognition rules, thresholds are set for the spatial state of the spatial unit, and the obstacle state within the spatial unit is determined based on different thresholds. In this way, based on the data detected by the obstacle recognition equipment (such as infrared signal sensors, radars, cameras, etc.) deployed in the corresponding spatial unit in the building, the spatial state evaluation value of each spatial unit is calculated, and then the obstacle state within each spatial unit is determined according to the obstacle judgment rules.

[0098] For example, for a certain space unit, the spatial state evaluation value of the space unit is calculated based on the continuous interruption duration of the infrared radio signal deployed at the space unit site in the building, radar data and / or image recognition results. The obstacle status in the space unit is determined by combining the spatial state evaluation value of the space unit with the set threshold:

[0099] If the spatial channel width evaluation value in the spatial state evaluation value of the spatial unit is less than the set safe passage threshold (such as 0.8 meters), the area corresponding to the spatial unit is marked as a "temporary blocking area" and the corresponding mark in the passage map is an unreachable or high-penalty edge;

[0100] If the infrared signal continuous interruption evaluation value in the spatial state evaluation value of the spatial unit exceeds a predetermined value (e.g., 10 seconds), the area corresponding to the spatial unit is marked as a "temporary blocking area" and marked as an unreachable or high-penalty edge in the traffic map;

[0101] If the obstacle blocking evaluation value in the image recognition result of the spatial state evaluation value of the spatial unit is positive, the area corresponding to the spatial unit is marked as a "temporary blocking area" and the corresponding mark in the traffic map is an unreachable or high-penalty edge.

[0102] Furthermore, for the spatial unit personnel positioning rules, the personnel density and / or residence time per unit area are determined based on the density of personnel positioning data, corresponding thresholds are set for the personnel density and / or residence time per unit area, and then the personnel congestion status in the spatial unit is set based on the personnel density and / or residence time thresholds per unit area.

[0103] For example, for a certain spatial unit, based on the personnel positioning data fed back by the UWB or Bluetooth AOA positioning base station deployed at the spatial unit site in the building, the personnel positioning status evaluation value in the spatial unit (including the personnel density per unit area and the residence time at the positioning point) is calculated. If the personnel density per unit area evaluation value in the personnel positioning status evaluation value is higher than the set threshold, and the residence time evaluation value of the positioning point exceeds the set threshold, the area corresponding to the spatial unit is marked as a "congested area" and the traffic weight in the path calculation is increased.

[0104] On this basis, this method further writes the above recognition results into a unified spatial state diagram, which can be called during subsequent path calculation to achieve risk avoidance, obstacle avoidance and dynamic replanning of the path, thereby ensuring the system's rapid response and accurate modeling of dangerous areas and passage obstacles in emergencies.

[0105] The aforementioned real-time data based on multi-type IoT sensors is coordinate mapped and semantically bound to the spatial units in the BIM model to form a basic spatial state diagram. On this basis, the recognition results here will be uniformly written into the spatial state diagram to give real-time status information to each spatial unit, and give each node and edge in each spatial unit the current risk level, obstacle status, congestion status and other traffic status labels and corresponding traffic weights. All information is written into the structured graph data. The constructed spatial state diagram not only contains static geometric topology, but also dynamically expresses the real-time status of each traffic path and node, which can be used to obtain effective path information in subsequent path calculations.

[0106] The spatial state diagram of the recognition results written in this way will be able to realize functions such as dynamic update, semantic expression and real-time response. Based on this, the structured expression and dynamic update of the real-time risk status inside the building can be realized, providing a complete and computable graph theory model support for the path planning module.

[0107] As an example, based on this spatial state diagram, by encoding and setting weights for different spatial states, it is possible to intelligently avoid high-risk areas or blocked channels according to the real-time environment, generate the optimal escape path, and support local replanning, thereby greatly improving emergency response efficiency and the reliability and safety of evacuation paths. Such a path planning mechanism can achieve a closed-loop linkage from perception-recognition-mapping-decision-making.

[0108] Furthermore, to enhance the system's intelligence and foresight, this method also introduces a state update and adaptive mechanism at this stage to dynamically maintain and intelligently predict the spatial state graph. This state update and adaptive mechanism integrates a sliding time window judgment mechanism, a data trend extrapolation mechanism, and a pre-closed warning mechanism.

[0109] Specifically, the state update and adaptation mechanism, when introducing a sliding time window judgment mechanism, will continuously monitor changes in various sensor data sequences and determine the risk level based on the changing state. For example, if the temperature or smoke concentration in a certain spatial unit exceeds the threshold multiple times within a set time, it will be judged as a persistent danger rather than a momentary fluctuation, and the risk level will be promptly increased.

[0110] When introducing the trend extrapolation mechanism, the status update and adaptive mechanism will analyze the changing trends of historical data (such as the slope of smoke concentration and the growth rate of carbon monoxide concentration) to predict its status trend in the short term in the future and mark in advance the areas that are about to enter high-risk areas.

[0111] When the state update and adaptive mechanism introduces pre-emptive closure warnings, if the system predicts that a channel or exit will become impassable within tens of seconds, it sends a warning signal to the path calculation module in advance, activating an alternative route or performing local replanning. This module transitions from "real-time monitoring" to "preemptive intervention," ensuring the system's proactive risk identification and dynamic path adjustment capabilities.

[0112] For example, when dynamically maintaining and intelligently predicting the spatial state diagram based on this state update and adaptive mechanism, if it is detected that the temperature in a certain area has risen continuously within 3 minutes and the smoke has increased simultaneously, an early warning will be issued and its risk level will be increased, driving the recalculation of the path.

[0113] In this stage, this method forms a closed loop of hazard identification driven by perception through rule fusion, spatial empowerment and dynamic update.

[0114] When calculating and planning the optimal emergency escape path in stage S4, the method first calls the fused BIM topology graph and the spatial state graph to construct a weighted traffic graph, in which the nodes in the weighted traffic graph are configured to represent spatial units or key channel intersections, and the edges in the graph are configured to represent traffic paths.

