Intelligent fire-fighting management system and method based on AI analysis

Abstract

The invention discloses an intelligent fire-fighting management system and method based on AI analysis, and relates to the technical field of fire-fighting management. The intelligent fire-fighting management method based on AI analysis comprises the following steps: acquiring fire basic risk data and fire correction risk data of each to-be-monitored area in real time, analyzing a fire correction risk index of each to-be-monitored area, and performing judgment and analysis on the fire correction risk index and a preset fire risk assessment interval; the to-be-monitored area with the fire correction risk index within the preset fire risk assessment interval is marked as the fire risk area, dynamic calculation is carried out by combining the real-time fire basic risk data and the correction risk data, the change trend of the fire risk can be accurately reflected, and the risk assessment accuracy is improved. When the fire risk index of a certain area exceeds the preset danger threshold value, fire alarm can be automatically triggered and a power supply cut-off measure is taken, so that the mechanism avoids the delay of traditional manual intervention, and the efficiency and the accuracy of fire early warning are improved.

Description

Technical Field

[0001] The present invention relates to the field of fire management technology, and specifically to an intelligent fire management system and method based on AI analysis. Background Art

[0002] With the acceleration of urbanization and the increase in building density, fire, as a common disaster, is occurring more frequently and causing more losses. Traditional fire monitoring and emergency response methods mainly rely on manual inspections and static fire risk assessments. These methods often have problems such as delayed response and low efficiency. After a fire occurs, timely acquisition of accurate fire risk information, prediction of fire spread trends and rapid response become the key to reducing fire losses and protecting people's lives and property. Therefore, smart fire management methods based on artificial intelligence (AI) technology have emerged, aiming to improve the accuracy, efficiency and response speed of fire management through real-time data collection, dynamic analysis and intelligent decision-making.

[0003] In addition, the limitations of existing technologies include at least the following problems. First, most existing smart fire protection systems rely on traditional risk assessment methods, which only consider parameters at a specific point in time, while ignoring the continuous impact of dynamic and changing environmental factors on fire risks, resulting in a relatively static judgment of fire risks and difficulty in accurately predicting the spread trend of fires. Second, fire risk assessments in existing technologies are mostly limited to a single basic data indicator and fail to fully integrate multi-dimensional influencing factors, such as the interaction of multiple factors such as air humidity, smoke concentration, and carbon monoxide concentration. This can easily lead to fire risk assessment results that are not comprehensive and accurate, making it difficult to respond to changes in sudden fire incidents in a timely manner. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention provides an intelligent fire management system and method based on AI analysis, which solves the static and single problems of fire assessment in the existing technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent fire management method based on AI analysis, comprising the following steps: real-time acquisition of basic fire risk data and fire modified risk data for each monitored area, and pre-processing, wherein the basic fire risk data includes regional air temperature value, regional air humidity value, regional smoke concentration value, and regional carbon monoxide concentration value; comprehensive analysis of the pre-processed basic fire risk data of each monitored area to obtain a basic fire risk index for each monitored area, which is calculated as follows: Among them, HjC i is the basic fire risk index of the i-th monitored area, KqW iis the regional air temperature value of the i-th monitored area, YwN i is the regional smoke concentration value of the i-th monitored area, ω1 is the smoke concentration influence coefficient stored in the database, KqS i is the regional air humidity value of the i-th monitored area, ω2 is the air humidity influence coefficient stored in the database, YyH i is the regional carbon monoxide concentration value of the i-th monitored area, ω3 is the carbon monoxide concentration influence coefficient stored in the database, i = 1, 2, 3, ..., i0, i0 is the number of monitored areas; for each monitored area, the basic fire risk index is corrected and analyzed based on the pre-processed fire-corrected risk data to obtain the fire-corrected risk index of each monitored area, and the fire-corrected risk index is judged and analyzed with the preset fire risk assessment interval; the monitored area whose fire-corrected risk index is within the preset fire risk assessment interval is marked as a fire risk area, and power cut-off and alarm notification measures are taken.

[0006] Furthermore, the fire correction risk data includes the regional structural thermal response coefficient, the environmental abnormal sound signal intensity value, and the electrical abnormal electromagnetic disturbance intensity value. The specific steps for obtaining the fire correction risk index of each area to be monitored are as follows: obtaining the structural thermal expansion stress activity value of each area to be monitored; reading the basic fire risk index of each area to be monitored, and performing a comprehensive analysis based on the regional structural thermal response coefficient, environmental abnormal sound signal intensity value, electrical abnormal electromagnetic disturbance intensity value, and structural thermal expansion stress activity value of the corresponding area to be monitored to obtain the fire correction risk index of each area to be monitored.

[0007] Furthermore, the specific steps for obtaining the active value of the structural thermal expansion stress of each area to be monitored are as follows: obtaining the number of acoustic emission occurrences in each area to be monitored; inputting the number of acoustic emission occurrences in each area to be monitored into a preset judgment model for judgment analysis to obtain the active value of the structural thermal expansion stress of each area to be monitored.

[0008] Furthermore, the specific formula for calculating the fire modified risk index of each monitored area is as follows: Among them, HxZ i is the fire risk index of the ith monitored area, HjC i is the basic fire risk index of the i-th monitored area, RxY i is the active value of the structural thermal expansion stress of the i-th monitored area, ε is the thermal expansion stress adjustment factor stored in the database, YcX i is the abnormal sound signal intensity value of the environment in the i-th monitored area, μ1 is the acoustic signal influence coefficient stored in the database, DcQ iis the electrical abnormal electromagnetic disturbance intensity value of the i-th area to be monitored, μ2 is the electromagnetic disturbance influence coefficient stored in the database, i = 1, 2, 3, ..., i0, i0 is the number of areas to be monitored.

