Multi-source fire data fusion and fire analysis and prediction system

Abstract

The invention relates to the technical field of fire monitoring and prediction, in particular to a multi-source fire data fusion and fire analysis and prediction system which comprises a data acquisition module, a data preprocessing module, a data fusion module, a fire analysis module and an early warning module. Wherein the data acquisition module is used for acquiring fire data from various sources; the data preprocessing module is used for preprocessing the fire data; the data fusion module is used for carrying out fusion processing on the multi-source data output by the data preprocessing module; the fire analysis module is used for identifying the current situation of the fire and the position of a fire source and predicting the spreading direction of the fire; and the early warning module is used for sending out an early warning signal in time. According to the invention, through the multi-source fire data fusion and intelligent analysis and prediction technology, the occurrence, spreading and risk level of the fire can be accurately monitored in real time, so that the timeliness and accuracy of fire emergency response are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fire monitoring and prediction, and in particular to a multi-source fire data fusion and fire situation analysis and prediction system. Background Art

[0002] With the acceleration of urbanization and the improvement of industrialization, fire, as a serious public emergency, poses a huge threat to the safety of human society; especially under extreme climatic conditions such as high temperature, high humidity and wind speed, the fire spreads quickly, bringing great challenges to emergency rescue; most of the existing fire monitoring systems rely on a single data source for fire detection, such as temperature and humidity sensors, smoke sensors, etc., but these single sensors cannot comprehensively and accurately evaluate the occurrence and spread of fire; in addition, traditional fire prediction methods are mostly based on historical data, ignoring the dynamic changes of real-time environmental data, resulting in untimely emergency response after the fire occurs and missing the best time to deal with it.

[0003] The existing fire monitoring and prediction systems have the following problems: First, most systems rely on limited data sources and fail to fully integrate data from multiple sources such as temperature and humidity, smoke, wind speed, and air pressure, making it difficult to fully grasp the entire process of fire occurrence and spread; second, the existing fire prediction methods are mostly based on static models, and cannot dynamically adjust the prediction results in real time, resulting in low prediction accuracy of fire spread speed and direction, and unable to accurately guide emergency response; finally, the warning mechanism of the existing system is too simple, lacks refined fire risk assessment, and cannot calculate the probability of fire occurrence and the possibility of spread in real time. Therefore, there is an urgent need for a multi-source fire data fusion and fire analysis and prediction system to solve the above problems and protect people's lives and property. Summary of the invention

[0004] Based on the above objectives, the present invention provides a multi-source fire data fusion and fire situation analysis and prediction system.

[0005] The multi-source fire data fusion and fire situation analysis and prediction system includes a data acquisition module, a data preprocessing module, a data fusion module, a fire situation analysis module and an early warning module; among which: Data collection module: used to collect fire data from multiple sources, including temperature and humidity sensor data, smoke sensor data, real-time environmental data and historical fire data; Data preprocessing module: used to receive fire data from the data acquisition module and perform preprocessing operations on it, including denoising, standardization and missing value filling, and output the cleaned multi-source data for use by subsequent modules; Data fusion module: used to fuse the multi-source data output by the data preprocessing module, using data fusion algorithm to integrate data information from different data sources into a unified data representation, and output fused data for use by the fire analysis module; Fire situation analysis module: Based on the fusion data output by the data fusion module, the Bayesian network algorithm is used to perform real-time analysis on the current fire situation, identify the current status and location of the fire source, predict the direction of fire spread, and output the fire risk prediction results; Early warning module: It is used to issue early warning signals in a timely manner and notify relevant personnel to take measures based on the fire risk prediction results output by the fire situation analysis module and the set fire risk threshold.

[0006] Optionally, the data acquisition module includes a temperature and humidity sensor data acquisition unit, a smoke sensor data acquisition unit, a real-time environment data acquisition unit, and a historical fire data acquisition unit; wherein: Temperature and humidity sensor data acquisition unit: used to collect environmental temperature and humidity information and convert it into digital signals; the temperature and humidity sensor is connected to each monitoring point at the fire scene and transmits data in real time by wireless or wired means; Smoke sensor data acquisition unit: used to monitor the smoke concentration changes at the fire scene and convert them into digital signals; Real-time environmental data acquisition unit: used to collect real-time environmental data, including wind speed, air pressure and temperature; Historical fire data acquisition unit: used to obtain relevant data of previous fire incidents from the database or historical record system, including temperature and humidity, smoke concentration, fire location and spread path information when the fire occurred.