[0115] When constructing the weighted access graph, each spatial node and the access edges between them are converted into a graph structure based on the fused BIM topology graph and spatial state graph, where the weight value W of the edge in the graph is calculated using a multi-factor weighted model.

[0116] Here, the edge weights in the weighted traffic graph are calculated by combining geometric distance, risk level, obstacle resistance, and personnel density. The calculation formula is as follows:

[0117] W = D × (1 + αR + βC + γB);

[0118] Where D is the physical distance of the edge, R is the risk coefficient (derived from sensor judgment), C is the personnel congestion density coefficient, B is the obstacle judgment result, and α, β, and γ are adjustment weights.

[0119] This weighted mechanism not only considers the shortest distance, but also integrates risk avoidance, obstacle avoidance and traffic efficiency to form a dynamically adjustable disaster response traffic map.

[0120] Furthermore, after obtaining the weighted traffic map, the improved A* algorithm or the dynamic Dijkstra algorithm path search algorithm is called to perform path search, and the optimal emergency escape path is calculated based on the weighted traffic map according to this algorithm.

[0121] The improved A* algorithm in this method introduces a multidimensional heuristic function. Based on the original distance estimation, it incorporates a risk penalty function, congestion weights, and obstacle cost factors, making the path search more inclined to choose low-risk, low-resistance areas. The dynamic Dijkstra algorithm supports rapid replanning in the event of path interruptions, avoiding full map recalculation, improving algorithm execution efficiency and response speed, and enhancing the system's real-time adaptability in emergency situations.

[0122] Therefore, the search process no longer relies solely on Euclidean distance; the heuristic function also incorporates the current path's risk level, traffic conditions, and expected time, ensuring that high-risk (or blocked) paths are prioritized. Among multiple candidate exit paths, the system selects the path with the lowest total cost and best status as the optimal escape path, outputting it as a coordinate sequence for the display module.

[0123] Furthermore, when calculating the optimal emergency escape path, this method can also combine heuristic functions to speed up the optimization as needed, and support target switching and multi-target redundant path generation (spare exits).

[0124] Specifically, this method further improves the efficiency and reliability of path calculation by introducing a heuristic function, a target switching mechanism, and a multi-target redundant path generation strategy to achieve more intelligent escape path planning. First, the heuristic function, based on a weighted combination of the estimated cost (such as Euclidean distance or floor distance) from the current node to the target exit and the risk assessment value, guides the A* algorithm to prioritize nodes that are more likely to become the optimal path during the search process, significantly reducing the invalid search area and accelerating the optimization process. Second, when constructing the weighted access graph, multiple exit nodes can be set as reachable targets. If the current target exit is detected to be impassable or the risk level increases during path planning, the system automatically switches to a suboptimal exit and recalculates the path, ensuring escape continuity and fault tolerance. Finally, under normal conditions, the system can also concurrently calculate the primary path and one or two backup paths, forming a multi-target path set based on different exits or avoidance strategies, and monitor the traffic status of each path in real time. If the primary path is blocked, it can immediately switch to the backup path without re-searching the entire system. This strategy presets multiple path options in the path calculation phase and implements fast switching in the execution phase, significantly enhancing the system's robustness and response efficiency in dealing with dynamic emergencies.

[0125] Furthermore, when the system detects that an edge in the path becomes blocked or in a high-risk state, it immediately triggers a local replanning mechanism to avoid recalculating the entire map and improve response speed.

[0126] Specifically, this method continuously monitors data streams from various networked sensors during the data fusion and hazard identification phase (S3). It then uses pre-set rules to determine whether the status of path edges has changed. For example, if smoke concentration in a passage rises sharply, the temperature exceeds a threshold, the door sensor is closed, or the infrared is blocked, the edge is marked as "blocked" or "high risk." This state change is recorded in the spatial state graph, and a state change event is generated for subsequent interaction. Simultaneously, this method triggers a local replanning mechanism by responding to state graph change events.

[0127] For example, when this method detects that the state of a certain side in the original path becomes impassable or the danger level increases, it writes the state change into the spatial state graph and generates a state change event at the same time; after responding to the state graph change event, the corresponding path calculation unit immediately calls the local replanning mechanism and performs local planning of the path based on the state change event.

[0128] The specific response mechanism involves locking the affected path segment between the start and destination nodes of the current path based on the state change identified by the state change event. Re-running the path search for this segment based on the weighted transit graph replaces only the affected path segment, while preserving the previous and next paths. This improves response efficiency and stability. This process is completed autonomously by the path computation unit, achieving a closed-loop linkage between edge state perception, recalculation triggering, and path updates.

[0129] This method outputs the calculated optimal emergency escape path in the form of a node sequence for display in subsequent practical applications.

[0130] Furthermore, this method also supports path validity verification for the calculated optimal emergency escape path, such as travel time rationality, boundary accessibility, etc., and writes the path that passes the validity verification into the cache for multi-terminal call.

[0131] Specifically, this method innovatively introduces a multi-dimensional dynamic verification mechanism based on a real-time state diagram. Compared to traditional methods that only statically determine path accessibility, this method combines real-time sensor data and spatial state diagrams to dynamically assess a path's travel time rationality, boundary accessibility, and future traversability. Travel time rationality verification not only considers path length but also occupant density, obstruction coefficient, and risk level, employing a dynamic traffic rate model to determine whether the path can be evacuated within the safe time limit. Boundary accessibility combines topological relationships and physical status within the BIM model, such as whether exit doors are open or passageways are closed, to ensure that the path endpoint is physically reachable. Furthermore, this method integrates trend prediction results from state updates and adaptive modules to proactively identify paths that may become high-risk or blocked, applying penalty weighting or replacement, thereby constructing a dynamic verification mechanism with forward-looking judgment capabilities. This mechanism significantly enhances the reliability and robustness of the path, a key technical advantage that distinguishes it from traditional path verification schemes, effectively improving safety and response efficiency in complex emergency scenarios.

