A bim-based urban and building emergency resource overall arrangement and deployment system

Abstract

The application belongs to the technical field of computer technology and emergency management, and specifically discloses a city and building emergency resource overall deployment system based on BIM, which comprises the following steps: system acquisition sensor data generates component state vector, converts local coordinates into absolute coordinates to generate a three-dimensional envelope and establish a spatial index; fusion road network and building channel to build a navigation network diagram to initialize the scheduling matrix; based on the state vector and the spatial index, the affected network edge is identified, the matrix composite weight value is dynamically updated, and the optimal path with the minimum comprehensive cost is solved in combination with the task weight. The system also realizes state comparison and weight rollback through log recording, and re-plans the local path of the affected task when the risk mutates. The application converts physical disasters into network constraints in real time, and improves the safety and adaptability of resource deployment in a complex dynamic environment.

Description

Technical Field

[0001] This invention belongs to the technical field of the intersection of computer technology and emergency management, and relates to a BIM-based urban and building emergency resource coordination and allocation system. Background Technology

[0002] In city-level emergency response scenarios, the core challenge of resource allocation lies in achieving cross-domain, collaborative deployment from the macro-scale of the city to the micro-scale of buildings under dynamically changing disaster environments. Currently, the technical solutions commonly used in the industry typically rely on Geographic Information Systems (GIS) for city-level route planning and independently use Building Information Modeling (BIM) or its two-dimensional drawings to manage the distribution of people and objects within buildings. This separated management model—where macro-level scheduling depends on the GIS road network and micro-level execution references BIM drawings—forms the basis for current emergency command and decision-making.

[0003] However, the aforementioned traditional methods have inherent technical limitations when dealing with complex disasters requiring detailed analysis down to the building's interior. First, GIS and BIM differ fundamentally in data structure, coordinate reference systems, and update mechanisms, leading to delays and semantic losses in data fusion. When a disaster occurs and the physical state inside a building (such as collapsed passageways or power outages) changes instantaneously, dispatch paths generated based on static GIS road networks cannot perceive these microscopic obstacles in real time, potentially causing dispatch instructions to be inconsistent with the physical situation. Second, converting BIM geometric information into GIS-recognizable network elements incurs computational overhead. Within the tight time window of emergency response, this data conversion delay can cause dispatch decisions to lag behind the actual evolution of the disaster.

[0004] Existing technologies attempt to abstract elements such as passageways and doors in BIM models into network nodes and edges through preprocessing and integrate them into a GIS road network database. This approach can provide a foundation for integrated indoor and outdoor path planning in static scenarios. However, in dynamic disaster scenarios, this approach has limitations. Its network topology is generated only once and cannot respond to real-time changes in component states sensed by physical sensors. For example, when the state of a BIM component (such as a fire door) changes from "open" to "closed," this change cannot be automatically and in real-time reflected in the corresponding network edge weights. The scheduling system will still perform calculations based on outdated network information, and the on-site usability of its planning results needs improvement.

[0005] Based on the above problems, the present invention aims to solve the problem that in emergency response, the dynamic changes in the physical state inside the building are not accurately mapped to the macro scheduling network model in real time, resulting in a discrepancy between the resource allocation plan and the actual situation on site. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention proposes the following technical solution: a BIM-based urban and building emergency resource coordination and allocation system, comprising: a vector generation module, which generates a state vector containing component identifiers, status codes, and damage measurement values ​​based on sensor data streams.

[0007] The spatial anchor point module obtains the local coordinates of the building information model components, converts them into absolute coordinates to generate a three-dimensional envelope, and binds the three-dimensional envelope with the component identifier and stores it in the spatial index table.

[0008] The matrix initialization module integrates urban road network and building internal passage topology data to construct a multi-layer navigation network graph, initializes the scheduling matrix, and assigns a composite initial weight value to each element in the matrix, including passage cost, time penalty, and safety risk coefficient.

[0009] The matrix reconstruction module retrieves the spatial index table based on the component identifier to determine the three-dimensional envelope, identifies the target network edges affected by the three-dimensional envelope, and modifies the composite weight value of the target network edges in the scheduling matrix based on the status code and damage metric.

[0010] The path calculation module obtains the task and the preset task weight coefficients, minimizes the comprehensive cost generated by the composite weight value and the task weight coefficients in the scheduling matrix, and calculates the optimal path.

[0011] The consistency verification module generates a transaction log when modifying the composite weight value, compares the subsequently obtained state vector with the transaction log, and rolls back the composite weight value in the scheduling matrix based on the comparison result.

[0012] The incremental replanning module monitors the scheduling matrix to identify local subgraphs where weights change. For tasks where the optimal path passes through a local subgraph, it replans and generates a locally optimized path starting from the current position. The locally optimized path is then combined with the optimal path to generate a solution.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention generates standardized building information model component state vectors and dynamically reconstructs the scheduling matrix based on these vectors, thereby enabling the real-time conversion of the anomalies of the underlying physical sensing signals into computable constraints of the upper-level scheduling network. This mechanism directly maps the component's state code and damage measurement value to the atomic modification of the network edge composite weight value through the state constraint mapping rule base. This process bypasses the traditional geometric transformation and data synchronization process between BIM and GIS models, enabling disaster events in the physical world to be reflected in the mathematical model almost synchronously, thereby improving the consistency between the network topology and the physical site state on which the scheduling decision is based.

[0014] (2) This invention constructs a multi-layered navigation network covering both indoor and outdoor areas and assigns composite initial weight values, including travel costs, time penalties, and safety risk coefficients, to the elements in the scheduling matrix. This provides a multi-dimensional decision-making benchmark for resource scheduling path calculation. During path optimization, the multi-objective optimization algorithm aims to minimize the comprehensive cost value determined by the composite weight value and the task type weight coefficient. This mechanism enables path calculation to consider not only distance or time but also safety risk as a core optimization variable, ensuring that the generated allocation scheme can proactively avoid areas marked as high-risk or blocked in the scheduling matrix, thus enhancing the safety of the scheduling scheme in complex disaster environments.

