CN115569338B - Multi-category fire early warning data distributed training and local extinguishing method and system - Google Patents
Multi-category fire early warning data distributed training and local extinguishing method and system Download PDFInfo
- Publication number
- CN115569338B CN115569338B CN202211398580.5A CN202211398580A CN115569338B CN 115569338 B CN115569338 B CN 115569338B CN 202211398580 A CN202211398580 A CN 202211398580A CN 115569338 B CN115569338 B CN 115569338B
- Authority
- CN
- China
- Prior art keywords
- fire
- data
- detector
- time
- detectors
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A62—LIFE-SAVING; FIRE-FIGHTING
- A62C—FIRE-FIGHTING
- A62C31/00—Delivery of fire-extinguishing material
- A62C31/02—Nozzles specially adapted for fire-extinguishing
- A62C31/03—Nozzles specially adapted for fire-extinguishing adjustable, e.g. from spray to jet or vice versa
-
- A—HUMAN NECESSITIES
- A62—LIFE-SAVING; FIRE-FIGHTING
- A62C—FIRE-FIGHTING
- A62C37/00—Control of fire-fighting equipment
- A62C37/36—Control of fire-fighting equipment an actuating signal being generated by a sensor separate from an outlet device
- A62C37/38—Control of fire-fighting equipment an actuating signal being generated by a sensor separate from an outlet device by both sensor and actuator, e.g. valve, being in the danger zone
- A62C37/40—Control of fire-fighting equipment an actuating signal being generated by a sensor separate from an outlet device by both sensor and actuator, e.g. valve, being in the danger zone with electric connection between sensor and actuator
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Public Health (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Fire Alarms (AREA)
Abstract
The invention relates to a multi-category fire early warning data distributed training and local fire extinguishing method, which comprises the following steps: the main detector collects data when fire occurs; the fire control controller wirelessly transmits data to an edge server, and a fire judgment model is built on the edge server and distributed training is carried out; when the edge server does not obtain a trained fire judgment model yet, the fire controller judges fire through a fire judgment threshold value table, takes the ignition point detected by the main detector as the center of a circle, opens a control valve of the spray head, and sends fire judgment parameters to surrounding detectors; when the edge server obtains a trained fire judgment model, fire judgment is respectively carried out on all the detectors and the edge server according to the optimal segmentation points, and fire judgment parameters are obtained. The invention also discloses a multi-category fire early warning data distributed training and local fire extinguishing system. The invention can protect the object to be damaged to the maximum extent and save the fire extinguishing agent on the premise of high-efficiency and accurate fire extinguishment.
Description
Technical Field
The invention relates to the technical field of local fire extinguishment, in particular to a multi-category fire early warning data distributed training and local fire extinguishment method and system.
Background
Fire extinguishing systems are classified according to application modes, and can be classified into total flooding fire extinguishing systems and local application fire extinguishing systems. For fire in large space or long and narrow space such as cable interlayer, comprehensive pipe gallery and cable pit, etc., local application fire extinguishing system is adopted to directly spray gas to the protected object in specified time at designed spraying rate to form local high concentration around the protected object for a certain period of time to reach the aim of fire extinguishing.
The spray heads of the local application fire extinguishing system are uniformly arranged around the protection object, and when a fire disaster occurs, the fire extinguishing agent is directly and intensively sprayed onto the protection object, so that the fire extinguishing agent covers the whole outer surface of the protection object, namely, the fire extinguishing agent is implemented by achieving higher concentration of the fire extinguishing agent gas in a local range around the protection object. The existing local application fire extinguishing system adopts an action area method or fully covers a protected object for calculating the consumption of the fire extinguishing agent, and the method can simply and quickly realize fire extinguishing, but has some problems: firstly, the calculated fire extinguishing agent consumption is much larger than the actual fire extinguishing amount by adopting an action area method or fully covering the protected object, so that the fire extinguishing agent is wasted; on the other hand, the fire extinguishing agent is directly sprayed on the areas not at risk of fire by the existing action area method or the protection method of the whole coverage of the protected objects, so that the objects not at risk of fire are polluted. In addition, most of the existing fire extinguishing methods are judged based on sensor threshold values, and the mode may not be well applicable to local fire extinguishing.
Disclosure of Invention
The invention aims at providing the multi-category fire early warning data distributed training and local fire extinguishing method which can protect the disaster object to the maximum extent and save the fire extinguishing agent on the premise of high-efficiency and accurate fire extinguishing, improves the accuracy of fire judgment and ensures that the local fire extinguishing method has higher response speed.