[0132] This method builds a system response decision engine at this stage to ensure that the escape route remains feasible and timely under the dynamic changes of the disaster situation.

[0133] As a further optimization setting, this method can also add a stage S5 after stage S4. In this stage, the optimal emergency escape path calculated and determined in stage S4 is displayed and interactively processed. The optimal emergency escape path calculated and determined in stage S4 is displayed in a visual form in the BIM three-dimensional model, and personalized evacuation guidance information is pushed to users through multiple terminals.

[0134] Specifically, this stage is based on three-dimensional rendering engines such as Cesium, Unity or Three.js, and the path is superimposed and displayed in the form of linear tracks, flow arrows or light beams on the original BIM model.

[0135] Furthermore, this method supports segmented highlighting of the generated path, such as green for normal areas, yellow for congested areas, and red for dangerous detour areas, thereby improving user recognition efficiency.

[0136] On this basis, the Bezier curve interpolation method is used to smooth the path trajectory, thereby ensuring visual beauty and continuity.

[0137] Furthermore, for mobile terminals, this method pushes path information in GeoJSON format to apps, mini-programs, or wearable devices through WebSocket or REST interfaces, providing point-by-point navigation with turn reminders and risk warnings.

[0138] At the interactive level, users can view information such as current location, target exit, current risk level, etc. through the interface, and it supports functions such as voice broadcast and light linkage prompts to meet the needs of special groups such as the elderly and children.

[0139] For building managers, this method can provide multi-path diversion maps, personnel heat maps, and evacuation progress analysis panels for on-site command and logistics scheduling.

[0140] At this stage, this method uses spatial visualization and human-computer interaction technology to open up the terminal link of "path generation-user acquisition", thereby improving the practicality and operability of the entire system.

[0141] The BIM and IoT-based dynamic identification emergency escape route planning method presented in this example solution can, when applied, be implemented as a corresponding software program, forming a corresponding BIM and IoT-based dynamic identification emergency escape route planning system. When running, this software program will execute the BIM and IoT-based dynamic identification emergency escape route planning method and store the data in a corresponding storage medium for access and execution by a processor.

[0142] Combine Figure 2 As shown, the thus formed BIM and IoT dynamic identification emergency escape path planning system 100 is functionally mainly composed of a BIM three-dimensional space modeling unit 110, an IoT perception unit 120, a data fusion and hazard identification unit 130, a path calculation unit 140, and a path display and interaction unit 150 that cooperate with each other.

[0143] Among them, the BIM three-dimensional space modeling unit 110 in this system is set to construct a BIM model that reflects the building's geometric structure, topological relationship and semantic information, thereby providing a digital foundation for subsequent systems to perform data fusion mapping, path calculation and visualization.

[0144] This unit performs spatial segmentation and topology extraction on the corresponding spatial units in the constructed BIM model to form a directed graph structure. The nodes in the directed graph structure are configured as functional areas or spatial units, and the edges are configured as traffic paths; and the geometric information between each node is extracted.

[0145] On this basis, this unit also constructs a corresponding spatial mapping table between the spatial units in the BIM model and the networked sensors in the IoT perception unit.

[0146] The Internet of Things perception unit 120 in this system includes a number of networked sensors deployed inside the building, and a spatial mapping relationship is formed between the deployed number of networked sensors and the corresponding spatial units in the BIM model constructed by the BIM three-dimensional space modeling unit 110.

[0147] On this basis, multiple networked sensors are deployed to collect real-time multi-dimensional information about the building's internal environment, including environmental parameters, occupant location, and structural status, forming a real-time dynamic data source for the building's interior. Environmental parameters include temperature, humidity, smoke, carbon monoxide concentration, and light intensity. Structural status includes temporary obstacles or blocked passageways within the building.

[0148] The data fusion and hazard identification unit 130 in this system is configured to interact with the BIM three-dimensional space modeling unit 110 and the Internet of Things perception unit 120 for data, and is used to map the dynamic real-time data collected by the Internet of Things perception unit 120 to the corresponding space unit of the BIM model constructed by the BIM three-dimensional space modeling unit 110 to form a real-time space status diagram, and dynamically identify dangerous areas and obstacles based on rules or models.

[0149] This data fusion and hazard identification unit 130 coordinate maps and semantically binds the real-time data of various types of IoT sensors such as smoke, temperature, access control, and positioning deployed by the IoT perception unit 120 to the spatial units in the BIM model, and assigns real-time status information to each spatial unit, including labels such as risk level, whether it is passable, and whether it is congested. All information is written into the structured graph data to construct a real-time spatial status graph. This real-time spatial status graph not only contains static geometric topology, but also dynamically expresses the real-time status of each passage path and node, which will serve as the basis for obtaining effective path information in the subsequent emergency escape path calculation stage.

[0150] The path calculation unit 140 in this system is configured to interact with the data fusion and hazard identification unit 120 and the data fusion and hazard identification unit 130. It can automatically generate an optimal escape path that avoids high-risk areas based on a real-time spatial state diagram and in combination with an improved path search algorithm, and can also perform dynamic replanning.

[0151] The path calculation unit 140 integrates the real-time spatial status diagram generated by the data fusion and hazard identification unit 130 with the BIM spatial topology to form a weighted graph model, and adopts an improved A* path search algorithm that introduces risk factors, congestion coefficients, and traffic probabilities, and introduces a multi-factor path evaluation mechanism to generate an optimal escape path that avoids high-risk areas.

[0152] Specifically, in the weighted graph model formed in the path calculation unit 140, the weight of each path edge not only includes the distance, but also the risk factor, the personnel density factor and the obstacle penalty coefficient, and the low-risk passable path is searched preferentially by setting the heuristic function; at the same time, when multiple feasible paths exist, the optimal escape path is dynamically selected by comprehensively evaluating factors such as path continuity, replanning cost and travel time.