[0007] Optionally, the data preprocessing module includes a denoising unit, a standardization unit and a missing value filling unit; wherein: De-noising unit: used to receive fire data from the data acquisition module and filter out noise in the fire data; the noise includes abnormal data caused by environmental interference or sensor error, and the de-noising unit uses a mean filter algorithm or a median filter algorithm to remove the noise data; Standardization unit: used to standardize the cleaned data from the denoising unit. The standardization operation includes normalizing each data item and scaling the data to a unified standard range, thereby eliminating the dimensional differences between different data sources. Missing value filling unit: used to process missing values ​​in the fire data from the data acquisition module, and replace the missing data with estimated values ​​using interpolation algorithms or mean filling methods to ensure that the processed data is complete and without missing values.

[0008] Optionally, the data fusion module includes a data fusion algorithm unit and a data integration unit; wherein: Data fusion algorithm unit: used to receive multi-source data from the data preprocessing module, and use a weighted average data fusion algorithm to fuse various types of data; the weighted average assigns different weights to different data sources, and according to the importance of each data source, performs weighted calculation on the information of multiple data sources such as temperature and humidity sensor data, smoke sensor data, real-time environmental data and historical fire data to obtain a comprehensive fused data; Data integration unit: used to receive the fused data output by the data fusion algorithm unit and convert it into a unified data representation format.

[0009] Optionally, the data fusion algorithm unit includes: Calculate the weight of each data source: According to the importance of each data source, assign weights to different sensor data. The preset weights are , the calculation formula is: ,in, Indicates The weight of the data source, For the The variance of the data source; Normalize weights: To ensure that the sum of all weights is 1, the calculated weights are normalized. The normalization formula is: ,in, Represents the normalized The weight of the data source, is the total number of data sources; Weighted average calculation of fused data: according to the normalized weights , weighted average the observation values ​​of each data source to obtain comprehensive fusion data , the weighted average calculation formula is: ,in, represents the final fused data, For the Observations from a data source.

[0010] Optionally, the fire situation analysis module includes a fire state identification unit, a fire source location unit and a fire spread prediction unit; wherein: Fire status recognition unit: Based on the fusion data output by the data fusion module, the Bayesian network algorithm is used to build a probability model. By analyzing historical fire data and current environmental data, the fire probability and current fire status are inferred in real time, including whether the fire source exists and whether it spreads, and real-time fire status information is output; Fire source positioning unit: Based on the fire status information output by the fire status recognition unit and combined with the sensor data at the fire scene, the fire source is accurately located through the Bayesian network to infer the location with the highest probability of the fire source appearing; Fire spread prediction unit: Based on the fire source location and current environmental data inferred by the fire source positioning unit, it predicts the direction and speed of fire spread and outputs the fire spread prediction results for a period of time in the future.

[0011] Optionally, the fire status identification unit includes: Constructing a Bayesian network model: Based on the temperature and humidity sensor data, smoke sensor data, real-time environmental data, and historical fire data in the fused data, a Bayesian network model with conditional dependencies is constructed. The Bayesian network model includes multiple nodes, each of which represents a fire-related random variable, including the location of the fire source, the fire state, or the environmental conditions, and defines its edge connections based on the conditional dependencies between the random variables. Inferring the probability of fire occurrence: Through the inference mechanism of the Bayesian network, the probability of fire occurrence is inferred in real time based on the input current fire environment data and historical fire data; is a random variable for fire occurrence, and the currently known environmental conditions are , then the conditional probability of fire occurrence is: ,in, is the prior probability of fire occurrence; Given the fire conditions, the current environmental data is observed The likelihood function of ; is the marginal probability of the current environmental data; Infer the fire status and whether the fire source exists: the probability of fire occurrence obtained based on the inference , through Bayesian reasoning, it is determined whether a fire is occurring and whether a fire source exists; if the probability of a fire occurring is higher than a set threshold, it is determined that a fire is occurring; Real-time update of fire status information: During actual operation, as new data is continuously input, the fire status recognition unit updates the conditional probability distribution in the Bayesian network in real time, and promptly adjusts the probability prediction of fire occurrence and the judgment results of the fire source, thereby dynamically reflecting the current fire status.

