Intelligent fireproof door triggering alarm system and method

Abstract

The invention relates to the technical field of fire alarm, in particular to an intelligent fireproof door trigger alarm system and method, and the method comprises the steps: taking a fireproof door as a configuration unit according to a sensor disposed on the fireproof door, and obtaining the early warning data of each configuration unit; taking the early warning data as a decision basis, performing event triggering rule extraction on the early warning data on each configuration unit in a fuzzy theory mode, and identifying an event set at corresponding time during triggering; sorting according to the triggering time of each data in the event set, and extracting a state tag and an event influence range of each configuration unit according to the event duration of each sorting to form a range simulation diagram; and according to the position of each configuration unit in the range simulation diagram, the state of the configuration unit is updated according to the difference of the early warning data at the two sides of the configuration unit, an early warning path is formed by the updated configuration unit, and an early warning strategy corresponding to the early warning path is set. The efficiency and accuracy of fire early warning processing are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of fire alarms, and in particular to an intelligent fire door triggering alarm system and method. Background Art

[0002] Fire doors serve as fire barriers within buildings. Fire alarms are typically installed outside these doors. Fire doors and fire alarms are essential components of modern building fire safety. Early warning devices installed on fire doors can be used to monitor the spread of fires, based on the door's status, providing a basis for fire response.

[0003] For example, Chinese patent publication number CN118570960A discloses an intelligent alarm triggering device and fire door, which relates to the field of signal alarm technology. The method includes: a stairwell positioning module obtains a building structure design drawing and extracts multiple stairwells of the target building, including multiple sets of fire doors on multiple floors; a door status recognition module retrieves multiple surveillance video sets to identify the fire door status and obtain multiple first safety factor sets; a floor status information acquisition module collects floor status information and floor evacuation partition information sets; an environmental safety recognition module performs environmental safety recognition and obtains multiple second safety factor sets; a safety level matching module matches fire safety levels based on multiple first safety factor sets and multiple second safety factor sets to obtain multiple fire safety levels; and an early warning module is used to issue early warnings for multiple stairwells based on multiple fire safety levels and generate early warning instructions. The prior art describes evacuation alarms in the event of an open flame fire, mainly for the rapid evacuation of multiple corridors in the event of a real-time open flame fire.

[0004] For example, Chinese Patent Publication No. CN111325940A discloses a fuzzy-theory-based intelligent fire linkage method and system for substations. The method includes: obtaining status monitoring data for substation main equipment; obtaining status monitoring data for fire monitoring sensors; estimating the fire development trend based on the substation main equipment status monitoring data and the fire monitoring sensor status monitoring data based on fuzzy theory, and generating fire linkage decision information; issuing an alarm based on the fire linkage decision information; obtaining security equipment status monitoring data; opening doors on evacuation paths and closing fire doors based on the fire linkage decision information; and performing fire extinguishing actions based on the fire linkage decision information. The method includes storing event triggering conditions, event action interruption conditions, and event action resumption condition rules in a rule base. Prior art describes relative time under fire progression, primarily focusing on fire development trends to identify fire propagation and spread at each point under fire extinguishing actions.

[0005] However, the above-mentioned existing technologies ignore the regional risks and the evolution of isolated events in a single area during fire handling, resulting in the inability to timely update the warning content at the location of the fire door when identifying the evolution of the fire, leading to errors in identifying the fire area and reducing the efficiency and accuracy of fire warning. Summary of the Invention

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: an intelligent fire door trigger alarm system, characterized in that it includes: a fire protection construction module, which is used to configure the fire door as a configuration unit based on the sensors configured on the fire door and the data received by the sensors, and obtain the early warning data of each configuration unit.

[0007] The event division module is used to use the warning data as the basis for decision-making and determine the warning data that falls within the warning range. When the coordinates of the warning data are the same, the event triggering rules of the warning data on each configuration unit are extracted using fuzzy theory to identify the event set at the corresponding time when the trigger occurs.

[0008] The event association module is used to sort the event set according to the trigger time of each data, extract the status label of each configuration unit and the event impact range based on the event duration of each sort, and form a range simulation diagram.

[0009] The warning path module is used to update the status of the configuration units according to the position of each configuration unit in the range simulation diagram and the difference in warning data on both sides of the configuration unit, form a warning path with the updated configuration units, and set the warning strategy corresponding to the warning path.

[0010] An intelligent fire door triggering alarm method is characterized by comprising: S1, taking the fire door as a configuration unit and obtaining early warning data of each configuration unit.