[0153] The optimal escape route generated in this way takes into account path safety, real-time performance and flexibility, significantly improving emergency escape response capabilities.

[0154] The path display and interaction unit 150 in this system is configured to interact with the BIM three-dimensional space modeling unit 110 and the path calculation unit 140 to display the path calculation results of the path calculation unit 140 in a visual form in the BIM model, and can also push personalized evacuation guidance information through multiple terminals.

[0155] The following further explains the specific structure of each functional unit in this system.

[0156] See also Figure 3 The BIM three-dimensional space modeling unit 110 in this system is specifically composed of a BIM three-dimensional space modeling module 111, a directed graph generation module 112, and a mapping module 113 that cooperate with each other.

[0157] The BIM three-dimensional space modeling module 111 in this unit can establish a complete building structure model based on BIM modeling platforms such as Revit and ArchiCAD, and can export the constructed model in IFC or custom format for subsequent system integration.

[0158] Furthermore, the BIM model constructed by this BIM three-dimensional space modeling module 111 includes components such as walls, floor slabs, doors and windows, evacuation exits, stairs, elevators, and evacuation signs, and each component is assigned a unique code (i.e., ID) and semantic attributes. The BIM model constructed in this way accurately reflects the building's geometric structure, topological relationship, and semantic information.

[0159] The description of semantic attributes and topological relationships has been given above and will not be repeated here.

[0160] The directed graph generation module 112 in this unit is configured to interact with the BIM three-dimensional space modeling module 111 for data, and can perform spatial segmentation and topology extraction on the rooms, passages, floors and other spatial areas (i.e., spatial units) in the BIM model constructed by the BIM three-dimensional space modeling module 111 to form a directed graph structure. The nodes in the directed graph structure are configured as functional areas or other corresponding spatial units in the BIM model, and the edges in the directed graph structure are configured as passage paths in the BIM model.

[0161] On this basis, the directed graph generation module 112 can also synchronously extract geometric space information such as geometric distance between nodes, floor height difference, passage width, etc. during the construction process of the BIM model, as basic parameters for subsequent escape path calculation.

[0162] The mapping module 113 in this unit is configured to perform data interaction with the BIM three-dimensional space modeling module 111, the directed graph generation module 112, and the Internet of Things perception unit 120, and can preset a binding relationship between each spatial unit in the BIM model and the networked sensors deployed on-site in the building, thereby forming a spatial mapping table between the spatial units in the BIM model and the networked sensors deployed on-site in the building.

[0163] For example, the spatial unit in the BIM model includes the corresponding spatial area, and the corresponding spatial area includes the corresponding components, thereby a mapping table of "room ID-component ID-sensor ID" can be preset, thereby enabling spatial mapping between the BIM model and the IoT sensor.

[0164] On this basis, an engine import module 114 can be further set in this unit. The engine import module 114 interacts with the BIM three-dimensional space modeling module 111, the directed graph generation module 112, and the mapping module 113. It can import the completed BIM model into the Cesium, Unity or WebGL engine, and realize lightweight three-dimensional display through glTF or a custom format, which facilitates subsequent path rendering and human-computer interaction.

[0165] On this basis, an interface encapsulation module 115 can be further set in this unit. The interface encapsulation module 115 interacts with the BIM three-dimensional space modeling module 111, the directed graph generation module 112, and the mapping module 113. It can encapsulate the BIM model interface into a RESTful API for the constructed BIM model, and provide services such as component attribute query, channel connectivity retrieval, and evacuation path query.

[0166] See also Figure 4 The IoT perception unit 120 in this system is specifically composed of a number of multi-type networked sensors 121, an edge computing gateway 122 and a perception data processing module 123.

[0167] Among them, several multi-type networked sensors 121 are deployed in corresponding spatial units inside the building, which are used for real-time collection of multi-dimensional information such as temperature, humidity, smoke concentration, carbon monoxide concentration, light intensity, personnel positioning and structural status in the corresponding spatial units, forming a dynamic data input source.

[0168] The multi-type networked sensors 121 here include temperature and humidity sensors, smoke sensors, UWB or Bluetooth AOA positioning base stations, laser radars and infrared sensors, etc. It is understandable that the multi-type networked sensors 121 in this solution are not limited to these, and any other feasible sensors can be used as needed.

[0169] For example, in specific deployments, temperature and humidity sensors and smoke sensors should be installed in high-risk areas such as corridors, computer rooms, and storage areas within a building;

[0170] UWB or Bluetooth AOA positioning base stations are deployed in public areas, entrances and exits, and evacuation passages within the building to achieve accurate positioning of personnel;

[0171] LiDAR and infrared sensors are used to identify temporary obstacles or channel blockages.

[0172] The edge computing gateway 122 in this unit interacts with the networked sensors 121 for edge data processing of the networked sensors deployed on site.

[0173] The edge computing gateway 122 is deployed at or near the building site, and all sensors deployed at the site are connected to the edge computing gateway via wired RS485 or wireless LoRa, Wi-Fi, NB-IoT, etc.

[0174] The edge computing gateway 122 can perform preliminary data filtering and anomaly identification on the collected data uploaded by networked sensors, and cache the processed data; on this basis, the edge computing gateway 122 further encapsulates the processed collected data through the MQTT protocol, and uniformly uploads it to the back-end system for subsequent processing.

[0175] Furthermore, the edge computing gateway 122 encapsulates the perception data collected by the networked sensors in a unified structure at the front end, including fields such as device number, data type, collection time, spatial coordinates, numerical value and status code, thereby ensuring the accuracy and reliability of data transmission and improving the efficiency of subsequent data fusion processing.

[0176] The perception data processing module 123 in this unit interacts with the edge computing gateway 122 and the BIM three-dimensional space modeling unit 110 to establish a mapping relationship table between the deployment coordinates of the networked sensors deployed in the building and the component ID in the BIM model, and adopts three-dimensional coordinate matching or topological alignment technology to ensure that the perception data is accurately mapped to the corresponding spatial unit.