[0015] (3) This invention achieves bidirectional synchronization between the digital model and the physical site state by synchronously generating network fission transaction logs when the weights of the scheduling matrix are modified, and performing state consistency verification and repair based on these logs. This mechanism can identify restorative changes in the physical state of components by comparing subsequently received component state vectors with historical logs in real time, and trigger rollback operations on the corresponding weight values ​​in the scheduling matrix accordingly. This ensures that the scheduling network can automatically restore its original accessibility after physical obstacles are removed, realizing the dynamic self-healing capability of the scheduling model and improving the system's adaptability in continuous and changing disaster scenarios. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the system module connections of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0019] Please see Figure 1 As shown, the present invention proposes a BIM-based urban and building emergency resource coordination and allocation system, which includes: a vector generation module, a spatial anchor point module, a matrix initialization module, a matrix reconstruction module, a path calculation module, a consistency verification module, and an incremental replanning module.

[0020] The vector generation module generates a state vector containing component identifiers, status codes, and damage measurement values ​​based on the sensor data stream.

[0021] In a preferred embodiment, the vector generation module further includes: storing the raw sensor data stream into a first-in-first-out circular buffer, extracting data with a preset sampling period through a decision thread, and comparing the data with a preset logical threshold to determine the current physical properties;

[0022] The component's globally unique identifier is retrieved from a preset mapping table based on the sensor identifier, and a damage metric is calculated in combination with the current physical properties. This damage metric is a dimensionless value normalized to the range of 0 to 1.

[0023] Specifically, this embodiment is executed by a vector generation module deployed on edge computing nodes or a central server within the building. First, the vector generation module uses the MQTT or OPCUA protocol to collect raw data streams generated in real time from various sensors distributed throughout the building at a sampling frequency of 10Hz to 100Hz. These sensors include, but are not limited to, Hall effect sensors installed on access control systems, piezoelectric stress gauges embedded in load-bearing walls, or infrared thermal imaging sensors deployed within passageways.

[0024] Next, the vector generation module stores the received raw sensor data stream into a FIFO circular buffer and starts an independent decision thread. The decision thread extracts data from the most recent N sampling periods from the circular buffer and determines the current physical properties of the corresponding BIM component based on a preset logical threshold. For example, for a stress gauge's voltage signal, the system converts it into a stress value and compares it with a preset structural safety stress threshold; for a Hall sensor's switching signal, the system directly determines the door's open / closed state. After determining the current physical properties, the vector generation module retrieves and extracts the component's globally unique identifier, preset initial status code, and functional parameters from the local BIM component attribute mapping table based on the sensor ID. Subsequently, the system calculates and generates a damage metric based on the current physical properties and updates the initial status code to the current status code.

[0025] Damage measurement value It is a dimensionless value normalized to the interval between 0 and 1, used to quantify the degree to which a component deviates from its normal operating condition. Its calculation method varies depending on the sensor type and physical properties. For a load-bearing component monitored by a stress sensor, its damage metric is... The calculation process is as follows:

[0026]

[0027] In the formula, This represents the final calculated damage metric. This represents the current stress value, which is collected and converted in real time by the sensor. This represents the safe stress threshold defined in the design specifications for a component. When the stress is below this value, the component is considered to be in a completely safe state. This represents the ultimate stress threshold of a component. When the stress reaches or exceeds this value, the component is considered to have completely failed or is in an extremely dangerous state. The formula linearly maps the current stress value to the interval between 0 and 1 using a piecewise function. Furthermore, it has preset logical requirements. The system performs the initialization phase... and Perform a consistency check; if a consistency check is detected... If the preset threshold is found to have a logical anomaly, the system will automatically... Revised to ,in The preset non-zero constant offset (e.g., set to 0.1 MPa based on stress units, or taking...) (20% of the sum of a fixed scalar) to ensure the denominator is... It is always a non-zero positive number, thus eliminating the mathematical risk of dividing by zero.

[0028] Finally, the vector generation module encapsulates the extracted globally unique component identifier, updated status code, calculated damage metric, and extracted functional parameters, along with the current system time (compliant with ISO 8601) obtained from the operating system kernel, into a standardized building information model (BIM) component status vector with a fixed data structure. The vector generation module then broadcasts the BIM component status vector to the system bus via UDP or DDS protocols, either broadcast or multicast, providing a real-time trigger source for subsequent scheduling matrix reconstruction.

[0029] The Building Information Modeling (BIM) component state vector is a JSON object or a fixed-length binary structure containing the following fields: A globally unique identifier for the component is a UUID string conforming to RFC4122, used to uniquely identify a BIM component throughout the system. The status code is an integer whose value and meaning are defined in a preset status code table; for example, 0 represents normal, 1 represents closed or blocked, 2 represents open or connected, and 3 represents structural risk. The damage metric is a floating-point number calculated using the aforementioned formula, ranging from 0 to 1. The functional parameter is a numerical value or string that varies depending on the component type; for example, for a stress sensor, this parameter is the measured raw stress value in Pa; for a smart access control system, this parameter may be empty. The logical threshold is a key parameter for determining changes in component state, set based on relevant building design codes, material mechanical performance standards, or empirical values ​​derived from statistical analysis of historical operating data. For example, for a C30 concrete load-bearing wall, its safe stress threshold... The ultimate stress threshold can be set at 14.3 MPa according to the "Code for Design of Concrete Structures" GB50010. It can be set to 20.1MPa. Dedicated communication protocols specifically refer to communication technologies selected in a system's internal local area network environment to meet the requirements of low latency and high reliability. For example, using the User Datagram Protocol (UDP) for broadcasting to reduce connection establishment overhead, or using the Data Distribution Service (DDS) protocol to implement a topic-based publish-subscribe communication model.

[0030] For example, suppose the system monitors a load-bearing wall component with the ID "UUID-Wall-C3-F10", which is equipped with a stress sensor. According to building design codes, the safe stress threshold for this component... The ultimate stress threshold is set at 14.3 MPa. The pressure was set to 20.1 MPa. At the initial moment, the sensor collected and calculated the current stress value. It is 10.5 MPa. Since 10.5 MPa is less than... The system calculates the damage measurement value based on the set pressure of 14.3 MPa using a formula. The status code is 0. At this point, the system generates a status code of 0, indicating normal operation. Finally, the system generates and broadcasts the building information model component status vector to the bus as {“guid”:“UUID-Wall-C3-F10”,“statusCode”:0,“damageValue”:0.0,“funcParam”:“10.5MPa”,“timestamp”:“2023-10-27T10:30:00Z”}. Subsequently, due to the impact of the disaster, the sensors detected a continuous increase in stress, and the current stress value collected at a certain moment... It is 18.0 MPa. At this time, Between and In between, the system calculates the damage metric value according to the formula. Simultaneously, the system updates the status code to 3, representing structural risk, according to internal rules. Finally, the system generates and broadcasts a new building information model component status vector as {“guid”:“UUID-Wall-C3-F10”,“statusCode”:3,“damageValue”:0.638,“funcParam”:“18.0MPa”,“timestamp”:“2023-10-27T10:35:10Z”}.