[0011] S2, taking the warning data as the basis for decision making, extracts event triggering rules from the warning data on each configuration unit using fuzzy theory, and identifies the event set corresponding to the time when the trigger occurs.

[0012] S3, sort the event set according to the trigger time of each data, and extract the status label and event impact range of each configuration unit based on the event duration of each sort, to form a range simulation diagram.

[0013] S4, according to the position of each configuration unit in the range simulation diagram, the configuration unit status is updated based on the difference in warning data on both sides of the configuration unit, a warning path is formed with the updated configuration unit, and a warning strategy corresponding to the warning path is set.

[0014] The beneficial effects of the present invention are: 1. The present invention collects data through temperature sensors, smoke sensors and fire sensors, uses the moving average of a time sliding window for threshold recognition, and through multi-source data fusion and priority mechanism, merges and aggregates multiple warning clear data according to their warning status to determine fire events in different states.

[0015] 2. The present invention describes event triggering rules based on fuzzy theory, and uses the action entity and replacement entity of the event triggering rules to judge and identify the interruption conditions, maintenance conditions and end conditions of each configuration unit, and then defines an event set with different state label combinations; according to the triggering time of each data in the event set, the events are aggregated in a sliding time window to determine the influence range of each location during the fire evolution process, further verify the fire change trend and fire source location, and dynamically adjust the warning range under the fire warning.

[0016] 3. The present invention searches for adjacent configuration units and identifies state transition paths through the positions of each configuration unit in the range simulation diagram. After clarifying the regional gaps between the configuration units, the path information of multiple corridors and intersections contained in the warning path is combined with the strategy of the current configuration unit to output the relative processing method of the warning path in the fire scenario, thereby improving the accuracy of the warning path processing when an alarm is triggered under a fire. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below with reference to the accompanying drawings and examples.

[0018] Figure 1 It is a system diagram of an intelligent fire door trigger alarm system.

[0019] Figure 2 The present invention is a flow chart of an event division module of an intelligent fire door triggering alarm system.

[0020] Figure 3 The present invention is a flowchart of an event correlation module of an intelligent fire door trigger alarm system.

[0021] Figure 4 The invention is a flowchart of a method for triggering an alarm of an intelligent fire door. DETAILED DESCRIPTION

[0022] The following embodiments of the present invention are described in detail. The embodiments described below are exemplary and are only used to explain the present invention, and are not to be construed as limiting the present invention. Where specific techniques or conditions are not specified in the embodiments, the techniques or conditions described in the literature in the art or in the product specifications shall be followed.

[0023] See Figure 1An intelligent fire door trigger alarm system includes: a fire protection construction module, an event division module, an event association module and an early warning path module; wherein the output end of the fire protection construction module is connected to the event division module, the output end of the event division module is connected to the event association module, and the output end of the event association module is connected to the early warning path module.

[0024] The fire protection construction module is used to obtain early warning data of each configuration unit based on the sensors configured on the fire door and the data received by the sensors, taking the fire door as the configuration unit.

[0025] The event division module is used to use the warning data as the basis for decision-making and determine the warning data that falls within the warning range. When the coordinates of the warning data are the same, the event triggering rules of the warning data on each configuration unit are extracted using fuzzy theory to identify the event set at the corresponding time when the trigger occurs.

[0026] The event association module is used to sort the event set according to the trigger time of each data, extract the status label of each configuration unit and the event impact range based on the event duration of each sort, and form a range simulation diagram.

[0027] The warning path module is used to update the status of the configuration units according to the position of each configuration unit in the range simulation diagram and the difference in warning data on both sides of the configuration unit, form a warning path with the updated configuration units, and set the warning strategy corresponding to the warning path.

[0028] The configuration unit receives the sensing data from the positioning sensors at various locations, uses the sensors corresponding to temperature, smoke, and fire, receives any one of the temperature, smoke, and fire states, forms an early warning basis, and uses temperature sensors, smoke sensors, and flame detectors to complete the environmental information of the scene where the current fire door is located.

[0029] Smoke sensors, such as photoelectric smoke detectors, detect smoke particles produced by combustion and trigger an alarm based on the principle of light scattering. Temperature sensors can take the form of fixed-temperature detectors or infrared detection arrays to monitor ambient temperature, triggering an alarm when a threshold is reached to indicate a significant temperature rise at the current location. Flame detectors, such as ultraviolet flame detectors or infrared beam smoke detectors, identify flames by detecting the corresponding light beam of burning substances. These sensors can be placed on one side of a fire door or on both the inside and outside of the corridor where the fire door is located to simultaneously detect potential fires in the corridor and units.