[0009] Furthermore, for fire risk areas, fire growth trend analysis is performed based on a preset AI algorithm, and when the fire growth trend is abnormal, fire alarm processing is performed. The specific steps are as follows: for fire risk areas, the fire basic change time series index and the fire area change intensity time series index are analyzed separately, and a comprehensive analysis is performed to obtain the fire growth trend index of the fire risk area; the fire growth trend index of the fire risk area is judged and analyzed with the preset growth trend interval; and when the fire growth trend index of the fire risk area is within the preset growth trend interval, it is regarded that the fire growth trend of the fire risk area is abnormal, and fire alarm processing is performed.

[0010] Furthermore, the specific formula for calculating the fire growth trend index of the fire risk area is as follows: Among them, HqS is the fire growth trend index of the fire risk area, QbQ is the time series index of the fire regional change intensity of the fire risk area, JcB is the time series index of the fire regional change intensity of the fire risk area, λ1 is the regional change suppression coefficient stored in the database, α is the interaction influence coefficient stored in the database, and λ2 is the basic change reinforcement coefficient stored in the database.

[0011] Furthermore, the specific steps for analyzing the fire basic change time series index of the fire risk area are as follows: continuously obtain the regional air temperature values, regional air humidity values, regional smoke concentration values, regional carbon monoxide concentration values, and visible light intensity values of the fire risk area at several time points, and perform change rate analysis on each of them to obtain the regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index of the fire risk area; conduct a comprehensive analysis of the regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index of the fire risk area to obtain the fire basic change time series index of the fire risk area.

[0012] Furthermore, the specific formula for calculating the fire base change time series index of the fire risk area is as follows: Among them, QbQ is the time series index of the fire regional change intensity in the fire risk area, WbH is the regional air temperature change index in the fire risk area, ξ1 is the temperature change influence coefficient stored in the database, SbH is the regional air humidity change index in the fire risk area, ξ2 is the humidity change influence coefficient stored in the database, YbH is the regional smoke concentration change index in the fire risk area, ξ3 is the smoke concentration change influence coefficient stored in the database, TbH is the regional carbon monoxide concentration change index in the fire risk area, ξ4 is the carbon monoxide concentration change influence coefficient stored in the database, KbH is the visible light intensity change index in the fire risk area, and ξ5 is the visible light intensity change influence coefficient stored in the database.

[0013] Furthermore, the specific steps for analyzing the fire regional change intensity time series index of the fire risk area are as follows: continuously obtain the regional air disturbance intensity values, micro-vibration intensity values, smoke color values, and ground resistance values at several time points in the fire risk area, and perform change rate analysis on each of them to obtain the regional air disturbance intensity change index, micro-vibration intensity change index, smoke color change index, and ground resistance change index of the fire risk area; conduct a comprehensive analysis of the regional air disturbance intensity change index, micro-vibration intensity change index, smoke color change index, and ground resistance change index of the fire risk area to obtain the fire basic change time series index of the fire risk area.

[0014] A smart fire management system based on AI analysis includes: a risk data acquisition unit, which is used to acquire the basic fire risk data and fire correction risk data of each monitored area in real time and perform pre-processing, wherein the basic fire risk data includes regional air temperature value, regional air humidity value, regional smoke concentration value, and regional carbon monoxide concentration value; a basic fire risk analysis unit, which is used to perform comprehensive analysis on the pre-processed basic fire risk data of each monitored area to obtain a basic fire risk index for each monitored area; a basic fire risk judgment unit, which is used to perform correction analysis on the basic fire risk index of each monitored area based on the pre-processed fire correction risk data to obtain a fire correction risk index for each monitored area, and perform judgment analysis with a preset fire risk assessment interval, and mark the monitored area whose fire correction risk index is within the preset fire risk assessment interval as a fire risk area, and take power cut-off and alarm notification measures; a fire growth trend analysis unit, which is used to perform fire growth trend analysis on the fire risk area based on a preset AI algorithm, and perform fire alarm processing when the fire growth trend is abnormal.

[0015] The present invention has the following beneficial effects:

[0016] (1) This intelligent fire management method based on AI analysis can accurately reflect the changing trend of fire risk by combining real-time basic fire risk data, such as regional air temperature, humidity, smoke concentration and carbon monoxide concentration, and corrected risk data, such as structural thermal response, environmental abnormal sound signals, micro-vibration intensity, etc. for dynamic calculation. Traditional fire risk assessment often relies on a single, static indicator, which makes risk judgment delayed after a fire occurs and fails to respond effectively in time. This method can more comprehensively assess the changes in fire risk and respond quickly by monitoring various environmental data in real time and deriving a fire risk index through comprehensive analysis. Specifically, when the fire risk index of a certain area exceeds the preset danger threshold, it can automatically trigger a fire alarm and take power-off measures. This mechanism avoids the delay of traditional manual intervention and improves the efficiency and accuracy of fire warning. By dynamically adjusting and updating the fire risk assessment results, it ensures that the most accurate risk data can be obtained in time at each stage of fire development, minimizing the spread of fire.

[0017] (2) This intelligent fire management method based on AI analysis significantly improves the accuracy of fire risk assessment by combining multiple physical parameters in the environment, such as structural thermal expansion stress activity value, abnormal audio signals, electrical abnormal electromagnetic disturbance and other multi-dimensional data for analysis. Specifically, the introduction of these parameters can identify potential early signals of fire. For example, the structural thermal expansion stress activity value reflects the thermal expansion of the material, micro-vibration signals can help identify signs of damage to the building structure due to fire, and electrical abnormal electromagnetic disturbance can serve as an early warning of electrical fire. These factors are rarely considered in traditional fire assessment systems. This method can accurately identify the occurrence and development trend of fire in real time through the fusion analysis of these factors, significantly improving the accuracy of fire risk prediction. Through multi-dimensional dynamic monitoring and fusion, it can warn of fire more timely and accurately, providing more response time for the fire department, thereby effectively avoiding the spread and diffusion of fire and reducing manpower and property losses.