[0031] The spatial anchor point module obtains the local coordinates of the building information model components, converts them into absolute coordinates to generate a three-dimensional envelope, and binds the three-dimensional envelope with the component identifier and stores it in the spatial index table.

[0032] In a preferred embodiment, the spatial anchor point module further includes: using a pre-calibrated parameter model in the project base point conversion algorithm to convert the local coordinates of the building information model components into absolute coordinates under the unified geographic coordinate system of the city;

[0033] Using an axis-aligned bounding box algorithm or a directed bounding box algorithm, a three-dimensional spatial envelope volume covering the geometric contour is generated based on the absolute coordinates, and the three-dimensional spatial envelope volume is stored in a spatial anchor point index table using an R-tree or quadtree structure.

[0034] Specifically, this embodiment is executed by the spatial anchor point module deployed on the central server. First, the spatial anchor point module retrieves the local coordinates and geometric contour information of the BIM component in the original design model, corresponding to the globally unique identifier of the component in the component's state vector, from the BIM design database or preprocessing cache. The local coordinates are typically three-dimensional Cartesian coordinates relative to the building project's base point, while the geometric contour information defines the component's precise three-dimensional shape, for example, described through boundary representation (B-Rep) or construction solid geometry (CSG).

[0035] Next, the spatial anchor point module invokes a project base point transformation algorithm to convert local coordinates into absolute coordinates under the city's unified geographic coordinate system. The core of this algorithm is the application of a pre-calibrated seven-parameter Bursaw-Wolf model or a simplified three-parameter translation model. This model includes the translation vector, rotation matrix, and scale factor from the BIM local coordinate system origin to the city's unified geographic coordinate system origin. The city's unified geographic coordinate system is a standard geographic reference system covering the entire urban area, such as the CGCS2000 national geodetic coordinate system or a local independent coordinate system. The transformation parameters required by the project base point transformation algorithm are calculated after the coordinates of the project base points under the city's unified geographic coordinate system are accurately determined using measurement methods such as GPS-RTK during the project planning or construction phase. After completing the coordinate transformation, the spatial anchor point module calculates and generates a three-dimensional spatial envelope based on the geometric contour information. The three-dimensional spatial envelope is composed of the absolute coordinates of its minimum and maximum vertices. and To reduce the complexity of subsequent spatial calculations and improve real-time performance, this calculation process typically employs axis-aligned bounding box (AABB) or directed bounding box (OBB) algorithms to generate the most compact cube or cuboid that can completely encompass the original complex geometric contour.

[0036] Finally, the spatial anchor module binds the calculated 3D spatial envelope to the globally unique identifier of the component and stores it as a new record in a spatial anchor index table. The spatial anchor index table is an efficient spatial index database organized using an R-tree or quadtree data structure. Its key is the globally unique identifier of the component, and its value is a data structure containing the coordinates of the 3D spatial envelope and related attributes. This table provides fast retrieval capabilities for subsequent spatial intersection calculations.

[0037] For example, following the aforementioned steps, when the spatial anchor point module receives the building information model component state vector {“guid”:“UUID-Wall-C3-F10”,…}, the system initiates this step based on the component's globally unique identifier “UUID-Wall-C3-F10”.

[0038] The system retrieves the geometric information of component "UUID-Wall-C3-F10" from the BIM database. Assume its geometric profile in the BIM local coordinate system is defined by two vertices, with local coordinates of... Mihe rice.

[0039] The system invokes a preset project base point transformation algorithm. The algorithm's transformation parameter library stores transformation parameters specific to this building project. Assuming the translation vector is... Meters. The system performs a coordinate transformation and calculates the absolute coordinates as follows: rice.

[0040] rice.

[0041] Thus, the system generates a three-dimensional spatial envelope, which consists of the smallest vertex. and the largest vertex It is uniquely determined under the city's unified geographical coordinate system.

[0042] Finally, the system stores this entry in the spatial anchor index table, forming a record: {“key”:“UUID-Wall-C3-F10”,“value”:{“envelope”:[(354130.0,4186525.0,75.5),(354130.2,4186535.0,78.5)]}}. This record establishes a mapping benchmark between building information model components and their precise physical locations in macro-geographic space.

[0043] The matrix initialization module integrates urban road network and building internal passage topology data to construct a multi-layer navigation network graph, initializes the scheduling matrix, and assigns a composite initial weight value to each element in the matrix, including passage cost, time penalty, and safety risk coefficient.

[0044] In a preferred embodiment, the matrix initialization module further includes: uniformly converting the node coordinates in the building internal passage topology data to the unified urban geographic coordinate system, identifying building entrance and exit nodes, and connecting the building entrance and exit nodes with the geographically closest nodes in the urban road network data through virtual connecting edges to form a connected multi-layer navigation network graph.

[0045] The passage cost is calculated based on the edge length of the multi-layer navigation network graph and the preset average passage speed. A time penalty is set, and the initial safety risk coefficient corresponding to the environmental attribute is queried from the preset safety benchmark table to generate a composite initial weight value.

[0046] Specifically, in this embodiment, the matrix initialization module, deployed on a central server, performs the initialization. First, the matrix initialization module integrates urban road network data and building internal passageway topology data from the urban transportation information center database and the building design database, respectively, through standard data interfaces. The urban road network data is presented in a diagram... The form in which nodes Represents the endpoints of intersections or key road segments, edges Representing roads, their attributes include length, speed limit, etc. Internal building passageway topology data is presented in a graph. The form in which nodes Represents a room, stairwell, or elevator entrance, side This represents a corridor, staircase, or elevator shaft, and its attributes include passage width, vertical height variation, etc.

[0047] Next, the matrix initialization module constructs a multi-layered navigation network graph covering both indoor and outdoor areas through node fusion and edge connection operations within the city's unified geographic coordinate system. Specifically, the system uses the aforementioned project base point transformation algorithm to uniformly transform the coordinates of all nodes in the building's internal passageway topology data to a unified urban geographic coordinate system. Subsequently, the system identifies the building's entrance and exit nodes and connects these nodes with the geographically nearest nodes in the urban road network data using virtual connecting edges, thus forming a unified, connected graph structure.