[0030] When obtaining the early warning data of each configuration unit, its implementation method also includes: performing threshold recognition on the temperature, smoke and fire-related data collected in each configuration unit, taking temperature as the first priority, smoke as the second priority, and fire as the third priority, and judging each priority in turn with the moving average of the time sliding window, and outputting the part that meets each priority as the early warning data of each configuration unit.

[0031] Preferably, when temperature is used as the first priority for judgment, the measured single-point continuous temperature is used as the judgment standard, the time series is divided into multiple time sliding windows, and the judgment is made based on the moving average of the single-point continuous temperature. If the value meets the average value of the temperature that can be measured by the fire door when a fire occurs, it means that with temperature as the first priority, a high-temperature area appears at the current fire door position, and a fire may occur. At the same time, for the identified temperature value, the current temperature value will be output to the early warning data with the collected coordinates.

[0032] Preferably, when smoke is taken as the second priority, the concentration value is collected in real time, and the critical value of the risk of suffocation of personnel is reached as the content of the smoke judgment. At the same time, when judging the second priority, it is necessary to first identify the temperature related to the first priority, and then identify the smoke-related value, and mark whether the two data meet the fire-related types in the early warning data.

[0033] Preferably, when the fire is taken as the third priority, the flame sensor is used to identify the relevant content in the ultraviolet band and the infrared band, and use it as the third parameter for identification. When there are abnormalities in the current two, the data related to the third priority will be obtained, and then the part with the flickering flame will be used as the warning data output at this time.

[0034] Preferably, when checking the data according to the three priorities, the acquired data also needs to be divided into the collected data corresponding to warning, normal and fire, and the upper and lower limit values of the fire door warning in the historical data are used to distinguish its status to illustrate the corresponding data of the current fire door being in any one of the fire state, normal state and warning state.

[0035] In one embodiment of the present invention, when dividing the warning data by time, a unified coordinate system transformation is performed based on the location of each configuration unit, and the warning data of the same coordinates are merged. For example, if a fire door triggers a smoke alarm and an abnormal temperature alarm at the same time, the corresponding warning data are merged to generate a composite event collection. If there is a contradiction in the data of the same coordinate afterwards, such as smoke exceeding the limit but the temperature is normal, the corresponding data needs to be checked and the warning range set when the current fire door is warning is obtained. The warning range indicates that some parameters in the physical area that the sensor can monitor have entered the smoke warning, temperature warning, etc., which require real-time monitoring data, and some data for alarm processing of the upper equipment, such as in corridors, stairwells or the entire floor, to identify the current fire door. The fire door obtains whether there are anomalies in the data, and processes these parts with event triggering rules; the event triggering rules represent single-parameter triggered warnings and multi-parameter triggered warnings, and record the parameters that continuously maintain the warning as their maintenance conditions; when the warning ends, the parameters return to the safety threshold part as their end conditions; the conditions that lead to a surge in parameter identification values and abnormal states of fire doors, such as the fire door being opened abnormally, or a surge in temperature and smoke collected by the fire door, are used as interruption conditions to illustrate that the fire is gradually worsening; the data corresponding to these three conditions are used as the main events currently identified to illustrate the relative situation of the fire identified at the location of each fire door, and the time point when each condition is extracted is used as its trigger time to illustrate the components of the event set.

[0036] The implementation method of determining the warning data that falls within the warning range also includes: performing a circumferential determination based on the model corresponding to the current fire door, the monitoring radius and the distance between multiple fire doors, and identifying a single warning data that falls within the warning range.

[0037] Based on the acquired single warning data, the values of each configuration unit in the warning state are verified respectively, and the warning data with the same coordinates are merged.

[0038] Preferably, it is necessary to check the fire doors currently installed in the corridors and hallways according to their models, the area that can be monitored, and the distance of each fire door in the corresponding floor and corridor. The fire doors in a single corridor or hallway should be checked in a circular manner to check whether there is a fire warning in the area where the fire door is set. If so, single warning data related to temperature, fire and smoke are obtained, and then the source coordinates of these warning data are determined. If there is a consistency, they are merged to ensure that there is no contradiction in the status of adjacent fire doors.

[0039] When extracting event trigger rules, its implementation method also includes: receiving the interruption conditions, maintenance conditions and end conditions after the event triggering rules, judging whether the states identified by each configuration unit are consistent, and if consistent, based on any adjacent configuration units in the same interruption conditions, maintenance conditions and end conditions, counting the number and triggering time of corresponding configuration units, and setting a status label for each configuration unit.

