CN116884159A - Intelligent fire-fighting early warning method and system based on AI (advanced technology attachment) identification - Google Patents
Intelligent fire-fighting early warning method and system based on AI (advanced technology attachment) identification Download PDFInfo
- Publication number
- CN116884159A CN116884159A CN202310755223.8A CN202310755223A CN116884159A CN 116884159 A CN116884159 A CN 116884159A CN 202310755223 A CN202310755223 A CN 202310755223A CN 116884159 A CN116884159 A CN 116884159A
- Authority
- CN
- China
- Prior art keywords
- fire
- factor
- induction
- historical
- sets
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000005516 engineering process Methods 0.000 title description 5
- 230000006698 induction Effects 0.000 claims abstract description 231
- 238000013210 evaluation model Methods 0.000 claims abstract description 40
- 238000011156 evaluation Methods 0.000 claims abstract description 28
- 230000005540 biological transmission Effects 0.000 claims description 6
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 238000000556 factor analysis Methods 0.000 claims description 6
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 101710154918 Trigger factor Proteins 0.000 claims description 3
- 230000001939 inductive effect Effects 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 claims description 2
- 238000012360 testing method Methods 0.000 description 7
- 238000012549 training Methods 0.000 description 7
- 238000012795 verification Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 231100001267 hazard identification Toxicity 0.000 description 1
- 238000001746 injection moulding Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
- G08B17/12—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions
- G08B17/125—Actuation by presence of radiation or particles, e.g. of infrared radiation or of ions by using a video camera to detect fire or smoke
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Emergency Management (AREA)
- Business, Economics & Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Alarm Systems (AREA)
Abstract
The application relates to the technical field of intelligent fire control, and provides a fire control intelligent early warning method and system based on AI identification, wherein the method comprises the following steps: acquiring a plurality of historical fire information sets; obtaining a plurality of historical fire induction factor sets; obtaining a plurality of first fire induction factor sequences; extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets; acquiring a fire induction key factor set according to a target fire scene; constructing a factor image evaluation model, and inputting an image acquisition result into the factor image evaluation model for evaluation to obtain a risk identification result; and when the dangerous recognition result meets the triggering threshold value of the induction factor, generating a fire early warning signal. The problem of low quality of the fire-fighting intelligent early warning result can be solved, and the accuracy and timeliness of the early warning result can be improved.
Description
Technical Field
The application relates to the technical field of intelligent fire protection, in particular to a fire protection intelligent early warning method and system based on AI identification.
Background
The intelligent fire control is an intelligent fire control scheme combining advanced technologies such as Internet of things and 5G, and compared with traditional fire control, the intelligent fire control scheme focuses on opening information islands among systems, improving perception early warning capability and emergency command intelligent capability. By finding earlier, handling faster, the risk and impact of fire is minimized.
The traditional fire-fighting early-warning method usually only starts alarming when a fire signal appears, and measures are difficult to take in advance to stop, and the current intelligent fire-fighting early-warning method usually carries out early warning according to a single factor, does not carry out comprehensive analysis on a plurality of reasons generated by the fire, and causes lower early-warning result quality.
In summary, the problem of low quality of the fire-fighting intelligent early warning result exists in the prior art.
Disclosure of Invention
Based on the above, it is necessary to provide a fire-fighting intelligent early warning method and system based on AI identification aiming at the technical problems.
An intelligent fire-fighting early warning method based on AI identification, the method comprises the following steps: acquiring historical fire events according to a preset acquisition unit and an acquisition scene to obtain a plurality of historical fire information sets, wherein the acquisition scene is a fire occurrence scene, and the historical fire information sets are divided according to the acquisition scene; carrying out historical fire event induction factor analysis on the plurality of historical fire information sets to obtain a plurality of historical fire induction factor sets, wherein the historical induction factor sets and the historical fire events have corresponding relations, and the number of the historical fire induction factor sets is equal to the number of the historical fire events; based on a fire expert system, arranging fire induction factors in the historical fire induction factor sets, and extracting N fire induction factors before ranking the fire induction factor sequences to obtain a first fire induction factor sequences, wherein N is an integer greater than 1; extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets, wherein the fire key induction factor sets have corresponding relations with the acquisition scene; acquiring a fire induction key factor set according to a target fire scene, wherein the target fire scene is any one scene in the acquisition scenes; acquiring image data and constructing a factor image evaluation model based on the fire induction key factor set, and inputting an image acquisition result into the factor image evaluation model for evaluation to obtain a hazard recognition result; judging the dangerous recognition result according to the triggering threshold of the triggering factor, generating a fire early-warning signal when the dangerous recognition result meets the triggering threshold of the triggering factor, and transmitting the fire early-warning signal step by step according to a preset transmission rule.