[0012] Optionally, the fire source locating unit includes: Input fire state information: According to the fire state information provided by the fire state recognition unit, identify the potential location area of ​​the fire source, and use this area information as one of the input data of the Bayesian network; Collect sensor data: Combine multiple sensor data at the fire scene to obtain environmental parameters at each point of the fire scene. The sensor data is represented as a random variable at each node in the network, specifically set to Indicates Data collected by sensors; Establish a Bayesian network model: Through the Bayesian network algorithm, a probability model of the fire source location is constructed based on the fire scene sensor data and fire status information; the Bayesian network represents the probability distribution of the fire source location through nodes and calculates the conditional probability of the fire source location; the probability model of the fire source location is expressed as: ; in, Fire source location Given the known sensor data The joint probability distribution under ; Indicates that the fire source is The probability of a position; Indicates the location of the fire source The relevant parent node; Indicates Data collected by sensors; Infer the location with the highest probability of the fire source: Based on the joint probability distribution of the fire source location calculated by the probability model of the fire source location, infer the area with the highest probability of the fire source appearing, that is, maximize the following conditional probability, expressed as: ,in, Indicates the location where the fire source is most likely to appear. Indicates the location of the fire Conditional probability given sensor data.

[0013] Optionally, the fire spread prediction unit includes: Application of heat conduction model: The spread of fire is affected by the heat conduction effect. The conduction model is used to predict the temperature change in the area around the fire source. The expression of the heat conduction model is: ,in, is temperature; For time; is the thermal diffusion coefficient; is the Laplace operator of temperature, which represents the spatial variation of temperature; Application of wind speed model: Combined with real-time environmental data of wind speed and air pressure, the wind speed model is used to predict the speed of fire spread; the wind speed model is based on the relationship between wind speed and fire spread, and uses the following formula to calculate the fire spread speed: ,in, The speed of fire spread; is the wind speed in the current environment; is the temperature of the fire area; is the ambient temperature; is a constant coefficient; Determine the direction of fire spread: The direction of fire spread is determined by the results of the wind speed model and the temperature gradient of the heat conduction model. The calculation formula is: ,in, is the angle of fire spread; and Respectively represent the wind speed in Axis and Axis components; combined with the fire spread speed and direction angle, the fire spread prediction unit outputs the fire spread prediction results for a period of time in the future, including the fire spread speed, direction and affected area.

[0014] Optionally, the early warning module includes a fire risk assessment unit, a threshold judgment unit and a notification unit; wherein: Fire risk assessment unit: used to comprehensively assess the current fire risk level based on the fire risk prediction results output by the fire situation analysis module. The fire risk prediction results include fire status information, fire spread speed and fire source location, and use the weighted summation method to determine the comprehensive fire risk level ; Threshold judgment unit: According to the set fire risk threshold, it judges whether the fire risk output by the fire risk assessment unit exceeds the predetermined threshold. The risk threshold is set to ;when When the fire is detected, it is judged to be in a high-risk state and the early warning mechanism is triggered; Notification unit: When the threshold judgment unit determines that the fire risk exceeds the set threshold, the notification unit sends a warning signal to relevant personnel or the control center according to the preset notification method. The preset notification method includes SMS, phone, mobile application push and email, and informs the specific fire location, risk level and emergency response measures.

[0015] Beneficial effects of the present invention: The present invention, by fusing temperature and humidity sensors, smoke sensors, real-time environmental data and historical fire data, can comprehensively and accurately monitor the occurrence and spread of fires. Through the processing and integration of multi-source data by the data fusion module, the system can achieve more accurate fire status identification and fire source positioning, improve the estimation accuracy of the probability of fire occurrence, and effectively reduce the data deviation problem faced by traditional single data source systems. This technical solution of multi-source data fusion ensures that the system can respond to dynamic changes in various stages of the fire in a timely manner, effectively improving the comprehensive performance and reliability of fire monitoring.

[0016] The present invention, by combining the Bayesian network algorithm with fire spread prediction, can predict the spread direction and speed of fire based on real-time environmental data and historical fire data, and calculate the fire risk level in real time. This technology makes fire risk assessment more accurate, can issue early warning signals in time according to the set risk threshold, and quickly inform relevant personnel to take emergency measures through the notification system, effectively improving the timeliness of fire emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A schematic diagram of a fire situation analysis and prediction system according to an embodiment of the present invention; Figure 2 Schematic diagram of a fire situation analysis module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0020] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0021] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0022] like Figure 1-Figure 2As shown, the multi-source fire data fusion and fire situation analysis and prediction system includes a data acquisition module, a data preprocessing module, a data fusion module, a fire situation analysis module and an early warning module; wherein: Data collection module: used to collect fire data from multiple sources, including temperature and humidity sensor data, smoke sensor data, real-time environmental data and historical fire data; Data preprocessing module: used to receive fire data from the data acquisition module and perform preprocessing operations on it, including denoising, standardization and missing value filling, and output the cleaned multi-source data for use by subsequent modules; Data fusion module: used to fuse the multi-source data output by the data preprocessing module, using data fusion algorithm to integrate data information from different data sources into a unified data representation, and output fused data for use by the fire analysis module; Fire situation analysis module: Based on the fusion data output by the data fusion module, the Bayesian network algorithm is used to perform real-time analysis on the current fire situation, identify the current status and location of the fire source, predict the direction of fire spread, and output the fire risk prediction results; Early warning module: It is used to issue early warning signals in a timely manner and notify relevant personnel to take measures based on the fire risk prediction results output by the fire situation analysis module and the set fire risk threshold.

[0023] The data acquisition module includes a temperature and humidity sensor data acquisition unit, a smoke sensor data acquisition unit, a real-time environment data acquisition unit, and a historical fire data acquisition unit; wherein: Temperature and humidity sensor data acquisition unit: used to collect environmental temperature and humidity information and convert it into digital signals; the temperature and humidity sensors are connected to various monitoring points at the fire scene and transmit data in real time via wireless or wired methods; Smoke sensor data acquisition unit: used to monitor the smoke concentration changes at the fire scene, convert them into digital signals, and transmit them to the data preprocessing module; Real-time environmental data acquisition unit: used to collect real-time environmental data, including wind speed, air pressure and temperature, and send the collected environmental data to the data preprocessing module; Historical fire data acquisition unit: used to obtain relevant data of previous fire events from the database or historical record system, including temperature and humidity, smoke concentration, fire location and spread path information when the fire occurred, for subsequent analysis; through the collaborative work of the above units, the data acquisition module can comprehensively and real-time collect fire-related data from different sensors and data sources, and provide the required raw data for the subsequent data preprocessing module.

[0024] The data preprocessing module includes a denoising unit, a standardization unit, and a missing value filling unit; among which: De-noising unit: used to receive fire data from the data acquisition module and filter out the noise in the fire data; the noise includes abnormal data caused by environmental interference or sensor error. The de-noising unit uses a mean filter algorithm or a median filter algorithm to remove the noise data, thereby ensuring the accuracy and reliability of the data; Standardization unit: used to standardize the cleaned data from the denoising unit. The standardization operation includes normalizing each data item and scaling the data to a unified standard range, thereby eliminating the dimensional differences between different data sources and improving the accuracy of subsequent data fusion and analysis. Missing value filling unit: used to process missing values ​​in the fire data from the data acquisition module, and replace the missing data with estimated values ​​using interpolation algorithms or mean filling methods to ensure that the processed data is complete and without missing values, thereby avoiding the impact of missing data on subsequent analysis results; through the above detailed preprocessing operations, the data preprocessing module can effectively clean and normalize the fire data from various data sources, providing high-quality data support for subsequent fusion and analysis.

[0025] The data fusion module includes a data fusion algorithm unit and a data integration unit; wherein: Data fusion algorithm unit: used to receive multi-source data from the data preprocessing module, and use the weighted average data fusion algorithm to fuse various types of data; weighted average assigns different weights to different data sources, and according to the importance of each data source, performs weighted calculation on the information of multiple data sources such as temperature and humidity sensor data, smoke sensor data, real-time environmental data and historical fire data to obtain a comprehensive fusion data; Data integration unit: used to receive the fused data output by the data fusion algorithm unit and convert it into a unified data representation format. The data representation format has a standardized structure, which is convenient for subsequent modules to perform efficient processing and analysis. The above unit adopts a data fusion algorithm unit using a weighted average method. It can optimize the effect of multi-source data fusion according to the reliability and weights of different data sources, ensure that the fused data is more accurate and stable, and suitable for subsequent fire analysis and prediction. The weighted average method effectively solves the impact of different information in multi-source data on the analysis results, and provides reliable data support for the system.

[0026] The data fusion algorithm unit includes: Calculate the weight of each data source: According to the importance of each data source, assign weights to different sensor data (including temperature and humidity sensor data, smoke sensor data, real-time environmental data, historical fire data, etc.). The preset weights are , the calculation formula is: ,in, Indicates The weight of the data source, For the The smaller the variance, the higher the reliability of the data source, and the corresponding weight The bigger; Normalize weights: To ensure that the sum of all weights is 1, the calculated weights are normalized. The normalization formula is: ,in, Represents the normalized The weight of the data source, is the total number of data sources; Weighted average calculation of fused data: according to the normalized weights , weighted average the observation values ​​of each data source to obtain comprehensive fusion data , the weighted average calculation formula is: ,in, represents the final fused data, For the The observation values ​​of each data source; through the steps of the weighted average method mentioned above, appropriate weights can be assigned according to the reliability of each data source, and the fire data from different sources can be effectively fused to generate accurate comprehensive data. This method can make full use of the advantages of each data source and improve the accuracy of data fusion.