[0040] If they are inconsistent, based on the time points at which the current configuration condition is in the interruption condition, maintenance condition, and end condition, a status tag is set for each configuration unit according to the time length of the interruption condition, maintenance condition, and end condition.

[0041] After the configuration unit has identified the status tag setting, the configuration unit updates the status of each configuration unit at the relative time points when the interruption condition, the maintenance condition, and the end condition are triggered.

[0042] Preferably, the status described above is used to indicate whether the area where the fire door is located is in any one of the states of normal, warning, and fire. These states will be explained through the interruption, maintenance and end related to the warning, as well as the interruption, maintenance and end corresponding to the fire, to illustrate the status of the area where the fire door is currently located; then the data corresponding to these interruptions, maintenance and ends will be used as its status label to indicate whether there is a warning, fire, etc. in the area where each fire door is located.

[0043] Preferably, if adjacent configuration units are in the same state, the primary focus is identifying the number of units triggering the alert and the time of the initial trigger to monitor abnormal conditions in the area where the fire door is located. For example, if smoke remains abnormally active, the corresponding configuration units are counted to filter out isolated false alarms and enhance the credibility of the alert in the area where the fire door is located. If the states are inconsistent, the duration of the persistent state is continuously monitored over time to distinguish between momentary disturbances and persistent fire risks, thus avoiding label confusion.

[0044] When each configuration unit is subsequently updated, the first triggering time of its interruption condition, maintenance condition, and end condition and the time of status update are used, and the relative time difference is used to illustrate the fire progress in the area where each fire door is located. When updating the status, it is updated in the form of interruption condition > maintenance condition > end condition to illustrate whether there is any data collection anomaly under the current fire door, as well as the status label of the fire duration status. The relative status of each configuration unit in the fire is described by using the new label to overwrite the old label.

[0045] As for the fuzzy theory method described above, it will use the form of membership to describe the warning data under different state labels. The data in the normal state will be represented by 0 as its membership, and the data in the fire state will be represented by 1 as its membership. The data in the warning state will be set as the ratio of the difference between the current warning value and the lower limit value in the warning state as its membership. According to the value of the membership, the currently identified warning data will be sorted and combined into an event set according to the states of the interruption condition, maintenance condition and end condition.

[0046] When sorting by membership, if it is a composite event combining multiple sets of data such as temperature, smoke, and fire, the membership is weighted and summed, and weights of 0.3, 0.5, and 0.2 are set for temperature, smoke, and fire, respectively. The weight setting will be based on the description of the possible spread state of the fire after the anomaly is identified in the area where the fire door is located. For example, in the early stages of a fire, the impact of smoke is more serious, and an alarm needs to be issued based on the smoke identification data, and identification is combined with obvious temperature anomalies. Finally, the overall situation is judged based on the fire situation.

[0047] The output event set at this time is sorted by membership and combined with the status label to generate a structured event set. For example, fire door A triggers a fire warning at 14:23 due to smoke and temperature anomalies, which lasts until 14:30. Each data in the event set is represented as a related data description form.

[0048] Preferably, the event triggering rule represents whether the current position of the fire door is in any one of the states of normal, warning, and fire, and the data triggered in this state are analyzed to obtain a relevant data set.

[0049] like Figure 2 As shown, the implementation method of the event partitioning module also includes: extracting the action entity of the event triggering rule, and determining each replacement entity associated with the action entity, performing comprehensive calculation according to the total relevant area of the replacement entity, and obtaining the configuration unit related to the current action entity.

[0050] At this time, the action entity represents the core action that triggers the event. For example, if the current alarm is identified and extracted through sensor data threshold judgment, and the corresponding smoke value is detected, then the action entity is the smoke detection action, which is used to segment the main action corresponding to each data in the event set; then the replacement entity is the alternative action that replaces the original action entity in a specific scenario. For example, when the fire door is damaged, the smoke alarm is replaced by the door magnetic alarm. That is, after identifying the current action entity and combining it with the relevant status reported by the fire door, based on the rule base mapping, check whether there is a replacement entity for the current fire door. That is, when the fire door status is abnormally open, smoke detection and other methods are mapped to door magnetic anomalies to verify whether the corresponding position of the fire door is damaged by fire or there is an abnormal fault itself.

[0051] Furthermore, the action entity represents the quantitative standards such as smoke concentration and temperature detected by the configured sensor, and describes the main action it implements with the actual operation and name of the detection. The replacement entity is an alternative solution to the original action entity in a specific scenario. Its existence is to solve the problem of failure or insufficient coverage of the original action entity. For example, when the sensor executing the action entity is abnormal, other sensors in the same fire zone are used as replacements. And when the door magnetic field of the fire door is abnormal, the door magnetic alarm is used to supplement the content detected by the action entity, thereby realizing the functional complementarity of the action entity and the replacement entity.

[0052] The total relevant area of the replacement entity represents the area of the fire door in the physical space, which is the superposition value of the influence range of the action entity and its replacement entity in the physical space. After superimposing these areas, the fire door position is used as the node, and the coverage area is generated according to the monitoring radius. The overlapping parts are merged and the corresponding configuration units are marked.

[0053] The interruption condition, maintenance condition and termination condition of each configuration unit are judged to identify whether the adjacent configuration units are in the same state, and the state of each configuration unit is updated, and the updated configuration unit is output as an event set.

[0054] Preferably, the monitoring radius is used to indicate the radius of a fixed range that the current fire door configuration sensor can identify. This radius is obtained based on the parameters of the set sensor and is used to indicate whether the data collected within the fixed range can be considered as a form of fire warning.

[0055] In one embodiment of the present invention, in the event association module, the trigger time of each data in the event set is mainly used as the benchmark, and after sorting according to the timestamp, a time series about the timeline is generated, and then the configuration units related to each event are aggregated to analyze the event impact range and the path under which the status label is propagated, so as to form a range simulation diagram in combination to complete the analysis of the abnormal propagation path corresponding to the fire trigger alarm in the event set.

[0056] like Figure 3 As shown, the implementation method of the event association module includes: sorting according to the triggering time of each data in the event set, arranging the event set in ascending order according to the timestamp, and generating an event time series corresponding to the event set.

[0057] Events are aggregated on the event time series using a sliding time window. The distances between multiple configuration units under the same event aggregation are determined. The event impact range under the same event aggregation is set based on the closest distance of the corresponding configuration units. A range simulation diagram is constructed using the closest distance and status label of the configuration units under the same event aggregation.

[0058] If there are no configuration units under the same event aggregation in the current event time sequence, the state transition paths of each configuration unit are extracted according to the transition of the state labels of each configuration unit, and the state transition paths are combined into a range simulation graph.

[0059] The aforementioned aggregation of identical events primarily uses spatiotemporal correlation to quickly locate regional risks, determine whether the current scenario involves fire spread, and address mid- and late-stage fire warnings. Aggregation of non-identical events tracks the evolution of isolated events at each fire door through state transitions to check for equipment anomalies in fire scenarios and the relative extent of certain areas in fire scenarios, thereby completing early-stage warning processing.

[0060] Preferably, in the same event aggregation scenario, the configuration units with early warnings and fire warnings in the same time window are regarded as the same event aggregation scenario in the form of a time window. Then, for any configuration unit, the configuration unit closest to the current configuration unit is found. This distance is regarded as the shortest distance between the configuration units. Then, the center point between the configuration units in the current same event scenario is found, and the distance from the center point to other configuration units is the minimum. Then, the minimum circumscribed circle of the composite shape composed of this center point and multiple other configuration units is used as the current event impact range. These configuration units are connected, and after marking the status labels, they are output as a range simulation diagram.

[0061] Preferably, the closest distance can be calculated using Euclidean distance, emphasizing the need to distribute the fire warning locations in the same event window. The Euclidean distance is directly calculated using the coordinates of the corresponding configuration unit, which can help understand the concentration trend of the fire in the overall space and then determine the spatial scope involved in the current fire.

[0062] Preferably, when there is a warning-related status tag, the time window can be set to 5 minutes to obtain a larger range of fire warning identification; when the status tag identified by the fire door is in a fire, the time window is set to 1 minute to quickly track the fire status.

[0063] Preferably, when they are not in the same event aggregation, it means that there are isolated warning and fire data in the time series. The data is usually represented on a single configuration unit. At this time, the impact range of the event generated will be located according to the monitoring radius of the configuration unit; and the state transition path of the corresponding configuration unit from normal to warning or fire state is viewed in the form of a time series to verify whether there is an abnormality in the corresponding fire door when obtaining data.

[0064] Preferably, when identifying the state transition path, the fire door state at the corresponding position in the time series is used for identification. As for the state label, it is a numerical value of the corresponding state that is combined into an event set. Here, the state label is used to indicate whether there is a state change in the parameters identified by the corresponding fire door.

[0065] The implementation method of the above-mentioned state transition path includes: searching for adjacent configuration units for the current configuration unit, starting from the state tags of the adjacent configuration units, and using the update time of each state tag as a state node. Each state node includes the timestamp, state type, temperature, smoke, and fire value of the corresponding configuration unit.

[0066] According to the timestamp of the state node, the time interval between each state node is recorded in a continuous verification manner. If the time interval of the state node is greater than the preset time interval and at least two parameters in each state node are abnormal, the state nodes are connected to form a state transition path, and the normal state of the current configuration unit is used as the starting point and end point of the state transition path.

[0067] At this time, a preset time interval is set to prevent instantaneous abnormal changes. The preset time interval can be set to 20s to obtain relatively complete data, and finally form a state transition path with normal starting point - temperature sudden rise - warning state - detection of open flame - interruption state - continuous deterioration of parameters - maintenance state - parameter recovery to safe value - end point. At least two parameter abnormalities are identified in the state transition path. Only when two parameter abnormalities occur can the data collected by the sensor configured at the current fire door be determined to be fire-related data. If there is only a single abnormality, it is easy to cause data point mutation and data collection abnormality, resulting in misidentification of the area where each configuration unit is located.

[0068] After completing the state transition path, the path display corresponding to the state transition path is output as a range simulation diagram to illustrate the evolution of fire warnings at different configuration units.

[0069] In one embodiment of the present invention, the early warning data on both sides of the configuration unit are used to identify the location of the abnormality identified by the configuration unit, such as the location of the fire source on the inside and outside of the door, and the abnormal conditions identified on the inside and outside of the door when a fire occurs, and update the status of the corresponding positions to assist in locating the fire source and the smoke spread path, as well as the description of the fire expansion path.

[0070] For example, if the temperature difference between the inside and outside of a door is greater than 35°C, the hotter side is the fire's source. Using smoke data, if the concentration difference is greater than 10% obs / m², the smoke's spread path can be determined, indicating that smoke will diffuse from the side with higher concentration. Furthermore, based on the updated status of the configuration unit, the fire intensity data can be added to the configuration unit to illustrate the fire's spread. The values of 35°C and 10% obs / m² are for illustrative purposes only. For specific identification scenarios, the average values of actual data at different locations should be used for judgment.

[0071] That is, when the configuration unit status is updated, its implementation method includes: extracting multiple target configuration units from the range simulation map, demarcating regional gaps based on the locations of the target configuration units, sorting the target configuration units based on the maximum recognition length under the regional gaps, and connecting the sorted target configuration units one by one to form an early warning path corresponding to the target configuration unit.

[0072] The target configuration units selected above are extracted from the range simulation map based on different combinations. The spatial distribution of these target configuration units is then determined. Specifically, the location between the two fire doors is located in the corridor corresponding to the target configuration unit. The corridors and intersections within the monitoring radius of the fire doors are then abstracted as directed edges and described as regional gaps. Regional gaps are the uncovered areas between units and are used to assess the severity of risks, such as the potential for fire spread or weak points in escape routes. After abstracting the corridors and intersections in the uncovered areas as edges, the uncovered areas are represented by the edge length to indicate their maximum recognition length.

[0073] The early warning path can quickly locate the current fire source by connecting fire doors with larger gaps, and prevent the fire from spreading by linking and closing the fire doors. For example, if the gap between fire doors A and B is too large during the current fire warning state, the system will generate a path connecting A→B and add evacuation instructions. At the same time, the regional gaps represent channels where the fire may spread, and the direction of fire spread needs to be monitored through the path to prevent the fire from spreading further.

[0074] Preferably, sorting is performed based on regional gaps. When completing the warning path, it is necessary to connect the location of the fire door with the corridor path, and use the maximum recognition length as the weight of the connected configuration unit. According to the weight, the uncovered area between each fire door in the current warning path is identified to facilitate the search for subsequent warning strategies.

[0075] When the early warning path module is implemented, its implementation method also includes: using the path information of multiple corridors and intersections contained in the early warning path to control the scene location of the target configuration unit, using the order of event triggering rules and status labels of corresponding events of multiple target configuration units in the controlled area to set the boundary information of each target configuration unit, and checking the early warning strategy under each target configuration unit, and gradually outputting the early warning strategy according to the preset time period to describe how to identify and process the spread of fire when a fire occurs at the location covered by the current fire door.

[0076] When generating warning paths, the building structure is abstracted into a node graph, with each fire door location as a node and hallways / corridors as edges, each assigned a direction of travel. A diffusion simulation is then performed on the currently received warning data, simulating the overall path of material combustion or fire spread based on the real-time data received. These are then graded to indicate the severity of the fire. For example, if a single unit warning is received, adjacent units are issued. Based on the updated data after the warning, the multiple units with updated warning data are connected. Changes in the positional data on both sides of the unit are then identified to identify the primary paths of smoke and fire spread, facilitating the generation of appropriate warning strategies.

[0077] At the same time, when the warning path is completed and its warning strategy is screened, the number and status labels of the obtained target configuration units are used to screen the warning strategies that are compatible with the current warning path from the database. This includes if the proportion of target configuration units in normal state on the warning path reaches more than 70%, then the warning strategy of directional evacuation based on the current warning path is executed; if the proportion of target configuration units in interrupted state on the warning path exceeds 40%, then the warning strategy of isolation based on the warning path is executed, and the strategies that are compatible with the currently generated warning path are checked in turn. It should be noted that the status described at this time represents the status label set by the interruption condition, end condition, and maintenance condition, and is used to illustrate the status of different target configuration units on the warning path.

[0078] As for the order of event triggering rules, the time and status change time of each target configuration unit on the warning path are extracted, and the changed status label is synchronized to the current target configuration unit. Combined with the number of target configuration units with the same status label at this time, the current warning strategy is checked and identified to illustrate the situation of the fire during its spread and after treatment.

[0079] Preferably, after receiving the early warning strategy of each target configuration unit, based on the status label of each target configuration unit, it is determined whether the status labels of adjacent target configuration units are consistent. When they are consistent, the distance between the adjacent target configuration units is obtained using the Manhattan distance algorithm, that is, the distance value identified by the axis of the coordinate system is directly calculated using the three-dimensional coordinates of the current target configuration unit, and only the size of the difference between the two points in each dimension is concerned. The sum of the absolute values of the coordinate differences corresponding to X, Y, and Z on the three-dimensional coordinates is used to describe the distance from a target configuration unit with the same status label to another target configuration unit.

[0080] Afterwards, the warning strategies are stored for the adjacent target configuration units with the same status labels until the status label of the target configuration unit on the current warning path is updated. The warning strategies before and after the update are compared and stored in the database as associated data.

[0081] If the status labels of adjacent target configuration units are inconsistent, the number of units with the same status label as the current target configuration unit on the warning path is checked, and the warning strategy of each target configuration unit is stored in the database according to the number of units. The corresponding data of the status labels of each target configuration unit belonging to fire and warning are set as the associated data of the warning strategy.

[0082] like Figure 4 As shown, the present invention also provides a method for triggering an alarm for an intelligent fire door, comprising: S1, taking the fire door as a configuration unit and obtaining early warning data of each configuration unit.

[0083] S2, taking the warning data as the basis for decision making, extracts event triggering rules from the warning data on each configuration unit using fuzzy theory, and identifies the event set corresponding to the time when the trigger occurs.

[0084] S3, sort the event set according to the trigger time of each data, and extract the status label and event impact range of each configuration unit based on the event duration of each sort, to form a range simulation diagram.

[0085] S4, according to the position of each configuration unit in the range simulation diagram, the configuration unit status is updated based on the difference in warning data on both sides of the configuration unit, a warning path is formed with the updated configuration unit, and a warning strategy corresponding to the warning path is set.

[0086] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered by the scope of protection of the present invention.

Claims

1. An intelligent fire door trigger alarm system, characterized in that: include: The fire protection construction module is used to obtain early warning data of each configuration unit based on the sensors configured on the fire door and the data received by the sensors; The event partitioning module is used to use the warning data as the basis for decision-making and determine the warning data that falls within the warning range. When the coordinates of the warning data are the same, the event triggering rules of the warning data on each configuration unit are extracted using fuzzy theory to identify the event set at the corresponding time when the trigger occurs. The event association module is used to sort the event set according to the trigger time of each data, extract the status label of each configuration unit and the event impact range based on the event duration of each sort, and form a range simulation diagram; The warning path module is used to update the status of the configuration units according to the position of each configuration unit in the range simulation diagram and the difference in warning data on both sides of the configuration unit, form a warning path with the updated configuration units, and set the warning strategy corresponding to the warning path.

2. The intelligent fire door trigger alarm system according to claim 1, characterized in that: When obtaining the warning data of each configuration unit, the implementation method also includes: Threshold recognition is performed using the temperature, smoke, and fire-related data collected in each configuration unit. Temperature is taken as the first priority, smoke as the second priority, and fire as the third priority. The moving average of the time sliding window is used to judge each priority in turn, and the parts that meet each priority are output as warning data for each configuration unit.

