Emergency rescue command system based on building safety requirements

Abstract

The application discloses an emergency rescue command system based on building safety requirements, comprising data acquisition, risk analysis, situation visualization, scheme generation, deduction optimization and instruction issuing modules. The system collects building deformation, vibration and environmental data through the deployment of sensors, identifies risk points and levels through feature extraction, maps risk information with a three-dimensional building model space, generates a dynamic visual safety situation layer, generates a preliminary scheme according to the layer matching emergency plan, integrates real-time rescue resource information for reachability and timeliness deduction, outputs an optimized command task sequence, and finally decomposes into specific instructions and issues to the rescue terminal. The scheme realizes three-dimensional intuitive presentation of risk situation and dynamic intelligent optimization of emergency command, improves the risk analysis efficiency and the scientificity of rescue resource scheduling.

Description

Technical Field

[0001] This invention relates to the field of building safety and emergency command technology, specifically to an emergency rescue command system based on building safety requirements. Background Technology

[0002] In the field of building safety monitoring, existing technologies mainly rely on deploying sensor networks to continuously collect parameters such as structural deformation and vibration, and then setting thresholds to trigger alarms. Alarm information is usually presented in the form of data lists, floor plans, or separate charts. This method of information presentation is disconnected from the actual three-dimensional spatial structure of the building, making it difficult to intuitively and quickly understand the specific spatial location, distribution, and severity of risk points. Command personnel need to perform complex mental conversions between abstract alarm data and the physical building, delaying the assessment of the overall spatial pattern of the hazard.

[0003] In the emergency response phase, existing solutions largely rely on pre-defined static emergency plan libraries. Systems or personnel invoke corresponding text-based plans based on alarm types. However, these plans fail to incorporate real-time information such as the location and availability of emergency resources, as well as on-site traffic conditions. The generation and execution of these plans lack a dynamic calculation and process simulation step based on real-time resource constraints, which may lead to unreasonable resource scheduling paths, conflicting or wasted task sequences for multiple risk points, ultimately impacting overall response efficiency.

[0004] There is a need for a technology that can transform discrete risk alarm information into an intuitive situation that is precisely bound to the three-dimensional space of a building, and can dynamically extrapolate and optimize based on pre-planned scenarios and real-time resource information to generate a command plan that can be executed efficiently. Summary of the Invention

[0005] This invention aims to solve at least one of the technical problems existing in the prior art;

[0006] Therefore, this invention proposes an emergency rescue command system based on building safety requirements, comprising:

[0007] The data acquisition module collects real-time building deformation data, vibration data, and ambient gas concentration data through a sensor array deployed at key nodes of the target building structure, forming a basic data stream of building safety status.

[0008] The risk analysis module performs feature extraction and abnormal pattern recognition on the basic data stream of building safety status to locate potential or existing building structural risk points and their risk levels.

[0009] The situation visualization module loads a 3D model of the target building based on a geographic information system, maps the risk points and their risk levels to the corresponding spatial locations of the 3D model, and generates a 3D visualization layer of building safety situation with risk labels.

[0010] The scheme generation module automatically matches and generates targeted preliminary emergency rescue action plans based on the three-dimensional visualization layer of the building safety situation and in combination with the preset emergency plan library.

[0011] The simulation and optimization module integrates the status and location information of available emergency rescue resources, performs resource accessibility and timeliness simulations on the preliminary emergency rescue action plan, and generates an optimized emergency rescue command task sequence.

[0012] The instruction issuing module decomposes the emergency rescue command task sequence into specific navigation and operation instructions, and issues them to the corresponding rescue robots and personnel smart terminals to guide them to designated risk points to perform tasks.

[0013] Preferably, the step of performing feature extraction and anomaly pattern recognition on the basic data stream of building safety status to locate potential or existing building structural risk points and their risk levels specifically involves:

[0014] The raw building deformation data collected by the sensor array is filtered and denoised to extract the displacement time-series curves of key structural points and calculate the displacement velocity and acceleration variation characteristics.

[0015] Spectral analysis is performed on the raw vibration data collected by the sensor array to extract the structural principal frequency, amplitude and damping ratio characteristics, and to identify abnormal vibration modes.

[0016] Trend analysis is performed on the ambient gas concentration data collected by the sensor array to identify abnormal increases in the concentration of flammable, toxic, or asphyxiating gases.

[0017] The displacement velocity and acceleration change characteristics, abnormal vibration mode characteristics, and abnormal gas concentration increase characteristics are input into a pre-trained risk assessment classification model.

[0018] The risk assessment classification model outputs a structural risk probability value and a risk type label for each sensor monitoring point location.

[0019] The monitoring points are sorted according to their risk probability values, and those with risk probability values ​​exceeding a preset threshold are selected as candidate risk points.

[0020] Based on the aforementioned risk type labels, each candidate risk point is assigned a quantitative risk level, which comprehensively considers the probability of the risk occurring and the expected severity of the consequences.

[0021] The spatial coordinates, risk types, and risk levels of all candidate risk points are packaged to form a risk point location and level list.

[0022] Preferably, the step of loading the 3D model of the target building based on the geographic information system, mapping the risk points and their risk levels to the corresponding spatial locations of the 3D model, and generating a 3D visualization layer of the building safety status with risk labels, specifically involves:

[0023] Calling a 3D model of the target building with structural component information from a building information model or reconstructing it from survey data, and importing it into a geographic information system platform;

[0024] Obtain the spatial coordinate information of each candidate risk point in the risk point location and level list;

[0025] In the geographic information system platform, the spatial coordinates of each candidate risk point are matched with the corresponding structural components or spatial locations in the three-dimensional model;

[0026] Based on the matching results, risk points are visually marked on the corresponding positions of the three-dimensional model using three-dimensional icons of different colors, shapes, or sizes, wherein the visual attributes of the icons are associated with the risk level.

[0027] Next to the risk point annotations in the 3D model, the risk type label and the specific risk probability value are displayed in a floating information box.

[0028] Render the annotated 3D model into an interactive 3D visualization scene;

[0029] Overlay the physical location distribution map of the sensor array, the internal passage network layer of the building, and the location layer of key facilities;

[0030] The interactive 3D visualization scene and all overlay layers are integrated and packaged into a 3D visualization layer data package for building safety status, which is then called and displayed by the command platform.

[0031] Preferably, the step of automatically matching and generating a targeted preliminary emergency response plan based on the three-dimensional visualization layer of the building safety situation and a preset emergency plan library is as follows:

[0032] Analyze the risk point information marked in the three-dimensional visualization layer of the building safety situation, and extract the dominant risk type, the highest risk level, and the spatial distribution characteristics of the risk points;

[0033] Using the dominant risk type and highest risk level as query keywords, a full-text search and matching degree calculation are performed in the preset emergency plan database;

[0034] The emergency response plan database contains standard handling procedures, required resource lists, and safety precautions for dealing with different types of building risks.

[0035] Retrieve and filter out several candidate emergency response plan entries whose matching degree exceeds the preset standard;

[0036] Analyze the spatial distribution characteristics of the risk points. If the risk points are concentrated, they are determined to be single-point or local events; if the risk points are dispersed, they are determined to be multi-point or systemic events.

[0037] Based on the determination of the event type, a basic task framework is generated by selecting or combining candidate emergency response plan items. The task framework includes four stages: isolation, assessment, handling, and verification.

[0038] The information of the passage network layer and key facility location layer superimposed on the three-dimensional visualization layer of the building safety situation is integrated into the basic task framework to plan the approximate approach route and operation points of the emergency rescue force.

[0039] Based on the initial status information of available emergency rescue resources, task execution entities are assigned to each stage in the basic task framework, forming a preliminary emergency rescue action plan draft that includes descriptions of task stages, action routes, work locations, and execution entities.

[0040] The draft preliminary emergency rescue action plan is subjected to logical consistency verification and rough time estimation, and an optimized preliminary emergency rescue action plan is output.

