A fire intelligent early warning method and system

By combining intelligent fire warning methods with video and IoT sensing devices, the problem of strong passivity of traditional fire warning technology is solved, early identification and escape channel planning of fires are realized, and fire response speed and safety are improved.

CN118172884BActive Publication Date: 2024-12-06WUHAN CHUYU TECH CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202410303487.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-18
Publication Date
2024-12-06
Estimated Expiration
2044-03-18

AI Technical Summary

Technical Problem

Traditional fire early warning technology is highly passive and cannot detect fires in time in the early stage. The operating conditions of firefighting facilities are difficult to grasp in a timely manner, resulting in unknown locations of the fire, causing difficulties in rescue and escape.

Method used

Using intelligent fire warning methods combining video and IoT sensing equipment, we can obtain on-site monitoring information in different areas of the building and the operating status of fire-fighting facilities, judge the severity of the fire, and plan escape channels to the management personnel and fire department. Issue a warning message.

Benefits of technology

It realizes early identification and remote monitoring of fires, improves fire recognition capabilities and response speed, ensures safe escape of personnel and effective operation of firefighting facilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118172884B_ABST
    Figure CN118172884B_ABST
Patent Text Reader

Abstract

The present invention proposes a fire intelligent early warning method and system, which belongs to the field of fire early warning technology; it comprises the following steps: S1: obtaining on-site monitoring information of different areas of a building and the operating status of fire-fighting facilities; S2: after obtaining the on-site monitoring information and the operating status of the fire-fighting facilities, the local data center determines whether there is a fire in the building, if there is a fire, executes step S3, and continues to execute step S2; S3: the local data center determines the severity of the fire, predicts the spread trend of the fire, generates warning information, and plans the escape route inside the building; S4: configuring a remote service unit, the remote service unit is connected to the local data center in communication; the remote service unit further sends a warning message to the management personnel, residents or emergency fire department of the building according to the warning information sent by the local data center, prompting emergency evacuation or disaster relief.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

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

[0002] From the analysis of the causes of fires, fires in rural areas, commercial places and residential places accounted for 245,000, 9,200 and 101,000 respectively. The main cause of fire was electrical fire, such as aging of lines, overload of electricity, unauthorized wiring or illegal charging. The population in cities is relatively concentrated, and the number of electric bicycles parked and charged has increased. In addition, the abnormally high temperature climate continues, and the electricity load continues to increase, resulting in a continuous increase in the risk of fire. The main management responsibility of the disaster is that the management party has a weak sense of responsibility, lacks awareness of safety work, does not strictly implement the safety management system, and does not conduct timely inspections of key fire safety areas, resulting in hidden dangers being buried, which ultimately causes irreparable losses.

[0003] Traditional fire warning methods are relatively passive and cannot detect fires in the early stages. Firefighters passively wait for people to call the police, which delays precious firefighting opportunities. Moreover, the operating status of firefighting facilities cannot be timely and reliably grasped. There is no reliable overall perception method, which leads to the unknown specific floor and location of the fire, which poses a great challenge to rescuers and escapees. Therefore, it is very necessary to provide a fire intelligent warning method and system that can inspect the firefighting facilities of buildings, conduct real-time fire monitoring of different parts of buildings, and guide people to enter and exit the fire scene safely. Summary of the invention

[0004] In view of this, the present invention proposes a fire intelligent warning method and system that combines video or IoT sensing equipment to centrally analyze or warn of building fire conditions, judge the fire conditions and issue warnings in a timely manner.

[0005] The technical solution of the present invention is achieved in this way:

[0006] In one aspect, the present invention provides a fire intelligent early warning method, comprising the following steps:

[0007] S1: Obtain on-site monitoring information of different areas of the building and the operating status of fire-fighting facilities;

[0008] S2: After obtaining the on-site monitoring information and the operating status of the fire-fighting facilities, the local data center also determines whether there is a fire in the building based on the on-site monitoring information. If there is a fire, step S3 is executed. If there is no fire, data from the local fire monitoring unit and the fire-fighting unit are continuously received;

[0009] S3: The local data center determines the severity of the fire, predicts the spread of the fire, generates warning information, and plans escape routes inside the building;

[0010] S4: Configure a remote service unit, which is connected to the local data center in communication; based on the warning information sent by the local data center, the remote service unit further sends warning information to the building's management personnel, residents or emergency fire department, prompting emergency evacuation or disaster relief.

[0011] On the basis of the above technical scheme, preferably, the obtaining of on-site monitoring information of different areas of the building described in step S1 is to configure local fire monitoring units in different areas of the building, the local fire monitoring units include an RGB camera and a number of detectors, the RGB camera and the number of detectors are arranged inside the distribution box, at the interface of the energy-consuming equipment or indoors, the RGB camera is used to obtain RGB video images inside the distribution box, at the interface or indoors; the number of detectors are used to obtain ambient temperature signals, smoke concentration signals and current signals of energy-consuming equipment; the RGB video images, ambient temperature signals, smoke concentration signals and current signals of energy-consuming equipment together constitute the on-site monitoring information, and the local fire monitoring unit sends the on-site monitoring information to the local data center.

[0012] Preferably, the operating status of the fire-fighting facilities is obtained by configuring fire-fighting units in different areas of the building; the fire-fighting units include a cabinet, fire-fighting equipment, an active fire-fighting device and a number of sensors, the cabinet is arranged on each floor of the building, and the cabinet is equipped with fire-fighting equipment and a number of sensors, the fire-fighting equipment is used for manual fire-fighting; the active fire-fighting device is used for active fire-fighting; a number of sensors are used to monitor the operating status of the fire-fighting equipment, and send the operating status of the fire-fighting equipment to the local data center.

[0013] Further preferably, the local data center described in step S2 further determines whether there is a fire in the building based on the on-site monitoring information by executing the following contents:

[0014] 1) When the ambient temperature signal, smoke concentration signal or current signal of the energy-consuming device obtained by several detectors exceeds the set threshold and maintains for a first set time, the local data center selects the first RGB video image at a certain moment before the first set time and the second RGB video image within a period of time after the first set time;

[0015] 2) The local data center selects a framed image of the selected first RGB video image at random and caches it, and frames the second RGB video image, and selects a framed image every certain number of frames and caches it;

[0016] 3) Converting each cached frame image from RGB space into a grayscale image, normalizing, gamma correcting and denormalizing the pixel values ​​of each frame image to obtain each gamma-corrected frame image;

[0017] 4) Using the gamma-corrected framed images corresponding to the first RGB video image as the background image, and using the gamma-corrected framed images corresponding to the second RGB video image as the images to be screened; performing Boolean operations on the images to be screened and the background image one by one, retaining the parts of the images to be screened that are different from the background image, removing discontinuous discrete areas, and using the remaining parts of the images to be screened as a number of potential flame areas, so that the potential flame areas have the same starting boundary, and arranging the potential flame areas in the time sequence of the frames;

[0018] 5) Use polygons to fit the boundary points of several potential flame areas and calculate the average change rate of the area of ​​the potential flame areas corresponding to adjacent frames S i , S i-1 and ΔS i Indicates the area and area change of the potential flame area in two adjacent frame images of the second RGB video image, i=2, 3, ..., N, and the time interval between adjacent frame images is T0; determines whether the average change rate of the area of ​​the potential flame area exceeds the set change rate threshold of 0.05, if yes, it indicates that there is a flame area, if no, it indicates that there is no flame;

[0019] 6) The duration of the selected second RGB video image does not exceed the duration of the first set time;

[0020] 7) The time interval T0 between adjacent frame images is [0.1, 0.12] ± 0.015 seconds.

[0021] Preferably, the local data center described in step S3 determines the severity of the fire, predicts the spread trend of the fire, and generates a warning message. After confirming the existence of the flame area, the fire is divided into the following severity levels according to the color and diffusion speed of the smoke in the indoor RGB video image: 1) If the interval of the smoke RGB value is R∈[237,255], G∈[207,255], B∈[0,87], R∈[0,127], G∈[201,255], B∈[0,60], R∈[0,41], G∈[0,36], B∈[0,33], or the current diffusion speed of the smoke exceeds 0.8m / s, it is considered to be severity level one; 2) If the interval of the RGB value corresponding to the smoke is 1) If the RGB value of the smoke is in the range of R∈[112, 192], G∈[128, 192], B∈[105, 192], or the current diffusion speed of the smoke does not exceed 0.3 m / s, then the severity level is determined to be level 3; severity level 1 > severity level 2 > severity level 3; 4) The local data center generates a warning message based on the determined severity level of the fire and sends it to the building’s management personnel, residents and fire department through short messages, AI voice calls or forced activation of a mobile phone APP for screen flashing warnings.

[0022] More preferably, the diffusion speed of the smoke is obtained by the local data center by obtaining the RGB video image of the room closest to the flame area, taking the wall corner line as a reference, drawing a length reference line extending to the boundary of the RGB video image that coincides with the reference, converting the horizontal length of the area corresponding to the RGB video image, and obtaining the video duration T1 corresponding to the diffusion of the smoke to the entire RGB video image range, and obtaining the diffusion speed of the smoke: Where L is the pixel length of the length reference line; α is the inclination angle between the RGB camera for acquiring the indoor RGB video image and the ground; β is the deflection angle of the length reference line relative to the vertical direction; λ is the distortion coefficient of the indoor RGB video image; η is the scaling ratio of the pixel size of the indoor RGB video image to the actual size.

[0023] Further preferably, the planning of the escape route inside the building described in step S3 includes the following contents: 1) using a chain sequence numbering method to sequentially number the monitoring areas of the RGB cameras on each floor of the building, and the numbering order of the monitoring areas on each floor is in a clockwise direction from the inside to the outside; 2) when the local data center confirms the existence of a flame area, determine the number of the monitoring area corresponding to the flame demand, and mark the fire spread area on the same floor or adjacent floors within 5 minutes, 10 minutes and 20 minutes according to the smoke spread direction in the fire severity level and the flame spread direction obtained by the second RGB video image; 3) turn on the active fire extinguishing device of the fire extinguishing unit in the fire spread area to reduce the spread rate of the fire; 4) according to the location of the flame area, plan the evacuation route in the order of near first and far later on the same floor, and high first and low later on different floors, and the planned evacuation route is away from the power shaft, ventilation shaft or elevator room near the fire spread area.

[0024] Preferably, a card reader is provided in the cabinet within the fire extinguishing unit, and a recognizable graphic code is provided on the fire-fighting equipment. When the fire-fighting equipment is placed in or taken out of the cabinet, it will be read by the card reader and the equipment status in the fire-fighting unit will be updated, and the new operating status of the fire-fighting equipment will be generated and sent to the local data center.

[0025] Further preferably, the local data center is also provided with a duty supervision unit; the building management personnel regularly inspect the local fire monitoring unit and the fire extinguishing unit, return the fire-fighting equipment to its place or replenish the quantity, and swipe the card to sign in at the card reader, and the card reader generates sign-in information and updated operating status information of the fire-fighting equipment and sends it to the duty supervision unit of the local data center; the duty supervision unit also supervises the on-duty status of the staff of the local data center, and sends non-fire alarm information to the on-duty staff and their superiors in the event of absence or absence from duty for a certain period of time.

[0026] On the other hand, the present invention provides a fire intelligent early warning system for implementing the above-mentioned fire intelligent early warning method, comprising:

[0027] Local fire monitoring units are deployed in different areas of each floor of the building to obtain on-site monitoring information from different areas of the building;

[0028] Fire extinguishing units are located in different areas on each floor of the building and are used to obtain the operating status of fire extinguishing facilities;

[0029] The local data center is set up near the building. It is used to obtain on-site monitoring information and the operating status of fire-fighting facilities. It also determines whether there is a fire in the building based on the on-site monitoring information. If there is a fire, the local data center determines the severity of the fire, predicts the spread trend of the fire, generates warning information, and plans escape routes inside the building;

[0030] The remote service unit is connected to the local data center for receiving warning information from the local data center to prompt emergency evacuation or disaster relief.

[0031] The present invention provides a fire intelligent early warning method and system, which has the following advantages over the prior art:

[0032] Beneficial effects:

[0033] (1) This application combines the local fire monitoring unit and fire extinguishing unit related to the Internet of Things to identify the fire point and smoke diffusion direction in the defense area, thereby determining the fire location, diffusion speed and predicting potential fire areas, and can achieve remote monitoring and network control, improving fire identification capabilities and response speed;

[0034] (2) Based on the location of the fire, the speed of smoke diffusion and the color of the smoke, determine the possible type of fire, the degree of harm caused by the smoke to the human body, and comprehensively judge the severity of the fire. Provide a plan for the corresponding escape route within the corresponding emergency escape time and provide it to the corresponding personnel to facilitate rescue and escape. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1 A method flow chart of a fire intelligent early warning method and system of the present invention;

[0037] Figure 2 A schematic diagram of the numbering order of monitoring areas on each floor of a fire intelligent early warning method and system of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with 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.

[0039] like Figure 1 As shown, on the one hand, the present invention provides a fire intelligent early warning method, comprising the following steps:

[0040] S1: Obtain on-site monitoring information of different areas of the building and the operating status of fire-fighting facilities.

[0041] Among them, obtaining on-site monitoring information of different areas of the building is to configure local fire monitoring units in different areas of the building. The local fire monitoring units include RGB cameras and several detectors. The RGB cameras and several detectors are set inside the distribution box, at the interface of the energy-consuming equipment or indoors. The RGB camera is used to obtain RGB video images inside the distribution box, at the interface or indoors; several detectors are used to obtain ambient temperature signals, smoke concentration signals and current signals of energy-consuming equipment; RGB video images, ambient temperature signals, smoke concentration signals and current signals of energy-consuming equipment together constitute on-site monitoring information, and the local fire monitoring unit sends the on-site monitoring information to the local data center. The local fire monitoring unit is a medium- and short-distance video acquisition device based on local Internet of Things technology.

[0042] To obtain the operating status of fire-fighting facilities, fire-fighting units are configured in different areas of the building; the fire-fighting units include cabinets, fire-fighting equipment, active fire-fighting devices and several sensors. The cabinets are set up on each floor of the building, and fire-fighting equipment and several sensors are configured in the cabinets. The fire-fighting equipment is used for manual fire-fighting; the active fire-fighting device is used for active fire-fighting; several sensors are used to monitor the operating status of fire-fighting equipment and send the operating status of fire-fighting equipment to the local data center. The fire-fighting unit is an interactive fusion of IoT sensing devices with sensing devices and traditional fire-fighting equipment. Fire-fighting equipment can be one or more of fire extinguishers, fire hydrants, water hoses and fire nozzles. Several sensors can obtain whether each fire-fighting equipment is placed in the corresponding initial position, or determine whether each fire-fighting equipment has been used based on the current weight.

[0043] As a preferred embodiment, a card reader is provided in the cabinet of the fire-fighting unit, and a recognizable graphic code, such as a barcode or a QR code, is provided on the fire-fighting equipment. When the fire-fighting equipment is placed in or taken out of the cabinet, it will be read by the card reader and the equipment status in the fire-fighting unit will be updated, and the new operating status of the fire-fighting equipment will be generated and sent to the local data center. The card reader can be an RFID reader / writer. If the fire-fighting equipment is taken out or returned without being placed in the initial position, the card reader will generate a new operating status of the fire-fighting equipment based on the detection results of several sensors and send it to the local data center, indicating that the fire-fighting equipment has been used or has not been put back to its original position, and timely intervention is required.

[0044] A local data center with an integrated IoT gateway device is installed outside the building. Using IoT for data transmission does not rely on the operator network and its base stations, making data transmission more secure and reliable, and network expansion is also convenient.

[0045] S2: After obtaining the on-site monitoring information and the operating status of the fire-fighting facilities, the local data center also determines whether there is a fire in the building based on the on-site monitoring information. If there is a fire, execute step S3. If there is no fire, continue to receive data from the local fire monitoring unit and the fire-fighting unit.

[0046] Step S2 specifically performs the following:

[0047] 1) When the ambient temperature signal, smoke concentration signal or current signal of the energy-consuming equipment obtained by several detectors exceeds the set threshold and lasts for the first set time, the local data center selects the first RGB video image at a certain moment before the first set time and the second RGB video image within a period of time after the first set time; if the electrical equipment has an overcurrent, causing the cable or contact to continue to heat up, a short circuit or spontaneous combustion may occur. Combined with the sensor detection signals obtained by the IoT terminal that obtains the ambient temperature and the IoT terminal that obtains the smoke concentration, if one or more of the above-detected current signals, ambient temperature signals or smoke concentration signals exceeds the set threshold and lasts for the first set time, it indicates that there is a potential fault and needs to be checked. Specifically, a frame-by-frame comparative analysis is performed based on the state of the first RGB video image at a certain moment before the first set time and several second RGB video images within a time interval at one end after the first set time.

[0048] 2) The local data center selects a framed image of the selected first RGB video image at random and caches it, and frames the second RGB video image. Every certain number of frames, a framed image is selected and cached. The flame image does not remain unchanged, but jumps regularly, and the jumping frequency usually does not exceed 20Hz, so the second RGB video image can be framed and cached according to integer multiples of the jumping frequency.

[0049] 3) Convert each cached sub-frame image from the RGB space into a grayscale image, normalize, gamma correct and denormalize the pixel value of each sub-frame image, and obtain each gamma-corrected sub-frame image.

[0050] The transformation process is to correct the background and reduce the influence of the ambient brightness on the frame images corresponding to the first RGB video image and the second RGB video image.

[0051] First, after converting each frame image into a grayscale image, some area pixels are too bright, which affects the judgment of whether it is a potential flame area. By normalizing each pixel value x of the grayscale image, the pixel value of each current frame image is normalized to a real number x0 between [0, 1]. where x max is the maximum grayscale value of the current frame image, x min is the minimum grayscale value of the current frame image; gamma correction is performed on the result x0 after the pixel value normalization to obtain the corrected result x′0, x′0=(x0+e) γ , where e is the compensation amount and γ is the gamma value; the compensation amount is inversely proportional to the exposure of the RGB camera. If the exposure is set to the default level, the compensation amount is 0, if the exposure is set to +1 level, the compensation is -0.1, and so on; the gamma value range is [2.5, 4]. The corrected result is denormalized to obtain the denormalized grayscale value x′, x′=x′0x max -x min (x′0-1), each frame image after gamma correction can be obtained, reducing the influence of exposure and ambient light on the frame image.

[0052] 4) The gamma-corrected framed images corresponding to the first RGB video image are used as the background image, and the gamma-corrected framed images corresponding to the second RGB video image are used as the images to be screened; Boolean operations are performed on the images to be screened and the background image one by one, and the parts of the images to be screened that are different from the background image are retained. After removing the discontinuous discrete areas, the remaining parts of the images to be screened are used as several potential flame areas, so that the several potential flame areas have the same starting boundary, and the several potential flame areas are arranged in the time sequence of the frames; the Boolean operation here is Boolean subtraction, that is, the content that is the same as the background image is removed from the images to be screened, and the different content is retained as the images to be screened.

[0053] The principle of eliminating discrete areas is to meet any of the following conditions: A) determine that the outline of the same discrete area in each image to be screened does not change; B) determine whether the distance between the same discrete area and the potential flame area in each image to be screened increases with the time sequence of the frame division.

[0054] 5) Use polygons to fit the boundary points of several potential flame areas and calculate the average change rate of the area of ​​the potential flame areas corresponding to adjacent frames S i , S i-1 and ΔS i Indicates the area and area change of the potential flame area in two adjacent frame images of the second RGB video image, i = 2, 3, ..., N, N is not less than 40; the time interval between adjacent frame images is T0; judge whether the average change rate of the area of ​​the potential flame area exceeds the set change rate threshold of 0.05, if yes, it means there is a flame area, if not, it means there is no flame. If it is a stroboscopic flame, the edge of the flame may be irregular, with obvious concave and convex feeling and intermittent fluctuations, and the area of ​​the potential flame area is increasing, then it is determined that there is a flame and the boundary of the flame is gradually expanding.

[0055] In order to obtain enough sub-frame images, the duration of the selected second RGB video image does not exceed the duration of the first set time. As a preferred embodiment, the time interval T0 between adjacent sub-frame images is [0.1, 0.12] ± 0.015 seconds.

[0056] S3: The local data center determines the severity of the fire, predicts the spread of the fire, generates warning information, and plans escape routes inside the building.

[0057] The local data center predicts the spread trend of the fire and generates warning information. After confirming the existence of the flame area, the fire is divided into the following severity levels based on the color and diffusion speed of the smoke in the indoor RGB video image:

[0058] 1) If the smoke RGB value is in the interval of R∈[237, 255], G∈[207, 255], B∈[0, 87], R∈[0, 127], G∈[201, 255], B∈[0, 60], R∈[0, 41], G∈[0, 36], B∈[0, 33], or the current smoke diffusion speed exceeds 0.8m / s, it is considered to be severity level 1; this situation corresponds to a fast smoke diffusion speed and the most toxic smoke types, such as black smoke, yellow-green smoke and yellow smoke.

[0059] 2) If the RGB value corresponding to the smoke is in the interval of R∈[225, 255], G∈[225, 255], B∈[205, 255], or the current diffusion speed of the smoke is in the interval of 0.3-0.8m / s, then the severity level is determined to be level 2; the corresponding smoke color in this case is white smoke.

[0060] 3) If the RGB value corresponding to the smoke is in the interval R∈[112, 192], G∈[128, 192], B∈[105, 192], or the current diffusion speed of the smoke does not exceed 0.3m / s, then the severity level is determined to be level 3; the corresponding smoke in this case is gray smoke. Severity level 1> Severity level 2> Severity level 3.

[0061] 4) The local data center generates a warning message based on the determined severity level of the fire and sends it to the remote service unit.

[0062] The diffusion speed of the smoke is obtained by the local data center by obtaining the RGB video image of the room closest to the flame area, taking the corner line as a reference, drawing a length reference line extending to the boundary of the RGB video image that coincides with the reference, converting the horizontal length of the area corresponding to the RGB video image, and obtaining the video duration T1 corresponding to the diffusion of smoke to the entire RGB video image range, and obtaining the diffusion speed of the smoke: Wherein L is the pixel length of the length reference line; α is the inclination angle between the RGB camera for acquiring the indoor RGB video image and the ground; β is the deflection angle of the length reference line relative to the vertical direction; λ is the distortion coefficient of the indoor RGB video image, which is related to the lens type of the RGB camera and is preferably a sine trigonometric function; η is the scaling ratio of the pixel size of the indoor RGB video image to the actual size.

[0063] Further preferably, the planning of the escape route inside the building described in step S3 includes the following contents: 1) using a chain sequence numbering method to sequentially number the monitoring areas of the RGB cameras on each floor of the building, and the numbering order of the monitoring areas on each floor is in a clockwise direction from the inside to the outside; 2) when the local data center confirms the existence of a flame area, determine the number of the monitoring area corresponding to the flame demand, and mark the fire spread area on the same floor or adjacent floors within 5 minutes, 10 minutes and 20 minutes according to the smoke spread direction in the fire severity level and the flame spread direction obtained by the second RGB video image; 3) turn on the active fire extinguishing device of the fire extinguishing unit in the fire spread area to reduce the spread rate of the fire; 4) according to the location of the flame area, plan the evacuation route in the order of near first and far later on the same floor, and high first and low later on different floors, and the planned evacuation route is away from the power shaft, ventilation shaft or elevator room near the fire spread area.

[0064] by Figure 2For example, the building has N floors, and the Mth floor is divided into 9 areas from the inside to the outside in a clockwise direction, M<N, respectively M1, M2, M3, ..., M9. Each area corresponds to one or more local fire monitoring units and fire extinguishing units. If the M2 area of ​​the M floor is the initial fire point, and the fire is expected to spread in the direction of M2, M9, and M8, then M9 / M3 is the first priority on the same floor, M8 / M4 is the second priority on the same floor, M7 / M5 is the third priority on the same floor, and M6 is the fourth priority on the same floor. Evacuation is carried out in an orderly manner. During evacuation, only stairs can be used, and the running elevator will stop at the nearest staircase and stop.

[0065] For different floors, high-rise first means that residents who are several floors below the top floor, such as N, N-1, N-2, ..., N-F+1, and residents who are several floors above the floor where the fire started, such as M+1, M+2, ..., M+F, are evacuated first, ensuring that the evacuation passage is used by high-rise residents first for a period of time, and then there is a floor difference between the two using the stairs, and there will be no congestion; low-rise later means that after a batch of high-rise residents are evacuated, the second set time is interval to remind residents who are several floors below the floor where the fire started, such as M-1, M-2, ..., M-F+1, and residents who are several floors above the first floor of the building to evacuate next, such as the first floor, the second floor, ..., the Fth floor; and then after the second set time, the residents of the floors above the fire started to evacuate, and so on. The second set time shall not exceed 2 minutes.

[0066] S4: Configure a remote service unit, which is connected to the local data center in communication; based on the warning information sent by the local data center, the remote service unit further sends warning information to the building's management personnel, residents or emergency fire department, prompting emergency evacuation or disaster relief.

[0067] The remote service unit further sends warning information including escape routes to the building's managers, residents and fire department via short messages, AI voice calls or forced activation of a mobile phone APP for screen flashing warnings.

[0068] In order to monitor the operation of the data center, the local data center also has a duty supervision unit; the information obtained by the duty supervision unit is directly reflected in the remote service unit. The building management personnel regularly inspect the local fire monitoring unit and fire extinguishing unit, return the fire equipment to its place or replenish the number, and swipe the card to sign in at the card reader. The card reader generates sign-in information and updated fire equipment operation status information and sends it to the duty supervision unit of the local data center; the duty supervision unit also monitors the on-duty status of the staff of the local data center, and sends non-fire alarm information to the on-duty staff and their superiors for the phenomenon of absence or absence from work for a certain period of time, prompting the staff to return to their posts as soon as possible.

[0069] On the other hand, the present invention provides a fire intelligent early warning system for implementing the above-mentioned fire intelligent early warning method, comprising:

[0070] Local fire monitoring units are deployed in different areas of each floor of the building to obtain on-site monitoring information from different areas of the building;

[0071] Fire extinguishing units are located in different areas on each floor of the building and are used to obtain the operating status of fire extinguishing facilities;

[0072] The local data center is set up near the building. It is used to obtain on-site monitoring information and the operating status of fire-fighting facilities. It also determines whether there is a fire in the building based on the on-site monitoring information. If there is a fire, the local data center determines the severity of the fire, predicts the spread trend of the fire, generates warning information, and plans escape routes inside the building;

[0073] The remote service unit is connected to the local data center for receiving warning information from the local data center to prompt emergency evacuation or disaster relief.

[0074] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A fire intelligent early warning method, characterized in that: The steps include: S1: Obtain on-site monitoring information of different areas of the building and the operating status of fire-fighting facilities; S2: After obtaining the on-site monitoring information and the operating status of the fire-fighting facilities, the local data center also determines whether there is a fire in the building based on the on-site monitoring information. If there is a fire, step S3 is executed. If there is no fire, data from the local fire monitoring unit and the fire-fighting unit are continuously received; S3: The local data center determines the severity of the fire, predicts the spread of the fire, generates warning information, and plans escape routes inside the building; S4: configuring a remote service unit, which is connected to the local data center in communication; the remote service unit further sends warning information to the building's management personnel, residents or emergency fire department based on the warning information sent by the local data center, prompting emergency evacuation or disaster relief; The acquisition of on-site monitoring information of different areas of the building described in step S1 is to configure local fire monitoring units in different areas of the building, the local fire monitoring units include RGB cameras and several detectors, the RGB cameras and several detectors are arranged inside the distribution box, at the interface of the energy-consuming equipment or indoors, the RGB camera is used to obtain RGB video images inside the distribution box, at the interface or indoors; the several detectors are used to obtain ambient temperature signals, smoke concentration signals and current signals of energy-consuming equipment; the RGB video images, ambient temperature signals, smoke concentration signals and current signals of energy-consuming equipment together constitute the on-site monitoring information, and the local fire monitoring unit sends the on-site monitoring information to the local data center; The operation status of the fire-fighting facilities is obtained by configuring fire-fighting units in different areas of the building; the fire-fighting units include a cabinet, fire-fighting equipment, an active fire-fighting device and a number of sensors; the cabinet is arranged on each floor of the building, and the cabinet is equipped with fire-fighting equipment and a number of sensors; the fire-fighting equipment is used for manual fire-fighting; the active fire-fighting device is used for active fire-fighting; the several sensors are used to monitor the operation status of the fire-fighting equipment, and send the operation status of the fire-fighting equipment to the local data center; The local data center described in step S2 also determines whether there is a fire in the building based on the on-site monitoring information, and executes the following contents: 1) When the ambient temperature signal, smoke concentration signal or current signal of the energy-consuming device obtained by several detectors exceeds the set threshold and maintains for a first set time, the local data center selects the first RGB video image at a certain moment before the first set time and the second RGB video image within a period of time after the first set time; 2) The local data center selects a framed image of the selected first RGB video image at random and caches it, and frames the second RGB video image, and selects a framed image every certain number of frames and caches it; 3) Converting each cached frame image from RGB space into a grayscale image, normalizing, gamma correcting and denormalizing the pixel values ​​of each frame image to obtain each gamma-corrected frame image; 4) Using the gamma-corrected framed images corresponding to the first RGB video image as the background image, and using the gamma-corrected framed images corresponding to the second RGB video image as the images to be screened; performing Boolean operations on the images to be screened and the background image one by one, retaining the parts of the images to be screened that are different from the background image, removing discontinuous discrete areas, and using the remaining parts of the images to be screened as a number of potential flame areas, so that the potential flame areas have the same starting boundary, and arranging the potential flame areas in the time sequence of the frames; 5) Use polygons to fit the boundary points of several potential flame areas and calculate the average change rate of the area of ​​the potential flame areas corresponding to adjacent frames S i , S i-1 and ΔS i Indicates the area and area change of the potential flame area in two adjacent frame images of the second RGB video image, i=2, 3, ..., N, and the time interval between adjacent frame images is T0; determines whether the average change rate of the area of ​​the potential flame area exceeds the set change rate threshold of 0.05, if yes, it indicates that there is a flame area, if no, it indicates that there is no flame; 6) The duration of the selected second RGB video image does not exceed the duration of the first set time; 7) The time interval T0 between adjacent frame images is [0.1, 0.12] ± 0.015 seconds; The local data center described in step S3 determines the severity of the fire, predicts the spread trend of the fire, and generates warning information. After confirming the existence of the flame area, the fire is divided into the following severity levels according to the color and diffusion speed of the smoke in the indoor RGB video image: 1) If the interval of the smoke RGB value is R∈[237,255], G∈[207,255], B∈[0,87], R∈[0,127], G∈[201,255], B∈[0,60], R∈[0,41], G∈[0,36], B∈[0,33], or the current smoke diffusion speed exceeds 0.8m / s, it is considered to be severity level 1; 2) If the RGB value corresponding to the smoke is in the interval of R∈[225, 255], G∈[225, 255], B∈[205, 255], or the current diffusion speed of the smoke is in the range of 0.3-0.8m / s, then the severity level is determined to be level 2; 3) If the RGB value corresponding to the smoke is in the interval of R∈[112, 192], G∈[128, 192], B∈[105, 192], or the current diffusion speed of the smoke does not exceed 0.3m / s, then the severity level is determined to be level 3; Severity level 1> Severity level 2> Severity level 3; 4) The local data center generates a warning message based on the determined severity level of the fire and sends it to the building's management personnel, residents and fire department through short messages, AI voice calls or forced activation of a mobile phone APP for screen flashing warnings; The diffusion speed of the smoke is obtained by the local data center by obtaining the RGB video image of the room closest to the flame area, taking the wall corner line as a reference, drawing a length reference line extending to the boundary of the RGB video image that coincides with the reference, converting the horizontal length of the area corresponding to the RGB video image, and obtaining the video duration T1 corresponding to the diffusion of smoke to the entire RGB video image range, and obtaining the diffusion speed of the smoke: Where L is the pixel length of the length reference line; α is the inclination angle between the RGB camera for acquiring the indoor RGB video image and the ground; β is the deflection angle of the length reference line relative to the vertical direction; λ is the distortion coefficient of the indoor RGB video image; η is the scaling ratio of the pixel size of the indoor RGB video image to the actual size.

2. The intelligent fire warning method according to claim 1, characterized in that: The planning of the escape passage inside the building described in step S3 includes the following contents: 1) using a chain sequence numbering method to sequentially number the monitoring areas of the RGB cameras on each floor of the building, and the numbering order of the monitoring areas on each floor is in a clockwise direction from the inside to the outside; 2) when the local data center confirms the existence of a flame area, determine the number of the monitoring area corresponding to the flame demand, and mark the fire spread area on the same floor or adjacent floors within 5 minutes, 10 minutes and 20 minutes according to the smoke spread direction in the fire severity level and the flame spread direction obtained by the second RGB video image; 3) turn on the active fire extinguishing device of the fire extinguishing unit in the fire spread area to reduce the spread speed of the fire; 4) according to the location of the flame area, plan the evacuation route in the order of near first and far later on the same floor, and high first and low later on different floors, and the planned evacuation route is far away from the power shaft, ventilation shaft or elevator room near the fire spread area.

3. The intelligent fire warning method according to claim 1, characterized in that: A card reader is provided in the cabinet within the fire extinguishing unit, and a recognizable graphic code is provided on the fire-fighting equipment. When the fire-fighting equipment is placed in or taken out of the cabinet, it will be read by the card reader and the equipment status in the fire-fighting unit will be updated, and the new operating status of the fire-fighting equipment will be generated and sent to the local data center.

4. The intelligent fire warning method according to claim 3, characterized in that: The local data center is also equipped with a duty supervision unit; the building management personnel regularly inspect the local fire monitoring unit and the fire extinguishing unit, return the fire-fighting equipment to its place or replenish the quantity, and swipe the card to sign in at the card reader, and the card reader generates sign-in information and updated fire-fighting equipment operation status information and sends it to the duty supervision unit of the local data center; the duty supervision unit also monitors the on-duty status of the staff of the local data center, and sends non-fire alarm information to the on-duty staff and their superiors in the event of absence or absence from work for a certain period of time.

5. An intelligent fire warning system, used to implement an intelligent fire warning method according to any one of claims 1 to 4, characterized in that: include: Local fire monitoring units are deployed in different areas of each floor of the building to obtain on-site monitoring information from different areas of the building; Fire extinguishing units are located in different areas on each floor of the building and are used to obtain the operating status of fire extinguishing facilities; The local data center is set up near the building. It is used to obtain on-site monitoring information and the operating status of fire-fighting facilities. It also determines whether there is a fire in the building based on the on-site monitoring information. If there is a fire, the local data center determines the severity of the fire, predicts the spread trend of the fire, generates warning information, and plans escape routes inside the building; The remote service unit is connected to the local data center for receiving warning information from the local data center to prompt emergency evacuation or disaster relief.

Citation Information

Patent Citations

  • Intelligent fire fighting system based on Internet of things

    CN110496355A

  • Fire alarm method based on Internet of Things equipment and video image comprehensive identification

    CN115601919A

  • Automatic fire alarm linkage system

    CN116665395A

  • Method and system for monitoring fire

    US20210049885A1