3. The intelligent fire door trigger alarm system according to claim 1, characterized in that: Methods for determining the warning data that falls within the warning range also include: Based on the model of the current fire door, the monitoring radius and the distance between multiple fire doors, a cyclic determination is performed to identify single warning data that falls within the warning range; Based on the acquired single warning data, the values of each configuration unit in the warning state are verified respectively, and the warning data with the same coordinates are merged.

4. The intelligent fire door trigger alarm system according to claim 1, characterized in that: The implementation of the event partitioning module also includes: Extract the action entity of the event triggering rule, determine each replacement entity associated with the action entity, perform comprehensive calculation based on the total relevant area of the replacement entities, and obtain the configuration unit related to the current action entity; The interruption condition, maintenance condition and termination condition of each configuration unit are judged to identify whether the adjacent configuration units are in the same state, and the state of each configuration unit is updated, and the updated configuration unit is output as an event set.

5. The intelligent fire door trigger alarm system according to claim 4, characterized in that: When extracting event trigger rules, the implementation method also includes: Receive the interruption condition, maintenance condition, and termination condition after the event triggering rule, determine whether the states identified by each configuration unit are consistent, and if consistent, count the number and triggering time of corresponding configuration units based on any adjacent configuration units in the same interruption condition, maintenance condition, and termination condition, and set a state label for each configuration unit; If they are inconsistent, then based on the time points at which the current configuration condition is in the interruption condition, maintenance condition, and end condition, a status label is set for each configuration unit according to the time length of the interruption condition, maintenance condition, and end condition; After the configuration unit has identified the status tag setting, the configuration unit updates the status of each configuration unit at the relative time points when the interruption condition, the maintenance condition, and the end condition are triggered.

6. The intelligent fire door trigger alarm system according to claim 1, characterized in that: The implementation of the event correlation module includes: Sort the event set according to the trigger time of each data, arrange the event set in ascending order according to the timestamp, and generate the event time series corresponding to the event set; Aggregate the event time series using a sliding time window to determine the distances between multiple configuration units under the same event aggregation. Use the closest distances between the corresponding configuration units to set the event impact range under the same event aggregation. Use the closest distances and status labels of the configuration units under the same event aggregation to construct a range simulation diagram. If there are no configuration units under the same event aggregation in the current event time sequence, the state transition paths of each configuration unit are extracted according to the transition of the state labels of the configuration units, and the state transition paths are combined into a range simulation graph.

7. The intelligent fire door trigger alarm system according to claim 6, characterized in that: The implementation methods of state transition path include: Search for neighboring configuration units for the current configuration unit, starting from the state labels of the neighboring configuration units, and taking the update time of each state label as the state node; According to the timestamp of the state node, the time interval between each state node is recorded in a continuous verification manner. If the time interval of the state node is greater than the preset time interval and at least two parameters in each state node are abnormal, the state nodes are connected to form a state transition path, and the normal state of the current configuration unit is used as the starting point and end point of the state transition path.

8. The intelligent fire door trigger alarm system according to claim 1, characterized in that: When updating the configuration unit status, the implementation methods include: Multiple target configuration units are extracted from the range simulation map, and regional gaps are delineated based on the locations of the target configuration units. The target configuration units are sorted based on the maximum recognition length under the regional gaps, and the sorted target configuration units are connected one by one to form a warning path corresponding to the target configuration unit.

9. The intelligent fire door trigger alarm system according to claim 1, characterized in that: When the early warning path module is implemented, its implementation method also includes: The path information of multiple corridors and intersections included in the early warning path is used to control the scene location of the target configuration unit. The boundary information of each target configuration unit is set by using the order of event triggering rules and the status labels of corresponding events of multiple target configuration units in the controlled area. The early warning strategy under each target configuration unit is checked and the early warning strategy is gradually output according to the preset time period.

10. A method for triggering an alarm for an intelligent fire door, characterized in that: include: S1, taking the fire door as a configuration unit, obtains the early warning data of each configuration unit; S2, using the warning data as the basis for decision making, extracts event triggering rules from the warning data on each configuration unit using fuzzy theory to identify the event set corresponding to the trigger time; S3, sort the event set according to the trigger time of each data, extract the status label and event impact range of each configuration unit based on the event duration of each sort, and form a range simulation diagram; S4, according to the position of each configuration unit in the range simulation diagram, the configuration unit status is updated based on the difference in warning data on both sides of the configuration unit, a warning path is formed with the updated configuration unit, and a warning strategy corresponding to the warning path is set.

Images