[0041] Preferably, the integration of the status and location information of available emergency rescue resources, and the simulation of resource accessibility and timeliness of the preliminary emergency rescue action plan, generate an optimized emergency rescue command task sequence, specifically as follows:

[0042] A list of available emergency rescue resources is acquired and maintained in real time, which includes the current status, real-time location, functional capabilities and battery life information of emergency rescue robots, specialized engineering vehicles, rescue personnel teams and special equipment;

[0043] The planned work locations in the preliminary emergency rescue action plan are compared with the real-time locations in the list of available rescue resources;

[0044] Based on the channel network layer in the three-dimensional visualization layer of the building safety situation, considering the passage restrictions, damage and congestion of the channels, calculate the theoretical shortest path and estimated travel time for each available emergency rescue resource to reach the designated work point from its current location.

[0045] Based on the estimated travel time and the resource's remaining capacity, assess whether the resource has the ability to be put into operation immediately upon arrival, and screen out resources that do not meet the conditions.

[0046] For the same work site, if multiple available emergency rescue resources meet the criteria, they will be prioritized based on their functional capabilities and the degree of matching with the task, as well as the estimated passage time.

[0047] For tasks involving multiple work locations and resources, an optimization model for resource scheduling and task allocation is established, with the objective function being the shortest overall task completion time and the highest resource utilization efficiency.

[0048] The simulation calculation outputs a detailed time-resource-task association table, which is the optimized emergency rescue command task sequence;

[0049] The emergency rescue command task sequence clearly defines the content of each specific task, the work location to be executed, the allocated resource units, the start and end times of the plan, and the dependencies between tasks.

[0050] Preferably, the step of decomposing the emergency rescue command task sequence into specific navigation and operation instructions, and issuing them to the corresponding rescue robots and personnel smart terminals to guide them to designated risk points to perform tasks, specifically involves:

[0051] The emergency rescue command task sequence is analyzed, and for each task item assigned to the rescue robot in the sequence, the coordinates of its target work point, the planned task content, and the time window are extracted.

[0052] Based on the coordinates of the target work point and the current position of the rescue robot, combined with the latest information on the status of the internal passageways of the building, a segmented navigation path from the current position to the target work point is generated.

[0053] The segmented navigation path is encoded into a serialized movement instruction set that can be recognized by the rescue robot. The movement instruction set includes forward distance, turning angle, speed, and strategies for dealing with obstacles.

[0054] At the same time, for the content of the planned task, the operation control program of the built-in or external tools of the rescue robot is matched to generate a specific sequence of operation instructions, such as the movement trajectory of the robotic arm, cutting parameters, and detection mode;

[0055] The serialized movement instruction set and operation instruction sequence are packaged together, and a task number and timestamp are attached. The package is then sent to the corresponding emergency rescue robot control unit via a wireless communication network.

[0056] For tasks assigned to rescue teams, extract their target work area, task steps, and safety requirements;

[0057] Generate augmented reality navigation guidance information and task cards for personnel-oriented smart terminals. The guidance information is overlaid on the building floor plan or real-scene video on the terminal screen.

[0058] The augmented reality navigation guidance information, task cards, and related partial information of the 3D visualization layer of building safety status are packaged and sent to the smart terminals of the corresponding rescue personnel team.

[0059] Preferably, the system further includes:

[0060] The execution feedback module continuously receives on-site video streams, vital sign data, and environmental re-measurement data from the rescue robot and personnel's smart terminals during task execution, forming an on-site task execution feedback data stream, specifically:

[0061] Establish a stable data transmission link with the intelligent terminals of all dispatched emergency rescue robots and personnel;

[0062] It can receive real-time panoramic video streams and close-up video streams of the work site collected by the rescue robot through its multi-camera system, as well as the robot's own status monitoring data.

[0063] It can receive real-time video streams from the perspective of personnel through the front-facing camera of the personnel's smart terminal, as well as vital signs data such as heart rate, blood oxygen, and body temperature of rescue personnel collected through wearable devices;

[0064] The command system instructs emergency rescue robots and personnel intelligent terminals to re-measure the gas concentration, temperature, and radiation intensity of the local environment at the risk point through built-in or external sensors when they arrive at the work site or perform critical operations, and generate environmental re-measurement data.

[0065] The received panoramic video stream and partial close-up video stream are synchronized in time and registered in space, and their images are associated with the corresponding positions in the three-dimensional visualization layer of the building safety situation.

[0066] Real-time analysis of received vital sign data is performed to assess the physiological workload and safety status of rescue personnel.

[0067] All transmitted video stream data, robot status data, vital sign data, and environmental retest data are fused and tagged according to time sequence, spatial location, and data source to form a multi-dimensional, spatiotemporally indexed on-site task execution feedback data stream.

[0068] Preferably, the system further includes:

[0069] The dynamic evaluation module integrates and analyzes the on-site task execution feedback data stream with the building safety status basic data stream to dynamically evaluate the status changes of risk points and the effectiveness of emergency response tasks. Specifically:

[0070] From the on-site task execution feedback data stream, environmental retest data is extracted and compared with the original sensor array data of the risk point to determine whether the key parameters of gas concentration and temperature have changed to a safe range.

[0071] Analyze the video stream in the on-site task execution feedback data stream, and use computer vision algorithms to identify the work progress of the rescue robot or rescue personnel, such as whether the support structure is installed in place and whether the leak point is sealed.

[0072] By correlating newly observed cracks and debris falling from the structural surface in the video stream with the original deformation and vibration data, we can assess whether secondary or diffuse changes have occurred at the risk points.

[0073] Analyze the vital signs data of rescue personnel in the on-site mission execution feedback data stream to assess personnel safety risks and determine whether replacement or reinforcement is needed.

[0074] Based on the above analysis results, a dynamic assessment conclusion is generated for each risk point that is being addressed. The conclusion includes the degree of risk mitigation, the percentage of task completion, whether new risks have emerged, and the safety status of personnel.

[0075] The dynamic assessment conclusions are compared with the original plan for the risk points in the emergency rescue command task sequence to calculate the deviation of task execution.

[0076] Based on the dynamic assessment conclusions and the deviation of task execution, update the labeling status of the corresponding risk points in the 3D visualization layer of the building safety situation, such as color gradient and progress bar display.

[0077] Preferably, the system further includes:

[0078] The real-time control module, based on the dynamic evaluation results, adjusts and replans the emergency rescue command task sequence in real time, generates updated navigation and operation instructions, and issues them out. Specifically:

[0079] Receive dynamic assessment conclusions of risk points and reports on task execution deviations;

[0080] If the dynamic assessment concludes that the risk has been largely eliminated, the corresponding task item is removed from the emergency rescue command task sequence, and the relevant resources are instructed to enter standby or return status.

[0081] If the dynamic assessment results show that the task execution progress is lagging behind or encounters unforeseen difficulties, the resources and time required for the task will be reassessed, the time arrangement in the emergency rescue command task sequence will be adjusted, or additional resources will be allocated for support.

[0082] If the dynamic assessment results show that new risk points have emerged or existing risks have spread, a new risk point location and level assessment process will be initiated, and the newly confirmed risk points and levels will be integrated into the three-dimensional visualization layer of the building safety situation.

[0083] Based on the updated risk situation and resource status, the resource accessibility and timeliness simulation is re-executed to generate an adjusted emergency rescue command task sequence;

[0084] Based on the adjusted emergency rescue command task sequence, navigation and operation instructions for newly added, changed, or continued tasks are re-decomposed and generated.

[0085] Updated navigation and operation instructions will be prioritized and sent to the smart terminals of emergency rescue robots and personnel affected by the adjustments.

[0086] The command platform interface highlights the changes to the task sequence and records the reasons for the adjustments and the basis for the decisions.

[0087] Preferably, the system further includes:

[0088] The report archiving module generates a complete emergency response report after confirming that the risk points have been effectively controlled or eliminated, updates the emergency plan database, and archives all process data. Specifically:

[0089] When the dynamic assessment of all risk points shows that the risks have been eliminated or reduced to an acceptable level, and no new risks have emerged, the commander shall confirm the end of the emergency rescue operation.