An AI identification-based fire-fighting intelligent early warning system, comprising:
the system comprises a historical fire information set acquisition module, a data acquisition module and a data processing module, wherein the historical fire information set acquisition module is used for acquiring historical fire events according to a preset acquisition unit and an acquisition scene to acquire a plurality of historical fire information sets, the acquisition scene is a fire occurrence scene, and the historical fire information sets are divided according to the acquisition scene;
a historical fire induction factor set obtaining module, configured to perform historical fire event induction factor analysis on a plurality of the historical fire information sets to obtain a plurality of historical fire induction factor sets, where the historical induction factor sets and the historical fire events have a correspondence, and the number of the historical fire induction factor sets is equal to the number of the historical fire events;
the first fire induction factor sequence obtaining module is used for arranging fire induction factors in the historical fire induction factor sets based on a fire expert system, extracting N fire induction factors before the ranking of the fire induction factor sequences, and obtaining a plurality of first fire induction factor sequences, wherein N is an integer greater than 1;
the fire induction key factor set obtaining module is used for extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets, wherein the fire induction key factor sets have a corresponding relation with the acquisition scene;
the fire induction key factor set matching module is used for obtaining a fire induction key factor set according to a target fire scene, wherein the target fire scene is any one of the acquisition scenes;
the dangerous identification result obtaining module is used for collecting image data and constructing a factor image evaluation model based on the fire induction key factor set, inputting an image collection result into the factor image evaluation model for evaluation, and obtaining a dangerous identification result;
and the fire early-warning signal generation module is used for judging the dangerous identification result according to the triggering threshold of the induction factor, generating a fire early-warning signal when the dangerous identification result meets the triggering threshold of the induction factor, and transmitting the fire early-warning signal step by step according to a preset transmission rule.
According to the intelligent fire-fighting early-warning method and system based on AI identification, the problem of low quality of intelligent fire-fighting early-warning results can be solved, and a plurality of historical fire information sets are obtained; obtaining a plurality of historical fire induction factor sets; obtaining a plurality of first fire induction factor sequences; extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets; acquiring a fire induction key factor set according to a target fire scene; constructing a factor image evaluation model, and inputting an image acquisition result into the factor image evaluation model for evaluation to obtain a risk identification result; and when the dangerous recognition result meets the triggering threshold value of the induction factor, generating a fire early warning signal. The accuracy and timeliness of the early warning result can be improved, so that measures can be taken in time to protect against fire.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
FIG. 1 is a schematic flow chart of a fire-fighting intelligent early warning method based on AI identification;
FIG. 2 is a schematic flow chart of acquiring a plurality of fire induction key factor sets in a fire-fighting intelligent early warning method based on AI identification;
FIG. 3 is a schematic flow chart of a sample data set constructed in the intelligent fire-fighting early warning method based on AI identification;
fig. 4 is a schematic structural diagram of a fire-fighting intelligent early warning system based on AI identification.
Reference numerals illustrate: the system comprises a historical fire information set acquisition module 1, a historical fire induction factor set acquisition module 2, a first fire induction factor sequence acquisition module 3, a fire induction key factor set acquisition module 4, a fire induction key factor set matching module 5, a hazard identification result acquisition module 6 and a fire early warning signal generation module 7.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
As shown in fig. 1, the application provides a fire-fighting intelligent early warning method based on AI identification, which comprises the following steps:
step S100: acquiring historical fire events according to a preset acquisition unit and an acquisition scene to obtain a plurality of historical fire information sets, wherein the acquisition scene is a fire occurrence scene, and the historical fire information sets are divided according to the acquisition scene;
specifically, based on big data technology, according to predetermineeing collection unit and collection scene and carrying out historical fire event and gathering, predetermineeing collection unit and being referred to the number that the fire event gathered, predetermineeing collection unit field technicians can customize the assignment, for example: 100 pieces. The historical fire event collection is carried out by setting the same collection unit, so that the uniformity of sample data can be ensured. The specific scene position of scene fire occurrence is collected, for example: residential areas, office buildings, chemical plants, shops, ordinary factories, gas stations and the like. A plurality of historical fire information sets are obtained, wherein the historical fire information sets are divided according to acquisition scenes. By obtaining a plurality of historical fire information sets, support is provided for further incentive analysis of fire events.