[0177] Specifically, the perception data processing module 123 pre-establishes a set of "networked sensor deployment coordinates-component ID" mapping relationship tables based on the unique geometric boundaries and component IDs of all spatial units (such as rooms, corridors, and stairwells) in the BIM model.

[0178] On this basis, when the networked sensors 121 are deployed, the three-dimensional installation coordinates of each sensor are recorded in the perception data processing module 123 and the coordinates are uniformly converted into a building coordinate system consistent with the BIM model;

[0179] Furthermore, the perception data processing module 123 performs a spatial containment judgment on the spatial position of the sensor and the geometric boundaries of the components or spatial units in the BIM model based on the three-dimensional coordinate matching method, and uses a three-dimensional spatial containment judgment algorithm to compare the spatial position of each sensor with the room boundary, component volume or channel geometry in the BIM model. Through spatial envelope operations (such as point-in-polyhedron algorithm), it is determined whether the sensor is inside a specific spatial unit, and then the spatial unit ID or component ID to which it belongs is determined; if multiple spaces are adjacent or overlapping, the closest distance priority or weight rule is used to determine the unique mapping relationship.

[0180] For sensors deployed in continuous spaces such as passages and stairs, topological alignment technology is used to map their positions to adjacent nodes or edges in the spatial topology map generated by the BIM model, ensuring that the data is bound to the key structural units of the spatial passage path to ensure that dynamic state changes can be accurately reflected during path calculation.

[0181] This perception data processing module 123 will eventually establish an association relationship table of "sensor ID-coordinate-component ID-spatial unit ID" to achieve accurate one-to-one mapping of perception data and spatial semantic units, providing a data basis for subsequent spatial state assignment, hazard identification and path planning, and ensuring the spatial accuracy and real-time performance of subsequent hazard identification and path generation.

[0182] Furthermore, the sensor data processing module 123 can further set corresponding data sampling and uploading frequencies for the deployed networked sensors to ensure real-time data. The specific setting scheme is not limited here and can be determined according to actual needs.

[0183] See also Figure 5 The data fusion and hazard identification unit 130 in this system is specifically composed of a data analysis module 131, a data fusion module 132 and a hazard identification module 133.

[0184] Among them, the data analysis module 131 in this unit is configured to interact with the Internet of Things perception unit 120 for data, and can receive dynamic real-time perception data streams with spatial coding from the deployed networked sensors and edge computing gateways; on this basis, the spatial unit in the BIM model corresponding to the collected dynamic real-time data is located according to the mapping relationship table between the deployment coordinates of the networked sensors and the component ID in the BIM model, that is, the spatial unit to which the dynamic real-time perception data stream belongs, such as a room, channel, node, etc.

[0185] The spatial coding scheme has been described above and will not be elaborated here.

[0186] The data fusion module 132 in this unit interacts with the data analysis module 131 to perform weighted averaging, maximum value extraction, or time trend analysis on the values ​​collected by multiple networked sensors corresponding to each spatial unit to obtain a comprehensive status evaluation value of the space.

[0187] The hazard identification module 133 in this unit is configured with a risk judgment rule library, in which a variety of risk judgment rules are preset. The risk judgment rule library and the settings of the risk judgment rules therein are as mentioned above and will not be elaborated here.

[0188] On this basis, the data fusion module 132 in this unit interacts with the hazard identification module 133 to dynamically identify dangerous areas and obstacles through a preset risk judgment rule library and semantic threshold model after calculating the comprehensive status assessment value of each spatial unit, thereby realizing hazard identification of the corresponding spatial units in the building and writing the identification results into the spatial status diagram in real time.

[0189] In this unit, the organic cooperation among the data analysis module 131, the data fusion module 132 and the hazard identification module 133 constitutes a logical chain of hazard identification, ensuring that the system can quickly and accurately identify risk areas and update the spatial status map based on multi-source perception results.

[0190] As a further example, the multiple risk judgment rules here include space unit risk judgment rules, space unit obstacle identification rules, and space unit personnel positioning rules.

[0191] The spatial unit risk assessment rules pre-set thresholds for various environmental parameters within each spatial unit, and set corresponding risk levels based on different environmental parameter threshold conditions. This is achieved by using environmental parameter sensors deployed in the corresponding spatial units within the building to obtain the corresponding environmental parameters of each spatial unit, and then determining the risk level of each spatial unit based on these parameters according to the assessment rules.

[0192] For example, if the smoke concentration collected on site in a certain spatial unit exceeds the set threshold, it will be marked as a "suspicious fire area"; if the temperature collected on site in a certain spatial unit exceeds 60°C, the spatial unit will be marked as a "high temperature danger zone"; if the smoke concentration collected on site in a certain spatial unit exceeds the set threshold and the temperature collected on site exceeds 60°C, that is, if multiple judgment conditions are met at the same time, the spatial unit will be marked as a "high-risk closed area".

[0193] For the spatial unit obstacle recognition rule, the obstacle status in the spatial unit is set according to the width value of the channel in the spatial unit detected by the sensors deployed on site.

[0194] For example, if the infrared radio signal deployed at the site of the space unit in the building is continuously interrupted or the radar detects that the channel width in the current space unit is less than 0.8 meters, the space unit area will be marked as a "temporary blocking area".

[0195] For the spatial unit personnel positioning rules, the personnel congestion status is determined based on the density of positioning data and the residence time.

[0196] For example, if the personnel positioning data in a certain spatial unit is dense and the stay time is long, the current spatial unit can be identified as a "personnel congestion area".

[0197] On this basis, the hazard identification module 133 further writes the above identification results into a unified spatial state diagram, which can be called during subsequent path calculation to achieve risk avoidance, obstacle avoidance and dynamic replanning of the path, thereby ensuring the system's rapid response and accurate modeling of dangerous areas and traffic obstacles in emergencies.

[0198] The implementation of the spatial state graph and the specific fusion scheme between the recognition results and the spatial state graph are not described here in detail. Please refer to the aforementioned scheme.