[0090] Automatically summarize all key data, decision records, instruction logs, and assessment reports generated throughout the entire process from the generation of basic data streams on building safety status to the final issuance of task adjustments;

[0091] According to the preset report template, the summarized data is organized into chapters, which shall include at least the event overview, risk analysis process, command and decision sequence, resource allocation status, task execution details, verification of handling effect, and summary recommendations;

[0092] Generate a complete emergency rescue operation report document with rich graphics and text, including data charts and key video screenshots;

[0093] Based on the successful verification of the handling procedures and resource allocation experience in this operation, as well as the newly discovered characteristics of the hazards, the emergency plan database will be supplemented and revised, adding new case entries or optimizing the original plans.

[0094] The complete emergency rescue operation report, the updated emergency plan library version, and all process data of the original building safety status basic data stream and on-site task execution feedback data stream are associated and packaged together.

[0095] Assign a unique event number to the associated data packet, add metadata tags, and store it in the historical database of building safety emergency cases to complete the archiving.

[0096] Compared with the prior art, the beneficial effects of the present invention are:

[0097] By precisely matching and binding the structural risk points and their assessment levels identified by the risk analysis module with the spatial coordinates of the target building's 3D digital model, a dynamic visualization layer integrating risk annotation is generated. This technology realizes the transformation of monitoring data from numerical alarms to spatial status quo, allowing the spatial distribution, clustering, and level differences of risks to be directly displayed in the 3D model. Command personnel can immediately understand the correspondence between risks and the building's physical structure, shortening the decision-making preparation process from receiving information to forming spatial understanding.

[0098] By integrating real-time location, status, and on-site environmental information of rescue robots, personnel, and equipment, digital simulations are performed on the preliminary action plans generated based on the contingency plans. The simulation process calculates the path time for resources to reach different risk points, the task execution sequence, and the probability of resource conflicts, thereby optimizing and generating a task sequence with shorter total time or more balanced resource allocation. This transforms static text-based contingency plans into dynamic, executable plans that adapt to real-time constraints, avoiding path redundancy and task waiting in resource scheduling, and improving the overall timeliness and reliability of multi-unit collaborative operations.

[0099] The optimized command task sequence is automatically parsed into specific navigation paths, operational instructions, and control parameters, and then distributed to the corresponding mobile robot platforms and personnel smart terminals. This process achieves seamless integration from macro-level command plans to specific execution actions, guiding on-site forces to accurately reach predetermined locations and carry out rescue tasks in sequence, reducing delays and errors that may be caused by command transmission levels and human intervention. Attached Figure Description

[0100] Figure 1 This is a timing diagram of the emergency rescue command system based on building safety requirements described in this invention;

[0101] Figure 2 A flowchart illustrating the workings of the risk analysis module;

[0102] Figure 3 A flowchart illustrating the operation of the situation visualization module;

[0103] Figure 4 A dual-axis analysis chart of task progress and time deviation in building emergency rescue scenarios;

[0104] Figure 5 A comparative chart verifying the multi-dimensional handling effects of building emergency rescue operations. Detailed Implementation

[0105] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0106] See Figure 1 This emergency rescue command system, based on building safety requirements, constructs a closed-loop command system from perception to execution by sequentially executing data acquisition, risk analysis, situation visualization, plan generation, simulation optimization, and command issuance. Sensor arrays deployed at key nodes of the target building structure continuously collect data on building deformation, vibration, and ambient gas concentration, forming a basic data stream of building safety situation. The risk analysis module extracts features and identifies abnormal patterns from this data stream, locating potential or existing structural risks and assessing their risk levels. The situation visualization module loads a 3D model of the target building based on a geographic information system, mapping the identified risk points and their levels to corresponding locations on the model, generating a 3D visualization layer of the building safety situation with risk labels. The plan generation module, based on this visualization layer and a pre-set emergency plan library, automatically matches and generates a preliminary emergency rescue action plan. The simulation optimization module integrates the status and location information of available rescue resources, performs resource accessibility and timeliness simulations on the preliminary plan, and generates an optimized emergency rescue command task sequence. The instruction issuing module breaks down the task sequence into specific navigation and operation instructions, which are then sent to the corresponding emergency rescue robots and personnel smart terminals to guide them to designated risk points to perform the tasks.

[0107] In one embodiment of the present invention, see [reference] Figure 2In practical implementation, taking the structural safety assessment and emergency rescue required after a fire in a high-rise building as an example, the risk analysis module begins processing the basic data stream of building safety status collected by the sensor array. The sensor array is deployed in the building's key load-bearing columns, beam nodes, and main escape routes. Deformation sensors collect the original displacement data sequence of key column nodes on a certain floor. The original displacement data sequence shows a continuous and increasing offset over a period of time after the fire. The risk analysis module filters and denoises the original building deformation data, eliminating minor fluctuations caused by temperature changes or measurement noise, and extracts a clear position displacement time-series curve of the column node. Based on the position displacement time-series curve, the displacement velocity and acceleration change characteristics are calculated. The acceleration value increases from 0.02 mm / s² to 0.15 mm / s² within ten minutes. At the same time, vibration sensors collect the original vibration data of the building floor slab. Spectrum analysis shows that the main vibration frequency of the structure shifts from the normal 2.5 Hz to 1.8 Hz, and at the same time, a high-frequency, atypical impact vibration mode is identified, which is the characteristic of abnormal vibration mode. Environmental sensors deployed in the corridors of the damaged floors collected raw environmental gas concentration data. After trend analysis, the data revealed that the carbon monoxide concentration rose from 30 ppm to 90 ppm in a short period of time, showing an abnormal increase.

[0108] The characteristics of displacement velocity and acceleration changes, abnormal vibration modes, and abnormal increases in gas concentration are input into a pre-trained risk assessment classification model. This model has been trained using a large amount of historical incident data and can correlate multimodal features with risk types. For column node monitoring locations, the risk assessment classification model outputs a structural risk probability value of 0.87 and assigns a risk type label of "load-bearing structure instability." For areas with abnormal gas concentrations, the output structural risk probability value is 0.65, and the risk type label is "toxic gas accumulation." In some embodiments, the risk analysis module sorts all monitoring points according to the magnitude of the structural risk probability value, sets a preset threshold of 0.6, and filters monitoring points with a structural risk probability value exceeding 0.6 as candidate risk points. Combining the risk type label output by the risk assessment classification model, the module assigns a quantified risk level to each candidate risk point. The risk level integrates the probability of the risk occurring (structural risk probability value) with the expected severity of the consequences. The "load-bearing structure instability" risk, which may lead to local collapse, has a high weighting for severity and is therefore assigned a "high risk" level, while the "toxic gas accumulation" risk is assigned a "medium risk" level. The spatial coordinates, risk types, and risk levels of all candidate risk points are packaged to form a risk point location and level list containing entries such as "coordinates (X1, Y1, Z1), risk type: load-bearing structure instability, risk level: high risk".

[0109] In practical implementation, the quantification of risk levels provides a direct basis for subsequent resource allocation. A comprehensive evaluation function is introduced during the risk level calculation process, with the structural risk probability value and consequence severity coefficient as the core variables. The structural risk probability value is directly given by the risk assessment classification model, while the consequence severity coefficient is obtained from a pre-defined mapping table based on the risk type label. For the "load-bearing structure instability" type, the consequence severity coefficient is set to 1.0; for the "toxic gas accumulation" type, the consequence severity coefficient is set to 0.7.

[0110] In some embodiments, the comprehensive evaluation function The calculation is expressed as:

[0111]

[0112] in: This represents the final calculated quantitative value of the risk level; This represents the structural risk probability value output by the risk assessment classification model; This represents the severity coefficient of the consequences obtained based on the risk type label; and These represent the structural risk probability values. With the severity coefficient of consequences The weighting factor. Based on the quantized value. The numerical range in which the data falls is mapped to discrete risk level labels such as "low risk," "medium risk," and "high risk." Optionally, the risk analysis module's processing of vibration data is not limited to spectrum analysis; it also needs to calculate the time-history variation of vibration energy. When vibration energy increases sharply in a short period of time, the time-history variation characteristics of vibration energy will also be input into the risk assessment classification model as part of the abnormal vibration mode characteristics.