[0048] Then, the matrix initialization module initializes and generates a scheduling matrix (DFM) based on the node connectivity relationships of the multi-layer navigation network graph. This matrix is ​​a... The square array, in which It is a multi-layer navigation network diagram Total number of nodes in the matrix. Matrix elements. Indicates from node to adjacent nodes The feasibility. If the node and nodes If there is an edge between them, then An initial composite weight value is assigned; if there are no directly connected edges, the travel cost in its composite weight value is set to infinity, indicating that it is unreachable.

[0049] Finally, the matrix initialization module initializes each passable element in the scheduling matrix. Assign a composite initial weight value that includes passage cost, time penalty, and safety risk coefficient. Passage cost is primarily calculated based on the physical length of the edge and the average passage speed. The time penalty is an initial zero-based additional delay used for subsequent dynamic adjustment. The safety risk coefficient is a basic risk assessment value set based on historical data or environmental type; for example, the initial risk coefficient for an open outdoor road is lower than that for a narrow indoor passage.

[0050] The composite initial weight value is a structured data object, and its specific calculation method is as follows: For connection nodes and nodes edge Its toll costs The calculation formula is ,in It is a connection node and nodes The length of the side, in meters; This is the average travel speed, measured in meters per second, for the environment type of the edge connecting the i-th and j-th nodes. This value is obtained from the system's preset travel rule base based on the edge type (e.g., urban main road, indoor corridor). Time penalty The initial value is set to 0. (Edge) Safety risk coefficient The initial values ​​are obtained from the safety baseline table based on the edge's environmental attributes. To ensure the scientific nature of subsequent path optimization, the system establishes a baseline normalized coefficient library, storing the normalized coefficients of the travel cost baseline for edges in various environments. For example, for an indoor corridor, if the maximum expected travel time is preset to 20 seconds, then the corresponding... Set to 20.0.

[0051] For example, following the preceding steps, the system begins to construct a multi-layered navigation network map that includes buildings. The system loads a data set containing nodes from the city's transportation database. The city road network was analyzed, and data containing nodes was loaded from the BIM database. The system retrieves the building's internal passageway topology. It queries the traffic rule base and finds that the average travel speed on city roads is 15 meters per second, while the average travel speed in indoor corridors is 1.5 meters per second.

[0052] The system transforms the coordinates of all nodes inside the building to a unified urban geographic coordinate system and identifies the main entrance node of the building. It is compared with the nearest urban road node in terms of geographical location. Connected by a 5-meter-long virtual pedestrian walkway. This generates a [structure / feature]. A multi-layered navigation network diagram with nodes.

[0053] The system initializes a based on this. The scheduling matrix. For a multi-layer navigation network graph, a connection between indoor nodes... and The corridor side is 12 meters long. The system calculates its composite initial weight values ​​as follows: . The system queries the safety benchmark table and finds that the default safety risk coefficient for the indoor corridor is 0.1.

[0054] Therefore, the corresponding element in the scheduling matrix The initial values ​​are set to {cost:8.0, penalty:0.0, risk:0.1}.

[0055] For pairs of nodes in the matrix that are not directly connected by an edge, for example Its initial value is set to {cost:Infinity,penalty:0.0,risk:0.0}.

[0056] By performing this process on all node pairs, the system completes the initialization of the scheduling matrix, establishing a bottom-level topology benchmark for subsequent dynamic reconstruction and path calculation based on the state vectors of building information model components.

[0057] The matrix reconstruction module determines the three-dimensional envelope by retrieving the spatial index table based on the component identifier, identifies the target network edge affected by the three-dimensional envelope, and modifies the composite weight value of the target network edge in the scheduling matrix based on the status code and damage metric value.

[0058] In a preferred embodiment, the matrix reconstruction module further includes: representing network edges in the multi-layer navigation network graph as three-dimensional spatial line segments, and applying the separating axis theorem to determine whether the three-dimensional spatial line segments geometrically intersect with the three-dimensional spatial envelope, so as to determine the target network edge set;

[0059] The system calls a pre-defined state constraint mapping rule library, matches the physical risk transformation factor corresponding to the state code, and updates the composite weight value of the target network edge set in the scheduling matrix through atomic modification operations.

[0060] Specifically, this embodiment is executed by the matrix reconstruction module deployed on the central server. First, the matrix reconstruction module continuously listens to the system bus and receives the building information model component status vector broadcast by the vector generation module. Following the aforementioned steps, the matrix reconstruction module receives a building information model component status vector with the content {“guid”:“UUID-Wall-C3-F10”,“statusCode”:3,“damageValue”:0.638,“funcParam”:“18.0MPa”,“timestamp”:“2023-10-27T10:35:10Z”}.

[0061] Next, the matrix reconstruction module retrieves the corresponding 3D spatial envelope from the spatial anchor index table based on the globally unique identifier "UUID-Wall-C3-F10" of the component in the building information model component state vector. Following the previous steps, the system successfully retrieves the record {"key":"UUID-Wall-C3-F10","value":{"envelope":[(354130.0,4186525.0,75.5),(354130.2,4186535.0,78.5)]}}, thus obtaining the 3D spatial envelope of the component in the city's unified geographic coordinate system.

[0062] Then, the matrix reconstruction module performs spatial intersection calculations between the 3D spatial envelope and each network edge in the multi-layer navigation network graph to identify the target network edge set affected by the building information model components. This calculation process involves traversing each edge in the multi-layer navigation network graph, representing it as a line segment in 3D space, and applying the separating axis theorem or the AABB intersection detection algorithm to determine whether the line segment geometrically intersects with the 3D spatial envelope. All intersecting network edges are added to a temporary target network edge set.

[0063] Finally, the matrix reconstruction module invokes a pre-defined state constraint mapping rule base and, based on the state codes and damage metrics in the building information model component state vectors, atomically modifies the composite weight values ​​in the scheduling matrix corresponding to the target network edge set. This atomic modification ensures that the scheduling matrix does not enter an inconsistent state during the modification process, guaranteeing data integrity.

[0064] The state constraint mapping rule base is a database that stores a series of condition-action rules. Each rule defines how the composite weight value of the corresponding element in the scheduling matrix should be modified when the state vector of a building information model component meets specific conditions.

[0065] For example, following the steps described above, the system continues execution. The system performs spatial intersection calculations and discovers indoor nodes connected in the multi-layer navigation network graph. and The corridor edge, whose geometric path passes through the three-dimensional spatial envelope of the component "UUID-Wall-C3-F10", is therefore identified and added to the target network edge set.

[0066] The system invokes the state constraint mapping rule base and performs rule matching based on the received building information model component state vectors, where "statusCode" is 3 and "damageValue" is 0.638. The system matches the following rules in the rule base:

[0067] Rule IDR-03. Condition. If the component type is "load-bearing wall" and the status code is 3 (structural risk), then execute the action.

[0068] Action. For all affected target network edges, the security risk coefficient in their composite weight values. According to the formula Update, among which It is the original safety risk coefficient. It is the damage metric value in the component state vector of the building information model. This is a physical risk conversion factor used to quantify the degree to which physical damage to components translates into network access risk. Additionally, a synchronization update time penalty value is included. ,in A maximum passage delay benchmark is preset for this type of path, thereby closing the logic of physical damage into the passage efficiency loss.

[0069] The system retrieves the physical risk conversion factor for structural risk events from the configuration database. The value is 0.8. The system then begins to atomically modify the scheduling matrix corresponding to the nodes. and elements and .

[0070] Before the modification, The composite weight value is {cost:8.0, penalty:0.0, risk:0.1}.

[0071] The system calculates the new security risk coefficient based on the matching rules and formulas. .

[0072] The system will The composite weight value is updated to {cost:8.0, penalty:0.0, risk:0.6104}. Since this edge is bidirectional, The same update is also performed.

[0073] Through this process, the local weights of the scheduling matrix are dynamically reconstructed, successfully transforming the physical disaster events of the building information model components (excessive stress in load-bearing walls) into mathematical constraints that the scheduling algorithm can calculate (significantly increased traffic risk), providing a real-time updated decision-making basis for the subsequent generation of disaster-adaptive allocation schemes.

[0074] The path calculation module obtains the task and the preset task weight coefficient, minimizes the comprehensive cost generated by the composite weight value and the task weight coefficient in the scheduling matrix, and calculates the optimal path.

[0075] In a preferred embodiment, the path calculation module further includes: retrieving task type weight coefficients that match the emergency resource scheduling task from a preset task type configuration library, wherein the task type weight coefficients include cost weight, penalty weight and risk weight.

[0076] By using a multi-objective optimization algorithm, the passage cost and time penalty in the composite weight value are divided by the normalization coefficient of the passage cost benchmark and the normalization coefficient of the time penalty benchmark under the corresponding environment type, respectively, so as to achieve dimensionless processing of the indicators of each dimension in the same order of magnitude range. Then, combined with the safety risk coefficient, the corresponding cost weight, penalty weight and risk weight are multiplied and summed to obtain the comprehensive cost value.

[0077] Specifically, in this embodiment, the path calculation module is deployed on a central server. First, the path calculation module obtains the target start and end point information of the emergency resource dispatching task to be executed from the emergency task management system. Following the previous steps, suppose the system receives an emergency medical rescue task whose target start point is a medical center in the city, and the corresponding node in the multi-layer navigation network graph... Its target destination is a room inside the disaster-stricken building, corresponding to the node. Simultaneously, the path calculation module determines the task type weight coefficient of the emergency resource dispatching task from the system's built-in task type configuration library based on the type of the task. For this emergency medical rescue task, the system queries the task type configuration library and retrieves the corresponding task type weight coefficient. This combination of coefficients indicates that this task prioritizes security, followed by travel costs.

[0078] Next, the path calculation module uses the scheduling matrix as the sole basis for path search and employs a multi-objective optimization algorithm to calculate the optimal path from the target origin to the target destination. This multi-objective optimization algorithm is an improved version of the A* algorithm, which, when evaluating the heuristic function and actual cost of each node, no longer uses a single distance or time value, but instead uses a comprehensive cost value.

[0079] The optimal path calculation process minimizes the comprehensive cost generated by weighting the composite weight value and the task type weight coefficient, ensuring that the path scheme avoids high-risk or blocking areas marked in the scheduling matrix. For connecting any adjacent nodes in a multi-layer navigation network graph... and nodes The edge, its comprehensive value The calculation formula is as follows:

[0080]

[0081] In the formula, For the normalization function, use The saturation process is performed in a saturation manner to ensure that individual cost items do not grow indefinitely due to excessively long paths, thereby guaranteeing that... When the risk level is high, the safety risk factor is... It can play a leading role in correcting path decisions. It represents the overall value of that side. , and These represent the task type weight coefficients obtained from the task type configuration library. They are retrieved from the preset task type configuration library based on the characteristics of the current emergency task. Their values ​​are in the range of [0,1] and the sum of them with the other weight coefficients is 1. For example, the values ​​are set to 0.3, 0.2, and 0.5 respectively. , and These represent the elements of the scheduling matrix. The values ​​of passage cost, time penalty, and security risk coefficient are read from the data. and These represent the normalization coefficients for the toll cost baseline and the time penalty baseline, respectively. The key technical logic lies in scaling large absolute values ​​of time costs (e.g., 8-100 seconds) to a dimensionless proportional value within the range [0,1] through division of these baseline values. At this point, when... When set to 0.5, the safety risk coefficient The changes will affect This has a dominant influence, thus mathematically supporting the "safety first" scheduling strategy.

[0082] The normalization function used in this invention Unified definition The physical meaning of using an upper limit cutoff value of 2 is that when the actual toll cost or time penalty exceeds twice the benchmark value, the marginal contribution of this item to the overall cost reaches saturation, thus preventing a single extreme road condition from infinitely amplifying the cost deviation. The leading role of [the government / organization] ensures that emergency resources always prioritize the safest route.

[0083] For example, following the aforementioned steps, the path calculation module begins by working with the slave node. To the node The optimal path is calculated for emergency medical rescue missions. The path calculation module initiates a multi-objective optimization algorithm, using a scheduling matrix as the search space. When the algorithm explores a connecting node... and When determining the path, it needs to calculate the comprehensive cost of that path segment.

[0084] The algorithm reads from the scheduling matrix The composite weight value, which has been dynamically reconstructed in the aforementioned steps, is {cost:8.0, penalty:0.0, risk:0.6104}.

[0085] The algorithm uses the task type weight coefficient of the current task. and call the normalization coefficients (Indoor corridor scene), calculate the comprehensive cost based on the formula: .

[0086] Meanwhile, the algorithm explores another alternative path, assuming that this path passes through the connecting nodes. and An edge, with a composite weight of {cost:10.0, penalty:0.0, risk:0.1} in the scheduling matrix, represents a longer channel unaffected by structural risk. The algorithm calculates its comprehensive cost: .