In order to achieve the above purpose, the present invention adopts the following technical scheme: a method for distributed training and local fire suppression of multi-category fire early warning data, the method comprising the sequential steps of:
(1) The main detector collects the data of wind speed, CO concentration and temperature when fire disaster occurs;
(2) The main detector transmits the acquired data to the fire control controller, the fire control controller wirelessly transmits the received data to the edge server, and a fire judgment model is built on the edge server and distributed training is carried out;
(3) When the edge server does not obtain a trained fire judgment model yet, the fire controller judges fire through a built-in fire judgment threshold value table and sends fire judgment parameters to the main detector, meanwhile, the fire controller takes a fire point detected by the main detector as a circle center, a control valve of a spray head on the main detector is started, the main detector sends the fire judgment parameters to surrounding detectors, the surrounding detectors collect data of wind speed, CO concentration and temperature when fire occurs after receiving the fire judgment parameters and send the data to the fire controller, and the fire controller judges fire through the built-in fire judgment threshold value table and determines whether the control valve of the spray head on the surrounding detectors is started or not; when the fire is extinguished, if the fire controller judges that the fire is extinguished, the fire controller closes the control valves of the spray heads on all the detectors;
(5) When the edge server obtains a trained fire judgment model, the edge server sends the fire judgment model to all the detectors, wherein all the detectors comprise a main detector and surrounding detectors, and fire judgment is respectively carried out on all the detectors and the edge server according to the optimal segmentation points to obtain fire judgment parameters; the edge server sends the fire judgment parameters to the fire controller; the fire control controller takes the ignition point detected by the main detector as the center of a circle, a control valve of a spray head on the main detector is opened, the main detector sends fire judgment parameters to surrounding detectors, the surrounding detectors collect data of wind speed, CO concentration and temperature when a fire disaster occurs after receiving the fire judgment parameters and send the data to the fire control controller, and the fire control controller judges the fire through a built-in fire judgment threshold value table and determines whether to open the control valve of the spray head on the surrounding detectors; when the fire is extinguished, if the fire controller judges that the fire is extinguished, the fire controller closes the control valves of the spray heads on all the detectors.
The step (2) specifically comprises the following steps:
(2a) Establishing a fire judgment model, wherein the fire judgment model adopts a non-premixed combustion model, namely, the influence of turbulence is considered through a probability density function, a mixed fraction transportation equation and a conservation scalar equation are solved, then the concentration of each component is deduced from the predicted mixed fraction distribution, the fire condition is further judged, and a heat radiation model is adopted to calculate the fire spreading trend;
(2b) On the basis of building a fire judgment model, the data of the wind speed, the CO concentration and the temperature, which are acquired by the main detector, during the occurrence of the fire are input to the fire judgment model for training, so that a trained fire judgment model is obtained.
In step (4), the fire judgment is performed on all the detectors and the edge server according to the optimal division point, and the fire judgment parameter is obtained specifically including the following steps:
(4a) Collecting data of wind speed, CO concentration and temperature when a fire disaster occurs by all detectors, namely first data, selecting a layer of neural network in a fire judgment model as a prediction division point, wherein one part of the fire judgment model before the layer of neural network is executed by all detectors, and the other part of the fire judgment model after the layer of neural network is executed by an edge server; inputting the first data acquired by all the detectors into a fire judgment model on the detector, executing the first data to a predicted division point of the fire judgment model, stopping executing the first data, outputting a result, namely second data, and simultaneously obtaining first processing time, wherein the first processing time refers to the time for obtaining the second data from the first data; the detector uploads the second data to the fire-fighting controller, the fire-fighting controller uploads the received second data to the edge server to obtain first transmission time, and the first transmission time refers to the time from the detector uploading the second data to the edge server receiving the second data; the edge server receives the uploaded second data, executes a fire judgment model after predicting the dividing points to obtain fire judgment parameters, and the period from the time when the edge server receives the uploaded second data to the time when the fire judgment model is executed is a second processing time; the edge server transmits the fire judgment parameters back to the fire control controller, wherein the back transmission time is the second transmission time;
(4b) Based on the first processing time, the second processing time, the first transmission time and the second transmission time, an optimal division point of the fire judgment model is obtained, the fire judgment models in all the detectors and the edge servers are divided at the optimal division point, the fire judgment model is a first block before the optimal division point and a second block after the optimal division point, the first block of the fire judgment model processes first data in the detectors, and the second block of the fire judgment model processes second data in the edge servers.
In step (4 b), the obtaining the optimal division point of the fire judgment model based on the first processing time, the second processing time, the first transmission time and the second transmission time specifically refers to:
definition of the definitionWhich represents the first data generated by the j-th detector in the corresponding deployment space in the i-th slot, i= … nJ= … m, definition +.>The total time required for fire determination by this data is represented, and consists of four parts: first treatment time->First transmission time->Second treatment time->And a second transmission time->
The minimum is obtained by the following methodValue:
definition of the definitionRepresenting an optimal division point of a fire judgment model to be adopted by a detector j in a corresponding deployment space in the i time slot; definition C i [j]In the time slot i, the minimum time of the detector j is valued; first data generated on the first detector i.e. j=1 of the first time slot i.e. i=1 +.>In the minimization +.>On the basis of which the dividing point is selected to obtain the firstFire determination time of detector> Optimal division point of fire judgment model +.>