[0018] (3) The intelligent fire management method based on AI analysis can dynamically track the changes in the fire area and calculate the fire growth trend index through the fire growth trend analysis based on AI technology. The index combines the fire area change intensity time series index and the fire base change time series index. Through the dynamic analysis of these indices, the development trend of the fire can be predicted and the possible spread trajectory and development speed of the fire can be identified in advance. For example, in the early stage of the fire, by continuously monitoring the changes in temperature, humidity, smoke concentration and other environmental parameters, it can be predicted whether the fire will show explosive growth. If the fire growth trend index exceeds the preset threshold, an alarm will be automatically triggered and emergency response measures will be implemented. This intelligent prediction based on fire growth trend can not only automatically judge the expansion of the fire, but also deploy emergency response in advance when the fire just occurs, greatly improving the efficiency and response speed of fire handling, and reducing the delays and omissions of traditional manual operations. In addition, through fire trend prediction, the fire department can make more comprehensive resource allocation and personnel arrangements in advance, improve the organization and effectiveness of firefighting operations, and optimize the overall emergency management process.

[0019] (4) The intelligent fire management system based on AI analysis has significantly improved the automation level of fire warning and emergency response by integrating AI analysis and multiple intelligent units. The risk data acquisition unit of the system obtains the basic fire risk data and fire correction risk data of each monitored area in real time, including basic data such as regional air temperature, humidity, smoke concentration and carbon monoxide concentration, as well as correction data such as structural thermal response, abnormal sound signal and electromagnetic interference. The basic fire risk analysis unit and the basic fire risk judgment unit process, pre-process and comprehensively analyze the data to accurately calculate the fire risk index of each area. When the risk index exceeds the safety range, the system automatically marks the area as a fire risk area and automatically triggers power cut-off and alarm notification measures to reduce the need for human intervention and improve the speed and accuracy of response. In addition, the fire growth trend analysis unit in the system uses AI algorithms to intelligently analyze the growth trend of the fire area based on real-time monitoring data and historical fire development data. The unit can accurately predict the development direction and speed of the fire, immediately trigger fire alarm processing when the fire growth trend is abnormal, and take corresponding emergency response measures. Through multi-dimensional dynamic monitoring and automated decision-making, the system can quickly identify and initiate responses in the early stages of a fire, greatly optimizing the deployment of emergency resources and personnel arrangements.

[0020] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flow chart of the intelligent fire management method based on AI analysis in the present invention.

[0022] Figure 2 This is a flowchart of the specific steps for obtaining the fire correction risk index of each monitored area in an intelligent fire management method based on AI analysis of the present invention.

[0023] Figure 3 This is a block diagram of an intelligent fire management system based on AI analysis in the present invention. DETAILED DESCRIPTION

[0024] See also Figure 1 , an embodiment of the present invention provides a technical solution: an intelligent fire management method based on AI analysis, comprising the following steps: real-time acquisition of basic fire risk data and corrected fire risk data of each area to be monitored, and pre-processing of the data, the basic fire risk data including regional air temperature value, regional air humidity value, regional smoke concentration value, and regional carbon monoxide concentration value; comprehensive analysis of the pre-processed basic fire risk data of each area to be monitored to obtain a basic fire risk index for each area to be monitored; for each area to be monitored, correction analysis of the basic fire risk index is performed based on the pre-processed corrected fire risk data to obtain a corrected fire risk index for each area to be monitored, and judgment analysis is performed with a preset fire risk assessment interval; the area to be monitored whose corrected fire risk index is within the preset fire risk assessment interval is marked as a fire risk area, and power cut-off and alarm notification measures are taken, that is, power cut-off of all electrical appliances in the fire risk area, and a fire warning is sent to relevant staff.

[0025] The specific formula for calculating the basic fire risk index of each monitored area is as follows: Among them, HjC i is the basic fire risk index of the i-th monitored area, KqW i is the regional air temperature value of the i-th monitored area, YwN i is the regional smoke concentration value of the i-th monitored area, ω1 is the smoke concentration influence coefficient stored in the database, KqS i is the regional air humidity value of the i-th monitored area, ω2 is the air humidity influence coefficient stored in the database, YyH i is the regional carbon monoxide concentration value of the i-th area to be monitored, ω3 is the carbon monoxide concentration influence coefficient stored in the database, i=1, 2, 3, ..., i0, i0 is the number of areas to be monitored.

[0026] It should be explained that the specific steps for obtaining the smoke concentration influence coefficient ω1, air humidity influence coefficient ω2, and carbon monoxide concentration influence coefficient ω3 stored in the database are: collecting corresponding real-time data from the actual monitoring system, and comparing and analyzing these data with the historical database, and further calculating these coefficients. These coefficients are determined through model learning and associated with the specific environmental conditions of the area (such as temperature, humidity, etc.) and historical fire data, so that they are continuously updated through data-driven methods. Through the role of these coefficients, the specific impact of different environmental factors on fire risks can be adjusted to ensure the accuracy of fire risk assessment.

[0027] Specifically, if Figure 2 As shown, the fire correction risk data includes the regional structural thermal response coefficient, the environmental abnormal sound signal intensity value, and the electrical abnormal electromagnetic disturbance intensity value. The specific steps for obtaining the fire correction risk index of each area to be monitored are as follows: obtain the structural thermal expansion stress activity value of each area to be monitored; read the fire basic risk index of each area to be monitored, and perform a comprehensive analysis based on the regional structural thermal response coefficient, environmental abnormal sound signal intensity value, electrical abnormal electromagnetic disturbance intensity value, and structural thermal expansion stress activity value of the corresponding area to be monitored to obtain the fire correction risk index of each area to be monitored.

[0028] Among them, the regional structural thermal response coefficient reflects the temperature rise rate and hysteresis response degree of the wall / ceiling / ground in the area. The higher the value, the material is easy to accumulate heat but the temperature rise is slow, and the fire may develop more covertly. The regional structural thermal response coefficient can be obtained by calculating the temperature difference sampled per second through a high-frequency infrared thermal imager / surface patch temperature sensor.

[0029] The abnormal environmental sound signal intensity value refers to the energy intensity of characteristic sounds related to fire in the area (such as arc bursts, glass shattering, and equipment burning sounds) within a specific frequency range, specifically the sound intensity in the target frequency band (such as 2kHz-8kHz).

[0030] The electrical abnormal electromagnetic disturbance intensity value reflects whether there are obvious electrical system abnormalities in the area, such as sudden electromagnetic signal intensity caused by arcing, short circuit, overload discharge, etc. It can read the target frequency band dB value in real time through SDR (software defined radio) and analyze the signal difference between two consecutive moments to obtain the electrical abnormal electromagnetic disturbance intensity value.

[0031] The specific formula for calculating the fire modified risk index for each monitored area is as follows: Among them, HxZ i is the fire risk index of the ith monitored area, HjC i is the basic fire risk index of the i-th monitored area, RxY iis the active value of the structural thermal expansion stress of the i-th monitored area, ε is the thermal expansion stress adjustment factor stored in the database, which is used to prevent the denominator from being zero, YcX i is the abnormal sound signal intensity value of the environment in the i-th monitored area, μ1 is the acoustic signal influence coefficient stored in the database, DcQ i is the electrical abnormal electromagnetic disturbance intensity value of the i-th area to be monitored, μ2 is the electromagnetic disturbance influence coefficient stored in the database, i = 1, 2, 3, ..., i0, i0 is the number of areas to be monitored.

[0032] It needs to be explained that the specific steps for obtaining the acoustic signal influence coefficient μ1 and the electromagnetic disturbance influence coefficient μ2 stored in the database are: by real-time monitoring of the acoustic signals and electromagnetic signals in the environment and comparing them with historical data, first, by deploying acoustic sensors (such as microphone arrays) to collect the sound signal intensity in the area in real time, analyze and extract characteristic signals, and calculate the change in acoustic signal intensity; then, use electromagnetic interference (EMI) detectors to monitor the electromagnetic signal anomalies emitted by electrical equipment in real time, and identify possible faults or arcing phenomena in the electrical system. Then, these real-time data will be compared with the historical acoustic and electromagnetic data in the database. Through data learning and model optimization, the acoustic signal influence coefficient and the electromagnetic disturbance influence coefficient will be calculated, thereby obtaining an adjustment effect on the fire risk assessment.

[0033] In this implementation plan, by introducing multiple physical parameters, such as structural thermal response coefficient, environmental abnormal sound signal intensity value and electrical abnormal electromagnetic disturbance intensity value, the accuracy and timeliness of fire risk assessment are significantly improved. Traditional fire risk assessment mostly relies on basic data such as temperature and humidity, but ignores complex factors such as structural changes, equipment failures and sound signals. Through real-time monitoring and analysis of these factors, potential fire hazards can be identified in the early stages of a fire, such as the thermal response characteristics of building materials, electrical abnormal signals of equipment and abnormal sound frequencies in the environment, thereby making more accurate corrections to fire risks. Specifically, the regional structural thermal response coefficient can reveal the changes in the material after being heated. The response speed provides a predictive basis for the development of hidden fires; the intensity of abnormal environmental sound signals can help identify abnormal phenomena such as equipment failure or explosion sounds during the fire process, and promptly reflect changes in the fire; the intensity of abnormal electrical electromagnetic disturbances can detect potential faults in the electrical system, such as arcs or short circuits, thereby providing early warning of the outbreak of fire. By comparing with historical data and intelligent learning, it can dynamically optimize fire risk assessment, avoiding the limitations of traditional methods. Taking all these factors into consideration, fire risk assessment is more comprehensive, real-time and accurate, thereby effectively improving the response speed and accuracy of the fire prevention and control system, and helping to quickly and effectively take emergency measures to reduce fire losses.

[0034] Specifically, the specific steps for obtaining the active value of structural thermal expansion stress in each area to be monitored are as follows: obtaining the number of acoustic emission (cracking sound) occurrences in each area to be monitored, which reflects whether the structure frequently experiences thermal stress damage phenomena such as microcracks and peeling in a short period of time, and represents whether the building is subjected to high thermal shock; inputting the number of acoustic emission (cracking sound) occurrences in each area to be monitored into a preset judgment model for judgment analysis to obtain the active value of structural thermal expansion stress in each area to be monitored.

[0035] The preset judgment model is specifically: Where RzY is the active value of the structural thermal expansion stress, x is the number of acoustic emission (cracking sound) occurrences, and CsY is the preset threshold value of the number of acoustic emission (cracking sound) occurrences.

[0036] In this implementation scheme, by monitoring and analyzing the number of occurrences of acoustic emission (crackling sounds), the changes in the structural thermal expansion stress of the building under the high temperature of the fire can be accurately reflected, thereby identifying in advance the structural damage that may occur during the fire. Under the action of high heat, the materials of the building may crack, peel off, etc. These tiny cracking sounds can be captured in real time by sensors. The number of occurrences of these cracking sounds is input into a preset judgment model, and the active value of the structural thermal expansion stress can be quantitatively analyzed, thereby identifying the structural risks brought by the fire. This method can effectively predict the changes in the building structure during a fire, discover hidden dangers in time, and avoid major structural damage or collapse caused by the fire. At the same time, the combination of acoustic emission data and intelligent models ensures accurate monitoring of the thermal response of the building, provides a more scientific and reliable decision-making basis for fire prevention and control, and improves the predictability and emergency response capabilities of fire safety management.

[0037] Specifically, for fire risk areas, fire growth trend analysis is performed based on a preset AI algorithm, and fire alarm processing is performed when the fire growth trend is abnormal. The specific steps are as follows: for fire risk areas, the fire basic change time series index and the fire area change intensity time series index are analyzed separately, and a comprehensive analysis is performed to obtain the fire growth trend index of the fire risk area; the fire growth trend index of the fire risk area is judged and analyzed with the preset growth trend interval; and when the fire growth trend index of the fire risk area is within the preset growth trend interval, it is regarded that the fire growth trend of the fire risk area is abnormal, and fire alarm processing is performed.

[0038] The specific formula for calculating the fire growth trend index of a fire risk area is as follows: Among them, HqS is the fire growth trend index of the fire risk area, QbQ is the fire area change intensity time series index of the fire risk area, JcB is the fire area change intensity time series index of the fire risk area, λ1 is the regional change suppression coefficient stored in the database, α is the interaction influence coefficient stored in the database (used to adjust the interaction between the fire area change intensity time series index and the fire basic change time series index, and determine the amplification amplitude of the trend index), and λ2 is the basic change reinforcement coefficient stored in the database.

[0039] It should be explained that the specific steps for obtaining the regional change suppression coefficient λ1, interaction influence coefficient α, and basic change reinforcement coefficient λ2 stored in the database are as follows: the regional change suppression coefficient calculates the suppression effect of the region under different conditions by analyzing the relationship between environmental changes in the region (such as temperature, humidity, wind speed and other factors) and fire risks. Then, the interaction influence coefficient is calculated by analyzing the interaction between different environmental factors (such as the joint impact of humidity and smoke concentration on fire). Finally, the basic change reinforcement coefficient is based on the impact of basic changes in the regional environment (such as temperature changes) on the fire growth trend, and is obtained through training with historical fire data. The calculation of these coefficients is based on the interaction between environmental change data and fire progress data in the region to help adjust the prediction of fire growth trends.

[0040] In this implementation, through the fire growth trend analysis based on AI algorithm, the dynamic changes of fire spread can be accurately predicted, which significantly improves the accuracy and timeliness of fire warning. The system first calculates the fire growth trend index by analyzing the fire basic change time series index and the fire area change intensity time series index, combined with the changes in environmental factors in the area (such as temperature, humidity, wind speed, etc.). This index can not only reflect the current state of the fire, but also predict the changing trend of the fire in the future, and identify the potential risk of fire spread in advance. By combining the regional change suppression coefficient, the interaction influence coefficient and the basic change reinforcement coefficient, this method can better reflect the complex interaction in the process of fire growth. Interactions, for example, the combined effects of humidity and smoke concentration may lead to the acceleration of fire spread, while the suppressive effect of regional environmental changes on fire may delay the development of fire. These coefficients make the prediction of fire growth trend more accurate and dynamic through learning and optimization of historical data. When the fire growth trend index is in the preset abnormal range, the system can immediately start fire alarm processing and automatically take emergency response measures, such as cutting off power supply and notifying relevant personnel. This greatly improves the efficiency of fire emergency response, reduces the lag of human operation, ensures rapid and scientific emergency decision-making, minimizes the possibility of fire spread, and improves the response speed and accuracy of fire safety management.

[0041] Specifically, the specific steps for analyzing the fire basic change time series index of the fire risk area are as follows: continuously obtain the regional air temperature values, regional air humidity values, regional smoke concentration values, regional carbon monoxide concentration values, and visible light intensity values of the fire risk area at several time points, and perform change rate analysis on them respectively to obtain the regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index of the fire risk area; conduct a comprehensive analysis of the regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index of the fire risk area to obtain the fire basic change time series index of the fire risk area.

[0042] Among them, the visible light intensity value refers to the light intensity in the fire area, which is usually related to the size of the flame and the concentration of the smoke. During a fire, the flame will produce a large amount of visible light, so its intensity change can reflect the intensity and spread of the fire. In the early stage of a fire, the color of the flame is usually brighter. As the fire spreads and the smoke increases, the visible light intensity will decrease. Especially in different stages of a fire, the speed and degree of change in visible light intensity can help us infer the development of the fire. It can be measured and obtained through a light sensor (photoelectric sensor).

[0043] The specific formula for calculating the fire base change time series index of the fire risk area is as follows: Among them, QbQ is the time series index of the fire regional change intensity in the fire risk area, WbH is the regional air temperature change index in the fire risk area, ξ1 is the temperature change influence coefficient stored in the database, SbH is the regional air humidity change index in the fire risk area, ξ2 is the humidity change influence coefficient stored in the database, YbH is the regional smoke concentration change index in the fire risk area, ξ3 is the smoke concentration change influence coefficient stored in the database, TbH is the regional carbon monoxide concentration change index in the fire risk area, ξ4 is the carbon monoxide concentration change influence coefficient stored in the database, KbH is the visible light intensity change index in the fire risk area, and ξ5 is the visible light intensity change influence coefficient stored in the database.

[0044] It should be explained that the specific steps for obtaining the temperature change influence coefficient ξ1, humidity change influence coefficient ξ2, smoke concentration change influence coefficient ξ3, carbon monoxide concentration change influence coefficient ξ4, and visible light intensity change influence coefficient ξ5 stored in the database are as follows: the temperature change influence coefficient is obtained by comparing the temperature change data provided by the real-time temperature sensor with the historical data of temperature changes during past fires to calculate the specific impact of temperature changes on fire risks. Similarly, the humidity change influence coefficient is obtained by combining the humidity change data obtained by the humidity sensor with historical data to analyze the inhibitory effect of humidity on the spread of fire. The smoke concentration change influence coefficient is obtained by monitoring the changes in smoke concentration in real time through the smoke sensor, and combining it with historical fire data to evaluate the impact of smoke concentration on fire. The carbon monoxide concentration change influence coefficient is obtained by providing data from the CO concentration sensor, and the influence coefficient is obtained by analyzing the correlation between CO concentration and fire development. Finally, the visible light intensity change influence coefficient is obtained by monitoring the light intensity changes in the fire area in real time through the light sensor, and combining it with the light intensity change data in historical fires for analysis. Through these steps, each influence coefficient can be dynamically adjusted to accurately predict the growth trend and risk of fire.

[0045] The specific implementation example of calculating the fire base change time series index of the fire risk area is as follows, with the following parameters:

[0046] The regional air temperature variation index in the fire risk area is approximately: 0.125.

[0047] The temperature change influence coefficient stored in the database is approximately: 0.156.

[0048] The regional air humidity change index in the fire risk area is approximately: 0.081.

[0049] The humidity change influence coefficient stored in the database is approximately: 0.142.

[0050] The regional smoke concentration change index in the fire risk area is approximately: 0.257.

[0051] The smoke concentration change influence coefficient stored in the database is approximately: 0.134.

[0052] The regional carbon monoxide concentration change index in the fire risk area is approximately: 0.05.

[0053] The carbon monoxide concentration change influence coefficient stored in the database is approximately: 0.249.

[0054] The visible light intensity variation index in the fire risk area is approximately: 0.103.

[0055] The visible light intensity change influence coefficient stored in the database is approximately: 0.347.

[0056] Substituting the above data into the specific formula for calculating the fire base change time series index of the fire risk area, we obtain:

[0057] Fire base change time series index of fire risk area = exp(((0.125)^0.156)+((0.081)^0.142)+((0.257)^0.134)+((0.05)^0.249)+

[0058] ((0.103)^0.347))≈24.169.

[0059] In this implementation plan, by real-time monitoring of multiple environmental parameters and performing rate of change analysis, the basic fire change time series index of the fire risk area can be dynamically and accurately calculated, thereby providing a scientific basis for the prediction of fire growth trends. Traditional fire monitoring systems often rely on a single indicator for evaluation, while ignoring the dynamic changes of multiple environmental factors during the fire process. By comprehensively analyzing multi-dimensional data such as temperature changes, humidity changes, smoke concentration changes, carbon monoxide concentration changes and visible light intensity changes, we can fully grasp the various changes and mutual influences in the development of the fire, greatly improving the accuracy of fire risk prediction. Among them, the monitoring of visible light intensity values is particularly important because the light intensity of the flame is closely related to the intensity and spread of the fire. The fire risk is closely related to the situation. In the early stage of a fire, the flame is bright. As the smoke concentration increases, the visible light intensity will gradually decrease. This change can reflect the spread speed and scope of the fire. Combined with the data from the light sensor, it is possible to monitor the changes in the fire in real time and adjust the fire risk assessment in time. In addition, the influence coefficients of temperature, humidity, smoke concentration, carbon monoxide concentration and visible light intensity stored in the database are dynamically adjusted by comparing with historical fire data, so that the system can accurately predict the fire risk according to the characteristics of different fire environments. This comprehensive, multi-dimensional time series analysis of fire basic changes not only improves the accuracy of fire assessment, but also provides more reliable data support for timely response, thereby effectively reducing the losses caused by fire.

[0060] Specifically, the specific steps for analyzing the fire area change intensity time series index of the fire risk area are as follows: continuously obtain the regional air disturbance intensity values, micro-vibration intensity values, smoke color values, and ground resistance values at several time points in the fire risk area, and perform change rate analysis on them respectively to obtain the regional air disturbance intensity change index, micro-vibration intensity change index, smoke color change index, and ground resistance change index of the fire risk area; conduct a comprehensive analysis of the regional air disturbance intensity change index, micro-vibration intensity change index, smoke color change index, and ground resistance change index of the fire risk area to obtain the fire basic change time series index of the fire risk area.

[0061] The regional air disturbance intensity value refers to the intensity of air movement (wind speed) within the fire area. Wind speed has a significant impact on the spread of fire. The stronger the wind, the faster the flames and smoke spread, and the faster the fire spreads. Inside a building, if windows are broken or doors are open, air movement can affect the progress of the fire. Air disturbance intensity is used to measure the impact of these environmental factors on fire development, and it can be measured using a micro-anemometer.

[0062] The micro-vibration intensity value refers to the intensity of tiny vibrations generated during a fire, usually caused by factors such as fire explosion, equipment damage, and thermal expansion of the structure. During the development of a fire, when electrical equipment fails or building materials rupture due to high-temperature expansion, tiny vibrations will be generated. Micro-vibrations can help us detect the explosiveness of the fire or structural changes. For example, electrical explosions caused by fire and thermal cracking of building materials will cause obvious vibrations, which can be measured and obtained using MEMS accelerometers.

[0063] Smoke color value refers to the color change of smoke during a fire. The color change of smoke usually changes from light gray, yellow, and orange to dark gray or black. This is closely related to the severity and temperature of the fire. Color change often signals the development status of the fire. For example, yellow and orange smoke may represent the early stages of a fire, while red or black smoke usually means that the fire has entered a more serious stage. It can be measured and obtained through image processing (using a camera or thermal imaging camera to capture the fire area and using computer vision algorithms to analyze the changes in smoke color. Color analysis generally focuses on the RGB channels, especially the changes in the red channel).

[0064] The grounding resistance value refers to the grounding resistance of the electrical system. Electrical fires often start from electrical equipment or line failures, and current will flow through the grounding system. Abnormal grounding resistance mutations usually mean that there is a fault in the electrical system. During a fire, equipment damage may cause changes in the current passing through the grounding system, which in turn leads to a sudden change in grounding resistance, especially in the case of electrical short circuits and arc discharges. It can be measured using a resistance probe.

[0065] The specific formula for calculating the fire base change time series index of the fire risk area is as follows: Among them, JcB is the time series index of the fire regional change intensity in the fire risk area, KrB is the regional air disturbance intensity change index in the fire risk area, δ1 is the air flow change influence coefficient stored in the database, WzB is the micro-vibration intensity change index in the fire risk area, δ2 is the micro-vibration intensity change influence coefficient stored in the database, YwB is the smoke color change index in the fire risk area, δ3 is the smoke color change influence coefficient stored in the database, JdB is the ground resistance change index in the fire risk area, and δ4 is the ground resistance change influence coefficient stored in the database.

[0066] It should be explained that the specific steps for obtaining the air flow change influence coefficient δ1, micro-vibration intensity change influence coefficient δ2, smoke color change influence coefficient δ3, and ground resistance change influence coefficient δ4 stored in the database are as follows: the air flow change influence coefficient is determined by analyzing the real-time data of wind speed or air flow in the area and combining it with the influence relationship of wind speed changes in historical fire data. The micro-vibration intensity change influence coefficient depends on the micro-vibration intensity data obtained by the vibration sensor, and compares it with the vibration pattern and fire progress relationship in the historical data. The smoke color change influence coefficient is captured in real time by image processing technology. The color change of smoke is calculated in combination with the records of the impact of smoke color changes on fire in the database. Finally, the ground resistance change influence coefficient obtains the data of ground resistance changes through the ground resistance sensor, and analyzes the effect of ground resistance changes on fire in historical data. All these coefficients rely on the combination of real-time monitoring and historical data, and are continuously optimized through data learning and model training.

[0067] This implementation, through comprehensive analysis of multiple environmental factors, enables a more accurate assessment of the dynamics of fire risk areas and the prediction of fire spread trends. Traditional fire monitoring systems typically rely solely on basic data such as temperature and humidity, while ignoring the impact of factors such as wind speed, microvibration, and smoke color changes on fire spread. By real-time monitoring and analyzing multidimensional data such as regional air disturbance intensity, microvibration intensity, smoke color changes, and ground resistance changes, this method can comprehensively reflect the multiple factors influencing fires. For example, wind speed changes directly affect the speed of fire spread, while microvibration can reveal potential risks of structural thermal expansion and equipment failure. Smoke color changes provide immediate information on the development stage of a fire, and ground resistance changes help detect potential faults in electrical systems and prevent the occurrence of electrical fires. Combining this real-time data with the influence coefficients in the database and continuously optimizing the intelligent model, fire risk assessments can be dynamically adjusted, providing more efficient and accurate fire warnings and emergency responses. This multi-level, multi-factor dynamic monitoring and analysis improves the accuracy of the fire warning system and the timeliness of emergency response, effectively reducing the time window for fire spread and providing a more reliable decision-making basis for fire prevention and control.

[0068] See also Figure 3 An embodiment of the present invention provides a technical solution: an intelligent fire management system based on AI analysis, comprising: a risk data acquisition unit, configured to acquire in real time basic fire risk data and modified fire risk data of each area to be monitored, and perform pre-processing, the basic fire risk data including regional air temperature value, regional air humidity value, regional smoke concentration value, and regional carbon monoxide concentration value; a basic fire risk analysis unit, configured to perform comprehensive analysis on the pre-processed basic fire risk data of each area to be monitored, to obtain a basic fire risk index for each area to be monitored; a basic fire risk judgment unit, configured to perform a correction analysis on the basic fire risk index of each area to be monitored based on the pre-processed modified fire risk data, to obtain a modified fire risk index for each area to be monitored, and to perform judgment analysis on the basic fire risk index of each area to be monitored, respectively, compared with a preset fire risk assessment interval, and mark the area to be monitored whose modified fire risk index is within the preset fire risk assessment interval as a fire risk area, and take power cut-off and alarm notification measures; a fire growth trend analysis unit, configured to perform fire growth trend analysis on the fire risk area based on a preset AI algorithm, and perform fire alarm processing when the fire growth trend is abnormal.

[0069] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0070] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A smart fire management method based on AI analysis, characterized in that: The following steps are involved: Obtaining the basic fire risk data and the modified fire risk data of each monitored area in real time and performing pre-processing. The basic fire risk data includes regional air temperature value, regional air humidity value, regional smoke concentration value, and regional carbon monoxide concentration value; The pre-processed fire basic risk data of each monitored area are comprehensively analyzed to obtain the fire basic risk index of each monitored area. The calculation formula is as follows: Among them, HjC i 、KqW i 、YwN i 、KqS i 、YyH i are the basic fire risk index, regional air temperature, regional smoke concentration, regional air humidity, and regional carbon monoxide concentration of the i-th monitored area, respectively. ω1, ω2, and ω3 are the smoke concentration influence coefficient, air humidity influence coefficient, and carbon monoxide concentration influence coefficient stored in the database, respectively. i = 1, 2, 3, ..., i0, where i0 is the number of monitored areas. For each area to be monitored, the basic fire risk index is corrected and analyzed based on the pre-processed fire risk correction data to obtain the fire risk correction index of each area to be monitored, and then compared with the preset fire risk assessment interval for judgment analysis; The monitored areas whose fire correction risk index is within the preset fire risk assessment range are marked as fire risk areas, and power cut-off and alarm notification measures are taken.

2. The intelligent fire management method based on AI analysis according to claim 1 is characterized in that: The fire risk correction data includes the regional structural thermal response coefficient, the environmental abnormal sound signal intensity value, and the electrical abnormal electromagnetic disturbance intensity value. The specific steps for obtaining the fire risk correction index of each monitored area are as follows: Obtain the active value of structural thermal expansion stress in each area to be monitored; The basic fire risk index of each area to be monitored is read, and a comprehensive analysis is performed based on the regional structural thermal response coefficient, environmental abnormal sound signal intensity value, electrical abnormal electromagnetic disturbance intensity value, and structural thermal expansion stress activity value of the corresponding area to be monitored to obtain the fire modified risk index of each area to be monitored.

3. The intelligent fire management method based on AI analysis according to claim 2 is characterized in that: The specific steps for obtaining the active value of the structural thermal expansion stress in each monitored area are as follows: Obtain the number of acoustic emission occurrences in each area to be monitored; The number of acoustic emission occurrences in each area to be monitored is input into a preset judgment model for judgment analysis to obtain the active value of structural thermal expansion stress in each area to be monitored.

4. The intelligent fire management method based on AI analysis according to claim 2 is characterized in that: The specific formula for calculating the fire modified risk index for each monitored area is as follows: Among them, HxZ i 、HjC i 、RxY i 、YcX i 、DcQ i They are, respectively, the fire modified risk index, fire basic risk index, structural thermal expansion stress activity value, environmental abnormal acoustic signal intensity value, and electrical abnormal electromagnetic disturbance intensity value of the i-th monitored area. ε is the thermal expansion stress adjustment factor stored in the database. μ1 and μ2 are, respectively, the acoustic signal influence coefficient and electromagnetic disturbance influence coefficient stored in the database. i = 1, 2, 3, …, i0, where i0 is the number of areas to be monitored.

5. The intelligent fire management method based on AI analysis according to claim 1 is characterized in that: For fire risk areas, the system analyzes fire growth trends based on a preset AI algorithm and initiates fire alarm processing when the fire growth trend becomes abnormal. The specific steps are as follows: For fire risk areas, the fire base change time series index and the fire area change intensity time series index are analyzed separately, and a comprehensive analysis is performed to obtain the fire growth trend index of the fire risk area; Analyze the fire growth trend index of the fire risk area and the preset growth trend range; When the fire growth trend index in the fire risk area is within the preset growth trend range, it is regarded that the fire growth trend in the fire risk area is abnormal, and a fire alarm is processed.

6. The intelligent fire management method based on AI analysis according to claim 5 is characterized in that: The specific formula for calculating the fire growth trend index of a fire risk area is as follows: Among them, HqS, QbQ, and JcB are the fire growth trend index, fire regional change intensity time series index, and fire regional change intensity time series index of the fire risk area, respectively. λ1, α, and λ2 are the regional change suppression coefficient, interaction influence coefficient, and basic change reinforcement coefficient stored in the database, respectively.

7. The intelligent fire management method based on AI analysis according to claim 5 is characterized in that: The specific steps for analyzing the time series index of fire base change in fire risk areas are as follows: Continuously obtain the regional air temperature values, regional air humidity values, regional smoke concentration values, regional carbon monoxide concentration values, and visible light intensity values at several time points in the fire risk area, and perform change rate analysis on each of them to obtain the regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index of the fire risk area; A comprehensive analysis is conducted on the regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index in the fire risk area to obtain the fire basic change time series index of the fire risk area.

8. The intelligent fire management method based on AI analysis according to claim 7 is characterized in that: The specific formula for calculating the fire base change time series index of the fire risk area is as follows: Among them, QbQ, WbH, SbH, YbH, TbH, and KbH are the fire area change intensity time series index, regional air temperature change index, regional air humidity change index, regional smoke concentration change index, regional carbon monoxide concentration change index, and visible light intensity change index of the fire risk area respectively; ξ1, ξ2, ξ3, ξ4, and ξ5 are the temperature change influence coefficient, humidity change influence coefficient, smoke concentration change influence coefficient, carbon monoxide concentration change influence coefficient, and visible light intensity change influence coefficient stored in the database respectively.

9. The intelligent fire management method based on AI analysis according to claim 5 is characterized in that: The specific steps for analyzing the time series index of fire area change intensity in fire risk areas are as follows: Continuously obtain the regional air disturbance intensity values, microvibration intensity values, smoke color values, and ground resistance values at several time points in the fire risk area, and perform change rate analysis on each value to obtain the regional air disturbance intensity change index, microvibration intensity change index, smoke color change index, and ground resistance change index in the fire risk area; A comprehensive analysis is conducted on the regional air disturbance intensity change index, microvibration intensity change index, smoke color change index, and ground resistance change index in the fire risk area to obtain the fire basic change time series index in the fire risk area.

10. An intelligent fire management system based on AI analysis, applying the intelligent fire management method based on AI analysis according to any one of claims 1 to 9, characterized in that: include: A risk data acquisition unit is used to acquire the basic fire risk data and the modified fire risk data of each monitored area in real time and perform pre-processing. The basic fire risk data includes the regional air temperature value, the regional air humidity value, the regional smoke concentration value, and the regional carbon monoxide concentration value; The basic fire risk analysis unit is used to comprehensively analyze the pre-processed basic fire risk data of each monitored area to obtain the basic fire risk index of each monitored area; A basic fire risk judgment unit is configured to perform a correction analysis on the basic fire risk index for each area to be monitored based on the pre-processed fire correction risk data, obtain a fire correction risk index for each area to be monitored, and perform judgment analysis on each area with a preset fire risk assessment interval. Areas to be monitored whose fire correction risk index falls within the preset fire risk assessment interval are marked as fire risk areas, and power supply and alarm notification measures are implemented. The fire growth trend analysis unit is used to analyze the fire growth trend of fire risk areas based on a preset AI algorithm, and to perform fire alarm processing when the fire growth trend becomes abnormal.

Images