[0087] In another scenario, if the task type is rapid material transportation, its task type weighting coefficient might be: Then the combined cost of the two paths mentioned above will become: . In this case, although path B5→B6 carries structural risk (the overall cost component of this contribution is...), However, due to the cost weighting of this task ( Furthermore, the normalized advantage in time efficiency of this path (0.28 vs. 0.35) offsets its risk disadvantage, thus the overall cost of path B5→B6 remains low. This demonstrates that the system can make a quantitative and scientific trade-off between safety and efficiency based on different task characteristics.

[0088] Returning to the scenario of emergency medical rescue missions, the multi-objective optimization algorithm, during its iterative process, due to... Increasing the overall cost to 0.5 makes the overall cost of path B5→B6 (0.4252) significantly higher than that of path B5→B7 (0.20), thus favoring other paths with lower overall costs, even if these paths have higher initial travel costs. Ultimately, the algorithm outputs an optimal path composed of a sequence of nodes that minimizes the overall cost globally, generating a disaster-adaptive allocation scheme that considers both efficiency and proactively avoids the physical risks revealed by the state vectors of building information model components.

[0089] The consistency verification module generates a transaction log when modifying the composite weight value, compares the subsequently obtained state vector with the transaction log, and rolls back the composite weight value in the scheduling matrix based on the comparison result.

[0090] In a preferred embodiment, the consistency verification module further includes: a copy of the building information model component state vector that records the modification time, the affected edge identifier, and the modification of the composite weight value in the network fission transaction log;

[0091] When the status code and damage metric value in the state vector of a newly received building information model component are lower than the corresponding values ​​recorded in the network fission transaction log, a restorative change is determined to have occurred, and the composite weight value corresponding to the affected edge identifier in the scheduling matrix is ​​restored to the composite initial weight value.

[0092] Specifically, in this embodiment, the consistency verification module is deployed on a central server. First, whenever the scheduling matrix undergoes a weight modification, such as during the reconstruction process described in the previous steps due to the receipt of the building information model component state vectors,... When the composite weight value is determined, the consistency verification module synchronously generates a network fission transaction log containing modification time, affected edge identifiers, and trigger vector information. The network fission transaction log is an immutable sequence of records stored in chronological order, used to track all dynamic changes in the scheduling matrix.

[0093] Next, the consistency verification module compares in real time the state vectors of subsequently received building information model components with globally unique identifiers for the same component with historical records in the network fission transaction log. This comparison process aims to detect potential inconsistencies between the physical world state and the digital model state.

[0094] Finally, if the consistency verification module determines that the physical state of the building information model component has undergone a restorative change, it will complete the dynamic repair of the network topology according to the corresponding composite weight value in the rollback scheduling matrix of the network fission transaction log.

[0095] Each record in the network fission transaction log is a data structure containing the following fields: Modification Time, which is the system timestamp when the weight modification was performed; Affected Edge Identifier, which is the unique identifier of the network edge corresponding to the modified scheduling matrix element; Trigger Vector Information, which is a copy of the complete building information model component state vector that triggered this modification.

[0096] Restorative change refers to the transformation of a component's state from abnormal or damaged to normal. The determination is based on the fact that the status code and damage metric in the newly received building information model component state vector are significantly lower than the corresponding values ​​in the trigger vector information that triggered the increase in the edge weight recorded in the network fission transaction log.

[0097] For example, following the steps described above, the system continues to run. In the aforementioned steps, when the elements of the scheduling matrix... When the status vector of a Building Information Model (BIM) component, {“guid”:“UUID-Wall-C3-F10”,“statusCode”:3,“damageValue”:0.638,…}, is modified, the consistency verification module synchronously generates and records a network fission transaction log. The log content is {logTimestamp:“2023-10-27T10:35:11Z”,affectedEdgeID:“Edge-B5-B6”,triggerVector:{“guid”:“UUID-Wall-C3-F10”,“statusCode”:3,“damageValue”:0.638,…}}. At this time, The composite weight value is {cost:8.0, penalty:0.0, risk:0.6104}.

[0098] After a period of time, assuming the on-site emergency response measures take effect and the stress in the load-bearing wall is relieved, sensors deployed on the component collect new current stress values. The value is 12.0 MPa. This value is below the safe stress threshold. At 14.3 MPa, the vector generation module calculates a new damage metric. The status code is set to 0, and the status code is updated to 0 (normal). Based on this, the system generates and broadcasts a new building information model component status vector, the content of which is {"guid":"UUID-Wall-C3-F10","statusCode":0,"damageValue":0.0,"funcParam":"12.0MPa","timestamp":"2023-10-27T11:00:00Z"}.

[0099] The consistency verification module receives this new Building Information Model (BIM) component state vector. Based on the component's globally unique identifier "UUID-Wall-C3-F10", it retrieves the aforementioned record from the network fission transaction log.

[0100] The system performed a comparison and found that the status code in the new building information model (BIM) component's status vector was 0, which was much lower than the status code 3 recorded in the trigger vector information in the log; at the same time, the new damage metric value was 0.0, which was also much lower than the 0.638 recorded in the log. Based on this, the system determined that the physical state of the BIM component had undergone a restorative change.

[0101] The consistency verification module triggered a rollback operation. It located the element in the scheduling matrix based on the affected edge identifier "Edge-B5-B6" in the network fission transaction log. The system retrieves the composite initial weight value of this type of edge from the configuration library, i.e., {cost:8.0, penalty:0.0, risk:0.1}.

[0102] The system will The composite weight value is rolled back from the current {cost:8.0,penalty:0.0,risk:0.6104} to its initial composite weight value {cost:8.0,penalty:0.0,risk:0.1}.

[0103] Through this process, the system completed the state consistency verification and dynamic repair of the multi-layer navigation network, ensuring that the digital model of the scheduling matrix can keep pace with the state changes of the physical site, and avoiding persistent deviations in path planning caused by failure to update in time due to state recovery.

[0104] The incremental replanning module monitors the scheduling matrix to identify local subgraphs with changing weights. For tasks whose optimal path passes through a local subgraph, it replans and generates a locally optimized path starting from the current position. The locally optimized path is then combined with the optimal path to generate a solution.

[0105] In a preferred embodiment, the incremental replanning module further includes: calculating the increment of the security risk coefficient in the scheduling matrix in a single update; if the increment exceeds a preset risk mutation threshold, then identifying the network edge with changed weight and the area covered by the predetermined number of hops extending outward as a local subgraph region.