[0027] The fire situation analysis module includes a fire status identification unit, a fire source location unit and a fire spread prediction unit; wherein: Fire status recognition unit: Based on the fusion data output by the data fusion module, the Bayesian network algorithm is used to build a probability model. By analyzing historical fire data and current environmental data (such as temperature, humidity, smoke concentration, etc.), the probability of fire occurrence and the current fire status are inferred in real time, including whether the fire source exists and whether it spreads, and real-time fire status information is output; Fire source positioning unit: Based on the fire status information output by the fire status recognition unit and combined with the sensor data at the fire scene, the fire source is accurately located through the Bayesian network to infer the location with the highest probability of the fire source appearing; Fire spread prediction unit: Based on the fire source location and current environmental data inferred by the fire source positioning unit, the fire spread direction and speed are predicted, and the fire spread prediction results for a period of time in the future are output; through the collaborative work of the above units, the fire status analysis, fire source location and fire spread prediction can be performed efficiently and accurately based on the fused data, thereby providing real-time and reliable data support for fire emergency response, and improving the accuracy of fire prediction and the timeliness of response.

[0028] The fire status recognition unit includes: Constructing a Bayesian network model: Based on the temperature and humidity sensor data, smoke sensor data, real-time environmental data, and historical fire data in the fused data, a Bayesian network model with conditional dependencies is constructed. The Bayesian network model includes multiple nodes, each of which represents a fire-related random variable, including the location of the fire source, the fire state, or the environmental condition, and defines its edge connection based on the conditional dependency between the random variables; the structure of the Bayesian network is expressed as: ,in, is the joint probability distribution of all random variables in the network; Indicates A random variable, such as is the location of the fire source; It is a fire state; For the fire to spread; for other fire-related variables; Indicates The parent node of the node, that is, Other variables with conditional dependencies, such as the fire source location depends on environmental conditions such as temperature and humidity, and the fire state depends on the fire source location and historical fire data; Indicates that at a given parent node Conditions, No. Nodes The conditional probability of represents the probability of fire occurrence; Inferring the probability of fire occurrence: Through the inference mechanism of the Bayesian network, the probability of fire occurrence is inferred in real time based on the input current fire environment data (such as temperature and humidity, smoke concentration, wind speed, etc.) and historical fire data; is a random variable for fire occurrence, and the currently known environmental conditions are , then the conditional probability of fire occurrence is: ,in, is the prior probability of fire occurrence, which represents the initial belief that fire occurs; Given the fire conditions, the current environmental data is observed The likelihood function of ; is the marginal probability of the current environmental data; Infer the fire status and whether the fire source exists: the probability of fire occurrence obtained based on the inference , through Bayesian reasoning, it is determined whether a fire is occurring and whether a fire source exists; if the probability of a fire occurring is higher than a set threshold, it is determined that a fire is occurring; Real-time update of fire status information: During the actual operation process, as new data is continuously input, the fire status recognition unit updates the conditional probability distribution in the Bayesian network in real time, and promptly adjusts the probability prediction of fire occurrence and the judgment result of fire source, thereby dynamically reflecting the current fire status; through the above steps, the fire status recognition unit can use the Bayesian network algorithm to perform efficient fire occurrence probability inference based on the real-time collected environmental data and historical data, accurately judge whether a fire has occurred, whether a fire source exists and its spread, thereby providing a scientific decision-making basis for fire emergency response.

[0029] The fire source location unit includes: Input fire state information: According to the fire state information provided by the fire state recognition unit, identify the potential location area of ​​the fire source, determine the suspicious area of ​​the fire scene (for example, the area with high smoke density or drastic temperature change), and use this area information as one of the input data of the Bayesian network; Collect sensor data: Combine multiple sensor data at the fire scene (including smoke sensor data, temperature and humidity sensor data, real-time environmental data, etc.) to obtain environmental parameters at each point in the fire scene. The sensor data is represented as a random variable at each node in the network, specifically set to Indicates Data collected by sensors (such as temperature, humidity, etc.); Establish a Bayesian network model: Through the Bayesian network algorithm, a probability model of the fire source location is constructed based on the fire scene sensor data and fire status information; the Bayesian network represents the probability distribution of the fire source location through nodes and calculates the conditional probability of the fire source location; the probability model of the fire source location is expressed as: ; in, Fire source location Given the known sensor data The joint probability distribution under ; Indicates that the fire source is the probability of different locations of the fire (e.g. the fire source may be located at different locations in the monitoring area); Indicates the location of the fire source The relevant parent node reflects the impact of environmental conditions (such as temperature, humidity, etc.) on the location of the fire source; Indicates Data collected by sensors (such as temperature, humidity, smoke concentration, etc.); Infer the location with the highest probability of the fire source: Based on the joint probability distribution of the fire source location calculated by the probability model of the fire source location, infer the area with the highest probability of the fire source appearing, that is, maximize the following conditional probability, expressed as: ,in, Indicates the location where the fire source is most likely to appear. Indicates the location of the fire Conditional probability under given sensor data; through the above steps, the fire source positioning unit can use the Bayesian network to accurately infer the fire source location based on the fire state information provided by the fire state identification unit and the sensor data at the fire scene, thereby ensuring the accuracy of fire source positioning and providing a reliable basis for subsequent fire spread prediction and emergency response.