[0199] On this basis, a state update and adaptation module 134 can be further set up in this unit. The state update and adaptation module 134 interacts with the data analysis module 131, the data fusion module 132 and the hazard identification module 133, and dynamically maintains and intelligently predicts and adjusts the spatial state diagram through mechanisms including sliding time window judgment, data trend extrapolation, and pre-closure warning, thereby improving the system's response capability to sudden risk changes.

[0200] Specifically, when processing data, the state update and adaptation module 134, using a sliding time window judgment mechanism, continuously monitors changes in various sensor data sequences and determines risk levels based on these changes. For example, if the temperature or smoke concentration in a certain spatial unit exceeds a threshold multiple times within a set timeframe, the risk is determined to be persistent rather than transient, and the risk level is promptly raised.

[0201] When introducing the trend extrapolation mechanism, the state update and adaptation module 134 will analyze the trend of historical data changes (such as the slope of smoke concentration and the growth rate of carbon monoxide concentration) to predict its state trend in the short term in the future and mark in advance the area that is about to enter a high-risk area.

[0202] When the state update and adaptation module 134 introduces a preemptive closure warning, if the system predicts that a certain passage or exit will become impassable within tens of seconds, it sends a warning signal to the path calculation module in advance, activating an alternative route or performing local replanning. This module achieves a shift from "real-time monitoring" to "preemptive intervention," ensuring the system's ability to proactively identify risks and dynamically adjust routes.

[0203] For example, if the state update and adaptation module 134 identifies that the temperature in a certain area rises continuously within 3 minutes and the smoke increases synchronously, it will issue an early warning and increase its risk level, driving the path calculation unit 140 to recalculate.

[0204] See also Figure 6 The path calculation unit 140 in this system is mainly composed of a weighted traffic graph construction module 141, a path search calculation module 142 and a path output module 143.

[0205] The weighted access graph construction module 141 in this unit calls the fused BIM topology graph and the spatial status graph to construct a weighted access graph, in which the nodes are configured to represent spatial units or key channel intersections, and the edges are configured to represent access paths.

[0206] When constructing the weighted access graph, the weighted access graph construction module 141 converts each spatial node and the access edges between them into a graph structure based on the fused BIM topology graph and the spatial state graph, and the weight value W of the edge in the graph is calculated using a multi-factor weighted model.

[0207] Furthermore, for the edges in the weighted traffic graph, the edge weights are calculated by combining geometric distance, risk level, obstacle resistance, and personnel density. The calculation formula is as follows:

[0208] W = D × (1 + αR + βC + γB);

[0209] Where D is the physical distance of the edge, R is the risk coefficient (derived from sensor judgment), C is the personnel congestion density coefficient, B is the obstacle judgment result, and α, β, and γ are adjustment weights.

[0210] This weighted mechanism not only considers the shortest distance, but also integrates risk avoidance, obstacle avoidance and traffic efficiency to form a dynamically adjustable disaster response traffic map.

[0211] The path search calculation module 142 in this unit interacts with the weighted pass graph construction module 141. After obtaining the weighted pass graph, the path search calculation module 142 uses the improved A* algorithm or the dynamic Dijkstra algorithm as the path search algorithm, and calculates the optimal emergency escape path based on the weighted pass graph according to this algorithm.

[0212] Furthermore, the path search calculation module 142 is also configured with a heuristic function. The path search calculation module 142 can combine the heuristic function to accelerate the optimization when calculating the optimal emergency escape path, and support target switching and multi-target redundant path generation (such as backup exits).

[0213] Specifically, this path search calculation module 142 introduces a multi-dimensional heuristic function to the improved A* algorithm. Based on the original distance estimation, it incorporates a risk penalty function, congestion weights, and obstacle cost factors, making the path search more inclined to select low-risk, low-resistance areas. The dynamic Dijkstra algorithm supports rapid replanning in the event of path interruptions, avoiding full map recalculation, improving algorithm execution efficiency and response speed, and enhancing the system's real-time adaptability in emergency situations.

[0214] Furthermore, this path search calculation module 142 can also further introduce a target switching mechanism and a multi-target redundant path generation strategy on the basis of introducing the heuristic function, perform target switching and multi-target redundant path generation, realize more intelligent escape path planning, and improve the efficiency and reliability of path calculation.

[0215] As a further illustration, this path search calculation module 142 first uses a heuristic function based on a weighted combination of the estimated cost (e.g., Euclidean distance or floor distance) from the current node to the target exit and the risk assessment value to guide the A* algorithm to prioritize nodes that are more likely to become the optimal path during the search process, significantly reducing the ineffective search area and accelerating the optimization process. Next, when constructing the weighted access graph, it allows multiple exit nodes to be set as reachable targets. If the current target exit is detected to be impassable or the risk level increases during the path planning process, the system automatically switches targets and selects a suboptimal exit to recalculate the path, ensuring escape continuity and fault tolerance. Finally, under normal conditions, the system can also calculate the primary path and one or two backup paths in parallel, forming a multi-target path set based on different exits or avoidance strategies, and monitor the traffic status of each path in real time. If the primary path is blocked, it can immediately switch to the backup path without re-searching the entire path. This strategy presets multiple path options during the path calculation phase and enables rapid switching during the execution phase, significantly enhancing the system's robustness and response efficiency to dynamic emergencies.

[0216] Furthermore, the path search and calculation module 142 is also configured with a local replanning mechanism. Based on this local replanning mechanism, the path search and calculation module 142 can immediately trigger the local replanning mechanism when the system detects that an edge in the path becomes blocked or high-risk, thereby avoiding recalculation of the entire map and improving response speed.

[0217] Specifically, the path search calculation module 142 is configured to cooperate with the data fusion and hazard identification unit 130. The data fusion and hazard identification unit 130 here is configured to continuously monitor the data streams from various networked sensors, and determine whether the state of the path edge has changed through preset rules. For example, after an event such as a sharp rise in smoke concentration in the channel, temperature exceeding the threshold, door magnet closing, or infrared being blocked, the edge is marked as "blocked" or "high risk", and the state change is written into the spatial state diagram, and a state change event is generated for subsequent linkage; the path search calculation module 142 is configured to subscribe to the change event of the spatial state diagram.

[0218] In this way, after the data fusion and hazard identification unit 130 generates a state change event, the path search calculation module 142 will trigger the configured local replanning mechanism in response to the state change event generated by the data fusion and hazard identification unit 130, and immediately call the replanning logic according to the state change corresponding to the state change event.

[0219] Furthermore, this path search and calculation module 142, using a local replanning mechanism, first locates the affected path segment between the start and destination nodes in the current path. It then re-executes a path search for this segment based on the weighted transit graph, replacing only the affected path segment while maintaining the previous and subsequent paths unchanged, thereby improving response efficiency and stability. This enables a closed-loop linkage between edge state awareness, recalculation triggering, and path updating.

[0220] The path output module 143 in this unit interacts with the path search and calculation module 142 to output the optimal emergency escape path calculated and determined by the path search and calculation module 142 in the form of a node sequence for the display module to call.

[0221] On this basis, a path validity verification module 144 can be further set in this unit. The path validity verification module 144 interacts with the path search calculation module 142 data, and can verify the validity of the optimal emergency escape path calculated and determined by the path search calculation module 142, such as the rationality of the travel time, boundary accessibility, etc., and write the verified path into the cache for multi-terminal calls.

[0222] Furthermore, the validity of the optimal emergency escape path is verified in the path validity verification module 144 by setting up a multi-dimensional dynamic verification mechanism based on a real-time state diagram. When verifying the validity of the optimal emergency escape path based on the multi-dimensional dynamic verification mechanism, the path validity verification module 144 combines real-time sensor data and spatial state diagrams to dynamically evaluate the rationality of the path's travel time, boundary accessibility, and future passability. Among them, the verification of the rationality of the travel time is not only based on the length of the path, but also takes into account the density of people, obstacle coefficient, and risk level, and uses a dynamic travel rate model to determine whether the path can be evacuated within the safe time limit; the boundary accessibility is combined with the topological relationship and physical status in the BIM model, such as whether the exit door is open, whether the passage is closed, etc., to ensure that the end of the path is physically reachable.

[0223] Furthermore, this path validation module 144 also incorporates a dynamic validation mechanism. This mechanism integrates state updates with the trend prediction results of the adaptive module to proactively identify high-risk or blocked edges in the path and perform weighted penalty or replacement processing, thereby achieving dynamic validation with forward-looking judgment capabilities. This dynamic validation mechanism significantly enhances the reliability and robustness of the path, completely contradicting the static nature of traditional path validation schemes that only determine path reachability. This effectively improves safety and response efficiency in complex emergency scenarios.

[0224] See also Figure 7 The path display and interaction unit 150 in this system is specifically composed of a path display module 151 and an interaction module 152.

[0225] Among them, the path display module 151 is based on a three-dimensional rendering engine such as Cesium, Unity or Three.js, and displays the path in the form of linear tracks, flowing arrows or light beams on the basis of the original BIM model.

[0226] Furthermore, the path display module 151 supports highlighting of path segments for the generated path, such as green for normal areas, yellow for congested areas, red for dangerous detour areas, etc., thereby improving user recognition efficiency.

[0227] On this basis, the path display module 151 preferably uses the Bezier curve interpolation method to smooth the path trajectory, thereby ensuring visual beauty and continuity.

[0228] Furthermore, the path display module 151 pushes path information in GeoJSON format to apps, mini-programs, or wearable devices via WebSocket or REST interfaces for mobile terminals, providing point-by-point navigation with turn reminders and risk warnings.

[0229] The interactive module 152 in this unit forms a human-computer interaction interface, allowing users to view information such as current location, target exit, current risk level, etc. through the interface, and supports functions such as voice broadcast and light linkage prompts to meet the needs of special groups such as the elderly and children.

[0230] Furthermore, the interactive module 152 can also provide a multi-path diversion diagram, a personnel heat map, and an evacuation progress analysis panel for on-site command and logistics scheduling.

[0231] The BIM and IoT-based dynamic identification emergency escape path planning system, developed based on the aforementioned solution, constructs a three-dimensional BIM model of the building and deploys smoke, temperature, and positioning sensors on-site to collect real-time information on the building's internal environment and occupant distribution. This data fusion and hazard identification unit then maps the sensory data into the BIM three-dimensional model space, enabling dynamic identification of hazardous areas such as fire sources, smoke, and obstacles. The path calculation unit, combined with an improved path search algorithm, automatically generates an optimal escape path that avoids high-risk areas. This is then visualized through a three-dimensional model and simultaneously pushed to user terminals and evacuation indicator devices. The system supports dynamic path updates and emergency dispatch assistance, improving the efficiency and safety of personnel evacuations in emergencies.

[0232] Based on the above-mentioned emergency escape path planning scheme based on dynamic identification of BIM and the Internet of Things, an embodiment of the present invention also provides a computer-readable storage medium on which a program is stored. When the program is executed by a processor, the steps of the above-mentioned method for dynamic identification of emergency escape path planning based on BIM and the Internet of Things are implemented.

[0233] An embodiment of the present invention also provides a processor, which is used to run a program, wherein when the program is running, the steps of the above-mentioned method for dynamic identification of emergency escape path planning based on BIM and the Internet of Things are executed.

[0234] An embodiment of the present invention also provides a terminal device, which includes a processor, a memory, and a program stored in the memory and runnable on the processor. The program code is loaded and executed by the processor to implement the steps of the above-mentioned method for dynamic identification of emergency escape path planning based on BIM and the Internet of Things.

[0235] The present invention also provides a computer program product, which, when executed on a data processing device, is suitable for executing the steps of the above-mentioned method for dynamic identification of emergency escape path planning based on BIM and the Internet of Things.

[0236] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0237] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0238] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0239] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A dynamic identification emergency escape route planning system based on BIM and the Internet of Things, characterized by: The system comprises: A BIM three-dimensional spatial modeling unit, wherein the BIM three-dimensional spatial modeling unit is configured to construct a three-dimensional building information model that reflects the building's geometric structure, topological relationship, and semantic information. The three-dimensional building information model performs spatial segmentation and topological extraction on corresponding spatial units to form a directed graph structure. The nodes in the directed graph structure are configured as functional areas or spatial units, and the edges are configured as traffic paths. The geometric information between each node is extracted. The BIM three-dimensional spatial modeling unit also constructs a spatial mapping table between the spatial units in the three-dimensional building information model and the networked sensors in the Internet of Things perception unit. An IoT sensing unit, which collects multi-dimensional information about the building's interior, including environmental parameters, occupant location, and structural status, in real time through a number of networked sensors deployed inside the building, thereby forming a real-time dynamic data source within the building; A data fusion and hazard identification unit configured to interact with the BIM three-dimensional spatial modeling unit and the Internet of Things perception unit to map the dynamic real-time data collected by the Internet of Things perception unit to the corresponding spatial unit constructed by the BIM three-dimensional spatial modeling unit to generate a real-time spatial state diagram, and to dynamically identify dangerous areas and obstacles based on rules or models; A path calculation unit is configured to interact with the data fusion and hazard identification unit to construct a multidimensional path evaluation function based on the real-time spatial state diagram dynamically generated in the data fusion and hazard identification unit, thereby automatically generating an optimal escape path that avoids high-risk areas and enabling dynamic replanning.

2. The BIM and IoT-based dynamic identification emergency escape route planning system according to claim 1 is characterized in that: The IoT perception unit establishes a mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the three-dimensional building information model for the deployed networked sensors, and maps the perception data of the networked sensors to the corresponding spatial units in the three-dimensional building information model through three-dimensional coordinate matching or topological alignment.

3. The BIM and IoT-based dynamic identification emergency escape route planning system according to claim 1 is characterized in that: The data fusion and hazard identification unit first locates the spatial units in the three-dimensional building information model corresponding to the dynamic real-time data collected by the Internet of Things sensing unit based on a mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the three-dimensional building information model; then, weighted averaging, maximum value extraction and / or time trend analysis are performed on the values ​​collected by the networked sensors corresponding to each spatial unit to obtain a comprehensive status evaluation value for each spatial unit, thereby generating a real-time spatial status diagram.

4. The BIM and IoT-based dynamic identification emergency escape route planning system according to claim 1 is characterized in that: When calculating the optimal escape path, the path calculation unit first calls the fused BIM topology map and the real-time spatial state map to construct a weighted access map; then the constructed multidimensional path evaluation function calculates the optimal emergency escape path based on the weighted access map.

5. The BIM and IoT-based dynamic identification emergency escape route planning system according to claim 1 is characterized in that: The emergency escape path planning system also includes a path display and interaction unit, which is configured to interact with the BIM three-dimensional space modeling unit and the path calculation unit data, and is used to display the path calculation results of the path calculation unit in a visual form in the three-dimensional building information model, and can also push personalized evacuation guidance information through multiple terminals.

6. A method for dynamic identification of emergency escape path planning based on BIM and Internet of Things, characterized in that: The method comprises: A three-dimensional building information model is constructed that can reflect the building's geometric structure, topological relationship, and semantic information. Spatial segmentation and topological extraction are performed on corresponding spatial units in the constructed three-dimensional building information model to form a directed graph structure. Nodes in the directed graph structure are configured as functional areas or spatial units, and edges are configured as passage paths. Geometric information between each node is extracted. A spatial mapping table between the spatial units and networked sensors deployed in the building is also constructed for the corresponding spatial units in the constructed three-dimensional building information model. Deploy several networked sensors inside the building to collect real-time multi-dimensional information including environmental parameters, occupant location, and structural status, forming a real-time dynamic data source inside the building; Mapping the collected dynamic real-time data to the real-time spatial status diagram generated in the corresponding spatial unit in the constructed 3D building information model, and dynamically identifying dangerous areas and obstacles based on rules or models; A multi-dimensional path evaluation function is constructed based on the dynamically generated real-time spatial state diagram, which automatically generates the optimal escape path avoiding high-risk areas and can perform dynamic replanning.

7. The method for planning emergency escape routes based on BIM and Internet of Things dynamic identification according to claim 6 is characterized in that: The method establishes a mapping relationship table between the deployment coordinates of the networked sensors and the component IDs in the three-dimensional building information model for the deployed networked sensors, and maps the perception data of the networked sensors to the corresponding spatial units in the three-dimensional building information model through three-dimensional coordinate matching or topological alignment.

8. The method for planning emergency escape routes based on BIM and Internet of Things dynamic identification according to claim 6 is characterized in that: The method, based on dynamic real-time data collected by IoT sensing units, first locates the spatial units in the 3D building information model corresponding to the collected dynamic real-time data based on a mapping table between the networked sensor deployment coordinates and component IDs in the 3D building information model. Next, weighted averaging, maximum value extraction, and / or temporal trend analysis are performed on the values ​​collected by the networked sensors corresponding to each spatial unit to obtain a comprehensive status assessment value for each spatial unit, thereby generating a real-time spatial status diagram.

9. The method for planning emergency escape routes based on BIM and Internet of Things dynamic identification according to claim 6 is characterized in that: When calculating the optimal escape path, the method first calls the fused BIM topology map and the real-time spatial state map to construct a weighted access map; then the constructed multidimensional path evaluation function calculates the optimal emergency escape path based on the weighted access map.

10. The method for planning emergency escape routes based on dynamic identification of BIM and Internet of Things according to claim 6, characterized in that: The method also includes a path display and interaction step, in which the path calculation results of the path calculation unit can be displayed in a visual form in a three-dimensional building information model, and personalized evacuation guidance information can be pushed through multiple terminals.

Images