[0113] In one embodiment of the present invention, see [reference] Figure 3 In its implementation, the situation visualization module operates on a Geographic Information System (GIS) platform. The module retrieves a 3D model containing structural component information from the target building's Building Information Model (BIM). This 3D model includes geometric and attribute data for components such as beams, columns, slabs, and walls. This 3D model is imported into the GIS platform as a baseline scene. The module then obtains a list of risk point locations and risk levels generated by the risk analysis module. This list includes candidate risk points with coordinates (105.2, 78.5, 15.3), along with their spatial coordinates, risk type, and risk level. Within the GIS platform, the spatial coordinates of the candidate risk points are precisely matched with the corresponding structural components or spatial locations in the 3D model. The coordinates (105.2, 78.5, 15.3) are matched to the load-bearing column component numbered C-15 in the 3D model.

[0114] Based on the matching results, 3D icons are used to visually annotate the corresponding locations on the 3D model. The visual attributes of the icons are associated with the risk level. Candidate risk points marked as "high risk" are marked with a flashing red pyramid icon, and candidate risk points marked as "medium risk" are marked with an orange cube icon. Next to the risk point annotations on the 3D model, a floating information box displays the risk type label and the specific risk probability value. When the cursor hovers over the red pyramid icon, the information box displays "Risk type: load-bearing structure instability, risk probability: 0.87". The module renders the annotated 3D model into an interactive 3D visualization scene, supporting rotation, scaling, and translation. In the 3D visualization scene, a physical location distribution map of the sensor array is overlaid, with blue dots indicating the installation location of each sensor; an internal building passage network layer is overlaid, highlighting all available corridors and staircases with a semi-transparent green highlight; and a critical facility location layer is overlaid, marking the locations of the power distribution room, pump room, and fire hydrants.

[0115] In some embodiments, the interactive 3D visualization scene is integrated with the physical location distribution map of the sensor array, the building's internal passage network layer, and the key facility location layer, and encapsulated into a single data package containing multi-layer data. This data package is the 3D visualization layer data package for building safety status, which is accessed and displayed in 3D by the main interface of the command platform through a standard interface. The scheme generation module is activated based on the 3D visualization layer of building safety status. The module parses all risk point information marked in the 3D visualization layer of building safety status, extracts the dominant risk type as "load-bearing structure instability", and the highest risk level as "high risk". The spatial distribution characteristics of the risk points show that most points are concentrated in the southwest corner of area C of the building. Using "load-bearing structure instability" and "high risk" as query keywords, a full-text search and matching degree calculation are performed in a preset emergency plan database. The emergency plan database stores entries such as "emergency support plan after fire in the load-bearing structure of a high-rise building" and "toxic gas leak sealing and ventilation plan".

[0116] Several candidate emergency response plans with a matching degree exceeding the preset standard were retrieved and filtered. For example, the matching degree of "Emergency Support Plan for Fire on Load-Bearing Structure of High-Rise Building" was calculated to be 0.92. The spatial distribution characteristics of the risk points were analyzed. Since the risk points were concentrated in the southwest corner of Zone C, the module determined that this event was a single-point or localized event. Based on the event type determination, "Emergency Support Plan for Fire on Load-Bearing Structure of High-Rise Building" was selected from the candidate emergency response plans as the basic task framework. The basic task framework includes four stages: isolating the hazardous area, assessing the degree of damage, installing temporary supports, and verifying structural stability. The information of the passage network layer and key facility location layer superimposed on the 3D visualization layer of the building safety situation was integrated into the basic task framework. The passage network layer showed that the path from the east entrance to the southwest corner of Zone C was unobstructed. Based on this, an approach route was planned for the rescue force to enter from the east entrance, pass through passage A, and reach the work site via staircase B.

[0117] Understandably, the solution generation module, combining the initial status information of available emergency rescue resources, assigns task execution entities to each stage in the basic task framework. The initial status information shows that "Heavy Support Robot Unit-01" is in standby mode and located at the base, while "Structural Assessment Team-Alpha" is located on the third floor of the building. A preliminary emergency rescue action plan draft is generated, containing descriptions of task stages, action routes, work locations, and execution entities. For example, "Phase 1: Isolation, the Structural Assessment Team-Alpha enters first along the predetermined route to delineate the work area; Phase 2: Assessment, the Structural Assessment Team-Alpha conducts on-site surveys; Phase 3: Disposal, the Heavy Support Robot Unit-01 enters and installs supports; Phase 4: Verification, the Structural Assessment Team-Alpha collaborates with the robot to conduct stability tests." The preliminary emergency rescue action plan draft undergoes logical consistency verification and a rough time estimate. The logical consistency verification ensures a reasonable sequence of task stages and no resource conflicts. The rough time estimate indicates that completing all stages will take approximately 85 minutes, outputting a preliminary emergency rescue action plan that can be optimized. Optionally, the matching degree calculation of emergency response plans not only relies on keyword retrieval but also involves a structured comparison of the applicable conditions of the plans, and the matching degree... The calculation can be expressed as:

[0118]

[0119] in: This represents the final calculated matching degree of the proposed plan; The semantic similarity score represents the risk type and the contingency plan type. The score is calculated using a natural language processing model. The score represents the degree of conformity between the risk level and the recommended activation level of the contingency plan. The conformity score is determined according to the level mapping table. and These represent semantic similarity scores. Compliance score The harmonic weighting coefficients are set during system initialization.

[0120] It is understandable that when planning the approach route for emergency rescue forces, if the channel network layer information shows that a certain path has a red warning mark indicating congestion, the plan generation module will automatically exclude the congested path, reselect the unobstructed path marked with green, and record the path change information in the remarks of the preliminary emergency rescue action plan draft.

[0121] In one embodiment of the present invention, in a specific implementation, the simulation optimization module receives a preliminary emergency rescue action plan output by the scheme generation module. The preliminary emergency rescue action plan includes the task of handling the high-risk point in the southwest corner of building C area. The task requires going to the work point near coordinates (105.2, 78.5, 15.3) to carry out temporary support work. The simulation optimization module acquires and maintains a list of available rescue resources in real time. The list of available rescue resources includes items such as "Heavy support robot unit-01: Status-Standby, Location-Base garage coordinates (X0, Y0, Z0), Function-Lifting and support, Endurance-4 hours", "Structural assessment team-Alpha: Status-Standby, Location-Building third floor rest area coordinates (X3, Y3, Z3), Function-Structural survey and assessment, Endurance-6 hours", and "Multifunctional detection robot-Beta: Status-Working, Location-Building second floor corridor coordinates (X2, Y2, Z2), Function-Gas and image detection, Endurance-1.5 hours". The module compares the coordinates of the planned work points (105.2, 78.5, 15.3) in the preliminary emergency rescue action plan with the real-time locations in the list of available rescue resources. The current position coordinates (X0, Y0, Z0) of the heavy support robot unit-01 are about 120 meters away from the target work point, and the current position coordinates (X3, Y3, Z3) of the structural assessment team-Alpha are about 35 meters away from the target.

[0122] Based on the passage network layer in the 3D visualization layer of building safety status, considering passage restrictions, damage, and congestion, the passage network layer shows that a section of passage B is marked with an orange warning, indicating limited passage capacity. The module calculates the theoretical shortest path and estimated travel time for each available emergency rescue resource from its current location to the designated work point. The estimated travel time for the heavy support robot unit-01 via passage A, passage B, and staircase C is 18 minutes; the estimated travel time for the structural assessment team-Alpha via corridor D and staircase C is 7 minutes. Based on the estimated travel time and resource endurance information, the multi-functional detection robot-Beta has only 1.5 hours of endurance remaining and is currently far away. It is deemed unsuitable for immediate deployment in subsequent long-term operations and is therefore eliminated. For the work site coordinates (105.2, 78.5, 15.3), both the heavy support robot unit-01 and the structural evaluation team-Alpha meet the requirements. Prioritization is based on the matching degree between their functional capabilities and the task, and the estimated travel time. The structural evaluation team-Alpha, due to its high task matching degree and short travel time, is assigned to perform the evaluation task first, followed by the heavy support robot unit-01. In some embodiments, for tasks involving multiple work sites and multiple resources, the deduction and optimization module establishes an optimization model for resource scheduling and task allocation, using the shortest overall task completion time and the highest resource utilization efficiency as the objective function for deduction and calculation. The construction takes into account task completion time and resource idle cost, and is expressed as:

[0123]

[0124] in: This represents the value of the objective function that needs to be minimized. Represents the total number of task items; Representing the The absolute timeframe for the planned completion of the task; This represents the total number of units of available emergency relief resources; Representing the Idle time of each resource unit during the execution of a task; The penalty coefficient representing resource idle time is used to balance time optimization and resource efficiency. The extrapolation calculation is based on constraints such as resource location, path travel time, and estimated task execution time, and is solved iteratively. The extrapolation calculation outputs a detailed time-resource-task association table, which is the optimized emergency rescue command task sequence. The emergency rescue command task sequence contains entries such as "Task ID: T001, Content: On-site structural safety assessment, Work location: near coordinates (105.2, 78.5, 15.3), Resource allocation: Structural assessment team - Alpha, Planned start time: 09:15, Planned end time: 09:45, Preceding tasks: None".