Step S200: carrying out historical fire event induction factor analysis on the plurality of historical fire information sets to obtain a plurality of historical fire induction factor sets, wherein the historical induction factor sets and the historical fire events have corresponding relations, and the number of the historical fire induction factor sets is equal to the number of the historical fire events;
specifically, the fire occurrence cause analysis is performed on the historical fire events in the plurality of the sets of historical fire information, for example: when a fire occurs in a factory, the cause of the fire may be various reasons such as short circuit of the electric wires due to improper operation of the machine, inflammables being piled up in the short circuit area, security personnel not being patrol in time, etc. Wherein the historical fire event corresponds to one of the set of historical fire inducing factors, the number of the set of historical fire inducing factors being equal to the number of the historical fire events.
Step S300: based on a fire expert system, arranging fire induction factors in the historical fire induction factor sets, and extracting N fire induction factors before ranking the fire induction factor sequences to obtain a first fire induction factor sequences, wherein N is an integer greater than 1;
specifically, a fire control expert system is constructed, wherein the fire control expert system is a fire control expert database combining artificial intelligence with a database, and comprises a fire cause unit, a fire cause evaluation unit, a daily precaution measure unit and a rescue scheme matching unit, a large number of fire cases and rescue schemes are stored in the fire control expert system, and the fire control expert system can be updated through continuous learning. And arranging fire influence degrees of the historical fire induction factor sets in the historical fire events through the fire expert system, wherein the higher the influence degrees of the historical fire induction factors are, the higher the ranking is, the first fire induction factor with the highest influence degree is ranked, and a fire induction factor sequence is constructed. And then extracting the top N fire induction factors in the fire induction factor sequence to obtain a plurality of first fire induction factor sequences, wherein N is an integer greater than 1, and the specific value of N can be set by a person skilled in the art in a self-defined way, for example: 3. By extracting the fire induction factors with high influence degree in the fire induction factor sequence, the accuracy of obtaining the fire induction factors can be improved.
Step S400: extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets, wherein the fire key induction factor sets have corresponding relations with the acquisition scene;
as shown in fig. 2, in one embodiment, the step S400 of the present application further includes:
step S410: performing weight distribution on fire disaster induction factors in the first fire disaster induction factor sequences according to preset weight dividing proportion, and classifying the first fire disaster induction factor sequences according to acquisition scenes to obtain a plurality of first fire disaster induction factor sequence sets;
step S420: and extracting similar fire induction factors from the first fire induction factor sequence sets to obtain the number of the similar fire induction factors and the weight value corresponding to each fire induction factor.
Specifically, a weight dividing ratio is preset, and a person skilled in the art can perform custom setting based on the number of fire induction factors in the first fire induction factor sequence, for example: when the fire induction factors are 3, the weights are respectively 50%, 30% and 20% in sequence, wherein the earlier the sequence is, the larger the weight ratio is. And classifying the first fire induction factor sequences according to the acquisition scene to obtain a plurality of first fire induction factor sequence sets, wherein the first fire induction factor sequence sets have corresponding relations with the acquisition scene.
Then performing homogeneous induction factor extraction on a plurality of the first fire induction factor sequence sets, for example: in the wire manufacturing factory, the wire short circuit in the injection molding workshop and the wire short circuit in the stranded wire workshop are the same kind of induction factors, and a plurality of the same kind of fire induction factors and weight values corresponding to the fire induction factors are obtained.
In one embodiment, step S400 of the present application further includes:
step S430: obtaining evaluation values of a plurality of similar fire induction factors through weighted calculation, and arranging the similar fire induction factors from large to small according to the evaluation values to obtain M similar fire induction factors with the same class in the sequence ranking of the similar fire induction factors, wherein M is an integer greater than 1 and M is smaller than N;
step S440: and ranking the same-kind fire induction factor sequences to obtain M same-kind fire induction factors as a plurality of fire induction key factor sets, wherein the fire key induction factor sets have corresponding relations with the acquired scene.
