KR102369267B1 - Fire risk optimization value derivation method and a fire alarm system that transmits data for simulation processing using the same - Google Patents

Abstract

According to an embodiment of the present invention, a fire alarm system includes a plurality of sensors which detects the occurrence of a fire and generates fire information, respectively, a repeater, a receiver, and a first server. The first server includes: a building information modeling unit which provides modeling information by implementing a virtual building; a big data receiving unit which receives big data from an external second server; a server memory which collects and stores the fire information received from the plurality of sensors; a server analysis unit which generates analysis data based on the modeling information, the big data, and the fire information; and a risk calculation unit which calculates a risk based on the analysis data, wherein the risk calculation unit may calculate a different risk according to a type of the building. The present invention provides the fire alarm system capable of deriving a risk optimization value.

Description

FIRE RISK OPTIMIZATION VALUE DERIVATION METHOD AND A FIRE ALARM SYSTEM THAT TRANSMITS DATA FOR SIMULATION PROCESSING USING THE SAME

The present invention relates to a method of deriving a fire risk optimization value and a fire alarm system using the same, and more particularly, to a fire alarm system that transmits data for simulation processing using the method of deriving a fire risk optimization value.

In general, a fire detector may notify the risk of fire by transmitting a change in a state of a fire, a temperature, or smoke to a receiving device through a voltage. However, when transmitted through voltage, the signal that a fire has occurred transmitted from the fire detector is information that is used and disappears as soon as a fire occurs, and it is difficult to use it later. In this case, when the fire detector malfunctions due to certain conditions, the fire detector may malfunction for the same reason if the same conditions are given next time. Users may lose confidence in the fire alarm system and are more likely to turn it off normally. In this case, users may be exposed to the risk of fire even if an actual fire occurs in the area where the fire alarm system is installed.

An object of the present invention is to provide a fire alarm system that transmits data for simulation processing with the aim of deriving a risk optimization value.

A fire alarm system according to an embodiment of the present invention includes a plurality of sensors, each of which generates fire information by detecting the occurrence of a fire, and performs RF communication (Radio Frequency communication) with each other, the plurality of sensors and the RF communication A building comprising a repeater for performing a, a receiver for performing the RF communication with the repeater, and a first server for performing the RF communication with the receiver, wherein the first server provides modeling information by virtualizing a building An information modeling unit, a big data receiving unit for receiving big data from an external second server, a server memory for collecting and storing the fire information received from the plurality of sensors, the modeling information, the big data, and the fire A server analysis unit generating analysis data based on the information, and a risk calculation unit calculating a degree of risk based on the analysis data, wherein the risk calculation unit may calculate a different level of risk depending on the type of the building.

The risk calculator may calculate the risk in real time.

The first server may transmit an upgrade signal generated based on the analysis data to the plurality of sensors, and each of the plurality of sensors may generate the fire information based on the upgrade signal.

Each of the plurality of sensors may generate different fire information for each building based on the upgrade signal.

It may further include a data extraction unit for extracting simulation data by reinforcement learning the analysis data.

The first server may receive information from the outside through a variable value adjustment dialog, and the risk calculator may calculate the risk in real time based on the simulation data and the information.

The first server may calculate a fire escape route based on the simulation data.

The fire escape route may be different depending on the type of the building.

The big data includes information on a first building and information on a second building different from the first building, and the server analysis unit includes the modeling information, information on the first building, and information on the second building. can be compared to determine whether the building is the first building or the second building.

When it is determined that the building is the first building, the first server transmits a first upgrade signal for setting the sensing sensitivities of the plurality of sensors to the first sensitivity to each of the plurality of sensors, and the building is the first building. When it is determined that the second building is a building, a second upgrade signal for setting the sensing sensitivity to a second sensitivity different from the first sensitivity may be transmitted to each of the plurality of sensors.

A method of deriving a fire risk optimization value according to an embodiment of the present invention includes generating fire information by a plurality of sensors, providing modeling information by realizing a building virtually, and receiving and storing big data from an external server. step, generating analysis data based on the fire information, the modeling information, and the big data, and calculating the risk of fire in real time based on the analysis data, and calculating the risk level may include calculating a different degree of risk according to the type of the building.

extracting simulation data by reinforcement learning the analysis data; receiving information through a variable value adjustment dialog window from the outside; and calculating the risk based on the simulation data, the information, and the analysis data. may include more.

The method may further include generating an upgrade signal based on the analysis data, transmitting the upgrade signal to the plurality of sensors, and generating different fire information for each building based on the upgrade signal.

As described above, the first server may calculate the analysis data. The analysis data may include an optimal value for judging the risk of fire according to a type of a building in which a plurality of sensors are installed by combining modeling information, fire information, and big data. Accordingly, it is possible to provide a fire alarm system capable of deriving a risk optimization value.

1 illustrates fire alarm systems according to an embodiment of the present invention.
2 is a perspective view illustrating one sensor among a plurality of sensors according to an embodiment of the present invention.
3 is a perspective view showing a repeater according to an embodiment of the present invention.
4 shows a receiver according to an embodiment of the present invention.
5 illustrates a first server according to an embodiment of the present invention.
6 illustrates an operation of a server analyzer according to an embodiment of the present invention.
7 is a diagram illustrating an operation of a risk calculator according to an embodiment of the present invention.
8 is a diagram illustrating reinforcement learning according to an embodiment of the present invention.
9 illustrates a sensor and a first server according to an embodiment of the present invention.
10 illustrates buildings and a first server according to an embodiment of the present invention.
11 shows a fire escape route according to an embodiment of the present invention.