[0125] It is understandable that the emergency rescue command task sequence clearly defines the content of each specific task, the work location, the allocated resource units, the planned start and end times, and the dependencies between tasks. For example, "Task ID: T002, Content: Install temporary hydraulic supports, Work location: Coordinates (105.2, 78.5, 15.3), Allocated resources: Heavy support robot unit-01, Planned start time: 09:50, Planned end time: 10:30, Prerequisite task: T001". The instruction issuance module receives and parses the emergency rescue command task sequence, extracting the target work location coordinates, planned task content, and time window for each task item assigned to the rescue robot in the sequence. For task T002, the target work location coordinates (105.2, 78.5, 15.3), the planned task content "Install temporary hydraulic supports", and the time window 09:50 to 10:30 are extracted. Based on the target work point coordinates (105.2, 78.5, 15.3) and the current position coordinates (X0, Y0, Z0) of the heavy support robot unit-01, combined with the latest building internal passage status information, the latest status information confirms that the orange warning for passage B is still valid but the robot can pass through. A segmented navigation path from the current position to the target work point is generated, and the path description is "go straight ahead for 20 meters from (X0, Y0, Z0), turn left into passage A, and travel along passage A for 60 meters...".

[0126] The segmented navigation path is encoded into a serialized movement command set recognizable by the rescue robot. This serialized movement command set includes commands such as "MOVE_FORWARD,distance=20m,speed=0.5m / s", "TURN_LEFT,angle=90deg", and "MOVE_FORWARD,distance=60m,speed=0.5m / s,obstacle_avoidance=ON". For the planned task "installing temporary hydraulic support", the hydraulic support arm operation control program built into the heavy-duty support robot unit-01 is matched to generate a specific sequence of operation commands. This sequence includes "EXTEND_ARM,position=coordinates(105.2,78.5,15.3),force=5000N", "LOCK_JOINT,joint_id=3", and "ACTUATE_HYDRAULIC,pressure=20MPa". The serialized movement command set and the operation command sequence are packaged together, appended with the task number "T002" and a timestamp, and then transmitted to the control unit of the heavy-duty support robot unit-01 via a wireless communication network. Optionally, for task T001 assigned to the Structural Assessment Team-Alpha, the instruction delivery module extracts its target work area, task steps, and safety requirements. The target work area is a radius of 5 meters centered at coordinates (105.2, 78.5, 15.3). Augmented reality navigation guidance information and task cards are generated for personnel smart terminals. The augmented reality navigation guidance information is overlaid on the building floor plan or real-world video on the terminal screen. When rescue personnel raise their smart terminals, a green arrow indicating the direction and the text "Turn left 30 meters ahead" are overlaid on the real corridor scene on the screen. The augmented reality navigation guidance information, task cards, and related partial information from the 3D visualization layer of the building safety situation are packaged and sent to the Structural Assessment Team-Alpha's smart terminal. The task card displays "Task ID: T001, Steps: 1. Scan the surface of column C-15 using a flaw detector; 2. Record the crack width and depth; 3. Send the data back to the command center; Safety requirements: Wear a respirator, stay away from obviously loose components."

[0127] In one embodiment of the present invention, the execution feedback module is activated after the emergency rescue command task sequence is issued. The module establishes a stable data transmission link with all dispatched rescue robots and personnel smart terminals. For the structural assessment team-Alpha and heavy support robot unit-01 performing tasks at risk point P-15 (coordinates (105.2,78.5,15.3)), the execution feedback module receives in real time panoramic video streams of the work site and close-up video streams of the robotic arm end effector captured by the multi-camera system on the heavy support robot unit-01. Simultaneously, it receives battery power and joint torque status monitoring data of the heavy support robot unit-01 itself. The module also receives in real time on-site video streams from the personnel's perspective captured by the front-facing camera of the structural assessment team-Alpha personnel smart terminal. The video streams show the assessment personnel inspecting the surface of the column, and simultaneously receive vital sign data such as heart rate, blood oxygen, and body temperature through the smart bracelet worn by the assessment personnel. The heart rate is 112 beats per minute, and the blood oxygen saturation is 97%.

[0128] The system instructs the intelligent terminals of the Heavy Support Robot Unit-01 and the Structural Assessment Team-Alpha to re-measure the local environment at the risk point when they arrive at the work location near coordinates (105.2, 78.5, 15.3). The composite gas sensor of the Heavy Support Robot Unit-01 generates environmental re-measurement data showing a carbon monoxide concentration of 65 ppm and a thermal imager showing a column surface temperature of 52 degrees Celsius. The portable thermometer of the Structural Assessment Team-Alpha generates environmental re-measurement data showing an ambient air temperature of 41 degrees Celsius. The execution feedback module performs time synchronization and spatial registration of the received panoramic video stream and local close-up video stream, associating the support operation footage captured by the Heavy Support Robot Unit-01's camera with the coordinates (105.2, 78.5, 15.3) in the 3D visualization layer of the building safety situation. The module performs real-time analysis of the received vital sign data; a heart rate of 112 beats per minute exceeds the preset static warning threshold of 100 beats per minute, indicating that the rescue personnel are in a state of high physiological stress. The execution feedback module integrates and tags all returned video stream data, robot status data, vital sign data, and environmental retest data according to time sequence, spatial location, and data source, forming a multi-dimensional, spatiotemporally indexed on-site task execution feedback data stream.

[0129] The dynamic assessment module receives and processes the on-site task execution feedback data stream. It extracts environmental re-measurement data from this data stream and compares it with the original sensor array data from risk point P-15. The original sensor array data showed a carbon monoxide concentration of 90 ppm and an air temperature near the column of 55 degrees Celsius. The on-site re-measurement data showed that the carbon monoxide concentration had decreased to 65 ppm and the air temperature near the column had decreased to 41 degrees Celsius. The module determined that the key parameters of gas concentration and temperature were shifting towards safe ranges. Analyzing the video stream within the on-site task execution feedback data stream, the module uses computer vision algorithms to identify the work progress of the heavy-duty support robot unit-01. The algorithm analyzes the video frames and identifies that three of the four fixed points of the temporary hydraulic support have been installed, determining that the support installation task is approximately 75% complete. The dynamic assessment module correlates a newly appearing fine vertical crack on the side of the column observed in the video stream with the original deformation and vibration data. An anomaly was found in the vibration spectrum of this area in the original data, suggesting that the appearance of the new crack may indicate a secondary change at the risk point.

[0130] In some embodiments, the vital signs data of rescue personnel in the on-site task execution feedback data stream are analyzed to assess personnel safety risks. If the heart rate of the structured assessment team - Alpha remains above 110 beats per minute, the system determines that personnel fatigue has accumulated, posing a safety risk and potentially requiring replacement or reinforcement. Based on the above analysis results, the dynamic assessment module generates a dynamic assessment conclusion for the risk point P-15 being addressed. The dynamic assessment conclusion includes the degree of risk mitigation, the percentage of task completion, whether new risks have emerged, and the personnel safety status. The dynamic assessment conclusion is recorded as structured data, as shown in Table 1.

[0131] Table 1: Summary of Dynamic Assessment Conclusions for Risk Point P-15

[0132]

[0133] The dynamic assessment conclusions are compared with the original plan for risk point P-15 in the emergency rescue command task sequence. The original plan required the support installation to be completed between 09:50 and 10:30. At the current time of 10:05, the task progress is 75%, and the deviation in task execution is calculated to be approximately 5 minutes behind schedule. Based on the dynamic assessment conclusions and the task execution deviation, the dynamic assessment module updates the labeling status of the corresponding risk point P-15 in the 3D visualization layer of the building safety situation. The 3D icon color of risk point P-15 changes from flashing red to alternating red and yellow flashing, and a progress bar displaying "75%" is added next to the icon. The newly added crack location is marked with a small yellow triangle icon. Optionally, when generating the comprehensive assessment conclusion, the dynamic assessment module uses a quantified risk change index. To integrate multi-dimensional information, risk change index The calculation is expressed as:

[0134] in: This represents the calculated risk change index; a positive value indicates increased risk, while a negative value indicates reduced risk. This represents the number of key environmental and status parameters being monitored. Representing the Weighting of changes in parameters (such as temperature, gas concentration, stress); Representing the The normalized difference between the field remeasured values ​​of each parameter and the original baseline values; The penalty coefficient representing the emergence of new risks (such as new cracks); Represents the percentage of completion for key tasks (such as installation support) derived from computer vision analysis. Modules are based on... The range and trend of values ​​help generate dynamic assessment conclusions. It can be understood that, for the analysis of personnel vital sign data, the dynamic assessment module not only monitors instantaneous values ​​but also calculates trends over a period of time. For example, it calculates the average and standard deviation of heart rate over 10 consecutive minutes. If the average continuously rises and the standard deviation increases, it may indicate that the person is in a state of heightened stress or fatigue. This trend analysis result will also be incorporated into the assessment conclusion of the personnel's safety status. The dynamic assessment module will update the 3D visualization layer of the building safety situation and include risk change indices. The dynamic evaluation results report is pushed to the command platform and real-time control module in real time.

[0135] See Figure 4 This is a dual-axis analysis chart of task progress and time deviation in a building emergency rescue scenario. It visually displays the difference between the planned and actual progress of core tasks such as support installation, as well as the cumulative trend of the progress deviation over time. The chart clearly shows that at all monitoring time points, the actual progress lags behind the planned progress, and the gap gradually widens over time. This trend can directly trigger a progress warning from the emergency command system, prompting commanders to assess the reasons for the delay. The time deviation increases continuously from an initial 0 minutes to a final 7 minutes, reflecting a continuous deterioration in the task delay without any signs of mitigation. This linear cumulative trend can help predict the final completion time of the task, providing a basis for subsequent resource allocation. Between monitoring time points 2 and 3, the increase in actual progress is significantly lower than the increase in planned progress, leading to a faster rate of increase in time deviation, reflecting potential unforeseen difficulties on-site, which can guide commanders to take targeted interventions.

[0136] In one embodiment of the present invention, in a specific implementation, the real-time control module receives the dynamic assessment conclusion and task execution deviation report of risk point P-15 from the dynamic assessment module. The dynamic assessment conclusion shows that the support installation task is 75% complete, but new minor cracks have appeared in the column, and the personnel show signs of fatigue. The task execution deviation report shows that the progress is lagging by about 5 minutes. The real-time control module analyzes the dynamic assessment conclusion and finds that the support installation task is not yet complete and a new risk point has appeared. The risk has not been basically eliminated, so the task item removal operation is not performed. The module determines that the task execution progress is lagging due to the discovery of new risks and the status of personnel. It reassesses the resources and time required for the task. Originally, the heavy support robot unit-01 was planned to complete the support installation independently. Now, it is assessed that an additional group of personnel is needed to assist in monitoring and fixation, and the structural assessment team-Alpha will be replaced and rotated to alleviate fatigue. The real-time control module dispatches the "structural assessment team-Bravo" which is in standby status to the area at coordinates (105.2,78.5,15.3) for reinforcement and rotation.

[0137] The dynamic assessment revealed new, fine vertical cracks on the side of the column. The real-time control module identified this as a new potential risk point and initiated a new risk point location and level assessment process. Based on the video coordinates of the crack's appearance and the matching with the 3D model, the new risk point was assigned coordinates (105.3, 78.6, 15.3) and designated as risk point P-16. After a simplified risk assessment process, risk point P-16 was assigned the risk type "local cracking" and the risk level "medium risk." The newly confirmed risk point P-16 and its level information were integrated into the 3D visualization layer of the building safety situation, with a new yellow triangle icon added to the corresponding position in the layer. Based on the updated risk situation and resource status, the updated risk situation includes the original risk point P-15, which still requires attention, and the newly added risk point P-16. The resource status includes Heavy Support Robot Unit-01, Structural Assessment Team-Alpha, and Structural Assessment Team-Bravo. The real-time control module triggered the simulation optimization module to re-execute resource accessibility and timeliness simulations.

[0138] In some embodiments, the re-executed simulation considers rotation and collaborative operations to generate an adjusted emergency rescue command task sequence. The adjusted emergency rescue command task sequence includes new task items, such as "Task ID: T001-B, Content: Rotate and assist in on-site monitoring, Work location: near coordinates (105.2, 78.5, 15.3), Allocated resources: Structural Assessment Team - Bravo, Planned start time: 10:10, Preceding task: None" and "Task ID: T002-Cont, Content: Continue to complete support installation and monitor new cracks, Work location: coordinates (105.2, 78.5, 15.3) and (105.3, 78.6, 15.3), Allocated resources: Heavy Support Robot Unit - 01, Structural Assessment Team - Bravo, Planned start time: 10:15, Preceding task: T001-B". Based on the adjusted emergency rescue command task sequence, the instruction issuance module re-decomposes and generates navigation and operation instructions for newly added, modified, or continued tasks. Augmented reality navigation guidance is generated for the Structural Assessment Team-Bravo, guiding it from the assembly point on the first floor of the building to the target area. Additional instructions are generated for the Heavy Support Robot Unit-01, instructing it to move to the vicinity of coordinates (105.3, 78.6, 15.3) after completing support installation to perform image scanning of the cracks. The updated navigation and operation instructions are prioritized and issued to the intelligent terminals of the affected rescue robots and personnel. The terminals of the Heavy Support Robot Unit-01 and the Structural Assessment Team-Bravo receive the new instructions. The changed parts of the task sequence are highlighted on the command platform interface with a highlighted border and an "Updated" label, and the reasons for the adjustment and the basis for the decision are recorded. The recorded information is: "10:08 Adjustment Decision: Due to the occurrence of a secondary crack (P-16) at risk point P-15 and the fatigue of the original assessment team, the Bravo team is added for rotation, and the robot's task is expanded to crack scanning."

[0139] Understandably, when the dynamic assessment conclusions of all risk points indicate that the risks have been eliminated or reduced to an acceptable level, and no new risks have emerged, the commander clicks the confirmation button on the command platform interface to confirm the end of the emergency rescue operation. The report archiving module then activates, automatically summarizing all key data, decision records, instruction logs, and assessment reports generated throughout the entire process, from the generation of the basic building safety situation data stream to the final task adjustment and issuance. The summarized data includes the initial risk point location and level list, various versions of the emergency rescue command task sequence, dynamic assessment conclusion record table, and all issued and returned instructions and feedback data packets. Following the preset report template, the report template requires seven chapters: event overview, risk analysis process, command decision sequence, resource allocation status, task execution details, verification of the handling effect, and summary recommendations, organizing the summarized data into these chapters. Generate a complete emergency response report document with rich graphics and text, including data charts and key video screenshots. The event overview section records the event as occurring on January 26, 2026 at 09:00, in Zone C of a high-rise commercial building, and the primary risk as structural instability after a fire. The section on verifying the effectiveness of the response includes charts comparing the stability test data of the columns after the support installation was completed, as well as screenshots showing no expansion of cracks in the crack scan results.

[0140] Optionally, based on the successfully validated response procedures, resource allocation experience, and newly discovered hazard characteristics from this operation, the report archiving module supplements and revises the emergency plan database. A supplementary entry is added to the "Emergency Support Plan after Fire in Load-Bearing Structures of High-Rise Buildings," described as follows: "If new minor cracks are found in load-bearing components during the response process, detection equipment should be immediately assigned to scan and monitor the cracks after the main support is completed, assess their stability, and mark the area as a secondary monitoring point." The report archiving module associates and packages the complete emergency rescue operation report, the updated version of the emergency plan database, and all process data from the original building safety status basic data stream and on-site task execution feedback data stream. This association and packaging is achieved by assigning the same event number "INC-20260126-001" to all data files. A unique event number "INC-20260126-001" is assigned to the associated data package, metadata tags such as "Building Type: High-Rise Commercial," "Risk Type: Fire Structure," and "Response Result: Success" are added, and the data is stored in the building safety emergency case history database, completing the archiving process. In some embodiments, the report archiving module calculates a data integrity check code when generating archived data packets. Data integrity check codes are used to ensure the integrity of archived data and to enable subsequent retrieval and verification. The calculation is based on the content hash values ​​of all associated files, expressed as:

[0141] in: This represents the final generated data integrity check code; Represents the selected cryptographic hash function; This represents a bitwise XOR operation; This represents the total number of files in the associated package; Representing the The specific binary data content of each file. Calculated. The value is stored in the database as part of the metadata and is associated with the event number "INC-20260126-001".

[0142] See Figure 5 This is a multi-dimensional comparative chart verifying the effectiveness of a building emergency rescue operation. It quantifies the changes in state before and after the emergency response across five key dimensions and compares them with preset safety thresholds, intuitively reflecting the overall effectiveness of the rescue operation. All dimensions' scores exceeded the safety thresholds after the response, validating the effectiveness of the emergency rescue operation and serving as a reference for similar future events. "Structural stability" is only slightly above the safety threshold, indicating that this dimension remains a key focus for future monitoring, requiring continuous structural health monitoring. The significant improvements in "environmental parameter recovery" and "task completion rate" can be extracted as successful experiences from this operation and used to optimize future emergency plans. Through this intuitive comparison, commanders can quickly confirm whether the rescue effects in each dimension have met the standards, providing a quantitative basis for deciding "whether to terminate the emergency response." For example, all dimensions' scores exceeding the safety thresholds after the response support the decision that "the emergency operation has achieved its expected goals."

Claims

1. An emergency rescue command system based on building safety requirements, characterized in that, Includes the following modules: The data acquisition module collects real-time building deformation data, vibration data, and ambient gas concentration data through a sensor array deployed at key nodes of the target building structure, forming a basic data stream of building safety status. The risk analysis module performs feature extraction and abnormal pattern recognition on the basic data stream of building safety status to locate potential or existing building structural risk points and their risk levels. The situation visualization module loads a 3D model of the target building based on a geographic information system, maps the risk points and their risk levels to the corresponding spatial locations of the 3D model, and generates a 3D visualization layer of building safety situation with risk labels. The scheme generation module automatically matches and generates targeted preliminary emergency rescue action plans based on the three-dimensional visualization layer of the building safety situation and in combination with the preset emergency plan library. The simulation and optimization module integrates the status and location information of available emergency rescue resources, performs resource accessibility and timeliness simulations on the preliminary emergency rescue action plan, and generates an optimized emergency rescue command task sequence. The instruction issuance module decomposes the emergency rescue command task sequence into specific navigation and operation instructions, and issues them to the corresponding rescue robots and personnel smart terminals to guide them to designated risk points to perform tasks. Specifically: The emergency rescue command task sequence is analyzed, and for each task item assigned to the rescue robot in the sequence, its target operation point coordinates, planned task content and time window are extracted. Based on the coordinates of the target work point and the current position of the rescue robot, combined with the latest information on the status of the internal passageways of the building, a segmented navigation path from the current position to the target work point is generated. The segmented navigation path is encoded into a serialized movement instruction set that can be recognized by the rescue robot. The movement instruction set includes forward distance, turning angle, speed, and strategies for dealing with obstacles. Simultaneously, for the content of the planned task, the operation control program of the built-in or external tools of the rescue robot is matched to generate a specific sequence of operation instructions, including the robotic arm movement trajectory, cutting parameters, and detection mode; The serialized movement instruction set and operation instruction sequence are packaged together, and a task number and timestamp are attached. The package is then sent to the corresponding emergency rescue robot control unit via a wireless communication network. For tasks assigned to rescue teams, extract their target work area, task steps, and safety requirements; Generate augmented reality navigation guidance information and task cards for personnel-oriented smart terminals. The guidance information is overlaid on the building floor plan or real-scene video on the terminal screen. The augmented reality navigation guidance information, task cards, and related partial information of the 3D visualization layer of building safety status are packaged and sent to the smart terminals of the corresponding rescue personnel team.

2. The emergency rescue command system based on building safety requirements according to claim 1, characterized in that, The process of extracting features and identifying abnormal patterns from the basic data stream of building safety status to locate potential or existing building structural risk points and their risk levels specifically involves: The raw building deformation data collected by the sensor array is filtered and denoised to extract the displacement time-series curves of key structural points and calculate the displacement velocity and acceleration variation characteristics. Spectral analysis is performed on the raw vibration data collected by the sensor array to extract the structural principal frequency, amplitude and damping ratio characteristics, and to identify abnormal vibration modes. Trend analysis is performed on the ambient gas concentration data collected by the sensor array to identify abnormal increases in the concentration of flammable, toxic, or asphyxiating gases. The displacement velocity and acceleration change characteristics, abnormal vibration mode characteristics, and abnormal gas concentration increase characteristics are input into a pre-trained risk assessment classification model. The risk assessment classification model outputs a structural risk probability value and a risk type label for each sensor monitoring point location. The monitoring points are sorted according to their risk probability values, and those with risk probability values ​​exceeding a preset threshold are selected as candidate risk points. Based on the aforementioned risk type labels, each candidate risk point is assigned a quantitative risk level, which comprehensively considers the probability of the risk occurring and the expected severity of the consequences. The spatial coordinates, risk types, and risk levels of all candidate risk points are packaged to form a risk point location and level list.

3. The emergency rescue command system based on building safety requirements according to claim 2, characterized in that, The process involves loading a 3D model of the target building based on a geographic information system, mapping the risk points and their risk levels to the corresponding spatial locations of the 3D model, and generating a 3D visualization layer of the building's safety status with risk annotations. Specifically: Calling a 3D model of the target building with structural component information from a building information model or reconstructing it from survey data, and importing it into a geographic information system platform; Obtain the spatial coordinate information of each candidate risk point in the risk point location and level list; In the geographic information system platform, the spatial coordinates of each candidate risk point are matched with the corresponding structural components or spatial locations in the three-dimensional model; Based on the matching results, risk points are visually marked on the corresponding positions of the three-dimensional model using three-dimensional icons of different colors, shapes, or sizes, wherein the visual attributes of the icons are associated with the risk level. Next to the risk point annotations in the 3D model, the risk type label and the specific risk probability value are displayed in a floating information box. Render the annotated 3D model into an interactive 3D visualization scene; Overlay the physical location distribution map of the sensor array, the internal passage network layer of the building, and the location layer of key facilities; The interactive 3D visualization scene and all overlay layers are integrated and packaged into a 3D visualization layer data package for building safety status, which is then called and displayed by the command platform.

4. The emergency rescue command system based on building safety requirements according to claim 1, characterized in that, Based on the three-dimensional visualization layer of the building safety situation, and in conjunction with a pre-set emergency plan library, the system automatically matches and generates targeted preliminary emergency rescue action plans, specifically as follows: Analyze the risk point information marked in the three-dimensional visualization layer of the building safety situation, and extract the dominant risk type, the highest risk level, and the spatial distribution characteristics of the risk points; Using the dominant risk type and highest risk level as query keywords, a full-text search and matching degree calculation are performed in the preset emergency plan database; The emergency response plan database contains standard handling procedures, required resource lists, and safety precautions for dealing with different types of building risks. Retrieve and filter out several candidate emergency response plan entries whose matching degree exceeds the preset standard; Analyze the spatial distribution characteristics of the risk points. If the risk points are concentrated, they are determined to be single-point or local events. If the risk points are scattered, they are judged as multi-point or systemic events; Based on the determination of the event type, a basic task framework is generated by selecting or combining candidate emergency response plan items. The task framework includes four stages: isolation, assessment, handling, and verification. The information of the passage network layer and key facility location layer superimposed in the three-dimensional visualization layer of the building safety situation is integrated into the basic task framework to plan the approach route and operation points of the emergency rescue force. Based on the initial status information of available emergency rescue resources, task execution entities are assigned to each stage in the basic task framework, forming a preliminary emergency rescue action plan draft that includes descriptions of task stages, action routes, work locations, and execution entities. The draft preliminary emergency rescue action plan is subjected to logical consistency verification and rough time estimation, and an optimized preliminary emergency rescue action plan is output.

5. The emergency rescue command system based on building safety requirements according to claim 1, characterized in that, The integration of available emergency rescue resources' status and location information is used to perform resource accessibility and timeliness simulations on the preliminary emergency rescue action plan, generating an optimized emergency rescue command task sequence, specifically: A list of available emergency rescue resources is acquired and maintained in real time, which includes the current status, real-time location, functional capabilities and battery life information of emergency rescue robots, specialized engineering vehicles, rescue personnel teams and special equipment; The planned work locations in the preliminary emergency rescue action plan are compared with the real-time locations in the list of available rescue resources; Based on the channel network layer in the three-dimensional visualization layer of the building safety situation, considering the passage restrictions, damage and congestion of the channels, calculate the theoretical shortest path and estimated travel time for each available emergency rescue resource to reach the designated work point from its current location. Based on the estimated travel time and the resource's remaining capacity, assess whether the resource has the ability to be put into operation immediately upon arrival, and screen out resources that do not meet the conditions. For the same work site, if multiple available emergency rescue resources meet the criteria, they will be prioritized based on their functional capabilities and the degree of matching with the task, as well as the estimated passage time. For tasks involving multiple work locations and resources, an optimization model for resource scheduling and task allocation is established, with the objective function being the shortest overall task completion time and the highest resource utilization efficiency. The simulation calculation outputs a detailed time-resource-task association table, which is the optimized emergency rescue command task sequence; The emergency rescue command task sequence clearly defines the content of each specific task, the work location to be executed, the allocated resource units, the start and end times of the plan, and the dependencies between tasks.

6. The emergency rescue command system based on building safety requirements according to claim 1, characterized in that, The system also includes: The execution feedback module continuously receives on-site video streams, vital sign data, and environmental re-measurement data from the rescue robot and personnel's smart terminals during task execution, forming an on-site task execution feedback data stream, specifically: Establish a stable data transmission link with the intelligent terminals of all dispatched emergency rescue robots and personnel; It can receive real-time panoramic video streams and close-up video streams of the work site collected by the rescue robot through its multi-camera system, as well as the robot's own status monitoring data. It can receive real-time video streams from the perspective of personnel through the front-facing camera of the personnel's smart terminal, as well as vital signs data such as heart rate, blood oxygen, and body temperature of rescue personnel collected through wearable devices; The command system instructs emergency rescue robots and personnel intelligent terminals to re-measure the gas concentration, temperature, and radiation intensity of the local environment at the risk point through built-in or external sensors when they arrive at the work site or perform critical operations, and generate environmental re-measurement data. The received panoramic video stream and partial close-up video stream are synchronized in time and registered in space, and their images are associated with the corresponding positions in the three-dimensional visualization layer of the building safety situation. Real-time analysis of received vital sign data is performed to assess the physiological workload and safety status of rescue personnel. All transmitted video stream data, robot status data, vital sign data, and environmental retest data are fused and tagged according to time sequence, spatial location, and data source to form a multi-dimensional, spatiotemporally indexed on-site task execution feedback data stream.

7. The emergency rescue command system based on building safety requirements according to claim 6, characterized in that, The system also includes: The dynamic evaluation module integrates and analyzes the on-site task execution feedback data stream with the building safety status basic data stream to dynamically evaluate the status changes of risk points and the effectiveness of emergency response tasks. Specifically: Environmental retest data is extracted from the on-site task execution feedback data stream and compared with the original sensor array data of the risk point to determine whether the key parameters of gas concentration and temperature have changed to a safe range. Analyze the video stream in the on-site task execution feedback data stream, and use computer vision algorithms to identify the operation progress of the rescue robot or rescue personnel, including whether the support structure is installed in place and whether the leak point is sealed. By correlating newly observed cracks and debris falling from the structural surface in the video stream with the original deformation and vibration data, we can assess whether secondary or diffuse changes have occurred at the risk points. Analyze the vital signs data of rescue personnel in the on-site mission execution feedback data stream to assess personnel safety risks and determine whether replacement or reinforcement is needed. Based on the above analysis results, a dynamic assessment conclusion is generated for each risk point that is being addressed. The conclusion includes the degree of risk mitigation, the percentage of task completion, whether new risks have emerged, and the safety status of personnel. The dynamic assessment conclusions are compared with the original plan for the risk points in the emergency rescue command task sequence to calculate the deviation of task execution. Based on the dynamic assessment conclusions and the deviation of task execution, the labeling status of the corresponding risk points in the 3D visualization layer of the building safety situation is updated, including color gradient and progress bar display.

8. The emergency rescue command system based on building safety requirements according to claim 7, characterized in that, The system also includes: The real-time control module, based on the dynamic evaluation results, adjusts and replans the emergency rescue command task sequence in real time, generates updated navigation and operation instructions, and issues them out. Specifically: Receive dynamic assessment conclusions of risk points and reports on task execution deviations; If the dynamic assessment concludes that the risk has been largely eliminated, the corresponding task item is removed from the emergency rescue command task sequence, and the relevant resources are instructed to enter standby or return status. If the dynamic assessment results show that the task execution progress is lagging behind or encounters unforeseen difficulties, the resources and time required for the task will be reassessed, the time arrangement in the emergency rescue command task sequence will be adjusted, or additional resources will be allocated for support. If the dynamic assessment results show that new risk points have emerged or existing risks have spread, a new risk point location and level assessment process will be initiated, and the newly confirmed risk points and levels will be integrated into the three-dimensional visualization layer of the building safety situation. Based on the updated risk situation and resource status, the resource accessibility and timeliness simulation is re-executed to generate an adjusted emergency rescue command task sequence; Based on the adjusted emergency rescue command task sequence, navigation and operation instructions for newly added, changed, or continued tasks are re-decomposed and generated. Updated navigation and operation instructions will be prioritized and sent to the smart terminals of emergency rescue robots and personnel affected by the adjustments. The command platform interface highlights the changes to the task sequence and records the reasons for the adjustments and the basis for the decisions.

9. The emergency rescue command system based on building safety requirements according to claim 8, characterized in that, The system also includes: The report archiving module generates a complete emergency response report after confirming that the risk points have been effectively controlled or eliminated, updates the emergency plan database, and archives all process data. Specifically: When the dynamic assessment of all risk points shows that the risks have been eliminated or reduced to an acceptable level, and no new risks have emerged, the commander shall confirm the end of the emergency rescue operation. Automatically summarize all key data, decision records, instruction logs, and assessment reports generated throughout the entire process from the generation of basic data streams on building safety status to the final issuance of task adjustments; According to the preset report template, the summarized data is organized into chapters, which shall include at least the event overview, risk analysis process, command and decision sequence, resource allocation status, task execution details, verification of handling effect, and summary recommendations; Generate a complete emergency rescue operation report document with rich graphics and text, including data charts and key video screenshots; Based on the successful verification of the handling procedures and resource allocation experience in this operation, as well as the newly discovered characteristics of the hazards, the emergency plan database will be supplemented and revised, adding new case entries or optimizing the original plans. The complete emergency rescue operation report, the updated emergency plan library version, and all process data of the original building safety status basic data stream and on-site task execution feedback data stream are associated and packaged to obtain an associated data package; Assign a unique event number to the associated data packet, add metadata tags, and store it in the historical database of building safety emergency cases to complete the archiving.

Images