Specifically, the weighted calculation is performed according to the number of the similar fire induction factors and the weight value corresponding to each fire induction factor, namely, the weight values corresponding to the similar fire induction factors are added, the obtained sum is the evaluation value of the similar fire induction factors, then the similar fire induction factors are arranged from large to small according to the evaluation value to obtain a similar induction factor sequence, M similar fire induction factors in the similar induction factor sequence, which are ranked first, are extracted, wherein M is an integer greater than 1, M is smaller than N, the specific numerical value of M can be set by a person skilled in the art based on the number of N in a self-defining manner, and M can be set to 3 on the assumption that the N value is 5. Obtaining the top M same kind fire induction factors of the same kind fire induction factor sequences. And then, ranking the same-kind fire induction factors as a plurality of fire induction key factor sets, wherein the fire induction key factor sets have a corresponding relation with the acquired scenes, namely one fire scene corresponds to one fire induction key factor set.
Step S500: acquiring a fire induction key factor set according to a target fire scene, wherein the target fire scene is any one scene in the acquisition scenes;
specifically, the plurality of fire induction key factor sets are matched according to a target fire scene, wherein the target fire scene is a scene to be subjected to fire-fighting intelligent early warning, and the target fire scene is any one scene in the acquisition scene, so that the fire induction key factor sets are obtained.
Step S600: acquiring image data and constructing a factor image evaluation model based on the fire induction key factor set, and inputting an image acquisition result into the factor image evaluation model for evaluation to obtain a hazard recognition result;
as shown in fig. 3, in one embodiment, the step S600 of the present application further includes:
step S610: factor image acquisition is carried out based on the fire induction key factor set, and the fire control expert system scores the factor image acquisition result to obtain a factor image scoring result;
step S620: and constructing a sample data set based on the factor image acquisition result and the factor image scoring result.
Specifically, according to the fire induction factors in the fire induction key factor set, factor image data searching and querying are performed based on big data technology, for example: and if the fire induction factor in the fire induction key factor set is a wire short circuit, obtaining a plurality of wire circuit images. Scoring the factor image acquisition result by the fire expert system, i.e. scoring according to the safety of the factor image acquisition result, for example: and if the factor image acquisition result is a circuit image, the circuit image with good package has high score and the circuit image with aging phenomenon has low score. And constructing a sample data set according to the factor image acquisition result and the factor image scoring result, and providing data support for the next step of supervision training of the model by obtaining the sample data set.
In one embodiment, step S600 of the present application further includes:
step S630: constructing a factor image evaluation model based on a convolutional neural network, wherein input data of the factor image evaluation model is the factor image acquisition result, and output data is the factor image grading result;
step S640: the factor image evaluation model comprises a factor image matching module and a plurality of factor image evaluation sub-models, wherein the number of the factor image evaluation sub-models is equal to the number of the fire induction key factors.
Specifically, a factor image evaluation model is constructed based on a convolutional neural network, and the factor image evaluation model is a neural network model which can be continuously subjected to iterative optimization in machine learning and is obtained through supervised training by a training data set. The input data of the factor image evaluation model is the factor image acquisition result, and the output data is the factor image grading result. The factor image evaluation model comprises a factor image matching module and a plurality of factor image evaluation sub-models, wherein the factor image matching module is used for matching fire induction factor types, and the factor image evaluation sub-models are used for evaluating factor images, wherein the number of the factor image evaluation sub-models is equal to that of the fire induction key factors.
Presetting a sample data dividing ratio, wherein the sample data dividing ratio can be assigned by a user in the field, for example: 80%, 10% and 10%. And dividing the sample data set into a training set, a verification set and a test set according to the preset sample dividing proportion. And performing supervised training on the factor image evaluation model by using the training set, and updating the weight value of the network parameter in the factor image evaluation model according to the error of the actual output result and the expected output result in the process of performing the supervised training until the model output result tends to a convergence state, and then performing output result verification on the model by using the verification set, wherein a verification accuracy index and a test accuracy index are preset, and can be custom set by a person skilled in the art, and the test accuracy index is larger than the verification accuracy index, for example: the verification accuracy index is 95%, and the test accuracy index is 98%. And when the accuracy rate of the model output result is larger than or equal to the verification accuracy rate index, testing the model through the test set, and when the model output result is larger than or equal to the test accuracy rate index, obtaining the factor image evaluation model. The accuracy of the factor image evaluation result can be improved by constructing a factor image evaluation model based on the convolutional neural network to evaluate the factor image.
In one embodiment, step S600 of the present application further includes:
step S650: real-time factor image acquisition is carried out based on the fire induction key factor set, and the real-time factor image acquisition result is input into the factor image evaluation model to obtain a plurality of factor image scoring results;
step S660: judging a plurality of factor image scoring results according to a preset factor safety coefficient, and generating a dangerous triggering frequency when the factor image scoring results are smaller than the preset factor safety coefficient to obtain a dangerous recognition result, wherein the dangerous recognition result comprises a dangerous triggering factor type and a total dangerous triggering frequency.
Specifically, according to the fire induction key factor set, real-time factor image acquisition is carried out through an image acquisition device to obtain a real-time factor image acquisition result, and the real-time factor image acquisition result is input into the factor image evaluation model to obtain a plurality of factor image scoring results. The factor security factor is preset, and a person skilled in the art can customize assignment based on actual situations, for example: the safety is over 60 minutes. Judging a plurality of factor image scoring results according to the preset factor safety coefficient, generating one dangerous triggering frequency when the factor image scoring results are smaller than the preset factor safety coefficient, and obtaining a dangerous identification result according to the type of the dangerous triggering factor and the total dangerous triggering frequency. And by obtaining the dangerous identification result, support is provided for generating early warning information in the next step.
Step S700: judging the dangerous recognition result according to the triggering threshold of the triggering factor, generating a fire early-warning signal when the dangerous recognition result meets the triggering threshold of the triggering factor, and transmitting the fire early-warning signal step by step according to a preset transmission rule.
In one embodiment, step S700 of the present application further comprises:
step S710: judging the total dangerous trigger times in the dangerous recognition result according to the preset factor trigger threshold, and generating a fire early-warning signal when the total dangerous trigger times meet the preset factor trigger threshold, wherein the fire early-warning signal comprises the dangerous trigger factor type and is sent step by step according to a preset sending rule.
Specifically, a factor trigger threshold is preset, which can be set by one skilled in the art in a custom manner, for example: 2 times. Judging the total dangerous triggering times in the dangerous recognition result according to the preset factor triggering threshold, and generating a fire early-warning signal when the total dangerous triggering times are larger than or equal to the preset factor triggering threshold, wherein the fire early-warning signal comprises the dangerous triggering factor type, and a preset fire early-warning signal sending rule which can be set based on the class user definition of a company responsible person, for example: when the fire early-warning signal is generated, the fire early-warning signal can be sent to a first responsible person such as a company manager and the like at the first time and then sequentially sent to a main responsible person such as a department manager, a group leader, first-line staff and the like. The method solves the problem of low quality of the fire-fighting intelligent early-warning result, and can improve the accuracy and timeliness of the early-warning result by generating the fire-warning signal, thereby timely taking measures to protect against fire.
In one embodiment, as shown in fig. 4, there is provided a fire-fighting intelligent pre-warning system based on AI identification, including: a historical fire information set obtaining module 1, a historical fire induction factor set obtaining module 2, a first fire induction factor sequence obtaining module 3, a fire induction key factor set obtaining module 4, a fire induction key factor set matching module 5, a hazard recognition result obtaining module 6, a fire early warning signal generating module 7, wherein:
the system comprises a historical fire information set acquisition module 1, wherein the historical fire information set acquisition module 1 is used for acquiring historical fire events according to a preset acquisition unit and an acquisition scene to obtain a plurality of historical fire information sets, wherein the acquisition scene is a fire occurrence scene, and the historical fire information sets are divided according to the acquisition scene;
a historical fire induction factor set obtaining module 2, wherein the historical fire induction factor set obtaining module 2 is configured to perform historical fire event induction factor analysis on a plurality of the historical fire information sets to obtain a plurality of historical fire induction factor sets, the historical induction factor sets and the historical fire events have a corresponding relationship, and the number of the historical fire induction factor sets is equal to the number of the historical fire events;
a first fire induction factor sequence obtaining module 3, where the first fire induction factor sequence obtaining module 3 is configured to arrange fire induction factors in the plurality of historical fire induction factor sets based on a fire control expert system, and extract N fire induction factors before ranking the fire induction factor sequences, to obtain a plurality of first fire induction factor sequences, where N is an integer greater than 1;
the fire induction key factor set obtaining module 4 is configured to extract key factors from the plurality of first fire induction factor sequences to obtain a plurality of fire induction key factor sets, where the fire induction key factor sets have a correspondence with the acquisition scene;
the fire induction key factor set matching module 5 is used for obtaining a fire induction key factor set according to a target fire scene, wherein the target fire scene is any one scene in the acquisition scenes;
the dangerous identification result obtaining module 6 is used for carrying out image data acquisition and constructing a factor image evaluation model based on the fire induction key factor set, inputting an image acquisition result into the factor image evaluation model for evaluation, and obtaining a dangerous identification result;
the fire early-warning signal generation module 7 is used for judging the dangerous identification result according to the triggering threshold of the triggering factor, generating a fire early-warning signal when the dangerous identification result meets the triggering threshold of the triggering factor, and transmitting the fire early-warning signal step by step according to a preset transmission rule.
In one embodiment, the system further comprises:
the first fire induction factor sequence set obtaining module is used for distributing weights of fire induction factors in the first fire induction factor sequences according to preset weight dividing proportions, classifying the first fire induction factor sequences according to acquisition scenes and obtaining a first fire induction factor sequence set;
the same kind of induction factor extraction module is used for extracting the same kind of induction factors from the first fire induction factor sequence sets to obtain the number of the same kind of fire induction factors and the weight value corresponding to each fire induction factor.
In one embodiment, the system further comprises:
the system comprises a first M similar fire induction factors obtaining modules, a second M similar fire induction factors obtaining module and a third M similar fire induction factors determining module, wherein the first M similar fire induction factors obtaining modules are used for obtaining evaluation values of a plurality of similar fire induction factors through weighted calculation, and arranging the similar fire induction factors according to the evaluation values from large to small to obtain M similar fire induction factors with a first sequence ranking of the similar fire induction factors, wherein M is an integer larger than 1, and M is smaller than N;
and the fire induction key factor set obtaining module is used for ranking the same-kind fire induction factors with the same-kind fire induction factor sequences to obtain a plurality of fire induction key factor sets, wherein the fire induction key factor sets have a corresponding relation with the acquisition scene.
In one embodiment, the system further comprises:
the factor image scoring result obtaining module is used for carrying out factor image acquisition based on the fire induction key factor set, scoring the factor image acquisition result through the fire control expert system and obtaining a factor image scoring result;
and the sample data set construction module is used for constructing a sample data set based on the factor image acquisition result and the factor image scoring result.
In one embodiment, the system further comprises:
the factor image evaluation model construction module is used for constructing a factor image evaluation model based on a convolutional neural network, wherein input data of the factor image evaluation model is the factor image acquisition result, and output data is the factor image scoring result;
the factor image evaluation model module refers to that the factor image evaluation model comprises a factor image matching module and a plurality of factor image evaluation sub-models, wherein the number of the factor image evaluation sub-models is equal to that of the fire induction key factors.
In one embodiment, the system further comprises:
the factor image scoring result obtaining modules are used for carrying out real-time factor image acquisition based on the fire induction key factor set, inputting the real-time factor image acquisition result into the factor image evaluation model and obtaining a plurality of factor image scoring results;
the risk identification result obtaining module is used for judging a plurality of factor image scoring results according to a preset factor safety coefficient, and generating a risk triggering frequency when the factor image scoring results are smaller than the preset factor safety coefficient to obtain a risk identification result, wherein the risk identification result comprises a risk triggering factor type and a total risk triggering frequency.
In one embodiment, the system further comprises:
the fire early-warning signal generation module is used for judging the total dangerous trigger times in the dangerous identification result according to the preset factor trigger threshold, and generating a fire early-warning signal when the total dangerous trigger times meet the preset factor trigger threshold, wherein the fire early-warning signal comprises the dangerous trigger factor type and is sent step by step according to a preset sending rule.
In summary, the application provides a fire-fighting intelligent early warning method and a fire-fighting intelligent early warning system based on AI identification, which have the following technical effects:
1. the problem of fire control intelligent early warning result quality lower is solved, through generating the conflagration early warning signal, can improve the rate of accuracy and the timeliness of early warning result to in time take measures and carry out the conflagration protection.
2. The accuracy of the factor image evaluation result can be improved by constructing a factor image evaluation model based on the convolutional neural network to evaluate the factor image.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples illustrate only a few embodiments of the application, which are described in detail and are not to be construed as limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of protection of the present application is to be determined by the appended claims.
Claims (8)
1. An intelligent fire-fighting early warning method based on AI identification is characterized by comprising the following steps:
acquiring historical fire events according to a preset acquisition unit and an acquisition scene to obtain a plurality of historical fire information sets, wherein the acquisition scene is a fire occurrence scene, and the historical fire information sets are divided according to the acquisition scene;
carrying out historical fire event induction factor analysis on the plurality of historical fire information sets to obtain a plurality of historical fire induction factor sets, wherein the historical induction factor sets and the historical fire events have corresponding relations, and the number of the historical fire induction factor sets is equal to the number of the historical fire events;
based on a fire expert system, arranging fire induction factors in the historical fire induction factor sets, and extracting N fire induction factors before ranking the fire induction factor sequences to obtain a first fire induction factor sequences, wherein N is an integer greater than 1;
extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets, wherein the fire key induction factor sets have corresponding relations with the acquisition scene;
acquiring a fire induction key factor set according to a target fire scene, wherein the target fire scene is any one scene in the acquisition scenes;
acquiring image data and constructing a factor image evaluation model based on the fire induction key factor set, and inputting an image acquisition result into the factor image evaluation model for evaluation to obtain a hazard recognition result;
judging the dangerous recognition result according to the triggering threshold of the triggering factor, generating a fire early-warning signal when the dangerous recognition result meets the triggering threshold of the triggering factor, and transmitting the fire early-warning signal step by step according to a preset transmission rule.
2. The method of claim 1, wherein the obtaining a plurality of sets of fire-inducing key factors comprises:
performing weight distribution on fire disaster induction factors in the first fire disaster induction factor sequences according to preset weight dividing proportion, and classifying the first fire disaster induction factor sequences according to acquisition scenes to obtain a plurality of first fire disaster induction factor sequence sets;
and extracting similar fire induction factors from the first fire induction factor sequence sets to obtain the number of the similar fire induction factors and the weight value corresponding to each fire induction factor.
3. The method as claimed in claim 2, comprising:
obtaining evaluation values of a plurality of similar fire induction factors through weighted calculation, and arranging the similar fire induction factors from large to small according to the evaluation values to obtain M similar fire induction factors with the same class in the sequence ranking of the similar fire induction factors, wherein M is an integer greater than 1 and M is smaller than N;
and ranking the same-kind fire induction factor sequences to obtain M same-kind fire induction factors as a plurality of fire induction key factor sets, wherein the fire key induction factor sets have corresponding relations with the acquired scene.
4. The method as claimed in claim 1, comprising:
factor image acquisition is carried out based on the fire induction key factor set, and the fire control expert system scores the factor image acquisition result to obtain a factor image scoring result;
and constructing a sample data set based on the factor image acquisition result and the factor image scoring result.
5. The method as recited in claim 4, comprising:
constructing a factor image evaluation model based on a convolutional neural network, wherein input data of the factor image evaluation model is the factor image acquisition result, and output data is the factor image grading result;
the factor image evaluation model comprises a factor image matching module and a plurality of factor image evaluation sub-models, wherein the number of the factor image evaluation sub-models is equal to the number of the fire induction key factors.
6. The method as recited in claim 5, comprising:
real-time factor image acquisition is carried out based on the fire induction key factor set, and the real-time factor image acquisition result is input into the factor image evaluation model to obtain a plurality of factor image scoring results;
judging a plurality of factor image scoring results according to a preset factor safety coefficient, and generating a dangerous triggering frequency when the factor image scoring results are smaller than the preset factor safety coefficient to obtain a dangerous recognition result, wherein the dangerous recognition result comprises a dangerous triggering factor type and a total dangerous triggering frequency.
7. The method as recited in claim 6, comprising:
judging the total dangerous trigger times in the dangerous recognition result according to the preset factor trigger threshold, and generating a fire early-warning signal when the total dangerous trigger times meet the preset factor trigger threshold, wherein the fire early-warning signal comprises the dangerous trigger factor type and is sent step by step according to a preset sending rule.
8. Fire control intelligent early warning system based on AI discernment, characterized in that includes:
the system comprises a historical fire information set acquisition module, a data acquisition module and a data processing module, wherein the historical fire information set acquisition module is used for acquiring historical fire events according to a preset acquisition unit and an acquisition scene to acquire a plurality of historical fire information sets, the acquisition scene is a fire occurrence scene, and the historical fire information sets are divided according to the acquisition scene;
a historical fire induction factor set obtaining module, configured to perform historical fire event induction factor analysis on a plurality of the historical fire information sets to obtain a plurality of historical fire induction factor sets, where the historical induction factor sets and the historical fire events have a correspondence, and the number of the historical fire induction factor sets is equal to the number of the historical fire events;
the first fire induction factor sequence obtaining module is used for arranging fire induction factors in the historical fire induction factor sets based on a fire expert system, extracting N fire induction factors before the ranking of the fire induction factor sequences, and obtaining a plurality of first fire induction factor sequences, wherein N is an integer greater than 1;
the fire induction key factor set obtaining module is used for extracting key factors from the first fire induction factor sequences to obtain a plurality of fire induction key factor sets, wherein the fire induction key factor sets have a corresponding relation with the acquisition scene;
the fire induction key factor set matching module is used for obtaining a fire induction key factor set according to a target fire scene, wherein the target fire scene is any one of the acquisition scenes;
the dangerous identification result obtaining module is used for collecting image data and constructing a factor image evaluation model based on the fire induction key factor set, inputting an image collection result into the factor image evaluation model for evaluation, and obtaining a dangerous identification result;
and the fire early-warning signal generation module is used for judging the dangerous identification result according to the triggering threshold of the induction factor, generating a fire early-warning signal when the dangerous identification result meets the triggering threshold of the induction factor, and transmitting the fire early-warning signal step by step according to a preset transmission rule.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310755223.8A CN116884159A (en) | 2023-06-26 | 2023-06-26 | Intelligent fire-fighting early warning method and system based on AI (advanced technology attachment) identification |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310755223.8A CN116884159A (en) | 2023-06-26 | 2023-06-26 | Intelligent fire-fighting early warning method and system based on AI (advanced technology attachment) identification |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116884159A true CN116884159A (en) | 2023-10-13 |
Family
ID=88265302
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310755223.8A Pending CN116884159A (en) | 2023-06-26 | 2023-06-26 | Intelligent fire-fighting early warning method and system based on AI (advanced technology attachment) identification |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116884159A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117688485A (en) * | 2024-02-02 | 2024-03-12 | 北京中卓时代消防装备科技有限公司 | Fire disaster cause analysis method and system based on deep learning |
-
2023
- 2023-06-26 CN CN202310755223.8A patent/CN116884159A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117688485A (en) * | 2024-02-02 | 2024-03-12 | 北京中卓时代消防装备科技有限公司 | Fire disaster cause analysis method and system based on deep learning |
CN117688485B (en) * | 2024-02-02 | 2024-04-30 | 北京中卓时代消防装备科技有限公司 | Fire disaster cause analysis method and system based on deep learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110263172B (en) | Power grid monitoring alarm information evenized autonomous identification method | |
CN105184084B (en) | Method and system for predicting fault type of electric power metering automation terminal | |
CN113190421A (en) | Detection and analysis method for equipment health state of data center | |
CN112149967B (en) | Power communication network vulnerability assessment method and system based on complex system theory | |
CN108304567B (en) | Method and system for identifying working condition mode and classifying data of high-voltage transformer | |
CN116884159A (en) | Intelligent fire-fighting early warning method and system based on AI (advanced technology attachment) identification | |
CN109491339B (en) | Big data-based substation equipment running state early warning system | |
CN116884193A (en) | Chip factory intelligent production monitoring alarm system based on multi-terminal induction fusion | |
CN114553671A (en) | Diagnosis method for power communication network fault alarm | |
CN112329688A (en) | Intelligent risk management and control system for public places | |
CN113094244A (en) | Intelligent detection system for computer room operation of data center | |
CN115617784A (en) | Data processing system and processing method for informationized power distribution | |
CN113010394B (en) | Machine room fault detection method for data center | |
CN114743089A (en) | Image recognition GIS fault diagnosis device and method based on SSA-SVM | |
CN116054416B (en) | Intelligent monitoring operation and maintenance management system based on Internet of things | |
CN108446202B (en) | Method for judging safety condition of machine room equipment | |
CN117595504A (en) | Intelligent monitoring and early warning method for power grid running state | |
CN114936614B (en) | Operation risk identification method and system based on neural network | |
CN114167837B (en) | Intelligent fault diagnosis method and system for railway signal system | |
CN113706841B (en) | Intelligent monitoring system for safety of power utilization behavior based on edge calculation | |
CN111313355A (en) | Method for updating monitoring signal event rule under manual supervision | |
CN116756619B (en) | Equipment intelligent diagnosis method and system based on big data | |
CN117557415B (en) | Community resource management method and system based on intelligent property | |
CN118094531B (en) | Safe operation and maintenance real-time early warning integrated system | |
CN117560300B (en) | Intelligent internet of things flow prediction and optimization system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |