Distributed building escape route correction method

Abstract

This application relates to a distributed building escape path correction method. The method includes: a regional controller monitoring communication status and switching to an isolated autonomous mode when communication is interrupted; collecting data for the region and generating a boundary state vector containing real-time risk values, risk trend prediction values, and environmental airflow vectors, and broadcasting this vector to neighboring regional controllers; calculating a migration impedance value based on feedback data, which includes a coupling effect penalty term calculated based on airflow vectors and estimated pedestrian flow; and generating a boundary breaker command to lock the boundary and correct the escape path when the migration impedance value meets the blocking condition. This application has the advantages of enabling autonomous coordination between regions when central dispatch is lost and effectively preventing secondary disasters caused by the spread of smoke with pedestrian flow.

Description

Technical Field

[0001] This application relates to the field of security technology, and in particular to a distributed building escape route correction method. Background Technology

[0002] With the increasing complexity of large building structures, intelligent evacuation guidance systems have been widely used in building safety. Current mainstream solutions typically employ a centralized control architecture, relying on a central server or fire control room host to aggregate sensor data from across the building, perform global path planning, and send control commands to lower-level area controllers or terminal devices. This architecture can achieve relatively efficient evacuation scheduling when the communication network is unobstructed and the central equipment is functioning properly.

[0003] However, the environment at fire scenes is often complex and changeable. The accompanying high temperatures or physical collapses can easily lead to the interruption of the main communication network or cause the central server to fail due to power supply failure. In the existing technical system, once the area controller detects that the communication connection with the central server has been lost, it is often forced to switch to a preset static escape mode or make simple logical judgments based solely on limited sensor data within the area.

[0004] Due to the lack of mechanisms for lateral information exchange with neighboring areas, isolated area controllers struggle to obtain real-time risk information about adjacent areas. This information asymmetry can lead to serious decision-making errors. For example, when a hazard is detected in its own area, the controller might blindly guide people to a seemingly safe but actually blocked-off adjacent area, or direct them to a potentially high-risk area where the risk is rapidly deteriorating. Furthermore, in the absence of coordination, simply opening boundary passages may ignore the influence of airflow direction, causing toxic fumes to flow back from high-risk areas to safe evacuation zones along with the evacuation flow, thus triggering secondary disasters. Summary of the Invention

[0005] In order to achieve autonomous coordination and dynamic path correction among regional controllers in the absence of a central unified scheduling, this application provides a distributed building escape path correction method.

[0006] This application provides a distributed building escape path correction method, which adopts the following technical solution:

[0007] A distributed building escape route correction method, applied to a building escape guidance system containing multiple area controllers, characterized in that the method includes the following steps:

[0008] S1. The regional controller monitors the communication status with the central server in real time. When the duration of the communication interruption exceeds a preset threshold, it switches to the island autonomous mode.

[0009] S2. The area controller collects sensor data within its area and generates a boundary state vector, and broadcasts the boundary state vector to adjacent area controllers through physically adjacent communication links; wherein, the boundary state vector includes the area's instantaneous risk value, risk trend prediction value, environmental airflow vector, and exit distance potential energy value;

[0010] S3. The area controller receives the boundary state vector fed back by the adjacent area controllers, and calculates the migration impedance function based on the data of the current area and the adjacent areas; the calculation process of the migration impedance function includes: calculating the difference between the instantaneous risk value of the adjacent area and the instantaneous risk value of the current area as the basic risk gradient; extracting the environmental airflow vector and the estimated pedestrian flow, and calculating the product of the two to obtain the coupling effect penalty term; weighting and summing the coupling effect penalty term, the basic risk gradient and the exit distance potential energy value to obtain the migration impedance value for the adjacent area;

[0011] S4. The area controller determines whether the migration impedance value meets the preset blocking condition; if so, it generates a boundary fuse command, controls the access control system located at the boundary of the area to perform a locking action, and corrects the escape guidance path in the area to avoid the boundary of the area.

[0012] Optionally, the environmental airflow vector includes the pressure difference or wind direction angle between the inside and outside of the area;

[0013] The step of calculating the coupling effect penalty term in S3 includes:

[0014] S31. Determine whether the environmental airflow vector indicates that the airflow direction is from this area to the adjacent area;

[0015] S32. If the airflow direction is from this area to the adjacent area, and the instantaneous risk value of this area is higher than the instantaneous risk value of the adjacent area, then a preset positive weighting coefficient is used to weight the product of the environmental airflow vector and the estimated pedestrian flow to obtain the coupling effect penalty term.

[0016] Optionally, the boundary state vector further includes a capacity margin field, the value of which is equal to the maximum capacity of the area where the area controller is located minus the number of people being detected in the current area.

[0017] When calculating the migration impedance value in step S3, the carrying capacity field of the adjacent region is read; if the value of the carrying capacity field is less than or equal to zero, the migration impedance value is increased by the congestion penalty coefficient.

[0018] Optionally, the risk trend prediction value is a first derivative value calculated based on the historical time series of the sensor data;

[0019] The process of generating the exit distance potential energy value in S2 is as follows: the area controller reads the topological distance data stored locally, or iteratively accumulates the exit distance potential energy value received from other adjacent areas to characterize the logical distance from the area to the nearest effective safe exit.

[0020] The step of generating the boundary circuit breaker command in S4 includes:

[0021] S41. Determine whether the instantaneous risk value of the adjacent area is lower than a preset safety threshold, but whether the predicted risk trend value of the adjacent area exceeds a preset deterioration threshold;

[0022] S42. If so, the adjacent region is determined to be in a quasi-high-risk state, and the migration impedance value corresponding to the adjacent region is set to infinity to generate the boundary fuse command.

[0023] Optionally, the step of correcting the escape guidance path in this area in S4 includes:

[0024] S43. The area controller calculates an internal retention consumption function based on the current oxygen concentration and carrying capacity of the area; wherein the sensor data also includes the current oxygen concentration;

[0025] S44. Compare the internal retention consumption function with the lowest migration impedance value calculated for all adjacent regions;

[0026] S45. If the difference between the value of the minimum migration impedance and the value of the internal retention consumption function is greater than a preset hysteresis threshold, it is determined that the external migration risk significantly exceeds the internal defense risk, and the active isolation defense mode is triggered.

[0027] S46. In the active isolation and defense mode, the area controller calls the HVAC control interface to adjust the air supply parameters of the area to change the environmental airflow vector. The area controller identifies the adjacent area with the highest risk trend prediction value and adjusts the air supply parameters to make the area form a positive pressure difference relative to the adjacent area, thereby blocking the flue gas diffusion path. At the same time, the access control system at the boundary of the area is controlled to perform a locking action.

[0028] Optionally, the boundary state vector further includes a defense state identifier field; the distributed building escape path correction method further includes the following steps:

[0029] S47. The area controller sets the defense status identifier field in the boundary status vector to valid and updates the value of the bearing capacity field to zero;

[0030] S48. Broadcast the updated boundary state vector to all neighboring area controllers through the physically adjacent communication links to trigger the neighboring area controllers to cut off the path connection to their respective area in the corresponding local topology map, so as to prevent the generation of routing loops.

[0031] Optionally, the method further includes the following steps:

[0032] S5. The area controller determines whether the area is marked as a core refuge layer and whether the current air pressure is negative.

[0033] S6. If this area is a core refuge floor and is under negative pressure, and the boundary state vector of the adjacent area is shown as a high-risk state of positive pressure, then ignore the migration impedance value calculation result in S3, forcibly execute the boundary fuse command, and output a control signal to start the pressurization and ventilation equipment in this area until the air pressure in this area changes to positive pressure.

[0034] S7. The area controller sends a coordinated depressurization request to other adjacent areas that are downwind or in a low-risk state; the adjacent area controller that receives the coordinated depressurization request controls the smoke exhaust valve located in the adjacent area to open, so as to cooperate with the area to build a directional air pressure gradient and assist the area to change to a positive pressure state.

[0035] Optionally, the method further includes the following steps:

[0036] S81. The area controller obtains the channel length value and the visibility value within the channel leading to the adjacent area;

[0037] S82. Calculate the estimated travel time required for personnel to pass through the passage based on the passage length value and the visibility value;

[0038] S83. Compare the estimated passage time with the human tolerance limit time calculated based on the current toxic gas concentration value collected by the sensor data; wherein, the sensor data also includes the toxic gas concentration value;

[0039] S84. If the estimated passage time is greater than the human body's tolerance limit time, then skip the calculation in S3 and directly generate a boundary circuit breaker command for the path leading to the adjacent area.

[0040] In summary, this application includes at least one of the following beneficial technical effects:

[0041] 1. This application effectively addresses the risk of single-point failure in traditional centralized systems by constructing a distributed architecture and an islanded autonomous mode. When the area controller detects that the communication interruption with the central server exceeds a preset threshold, it can automatically switch roles to become the local decision-making center, interacting with neighboring nodes using physically adjacent communication links to exchange boundary state vectors. This mechanism ensures that even in extreme environments where a fire causes the backbone network to collapse or the central server to fail, the system can still generate effective evacuation guidance based on local topology and real-time data, significantly improving the system's survivability and business continuity.

[0042] 2. This application proposes a migration impedance function calculation method that includes a coupling effect penalty term, which can prevent secondary disasters caused by evacuation behavior. By calculating the product of the environmental airflow vector and the estimated pedestrian flow, this method quantifies the coupling risk of smoke and pedestrian flow. Unlike traditional path planning that only considers static distance, this application can identify and suppress paths that, although physically short, have downwind pressure differences at the algorithm level, thereby preventing toxic smoke from flowing back into safe areas with the evacuation flow and achieving a dynamic balance between evacuation efficiency and environmental safety.

[0043] 3. This application introduces a predictive boundary circuit breaker mechanism based on risk trend forecasts, enhancing the defense capability against potential hazards. By calculating the first derivative of sensor data, the system can identify near-high-risk states with low immediate risk values ​​but rapid deterioration rates. The boundary circuit breaker command generated based on this can forcibly control access control locks and correct guidance paths, implementing blocking of high-risk paths at both physical and logical levels. This effectively prevents evacuees from being misled by temporary safety data into areas about to spiral out of control, ensuring the substantial safety of escape routes.

[0044] 4. This application implements a survival environment reconstruction strategy based on an active isolation and defense mode, providing a reliable shelter barrier for personnel who cannot be evacuated immediately. When the risk of external migration is determined to be significantly higher than the risk of internal confinement, or when the core refuge floor is under negative pressure, the system automatically activates HVAC equipment to construct a positive pressure air shield relative to adjacent areas, and works in conjunction with the access control system to form a double isolation. This active defense method utilizes fluid dynamics principles to resist the intrusion of external smoke, and combined with physiological constraints on the human body's tolerance limits, it maximizes the survival time window for trapped personnel in the fire. Attached Figure Description

[0045] Figure 1 A schematic diagram illustrating the main flow of a distributed building escape path correction method in one embodiment of the present invention is shown.

[0046] Figure 2 A schematic diagram illustrating the distributed architecture and regional controller connection relationship of a building escape guidance system in one embodiment of the present invention is shown.

[0047] Figure 3 A flowchart illustrating the specific logic of calculating the migration impedance function and generating the coupling effect penalty term in one embodiment of the present invention is shown.

[0048] Figure 4 A flowchart illustrating an active isolation defense mode based on the comparison of internal retention consumption and external migration risk and the protection of the core refuge layer in one embodiment of the present invention is shown. Detailed Implementation

[0049] The present application will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the application and are not intended to limit the scope of the application.