In this specification, when a component (or region, layer, portion, etc.) is referred to as being “on,” “connected to,” or “coupled to” another component, it is directly disposed/on the other component. It means that it can be connected/coupled or a third component can be placed between them.

Like reference numerals refer to like elements. In addition, in the drawings, thicknesses, ratios, and dimensions of components are exaggerated for effective description of technical content.

“and/or” includes any combination of one or more that the associated configurations may define.

Terms such as first, second, etc. may be used to describe various elements, but the elements should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may also be referred to as a first component. The singular expression includes the plural expression unless the context clearly dictates otherwise.

In addition, terms such as "below", "below", "above", and "upper side" are used to describe the relationship of the components shown in the drawings. The above terms are relative concepts, and are described based on directions indicated in the drawings.

Unless defined otherwise, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Also, terms such as terms defined in commonly used dictionaries should be construed as having a meaning consistent with their meaning in the context of the relevant art, and unless they are interpreted in an ideal or overly formal sense, they are explicitly defined herein can be

Terms such as “comprise” or “have” are intended to designate that a feature, number, step, action, component, part, or combination thereof described in the specification is present, and includes one or more other features, number, or step. , it should be understood that it does not preclude the possibility of the existence or addition of an operation, a component, a part, or a combination thereof.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

1 illustrates a fire alarm system according to an embodiment of the present invention.

Referring to FIG. 1 , the fire alarm system 10 may detect a fire situation. The fire alarm system 10 may include a plurality of sensors SM, an image recording unit CT, a repeater 200 , a receiver 300 , and a first server 400 .

Each of the plurality of sensors SM may detect whether a fire has occurred. Also, the plurality of sensors SM may transmit state information (eg, information on current temperature, smoke, humidity, or gas of a place where the sensor is installed) even if a fire has not occurred. Although five sensors SM are illustrated in FIG. 1 by way of example, the present invention is not limited thereto. Each of the plurality of sensors SM may transmit the first fire information signal SG-1a including fire information to the adjacent sensors SM and/or the repeater 200 .

The first fire information signal SG-1a may be a signal generated by the sensor SM that detects whether a fire has occurred or a signal amplified by the sensor SM.

As a method of transmitting the first fire detection signal SG-1, a radio frequency (RF) communication method may be used. The RF communication method may be a communication method for exchanging information by radiating a radio frequency. It is a broadband communication method using a frequency, so it can have high stability because it is less affected by climate and environment. In the RF communication method, voice or other additional functions may be interlocked and the transmission speed may be high. For example, the RF communication method may use a frequency of 447 MHz to 924 MHz. However, this is exemplary and in an embodiment of the present invention, a communication method such as Ethernet, Wifi, LoRA, M2M, 3G, 4G, LTE, LTE-M, Bluetooth, or WiFi Direct may be used.

In an embodiment of the present invention, the RF communication method may include a Listen Before Transmission (LBT) communication method. This is a frequency selection method that determines whether the selected frequency is being used by another system, and selects another frequency when it is determined that it is occupied. For example, a node intending to transmit may first listen to the medium, determine if it is in an idle state, and then flush the backoff protocol prior to transmission. By distributing data using this LBT communication method, collisions between signals in the same band can be prevented.

The repeater 200 may perform RF communication with a plurality of sensors SM. For example, the repeater 200 may communicate with 40 sensors SM. The repeater 200 may receive the first fire information signal SG-1a from the plurality of sensors SM. The repeater 200 may convert the first fire information signal SG-1a into the second fire information signal SG-2a. The second fire information signal SG-2a may include the fire information. The repeater 200 may transmit the second fire information signal SG-2a to the receiver 300 .

The RF communication method may be used as a method of transmitting the second fire information signal SG-2a. That is, the repeater 200 and the receiver 300 may perform the RF communication.

The receiver 300 may receive the second fire information signal SG-2a from the repeater 200 . The receiver 300 may convert the second fire information signal SG-2a into the third fire information signal SG-3a. The third fire information signal SG-3a may include the fire information. The receiver 300 may transmit the third fire information signal SG-3a to the first server 400 .

The RF communication method may be used as a method of transmitting the third fire information signal SG-3a. That is, the receiver 300 and the first server 400 may perform the RF communication.

The first server 400 may determine the fire situation based on the third fire information signal SG-3a received from the receiver 300 .

The first server 400 may transmit the third response signal SG-3b to the receiver 300 upon receiving the third fire information signal SG-3a.

The receiver 300 may transmit the second response signal SG-2b to the repeater 200 upon receiving the second fire information signal SG-2a or the third response signal SG-3b.

The repeater 200 may transmit the first response signal SG-1b to the plurality of sensors SM upon receiving the first fire information signal SG-1a or the second response signal SG-2b. .

The first server 400 may receive big data BD from an external second server BS. The big data BD may be updated periodically. Big data (BD) is a means of predicting a diversified society, and may refer to data of a size that exceeds the ability of common software tools to collect, manage, and process within an acceptable hardening time. Such large amounts of data can provide more insight than traditional limited data.

The first server 400 may include an algorithm including an artificial intelligence model capable of calculating the degree of risk. This will be described later.

The first server 400 may calculate the risk based on the big data BD, the values detected by each of the plurality of sensors SM, and the image IM captured by the image recording unit CT. The risk level may be a value including whether there is a possibility and/or a risk of fire occurring in the building.

When the degree of risk exceeds a predetermined value, the first server 400 may transmit a warning message and location information to the plurality of parties 20 .

The plurality of persons 20 may include, for example, a fire station, persons concerned at a place where a fire occurred, a disaster prevention center (or a public institution related to public safety), and a monitoring screen disposed in a situation room of a building. there is. The plurality of parties 20 may receive the fire warning message in the form of a text message, a video message, or a voice message through a landline phone, a smart phone, or other portable terminal.