For first data generated on the second detector of the first time slot i=1 j=2Selecting optimal division point of fire judgment model>And gets the fire judgment time +.> Wherein->Meaning that the first allocation C1[1 ] is satisfied]The minimum +.>By analogy, for a certain detector q of the first time slot, first data +.q is generated for j=q>The optimal division point of the corresponding fire judgment model is +.>The fire condition judging time is +.>After this, the optimal division point set of the fire judgment model in all the detectors in the first time slot is obtained>And total fire determination time->
When the second time slot starts, if a certain detector q in the previous time slot executes fire judgment and then finds that a fire disaster occurs in a responsible area, the fire judgment needs to be preferentially carried out in the round, and the fire judgment is used as the first detector of the time slot; if the fire occurs in the area which is responsible for the plurality of detectors, sequencing the fire according to the severity of the fire, dividing the severity of the fire according to the threshold range of the fire judgment threshold table, otherwise, executing according to the sequence of the last time slot; thus, the division point and the fire judgment time of the fire judgment model of the first i.e., j=1 detector of i=2, which is the second time slot, are respectivelyAnd-> The second detector, i.e. the fire judgment model with j=2, has optimal division point and fire judgment time of +.>And->And so on, obtaining the optimal division point set of the fire judgment model in the second time slot +.>And total fire determination time->
And analogizing the time slots to obtain an optimal division point set X= { X of the fire judgment model of all the detectors in the final total time slot 1 ,X 2 ,……,X n Fire determination time in total time slot
Another object of the present invention is to provide a system for a multi-category fire early warning data distributed training and local extinguishing method, comprising:
the main detector is used for collecting and uploading parameters of wind speed, CO concentration and temperature when a fire disaster occurs, and the detector closest to the ignition point is the main detector;
the fire control controller is used for receiving the data uploaded by the detector, uploading the data to the edge server, transmitting the information sent by the received edge server back to the detector, and arranging a fire judgment threshold value table in the fire control controller;
the edge server is used for receiving parameters of wind speed, CO concentration and temperature and corresponding fire extinguishing operation when a fire disaster occurs, and training to obtain a fire judgment model capable of judging the fire and sending out corresponding fire extinguishing instructions;
the spray head is used for spraying fire extinguishing agent, has the spraying time and spraying force of different gears and is controlled by the fire control controller, and the spray head is arranged on the main detector and the surrounding detectors;
and the surrounding detector is used for collecting and uploading parameters of wind speed, CO concentration and temperature when the fire disaster occurs.
According to the space potential ignition condition, a fire extinguishing partition is arranged, a main detector, a fire control controller and surrounding detectors are deployed in the fire extinguishing area, an edge server is remotely arranged, the main detector, the surrounding detectors and the fire control controller are in wired connection with each other, and the fire control controller and the edge server are in wireless communication.
According to the technical scheme, the beneficial effects of the invention are as follows: firstly, compared with the traditional local application fire extinguishing system, the invention arranges the detector and the spray head in a mode of setting fire extinguishing areas, and implements a corresponding fire extinguishing method according to actual fire conditions, thereby maximally protecting the disaster-stricken object and saving the fire extinguishing agent on the premise of high-efficiency and accurate fire extinguishing; secondly, the detector can collect actual parameters of a fire extinguishing area for neural network training to obtain a fire judgment model, so that the accuracy of fire judgment is improved; thirdly, the invention adopts a model segmentation method to accelerate the speed of the fire judgment process, and the designed optimal segmentation point algorithm can lead the fire judgment model to be executed on the detector and the edge server in a blocking way so as to accelerate the speed of the whole fire judgment process, thereby leading the local fire extinguishing method to have faster response speed.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram of a large space partial fire;
FIG. 3 is a schematic view of a local fire in an elongated space;
FIG. 4 is a flow chart of the operation of the main detector and the surrounding detectors.
Detailed Description
As shown in fig. 1 and 2, a multi-category fire early warning data distributed training and local fire extinguishing method comprises the following sequential steps:
(1) The main detector collects the data of wind speed, CO concentration and temperature when fire disaster occurs;
(2) The main detector transmits the acquired data to the fire control controller, the fire control controller wirelessly transmits the received data to the edge server, and a fire judgment model is built on the edge server and distributed training is carried out;
(3) When the edge server does not obtain a trained fire judgment model yet, the fire controller judges fire through a built-in fire judgment threshold value table and sends fire judgment parameters to the main detector, meanwhile, the fire controller takes a fire point detected by the main detector as a circle center, a control valve of a spray head on the main detector is started, the main detector sends the fire judgment parameters to surrounding detectors, the surrounding detectors collect data of wind speed, CO concentration and temperature when fire occurs after receiving the fire judgment parameters and send the data to the fire controller, and the fire controller judges fire through the built-in fire judgment threshold value table and determines whether the control valve of the spray head on the surrounding detectors is started or not; when the fire is extinguished, if the fire controller judges that the fire is extinguished, the fire controller closes the control valves of the spray heads on all the detectors;
(6) When the edge server obtains a trained fire judgment model, the edge server sends the fire judgment model to all the detectors, wherein all the detectors comprise a main detector and surrounding detectors, and fire judgment is respectively carried out on all the detectors and the edge server according to the optimal segmentation points to obtain fire judgment parameters; the edge server sends the fire judgment parameters to the fire controller; the fire control controller takes the ignition point detected by the main detector as the center of a circle, a control valve of a spray head on the main detector is opened, the main detector sends fire judgment parameters to surrounding detectors, the surrounding detectors collect data of wind speed, CO concentration and temperature when a fire disaster occurs after receiving the fire judgment parameters and send the data to the fire control controller, and the fire control controller judges the fire through a built-in fire judgment threshold value table and determines whether to open the control valve of the spray head on the surrounding detectors; when the fire is extinguished, if the fire controller judges that the fire is extinguished, the fire controller closes the control valves of the spray heads on all the detectors.
