Intelligent fire warning system based on AI and data analysis

Abstract

The application relates to the technical field of fire-fighting early warning, and discloses an intelligent fire-fighting early warning system based on AI and data analysis, wherein an intelligent monitoring module numbers all fire-fighting indicator lights in a safety evacuation passage, then obtains i pieces of picture data of each fire-fighting indicator light, and stores the picture data under the corresponding fire-fighting indicator light number; a data processing module selects one piece of picture data most suitable for data analysis; a data analysis module calculates an indicator light fault parameter C n ; compares the fault parameter C n with a preset target threshold value C0, if C n is less than C0, further judgment needs to be made on the fire-fighting indicator light to determine whether the fire-fighting indicator light is abnormal, and if C n is greater than or equal to C0, the fire-fighting indicator light is abnormal, and corresponding early warning instructions are immediately sent through an early warning module.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fire warning technology, and particularly relates to a smart fire warning system based on AI and data analysis. BACKGROUND

[0002] "Fire control" means eliminating hidden dangers and preventing disasters (i.e. preventing and solving man-made and natural disasters encountered by people in the process of life, work and study), and of course, in the early stage of people's understanding, it means (extinguishing) fire. It mainly includes personnel rescue at the fire scene, rescue of important facilities, equipment and cultural relics, safety protection and rescue of important property, and extinguishing fire. The purpose is to reduce the damage caused by fire and reduce personnel casualties and property losses.

[0003] The current fire warning system is mostly to determine whether there is a fire hazard through temperature sensor data and / or smoke concentration sensor data. However, when a fire hazard is found, the safety evacuation passage is a special passage for guiding people to evacuate to a safe area. If the smoke caused by the fire is very large, the fire indicator light in the safety evacuation passage can guide people to evacuate from the safety evacuation passage. Therefore, it is necessary to ensure that the fire indicator light in the building is in a normal working state. When an abnormal state of the fire indicator light is found, fire warning must be carried out immediately. SUMMARY

[0004] The purpose of the present application is to provide a smart fire warning system based on AI and data analysis to solve the above technical problems.

[0005] The purpose of the present application can be achieved by the following technical solutions:

[0006] A smart fire warning system based on AI and data analysis, comprising an intelligent monitoring module, a data analysis module, a warning module and a main controller;

[0007] The intelligent monitoring module is used for patrolling along the safety evacuation passage and acquiring front picture data of all fire indicator lights in the safety evacuation passage;

[0008] The data processing module is used for processing the picture data acquired by the intelligent monitoring module and extracting picture data that can be used for data analysis;

[0009] The data analysis module is used for judging whether the fire indicator light is in a normal working state according to the picture data extracted by the data processing module;

[0010] The warning module is used for issuing corresponding warning instructions according to the result of the judgment of the data analysis module;

[0011] The main controller is used for controlling all modules to work normally.

[0012] Using the above technical solution, the intelligent monitoring module numbers all fire indicator lights in the safe evacuation route, then acquires i images of each fire indicator light and stores them under the corresponding fire indicator light number; the data processing module selects the most suitable image for data analysis; and the data analysis module calculates the indicator light fault parameter C. n ; The fault parameter C n Compare with the preset target threshold C0, if C n If the value is less than n0, further judgment is needed on the fire indicator light, and the fault parameter C is used. n Calculate the fault parameter measurement index σ n And through the fault parameter measurement index σ n Determine the deviation index D of the fault parameters n ; the deviation index D n Compared with the preset deviation threshold D0, if the deviation index D n If the deviation threshold D0 is greater than or equal to the deviation value, the fire indicator light is abnormal, and a corresponding warning command will be immediately issued through the warning module. If C n If the value is greater than or equal to C0, the fire indicator light is abnormal, and a corresponding warning command will be immediately issued through the warning module.

[0013] As a further description of the present invention, the working process of the intelligent monitoring module includes:

[0014] Number all fire indicator lights in the safe evacuation routes: 1, 2, 3, ..., n, where n is the number of fire indicator lights;

[0015] Periodically inspect the safety evacuation routes to obtain frontal image data of the fire indicator lights within the safety evacuation routes. Obtain i images for each fire indicator light and store them under the corresponding fire indicator light number.