[0050] In the following description, numerous specific details are set forth for purposes of explanation in order to provide a thorough understanding of the inventive concept. As part of this specification, some of the accompanying drawings of this disclosure are block diagrams illustrating structures and devices to avoid complicating the disclosed principles. For clarity, not all features of the actual embodiment need to be described. Furthermore, the language used in this disclosure has been primarily chosen for readability and instructional purposes and may not have been chosen to define or limit the subject matter of the invention, thus requiring the necessary claims to determine such inventive subject matter. References to “an embodiment” or “an embodiment” in this disclosure mean that a particular feature, structure, or characteristic described in connection with that embodiment is included in at least one embodiment, and multiple references to “an embodiment” or “an embodiment” should not be construed as necessarily referring to the same embodiment.

[0051] Unless explicitly defined, the terms “a,” “an,” and “the” are not intended to refer to a singular entity, but rather to include a general category whose specific examples can be used for illustration. Therefore, the use of the terms “a” or “an” can mean any number of at least one, including “a,” “one or more,” “at least one,” and “one or more.” The term “or” means any of the options and any combination of the options, including all options unless explicitly indicated that the options are mutually exclusive. The phrase “at least one of” when combined with a list of items refers to a single item in the list or any combination of items in the list. The phrase does not require all items listed unless explicitly defined as such.

[0052] As modern architecture develops towards higher floors, larger scales, and more complex functions, population density continues to increase, and the internal spatial structures of buildings are becoming increasingly complex. In the event of sudden emergencies such as fires and earthquakes, ensuring the rapid, orderly, and safe evacuation of people has become a key focus in the field of public safety.

[0053] Current building escape and security systems are mainly divided into two categories: traditional and intelligent. Traditional evacuation guidance systems rely primarily on fluorescent signs with fixed directions or constantly lit emergency lights. Their directional paths are fixed during the building design phase and cannot be adjusted according to the actual location of a fire. Existing intelligent evacuation guidance systems typically employ a centralized control architecture. This involves a central server or a host computer located in the fire control room, which collects signals from smoke detectors, heat detectors, and manual alarm buttons throughout the building. Based on a pre-set topology map and real-time fire alarm signals, the central server uses a global path planning algorithm to calculate the optimal escape route and sends instructions to the area controllers on the ground floor, thereby changing the direction of the intelligent evacuation indicator lights or controlling the opening and closing of access control systems.

[0054] However, the aforementioned existing implementation methods have significant limitations in practical applications. First, the centralized architecture is highly dependent on the central node and backbone communication network. In the event of a severe fire, high temperatures, explosions, or building collapses can easily lead to the interruption of backbone network cables or power failure of the central server. Once a regional controller loses contact with the central server due to a communication failure, the regional controller often can only execute preset static logic or maintain the state of the last received instruction, unable to adapt to the dynamic changes in the fire environment. Second, existing regional controllers typically lack horizontal information exchange mechanisms. Regional controllers in a communication island state have difficulty obtaining the risk status of adjacent areas, easily guiding evacuees to seemingly safe but actually blocked or impending flashover areas. Furthermore, existing systems often only consider distance and fire source location when planning routes, ignoring the physical characteristics of smoke flow in a fire. For example, opening a passage connecting adjacent areas without assessing the pressure difference may cause toxic smoke to flow back from high-risk areas to safe areas with the evacuation flow, thereby triggering secondary disasters.

[0055] Therefore, this application discloses a distributed building escape path correction method, referring to... Figure 1 The method, applicable to building escape guidance systems comprising multiple area controllers, is characterized by comprising the following steps:

[0056] S1. The regional controller monitors the communication status with the central server in real time. When the duration of the communication interruption exceeds a preset threshold, it switches to the island autonomous mode.

[0057] Reference Figure 2The area controller establishes a connection with the central server via Ethernet, fiber optic, or industrial fieldbus, and can be configured with a wireless network as a backup link. The area controller monitors the communication link by actively sending heartbeat packets and waiting for a reply, or by listening to the inspection signals broadcast by the central server.

[0058] The communication status monitored in real time by the area controller is divided into normal connection status, network jitter status, and communication interruption status. In the normal connection status, the heartbeat packet delay is within the standard range; the network jitter status is characterized by intermittent signal interruption, fluctuating response delay, or occasional packet loss; the communication interruption status refers to the inability to receive a valid signal for a continuous period of time, corresponding to a broken backbone line or failure of the central server.

[0059] Setting a preset threshold aims to distinguish between network jitter and communication interruptions, preventing the system from frequently switching between normal and isolated modes. A threshold set too short can cause network fluctuations to be misjudged as faults, leading to system instability; a threshold set too long will delay the activation of contingency plans. This threshold is typically set based on network latency characteristics, for example, between thirty and sixty seconds.

[0060] In normal network mode, the regional controller acts as the execution unit, uploading data and executing global commands issued by the central server. In island autonomous mode, the regional controller no longer relies on upper-level commands, activates local decision-making algorithms, and enables lateral communication interfaces with neighboring regional controllers. Switching to island autonomous mode after communication interruption eliminates the dependence on the central node, allowing each regional controller to continue providing evacuation guidance using local and neighboring data when the backbone network fails.

[0061] After switching to the isolated autonomous mode, step S2 is executed: the area controller collects sensor data within its area and generates a boundary state vector, and broadcasts the boundary state vector to the adjacent area controllers through physically adjacent communication links; wherein, the boundary state vector includes the area's instantaneous risk value, risk trend prediction value, environmental airflow vector, and exit distance potential energy value.

[0062] Optionally, the environmental airflow vector includes the pressure difference or wind direction angle between the inside and outside of the area;

[0063] Optionally, the risk trend prediction value is a first derivative value calculated based on the historical time series of the sensor data;

[0064] The process of generating the exit distance potential energy value in S2 is as follows: the area controller reads the topological distance data stored locally, or iteratively accumulates the exit distance potential energy value received from other adjacent areas to characterize the logical distance from the area to the nearest effective safe exit.

[0065] The area controller connects to various sensors distributed throughout the area's ceiling, doorways, and ventilation ducts, including smoke detectors, temperature detectors, carbon monoxide concentration sensors, and differential pressure sensors. The area controller acquires environmental parameter data in real time through analog sampling or digital bus polling. Boundary state vectors are standardized data frame structures that encapsulate the complex environmental states within the area into unified communication protocol packets. This structure masks the differences in equipment across different areas and supports efficient resolution between distributed nodes.

[0066] Physically adjacent communication links refer to the underlying communication buses that directly connect two spatially adjacent area controllers, such as RS485 buses, CAN buses, or point-to-point wireless mesh networks. In a network outage environment where the backbone network is paralyzed, this decentralized broadcast mechanism allows controllers to send status data directly to their neighbors without going through aggregation switches or routers, maintaining local communication capabilities.

[0067] The immediate risk value is a numerical value obtained after weighted normalization of multi-dimensional sensor data, representing the current level of danger. The risk trend prediction value is the first derivative calculated using a sliding window to differentiate the historical sensor data sequence, used to predict the direction of fire development. For example, when the temperature rises sharply but has not yet reached the alarm threshold, a larger derivative value can provide an early warning of flashover risk. The environmental airflow vector quantifies the physical characteristics of airflow through pressure difference or wind direction angle values. This parameter is directly related to the direction of smoke diffusion; a positive pressure difference indicates outward repulsion of airflow, while a negative pressure difference indicates inward inhalation, providing a physical basis for subsequent judgments on whether smoke will flow back into the area with people.

[0068] The potential energy value for the exit distance is generated using the distance vector routing algorithm. If the region has a direct safe exit, the potential energy value is initialized to the base value by reading the locally stored topology data. If the region has no exit, the potential energy values ​​broadcast from neighboring regions are received, the minimum value is selected, and one is added to it as the potential energy value for the region. This value represents the logical hop count to reach the nearest valid safe exit and is used to construct a dynamic steering gradient in the absence of a global map.

[0069] After completing the interaction of the boundary state vectors, the area controller executes step S3: the area controller receives the boundary state vectors fed back by the neighboring area controllers, and calculates the migration impedance function based on the data of the current area and the neighboring areas; the calculation process of the migration impedance function includes: calculating the difference between the instantaneous risk value of the neighboring area and the instantaneous risk value of the current area as the basic risk gradient; extracting the environmental airflow vector and the estimated pedestrian flow, and calculating the product of the two to obtain the coupling effect penalty term; weighted summing of the coupling effect penalty term, the basic risk gradient, and the exit distance potential energy value to obtain the migration impedance value for the neighboring area.

[0070] The area controller triggers the reception of feedback data through an interrupt service routine or polling mechanism on the communication port. Once the physical link layer detects the payload, the area controller performs integrity verification and parsing on the data frame, extracting the immediate risk value, risk trend prediction value, environmental airflow vector, and exit distance potential energy value of the adjacent area. The migration impedance function is defined as a mathematical model characterizing the comprehensive resistance or cost faced by personnel moving from their current area to an adjacent area. The calculated migration impedance value is used to quantify the difficulty of passage; the lower the impedance value, the more suitable the path is as an escape route.

[0071] Specifically, the migration impedance value Z for a given adjacent region can be calculated using the following formula:

[0072]

[0073] in, and These represent the immediate risk values ​​of adjacent areas and the current area, respectively. That is, the basic risk gradient; The magnitude or projected value representing the ambient airflow vector. The product of the two represents the estimated pedestrian flow through the boundary. This constitutes a coupling effect penalty term; This represents the potential energy value at the exit distance; These are the corresponding weighting coefficients.

[0074] The basic risk gradient reflects the difference in security potential. When the instantaneous risk values ​​of adjacent areas... Less than the immediate risk value of this area When the risk difference is negative, the total impedance value Z decreases, thus creating a driving force for evacuation and guiding the population to naturally move towards lower-risk areas. Conversely, if the risk of an adjacent area is higher, the risk difference is positive, increasing the impedance and inhibiting the population from moving there.

[0075] The construction of the coupling effect penalty term is based on the interaction principle of fluid mechanics and crowd dynamics. (Estimated pedestrian flow) This represents the likelihood and duration of opening a physical channel, and the ambient airflow vector. This represents the potential driving force of smoke diffusion. The product of the two quantifies the risk of smoke diffusion caused by evacuation behavior. If there is only a pressure difference but no one passes through, or if someone passes through but the pressure is balanced, the coupling risk is low; only when a large number of people pass through and there is an unfavorable pressure difference does this penalty term increase significantly, thereby suppressing path selection that may lead to smoke backflow at the algorithm level.

[0076] The integrated weighted summation model assigns different weight coefficients to the risk gradient, coupling effect penalty term, and exit distance potential energy value. This achieves a balance among multiple objectives. Weighting strategies are typically set based on the characteristics of the building; for example, increasing the weight of coupling effects in areas with poor smoke extraction conditions. The value is adjusted, while in areas with complex paths, the weight of the exit distance potential energy value is increased. The numerical value. This multi-parameter fusion calculation avoids the one-sidedness of single-index decision-making, so that the calculated migration impedance value can comprehensively reflect the safety of the path, the impact on the environment, and the escape efficiency.

[0077] Furthermore, the specific steps for calculating the coupling effect penalty term in S3 include S31-S32.

[0078] S31. Determine whether the environmental airflow vector indicates that the airflow direction is from this area to the adjacent area;

[0079] S32. If the airflow direction is from this area to the adjacent area, and the instantaneous risk value of this area is higher than the instantaneous risk value of the adjacent area, then a preset positive weighting coefficient is used to weight the product of the environmental airflow vector and the estimated pedestrian flow to obtain the coupling effect penalty term.

[0080] The determination of the direction of the ambient airflow vector is typically based on vector algebra operations. The area controller calculates the dot product of the ambient airflow vector and the normal vector leading to the boundary of the adjacent area. If the calculation result indicates the same direction, the airflow direction is determined to be downstream, i.e., flowing from this area to the adjacent area. This determination logic aims to identify the potential risk of downstream smoke diffusion and prevent toxic smoke from invading downstream areas by means of airflow potential energy. In the determination of risk gradient conditions, when the immediate risk value of this area is higher than that of the adjacent area, it means that there is a strong need to evacuate from the high-risk area to the low-risk area. However, if the airflow direction also points to the low-risk area at this time, simply opening the passage will cause the fire or smoke to spread into this area. Therefore, under this specific condition, the system calls a preset positive weighting coefficient to significantly amplify the product of the ambient airflow vector and the estimated flow of people, thereby quantifying the coupling effect of smoke diffusion and personnel flow. This mechanism significantly increases the migration impedance value of this path through mathematical means, thereby suppressing this path selection, which, although in line with escape intuition, may lead to secondary disasters, at the algorithm level.

[0081] Optionally, the boundary state vector further includes a capacity margin field, the value of which is equal to the maximum capacity of the area where the area controller is located minus the number of people being detected in the current area; when calculating the migration impedance value in S3, the capacity margin field of the adjacent area is read; if the value of the capacity margin field is less than or equal to zero, the value of the migration impedance value is increased by the congestion penalty coefficient.

[0082] The carrying capacity field is a dynamic indicator measuring the remaining capacity of an area. The maximum carrying capacity of an area is typically preset based on the personnel density limit per unit area in building fire protection design codes and the effective net area of ​​the area. The area controller uses camera counting or infrared beam sensors to count the number of people in the area in real time and calculates the carrying capacity by subtracting the current number of people detected from the maximum carrying capacity. The congestion penalty coefficient trigger mechanism is set to activate when the carrying capacity is less than or equal to zero. Once this condition is triggered, the algorithm adds a very large penalty coefficient (e.g., infinity or a specific high value) to the migration impedance value. This mechanism acts as a logical block in dynamic flow control, preventing subsequent flow of people from continuing to flood into adjacent areas that are already saturated or overloaded, thereby effectively avoiding stampedes caused by spatial congestion and achieving load balancing of evacuation flows.

[0083] Based on the calculated migration impedance value, the area controller executes step S4: the area controller determines whether the migration impedance value meets the preset blocking condition; if so, it generates a boundary fuse command, controls the access control system located at the boundary of the area to perform a locking action, and corrects the escape guidance path in the area to avoid the boundary of the area.

[0084] The determination logic for migration impedance values ​​typically employs a threshold comparison method. The area controller compares the calculated migration impedance value with a preset blocking threshold. This blocking threshold is set based on the building's safety level and personnel movement characteristics, representing the risk limit for path passage. When the migration impedance value exceeds this threshold, or is significantly higher than the impedance value of other optional paths in a relative comparison, the system determines that the path is no longer passable and meets the preset blocking conditions.

[0085] Once the blocking conditions are met, the area controller immediately generates a boundary circuit breaker command. This command has extremely high execution priority in the system control hierarchy, allowing it to bypass regular normally open signals or low-priority access requests and directly occupy the output channel of the underlying controller. This high-priority triggering mechanism ensures that the system can quickly cut off the propagation path of the hazard source in critical moments, rather than getting stuck in the waiting queue of conventional logic.

[0086] Upon receiving a boundary trip command, the access control system executes the specific locking action. For two-way access doors equipped with electromagnetic locks, the controller cuts off the power supply to the electromagnetic lock or sends a locking signal to make it engage; for open boundaries equipped with fireproof roller shutters, the controller drives the roller shutter motor to descend until it is fully closed. This physical action constructs a physical barrier at the boundary of the area, not only physically blocking the spread of external high-temperature smoke or open flames into this area, but also forcibly preventing panicked people in this area from rushing into high-risk areas.

[0087] Meanwhile, the area controller sends control data to the intelligent evacuation guidance terminals within the area via the underlying bus to correct the escape guidance path. Specific implementation methods include controlling the variable-directional indicator lights to change the direction of their arrows, directing them away from the fused boundary; or controlling the ground light guide lines to extinguish the light segments leading to the boundary and displaying a no-entry sign on the exit display screen in that direction. This guidance strategy, which avoids the area boundary, establishes a clear psychological boundary for evacuees at the visual level, effectively preventing people from blindly following the crowd into high-risk or severely congested areas, ensuring that the flow of people is guided to alternative exits with lower impedance and higher safety.

[0088] Furthermore, the specific steps for generating the boundary circuit breaker command in S4 include:

[0089] S41. Determine whether the instantaneous risk value of the adjacent area is lower than a preset safety threshold, but whether the predicted risk trend value of the adjacent area exceeds a preset deterioration threshold;

[0090] S42. If so, the adjacent region is determined to be in a quasi-high-risk state, and the migration impedance value corresponding to the adjacent region is set to infinity to generate the boundary fuse command.

[0091] Relying solely on immediate risk values ​​for path planning is inefficient and fails to adapt to the ever-changing fire scene environment. Before smoldering transitions to open flame or flashover occurs, the temperature or smoke concentration values ​​collected by sensors may be within safe ranges, but their rate of change is significant. Introducing risk trend prediction values, i.e., the first derivative, aims to obtain the dynamic evolution characteristics of environmental parameters and compensate for the shortcomings of static threshold determination.

[0092] A near-high-risk state indicates that the regional environment is in a critically unstable transitional state. Although the current environmental indicators in the area have not triggered conventional alarms, their rate of deterioration suggests that flashover or environmental runaway may occur in the short term. Step S41 employs a dual-determination logic to identify this state. If only the immediate risk value is detected, the system will classify the area as a safe node; if only the risk trend prediction value is detected, areas within the safe range but with fluctuating values ​​may be classified as dangerous areas. When both the immediate value and the trend value are met simultaneously, areas with low immediate values ​​but a high deterioration trend can be identified.

[0093] Setting the migration impedance to infinity effectively excludes the path in the mathematical model. Regardless of the path's distance or airflow conditions, the infinite impedance prevents it from being selected in the path optimization algorithm. This mechanism of generating boundary-breaking commands prevents evacuees from entering high-risk areas with low immediate risk but impending deterioration, ensuring the safety of escape routes.

[0094] Furthermore, referring to Figure 3 The step of correcting the escape guidance path in this area in S4 also includes S43-S46.

[0095] S43. The area controller calculates an internal retention consumption function based on the current oxygen concentration and carrying capacity of the area; wherein the sensor data also includes the current oxygen concentration.

[0096] S44. Compare the internal retention consumption function with the lowest migration impedance value calculated for all adjacent regions.

[0097] S45. If the difference between the minimum migration impedance value and the internal retention consumption function value is greater than a preset hysteresis threshold, it is determined that the external migration risk significantly exceeds the internal defense risk, and the active isolation defense mode is triggered.

[0098] S46. In the active isolation and defense mode, the area controller calls the HVAC control interface to adjust the air supply parameters of the area to change the environmental airflow vector. The area controller identifies the adjacent area with the highest risk trend prediction value and adjusts the air supply parameters to make the area form a positive pressure difference relative to the adjacent area, thereby blocking the flue gas diffusion path. At the same time, the access control system at the boundary of the area is controlled to perform a locking action.

[0099] The internal dwelling cost function quantifies the survival cost of personnel waiting for rescue in place within a given area. The function's value is negatively correlated with the current oxygen concentration and with carrying capacity. As oxygen concentration decreases or overcrowding increases, the dwelling cost function increases, indicating a reduction in the feasibility of survival within the area. This model integrates physiological maintenance needs with physical space constraints, providing a mathematical assessment of the safety of the internal environment.

[0100] As a specific implementation example, internal retention consumption value It can be represented as: .in, The current oxygen concentration, For normalized carrying capacity, and These are the preset weighting coefficients.

[0101] The internal and external risk game decision-making mechanism is implemented by comparing the minimum migration resistance value with the internal retention cost function. The minimum migration resistance value represents the minimum cost among all possible external escape paths, i.e., the lower bound of external migration risk. Comparing the two aims to determine which strategy has a lower overall risk: risking traversing the external region or staying in the local area.

[0102] Setting a preset hysteresis threshold is to prevent decision-making oscillations. The system only switches to defensive mode when the difference between the external migration risk value and the internal defense risk value reaches a certain threshold. This hysteresis comparison logic avoids frequent switching between evacuation and defense modes due to small fluctuations in sensor data, ensuring the continuity and stability of the control strategy.

[0103] In active isolation and defense mode, the area controller reconstructs the physical environment by adjusting air supply parameters. Based on fluid dynamics principles, the controller increases the airflow to its area, creating a positive pressure environment relative to high-risk adjacent areas. This artificially created pressure gradient forms an aerodynamic barrier that counteracts the natural diffusion force of flue gas, preventing external toxic fumes from entering the area through gaps or open channels. Simultaneously, in conjunction with the locking action of the access control system, a dual defense system of airflow pressure differential barrier and physical isolation is formed, cutting off the flue gas propagation path while preventing accidental damage to the sealed environment by personnel.

[0104] Optionally, the boundary state vector further includes a defense state identifier field; the distributed building escape path correction method further includes the following steps:

[0105] S47. The area controller sets the defense status identifier field in the boundary status vector to valid and updates the value of the bearing capacity field to zero;

[0106] S48. Broadcast the updated boundary state vector to all neighboring area controllers through the physically adjacent communication links to trigger the neighboring area controllers to cut off the path connection to their respective area in the corresponding local topology map, so as to prevent the generation of routing loops.

[0107] The defense status identifier field, serving as a state synchronization signal in the distributed system, indicates to other nodes in the network that this area has entered an impassable isolation mode. Updating the capacity field to zero implements a logical flow control strategy. This operation, at the data level, rejects external pedestrian input; regardless of whether there is still physical space available in this area, it sends an unreceivable signal to the outside, thus coordinating with the locking action of physical access control to prevent subsequent path planning algorithms from guiding pedestrian flow to this node.

[0108] The area controller broadcasts the updated boundary state vector through physically adjacent communication links, enabling neighboring area controllers to obtain the state changes of their respective areas and update their maintained local topology data. Triggering a neighboring area controller to sever path connections to its own area is equivalent to dynamically pruning the local topology graph in graph theory algorithms. This means treating the area as an invalid node and removing directed edges connected to it from the path search computation graph.

[0109] This mechanism prevents routing loops. In distributed routing computation, if connections to closed areas are not severed, adjacent nodes may use outdated state information to point to each other as their next-hop path, creating an infinite loop. By severing connections, the path planning algorithm is ensured to converge to other valid safe exits, avoiding unnecessary use of computational resources and incorrect navigation instructions.

[0110] Furthermore, in order to ensure the safety of special areas, refer to Figure 4 The method executes step S5: The area controller determines whether the area is marked as a core refuge layer and whether the current air pressure is negative.

[0111] A core refuge floor is a specific area within a building that possesses a higher fire resistance rating, an independent ventilation system, and reserves of survival supplies. It has the highest safety priority in the overall evacuation strategy and is used to provide temporary shelter when immediate evacuation to the outdoors is not possible. The area controller identifies whether its area is marked as a core refuge floor by reading preset configuration parameters in its local non-volatile memory or the status of hardware DIP switches.

[0112] Monitoring atmospheric pressure relies on micro-differential pressure sensors deployed at area boundaries, such as the inside of fire doors or ventilation openings. The area controller acquires the differential pressure values ​​fed back by the sensors in real time via an analog-to-digital converter interface. Negative pressure refers to a situation where the atmospheric pressure within the area is lower than the atmospheric pressure in adjacent areas or public corridors. In a fire environment, smoke flow follows the physical law of diffusion from high-pressure areas to low-pressure areas. If the core refuge floor is under negative pressure, an airflow inhalation effect will occur, causing toxic smoke from outside to seep into the refuge floor through door gaps or pipe openings, compromising the safety of the area.

[0113] When a critical situation is detected, continue to refer to... Figure 4 Step S6: If this area is a core refuge layer and is under negative pressure, and the boundary state vector of the adjacent area is shown as a high-risk state of positive pressure, then ignore the migration impedance value calculation result in S3, forcibly execute the boundary fuse command, and output a control signal to start the pressurization and ventilation equipment in this area until the air pressure in this area changes to positive pressure.

[0114] In a scenario where the local area is under negative pressure and the adjacent area is under positive pressure and high risk, according to fluid dynamics principles, high-pressure dense smoke from the outside will overcome resistance and flow back into the refuge floor, posing a risk of smoke inhalation. In this situation, the system employs a mechanism that ignores migration impedance calculations, switching from a cost-based algorithm optimization logic to a priority decision logic based on hard rule protection. A boundary fusion command is forcibly executed to immediately close fire doors or roller shutters, achieving physical isolation. Simultaneously, a pressurized air supply system is activated to construct a positive pressure barrier, actively injecting air to increase internal pressure and resist external smoke intrusion. This process uses a closed-loop control strategy, continuously operating the pressurized system until the differential pressure sensor indicates that the air pressure in this area is higher than that in the external area, thus completing the reconstruction of the refuge floor's safe environment.

[0115] S7. The area controller sends a coordinated depressurization request to other adjacent areas that are downwind or in a low-risk state; the adjacent area controller that receives the coordinated depressurization request controls the smoke exhaust valve located in the adjacent area to open, so as to cooperate with the area to build a directional air pressure gradient and assist the area to change to a positive pressure state.

[0116] The strategy for selecting target nodes for coordination by the area controller is based on airflow vectors and risk values, prioritizing adjacent areas downwind or with lower risk values ​​to avoid directing flue gas towards high-risk areas. Coordination request commands are encapsulated and transmitted via a cross-area communication protocol. Upon receiving the request, adjacent controllers open exhaust valves or activate exhaust fans. This mechanism utilizes fluid dynamics principles, combining pressurized air supply in the local area with depressurized exhaust in adjacent areas to create a directional air pressure gradient from the local area to adjacent areas. This distributed coordination mechanism accelerates the establishment of a positive pressure environment in the core refuge layer and improves exhaust efficiency.

[0117] Optionally, the method further includes the following steps S81-S84.

[0118] S81. The area controller obtains the channel length value and the visibility value within the channel leading to the adjacent area.

[0119] S82. Calculate the estimated travel time required for personnel to pass through the passage based on the passage length value and the visibility value.

[0120] S83. Compare the estimated travel time with the human tolerance limit time calculated based on the current toxic gas concentration value collected by the sensor data; wherein the sensor data also includes the toxic gas concentration value.

[0121] S84. If the estimated passage time is greater than the human body's tolerance limit time, then skip the calculation in S3 and directly generate a boundary circuit breaker command for the path leading to the adjacent area.

[0122] The regional controller jointly collects the static geometric parameters (such as length) and dynamic environmental parameters (such as visibility) of the passageway. The calculation model for the estimated travel time considers the impact of visibility on personnel movement speed. When visibility decreases, movement speed decreases non-linearly, leading to a longer travel time. The human tolerance limit time is calculated based on the currently collected toxic gas concentration values ​​using toxicological models (such as Hubble's Law), representing the maximum exposure time for a human body at a specific concentration without irreversible damage. The estimated travel time is compared with the human tolerance limit time. If the estimated travel time exceeds the tolerance limit, the path is determined to have an extremely high risk. In this case, the system directly generates a boundary circuit breaker command, skipping the impedance calculation in S3. This logic reflects the priority of physiological hard constraints over costly soft constraints, ensuring direct blocking when the path is infeasible, rather than performing invalid optimization calculations.

[0123] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0124] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0125] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A distributed building escape route correction method, applied to a building escape guidance system containing multiple area controllers, characterized in that, The method includes the following steps: S1. The regional controller monitors the communication status with the central server in real time. When the duration of the communication interruption exceeds a preset threshold, it switches to the island autonomous mode. S2. The area controller collects sensor data within its area and generates a boundary state vector, broadcasting the boundary state vector to adjacent area controllers via physically adjacent communication links; wherein, the boundary state vector includes the area's instantaneous risk value, risk trend prediction value, environmental airflow vector, and exit distance potential energy value; the environmental airflow vector includes the air pressure difference value or wind direction angle value inside and outside the area; the exit distance potential energy value is generated using the distance vector routing algorithm principle. If the area directly has a safe exit, the exit distance potential energy value is initialized to a base value by reading the locally pre-stored topology data; if the area has no safe exit, the exit distance potential energy value broadcast by the adjacent area is received, the minimum value is selected and incremented by one as the exit distance potential energy value of the area, representing the logical hop count to reach the nearest effective safe exit from the area; S3. The area controller receives the boundary state vector fed back by the adjacent area controllers, and calculates the migration impedance function based on the data of the current area and the adjacent area. The calculation process of the migration impedance function includes: calculating the difference between the instantaneous risk value of the adjacent area and the instantaneous risk value of the current area as the basic risk gradient; extracting the environmental airflow vector and the estimated pedestrian flow, and calculating the product of the two to obtain the coupling effect penalty term; weighting and summing the coupling effect penalty term, the basic risk gradient, and the exit distance potential energy value to obtain the migration impedance value for the adjacent area. The step of calculating the coupling effect penalty term includes: S31. determining whether the environmental airflow vector indicates that the airflow direction is from the current area to the adjacent area; S32. if the airflow direction is from the current area to the adjacent area, and the instantaneous risk value of the current area is higher than the instantaneous risk value of the adjacent area, then calling a preset positive weighting coefficient to weight the product of the environmental airflow vector and the estimated pedestrian flow to obtain the coupling effect penalty term. S4. The area controller determines whether the migration impedance value meets the preset blocking condition; if so, it generates a boundary fuse instruction. The boundary fuse instruction has a higher execution priority than the conventional normally open signal or low priority access request in the system control level. It directly occupies the output channel of the lower controller, controls the access control system located at the boundary of this area to perform a locking action, and corrects the escape guidance path in this area to avoid the boundary of this area.

2. The distributed building escape path correction method according to claim 1, characterized in that, The boundary state vector also includes a capacity margin field, the value of which is equal to the maximum capacity of the area where the area controller is located minus the number of people being detected in the current area. When calculating the migration impedance value in step S3, the carrying capacity field of the adjacent area is read; if the value of the carrying capacity field is less than or equal to zero, the value of the migration impedance value is increased by the congestion penalty coefficient, wherein the congestion penalty coefficient is used to logically block dynamic traffic.

3. The distributed building escape path correction method according to claim 2, characterized in that, The risk trend prediction value is the first derivative value calculated using a sliding window based on the historical time series of the sensor data. The step of generating the boundary circuit breaker command in S4 includes: S41. Determine whether the instantaneous risk value of the adjacent area is lower than a preset safety threshold, but whether the predicted risk trend value of the adjacent area exceeds a preset deterioration threshold; S42. If so, the adjacent region is determined to be in a quasi-high-risk state, and the migration impedance value corresponding to the adjacent region is set to infinity to generate the boundary fuse command.

4. The distributed building escape path correction method according to claim 3, characterized in that, The step of correcting the escape guidance path within this area in S4 includes: S43. The area controller calculates an internal retention consumption function based on the current oxygen concentration and carrying capacity of the area; wherein, the sensor data also includes the current oxygen concentration, and the internal retention consumption function is used to quantify the survival cost of personnel waiting for rescue in place within the area; S44. Compare the internal retention consumption function with the lowest migration impedance value calculated for all adjacent regions; S45. If the difference between the value of the minimum migration impedance and the value of the internal retention consumption function is greater than a preset hysteresis threshold, it is determined that the external migration risk significantly exceeds the internal defense risk, and the active isolation defense mode is triggered. S46. In the active isolation and defense mode, the area controller calls the HVAC control interface to adjust the air supply parameters of the area to change the environmental airflow vector. The area controller identifies the adjacent area with the highest risk trend prediction value and adjusts the air supply parameters to make the area form a positive pressure difference relative to the adjacent area, thereby blocking the flue gas diffusion path. At the same time, the access control system at the boundary of the area is controlled to perform a locking action.

5. The distributed building escape path correction method according to claim 4, characterized in that, The boundary state vector also includes a defense state identifier field; the distributed building escape path correction method further includes the following steps: S47. The area controller sets the defense status identifier field in the boundary status vector to valid and updates the value of the bearing capacity field to zero; S48. Broadcast the updated boundary state vector to all neighboring area controllers through the physically adjacent communication links to trigger the neighboring area controllers to cut off the path connection to their own area in the corresponding local topology map to prevent routing loops from occurring; wherein, cutting off the path connection to their own area is to perform dynamic pruning on the local topology graph in the graph theory algorithm, that is, to treat their own area as an invalid node and remove the directed edge connected to the node from the computation graph of path search.

6. The distributed building escape path correction method according to claim 5, characterized in that, The method further includes the following steps: S5. The area controller determines whether the area is marked as a core refuge layer and whether the current air pressure is negative. S6. If this area is a core refuge layer and is under negative pressure, and the boundary state vector of the adjacent area received at the same time shows a positive pressure high-risk state, then ignore the migration impedance value calculation result in S3, forcibly execute the boundary fuse command, and output a control signal to start the pressurization and ventilation equipment in this area until the air pressure in this area changes to positive pressure; wherein, the pressurization and ventilation equipment adopts a closed-loop control strategy and continues to operate until the micro-differential pressure sensor feedback value shows that the air pressure in this area is higher than that in the external area; S7. The area controller sends a coordinated depressurization request to other adjacent areas that are downwind or in a low-risk state; the adjacent area controller that receives the coordinated depressurization request controls the smoke exhaust valve located in the adjacent area to open, so as to cooperate with the area to build a directional air pressure gradient and assist the area to change to a positive pressure state.

7. The distributed building escape path correction method according to claim 6, characterized in that, The method further includes the following steps: S81. The area controller obtains the channel length value and the visibility value within the channel leading to the adjacent area; S82. Calculate the estimated passage time required for personnel to pass through the passage based on the passage length value and the visibility value; wherein, the calculation model for the estimated passage time takes into account the influence of visibility on the movement speed of personnel, and the movement speed decreases non-linearly when the visibility decreases; S83. Compare the estimated travel time with the human tolerance limit time calculated based on the current toxic gas concentration value collected by the sensor data; wherein, the sensor data also includes the toxic gas concentration value; the human tolerance limit time is calculated based on the currently collected toxic gas concentration value using a toxicological model, and represents the maximum exposure time for which the human body will not suffer irreversible damage at a specific concentration. S84. If the estimated passage time is greater than the human body's tolerance limit time, then skip the calculation in S3 and directly generate a boundary circuit breaker command for the path leading to the adjacent area.

8. The distributed building escape path correction method according to claim 1, characterized in that, The step in S31 to determine whether the environmental airflow vector indicates that the airflow direction is from this region to the adjacent region is as follows: calculate the dot product of the environmental airflow vector and the normal vector leading to the boundary of the adjacent region. If the calculation result indicates the same direction, then it is determined that the airflow direction is from this region to the adjacent region.

Images