The step (2) specifically comprises the following steps:
(2a) Establishing a fire judgment model, wherein the fire judgment model adopts a non-premixed combustion model, namely, the influence of turbulence is considered through a probability density function, a mixed fraction transportation equation and a conservation scalar equation are solved, then the concentration of each component is deduced from the predicted mixed fraction distribution, the fire condition is further judged, and a heat radiation model is adopted to calculate the fire spreading trend;
(2b) On the basis of building a fire judgment model, the data of the wind speed, the CO concentration and the temperature, which are acquired by the main detector, during the occurrence of the fire are input to the fire judgment model for training, so that a trained fire judgment model is obtained.
In step (4), the fire judgment is performed on all the detectors and the edge server according to the optimal division point, and the fire judgment parameter is obtained specifically including the following steps:
(4a) Collecting data of wind speed, CO concentration and temperature when a fire disaster occurs by all detectors, namely first data, selecting a layer of neural network in a fire judgment model as a prediction division point, wherein one part of the fire judgment model before the layer of neural network is executed by all detectors, and the other part of the fire judgment model after the layer of neural network is executed by an edge server; inputting the first data acquired by all the detectors into a fire judgment model on the detector, executing the first data to a predicted division point of the fire judgment model, stopping executing the first data, outputting a result, namely second data, and simultaneously obtaining first processing time, wherein the first processing time refers to the time for obtaining the second data from the first data; the detector uploads the second data to the fire-fighting controller, the fire-fighting controller uploads the received second data to the edge server to obtain first transmission time, and the first transmission time refers to the time from the detector uploading the second data to the edge server receiving the second data; the edge server receives the uploaded second data, executes a fire judgment model after predicting the dividing points to obtain fire judgment parameters, and the period from the time when the edge server receives the uploaded second data to the time when the fire judgment model is executed is a second processing time; the edge server transmits the fire judgment parameters back to the fire control controller, wherein the back transmission time is the second transmission time;
(4b) Based on the first processing time, the second processing time, the first transmission time and the second transmission time, an optimal division point of the fire judgment model is obtained, the fire judgment models in all the detectors and the edge servers are divided at the optimal division point, the fire judgment model is a first block before the optimal division point and a second block after the optimal division point, the first block of the fire judgment model processes first data in the detectors, and the second block of the fire judgment model processes second data in the edge servers.
In step (4 b), the obtaining the optimal division point of the fire judgment model based on the first processing time, the second processing time, the first transmission time and the second transmission time specifically refers to:
definition of the definitionIt means that in the ith time slot i= … n, the first data generated by the jth detector in the corresponding deployment space, j= … m, define +.>The total time required for fire determination by this data is represented, and consists of four parts: first treatment time->First transmission time->Second treatment time->And a second transmission time->
The minimum is obtained by the following methodValue:
definition of the definitionRepresenting the best of the fire decision models to be taken by the detector j in the corresponding deployment space in the i time slotDividing points; definition C i [j]In the time slot i, the minimum time of the detector j is valued; first data generated on the first detector i.e. j=1 of the first time slot i.e. i=1 +.>In the minimization +.>Selecting the dividing point to obtain the fire judging time of the first detector +.> Optimal division point of fire judgment model +.>
For first data generated on the second detector of the first time slot i=1 j=2Selecting optimal division point of fire judgment model>And gets the fire judgment time +.> Wherein->Meaning that the first allocation C1[1 ] is satisfied]The minimum +.>By analogy, for a certain detector q of the first time slot, first data +.q is generated for j=q>The optimal division point of the corresponding fire judgment model is +.>The fire condition judging time is +.>After this, the optimal division point set of the fire judgment model in all the detectors in the first time slot is obtained>And total fire determination time->
When the second time slot starts, if a certain detector q in the previous time slot executes fire judgment and then finds that a fire disaster occurs in a responsible area, the fire judgment needs to be preferentially carried out in the round, and the fire judgment is used as the first detector of the time slot; if the fire occurs in the area which is responsible for the plurality of detectors, sequencing the fire according to the severity of the fire, dividing the severity of the fire according to the threshold range of the fire judgment threshold table, otherwise, executing according to the sequence of the last time slot; thus, the division point and the fire judgment time of the fire judgment model of the first i.e., j=1 detector of i=2, which is the second time slot, are respectivelyAnd-> The second detector, j=2, is the best score for the fire judgment modelThe cutting point and the fire judgment time are respectively +.>And->And so on, obtaining the optimal division point set of the fire judgment model in the second time slot +.>And total fire determination time->
And analogizing the time slots to obtain an optimal division point set X= { X of the fire judgment model of all the detectors in the final total time slot 1 ,X 2 ,……,X n Fire determination time in total time slot
The system comprises:
the main detector is used for collecting and uploading parameters of wind speed, CO concentration and temperature when a fire disaster occurs, and the detector closest to the ignition point is the main detector;
the fire control controller is used for receiving the data uploaded by the detector, uploading the data to the edge server, transmitting the information sent by the received edge server back to the detector, and arranging a fire judgment threshold value table in the fire control controller;
the edge server is used for receiving parameters of wind speed, CO concentration and temperature and corresponding fire extinguishing operation when a fire disaster occurs, and training to obtain a fire judgment model capable of judging the fire and sending out corresponding fire extinguishing instructions;
the spray head is used for spraying fire extinguishing agent, has the spraying time and spraying force of different gears and is controlled by the fire control controller, and the spray head is arranged on the main detector and the surrounding detectors;
and the surrounding detector is used for collecting and uploading parameters of wind speed, CO concentration and temperature when the fire disaster occurs.