[0016] As a further description of the present invention, the working process of the data processing module includes:

[0017] Acquire historical image data collected by the intelligent monitoring module and generate training samples;

[0018] A model is built using convolutional neural networks, and the model is trained using training samples to obtain an image analysis model;

[0019] The image data of each fire indicator light is used as input data into the analysis model. After processing by the analysis model, the most suitable image data for data analysis is output.

[0020] Assign the fire indicator light number to the corresponding image data for data analysis.

[0021] Through the technical scheme, all the obtained picture data under each fire indicator light number is judged, and a picture data most suitable for data analysis is screened out through the learning model, so as to avoid the influence of external factors such as light on the judgment result.

[0022] As a further description of the present application, the working process of the data analysis module includes:

[0023] Step S100, dividing the picture data for data analysis into several regions of the same size, and numbering each region: 1, 2, 3, …, i, wherein i is the number of regions;

[0024] Step S200, obtaining the chrominance P i and the brightness L i of each region picture; i i

[0025] Step S300, according to the chrominance P i and the brightness L i of each region picture obtained in step S200, obtaining the average chrominance P n and the average brightness L n of the picture data for data analysis:

[0026]

[0027] Step S400, according to the average chrominance P n and the average brightness L n of the picture data for data analysis obtained in step 300, obtaining the indicator light fault parameter C n :

[0028]

[0029] Wherein, α is a compensation coefficient;

[0030] Step S500, comparing the fault parameter C n with the preset target threshold C0, if C n is less than C0, further judgment is performed, if C n is greater than or equal to C0, it is judged that the fire indicator light is abnormal;

[0031] Step S600, according to the judgment result in step S500, if the fire indicator light is abnormal, immediately issuing a corresponding early warning instruction through the early warning module.

[0032] As a further description of the present application, the further judgment includes:

[0033] According to the fault parameter C n obtained in the step S400, obtaining the measurement index σ of the fault parametern :

[0034]

[0035] in, This represents the average value of the fault parameters;

[0036] Based on the fault parameter measurement index σ n Determine the deviation index D of the fault parameters n :

[0037]

[0038] Deviation index D n Compared with the preset deviation threshold D0, if the deviation index D n If the deviation threshold D0 is greater than or equal to the deviation threshold, the fire indicator light is judged to be abnormal, and a corresponding warning command is immediately issued through the warning module.

[0039] Using the above technical solution, the image data used for data analysis is divided into i regions of equal size, and the chromaticity P of each region is obtained. i and brightness L i According to the chromaticity P of each acquired region image i and brightness L i Calculate the average chromaticity P of the image data used for data analysis. n and average brightness L n Through average chromaticity P n and average brightness L n Determine the fault parameter C of the indicator light n Fault parameter C n Compare with the preset target threshold C0, if C n If C is greater than or equal to C0, the fire indicator light is considered abnormal; if C... n If the value is less than C0, further judgment is made, and the fault parameter C is used. n Calculate the fault parameter measurement index σ n And through the fault parameter measurement index σ n Determine the deviation index D of the fault parameters n ; the deviation index D n Compared with the preset deviation threshold D0, if the deviation index D n If the deviation threshold D0 is greater than or equal to the deviation threshold, the fire indicator light is abnormal, and a corresponding warning command will be issued immediately through the warning module.

[0040] As a further description of the present invention, the working method of the fire early warning system includes the following steps:

[0041] Step SS100, the intelligent monitoring module numbers all the fire indicator lights in the safety evacuation passage, then acquires i pieces of picture data of each fire indicator light and stores them under the corresponding fire indicator light number;

[0042] Step SS200, the data processing module processes the picture data of each fire indicator light acquired and selects one picture data most suitable for data analysis;

[0043] Step SS300, the data analysis module calculates the average chroma P n and the average brightness L n of the picture data for data analysis, and calculates the indicator light failure parameter C n according to the average chroma P n and the average brightness L n ;

[0044] Step SS400, the failure parameter C n is compared with the preset target threshold C0, if C n is less than C0, step SS500 is entered, if C n is greater than or equal to C0, step SS700 is entered;

[0045] Step SS500, the failure parameter C n calculates the measurement index σ n of the failure parameter, and calculates the deviation index D n of the failure parameter through the measurement index σ n of the failure parameter;

[0046] Step SS600, the deviation index D n is compared with the preset deviation threshold D0, if the deviation index D n is greater than or equal to the deviation threshold D0, step SS700 is entered;

[0047] Step SS700, the data analysis module judges that the fire indicator light is abnormal and immediately issues a corresponding early warning instruction through the early warning module.

[0048] As a further description of the scheme of the application, when the early warning module issues a corresponding early warning instruction, the number of the abnormal fire indicator light is also issued.

[0049] As a further description of the scheme of the application, the intelligent monitoring module collects picture data based on an artificial intelligence robot.

[0050] The beneficial effects of this invention are as follows: 1. The intelligent monitoring module numbers all fire indicator lights in the safe evacuation route, then acquires i images of each fire indicator light and stores them under the corresponding fire indicator light number; the data processing module selects the most suitable image for data analysis; the data analysis module calculates the indicator light fault parameter C. n ; The fault parameter C n Compare with the preset target threshold C0, if C n If the value is less than C0, further assessment of the fire indicator lights is needed to determine if they are malfunctioning. If C... n If the value is greater than or equal to C0, the fire indicator light is abnormal, and a corresponding warning command will be immediately issued through the warning module.

[0051] 2. The data processing module evaluates all the images acquired under each fire indicator light number and uses a learning model to select the most suitable image for data analysis, thus avoiding interference from external factors such as lighting that could affect the evaluation results. Attached Figure Description

[0052] The invention will now be further described with reference to the accompanying drawings.

[0053] Figure 1 This is a block diagram of the intelligent fire early warning system based on AI and data analysis provided by the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Please see Figure 1 As shown, the present invention is a smart fire early warning system based on AI and data analysis, including an intelligent monitoring module, a data analysis module, an early warning module and a main controller;

[0056] The intelligent monitoring module is used to patrol along the safe evacuation route and acquire frontal image data of all fire indicator lights in the safe evacuation route;

[0057] The data processing module is used to process the image data acquired by the intelligent monitoring module and extract image data that can be used for data analysis.

[0058] The data analysis module is used to determine whether the fire indicator light is in normal working condition based on the image data extracted by the data processing module.

[0059] The early warning module is used to issue corresponding early warning commands based on the results interpreted by the data analysis module.

[0060] The main controller is used to control the normal operation of all modules.

[0061] Using the above technical solution, the intelligent monitoring module numbers all fire indicator lights in the safe evacuation route, then acquires i images of each fire indicator light and stores them under the corresponding fire indicator light number; the data processing module selects the most suitable image for data analysis; and the data analysis module calculates the indicator light fault parameter C. n ; The fault parameter C n Compare with the preset target threshold C0, if C n If the value is less than C0, further judgment is made on the fire indicator light, and the fault parameter C is used. n Calculate the fault parameter measurement index σ n And through the fault parameter measurement index σ n Determine the deviation index D of the fault parameters n ; the deviation index D n Compared with the preset deviation threshold D0, if the deviation index D n If the deviation threshold D0 is greater than or equal to the deviation value, the fire indicator light is abnormal, and a corresponding warning command will be immediately issued through the warning module. If C n If the value is greater than or equal to C0, the fire indicator light is abnormal, and a corresponding warning command will be immediately issued through the warning module.

[0062] The working process of the intelligent monitoring module includes:

[0063] Number all fire indicator lights in the safe evacuation routes: 1, 2, 3, ..., n, where n is the number of fire indicator lights;

[0064] Periodically inspect the safety evacuation routes to obtain frontal image data of the fire indicator lights within the safety evacuation routes. Obtain i images for each fire indicator light and store them under the corresponding fire indicator light number.

[0065] The working process of the data processing module includes:

[0066] Acquire historical image data collected by the intelligent monitoring module and generate training samples;

[0067] A model is built using convolutional neural networks, and the model is trained using training samples to obtain an image analysis model;

[0068] The image data of each fire indicator light is used as input data into the analysis model. After processing by the analysis model, the most suitable image data for data analysis is output.

[0069] The number of the fire indication lamp is assigned to the corresponding picture data for data analysis.

[0070] Through the above technical solution, all the pictures obtained under each fire indication lamp number are judged, and the picture data most suitable for data analysis is selected through the learning model, so as to avoid the influence of external factors such as light on the judgment result.

[0071] The working process of the data analysis module includes:

[0072] Step S100, divide the picture data for data analysis into several regions of the same size, and number each region: 1, 2, 3, …, i, wherein i is the number of regions;

[0073] Step S200, obtain the chrominance P i and the brightness L i of each region picture;

[0074] Step S300, according to the chrominance P i and the brightness L i of each region picture obtained in step S200, calculate the average chrominance P n and the average brightness L n of the picture data for data analysis:

[0075]

[0076] Step S400, according to the average chrominance P n and the average brightness L n of the picture data for data analysis obtained in step 300, calculate the indication lamp fault parameter C n :

[0077]

[0078] Wherein, α is a compensation coefficient;

[0079] Step S500, compare the fault parameter C n with the preset target threshold C0, if C n is less than C0, further judgment is performed, if C n is greater than or equal to C0, it is judged that the fire indication lamp is abnormal;

[0080] Step S600, according to the judgment result in step S500, if the fire indication lamp is abnormal, immediately issue the corresponding early warning instruction through the early warning module.

[0081] The further judgment includes:

[0082] According to the fault parameter C calculated in the step S400 n The measurement index σ of the fault parameter n :

[0083]

[0084] Wherein, is the mean value of the fault parameter;

[0085] According to the measurement index σ of the fault parameter n The deviation index D of the fault parameter n :

[0086]

[0087] The deviation index D n is compared with the preset deviation threshold D0, if the deviation index D n is greater than or equal to the deviation threshold D0, it is judged that the fire indicating lamp is abnormal, and the corresponding early warning instruction is immediately sent through the early warning module.

[0088] Through the above technical solution, the picture data used for data analysis is divided into i same size areas, the chroma P i and the brightness L i of each area picture are obtained, according to the obtained chroma P i and the brightness L i of each area picture, the average chroma P n and the average brightness L n of the picture data used for data analysis are obtained, the indicating lamp fault parameter C n is obtained through the average chroma P n and the average brightness L n , the fault parameter C n is compared with the preset target threshold C0, if C n is greater than or equal to C0, it is judged that the fire indicating lamp is abnormal, if C n is less than C0, further judgment is carried out, and the fault parameter C n is used to obtain the measurement index σ of the fault parameter n , the measurement index σ of the fault parameter n is used to obtain the deviation index D of the fault parameter n ; the deviation index D n is compared with the preset deviation threshold D0, if the deviation index D n is greater than or equal to the deviation threshold D0, the fire indicating lamp is abnormal, and the corresponding early warning instruction is immediately sent through the early warning module.

[0089] The working method of the fire early warning system comprises the following steps:

[0090] Step SS100, the intelligent monitoring module numbers all the fire indicator lights in the safety evacuation passage, then obtains i pieces of picture data for each fire indicator light, and stores them under the corresponding fire indicator light number;

[0091] Step SS200, the data processing module processes the picture data obtained for each fire indicator light, and selects one picture data most suitable for data analysis;

[0092] Step SS300, the data analysis module calculates the average chroma P n and the average brightness L n of the picture data used for data analysis, and calculates the indicator light failure parameter C n according to the average chroma P n and the average brightness L n ;

[0093] Step SS400, the failure parameter C n is compared with a preset target threshold C0, if C n is less than C0, step SS500 is entered, if C n is greater than or equal to C0, step SS700 is entered;

[0094] Step SS500, the failure parameter C n calculates the measurement index σ n of the failure parameter, and calculates the deviation index D n of the failure parameter through the measurement index σ n of the failure parameter;

[0095] Step SS600, the deviation index D n is compared with a preset deviation threshold D0, if the deviation index D n is greater than or equal to the deviation threshold D0, step SS700 is entered;

[0096] Step SS700, the data analysis module judges that the fire indicator light is abnormal, and immediately issues a corresponding early warning instruction through the early warning module.

[0097] When the early warning module issues a corresponding early warning instruction, the number of the abnormal fire indicator light will be issued together.

[0098] The intelligent monitoring module collects picture data based on an artificial intelligence robot.

[0099] The above has described one embodiment of the present application in detail, but the content described is only the preferred embodiment of the present application, and cannot be considered as used to limit the implementation range of the present application. Any equivalent changes and improvements made according to the application scope of the present application should still belong to the patent coverage range of the present application.

Claims

1. A smart fire early warning system based on AI and data analysis, characterized in that, It includes an intelligent monitoring module, a data processing module, a data analysis module, an early warning module, and a main controller; The intelligent monitoring module is used to patrol along the safe evacuation route and acquire frontal image data of all fire indicator lights in the safe evacuation route; The data processing module is used to process the image data acquired by the intelligent monitoring module and extract image data that can be used for data analysis. The data analysis module is used to determine whether the fire indicator light is in normal working condition based on the image data extracted by the data processing module. The early warning module is used to issue corresponding early warning commands based on the results interpreted by the data analysis module. The main controller is used to control the normal operation of all modules; The working process of the data analysis module includes: Step S100: Divide the image data used for data analysis into several regions of the same size, and number each region as follows: 1, 2, 3, ..., i, where i is the number of regions; Step S200: Obtain the chromaticity of each region of the image. and brightness ; Step S300: Based on the chromaticity of each region image obtained in step S200 and brightness Calculate the average chromaticity of the image data used for data analysis. and average brightness : ； Step S400: Based on the average chromaticity of the image data obtained in step 300 for data analysis and average brightness Determine the fault parameters of the indicator light : ； in, The compensation coefficient; Step S500: Transfer fault parameters With the preset target threshold If a comparison is made, Less than Then further judgment is made, if Greater than or equal to Then the fire indicator light is determined to be abnormal; Step S600: Based on the judgment result in step S500, if the fire indicator light is abnormal, immediately issue a corresponding warning command through the warning module; Further assessment of fire indicator lights includes: Based on the fault parameters obtained in step S400 Calculate the metrics for fault parameters : ； in, This represents the average value of the fault parameters; Based on the measurement indicators of fault parameters Determine the deviation index of fault parameters : ； Deviation index Deviation threshold from preset Comparison, if the deviation index Greater than or equal to the deviation threshold If the fire indicator light is detected as abnormal, a corresponding warning command will be immediately issued through the warning module.

2. The intelligent fire early warning system based on AI and data analysis according to claim 1, characterized in that, The working process of the intelligent monitoring module includes: Number all fire indicator lights in the safe evacuation routes: 1, 2, 3, ..., n, where n is the number of fire indicator lights; Periodically inspect the safety evacuation routes to obtain frontal image data of the fire indicator lights within the safety evacuation routes. Obtain i images for each fire indicator light and store them under the corresponding fire indicator light number.

3. The intelligent fire early warning system based on AI and data analysis according to claim 2, characterized in that, The working process of the data processing module includes: Acquire historical image data collected by the intelligent monitoring module and generate training samples; A model is built using convolutional neural networks, and the model is trained using training samples to obtain an image analysis model; The image data of each fire indicator light is used as input data into the analysis model. After processing by the analysis model, the most suitable image data for data analysis is output. Assign the fire indicator light number to the corresponding image data for data analysis.

4. The intelligent fire early warning system based on AI and data analysis according to claim 3, characterized in that, The working method of the fire early warning system includes the following steps: Step SS100: The intelligent monitoring module numbers all fire indicator lights in the safe evacuation route, then acquires i image data for each fire indicator light and stores them under the corresponding fire indicator light number; Step SS200: The data processing module processes the image data of each fire indicator light and selects the most suitable image data for data analysis. Step SS300: The data analysis module calculates the average chromaticity of the image data used for data analysis. and average brightness And based on average chromaticity and average brightness Determine the fault parameters of the indicator light ; Step SS400: Transfer fault parameters With the preset target threshold If a comparison is made, Less than Proceed to step SS500, if Greater than or equal to Proceed to step SS700; Step SS500, Fault Parameters Calculate the metrics for fault parameters And through the measurement indicators of fault parameters. Determine the deviation index of fault parameters ; Step SS600: Deviance Index Deviation threshold from preset Comparison, if the deviation index Greater than or equal to the deviation threshold Proceed to step SS700; Step SS700: The data analysis module determines that the fire indicator light is abnormal and immediately issues a corresponding warning command through the early warning module.

5. A smart fire early warning system based on AI and data analysis according to claim 4, characterized in that, When the early warning module issues a corresponding early warning command, it will also send out the abnormal fire indicator light number.

6. A smart fire early warning system based on AI and data analysis according to claim 5, characterized in that, The intelligent monitoring module collects image data based on an artificial intelligence robot.

Images