[0106] The real-time location of the affected scheduled task is obtained through the positioning system. The real-time location is used as the new starting point, and the first node in the original path located outside the local subgraph area is used as the temporary endpoint for local path replanning.

[0107] Specifically, this embodiment is executed by the incremental replanning module deployed on the central server. First, the incremental replanning module continuously monitors update events of the scheduling matrix and identifies local subgraph regions where weights have undergone drastic changes. A drastic weight change refers to the situation where the composite weight value of an element in the scheduling matrix increases in a single update event beyond the system's preset risk mutation threshold. The risk mutation threshold is a sensitivity parameter used to trigger replanning; its value is read from the system's emergency response plan configuration library and dynamically set according to the current emergency response level. A local subgraph region is a set of nodes and edges covered by a predetermined number of hops, centered on the network edge where the weight has drastically changed.

[0108] Next, the incremental replanning module retrieves the currently executing affected scheduled tasks whose paths pass through local subgraph regions, and extracts the current position of the affected scheduled tasks as the new starting point.

[0109] Finally, the incremental replanning module performs local path replanning only for local subgraph regions, and concatenates the generated local optimized path with the unaffected part of the original path to output the updated resource scheduling scheme.

[0110] For example, following all the preceding steps, the system continues execution. In the preceding steps, due to receiving the building information model component status vector {“guid”:“UUID-Wall-C3-F10”,“statusCode”:3,“damageValue”:0.638,…}, the elements of the scheduling matrix… The safety risk coefficient in the composite weight value jumps from 0.1 to 0.6104, an increment of 0.5104. The system reads the risk mutation threshold set at the current emergency response level from the emergency response plan configuration library, which is 0.4. Since 0.5104 is greater than 0.4, the incremental replanning module determines the connection node... and The weights of the edges changed drastically, and the region formed by the edge and its adjacent nodes was identified as a local subgraph region.

[0111] At this point, suppose there is an emergency resource scheduling task in progress, and its original optimal path is... The incremental replanning module retrieved the task and found that its planned path contained edges with drastically changed weights. .

[0112] The incremental replanning module obtains the real-time location of the task execution unit through GPS or an indoor positioning system, assuming that it has already reached the node. The system then extracts the current location node of the task. As a new starting point.

[0113] The incremental replanning module initiates local path replanning. It starts with nodes. Starting from the first node in the original path that is outside the local subgraph region, As a temporary endpoint, under the constraints of the currently updated scheduling matrix, recalculate a route from... arrive The optimal path. Because The safety risk coefficient in the composite weight value has increased significantly, and the newly calculated local optimal path is likely to be a path that bypasses the edge. The path, for example .

[0114] Finally, the incremental replanning module will generate the locally optimized path. This is concatenated with the unaffected portion of the original path. The executed portion of the original path is... The unaffected subsequent parts are After concatenation, the system outputs the updated resource scheduling scheme, the path of which is... .

[0115] Through this process, the system achieves rapid response and adaptive path adjustment to local emergencies without interrupting tasks or performing global recalculation, thereby improving the system's response efficiency under continuous disasters.

[0116] In a further preferred embodiment, an archiving module is also included: archiving the resource scheduling schemes generated by completed or terminated emergency resource scheduling tasks in association with the network fission transaction logs generated during the execution period;

[0117] Extract path nodes, timestamps, comprehensive cost value, and component state evolution sequences to construct structured event records and import them into a knowledge graph database to establish causal links between disaster events and scheduling decisions.

[0118] Specifically, this embodiment is executed by the archiving module deployed on the central server. First, the archiving module monitors the task status stream of the emergency task management system. When it detects that the status of any emergency resource scheduling task has changed to "completed" or "abnormally aborted," the system immediately locks the lifecycle time window of that task. The system extracts the resource scheduling plan for the final execution of the task from the task database. This plan includes the actual sequence of path nodes, the actual travel time for each road segment, and the final accumulated comprehensive cost.

[0119] Next, the archiving module uses the lifecycle time window as the search condition to batch retrieve all network fission transaction logs generated within that time period from historical storage. The system uses a time alignment algorithm to spatiotemporally correlate the spatial location in the path node sequence with the affected edge identifiers in the network fission transaction logs, filtering out the key log set that directly affects the scheduling decision.

[0120] Then, the archiving module performs feature extraction and structured reorganization. The system extracts trigger vector information from the key log set, i.e., copies of the building information model component state vectors, to form a sequence of component state evolution; at the same time, it extracts path decision points from the resource scheduling scheme. The system encapsulates the above information into a structured event record (e.g., using JSON-LD format), which includes: the source of the disaster (component ID and damage metric), the affected network area (set of edge IDs), the scheduling decision response (replanning actions and path changes), and the final consequences (task success or failure and cost).

[0121] Finally, the archiving module connects to a pre-defined knowledge graph database (such as Neo4j or RDF storage). The system maps structured event records to entity nodes ("disaster event nodes" and "scheduling decision nodes") and relational edges ("causing", "responding to") in the graph. By establishing causal links between disaster events and scheduling decisions, the system provides traceable, high-quality sample data for subsequent disaster simulation model training and automated optimization of emergency plans.

[0122] In a further preferred embodiment, the incremental replanning module is further used to: trigger incremental replanning logic in real time when the matrix reconstruction module performs atomic modifications that cause a sudden change in the safety risk coefficient, and realize a closed-loop response to physical disaster events through dynamic adjustment of local paths.

[0123] Specifically, this embodiment describes the event-driven and message-passing mechanism between modules within the system. First, the matrix reconstruction module incorporates a difference monitor when performing atomic modification operations on the scheduling matrix. Before each write operation, this monitor calculates the difference between the old and new composite weight values ​​of the target network edge, particularly the increment of the security risk coefficient. .

[0124] Next, when the calculated When the risk mutation threshold (e.g., 0.4) is exceeded, the matrix reconstruction module does not directly call the replanning function, but instead publishes a high-priority "risk blocking event" through the system's internal message bus. The event payload includes the affected edge identifier, the new weight value after the mutation, and the timestamp of the event occurrence.

[0125] Then, the incremental replanning module, as a resident subscriber to the message bus, captures the "risk blocking event" in real time. Once captured successfully, the incremental replanning module immediately interrupts non-urgent background computing tasks and uses the affected edge identifier in the event payload to quickly retrieve all currently executing active scheduled tasks whose planned paths contain that edge.

