IoT intelligent analysis and early warning system and method based on edge computing gateway

Through the edge computing gateway combining image recognition and sensor data, a fire hazard assessment model is established and the water spray volume and pressure are adaptively controlled, which solves the problem of false alarms and resource waste in the fire system, and improves fire safety and resource utilization efficiency.

CN118658285BActive Publication Date: 2024-11-05CHENGANSHENGBANG (CHANGCHUN) NETWORK TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411125332.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-05
Estimated Expiration
2044-08-16

AI Technical Summary

Technical Problem

The existing fire protection systems have frequent false alarms, waste of resources and inability to adapt to changes in the spray system, and have failed to effectively use material information to analyze fire hazards.

Method used

Using an IoT intelligent analysis and early warning system based on edge computing gateway, through image recognition technology and sensor data, a grayscale and radiation energy density fitting model is established, flame characteristic analysis and fire hazard evaluation are carried out, and total water spray and spray are adaptively controlled. pressure.

Benefits of technology

It improves system computing efficiency, accuracy and resource utilization, reduces false alarm processing, and ensures fire safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118658285B_ABST
    Figure CN118658285B_ABST
Patent Text Reader

Abstract

The present invention discloses an Internet of Things intelligent analysis and early warning system and method based on an edge computing gateway, and belongs to the technical field of the Internet of Things for fire protection. The present invention uses image recognition technology to perform image recognition on a test flame image, obtains characteristic information of the flame image, and constructs a fitting model according to the characteristic information of the flame image; collects data of the area to be tested in real time through sensors and cameras, transmits the data of the area to be tested to the edge computing gateway, obtains the location distribution of the area to be tested according to the sensor number, and extracts material information of the area to be tested according to the location distribution; the edge computing gateway extracts the data features of the area to be tested, and performs a hazard analysis according to the data features of the area to be tested; based on the hazard analysis results of the data features of the area to be tested, the administrator is warned and the fire extinguishing equipment is adaptively controlled. Improve the accuracy of the degree of fire hazards of central materials; improve the utilization efficiency of water resources and ensure the safety of materials.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of fire protection Internet of Things technology, and specifically to an Internet of Things intelligent analysis and early warning system and method based on an edge computing gateway. Background Art

[0002] With the acceleration of my country's urbanization process, the number of high-rise and large buildings and various types of venues in cities is increasing, and the fire safety situation is extremely severe. Fire supervision work is facing increasingly severe tests due to limited manpower and lack of effective technical support and social coordination.

[0003] Fire protection system facilities generally have a large number of false alarms. Administrators need to spend a lot of manpower and material resources to deal with false alarms in a timely manner, which easily leads to waste of resources. In addition, the fire protection system does not take the surrounding materials stacked into account in the fire hazard analysis. The sprinkler system only uses the preset water volume and pressure to extinguish the fire, and cannot achieve adaptive changes.

[0004] Therefore, the present invention discloses an IoT intelligent analysis and early warning system and method based on an edge computing gateway to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide an IoT intelligent analysis and early warning system and method based on edge computing gateway to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solution: an IoT intelligent analysis and early warning method based on an edge computing gateway, the method comprising the following steps:

[0007] S1: Use image recognition technology to perform image recognition on the test flame image, obtain feature information of the flame image, form a grayscale radiation energy density set based on the feature information of the flame image, and establish a grayscale and radiation energy density fitting model based on the grayscale radiation energy density set;

[0008] S2: Analyze the fire hazard degree of materials according to the material information and the impact hazard area, collect data of the area to be tested in real time through the sensor set and camera, transmit the data of the area to be tested to the edge computing gateway, obtain the location distribution of the area to be tested according to the sensor number, and extract the material information of the area to be tested according to the location distribution;

[0009] S3: The edge computing gateway uses image recognition technology to perform image recognition on the data of the area to be tested, extracts the data features of the area to be tested, and performs hazard analysis based on the data features of the area to be tested;

[0010] S4: Based on the hazard analysis results of the data characteristics of the area to be tested, the administrator is warned and the total amount of water sprayed and the injection pressure of the fire extinguishing equipment are adaptively controlled.

[0011] According to the above scheme, in S1, the following steps are included:

[0012] S101: graying the test flame image, extracting the flame contour of the grayed test flame image using a contour extraction algorithm, and obtaining the gray value of each pixel in the test flame contour image;

[0013] S102: Acquire a test thermal radiation image corresponding to the test flame outline image, and calculate the radiation energy density using Planck's law for each pixel in the test thermal radiation image;

[0014] S103: Correlate the grayscale values ​​corresponding to the pixels of the test flame outline image with the radiation energy density corresponding to the pixels in the test thermal radiation image to construct a grayscale radiation energy density set, denoted as GRED=={(GV j , E j )|j∈[1,M]};where GV j Indicates the gray value corresponding to the j-th pixel, E j represents the radiation energy density corresponding to the j-th pixel, and M represents the total number of pixels;

[0015] S104: According to the grayscale radiation energy density set, a grayscale and radiation energy density fitting model is established: y=β1×x+β2; wherein β1 and β2 represent fitting coefficients, x represents the independent variable of the pixel grayscale value, and y represents the dependent variable of the pixel radiation energy density. The least squares method is used to calculate and solve β1 and β2 in the grayscale and radiation energy density fitting model.

[0016] By constructing a grayscale and radiation energy density fitting model, the pixel points of the flame outline image and the radiation energy density can be associated, which can speed up the calculation speed of the system. There is no need to calculate the radiation energy density of each flame image in the actual analysis, thus improving the overall efficiency of the system.

[0017] According to the above solution, in S2, the sensors include a temperature sensor, a smoke sensor and a thermal radiation sensor;

[0018] Before collecting the data of the area to be measured in real time by the sensor, the following steps are included:

[0019] S201: different types of sensors within the same preset range in the building are grouped together, and different types of sensors in the same sensor group are installed at the same position; and an indoor planning map of the building is extracted;

[0020] S202: acquiring sensor set installation positions according to the indoor planning map, generating a sensor set position matrix according to the sensor set installation positions, and numbering the sensor sets according to the sensor set position matrix;

[0021] S203: Associating the material information with the sensor set installation location, and performing a first fire hazard level assessment on the material according to the material information. The specific calculation formula is as follows;

[0022] RI = α × (MG ÷ AR);

[0023] Wherein RI represents the first fire hazard level of the material, α represents the combustibility coefficient of the material, and the combustibility coefficient of the material is preset according to the chemical material of the material; MG represents the weight of the material; AR represents the area occupied by the material;

[0024] S204: according to the first fire hazard degree of the material, different impact distances are set for the material, and a circular area is divided for each material as an impact hazard area with the geometric center of the material's floor area as the center and the impact distance as the radius; the area of ​​the overlapping area with the impact hazard area of ​​the central material is extracted, and the first fire hazard degree of the surrounding materials belonging to the overlapping area of ​​the impact hazard area of ​​the central material is extracted; the second fire hazard degree of the central material is calculated according to the first fire hazard degree and the area of ​​the overlapping area of ​​the impact hazard area. The specific calculation formula is as follows:

[0025] ;

[0026] RII represents the second fire hazard level of the central materials, S i Represents the area of ​​the overlapping area of ​​the ith potential hazard zone, RI i It represents the first fire hazard level of the surrounding materials in the overlapping area of ​​the i-th impact hazard zone, and RI represents the first fire hazard level of the central materials; i∈[1,n], where n represents the total number of overlapping areas affecting the hazard zone, S0 represents the area of ​​the hazard zone affected by the central materials, and RI0 represents the first fire hazard level of the central materials.

[0027] Fire has the ability to spread quickly, and surrounding materials will have a certain impact on central materials. Calculating the fire hazard level of central materials and nearby surrounding materials can improve the accuracy of the fire hazard level of central materials.

[0028] According to the above scheme, in S3, the real-time collected image is grayed, and the flame contour is extracted from the grayed real-time collected image using a contour extraction algorithm to obtain the gray value of each pixel in the real-time flame contour image;

[0029] Substituting the grayscale value of each pixel in the real-time flame outline image into the grayscale and radiation energy density fitting model, the radiation energy density of each pixel in the real-time flame outline image is calculated; performing spectral analysis on all pixels of the real-time flame outline image, obtaining the wavelength range of each pixel, summing the radiation energy density of all wavelength ranges of each pixel, and then summing the radiation energy of all pixels to obtain the total radiation energy of the real-time flame outline image;

[0030] The hazard assessment of the real-time flame contour image is performed, and the specific calculation formula is as follows:

[0031] FAT=A1×RII+A2×TRE+A3×(B1×T×L);

[0032] Among them, FAT represents the risk score of the real-time flame profile image, TRE represents the total radiation energy of the real-time flame profile image, T represents the real-time indoor temperature, L represents the distance from the flame position to the sprinkler port, A1, A2 and A3 represent different risk coefficients, B1 represents the indoor temperature conversion coefficient, and A1, A2, A3 and B1 are all system preset constants;

[0033] The method for calculating the distance from the flame position to the spray port comprises: using image recognition technology to identify the pixel positions of the preset calibration object and the flame on the real-time collected image, and calculating the actual distance vector from the flame to the camera through the internal and external parameters of the camera; the position of the calibration object and the camera is fixed; and calculating the distance from the flame position to the spray port according to the actual distance vector from the flame to the camera and the distance vector from the camera to the spray port;

[0034] Determine whether the hazard score of the real-time flame is greater than the hazard score threshold. If it is greater than, the real-time flame will be taken as a key monitoring target. If it is less than or equal to, no processing will be done.

[0035] According to the above scheme, in S4, the total number of key monitoring targets is counted, and the total hazard score of all key monitoring targets is calculated; a total hazard score threshold is set for the total hazard score. If the total hazard score is greater than the total hazard score threshold, a first-level warning is issued to the administrator, and the specific distribution location and material information of the key monitoring targets are sent; according to the spatial position relationship between the sprinkler port and the key monitoring target, the sprinkler port is adaptively controlled and managed;

[0036] If the total hazard score is less than or equal to the total hazard score threshold, a secondary warning will be issued to the administrator; the sprinkler outlets will be controlled and managed adaptively based on the spatial position relationship between the sprinkler outlets and the key monitoring targets;

[0037] The emergency situation of the first-level warning is higher than that of the second-level warning;

[0038] The control and management of the spray port includes:

[0039] Obtain the spatial positions of all key monitoring targets and all sprinkler openings;

[0040] According to the spatial position of the key monitoring target and the spatial position of the spray outlet, the spray outlet closest to the key monitoring target is selected as the target spray outlet;

[0041] Calculate the total hazard score for the key monitoring targets of the same sprinkler outlet, and extract the distance from the key monitoring target farthest from the sprinkler outlet to the sprinkler outlet;

[0042] The total amount of water sprayed and the injection pressure of the sprinkler outlet are controlled according to the total distribution of the hazard scores of each sprinkler outlet and the distance from the key monitoring target farthest from the sprinkler outlet to the sprinkler outlet; the total amount of water sprayed and the injection pressure are system preset values.

[0043] The key monitoring targets of the same sprinkler outlet are assigned a hazard score. The total amount of water sprayed and the injection pressure of the sprinkler outlet are controlled according to the total hazard score distribution of each sprinkler outlet and the distance from the key monitoring target farthest from the sprinkler outlet. The total amount of water sprayed and the injection pressure can be adaptively adjusted according to the actual flame situation, thereby improving the utilization efficiency of water resources and further ensuring the safety of other materials.

[0044] Another aspect of the present application provides an IoT intelligent analysis and early warning system based on an edge computing gateway, the system is applied to the implementation of the IoT intelligent analysis and early warning method based on an edge computing gateway, the system includes a fitting model construction module, a fire hazard analysis module, a data acquisition module, a hazard analysis module and an early warning control module;

[0045] The fitting model building module is used to perform image recognition on the test flame image using image recognition technology, obtain feature information of the flame image, and build a fitting model according to the feature information of the flame image;

[0046] The fire hazard analysis module is used to obtain the sensor set installation location according to the indoor planning map, number the sensor set according to the sensor set location matrix, associate the material information with the sensor set installation location, and analyze the fire hazard degree according to the material information;

[0047] The data acquisition module is used to collect data of the area to be tested in real time through sensors and cameras, transmit the data of the area to be tested to the edge computing gateway, obtain the location distribution of the area to be tested according to the sensor number, and extract material information of the area to be tested according to the location distribution.

[0048] The risk analysis module is used in the edge computing gateway to use image recognition technology to perform image recognition on the data of the area to be tested, extract the data features of the area to be tested, and perform risk analysis based on the data features of the area to be tested;

[0049] The early warning control module warns the administrator and performs adaptive control on the fire extinguishing equipment based on the risk analysis results of the data characteristics of the area to be tested.

[0050] According to the above scheme, the fitting model construction module includes a flame characteristic analysis unit and a fitting model generation unit;

[0051] The flame feature analysis unit is used to grayscale the test flame image, extract the flame contour of the grayscale test flame image using a contour extraction algorithm, and obtain the grayscale value and radiation energy density of each pixel in the test flame contour image;

[0052] The fitting model generation unit is used to construct a grayscale radiation energy density set, establish a grayscale and radiation energy density fitting model according to the grayscale radiation energy density set, and solve the fitting coefficient.

[0053] According to the above scheme, the fire hazard analysis module includes an information integration unit and a fire hazard degree analysis unit;

[0054] The information integration unit is used to group different types of sensors within the same preset range in the building into a set, and the installation positions of different types of sensors in the same sensor set are consistent; extract the indoor planning map of the building; obtain the installation position of the sensor set according to the indoor planning map, generate a sensor set position matrix according to the installation position of the sensor set, number the sensor set according to the sensor set position matrix, and associate the material information with the installation position of the sensor set;

[0055] The fire hazard level analysis unit evaluates the first fire hazard level of the materials according to the material information, sets different impact distances for the materials according to the first fire hazard level of the materials, and calculates the second fire hazard level of the central materials according to the area of ​​the overlapping area between the first fire hazard level and the impact hazard zone.

[0056] According to the above scheme, the risk analysis module includes a risk characteristic calculation unit and a monitoring target processing unit;

[0057] The hazard characteristic calculation unit is used to grayscale the real-time collected image, extract the flame contour of the grayscaled real-time collected image using a contour extraction algorithm, and obtain the grayscale value of each pixel in the real-time flame contour image; substitute the grayscale value into the grayscale and radiation energy density fitting model to calculate the radiation energy density, and calculate the total radiation energy according to the radiation energy density; calculate the distance from the flame position to the spray port;

[0058] The monitoring target processing unit is used to determine whether the danger score of the real-time flame is greater than the danger score threshold. If it is greater, the real-time flame is taken as a key monitoring target; if it is less than or equal to the danger score threshold, no processing is performed.

[0059] According to the above solution, the early warning control module includes an early warning unit and a control unit;

[0060] The early warning unit is used to count the total number of key monitoring targets and calculate the total value of the risk scores of all key monitoring targets; set a total value threshold of the risk score for the total value of the risk score; if the total value of the risk score is greater than the total value threshold of the risk score, a first-level early warning is issued to the administrator; if the total value of the risk score is less than or equal to the total value threshold of the risk score, a second-level early warning is issued to the administrator;

[0061] The control unit is used to adaptively control and manage the spray port according to the spatial position relationship between the spray port and the key monitoring target.

[0062] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: based on the construction of a grayscale and radiation energy density fitting model, the pixel points of the flame outline image and the radiation energy density can be associated, the calculation speed of the system can be accelerated, and there is no need to calculate the radiation energy density of each flame image in the actual analysis, thereby improving the overall efficiency of the system; fire has the ability to spread rapidly, and the surrounding materials will have a certain impact on the central materials. The fire hazard degree of the central materials and the surrounding materials nearby is calculated to improve the accuracy of the fire hazard degree of the central materials; according to the actual flame situation, the total amount of water sprayed and the injection pressure are adaptively adjusted, which improves the utilization efficiency of water resources and can further ensure the safety of other materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0064] Figure 1 It is a flow chart of the IoT intelligent analysis and early warning method based on edge computing gateway of the present invention;

[0065] Figure 2It is a structural schematic diagram of the Internet of Things intelligent analysis and early warning system based on edge computing gateway of the present invention. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] See also Figure 1 The present invention provides a technical solution: an IoT intelligent analysis and early warning method based on an edge computing gateway, the method comprising the following steps:

[0068] S1: Use image recognition technology to perform image recognition on the test flame image, obtain feature information of the flame image, form a grayscale radiation energy density set based on the feature information of the flame image, and establish a grayscale and radiation energy density fitting model based on the grayscale radiation energy density set;

[0069] In S1, the following steps are included:

[0070] S101: graying the test flame image, extracting the flame contour of the grayed test flame image using a contour extraction algorithm, and obtaining the gray value of each pixel in the test flame contour image;

[0071] S102: Acquire a test thermal radiation image corresponding to the test flame outline image, and calculate the radiation energy density using Planck's law for each pixel in the test thermal radiation image;

[0072] S103: Correlate the grayscale values ​​corresponding to the pixels of the test flame outline image with the radiation energy density corresponding to the pixels in the test thermal radiation image to construct a grayscale radiation energy density set, denoted as GRED=={(GV j , E j )|j∈[1,M]};where GV j Indicates the gray value corresponding to the j-th pixel, E j represents the radiation energy density corresponding to the j-th pixel, and M represents the total number of pixels;

[0073] S104: According to the grayscale radiation energy density set, a grayscale and radiation energy density fitting model is established: y=β1×x+β2; wherein β1 and β2 represent fitting coefficients, x represents the independent variable of the pixel grayscale value, and y represents the dependent variable of the pixel radiation energy density. The least squares method is used to calculate and solve β1 and β2 in the grayscale and radiation energy density fitting model.

[0074] S2: Analyze the fire hazard degree of materials according to the material information and the impact hazard area, collect data of the area to be tested in real time through the sensor set and camera, transmit the data of the area to be tested to the edge computing gateway, obtain the location distribution of the area to be tested according to the sensor number, and extract the material information of the area to be tested according to the location distribution;

[0075] In S2, the sensors include a temperature sensor, a smoke sensor, and a thermal radiation sensor;

[0076] Before collecting data of the area to be measured in real time through sensors, the following steps are included:

[0077] S201: different types of sensors within the same preset range in the building are grouped together, and different types of sensors in the same sensor group are installed at the same position; and an indoor planning map of the building is extracted;

[0078] S202: acquiring sensor set installation positions according to the indoor planning map, generating a sensor set position matrix according to the sensor set installation positions, and numbering the sensor sets according to the sensor set position matrix;

[0079] S203: Associating the material information with the sensor set installation location, and performing a first fire hazard level assessment on the material according to the material information. The specific calculation formula is as follows;

[0080] RI = α × (MG ÷ AR);

[0081] Among them, RI represents the first fire hazard level of the material, α represents the combustibility coefficient of the material, which is preset according to the chemical material of the material; MG represents the weight of the material; AR represents the area occupied by the material;

[0082] S204: according to the first fire hazard degree of the material, different impact distances are set for the material, and a circular area is divided for each material as an impact hazard area with the geometric center of the material's floor area as the center and the impact distance as the radius; the area of ​​the overlapping area with the impact hazard area of ​​the central material is extracted, and the first fire hazard degree of the surrounding materials belonging to the overlapping area of ​​the impact hazard area of ​​the central material is extracted; the second fire hazard degree of the central material is calculated according to the first fire hazard degree and the area of ​​the overlapping area of ​​the impact hazard area. The specific calculation formula is as follows:

[0083] ;

[0084] RII represents the second fire hazard level of the central materials, S i Represents the area of ​​the overlapping area of ​​the ith potential hazard zone, RI iIt represents the first fire hazard level of the surrounding materials in the overlapping area of ​​the i-th impact hazard zone, and RI represents the first fire hazard level of the central materials; i∈[1,n], where n represents the total number of overlapping areas affecting the hazard zone, S0 represents the area of ​​the hazard zone affected by the central materials, and RI0 represents the first fire hazard level of the central materials.

[0085] Example 1: In this example, the impact hazard area of ​​the central material is 4π, and the first fire hazard level is 10; the overlapping area of ​​the surrounding materials that overlap with the impact hazard area of ​​the central material and the first fire hazard level are combined into a set; the set is ;

[0086] The specific calculation of the second fire hazard level RII of the central materials is as follows:

[0087] ;

[0088] S3: The edge computing gateway uses image recognition technology to perform image recognition on the data of the area to be tested, extracts the data features of the area to be tested, and performs hazard analysis based on the data features of the area to be tested;

[0089] In S3, the real-time collected image is grayed, and the flame contour is extracted from the grayed real-time collected image using a contour extraction algorithm to obtain the gray value of each pixel in the real-time flame contour image;

[0090] Substituting the grayscale value of each pixel in the real-time flame outline image into the grayscale and radiation energy density fitting model, the radiation energy density of each pixel in the real-time flame outline image is calculated; performing spectral analysis on all pixels of the real-time flame outline image, obtaining the wavelength range of each pixel, summing the radiation energy density of all wavelength ranges of each pixel, and then summing the radiation energy of all pixels to obtain the total radiation energy of the real-time flame outline image;

[0091] The hazard assessment of the real-time flame contour image is performed, and the specific calculation formula is as follows:

[0092] FAT=A1×RII+A2×TRE+A3×(B1×T×L);

[0093] Among them, FAT represents the risk score of the real-time flame profile image, TRE represents the total radiation energy of the real-time flame profile image, T represents the real-time indoor temperature, L represents the distance from the flame position to the sprinkler port, A1, A2 and A3 represent different risk coefficients, B1 represents the indoor temperature conversion coefficient, and A1, A2, A3 and B1 are all system preset constants;

[0094] The method for calculating the distance from the flame position to the sprinkler port includes: using image recognition technology to identify the pixel positions of the preset calibration object and the flame on the real-time collected image, and calculating the actual distance vector from the flame to the camera through the internal and external parameters of the camera; the position of the calibration object and the camera is fixed; and calculating the distance from the flame position to the sprinkler port according to the actual distance vector from the flame to the camera and the distance vector from the camera to the sprinkler port;

[0095] Determine whether the hazard score of the real-time flame is greater than the hazard score threshold. If it is greater than, the real-time flame will be taken as a key monitoring target. If it is less than or equal to, no processing will be done.

[0096] S4: Based on the hazard analysis results of the data characteristics of the area to be tested, the administrator is warned and the total amount of water sprayed and the injection pressure of the fire extinguishing equipment are adaptively controlled.

[0097] In S4, the total number of key monitoring targets is counted, and the total hazard score of all key monitoring targets is calculated; a total hazard score threshold is set for the total hazard score. If the total hazard score is greater than the total hazard score threshold, a first-level warning is issued to the administrator, and the specific distribution location and material information of the key monitoring targets are sent; according to the spatial position relationship between the sprinkler port and the key monitoring target, the sprinkler port is adaptively controlled and managed;

[0098] If the total hazard score is less than or equal to the total hazard score threshold, a secondary warning will be issued to the administrator; the sprinkler outlets will be controlled and managed adaptively based on the spatial position relationship between the sprinkler outlets and the key monitoring targets;

[0099] The emergency situation of the first-level warning is higher than that of the second-level warning;

[0100] Control and management of sprinkler ports include:

[0101] Obtain the spatial positions of all key monitoring targets and all sprinkler openings;

[0102] According to the spatial position of the key monitoring target and the spatial position of the spray outlet, the spray outlet closest to the key monitoring target is selected as the target spray outlet;

[0103] Calculate the total hazard score for the key monitoring targets of the same sprinkler outlet, and extract the distance from the key monitoring target farthest from the sprinkler outlet to the sprinkler outlet;

[0104] The total amount of water sprayed and the injection pressure of the sprinkler outlets are controlled according to the total distribution of the hazard scores of each sprinkler outlet and the distance from the key monitoring target farthest from the sprinkler outlet to the sprinkler outlet; the total amount of water sprayed and the injection pressure are system preset values.

[0105] See also Figure 2, the present invention provides a technical solution: an IoT intelligent analysis and early warning system based on an edge computing gateway, the system comprising a fitting model building module, a fire hazard analysis module, a data acquisition module, a hazard analysis module and an early warning control module;

[0106] The fitting model building module is used to perform image recognition on the test flame image using image recognition technology, obtain feature information of the flame image, and build a fitting model according to the feature information of the flame image;

[0107] The fire hazard analysis module is used to obtain the sensor set installation location according to the indoor planning map, number the sensor set according to the sensor set location matrix, associate the material information with the sensor set installation location, and analyze the fire hazard degree according to the material information;

[0108] The data acquisition module is used to collect data of the area to be tested in real time through sensors and cameras, transmit the data of the area to be tested to the edge computing gateway, obtain the location distribution of the area to be tested according to the sensor number, and extract the material information of the area to be tested according to the location distribution.

[0109] The hazard analysis module is used in the edge computing gateway to use image recognition technology to perform image recognition on the data of the area to be tested, extract the data features of the area to be tested, and perform hazard analysis based on the data features of the area to be tested;

[0110] The early warning control module warns the administrator and performs adaptive control of the fire-fighting equipment based on the hazard analysis results of the data characteristics of the area to be tested.

[0111] The fitting model building module includes a flame characteristic analysis unit and a fitting model generation unit;

[0112] The flame feature analysis unit is used to grayscale the test flame image, extract the flame contour of the grayscale test flame image using a contour extraction algorithm, and obtain the grayscale value and radiation energy density of each pixel in the test flame contour image;

[0113] The fitting model generation unit is used to construct a grayscale radiation energy density set, establish a grayscale and radiation energy density fitting model according to the grayscale radiation energy density set, and solve the fitting coefficient.

[0114] The fire hazard analysis module includes an information integration unit and a fire hazard degree analysis unit;

[0115] The information integration unit is used to group different types of sensors within the same preset range in the building into a set, and the installation positions of different types of sensors in the same sensor set are consistent; extract the indoor planning map of the building; obtain the installation position of the sensor set according to the indoor planning map, generate a sensor set position matrix according to the installation position of the sensor set, number the sensor set according to the sensor set position matrix, and associate the material information with the installation position of the sensor set;

[0116] The fire hazard degree analysis unit evaluates the first fire hazard degree of the materials according to the material information, sets different impact distances for the materials according to the first fire hazard degree of the materials, and calculates the second fire hazard degree of the central materials according to the area of ​​the overlapping area between the first fire hazard degree and the impact hazard zone.

[0117] The hazard analysis module includes a hazard characteristic calculation unit and a monitoring target processing unit;

[0118] The hazard characteristic calculation unit is used to grayscale the real-time collected image, extract the flame contour of the grayscaled real-time collected image using a contour extraction algorithm, and obtain the grayscale value of each pixel in the real-time flame contour image; substitute the grayscale value into the grayscale and radiation energy density fitting model to calculate the radiation energy density, and calculate the total radiation energy according to the radiation energy density; calculate the distance from the flame position to the spray port;

[0119] The monitoring target processing unit is used to determine whether the danger score of the real-time flame is greater than the danger score threshold. If it is greater, the real-time flame is taken as a key monitoring target; if it is less than or equal to the danger score threshold, no processing is performed.

[0120] The early warning control module includes an early warning unit and a control unit;

[0121] The early warning unit is used to count the total number of key monitoring targets and calculate the total value of the risk scores of all key monitoring targets; set a total value threshold for the risk score, if the total value is greater than the total value threshold, a first-level early warning is issued to the administrator; if the total value is less than or equal to the total value threshold, a second-level early warning is issued to the administrator;

[0122] The control unit is used to adaptively control and manage the sprinkler openings according to the spatial position relationship between the sprinkler openings and key monitoring targets.

[0123] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0124] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An IoT intelligent analysis and early warning method based on edge computing gateway, characterized in that: The method comprises the following steps: S1: Use image recognition technology to perform image recognition on the test flame image, obtain feature information of the flame image, form a grayscale radiation energy density set based on the feature information of the flame image, and establish a grayscale and radiation energy density fitting model based on the grayscale radiation energy density set; S2: Analyze the fire hazard degree of materials according to the material information and the impact hazard area, collect data of the area to be tested in real time through the sensor set and camera, transmit the data of the area to be tested to the edge computing gateway, obtain the location distribution of the area to be tested according to the sensor number, and extract the material information of the area to be tested according to the location distribution; S3: The edge computing gateway uses image recognition technology to perform image recognition on the data of the area to be tested, extracts the data features of the area to be tested, and performs hazard analysis based on the data features of the area to be tested; S4: Based on the hazard analysis results of the data characteristics of the area to be tested, the administrator is warned and the total amount of water sprayed and the spray pressure of the fire extinguishing equipment are adaptively controlled; In S1, the following steps are included: S101: graying the test flame image, extracting the flame contour of the grayed test flame image using a contour extraction algorithm, and obtaining the gray value of each pixel in the test flame contour image; S102: Acquire a test thermal radiation image corresponding to the test flame outline image, and calculate the radiation energy density using Planck's law for each pixel in the test thermal radiation image; S103: Correlate the grayscale values ​​corresponding to the pixels of the test flame outline image with the radiation energy density corresponding to the pixels in the test thermal radiation image to construct a grayscale radiation energy density set, denoted as GRED=={(GV j , E j )|j∈[1,M]};where GV j Indicates the gray value corresponding to the j-th pixel, E j represents the radiation energy density corresponding to the j-th pixel, and M represents the total number of pixels; S104: According to the grayscale radiation energy density set, a grayscale and radiation energy density fitting model is established: y=β1×x+β2; wherein β1 and β2 represent fitting coefficients, x represents the independent variable of the pixel grayscale value, and y represents the dependent variable of the pixel radiation energy density. The least square method is used to calculate and solve β1 and β2 in the grayscale and radiation energy density fitting model; In S2, the sensors include a temperature sensor, a smoke sensor, and a thermal radiation sensor; Before collecting the data of the area to be measured in real time by the sensor, the following steps are included: S201: different types of sensors within the same preset range in the building are grouped together, and different types of sensors in the same sensor group are installed at the same position; and an indoor planning map of the building is extracted; S202: acquiring sensor set installation positions according to the indoor planning map, generating a sensor set position matrix according to the sensor set installation positions, and numbering the sensor sets according to the sensor set position matrix; S203: Associating the material information with the sensor set installation location, and performing a first fire hazard level assessment on the material according to the material information. The specific calculation formula is as follows; RI = α × (MG ÷ AR); Wherein RI represents the first fire hazard level of the material, α represents the combustibility coefficient of the material, and the combustibility coefficient of the material is preset according to the chemical material of the material; MG represents the weight of the material; AR represents the area occupied by the material; S204: according to the first fire hazard degree of the material, different impact distances are set for the material, and a circular area is divided for each material as an impact hazard area with the geometric center of the material's floor area as the center and the impact distance as the radius; the area of ​​the overlapping area with the impact hazard area of ​​the central material is extracted, and the first fire hazard degree of the surrounding materials belonging to the overlapping area of ​​the impact hazard area of ​​the central material is extracted; the second fire hazard degree of the central material is calculated according to the first fire hazard degree and the area of ​​the overlapping area of ​​the impact hazard area. The specific calculation formula is as follows: ; RII represents the second fire hazard level of the central materials, S i Represents the area of ​​the overlapping area of ​​the ith potential hazard zone, RI i It represents the first fire hazard level of surrounding materials in the overlapping area of ​​the i-th impact hazard zone, and RI represents the first fire hazard level of central materials; i∈[1,n], where n represents the total number of overlapping areas of the impact hazard zone, S0 represents the area of ​​the impact hazard zone of the central material, and RI0 represents the first fire hazard level of the central material; In S3, the real-time collected image is grayed, and the flame contour is extracted from the grayed real-time collected image using a contour extraction algorithm to obtain the gray value of each pixel in the real-time flame contour image; Substituting the gray value of each pixel in the real-time flame outline image into the gray value and radiation energy density fitting model, the radiation energy density of each pixel in the real-time flame outline image is calculated; Perform spectral analysis on all pixels of the real-time flame outline image to obtain the wavelength range of each pixel, sum the radiation energy density of all wavelength ranges of each pixel, and then sum the radiation energy of all pixels to obtain the total radiation energy of the real-time flame outline image; The hazard assessment of the real-time flame contour image is performed, and the specific calculation formula is as follows: FAT=A1×RII+A2×TRE+A3×(B1×T×L); Among them, FAT represents the risk score of the real-time flame profile image, TRE represents the total radiation energy of the real-time flame profile image, T represents the real-time indoor temperature, L represents the distance from the flame position to the sprinkler port, A1, A2 and A3 represent different risk coefficients, B1 represents the indoor temperature conversion coefficient, and A1, A2, A3 and B1 are all system preset constants; The method for calculating the distance from the flame position to the spray port comprises: using image recognition technology to identify the pixel positions of the preset calibration object and the flame on the real-time collected image, and calculating the actual distance vector from the flame to the camera through the internal and external parameters of the camera; the position of the calibration object and the camera is fixed; and calculating the distance from the flame position to the spray port according to the actual distance vector from the flame to the camera and the distance vector from the camera to the spray port; Determine whether the hazard score of the real-time flame is greater than the hazard score threshold. If it is greater than, the real-time flame will be taken as a key monitoring target. If it is less than or equal to, no processing will be done.

2. The IoT intelligent analysis and early warning method based on edge computing gateway according to claim 1 is characterized in that: In S4, the total number of key monitoring targets is counted, and the total hazard score of all key monitoring targets is calculated; a total hazard score threshold is set for the total hazard score. If the total hazard score is greater than the total hazard score threshold, a first-level warning is issued to the administrator, and the specific distribution location and material information of the key monitoring targets are sent; according to the spatial position relationship between the sprinkler port and the key monitoring target, the sprinkler port is adaptively controlled and managed; If the total risk score is less than or equal to the total risk score threshold, a second-level warning will be issued to the administrator; According to the spatial position relationship between the sprinkler opening and the key monitoring target, the sprinkler opening is controlled and managed adaptively; The emergency situation of the first-level warning is higher than that of the second-level warning; The control and management of the spray port includes: Obtain the spatial positions of all key monitoring targets and all sprinkler openings; According to the spatial position of the key monitoring target and the spatial position of the spray outlet, the spray outlet closest to the key monitoring target is selected as the target spray outlet; Calculate the total hazard score for the key monitoring targets of the same sprinkler outlet, and extract the distance from the key monitoring target farthest from the sprinkler outlet to the sprinkler outlet; The total amount of water sprayed and the injection pressure of the sprinkler outlet are controlled according to the total distribution of the hazard scores of each sprinkler outlet and the distance from the key monitoring target farthest from the sprinkler outlet to the sprinkler outlet; the total amount of water sprayed and the injection pressure are system preset values.

3. An IoT intelligent analysis and early warning system based on edge computing gateway, wherein the system is applied to the IoT intelligent analysis and early warning method based on edge computing gateway according to any one of claims 1-2, characterized in that: The system includes a fitting model building module, a fire hazard analysis module, a data acquisition module, a hazard analysis module and an early warning control module; The fitting model building module is used to perform image recognition on the test flame image using image recognition technology, obtain feature information of the flame image, and build a fitting model according to the feature information of the flame image; The fire hazard analysis module is used to obtain the sensor set installation location according to the indoor planning map, number the sensor set according to the sensor set location matrix, associate the material information with the sensor set installation location, and analyze the fire hazard degree according to the material information; The data acquisition module is used to collect data of the area to be tested in real time through sensors and cameras, transmit the data of the area to be tested to the edge computing gateway, obtain the location distribution of the area to be tested according to the sensor number, and extract the material information of the area to be tested according to the location distribution; The risk analysis module is used in the edge computing gateway to use image recognition technology to perform image recognition on the data of the area to be tested, extract the data features of the area to be tested, and perform risk analysis based on the data features of the area to be tested; The early warning control module warns the administrator and performs adaptive control on the fire extinguishing equipment based on the risk analysis results of the data characteristics of the area to be tested.

4. The IoT intelligent analysis and early warning system based on edge computing gateway according to claim 3 is characterized in that: The fitting model building module includes a flame characteristic analysis unit and a fitting model generation unit; The flame feature analysis unit is used to grayscale the test flame image, extract the flame contour of the grayscale test flame image using a contour extraction algorithm, and obtain the grayscale value and radiation energy density of each pixel in the test flame contour image; The fitting model generation unit is used to construct a grayscale radiation energy density set, establish a grayscale and radiation energy density fitting model according to the grayscale radiation energy density set, and solve the fitting coefficient.

5. The IoT intelligent analysis and early warning system based on edge computing gateway according to claim 3 is characterized in that: The fire hazard analysis module includes an information integration unit and a fire hazard degree analysis unit; The information integration unit is used to group different types of sensors within the same preset range in the building into a set, and the installation positions of different types of sensors in the same sensor set are consistent; extract the indoor planning map of the building; obtain the installation position of the sensor set according to the indoor planning map, generate a sensor set position matrix according to the installation position of the sensor set, number the sensor set according to the sensor set position matrix, and associate the material information with the installation position of the sensor set; The fire hazard level analysis unit evaluates the first fire hazard level of the materials according to the material information, sets different impact distances for the materials according to the first fire hazard level of the materials, and calculates the second fire hazard level of the central materials according to the area of ​​the overlapping area between the first fire hazard level and the impact hazard zone.

6. The IoT intelligent analysis and early warning system based on edge computing gateway according to claim 3 is characterized in that: The risk analysis module includes a risk characteristic calculation unit and a monitoring target processing unit; The hazard characteristic calculation unit is used to grayscale the real-time collected image, extract the flame contour of the grayscaled real-time collected image using a contour extraction algorithm, and obtain the grayscale value of each pixel in the real-time flame contour image; substitute the grayscale value into the grayscale and radiation energy density fitting model to calculate the radiation energy density, and calculate the total radiation energy according to the radiation energy density; Calculate the distance from the flame position to the sprinkler opening; The monitoring target processing unit is used to determine whether the danger score of the real-time flame is greater than the danger score threshold. If it is greater, the real-time flame is taken as a key monitoring target; if it is less than or equal to the danger score threshold, no processing is performed.

7. The IoT intelligent analysis and early warning system based on edge computing gateway according to claim 3 is characterized in that: The early warning control module includes an early warning unit and a control unit; The early warning unit is used to count the total number of key monitoring targets and calculate the total value of the risk scores of all key monitoring targets; set a total value threshold of the risk score for the total value of the risk score; if the total value of the risk score is greater than the total value threshold of the risk score, a first-level early warning is issued to the administrator; if the total value of the risk score is less than or equal to the total value threshold of the risk score, a second-level early warning is issued to the administrator; The control unit is used to adaptively control and manage the spray port according to the spatial position relationship between the spray port and the key monitoring target.