[0030] The fire spread prediction unit includes: Application of heat conduction model: The spread of fire is affected by the heat conduction effect. The conduction model is used to predict the temperature change in the area around the fire source. The expression of the heat conduction model is: ,in, is temperature; For time; is the thermal diffusion coefficient; is the Laplace operator of temperature, which represents the spatial variation of temperature. Through this model, the temperature distribution of the area around the fire source can be calculated, and then the initial speed and direction of the fire spread can be inferred. Application of wind speed model: Combined with real-time environmental data of wind speed and air pressure, the wind speed model is used to predict the speed of fire spread; the wind speed model is based on the relationship between wind speed and fire spread, and uses the following formula to calculate the fire spread speed: ,in, The speed of fire spread; is the wind speed in the current environment; is the temperature of the fire area; is the ambient temperature; is a constant coefficient, which indicates the influence of wind speed on the spread of fire. Through this model, the spread speed of fire can be predicted; Determine the direction of fire spread: The direction of fire spread is determined by the results of the wind speed model and the temperature gradient of the heat conduction model. The calculation formula is: ,in, is the angle of fire spread; and Respectively represent the wind speed in Axis and Axis components; combined with the fire spread speed and direction angle, the fire spread prediction unit outputs the fire spread prediction results for a period of time in the future, including the fire spread speed, direction and affected area.

[0031] The early warning module includes a fire risk assessment unit, a threshold judgment unit and a notification unit; wherein: Fire risk assessment unit: used to comprehensively assess the current fire risk level based on the fire risk prediction results output by the fire situation analysis module. The fire risk prediction results include fire status information, fire spread speed and fire source location, and use the weighted summation method to determine the comprehensive fire risk level ; The comprehensive fire risk assessment formula is: ,in, is the comprehensive fire risk level; The probability of fire occurrence indicates the possibility of fire occurrence, usually ranging from 0 to 1; The fire spread speed indicates the speed at which the fire spreads from the fire source to other areas, in meters per second; is the distance between the fire source and sensitive areas (such as buildings, flammable items, etc.), in meters; is the weight of each factor, indicating the importance of each factor in the comprehensive risk assessment, and The weight value is set according to the actual application requirements; the comprehensive risk level of fire is obtained by weighted summation of various factors. , according to which the fire danger can be quantitatively assessed; Threshold judgment unit: According to the set fire risk threshold, it judges whether the fire risk output by the fire risk assessment unit exceeds the predetermined threshold. The risk threshold is set to , the calculation formula is: ,in, For the fire risk factors (such as fire probability, spread speed, distance from fire source, etc.); is the weight of the corresponding factor, indicating the impact of the factor on the overall risk; The total number of factors considered in risk assessment; When the fire is detected, it is judged to be in a high-risk state and the early warning mechanism is triggered; Notification unit: When the threshold judgment unit determines that the fire risk exceeds the set threshold, the notification unit sends a warning signal to relevant personnel or the control center according to the preset notification method. The preset notification method includes SMS, phone, mobile application push and email, and informs the specific fire location, risk level and emergency response measures; through the above technical solution, the warning module can accurately judge the degree of fire danger based on real-time fire risk analysis results and issue warning signals in a timely manner, helping relevant personnel to take timely measures and effectively reduce the losses caused by fire.

[0032] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0033] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Multi-source fire data fusion and fire situation analysis and prediction system, characterized by: It includes data acquisition module, data preprocessing module, data fusion module, fire situation analysis module and early warning module; among which: Data collection module: used to collect fire data from multiple sources, including temperature and humidity sensor data, smoke sensor data, real-time environmental data and historical fire data; Data preprocessing module: used to receive fire data from the data acquisition module and perform preprocessing operations on it, including denoising, standardization and missing value filling, and output the cleaned multi-source data for use by subsequent modules; Data fusion module: used to fuse the multi-source data output by the data preprocessing module, using data fusion algorithm to integrate data information from different data sources into a unified data representation, and output fused data for use by the fire analysis module; Fire situation analysis module: Based on the fusion data output by the data fusion module, the Bayesian network algorithm is used to perform real-time analysis on the current fire situation, identify the current status and location of the fire source, predict the direction of fire spread, and output the fire risk prediction results; Early warning module: It is used to issue early warning signals in a timely manner and notify relevant personnel to take measures based on the fire risk prediction results output by the fire situation analysis module and the set fire risk threshold.

2. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 1 is characterized in that: The data acquisition module includes a temperature and humidity sensor data acquisition unit, a smoke sensor data acquisition unit, a real-time environment data acquisition unit and a historical fire data acquisition unit; wherein: Temperature and humidity sensor data acquisition unit: used to collect environmental temperature and humidity information and convert it into digital signals; the temperature and humidity sensor is connected to each monitoring point at the fire scene and transmits data in real time by wireless or wired means; Smoke sensor data acquisition unit: used to monitor the smoke concentration changes at the fire scene and convert them into digital signals; Real-time environmental data acquisition unit: used to collect real-time environmental data, including wind speed, air pressure and temperature; Historical fire data acquisition unit: used to obtain relevant data of previous fire incidents from the database or historical record system, including temperature and humidity, smoke concentration, fire location and spread path information when the fire occurred.

3. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 1 is characterized in that: The data preprocessing module includes a denoising unit, a standardization unit and a missing value filling unit; wherein: De-noising unit: used to receive fire data from the data acquisition module and filter out noise in the fire data; the noise includes abnormal data caused by environmental interference or sensor error, and the de-noising unit uses a mean filter algorithm or a median filter algorithm to remove the noise data; Standardization unit: used to standardize the cleaned data from the denoising unit. The standardization operation includes normalizing each data item and scaling the data to a unified standard range, thereby eliminating the dimensional differences between different data sources. Missing value filling unit: used to process missing values ​​in the fire data from the data acquisition module, and replace the missing data with estimated values ​​using interpolation algorithms or mean filling methods to ensure that the processed data is complete and without missing values.

4. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 1 is characterized in that: The data fusion module includes a data fusion algorithm unit and a data integration unit; wherein: Data fusion algorithm unit: used to receive multi-source data from the data preprocessing module, and use a weighted average data fusion algorithm to fuse various types of data; the weighted average assigns different weights to different data sources, and according to the importance of each data source, performs weighted calculation on the information of multiple data sources such as temperature and humidity sensor data, smoke sensor data, real-time environmental data and historical fire data to obtain a comprehensive fused data; Data integration unit: used to receive the fused data output by the data fusion algorithm unit and convert it into a unified data representation format.

5. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 4 is characterized in that: The data fusion algorithm unit comprises: Calculate the weight of each data source: According to the importance of each data source, assign weights to different sensor data. The preset weights are , the calculation formula is: ,in, Indicates The weight of the data source, For the The variance of the data source; Normalize weights: To ensure that the sum of all weights is 1, the calculated weights are normalized. The normalization formula is: ,in, Represents the normalized The weight of the data source, is the total number of data sources; Weighted average calculation of fused data: according to the normalized weights , weighted average the observation values ​​of each data source to obtain comprehensive fusion data , the weighted average calculation formula is: ,in, represents the final fused data, For the Observations from a data source.

6. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 1 is characterized in that: The fire situation analysis module includes a fire state identification unit, a fire source location unit and a fire spread prediction unit; wherein: Fire status recognition unit: Based on the fusion data output by the data fusion module, the Bayesian network algorithm is used to build a probability model. By analyzing historical fire data and current environmental data, the fire probability and current fire status are inferred in real time, including whether the fire source exists and whether it spreads, and real-time fire status information is output; Fire source positioning unit: Based on the fire status information output by the fire status recognition unit and combined with the sensor data at the fire scene, the fire source is accurately located through the Bayesian network to infer the location with the highest probability of the fire source appearing; Fire spread prediction unit: Based on the fire source location and current environmental data inferred by the fire source positioning unit, it predicts the direction and speed of fire spread and outputs the fire spread prediction results for a period of time in the future.

7. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 6 is characterized in that: The fire state recognition unit comprises: Constructing a Bayesian network model: Based on the temperature and humidity sensor data, smoke sensor data, real-time environmental data, and historical fire data in the fused data, a Bayesian network model with conditional dependencies is constructed. The Bayesian network model includes multiple nodes, each of which represents a fire-related random variable, including the location of the fire source, the fire state, or the environmental conditions, and defines its edge connections based on the conditional dependencies between the random variables. Inferring the probability of fire occurrence: Through the inference mechanism of the Bayesian network, the probability of fire occurrence is inferred in real time based on the input current fire environment data and historical fire data; is a random variable for fire occurrence, and the currently known environmental conditions are , then the conditional probability of fire occurrence is: ,in, is the prior probability of fire occurrence; Given the fire conditions, the current environmental data is observed The likelihood function of ; is the marginal probability of the current environmental data; Infer the fire status and whether the fire source exists: the probability of fire occurrence obtained based on the inference , through Bayesian reasoning, it is determined whether a fire is occurring and whether a fire source exists; if the probability of a fire occurring is higher than a set threshold, it is determined that a fire is occurring; Real-time update of fire status information: During actual operation, as new data is continuously input, the fire status recognition unit updates the conditional probability distribution in the Bayesian network in real time, and promptly adjusts the probability prediction of fire occurrence and the judgment results of the fire source, thereby dynamically reflecting the current fire status.

8. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 7 is characterized in that: The fire source locating unit comprises: Input fire state information: According to the fire state information provided by the fire state recognition unit, identify the potential location area of ​​the fire source, and use this area information as one of the input data of the Bayesian network; Collect sensor data: Combine multiple sensor data at the fire scene to obtain environmental parameters at each point of the fire scene. The sensor data is represented as a random variable at each node in the network, specifically set to Indicates Data collected by sensors; Establish a Bayesian network model: Through the Bayesian network algorithm, a probability model of the fire source location is constructed based on the fire scene sensor data and fire status information; the Bayesian network represents the probability distribution of the fire source location through nodes and calculates the conditional probability of the fire source location; the probability model of the fire source location is expressed as: ； in, Fire source location Given the known sensor data The joint probability distribution under ; Indicates that the fire source is The probability of a position; Indicates the location of the fire source The relevant parent node; Indicates Data collected by sensors; Infer the location with the highest probability of the fire source: Based on the joint probability distribution of the fire source location calculated by the probability model of the fire source location, infer the area with the highest probability of the fire source appearing, that is, maximize the following conditional probability, expressed as: ,in, Indicates the location where the fire source is most likely to appear. Indicates the location of the fire Conditional probability given sensor data.

9. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 8 is characterized in that: The fire spread prediction unit comprises: Application of heat conduction model: The spread of fire is affected by the heat conduction effect. The conduction model is used to predict the temperature change in the area around the fire source. The expression of the heat conduction model is: ,in, is temperature; For time; is the thermal diffusion coefficient; is the Laplace operator of temperature, which represents the spatial variation of temperature; Application of wind speed model: Combined with real-time environmental data of wind speed and air pressure, the wind speed model is used to predict the speed of fire spread; the wind speed model is based on the relationship between wind speed and fire spread, and uses the following formula to calculate the fire spread speed: ,in, The speed of fire spread; is the wind speed in the current environment; is the temperature of the fire area; is the ambient temperature; is a constant coefficient; Determine the direction of fire spread: The direction of fire spread is determined by the results of the wind speed model and the temperature gradient of the heat conduction model. The calculation formula is: ,in, is the angle of fire spread; and Respectively represent the wind speed in Axis and Axis components; combined with the fire spread speed and direction angle, the fire spread prediction unit outputs the fire spread prediction results for a period of time in the future, including the fire spread speed, direction and affected area.

10. The multi-source fire data fusion and fire situation analysis and prediction system according to claim 1, characterized in that: The early warning module includes a fire risk assessment unit, a threshold judgment unit and a notification unit; wherein: Fire risk assessment unit: used to comprehensively assess the current fire risk level based on the fire risk prediction results output by the fire situation analysis module. The fire risk prediction results include fire status information, fire spread speed and fire source location, and use the weighted summation method to determine the comprehensive fire risk level ; Threshold judgment unit: According to the set fire risk threshold, it judges whether the fire risk output by the fire risk assessment unit exceeds the predetermined threshold. The risk threshold is set to ;when When the fire is detected, it is judged to be in a high-risk state and the early warning mechanism is triggered; Notification unit: When the threshold judgment unit determines that the fire risk exceeds the set threshold, the notification unit sends a warning signal to relevant personnel or the control center according to the preset notification method. The preset notification method includes SMS, phone, mobile application push and email, and informs the specific fire location, risk level and emergency response measures.

Images