According to the space potential ignition condition, a fire extinguishing partition is arranged, a main detector, a fire control controller and surrounding detectors are deployed in the fire extinguishing area, an edge server is remotely arranged, the main detector, the surrounding detectors and the fire control controller are in wired connection with each other, and the fire control controller and the edge server are in wireless communication.
The invention is further described below in connection with fig. 1 to 4.
The detector collects and uploads parameters such as wind speed, CO concentration, temperature and the like when a fire disaster occurs, and a singlechip is arranged in the detector, so that reasoning operation, namely fire judgment operation, can be performed, and the detectors can be linked; the main detector can only broadcast fire to surrounding detectors, and does not perform operations such as interaction with the surrounding detectors, and all the detectors have completely independent detection and algorithm execution.
As shown in fig. 4, the detector which detects the fire first is used as the main detector, and the main detector always broadcasts the fire judgment parameters according to the frequency of once per second in the period between the fire finding to the room temperature and the lowering to the threshold value; and the surrounding detector confirms the control of the surrounding detector on the spray head under the control of the surrounding detector after weighted average according to the received fire judgment parameters and the fire judgment parameters calculated by the surrounding detector.
The fire control controller can automatically receive the detector signals and automatically clear redundant information under normal conditions; when the fire control controller receives an abnormal signal of a certain detector, the fire control controller can automatically identify whether the signal is true or false according to the signals of other detectors in parallel and automatically clear false signals. When a fire occurs, the fire judgment parameters obtained according to a fire judgment threshold value table, namely a table 1 or an edge server are transmitted to a detector, and actions such as pre-starting, starting and closing of the fire extinguishing spray nozzle are controlled until the temperature is reduced to below 50 ℃ and the spray nozzle is kept unchanged for 10 minutes, wherein the table 1 is as follows:
TABLE 1
Spray head spray is controlled by two factors: the spray force is adjustable in multiple gears, and more spray schemes are provided for combined spray of different spray heads; secondly, when the spraying time is lower than the threshold value, all the spray heads are turned off, the fire is controlled, and the spraying gear can be regulated down or part of spray heads can be turned off at proper time according to the fire condition.
During the execution of the algorithm, the data transmission between the detector and the edge server and the data processing of the edge server both bring about certain time consumption. In a practical scenario, in order to ensure the timeliness of the fire response, it is necessary to make the response time of the fire extinguishing apparatus as small as possible, which requires that the time of both parts be as small as possible. Because the detector is connected with the fire control controller in a wired way, the transmission time between the detector and the fire control controller is negligible, and the data uploading and returning time only considers the wireless transmission time between the fire control controller and the edge server. The time of wireless transmission is mainly related to the network condition, and compared with the time of wireless transmission, the time of data processing is reduced as much as possible. In order to reduce the time consumption of the data processing process, it is considered to accelerate the inference process by using a model segmentation method in the inference phase, i.e., the fire decision phase. Specifically, the fire judgment model to be calculated is divided into two parts, one part is executed on a detector with weak calculation power, the other part is executed on an edge server with strong calculation power, and the calculation at the two ends is parallel.
In summary, the following several time consumptions need to be considered in the case of introducing model segmentation. I.e. the processing time (first processing time) t of the acquisition data at the detector s Data upload (first transmission time) t u And a backhaul time (second transmission time) t d Processing time (second processing time) t of data at edge server w 。
In the model segmentation process, when the original data volume or the intermediate data volume is large and complex, the detector or the edge server may not be able to process in time, and thus the resulting data latency time may affect the total time of the reasoning process. It is therefore necessary to find suitable segmentation points for model segmentation, minimizing the sum of the above-mentioned times.
Example 1
For local fire in large space, as shown in fig. 2, when a fire occurs at an ignition point, a detector sends out an alarm control signal, the system receives the position of the fire, firstly, a c2/d2/d3/d4/c4/c3/c2 nozzle is started to extinguish the fire, meanwhile, the detector in the area monitors the temperature of the area to be reduced in real time and is maintained below 100 ℃, the fire is considered to be controlled, the spreading phenomenon does not occur, the detector monitors the temperature of the area to be maintained below 50 ℃ in real time for 10 minutes, the fire is considered to be extinguished, and the control system sends out an instruction to stop spraying the fire extinguishing agent. If the detector in the area monitors the problem of the area continuously rising in real time, and the problem indicates that the fire has a spreading trend, starting a b1/c1/d1/e1/e2/e3/e4/e5/d5/c5/b5/b4/b3/b2 nozzle to start extinguishing, considering that the fire is extinguished when the temperature of the area is kept below 50 ℃ for 10 minutes, and the control system sends out an instruction to stop spraying the fire extinguishing agent, and so on.