[0126] Finally, for each affected task retrieved, the incremental replanning module takes its current real-time location as the starting point and nodes in the original path located outside the local risk area as anchor points, and performs a local path search (such as the D*Lite algorithm or the local A* algorithm) on the updated scheduling matrix. The generated locally optimized path will directly replace the blocked part in the original path, and the updated instructions will be sent to the front-end terminal.

[0127] Through the above mechanism, the system achieves a millisecond-level closed-loop response from "physical sensor perception" → "state vector generation" → "matrix weight atomization modification" → "event release" → "incremental replanning", ensuring that the emergency resource allocation plan can always be dynamically synchronized with the evolution of physical disasters inside the building.

[0128] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, and all such modifications and additions should fall within the protection scope of the present invention.

Claims

1. A BIM-based urban and building emergency resource coordination and allocation system, characterized in that, include: The vector generation module generates a state vector containing component identifiers, status codes, and damage measurement values ​​based on the sensor data stream. The spatial anchor point module obtains the local coordinates of the building information model components, converts them into absolute coordinates to generate a three-dimensional envelope, and binds the three-dimensional envelope with the component identifier and stores it in the spatial index table. The matrix initialization module integrates urban road network and building internal passage topology data to construct a multi-layer navigation network graph, initializes the scheduling matrix, and assigns a composite initial weight value to each element in the matrix, including passage cost, time penalty, and safety risk coefficient. The matrix reconstruction module retrieves the spatial index table based on the component identifier to determine the three-dimensional envelope, identifies the target network edges affected by the three-dimensional envelope, and modifies the composite weight value of the target network edges in the scheduling matrix based on the status code and damage metric. The path calculation module obtains the task and the preset task weight coefficients, minimizes the comprehensive cost generated by the composite weight value and the task weight coefficients in the scheduling matrix, and calculates the optimal path. The consistency verification module generates a transaction log when modifying the composite weight value, compares the subsequently obtained state vector with the transaction log, and rolls back the composite weight value in the scheduling matrix based on the comparison result. The incremental replanning module monitors the scheduling matrix to identify local subgraphs where weights change. For tasks where the optimal path passes through a local subgraph, it replans and generates a locally optimized path starting from the current position. The locally optimized path is then combined with the optimal path to generate a solution.

2. The BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The vector generation module further includes: The raw data stream from the sensor is stored in a first-in-first-out circular buffer. Data from a preset sampling period is extracted by a decision thread and compared with a preset logical threshold to determine the current physical properties. The component's globally unique identifier is retrieved from a preset mapping table based on the sensor identifier, and a damage metric is calculated in combination with the current physical properties. This damage metric is a dimensionless value normalized to the range of 0 to 1.

3. The BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The spatial anchor point module further includes: By using the pre-calibrated parameter model in the project base point transformation algorithm, the local coordinates of the building information model components are converted into absolute coordinates under the unified geographic coordinate system of the city; Using an axis-aligned bounding box algorithm or a directed bounding box algorithm, a three-dimensional spatial envelope volume covering the geometric contour is generated based on the absolute coordinates, and the three-dimensional spatial envelope volume is stored in a spatial anchor point index table using an R-tree or quadtree structure.

4. The BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The matrix initialization module further includes: The node coordinates in the internal passage topology data of the building are uniformly transformed to the unified urban geographic coordinate system, the building entrance and exit nodes are identified, and the building entrance and exit nodes are connected to the nodes with the closest geographical location in the urban road network data through virtual connecting edges to form a connected multi-layer navigation network map. The passage cost is calculated based on the edge length of the multi-layer navigation network graph and the preset average passage speed. A time penalty is set, and the initial safety risk coefficient corresponding to the environmental attribute is queried from the preset safety benchmark table to generate a composite initial weight value.

5. A BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The matrix reconstruction module further includes: The network edges in the multi-layer navigation network graph are represented as three-dimensional spatial line segments. The separating axis theorem is applied to determine whether the three-dimensional spatial line segments geometrically intersect with the three-dimensional spatial envelope, so as to determine the target network edge set. The system calls a pre-defined state constraint mapping rule library, matches the physical risk transformation factor corresponding to the state code, and updates the composite weight value of the target network edge set in the scheduling matrix through atomic modification operations.

6. A BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The path calculation module further includes: Retrieve the task type weight coefficient that matches the emergency resource dispatch task from the preset task type configuration library. The task type weight coefficient includes cost weight, penalty weight and risk weight. By using a multi-objective optimization algorithm, the passage cost and time penalty in the composite weight value are divided by the normalization coefficient of the passage cost benchmark and the normalization coefficient of the time penalty benchmark under the corresponding environment type, respectively, so as to achieve dimensionless processing of the indicators of each dimension in the same order of magnitude range. Then, combined with the safety risk coefficient, the corresponding cost weight, penalty weight and risk weight are multiplied and summed to obtain the comprehensive cost value.

7. A BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The consistency verification module further includes: Record the modification time, affected edge identifier, and a copy of the state vector of the building information model component that triggered the modification of the composite weight value in the network fission transaction log; When the status code and damage metric value in the state vector of a newly received building information model component are lower than the corresponding values ​​recorded in the network fission transaction log, a restorative change is determined to have occurred, and the composite weight value corresponding to the affected edge identifier in the scheduling matrix is ​​restored to the composite initial weight value.

8. A BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The incremental replanning module further includes: Calculate the increment of the safety risk coefficient in the scheduling matrix in a single update. If the increment exceeds the preset risk mutation threshold, the network edge with the changed weight and the area covered by the predetermined number of hops outward are identified as a local subgraph region. The real-time location of the affected scheduled task is obtained through the positioning system. The real-time location is used as the new starting point, and the first node in the original path located outside the local subgraph area is used as the temporary endpoint for local path replanning.

9. A BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, It also includes an archiving module: The resource scheduling schemes generated from completed or suspended emergency resource scheduling tasks are archived in association with the network fission transaction logs generated during the execution period; Extract path nodes, timestamps, comprehensive cost value, and component state evolution sequences to construct structured event records and import them into a knowledge graph database to establish causal links between disaster events and scheduling decisions.

10. A BIM-based urban and building emergency resource coordination and allocation system according to claim 1, characterized in that, The incremental replanning module is further used for: When the matrix reconstruction module performs atomic modifications that cause a sudden change in the safety risk coefficient, the incremental replanning logic is triggered in real time, and a closed-loop response to physical disaster events is achieved through dynamic adjustment of local paths.

Images