2 is a perspective view illustrating one sensor among a plurality of sensors according to an embodiment of the present invention.

1 and 2 , each of the plurality of sensors SM may include a fire sensor SS, a memory MM, an amplification unit AMP, a communication unit ATN, and a battery unit TT1. there is.

The fire sensor SS may detect at least one of smoke, temperature, humidity, and gas. The fire sensor SS may generate fire information by detecting at least one of smoke, temperature, humidity, and gas. The fire information may include a value measured by the fire sensor SS. In FIG. 2, one fire sensor SS is illustrated by way of example, but the present invention is not limited thereto. For example, each of the plurality of sensors SM may include a plurality of fire sensors, and each of the plurality of fire sensors may detect at least one of smoke, temperature, humidity, and gas.

The memory MM may store information on the fire sensor SS and unique address information of the sensor SM. The sensor SM may automatically determine a modulation method generated by the mounted fire sensor SS through information on the fire sensor SS. Through this automatic modulation method, the fire sensor SS simply generates fire information no matter what type of fire sensors are mounted on each of the plurality of sensors SM, and transmits the first fire information signal SG-1a. can do.

The memory MM may include a volatile memory or a non-volatile memory. The volatile memory may include DRAM, SRAM, flash memory, or FeRAM. The non-volatile memory may include an SSD or HDD.

The communication unit ATN may transmit the first fire information signal SG-1a to the repeater 200 . The communication unit ATN may also transmit the first fire information signal SG-1a to other adjacent sensors SM. The first fire information signal SG-1a may include fire information generated by the fire sensor SS and the address information. The communication unit ATN may periodically transmit the first fire information signal SG-1a to the repeater 200 .

When the sensor SM and the repeater 200 are far away from each other, and it is difficult for the repeater 200 to directly receive the first fire information signal SG-1a, the communication unit ATN transmits the first to another adjacent sensor SM. By transmitting the fire information signal SG-1a, signal transmission to the repeater 200 can be stably performed. The communication unit ATN may receive the first fire information signal SG-1a from another adjacent sensor SM.

The communication unit ATN may receive the first fire information signal SG-1a from another sensor SM. In the process of receiving the received first fire information signal SG-1a from another adjacent sensor SM, a transmission rate and/or accuracy may be reduced due to a transmission distance and noise. The amplifying unit AMP may amplify the degraded first fire information signal SG-1a. Transmission rate and/or accuracy of the amplified first fire information signal SG-1a may be improved. The communication unit ATN transmits the amplified first fire information signal SG- 1a) can be transmitted. The amplified first fire information signal SG-1a may increase the accuracy, transmission rate, and transmission distance of signals transmitted between the plurality of sensors SM and the repeater 200 .

The amplified first fire information signal SG-1a according to an embodiment of the present invention may be transmitted to another adjacent sensor SM and amplified again by the amplification unit AMP of the other adjacent sensor SM.

According to the present invention, the plurality of sensors SM may stably transmit data to the plurality of sensors SM and the repeater 200 using the amplification unit AMP. Accordingly, the reliability of the plurality of sensors SM may be improved.

The battery unit TT1 may supply power to the fire sensor SS, the memory MM, the amplification unit AMP, and the communication unit ATN.

The communication unit (ATN) according to an embodiment of the present invention may use an RF communication method. The RF communication method may consume less power. Power use of the sensor SM may be minimized, and the sensor SM may be driven with low power. Accordingly, the battery unit TT1 may stably supply power to the fire sensor SS, the memory MM, the amplification unit AMP, and the communication unit ATN for a long time.

In addition, according to the present invention, the plurality of sensors SM is divided into a power saving mode that does not consume power and a normal mode that operates in a fire situation to minimize the power use of each of the plurality of sensors SM there is. Accordingly, each of the plurality of sensors SM may be driven with low power.

3 is a perspective view showing a repeater according to an embodiment of the present invention.

1 and 3 , the repeater 200 may communicate with a plurality of sensors SM. For example, the repeater 200 may communicate with 40 sensors SM. The repeater 200 may include a communication unit ANT-G, a power supply unit PW-G, a battery unit BT-G, and a control unit CC-G.

The communication unit ATN-G may communicate with the plurality of sensors SM and the receiver 300 . The communication unit ATN-G may receive the first fire information signal SG-1a from each of the plurality of sensors SM. The communication unit ATN-G and the communication unit ATN (refer to FIG. 2 ) of the plurality of sensors SM may communicate wirelessly through an RF communication method. The communication unit ATN-G may transmit the second fire information signal SG-2a to the receiver 300 . The communication unit ATN-G and the communication unit ATN-R (refer to FIG. 4 ) of the receiver 300 may communicate wirelessly through an RF communication method.

The power supply unit PW-G may receive the first power from the outside. The first power may supply power to the communication unit ATN-G and the control unit CC-G.

The battery unit BT-G may supply the second power. The second power may supply power to the communication unit ATN-G and the control unit CC-G.

According to the present invention, even when the first power supplied from the power supply unit PW-G is cut off, the battery unit BT-G supplies the second power so that the repeater 200 can operate. The repeater 200 may stably receive the first fire information signal SG-1a from the plurality of sensors SM, and stably transmit the second fire information signal SG-2a to the receiver 300. there is. Accordingly, the reliability of signal transmission may be improved.

The controller CC-G may convert the first fire information signal SG-1a into the second fire information signal SG-2a. When the first power from the power supply unit PW-G is not supplied to the communication unit ATN-G, the control unit CC-G transmits the second power from the battery unit BT-G to the communication unit ATN-G. can be supplied to

4 shows a receiver according to an embodiment of the present invention.

1 and 4 , the receiver 300 may receive the second fire information signal SG-2a from the plurality of repeaters 200 . For example, the receiver 300 may communicate with 24 repeaters 200 . That is, the receiver 300 may communicate with 960 sensors SM. However, this is exemplary and the number of repeaters 200 through which the receiver 300 communicates according to an embodiment of the present invention is not limited thereto.

The receiver 300 includes a communication unit (ATN-R), a power supply unit (PW-R), a battery unit (BT-R), an intermediate memory (MM-R), a control unit (CT-R), and a display unit (DA-R) may include

The communication unit ATN-R may communicate with the repeater 200 and the first server 400 . The communication unit ATN-R may receive the second fire information signal SG-2a from the repeater 200 . The communication unit (ATN-R) and the communication unit (ATN-G, see FIG. 3 ) of the repeater 200 may communicate wirelessly through an RF communication method. The communication unit ATN-R may transmit the third fire information signal SG-3a to the first server 400 . The communication unit ATN-R and the communication unit 400 (refer to FIG. 5 ) of the first server 400 may communicate wirelessly through an RF communication method.

The power supply unit PW-R may receive the first power from the outside. The first power may supply power to the communication unit ATN-R, the intermediate memory MM-R, the control unit CT-R, and the display unit DA-R.

The battery unit BT-R may supply the second power. The second power may supply power to the communication unit ATN-R, the intermediate memory MM-R, the display unit DA-R, and the control unit CT-R.

According to the present invention, even when the first power supplied from the power supply unit PW-R is cut off, the battery unit BT-R supplies the second power to enable the receiver 300 to operate. The receiver 300 may stably receive the second fire information signal SG-2a from the repeater 200 and stably transmit the third fire information signal SG-3a to the first server 400 . . Accordingly, the reliability of signal transmission may be improved.

The display unit DA-R may provide image information corresponding to the state of the plurality of sensors SM or the state of the repeater 200 . The display unit DA-R may include a liquid crystal display panel or an organic light emitting display panel. The display unit DA-R may receive an external input provided by a user. For example, the display unit DA-R may further include a touch unit.

The controller CT-R may control each of the plurality of sensors SM. The user may provide an input to the display unit DA-R so that the control unit CT-R controls each of the plurality of sensors SM. For example, the control unit CT-R may include information on a place where each of the plurality of sensors SM is disposed, information against the type of a value sensed by each of the plurality of sensors SM, and/or a plurality of Information on whether each of the sensors SM operates normally may be controlled. The receiver 300 may control the plurality of sensors SM disposed in various places through the repeater 200 .

5 illustrates a first server according to an embodiment of the present invention.

1 and 5 , the first server 400 includes a building information modeling unit 410 , a big data receiving unit 420 , a server memory 430 , a server analyzing unit 440 , and a risk calculating unit 450 . , a communication unit 460 , a data extraction unit 470 , and a learning unit 480 .

The building information modeling unit 410 may implement a building virtually. The building may be a building in which a plurality of sensors SM and an image recording unit CT are disposed. The building information modeling unit 410 may provide modeling information.

The big data receiver 420 may receive the big data BD from the second server BS.

The big data BD may include information about a building. For example, big data (BD) includes information on large-scale complex buildings, information on apartment houses, information on commercial facilities, information on traditional markets, information on schools, information on village halls, information on military bases. , information on national parks, information on hospitals, information on apartments, and information on detached houses may be included.

The big data BD may include surrounding environment data for determining whether a fire has occurred. For example, the surrounding environment data includes data corresponding to the probability of occurrence of fire by date, data corresponding to the probability of occurrence of fire by time, data corresponding to the probability of occurrence of fire by space, data corresponding to the probability of occurrence of fire by temperature, and humidity. It may include at least one of data corresponding to the probability of occurrence of each fire, data corresponding to the probability of occurrence of fire by weather, data corresponding to the probability of occurrence of fire by industry, and data corresponding to the probability of occurrence of fire by user.

For example, the data corresponding to the fire occurrence probability for each date may include a fire occurrence probability for each day of the week and a fire occurrence probability for each month. The data corresponding to the fire occurrence probability for each time may include the fire occurrence probability divided into dawn, morning, afternoon, evening, or late night. The data corresponding to the fire occurrence probability for each space may include the fire occurrence probability divided into a city center, a mountainous area, a beach, or a rural area. The data corresponding to the fire occurrence probability for each temperature may include the fire occurrence probability divided into spring, summer, autumn, or winter. The data corresponding to the fire occurrence probability for each humidity may include a fire occurrence probability for each specific humidity value. The data corresponding to the fire occurrence probability for each weather may include a fire occurrence probability divided into a sunny day, a cloudy day, or a rainy day. The data corresponding to the fire occurrence probability for each industry may include a fire occurrence probability divided into homes, restaurants, factories, or offices. The fire occurrence probability for each user may include a fire occurrence probability divided by age, occupation, or gender.

The server memory 430 may store information (eg, contact information, address, or name) of the parties 20 . The server memory 430 may store data for calculating the fire risk. For example, the server memory 430 may collect and store fire information received from a plurality of sensors SM. This will be described later.

The server memory 430 may include volatile memory or non-volatile memory. Volatile memory may include DRAM, SRAM, flash memory, or FeRAM. Non-volatile memory may include SSD or HDD.

The server analysis unit 440 may generate analysis data based on the modeling information implemented by the building information modeling unit 410 , the big data BD received by the big data receiving unit 420 , and the fire information.

The risk calculator 450 may calculate the risk based on the analysis data.

The communication unit 460 may transmit a warning message to the plurality of parties 20 based on the degree of risk when there is a possibility and/or a risk of fire occurring in the building.

The data extraction unit 470 may extract simulation data by reinforcement learning the analysis data. This will be described later.

The learning unit 480 may provide a function necessary for reinforcement learning of the analysis data.

6 illustrates an operation of a server analyzer according to an embodiment of the present invention.

1 and 6 , the building information modeling unit 410 may generate modeling information MD by virtually realizing a building in which a plurality of sensors SM are installed. The modeling information MD may include information about electric wiring arranged inside and outside the building as well as information about gas piping and the like.

The plurality of sensors SM may generate fire information FI. The plurality of sensors SM may transmit the fire information FI to the first server 400 using the first to third fire information signals SG-1a, SG-2a, and SG-3a.

The big data receiver 420 may receive the big data BD from the second server BS. The big data BD may include information about a building. The information on the building may include information on the risk of fire for each building. For example, Big Data (BD) provides information that large-scale complexes are more likely to have a large number of people during the day, so fires are more likely to occur during the day than at night. Information on the degree of risk, information on the causes of fires that can occur in large complexes, information on firefighting facilities installed in accordance with the fire-fighting laws of large complexes, information on the characteristics of fires occurring in large complexes, Information that apartment houses are highly likely to cause fires due to electricity and gas, information on firefighting facilities installed in accordance with the fire-fighting laws for apartment houses, information on the characteristics of fires occurring in apartment houses, information on firefighting facilities in traditional markets Information on whether it is old or defective, information that a fire is likely to occur due to electrical factors in traditional markets, information that the damaged space may spread if a fire occurs because the space is open in traditional markets, and fires in traditional markets Information on the characteristics of the market, information on industries containing inflammable materials in traditional markets, information on the high probability of fires in classrooms and lunchrooms in schools, information on the frequency of fires by day of the week in schools, and lightning-induced fires in schools Information about which fires are most likely to occur, information about where firefighting systems are deployed in national parks, information about fire management offices in national parks, information that national parks are most likely to be ignited by wildfires, and when wildfires occur frequently. information about the national park, information about the most likely occurrence of fire by visitors to the national park, information on the time when the most visitors to the national park visit the national park, and information on firefighting facilities installed in accordance with the fire-fighting laws of hospitals. can

The server memory 430 may receive and store modeling information MD, fire information FI, and big data BD. The server memory 430 may provide modeling information MD, fire information FI, and big data BD to the server analysis unit 440 .

The server analyzer 440 may generate the analysis data DATA1 based on the modeling information MD, the fire information FI, and the big data BD.

For example, when the server analysis unit 440 determines that the building in which the plurality of sensors SM is installed is a school from the modeling information MD, the server analysis unit 440 receives the information about the school from the big data receiving unit 420 . information can be received. The information about the school may include information that a fire caused by gas is highly likely to occur in the school cafeteria, and the server analysis unit 440 may receive the fire information FI from the plurality of sensors SM. there is. When the fire information FI includes information that gas is detected in the school cafeteria, the server analysis unit 440 generates analysis data DATA1 including information that a fire is very likely to occur in the current cafeteria. can

For example, when the server analysis unit 440 determines from the modeling information MD that the building in which the plurality of sensors SM are installed is a large complex, the server analysis unit 440 receives the large data from the big data receiving unit 420 . You can receive information about complex buildings. The information on the large-scale complex may include information that the large-scale complex is more likely to have a large number of people during the day, and thus a fire is more likely to occur during the day than at night. The server analyzer 440 may receive the fire information FI from the plurality of sensors SM. When the fire information FI includes information that smoke has occurred in a crowded area, the server analysis unit 440 analyzes data DATA1 including information that a fire is likely to have occurred in a currently crowded area. can create

For example, when the server analysis unit 440 determines from the modeling information MD that the building in which the plurality of sensors SM are installed is an apartment house, the server analysis unit 440 receives the big data receiving unit 420 from the apartment house. information can be received. The information on the apartment house may include information indicating that the apartment house has a high probability of generating a fire due to electricity and gas. The server analysis unit 440 may extract the area in which electricity is used the most from the information on the electrical wiring of the modeling information MD, and the server analysis unit 440 receives information that a fire is likely to occur in the area. It is possible to generate the analysis data DATA1 that includes it.

According to the present invention, the analysis data DATA1 combines modeling information MD, fire information FI, and big data BD to determine the risk of fire depending on the type of building in which the plurality of sensors SM are installed. may include an optimal value for determining . Accordingly, it is possible to provide the fire alarm system 10 capable of deriving a risk optimization value.

7 illustrates an operation of the risk calculator according to an embodiment of the present invention, and FIG. 8 illustrates reinforcement learning according to an embodiment of the present invention.

1, 7, and 8 , the risk calculator 450 may calculate the risk RK based on the analysis data DATA1 . The risk calculator 450 may calculate a different risk RK according to the type of building. The risk calculator 450 may transmit the risk RK to the plurality of parties 20 . For example, the risk calculator 450 may transmit the risk RK to the monitoring screen, and the monitoring screen may visually display the risk RK and provide it to the user. That is, the risk calculator 450 may display the simulation processed based on the modeling information MD on the monitoring screen, and the risk calculator 450 displays the risk RK in the simulation to provide the user with can

The risk level RK may be information including a possibility that a fire may occur.

For example, if a plurality of sensors SM are installed in a school, and the analysis data DATA1 includes information that a fire is highly likely to occur in the current cafeteria, the risk calculation unit 450 provides modeling information ( MD), it is possible to inform the user of the possibility of a fire by displaying the risk (RK) that a fire may occur in the dining room.

For example, when the plurality of sensors SM is installed in a large complex and the analysis data DATA1 includes information that a fire is likely to have occurred in a currently crowded area, the risk calculation unit 450 ) may indicate a fire occurrence to the user by displaying the risk level RK that a fire has occurred in an area with many people shown in the modeling information MD. In addition, the communication unit 460 may transmit a warning message and a place where a fire occurred in the large complex to the plurality of parties 20 .

For example, when the plurality of sensors SM is installed in an apartment house and the analysis data DATA1 includes information that a fire is highly likely to occur in an area where electricity is used the most, the risk calculation unit 450 ) may indicate a risk level RK that a fire may occur in the area of the apartment house shown in the modeling information MD to inform the user of the possibility of a fire.

Contrary to the present invention, when an unexpected fire occurs, fire related personnel may be embarrassed and may not be able to take prompt action against the fire. In case of a fire, if the golden time exceeds 5 minutes, the casualties may increase by about two times or more, and the amount of property damage may increase by about three times or more. However, according to the present invention, the risk calculator 450 may calculate the risk RK based on the analysis data DATA1 in real time. A user can predict the possibility of a fire based on the real-time calculated risk level (RK) and easily respond when a fire occurs. In other words, it is possible to prevent a fire by taking the first action measures for an area with a high probability of fire from the risk level (RK) before a fire occurs, and the fact that a fire could occur based on the risk level (RK) when a fire occurs It was possible to extinguish the fire within the above golden time because it was aware of the Accordingly, it is possible to prevent the increase in casualties and the spread of property damage. Accordingly, it is possible to provide a method of deriving a fire risk optimization value with improved reliability and a fire alarm system 10 using the method.

The server analysis unit 440 may transmit the analysis data DATA1 to the server memory 430 . The server memory 430 may store the analysis data DATA1 . The server analysis unit 440 may generate new analysis data based on the analysis data DATA1 , the modeling information MD, the fire information FI, and the big data BD stored in the server memory 430 .

The server analysis unit 440 may transmit the analysis data DATA1 to the data extraction unit 470 . The data extraction unit 470 may extract the simulation data DATA2 by reinforcement learning the analysis data DATA1 .

The learning model may be generated by the learning unit 480 and provided to the data extracting unit 470 .

The learning unit 480 learns the learning data generated in advance, extracts the simulation data DATA2 from the analysis data DATA1 , generates a learning model for calculating the probability of occurrence of a fire, and provides it to the data extraction unit 470 . can do.

The learning unit 480 may continuously perform reinforcement learning on the generated learning model based on the simulation data DATA2 provided from the data extraction unit 470 to perform a function of upgrading the corresponding learning model. .

Reinforcement learning may be a learning method applied to the trial and error concept.

The data extracting unit 470 observes the state from the learning unit 480 and if an appropriate action is taken accordingly, the data extraction unit 470 receives a reward from the learning unit 480 based on the action. Observation is transformed into a state, and the data extraction unit 470 repeats this series of “observation-action-reward” interactions and must perform a task of maximizing the reward obtained from the learning unit 480, performing the above operation Reinforcement learning is a series of processes to do this.

The process of "observation-action-reward" can be said to be an experience.

Reinforcement learning may include a decision-making process based on Markov properties. In other words, when there is a series of temporal events, the expected state of the next stage from the state of the present stage is independent of the events in the past.

Here, the Markov reward process is to find the state value function as the value of the present state, which is the sum of the reward and the average value of the present state.

From this, it is proven that an optimal policy function exists, and the optimal policy can be obtained immediately by knowing the optimal state value. In the present invention, the optimal policy may be to extract the simulation data DATA2 from the analysis data DATA1 to calculate the probability of occurrence of a fire.

That is, the learning unit 480 may extract the simulation data DATA2 by transferring the learning result to the data extraction unit 470 using reinforcement learning. By continuously updating the learning model using the simulation data DATA2 , the learning unit 480 may perform reinforcement learning on the corresponding learning model to gradually upgrade the learning model. Through this, the fire alarm system 10 may extract different simulation data DATA2 for each building and calculate a highly reliable fire occurrence probability.

The data extractor 470 may transmit the simulation data DATA2 to the server memory 430 . The server memory 430 may store simulation data DATA2 . The data extraction unit 470 may extract the simulation data DATA2 stored in the server memory 430 and new simulation data extracted through reinforcement learning. Through the new simulation data, it is possible to calculate the probability of occurrence of a highly reliable fire.

The risk calculator 450 may calculate the risk RK in real time based on the simulation data DATA2 . A user can predict the possibility of a fire based on the real-time calculated risk level (RK) and easily respond when a fire occurs.

The first server 400 may receive the information IN through the variable value adjustment dialog DX from the outside. The information IN may include information additionally obtained by the user about the possibility that a fire may occur in addition to the plurality of sensors SM and the big data BD. The information IN may include weights required in real time for the site of the building. The user can input information IN in real time through the variable value adjustment dialog DX.

The risk calculator 450 may calculate the risk RK in real time based on the information IN, the analysis data DATA1 , and the simulation data DATA2 .

According to the present invention, the risk RK is determined by combining the information IN, the analysis data DATA1, and the simulation data DATA2 to determine the possibility of fire according to the type of the building in which the plurality of sensors SM are installed. It may include an optimal value for Accordingly, it is possible to provide the fire alarm system 10 capable of deriving a risk optimization value.

Also, according to the present invention, the risk calculator 450 may calculate the risk RK in real time based on the information IN, the analysis data DATA1 , and the simulation data DATA2 . A user can predict the possibility of a fire based on the real-time calculated risk level (RK) and easily respond when a fire occurs. In other words, it is possible to prevent a fire by taking the first action measures for an area with a high probability of fire from the risk level (RK) before a fire occurs, and the fact that a fire could occur based on the risk level (RK) when a fire occurs Because they were aware of this, they could extinguish the fire within the golden time. Accordingly, it is possible to prevent the increase in casualties and the spread of property damage. Accordingly, it is possible to provide a method of deriving a fire risk optimization value with improved reliability and a fire alarm system 10 using the method.

9 shows a sensor and a first server according to an embodiment of the present invention, and FIG. 10 shows buildings and a first server according to an embodiment of the present invention.

1, 9, and 10 , the first server 400 may generate an upgrade signal UP based on the analysis data DATA1 . The communication unit 460 may transmit the upgrade signal UP to the plurality of sensors SM. The upgrade signal UP may be transmitted while being included in the first to third response signals SG-1b, SG-2b, and SG-3b.

Each of the plurality of sensors SM may generate fire information FI-1 based on the upgrade signal UP. The fire information FI-1 may be generated differently for each building based on the upgrade signal UP.

The plurality of sensors SM may transmit the fire information FI to the first server 400 . The server analysis unit 440 may generate the analysis data DATA1 based on the fire information FI. The analysis data DATA1 may be stored in the server memory 430 . The first server 400 may generate an upgrade signal UP based on the analysis data DATA1 and modeling information MD (refer to FIG. 6 ). The plurality of sensors SM may receive the upgrade signal UP. The plurality of sensors SM may generate fire information FI-1 based on the upgrade signal UP. The fire information FI-1 may be different from the fire information FI previously transmitted by the plurality of sensors SM. The first server 400 may update the analysis data DATA1 based on the fire information FI-1. The risk calculator 450 may calculate the risk RK (refer to FIG. 7 ) in real time based on the updated analysis data DATA1 . Accordingly, the reliability of the risk level RK (refer to FIG. 7 ) may be improved.

For example, the big data BD may include information on the first building A1 , information on the second building A2 , and information on the third building A3 . The server analysis unit 440 (refer to FIG. 7 ) may determine the type of building by comparing the modeling information (MD, see FIG. 6 ) and big data.

The server analysis unit 440 (refer to FIG. 7 ) may determine the first building A1 as a large-scale complex based on the modeling information (MD, see FIG. 6 ) and the information on the first building A1 . The server analyzer 440 (refer to FIG. 7 ) may determine the second building A2 as a school based on the modeling information (MD, see FIG. 6 ) and the information on the second building A2 . In addition, the server analysis unit 440 (refer to FIG. 7 ) may determine the third building A3 as a hospital based on the modeling information (MD, see FIG. 6 ) and information on the third building A3 .

The first server 400 may generate different upgrade signals UP1 , UP2 , and UP3 for each building based on the analysis data DATA1 and the modeling information MD (refer to FIG. 6 ). The upgrade signals UP1 , UP2 , and UP3 may include a first upgrade signal UP1 , a second upgrade signal UP2 , and a third upgrade signal UP3 .

The first upgrade signal UP1 may include information optimized for the first building A1 . For example, the first upgrade signal UP1 may set the sensing sensitivities of the plurality of sensors SM installed in the first building A1 to the first sensitivity. The plurality of sensors SM installed in the first building A1 may generate the first fire information FI1 optimized for the first building A1 based on the first upgrade signal UP1 . The first server 400 may update the analysis data DATA1 based on the first fire information FI1, and the risk calculator 450 may update the second building A2 based on the updated analysis data DATA1. ) and different from the third building A3 and optimized for the first building A1 (RK, see FIG. 7 ) may be calculated.

The second upgrade signal UP2 may include information optimized for the second building A2 . For example, the second upgrade signal UP2 may set the sensing sensitivity of the plurality of sensors SM installed in the second building A2 to a second sensitivity different from the first sensitivity. The plurality of sensors SM installed in the second building A2 may generate the second fire information FI2 optimized for the second building A2 based on the second upgrade signal UP2 . The first server 400 may update the analysis data DATA1 based on the second fire information FI2, and the risk calculator 450 may update the first building A1 based on the updated analysis data DATA1. ) and the third building (A3) and different and optimized for the second building (A2) (RK, see FIG. 7 ) may be calculated.

The third upgrade signal UP3 may include information optimized for the third building A3 . For example, the third upgrade signal UP3 may set the sensing sensitivities of the plurality of sensors SM installed in the third building A3 to a third sensitivity different from the first sensitivity and the second sensitivity. The plurality of sensors SM installed in the third building A3 may generate the third fire information FI3 optimized for the third building A3 based on the third upgrade signal UP3 . The first server 400 may update the analysis data DATA1 based on the third fire information FI3, and the risk calculator 450 may update the first building A1 based on the updated analysis data DATA1. ) and different from the second building A2 and optimized for the third building A3 (RK, see FIG. 7 ) may be calculated. The user can predict the possibility of a fire based on the real-time calculated risk (RK, see FIG. 7), and can easily respond when a fire occurs. Accordingly, it is possible to provide a method of deriving a fire risk optimization value with improved reliability and a fire alarm system 10 using the method.

11 shows a fire escape route according to an embodiment of the present invention.

1, 6, 7, and 11 , for example, FIG. 11 may show a specific floor of a large-scale complex. A plurality of sensors SM and an image recording unit CT may be installed on the specific layer. Although not shown in FIG. 11 , the large complex according to an embodiment of the present invention may include a repeater 200 , a receiver 300 , and a first server 400 . The fire alarm system 10 may show a fire evacuation route on the monitoring screen. A user may predict a place (AR) where a fire may occur through the monitoring screen, and may be provided with a fire escape route (RT1).

The plurality of sensors SM may transmit the fire information FI to the first server 400 .

The server analyzer 440 may generate the analysis data DATA1 based on the modeling information MD, the fire information FI, and the big data BD.

The data extraction unit 470 may extract the simulation data DATA2 by reinforcement learning the analysis data DATA1 .

The risk calculator 450 may calculate the risk RK in real time based on the analysis data DATA1 and the simulation data DATA2 . A user can predict the possibility of a fire based on the real-time calculated risk level (RK) and easily respond when a fire occurs.

For example, based on the risk RK calculated by the risk calculator 450 , the first server 400 may determine that a fire is highly likely to occur in the area AR of the specific floor. The first server 400 may generate the fire escape route RT1 based on the risk level RK, the analysis data DATA1 , and the simulation data DATA2 . The fire escape route RT1 may include a route through which an evacuator can safely evacuate through an area where a fire is least likely to occur in consideration of the area AR. For example, the big data (BD) may include information on the risk of fire occurrence according to the material of the exterior material constituting the large-scale complex, and the first server 400 is a safe fire based on the big data (BD). An escape route RT1 can be created. The fire escape route RT1 may be provided differently depending on the type of the building.

According to the present invention, the first server 400 may extract the simulation data DATA2 through reinforcement learning for fire. The first server 400 predicts the fire situation of a specific building in advance based on modeling information (MD), big data (BD), fire information (FI), analysis data (DATA1), and simulation data (DATA2), , based on this, the fire escape route (RT1) is provided quickly so that evacuees can evacuate from the fire area within the golden time. Accordingly, it is possible to prevent the increase in casualties and the spread of property damage. Accordingly, it is possible to provide a method of deriving a fire risk optimization value with improved reliability and a fire alarm system 10 using the method.

Although the above has been described with reference to the preferred embodiment of the present invention, those skilled in the art or those having ordinary knowledge in the technical field will not depart from the spirit and technical scope of the present invention described in the claims to be described later. It will be understood that various modifications and variations of the present invention can be made without departing from the scope of the present invention. Accordingly, the technical scope of the present invention should not be limited to the contents described in the detailed description of the specification, but should be defined by the claims.

10: fire alarm system SM: multiple sensors
200: repeater 300: receiver
400: first server

Claims (13)

a plurality of sensors, each of which generates fire information by detecting the occurrence of a fire, and performs RF communication (Radio Frequency communication) with each other;
a repeater performing the RF communication with the plurality of sensors;
a receiver performing the RF communication with the repeater; and
A first server for performing the RF communication with the receiver,
The first server,
a building information modeling unit that provides modeling information by virtualizing a building;
a big data receiving unit for receiving big data from an external second server;
a server memory for collecting and storing the fire information received from the plurality of sensors;
a server analysis unit generating analysis data based on the modeling information, the big data, and the fire information; and
And a risk calculator for calculating the risk based on the analysis data,
The risk calculation unit calculates a different level of risk according to the type of the building,
The risk level calculation unit is a fire alarm system for calculating the level of risk in real time. delete According to claim 1,
The first server transmits an upgrade signal generated based on the analysis data to the plurality of sensors,
Each of the plurality of sensors is a fire alarm system for generating the fire information based on the upgrade signal. 4. The method of claim 3,
Each of the plurality of sensors is a fire alarm system for generating different fire information for each building based on the upgrade signal. According to claim 1,
The fire alarm system further comprising a data extraction unit for extracting simulation data by reinforcement learning the analysis data. 6. The method of claim 5,
The first server receives information from the outside through the variable value adjustment dialog,
The risk calculator calculates the risk in real time based on the simulation data and the information. 6. The method of claim 5,
The first server is a fire alarm system for calculating a fire escape route based on the simulation data. 8. The method of claim 7,
The fire escape route is different according to the type of the building fire alarm system. 6. The method of claim 5,
The big data includes information about a first building and information about a second building that is different from the first building,
The server analyzer compares the modeling information with the information on the first building and the information on the second building to determine whether the building is the first building or the second building. 10. The method of claim 9,
When it is determined that the building is the first building, the first server transmits a first upgrade signal for setting the sensing sensitivities of the plurality of sensors to the first sensitivity to each of the plurality of sensors, and the building is the first building. 2 When it is determined that the building is a building, a fire alarm system for transmitting a second upgrade signal for setting the sensing sensitivity to a second sensitivity different from the first sensitivity to each of the plurality of sensors. generating fire information by a plurality of sensors;
providing modeling information by implementing a building virtually;
receiving and storing big data from an external server;
generating analysis data based on the fire information, the modeling information, and the big data; and
Comprising the step of calculating the risk of fire in real time based on the analysis data,
Calculating the risk level is a fire risk optimization value deriving method comprising the step of calculating in real time a different level of risk according to the type of the building. 12. The method of claim 11,
extracting simulation data by reinforcement learning the analysis data;
Receiving information from the outside through a variable value adjustment dialog; and
The method of deriving a fire risk optimization value further comprising calculating the degree of risk based on the simulation data, the information, and the analysis data. 12. The method of claim 11,
generating an upgrade signal based on the analysis data;
transmitting an upgrade signal to the plurality of sensors; and
Fire risk optimization value deriving method further comprising the step of generating different fire information for each building based on the upgrade signal.

Images