Example two
For local fire in long and narrow space, as shown in fig. 3, when a fire occurs at an ignition point, a detector sends out an alarm control signal, the system receives the position of the fire, firstly, a b/c spray head is started to extinguish the fire, meanwhile, the detector in the area monitors the temperature drop of the area in real time and maintains the temperature lower than 100 ℃, the fire is considered to be controlled, the spreading phenomenon does not occur, the detector monitors the temperature of the area in real time and maintains the temperature lower than 50 ℃ for 10 minutes, the fire is considered to be extinguished, and the control system sends out an instruction to stop spraying the fire extinguishing agent. If the detector in the area monitors the problem in the area continuously rises in real time, and the problem indicates that the fire has a spreading trend, the a … d spray head is started to extinguish the fire, when the temperature of the area is kept below 50 ℃ for 10 minutes, the fire is considered to be extinguished, and the control system sends out a command to stop spraying the fire extinguishing agent, and so on.
In summary, the detector and the spray heads are arranged in a mode of setting fire extinguishing areas, and a corresponding fire extinguishing method is implemented according to actual fire conditions, so that fire extinguishing agents can be saved while objects suffering from disasters are protected to the maximum extent on the premise of high-efficiency and accurate fire extinguishing; the detector can collect actual parameters of a fire extinguishing area for neural network training to obtain a fire judgment model, so that the accuracy of fire judgment is improved; the invention adopts the method of model segmentation to accelerate the speed of the fire judgment process, and the designed optimal segmentation point algorithm can lead the fire judgment model to be executed on the detector and the edge server in a blocking way so as to accelerate the speed of the whole fire judgment process, thereby leading the local fire extinguishing method to have faster response speed.
Claims (2)
1. A multi-category fire early warning data distributed training and local fire extinguishing method is characterized in that: the method comprises the following steps in sequence:
(1) The main detector collects the data of wind speed, CO concentration and temperature when fire disaster occurs;
(2) The main detector transmits the acquired data to the fire control controller, the fire control controller wirelessly transmits the received data to the edge server, and a fire judgment model is built on the edge server and distributed training is carried out;
(3) When the edge server does not obtain a trained fire judgment model yet, the fire controller judges fire through a built-in fire judgment threshold value table and sends fire judgment parameters to the main detector, meanwhile, the fire controller takes a fire point detected by the main detector as a circle center, a control valve of a spray head on the main detector is started, the main detector sends the fire judgment parameters to surrounding detectors, the surrounding detectors collect data of wind speed, CO concentration and temperature when fire occurs after receiving the fire judgment parameters and send the data to the fire controller, and the fire controller judges fire through the built-in fire judgment threshold value table and determines whether the control valve of the spray head on the surrounding detectors is started or not; when the fire is extinguished, if the fire controller judges that the fire is extinguished, the fire controller closes the control valves of the spray heads on all the detectors;
(4) When the edge server obtains a trained fire judgment model, the edge server sends the fire judgment model to all the detectors, wherein all the detectors comprise a main detector and surrounding detectors, and fire judgment is respectively carried out on all the detectors and the edge server according to the optimal segmentation points to obtain fire judgment parameters; the edge server sends the fire judgment parameters to the fire controller; the fire control controller takes the ignition point detected by the main detector as the center of a circle, a control valve of a spray head on the main detector is opened, the main detector sends fire judgment parameters to surrounding detectors, the surrounding detectors collect data of wind speed, CO concentration and temperature when a fire disaster occurs after receiving the fire judgment parameters and send the data to the fire control controller, and the fire control controller judges the fire through a built-in fire judgment threshold value table and determines whether to open the control valve of the spray head on the surrounding detectors; when the fire is extinguished, if the fire controller judges that the fire is extinguished, the fire controller closes the control valves of the spray heads on all the detectors;
the step (2) specifically comprises the following steps:
(2a) Establishing a fire judgment model, wherein the fire judgment model adopts a non-premixed combustion model, namely, the influence of turbulence is considered through a probability density function, a mixed fraction transportation equation and a conservation scalar equation are solved, then the concentration of each component is deduced from the predicted mixed fraction distribution, the fire condition is further judged, and a heat radiation model is adopted to calculate the fire spreading trend;
(2b) On the basis of building a fire judgment model, inputting the data of the wind speed, the CO concentration and the temperature, acquired by the main detector, during the occurrence of the fire to the fire judgment model for training, so as to obtain a trained fire judgment model;
in step (4), the fire judgment is performed on all the detectors and the edge server according to the optimal division point, and the fire judgment parameter is obtained specifically including the following steps:
(4a) Collecting data of wind speed, CO concentration and temperature when a fire disaster occurs by all detectors, namely first data, selecting a layer of neural network in a fire judgment model as a prediction division point, wherein one part of the fire judgment model before the layer of neural network is executed by all detectors, and the other part of the fire judgment model after the layer of neural network is executed by an edge server; inputting the first data acquired by all the detectors into a fire judgment model on the detector, executing the first data to a predicted division point of the fire judgment model, stopping executing the first data, outputting a result, namely second data, and simultaneously obtaining first processing time, wherein the first processing time refers to the time for obtaining the second data from the first data; the detector uploads the second data to the fire-fighting controller, the fire-fighting controller uploads the received second data to the edge server to obtain first transmission time, and the first transmission time refers to the time from the detector uploading the second data to the edge server receiving the second data; the edge server receives the uploaded second data, executes a fire judgment model after predicting the dividing points to obtain fire judgment parameters, and the period from the time when the edge server receives the uploaded second data to the time when the fire judgment model is executed is a second processing time; the edge server transmits the fire judgment parameters back to the fire control controller, wherein the back transmission time is the second transmission time;
(4b) Based on the first processing time, the second processing time, the first transmission time and the second transmission time, an optimal division point of a fire judgment model is obtained, the fire judgment model in all the detectors and the edge servers is divided at the optimal division point, the fire judgment model is a first block before the optimal division point and a second block after the optimal division point, the first block of the fire judgment model processes first data in the detectors, and the second block of the fire judgment model processes second data in the edge servers;
in step (4 b), the obtaining the optimal division point of the fire judgment model based on the first processing time, the second processing time, the first transmission time and the second transmission time specifically refers to:
definition of the definitionIt means that in the ith time slot i= … n, the first data generated by the jth detector in the corresponding deployment space, j= … m, define +.>The total time required for fire determination by this data is represented, and consists of four parts: first treatment time->First transmission time->Second treatment time->And a second transmission time
The minimum is obtained by the following methodValue:
definition of the definitionRepresenting an optimal division point of a fire judgment model to be adopted by a detector j in a corresponding deployment space in the i time slot; definition C i [j]In the time slot i, the minimum time of the detector j is valued; first data generated on the first detector i.e. j=1 of the first time slot i.e. i=1 +.>In the minimization +.>Selecting the dividing point to obtain the fire judging time of the first detector +.> Optimal division point of fire judgment model +.>
For first data generated on the second detector of the first time slot i=1 j=2Selecting optimal division point of fire judgment model>And gets the fire judgment time +.> Wherein the method comprises the steps ofMeaning that the first allocation C1[1 ] is satisfied]The minimum +.>By analogy, for a certain detector q of the first time slot, first data +.q is generated for j=q>The optimal dividing point of the corresponding fire judgment model isThe fire condition judging time is +.> After this, the optimal division point set of the fire judgment model in all the detectors in the first time slot is obtained>And total fire determination time->
When the second time slot starts, if a certain detector q in the previous time slot executes fire judgment and then finds that a fire disaster occurs in a responsible area, the fire judgment needs to be preferentially carried out in the round, and the fire judgment is used as the first detector of the time slot; if the fire occurs in the area which is responsible for the plurality of detectors, sequencing the fire according to the severity of the fire, dividing the severity of the fire according to the threshold range of the fire judgment threshold table, otherwise, executing according to the sequence of the last time slot; thus, the division point and the fire judgment time of the fire judgment model of the first i.e., j=1 detector of i=2, which is the second time slot, are respectivelyAnd the second detector, i.e. the fire judgment model with j=2, has optimal division point and fire judgment time of +.>And->And so on, obtaining the optimal division point set of the fire judgment model in the second time slot +.>And total fire determination time->
And analogizing the time slots to obtain an optimal division point set X= { X of the fire judgment model of all the detectors in the final total time slot 1 ,X 2 ,……,X n Fire determination time in total time slot
2. A system for implementing the multi-category fire early warning data distributed training and localized fire suppression method of claim 1, characterized by: comprising the following steps:
the main detector is used for collecting and uploading parameters of wind speed, CO concentration and temperature when a fire disaster occurs, and the detector closest to the ignition point is the main detector;
the fire control controller is used for receiving the data uploaded by the detector, uploading the data to the edge server, transmitting the information sent by the received edge server back to the detector, and arranging a fire judgment threshold value table in the fire control controller;
the edge server is used for receiving parameters of wind speed, CO concentration and temperature and corresponding fire extinguishing operation when a fire disaster occurs, and training to obtain a fire judgment model capable of judging the fire and sending out corresponding fire extinguishing instructions;
the spray head is used for spraying fire extinguishing agent, has the spraying time and spraying force of different gears and is controlled by the fire control controller, and the spray head is arranged on the main detector and the surrounding detectors;
the surrounding detector is used for collecting and uploading parameters of wind speed, CO concentration and temperature when a fire disaster occurs;
according to the space potential ignition condition, a fire extinguishing partition is arranged, a main detector, a fire control controller and surrounding detectors are deployed in the fire extinguishing area, an edge server is remotely arranged, the main detector, the surrounding detectors and the fire control controller are in wired connection with each other, and the fire control controller and the edge server are in wireless communication.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211398580.5A CN115569338B (en) | 2022-11-09 | 2022-11-09 | Multi-category fire early warning data distributed training and local extinguishing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211398580.5A CN115569338B (en) | 2022-11-09 | 2022-11-09 | Multi-category fire early warning data distributed training and local extinguishing method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115569338A CN115569338A (en) | 2023-01-06 |
| CN115569338B true CN115569338B (en) | 2023-10-13 |
Family
ID=84589338
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211398580.5A Active CN115569338B (en) | 2022-11-09 | 2022-11-09 | Multi-category fire early warning data distributed training and local extinguishing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115569338B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB1474680A (en) * | 1974-12-19 | 1977-05-25 | Cerberus Ag | Fire extinguishing installations |
| JP2005168645A (en) * | 2003-12-09 | 2005-06-30 | Ohbayashi Corp | Fireproof compartment system |
| CN110046837A (en) * | 2019-05-20 | 2019-07-23 | 北京唐芯物联网科技有限公司 | A kind of fire management system based on artificial intelligence |
| CN111035872A (en) * | 2020-01-02 | 2020-04-21 | 中车青岛四方车辆研究所有限公司 | Battery box fire prevention and control system and method |
| CN111127811A (en) * | 2019-12-25 | 2020-05-08 | 上海船舶电子设备研究所(中国船舶重工集团公司第七二六研究所) | Fire early detection alarm system and method |
| CN111686392A (en) * | 2020-06-23 | 2020-09-22 | 海南科技职业大学 | Artificial intelligence fire extinguishing system is surveyed to full scene of vision condition |
| CN112156399A (en) * | 2020-07-22 | 2021-01-01 | 浙江蓝盾电工新材料科技有限公司 | Fire-fighting system for echelon early warning and multiple times of accurate discharge of lithium battery energy storage units |
| CN114528987A (en) * | 2022-02-15 | 2022-05-24 | 中国科学院上海微***与信息技术研究所 | Neural network edge-cloud collaborative computing segmentation deployment method |
| CN216676774U (en) * | 2021-12-22 | 2022-06-07 | 福建永福电力设计股份有限公司 | Water spray fire extinguishing system for transformer |
| CN114699701A (en) * | 2022-04-08 | 2022-07-05 | 中车大连机车研究所有限公司 | Fire prevention and control device capable of realizing accurate fire extinguishing |
-
2022
- 2022-11-09 CN CN202211398580.5A patent/CN115569338B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB1474680A (en) * | 1974-12-19 | 1977-05-25 | Cerberus Ag | Fire extinguishing installations |
| JP2005168645A (en) * | 2003-12-09 | 2005-06-30 | Ohbayashi Corp | Fireproof compartment system |
| CN110046837A (en) * | 2019-05-20 | 2019-07-23 | 北京唐芯物联网科技有限公司 | A kind of fire management system based on artificial intelligence |
| CN111127811A (en) * | 2019-12-25 | 2020-05-08 | 上海船舶电子设备研究所(中国船舶重工集团公司第七二六研究所) | Fire early detection alarm system and method |
| CN111035872A (en) * | 2020-01-02 | 2020-04-21 | 中车青岛四方车辆研究所有限公司 | Battery box fire prevention and control system and method |
| CN111686392A (en) * | 2020-06-23 | 2020-09-22 | 海南科技职业大学 | Artificial intelligence fire extinguishing system is surveyed to full scene of vision condition |
| CN112156399A (en) * | 2020-07-22 | 2021-01-01 | 浙江蓝盾电工新材料科技有限公司 | Fire-fighting system for echelon early warning and multiple times of accurate discharge of lithium battery energy storage units |
| CN216676774U (en) * | 2021-12-22 | 2022-06-07 | 福建永福电力设计股份有限公司 | Water spray fire extinguishing system for transformer |
| CN114528987A (en) * | 2022-02-15 | 2022-05-24 | 中国科学院上海微***与信息技术研究所 | Neural network edge-cloud collaborative computing segmentation deployment method |
| CN114699701A (en) * | 2022-04-08 | 2022-07-05 | 中车大连机车研究所有限公司 | Fire prevention and control device capable of realizing accurate fire extinguishing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115569338A (en) | 2023-01-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210027602A1 (en) | Enhanced audiovisual analytics | |
| CN106385457B (en) | Cloud service platform kitchen work environment intelligent warning method and system | |
| CN103830855A (en) | Dynamic firefighting emergency evacuation indicating system for large-scale public building | |
| JP6186409B2 (en) | Wireless communication network system and management method for managing multiple areas | |
| KR101180411B1 (en) | Automatic Control Apparatus of a fire door | |
| CN112116775A (en) | Fire detection method of multi-split air-conditioning equipment and multi-split air-conditioning system | |
| CN110011883B (en) | Intelligent building Internet of things information marginalization analysis and local control system | |
| CN115204701B (en) | Fire risk prevention and control method, system, equipment and storage medium for stadium | |
| CN115569338B (en) | Multi-category fire early warning data distributed training and local extinguishing method and system | |
| CN114748813B (en) | Quick fire extinguishing method, system, equipment and storage medium suitable for high-rise building | |
| CN111888675B (en) | Automatic fire-fighting air cushion control platform and method | |
| CN116052350A (en) | Fire disaster early warning and rescue system and method based on road-asking guiding navigation mode | |
| CN117482445A (en) | Intelligent simple spraying fire early-stage suppression system based on Internet of things | |
| JP5276764B2 (en) | Fire monitoring system | |
| RU2589617C1 (en) | Method for detection and control of abnormal situation, intelligent station to this end | |
| CN110486898A (en) | Air conditioning control method, device and air-conditioning system based on User Activity rule | |
| CN115662041A (en) | Fire extinguishing system based on fire sensing and consequence prediction of transformer substation | |
| CN114887267A (en) | Intelligent fire positioning method and partitioned fire extinguishing device for power cabin of comprehensive pipe rack | |
| CN110624193B (en) | Fire control method for preventing afterburning based on Internet of things | |
| CN208506834U (en) | A kind of building security system | |
| EP3448081B1 (en) | Device and method for automated tuning of auto-channel selection (acs) in wireless networks | |
| CN117236671B (en) | Dynamic real-time monitoring method and system for mountain fire of power transmission line | |
| CN117592768B (en) | Fire disaster management method and system for internal places of enterprises based on big data processing | |
| JPH0540887A (en) | Refuge simulation system and refuge guidance system | |
| KR20200027079A (en) | Apparatus and method for optimum control of air conditioner using artificial neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |