Artificial intelligence-based fire detection system and the method of thereof

Abstract

An artificial intelligence-based fire detection system and a method thereof are disclosed. An artificial intelligence-based fire detection system according to one embodiment of the present invention comprises: a sensor unit that collects fire factor data by providing at least one gas sensor and a temperature sensor; an environmental information collection unit that collects location information and time information by including at least one GPS module and receives atmospheric information including temperature, wind speed, fine dust concentration, and humidity from a preset weather observation server; a data learning unit that calculates fire occurrence condition data by applying a preset artificial intelligence learning algorithm to the data collected from the sensor unit and the environmental information collection unit; a fire detection unit that determines whether a fire has occurred by comparing the fire factor data collected through the sensor unit with the fire occurrence condition data calculated by the data learning unit; and a fire suppression unit that controls the execution of a fire suppression manual through a fire suppression module provided in a preset area when it is determined that a fire has occurred by the fire detection unit.

Description

Technology Field

[0001] The present invention relates to a fire detection system and a method thereof, and more specifically, to an artificial intelligence-based fire detection system and a method thereof capable of more precise fire detection and response to fire occurrence based on gas concentration and temperature information collected through a sensor and fire occurrence conditions calculated through an artificial intelligence learning algorithm. Background Technology

[0002] Generally, each building, including houses and apartments, is equipped with fire detection facilities. In particular, fire detection and fire prevention systems are critical in places with a high risk of fire, such as petrochemical plants, gas and oil refineries, ships, and thermal power plants, where there are many flammable hazardous materials on site.

[0003] To this end, various fire detection systems have been proposed and are in use. Among them, Korean Registered Patent Publication No. 10-2014-0029575 discloses a fire detection system for a ship that activates a fire suppression unit upon detecting a fire by utilizing a sensor that detects smoke and flames and a video camera. However, sensors and video cameras have the disadvantage that detection is only possible after the fire has spread to a certain extent from the location where it has already occurred. Additionally, they react sensitively to areas close to the sensor but fail to properly sense areas at a distance, making precise detection impossible.

[0004] Furthermore, Korean Registered Patent Publication No. 10-0844996 discloses a fire detection system and method that detects a fire by calculating the temperature through the detection of the frequency of the current passing through a wire; however, this method is ineffective in places such as factories where there are significant temperature differences depending on the location within the same space due to machinery or working conditions, as the frequency of the current can change due to the temperature, and it has the disadvantage of causing significant difficulties in system construction.

[0005] Therefore, there is a need for research on systems and methods capable of accurate and rapid fire detection, taking into account factors such as the distance from the sensor and the installation location. The problem to be solved

[0006] The present invention aims to provide an artificial intelligence-based fire detection system and method capable of more accurate fire detection and rapid response according to the characteristics of a location requiring fire detection, by considering gas concentration and temperature data collected through sensors, real-time atmospheric information, location information, and time information, and classifying and calculating fire occurrence conditions by time of day and season.

[0007] In addition, the purpose is to provide an artificial intelligence-based fire detection system and method capable of preventing safety accidents that could lead to a fire before an actual fire occurs by generating an early warning based on the gas concentration collected through the sensor and the duration of the gas concentration.

[0008] The problems that the present invention aims to solve are not limited to those mentioned above, and other problems that the present invention aims to solve that are not mentioned herein will be clearly understood by those skilled in the art to which the present invention belongs from the description below. means of solving the problem

[0009] An artificial intelligence-based fire detection system according to an embodiment of the present invention comprises: a sensor unit that collects fire factor data by providing at least one gas sensor and a temperature sensor; an environmental information collection unit that collects location information and time information by including at least one GPS module and receives atmospheric information including temperature, wind speed, fine dust concentration, and humidity from a preset weather observation server; a data learning unit that calculates fire occurrence condition data by applying a preset artificial intelligence learning algorithm to the data collected from the sensor unit and the environmental information collection unit; a fire detection unit that determines whether a fire has occurred by comparing the fire factor data collected through the sensor unit with the fire occurrence condition data calculated by the data learning unit; and a fire suppression unit that controls the execution of a fire suppression manual through a fire suppression module provided in a preset area when it is determined that a fire has occurred by the fire detection unit, wherein the fire factor data includes real-time gas concentration data including at least one of CO2 and CO collected through the gas sensor, duration data of the real-time gas concentration data, real-time temperature data and duration data of the real-time temperature data collected through the temperature sensor.

[0010] In addition, the data learning unit is characterized by calculating seasonal fire occurrence condition data based on atmospheric information and location information collected from the environmental information collection unit, and calculating hourly fire occurrence condition data based on time information collected from the environmental information collection unit and real-time gas concentration data and real-time temperature data collected from the sensor unit.

[0011] Additionally, the fire detection unit is characterized by including a first detection unit that detects whether a fire has occurred by determining whether fire factor data collected through a sensor unit satisfies fire occurrence condition data calculated by a data learning unit, wherein if real-time gas concentration data collected through the sensor unit exceeds a preset first standard concentration and the duration of the real-time gas concentration data exceeds a preset first standard time, the first detection unit generates a fire occurrence alarm signal and transmits an alarm signal to a preset administrator terminal; and a second detection unit that generates a fire occurrence alarm signal and transmits an alarm signal to a administrator terminal and generates a control signal to control a fire suppression module when, within a preset second standard time after the fire occurrence alarm signal is generated, real-time temperature data collected through the sensor unit exceeds a preset first standard temperature, the duration of the real-time temperature data exceeds a preset third standard time, and satisfies fire occurrence condition data calculated by the data learning unit.

[0012] Additionally, the fire suppression module includes at least one throwable fire extinguisher coupled to a housing in which a sensor unit is installed, and the fire suppression unit is characterized by controlling the fire suppression module so that, when it is determined through a fire detection unit that a fire has occurred, the throwable fire extinguisher is separated from the housing and thrown into the fire area based on a fire suppression manual.

[0013] In addition, it is characterized by further including a sensor monitoring unit that detects whether the sensor unit is faulty by calculating at least one of the average value, maximum value, minimum value, and rate of change of fire factor data collected from the sensor unit for a preset period of time. Effects of the invention

[0014] According to the present invention, by considering gas concentration and temperature data collected through sensors, along with real-time atmospheric information, location information, and time information, fire occurrence conditions are classified and calculated by time period and season, thereby enabling more accurate fire detection and rapid response according to the characteristics of the location where fire detection is required.

[0015] In addition, by generating an early warning based on the gas concentration collected through the sensor and the duration of the gas concentration, it has the effect of preventing safety accidents that could lead to a fire before an actual fire occurs. Brief explanation of the drawing

[0016] FIG. 1 is a configuration diagram of an artificial intelligence-based fire detection system according to an embodiment of the present invention. FIG. 2 is a drawing for explaining the fire detection unit of an artificial intelligence-based fire detection system according to an embodiment of the present invention. FIG. 3 is a drawing for explaining a fire suppression module of an artificial intelligence-based fire detection system according to an embodiment of the present invention. Specific details for implementing the invention

[0017] Specific details regarding the problem to be solved, the means for solving the problem, and the effects of the invention as described above are included in the embodiments and drawings to be described below. The advantages and features of the present invention, and the methods for achieving them, will become clear by referring to the embodiments described below in detail together with the accompanying drawings.

[0018] The scope of the present invention is not limited to the embodiments described below, and various modifications can be made by those skilled in the art within the scope of the technical essence of the present invention.

[0020] Hereinafter, the name of the invention, which is the present invention, will be explained in detail with reference to the attached FIG. 1.

[0021] FIG. 1 is a configuration diagram of an artificial intelligence-based fire detection system according to an embodiment of the present invention, FIG. 2 is a diagram for explaining a fire detection unit of an artificial intelligence-based fire detection system according to an embodiment of the present invention, and FIG. 3 is a diagram for explaining a fire suppression module of an artificial intelligence-based fire detection system according to an embodiment of the present invention.

[0023] <Example 1>

[0024] Referring to FIG. 1, the artificial intelligence-based fire detection system (100) of the present invention may include a sensor unit (110), an environmental information collection unit (120), a data learning unit (130), a fire detection unit (140), and a fire suppression unit (150).

[0026] More specifically, the sensor unit (110) is provided with at least one gas sensor and a temperature sensor to collect fire factor data.

[0027] At this time, the fire factor data may include real-time gas concentration data including at least one of CO2 and CO collected through the gas sensor, the duration of the real-time gas concentration data, real-time temperature data through the temperature sensor, and the duration of the real-time temperature data.

[0028] For example, the duration of the real-time gas concentration data can be calculated as the time during which the real-time concentration data measured through the sensor unit persists within a preset error range of ±10%, and the duration of the real-time temperature data can also be calculated in the same way as the method used to calculate the duration of the real-time gas concentration data.

[0030] The above environmental information collection unit (120) includes at least one GPS module to collect location information and time information, and can receive atmospheric information including temperature, wind speed, fine dust concentration and humidity from a preset weather observation server.

[0031] That is, the environment information collection unit (120) can collect data from the outside through a network server by including at least one communication module.

[0033] Meanwhile, the data learning unit (130) can learn by applying a preset artificial intelligence algorithm to the data collected from the sensor unit (110) and the environment information collection unit (120) and produce fire occurrence condition data.

[0034] More specifically, the data learning unit (130) can calculate seasonal fire occurrence condition data based on the atmospheric information and location information collected from the environmental information collection unit (120), and calculate hourly fire occurrence condition data based on time information collected from the environmental information collection unit (120) and the real-time gas concentration data and real-time temperature data collected from the sensor unit (110).

[0035] At this time, the initial values ​​of the above-mentioned seasonal fire occurrence condition data and the above-mentioned hourly fire occurrence condition data can be set in advance through a designated administrator terminal, and the seasonal fire occurrence condition data can be optimized by extracting and learning seasonal fire occurrence condition parameters (ambient temperature, humidity, wind speed, fine dust concentration, CO2 concentration, CO2 concentration, O2 concentration, etc.) from history data generated during a fire during a pre-set period (e.g., 1 year).

[0036] That is, the reference value of the parameter for determining fire occurrence during the relatively dry winter season and the reference value of the parameter for determining fire occurrence during the summer season, which has higher humidity compared to the winter season, can be set differently from each other, and accordingly, the fire detection unit (140) can determine whether a fire has occurred by comparing the data collected through the sensor unit (110) with the fire occurrence condition data calculated by the data learning unit (130).

[0038] In addition, the above-mentioned hourly fire occurrence condition data is calculated by learning the time information and the fire factor data through the data learning unit (130). For example, if the artificial intelligence-based fire detection system (100) is installed in a house and there is a large change in real-time gas concentration data and real-time temperature data due to cooking in the morning or evening, the parameter reference values ​​for the morning and evening time periods defined through a preset administrator terminal and the parameter reference values ​​for the daytime period may be set differently from each other.

[0040] More specifically, as illustrated in FIG. 2, the fire detection unit (140) detects whether a fire has occurred by determining whether the fire factor data collected through the sensor unit (110) satisfies the fire occurrence condition data calculated by the data learning unit (130); wherein, if the real-time gas concentration data collected through the sensor unit (110) exceeds a preset first reference concentration and the duration of the real-time gas concentration data exceeds a preset first reference time, the first detection unit (141) generates a fire early warning signal and transmits an alarm signal to a preset administrator terminal; and if, within a preset second reference time after the fire early warning signal is generated, the real-time temperature data collected through the sensor unit (110) exceeds a preset first reference temperature and the duration of the real-time temperature data exceeds a preset third reference time and satisfies the fire occurrence condition data calculated by the data learning unit (140), the first detection unit (141) generates a fire occurrence alarm signal and transmits an alarm signal to the administrator terminal and generates a control signal to control the fire suppression module. It may include a second sensing unit (142).

[0041] That is, the first detection unit (141) primarily determines the possibility of a fire based on real-time gas concentration data collected through the gas sensor of the sensor unit (110), and if the concentration of CO2 and CO0 is abnormally high, determines that a safety accident problem such as a gas leak has occurred and can notify the administrator terminal of the possibility of a safety accident.

[0042] In addition, the second detection unit (142) can more accurately determine whether a fire has occurred by comparing real-time temperature data collected through the temperature sensor of the sensor unit (110) and the duration of the real-time temperature data with fire occurrence condition data calculated by the data learning unit (130) after the fire early warning signal is generated, and if it is determined that a fire has occurred, it can transmit fire occurrence information to the administrator terminal.

[0043] Additionally, the second detection unit (142) generates a control signal to control the fire suppression module for rapid fire suppression and transmits it to the fire suppression unit (150). When the fire suppression unit (150) determines that a fire has occurred in the fire detection unit (140), it can control the fire suppression manual to be executed through the fire suppression module provided in a preset area.

[0044] That is, the fire detection unit (140) can perform the fire suppression manual based on the control signal of the fire suppression module generated by the second detection unit (142).

[0045] More specifically, as illustrated in FIG. 3 above, the fire suppression module (151) may include at least one throwable fire extinguisher (152) coupled to a housing (111) in which the sensor unit (110) is mounted.

[0046] Accordingly, the fire suppression unit (150) can control the fire suppression module (151) so that when it is determined through the fire detection unit (140) that a fire has occurred, the throwable fire extinguisher (152) is separated from the housing and thrown into the fire area based on the fire suppression manual.

[0047] For example, the housing (111) and the fire suppression module (151) may be connected via at least one spring pin, and when a fire occurs, the spring pin connecting the housing (111) and the fire suppression module (151) is released so that the throwable fire extinguisher (152) included in the fire suppression module (151) can fall vertically.

[0049] Meanwhile, the artificial intelligence-based fire detection system (100) may further include a sensor monitoring unit (not shown) that detects whether the sensor unit (110) is malfunctioning by calculating at least one of the average value, maximum value, minimum value, and rate of change of the fire factor data collected from the sensor unit (110) for a preset time.

[0050] For example, the sensor monitoring unit monitors the average value, maximum value, minimum value, and rate of change of the fire factor data in a normal state where no fire has occurred, and counts cases where the difference between the maximum value and the average value of the fire factor data exceeds a preset reference value, or cases where the difference between the minimum value and the average value of the fire factor data exceeds the reference value, and if the counted number exceeds the reference number, it can determine that a malfunction has occurred in the sensor unit (110).

[0051] In addition, the rate of change of the fire factor data collected through the sensor unit is monitored, and it may be determined that a malfunction has occurred in the sensor unit (110) even if the number of times the rate of change changes by more than a preset standard rate in a normal state where no fire has occurred exceeds the standard number of times.

[0052] As another example, if the sensor monitoring unit detects a malfunction of the sensor unit (110), it may transmit a notification signal to the administrator terminal requesting an inspection of the sensor unit (110).

[0054] As another example, the sensor monitoring unit may further include a noise analysis unit (not shown) that receives and analyzes data collected through the gas sensor to determine whether noise is generated due to foreign matter being attached to the surface of the gas sensor.

[0055] In other words, if foreign substances are attached to the surface of the gas sensor, the sensor measurements will exhibit an irregular or irregular pattern; by taking these characteristics into account, it is possible to determine whether foreign substances are attached to the sensor surface and, furthermore, to determine sensor errors.

[0056] To this end, the noise analysis unit calculates the average and standard deviation of multiple sensor values ​​measured over a preset period of time, and if there is a preset number (e.g., 5) or more of sensor values ​​(hereinafter referred to as 'noise data') that deviate from the average and standard deviation among the multiple sensor values, it can be estimated that foreign matter is attached to the surface of the sensor.

[0057] At the same time, the measured time interval for each of the above noise data is calculated, and if the time interval is not constant, it can be determined that foreign matter is attached to the sensor surface.

[0058] Here, whether the above time interval is non-constant is determined by the maximum difference in time intervals between noise data (M) calculated by [Mathematical Formula 1] below. d The average value of the time interval between noise data calculated according to [Equation 2] (T av If ) exceeds a constant multiple (e.g., 2) or more, it can be determined that the above time interval is not constant.

[0060] [Mathematical Formula 1]

[0061]

[0062] (Here, M d is the maximum difference value of the time interval between noise data, N max is the maximum value of the noise data time interval, N min )

[0064] [Mathematical Formula 2]

[0065]

[0066] (Here, T av is the average value of the time intervals between noise data, T (i-1) to i is the time interval between the i-1th noise data and the i-th noise data, T 0 to 1 represents the time interval between the sensor operation time and the appearance of the first noise data, and n represents the number of noise data.

[0068] For example, as shown in below, from T0 to T 20 When the data is 9, 10, 11, 12, 14, 18, 9, 9, 6, 13, 2, 8, 9, 9, 10, 17, 9, 10, 3, 12, the mean is 10 and the standard deviation is 3.42, so there are 6 noise data points: 14, 18, 6, 2, 17, 3. Since the number of noise points is 5 or more, which is the preset number, it can be presumed that foreign matter is attached to the surface of the sensor.

[0070]

[0071]

[0073] Subsequently, if we calculate whether the noise appears irregularly according to [Equation 1] and [Equation 2], the maximum difference in time intervals between noise data (M) as shown in below d ) becomes 8-1=7, and the average value of the time intervals between noise data (T av ) becomes 3.5.

[0075] Table 2

[0076]

[0078] Therefore, the maximum difference value of the time interval between noise data (M d ) 7 is the average value of the time intervals between noise data (T av Since it is more than twice 3.5, it can be determined that foreign matter is attached to the surface of the sensor in this case.

[0080] As described above, in one embodiment of the present invention, errors caused by foreign substances adhering to the sensor surface can be monitored and diagnosed through a noise analysis unit provided in the sensor monitoring unit, thereby enabling a more accurate determination of whether the sensor unit (110) is in a normal state.

[0082] According to the present invention as described above, by considering gas concentration and temperature data collected through a sensor, real-time atmospheric information, location information, and time information, and classifying and calculating fire occurrence conditions by time period and season, an artificial intelligence-based fire detection system and method capable of more accurate fire detection and rapid response according to the characteristics of a location requiring fire detection can be provided.

[0083] In addition, an artificial intelligence-based fire detection system and method can be provided that can prevent safety accidents that could lead to a fire before an actual fire occurs by generating an early warning based on the gas concentration collected through the sensor and the duration of the gas concentration.

[0085] In addition, a control method for an artificial intelligence-based fire detection system according to an embodiment of the present invention may be recorded on a computer-readable medium containing program instructions for performing operations implemented by various computers. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions on the medium may be those specifically designed and configured for the present invention, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0087] As described above, although an embodiment of the present invention has been explained by limited embodiments and drawings, the embodiment of the present invention is not limited to the embodiments described above, and various modifications and variations are possible from this description by those skilled in the art to which the present invention pertains. Accordingly, an embodiment of the present invention should be understood only by the claims described below, and all equivalent or analogous variations thereof shall be considered to be within the scope of the inventive concept. Explanation of the symbols

[0089] 110 : Sensor part 111 : Housing 120 : Environmental Information Collection Department 130 : Data learning section 140: Fire detection unit 141: First detection unit 142 : Second detection unit 150: Fire Suppression Unit

Claims

Claim 1 A sensor unit equipped with at least one gas sensor and one temperature sensor to collect fire factor data; an environmental information collection unit including at least one GPS module to collect location information and time information, and to receive atmospheric information including temperature, wind speed, fine dust concentration, and humidity from a preset weather observation server; a data learning unit that applies a preset artificial intelligence learning algorithm to the data collected from the sensor unit and the environmental information collection unit to calculate fire occurrence condition data; a fire detection unit that determines whether a fire has occurred by comparing the fire factor data collected through the sensor unit with the fire occurrence condition data calculated by the data learning unit; and a fire suppression unit that controls the execution of a fire suppression manual through a fire suppression module provided in a preset area when it is determined that a fire has occurred by the fire detection unit.The fire factor data includes real-time gas concentration data including at least one of CO2 and CO collected through the gas sensor, the duration of the real-time gas concentration data, real-time temperature data collected through the temperature sensor, and the duration of the real-time temperature data; the data learning unit calculates seasonal fire occurrence condition data based on the atmospheric information collected from the environmental information collection unit and the location information, and calculates hourly fire occurrence condition data based on the time information collected from the environmental information collection unit and the real-time gas concentration data and real-time temperature data collected from the sensor unit; and the fire detection unit detects whether a fire has occurred by determining whether the fire factor data collected through the sensor unit satisfies the fire occurrence condition data calculated by the data learning unit, wherein if the real-time gas concentration data collected through the sensor unit exceeds a preset first reference concentration and the duration of the real-time gas concentration data exceeds a preset first reference time, the first detection unit generates a fire early warning signal and transmits an alarm signal to a preset administrator terminal. and a second detection unit that, when the above fire early warning signal is generated and the real-time temperature data collected through the sensor unit within a preset second reference time exceeds a preset first reference temperature, the duration of the real-time temperature data exceeds a preset third reference time, and the fire occurrence condition data calculated by the data learning unit is satisfied, generates a fire occurrence alarm signal, transmits an alarm signal to the administrator terminal, and generates a control signal to control the fire suppression module;The fire suppression module comprises at least one throwable fire extinguisher coupled to a housing on which the sensor unit is mounted, and the fire suppression unit controls the fire suppression module so that, when it is determined through the fire detection unit that a fire has occurred, the throwable fire extinguisher is separated from the housing and thrown into the fire area based on the fire suppression manual, and the sensor monitoring unit detects whether the sensor unit is malfunctioning by calculating at least one of the average value, maximum value, minimum value, and rate of change of the fire factor data collected from the sensor unit for a preset time period;The sensor monitoring unit further includes the following: monitoring the average, maximum, minimum, and rate of change of the fire factor data in a normal state where no fire has occurred; counting cases where the difference between the maximum and average values ​​of the fire factor data exceeds a preset reference value, and when the counted number exceeds a reference number, determining that a malfunction has occurred in the sensor unit; monitoring the rate of change of the fire factor data collected through the sensor unit; and when the number of times the rate of change changes by more than a preset reference ratio in a normal state where no fire has occurred exceeds the reference number, determining that a malfunction has occurred in the sensor unit; the sensor monitoring unit further includes a noise analysis unit that receives and analyzes data collected through the gas sensor to determine whether noise is generated due to foreign matter adhering to the surface of the gas sensor; the noise analysis unit calculates the average and standard deviation of a plurality of sensor values ​​measured over a preset period of time, and among the plurality of sensor values, a sensor value that deviates from the average by a standard deviation (hereinafter referred to as 'noise data'). If the number of () exceeds a preset number, it is presumed that foreign matter is attached to the sensor surface, and the measured time interval for each of the noise data is calculated; if the time interval is not constant, it is confirmed that foreign matter is attached to the sensor surface, provided that whether the time interval is not constant is determined by the maximum difference value (M) between the time intervals of the noise data calculated by the following [Equation 1]. d The average value of the time interval between noise data calculated according to [Equation 2] (T av An AI-based fire detection system characterized by determining that the time interval is non-constant when ) exceeds a constant multiple or more. [Mathematical Formula 1] (Here, M d is the maximum difference value of the time interval between noise data, N max is the maximum value of the noise data time interval, N min ☐ represents the minimum value of the noise data time interval, respectively)[Equation 2] (Here, T av is the average value of the time intervals between noise data, T (i-1) to i is the time interval between the i-1th noise data and the i-th noise data, T 0 to 1 represents the time interval between the sensor operation time and the appearance of the first noise data, and n represents the number of noise data. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete

Images