Intelligent household fire automatic alarm and escape guidance system

Abstract

The invention relates to the technical field of fire monitoring and alarming, in particular to an intelligent household fire automatic alarming and escape guidance system, which comprises a time-sensitive escape window evaluation module, a time-sensitive escape window evaluation module, an escape window evaluation module and an escape guidance module, wherein the time-sensitive escape window evaluation module is based on fire source temperature, smoke concentration, humidity and oxygen content collected by a fire detection sensor network in combination with a fire source position; the severity of a current fire and the influence of the current fire on an escape window are evaluated in real time, and preliminary escape time early warning is provided; the fire spreading trend prediction module is used for predicting a fire spreading path based on fire detection sensor network data and a machine learning algorithm and generating a dynamic risk map; and the self-adaptive escape path optimization module is used for generating a plurality of escape paths with different priorities, optimizing the paths in real time according to the fire change, and ensuring that the escape paths are always in a low-risk area. According to the invention, it is ensured that family members can obtain timely alarm and action guidance in an escape time window, and the escape success rate is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fire monitoring and alarm, and in particular to an intelligent household fire automatic alarm and escape guidance system. Background Art

[0002] With the development of smart home and Internet of Things technology, home fire alarm and escape guidance systems are gradually being integrated into modern home environments, providing residents with more efficient fire warning and escape support. However, traditional fire alarm systems have some limitations, especially in terms of adaptive response to dynamic fire development and escape needs, which are specifically manifested in the following aspects: Traditional home fire alarm systems are mostly based on smoke and temperature detectors, which sound an alarm when a fire occurs. However, such systems usually only provide fixed alarm signals and cannot dynamically generate escape routes or adjust escape times based on changes in the location of the fire source, the speed of fire spread, and the severity of the fire. This static alarm mechanism may lead to inaccurate calculations of escape time and cannot cope with changes in the location of the fire source and the speed of fire spread, causing inconvenience to residents' escape decisions and even safety hazards.

[0003] Most of the current fire alarm systems are based on simple risk area divisions, lacking the prediction of fire spread paths and dynamic updates of high-risk areas. Fire spread has complex spatial and temporal dependencies and cannot be accurately represented by simple preset maps or static risk maps. During a fire, the location of the fire source, the direction of fire spread and the risk area are constantly changing. If residents follow the static risk distribution for escape guidance, they may choose paths in high-risk areas, resulting in escape failure.

[0004] Most existing systems only provide escape guidance for a single route and do not have the ability to optimize the route in real time. Once the route is blocked by the fire source or the risk level increases, residents may not be able to obtain alternative routes or adjust their escape plans in time, resulting in escape failure. Summary of the invention

[0005] The invention provides an intelligent household fire automatic alarm and escape guidance system.

[0006] Intelligent household fire automatic alarm and escape guidance system, including: Time-sensitive escape window assessment module: Based on the fire source temperature, smoke concentration, humidity and oxygen content collected by the fire detection sensor network, combined with the fire source location, it conducts real-time assessment of the severity of the current fire and its impact on the escape window, generates a dynamic change model of the escape time window size, and provides preliminary escape time warning; Fire spread trend prediction module: Based on the fire detection sensor network data and machine learning algorithms, it predicts the fire spread path and generates a dynamic risk map; Adaptive escape path optimization module: Based on the output results of the time-sensitive escape window assessment module and combined with the dynamic risk map, it generates multiple escape paths with different priorities and optimizes the paths in real time according to fire changes to ensure that the escape paths are always in low-risk areas.

[0007] Optionally, the time-sensitive escape window assessment module specifically includes: Data receiving unit: used to receive the fire source temperature collected by the fire detection sensor network , smoke concentration ,humidity and oxygen content ,The fire detection sensor network includes fire source temperature sensor, smoke sensor, and oxygen sensor; Fire trend analysis unit: based on the output results of the data receiving unit and combined with the location of the fire source , the speed of fire development, by analyzing the severity of the current fire, the spatial location of the fire source and the spread range, the severity of the current fire and fire development speed Conduct analysis; Dynamic escape window calculation unit: receives the output of the fire trend analysis unit, generates a dynamic escape time window size model based on the current fire situation, and calculates the escape time window and provide preliminary escape time warning .

[0008] Optionally, the fire severity The calculation is expressed as: ; in, is the maximum threshold of fire source temperature; is the maximum threshold of smoke concentration; is the maximum threshold of humidity; is the maximum oxygen content (21%), are weight coefficients for different parameters to ensure that the severity of the fire accurately reflects the impact of each parameter; is a minimum value, to prevent the denominator from being zero when the fire source and the personnel coincide with each other, and ; Indicates the location of the fire With personnel location spatial distance.

[0009] Optionally, the fire development speed The calculation is expressed as: ; in, is the weight coefficient of different parameters to ensure that the fire development speed can reasonably reflect the influence of various factors and meet ; is the temperature change rate, which indicates the rate of change of temperature over time; is the smoke concentration change rate, which indicates the rate of change of smoke concentration over time; is the humidity change rate, which indicates the rate of change of humidity over time; is the rate of change of oxygen content, which indicates the rate of change of oxygen content over time; It is the rate of change of fire source distance, which indicates the rate of change of the distance between the fire source and personnel over time.

[0010] Optionally, the escape time window It indicates the available time for personnel to escape from the fire source safely from the current time. It is calculated by combining the distance from the fire source to the escape area and the speed of fire development. The calculation is expressed as: ,in, is the initial safety distance, that is, the distance between the personnel position and the initial boundary of the fire source position, Indicates the current fire source distance, that is, the distance from the fire source to the personnel at the current time.

[0011] Optionally, the escape time warning :Determine whether to trigger the escape time warning based on the escape time window Wescape and the set minimum safety threshold Wthreshold.

[0012] Optionally, the minimum safety threshold Wthreshold is dynamically adjusted based on the fire severity R. When the fire severity R is too high, the minimum safety threshold Wthreshold is lowered so that an early warning is triggered within a short escape time window Wescape, thereby increasing sensitivity to severe fires.

[0013] Optionally, the fire spread trend prediction module specifically includes: Data preprocessing unit: Receives data collected by the fire detection sensor network, including temperature, smoke concentration, humidity, oxygen content and fire source location, and also receives the distance between the fire source location and the personnel , the data is normalized to form a structured fire status data set; Fire spread path prediction unit: Based on the machine learning algorithm, the fire status data set output by the data preprocessing unit is trained and predicted. By analyzing the time series changes of fire characteristics, a prediction model of the fire spread path is generated. The machine learning algorithm includes a CNN-LSTM combination model. CNN extracts the spatial characteristics of the fire status data. The time dependency of the features is modeled by LSTM to predict the fire spread path. CNN processes the input multi-dimensional fire status data and extracts the spatial characteristics of the data. Each convolution layer performs convolution operations on the collected temperature, humidity, smoke concentration, and oxygen content to generate a feature map to capture the spatial distribution information of the fire status. The feature map extracted by CNN is input into LSTM. LSTM processes the time series information of the feature data, captures the dynamic changes in the fire spread path, models the changing trend of fire characteristics over time, and predicts the future path of the fire expansion. Combined with the output of the CNN-LSTM combination model, the prediction of the future location of the fire source is obtained. ,in is the number of time steps for prediction; Dynamic risk map generation unit: receives the diffusion area output by the fire spread path prediction unit, maps the fire spread path to the home floor plan, generates a dynamic risk map, and Differentiate and display high-risk areas and relatively safe areas.

[0014] Optionally, the regional risk index The distance from each area to the fire source and the severity of the fire in each area Get, for map display: ,in, represents the regional risk index at the sth moment, is the predicted distance from the fire source to each area at the sth moment, is the minimum value that prevents the denominator from being zero.

[0015] Optionally, the adaptive escape path optimization module receives the escape time window Wescape and the dynamic risk map, and generates multiple initial escape paths of different priorities according to the current fire source location and its spread path, combined with the safe exit location in the home floor plan. , Indicates The generated escape paths contain continuous routes connected by multiple nodes (doors, corridors, exits, etc.); path priority Time required by route , Wescape, and the regional risk index on the path, giving priority to low-risk paths that can be completed within the escape time window; Path time (i.e. through the path The time required) is calculated as: ,in, For path The number of nodes, including doorways, corridors, and exits, For path Midpoint To Node The distance Through the node The speed of travel at Path priority Calculated as: ,in, represents the path at time s The sum of the regional risk indexes of all nodes above.

[0016] Beneficial effects of the present invention: The present invention uses a time-sensitive escape window assessment module to calculate a dynamic escape time window in real time according to the location of the fire source and the severity of the fire (temperature, smoke concentration, humidity, oxygen content and other parameters). The system can adjust the escape window at any time to adapt to the development speed of the fire and changes in the location of the fire source, thereby providing family members with real-time escape time warnings. This dynamic escape time management mechanism effectively avoids the problem of fixed preset escape time in traditional systems that does not adjust with changes in the fire, ensuring that family members can obtain timely alarms and action guidance within the escape time window, thereby improving the success rate of escape.

[0017] The present invention, through a fire spread trend prediction module, uses a machine learning model (such as CNN-LSTM) to predict the dynamic changes of the fire source location, and generates the future fire spread path and the distribution of risk areas. The generated dynamic risk map intuitively displays high-risk and relatively safe areas according to the fire source spread path, so that users can avoid high-risk areas and choose safer paths during the escape process. Compared with the limitations of traditional systems that rely on static risk maps, dynamic visualization of fire risks is achieved through machine learning algorithms, and the distribution of high-risk areas is updated in real time, which significantly improves the accuracy and adaptability of escape guidance.

[0018] In the present invention, the adaptive escape path optimization module combines the escape time window and the dynamic risk map to generate multiple escape paths with different priorities, and adjusts the paths in real time by real-time monitoring of changes in the fire source location and risk areas, effectively avoiding escape failures caused by blocking of a single path due to fire source, and ensuring that the escape path always remains relatively safe under changes in the fire environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] 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.

[0020] Figure 1 Schematic diagram of system function modules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0021] 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.

[0022] 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).

[0023] 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.

[0024] like Figure 1 As shown, the intelligent household fire automatic alarm and escape guidance system includes: Time-sensitive escape window assessment module: Based on the fire source temperature, smoke concentration, humidity and oxygen content collected by the fire detection sensor network, combined with the fire source location, it conducts real-time assessment of the severity of the current fire and its impact on the escape window, generates a dynamic change model of the escape time window size, and provides preliminary escape time warning; Fire spread trend prediction module: Based on the fire detection sensor network data and machine learning algorithms, it predicts the fire spread path and generates a dynamic risk map to display high-risk areas and relatively safe areas; Adaptive escape path optimization module: Based on the output results of the time-sensitive escape window assessment module and combined with the dynamic risk map, it generates multiple escape paths with different priorities and optimizes the paths in real time according to fire changes to ensure that the escape paths are always in low-risk areas.

[0025] The time-sensitive escape window assessment module specifically includes: Data receiving unit: used to receive the fire source temperature collected by the fire detection sensor network , smoke concentration ,humidity and oxygen content ,The fire detection sensor network includes fire source temperature sensor, smoke sensor, and oxygen sensor; Fire source temperature : The fire source temperature is collected in real time by the fire source temperature sensor near the fire source, indicating the temperature value of the fire source area; Smoke density : The smoke concentration is collected by the smoke sensor and indicates the smoke concentration per unit volume; humidity : Humidity is collected by humidity sensors to assess the water vapor content in the environment; Oxygen content : The oxygen content is collected by the oxygen sensor to determine the degree of oxygen consumption in the fire area; Fire location : The location of the fire source is determined by multiple position sensors, indicating the spatial coordinates of the fire source , using the triangulation positioning algorithm, combined with the smoke sensor and its concentration data and the fire source temperature sensor and its temperature data for positioning, the temperature or smoke concentration detected by the sensor near the fire source is higher, while the sensor farther away from the fire source has a lower detection value, thereby calculating the distance from the fire source to the smoke or fire source temperature sensor, knowing the distance between the fire source and multiple smoke or fire source temperature sensors, and using the triangulation positioning algorithm to find the fire source position; set The position coordinates of sensors are ,in , Each sensor records its own fire index (temperature or smoke density). The temperature or smoke density detected by the sensor near the fire source is higher, while the sensor farther away from the fire source has a lower detection value. The distance from the fire source to the sensor is expressed in the following way: : ; Use the relationship between temperature or concentration and distance to infer distance. For example, the decrease in temperature with distance is expressed as: ,in, For sensor The measured temperature, is the initial temperature of the fire source, from which the distance Estimates: , multiple measurements are needed to average out the errors, especially in the presence of environmental interference; Know the distance between the fire source and multiple sensors , use the triangulation algorithm to find the location of the fire source ; According to the principle of geometry, the location of the fire source The following set of equations should be satisfied: ; ;

[0026] ; By solving the above equations, the fire source location can be found .

[0027] Fire trend analysis unit: based on the output results of the data receiving unit and combined with the location of the fire source , the speed of fire development, by analyzing the severity of the current fire, the spatial location of the fire source and the spread range, the severity of the current fire and fire development speed Conduct analysis; Dynamic escape window calculation unit: receives the output of the fire trend analysis unit, generates a dynamic escape time window size model based on the current fire situation, and calculates the escape time window and provide preliminary escape time warning .

[0028] Fire severity The calculation is expressed as: ; in, is the maximum threshold of fire source temperature; is the maximum threshold of smoke concentration; is the maximum threshold of humidity; is the maximum oxygen content (21%), are weight coefficients for different parameters to ensure that the severity of the fire accurately reflects the impact of each parameter; is a minimum value, to prevent the denominator from being zero when the fire source and the personnel coincide with each other, and ; Indicates the location of the fire With personnel location Spatial distance: , multiple positioning sensors can be deployed inside the home, and micro positioning tags can be installed on the devices carried by residents (such as mobile phones, bracelets, etc.). These positioning tags communicate with sensors to obtain the location coordinates of family members when escaping from a fire, and update the location data in the system in real time.

[0029] Fire development speed The location of the fire source can also affect the speed of fire spread, especially in a spatial environment, where the speed of fire spread in different directions may be affected by the distance. The calculation is expressed as: ; in, is the weight coefficient of different parameters to ensure that the fire development speed can reasonably reflect the influence of various factors and meet ; is the temperature change rate, which indicates the rate of change of temperature over time; is the smoke concentration change rate, which indicates the rate of change of smoke concentration over time; is the humidity change rate, which indicates the rate of change of humidity over time; is the rate of change of oxygen content, which indicates the rate of change of oxygen content over time; It is the rate of change of fire source distance, which indicates the rate of change of the distance between the fire source and personnel over time.

[0030] Escape time window It indicates the available time for personnel to escape from the fire source safely from the current time. It is calculated by combining the distance from the fire source to the escape area and the speed of fire development. The calculation is expressed as: ,in, is the initial safety distance, that is, the distance between the personnel position and the initial boundary of the fire source position, Indicates the current fire source distance, that is, the distance from the fire source to the personnel at the current time.

[0031] Escape time warning :Determine whether to trigger the escape time warning according to the escape time window Wescape and the set minimum safety threshold Wthreshold: ; in, It is the safety threshold of the minimum escape time, ensuring that the system can issue timely warnings to personnel when the danger escalates.

[0032] As a comprehensive risk assessment indicator, the fire severity R can directly reflect the danger level of the fire. By combining the fire source temperature, smoke concentration, humidity, oxygen content and fire source location, the overall hazard of the fire can be determined. The value of the fire severity R can help the system decide whether to issue an early warning to family members. When R reaches a certain preset threshold, a high-priority escape notification can be triggered to remind family members to evacuate as soon as possible.

[0033] The minimum safety threshold Wthreshold is dynamically adjusted based on the fire severity R. When the fire severity R is too high, the minimum safety threshold Wthreshold is lowered so that an early warning is triggered within a short escape time window Wescape, thereby increasing sensitivity to severe fires.

[0034] Set dynamic threshold for escape window time Depending on the severity of the fire Varies with the changes of: ,in, It is the preset basic escape time threshold; is a function of the severity of the fire: ,in is the adjustment coefficient, when When increasing, will be reduced accordingly, allowing the system to speed up the warning as the severity of the fire increases.

[0035] The fire spread trend prediction module specifically includes: Data preprocessing unit: Receives data collected by the fire detection sensor network, including temperature, smoke concentration, humidity, oxygen content and fire source location, and also receives the distance between the fire source location and the personnel , the data is normalized to form a structured fire status data set; Temperature data :Depend on Temperature data collected by sensors; Smoke density data :Depend on Smoke concentration data collected by sensors; Humidity data :Depend on Humidity data collected by sensors; Oxygen content data :Depend on Oxygen content data collected by sensors; Fire source location data : The fire source location coordinates determined by the sensor network; Normalize each data item to obtain a standardized data set.

[0036] Fire spread path prediction unit: Based on the machine learning algorithm, the fire status data set output by the data preprocessing unit is trained and predicted. By analyzing the time series changes of fire characteristics, a prediction model of the fire spread path is generated. The machine learning algorithm includes a CNN-LSTM combination model. CNN extracts the spatial characteristics of the fire status data. The time dependency of the features is modeled by LSTM to predict the fire spread path. CNN processes the input multi-dimensional fire status data and extracts the spatial characteristics of the data. Each convolution layer performs convolution operations on the collected temperature, humidity, smoke concentration, and oxygen content to generate a feature map to capture the spatial distribution information of the fire status. The feature map extracted by CNN is input into LSTM. LSTM processes the time series information of the feature data, captures the dynamic changes in the fire spread path, models the changing trend of fire characteristics over time, and predicts the future path of the fire expansion. Combined with the output of the CNN-LSTM combination model, the prediction of the future location of the fire source is obtained. ,in is the number of time steps for prediction; Dynamic risk map generation unit: receives the diffusion area output by the fire spread path prediction unit, maps the fire spread path to the home floor plan, generates a dynamic risk map, and Differentiate and display high-risk areas and relatively safe areas.

[0037] The specific mapping is as follows: First, use the layout information of the home building (room size, door and window locations, corridors) to build a digital model of the home floor plan. This floor plan model should include the boundaries of each room, connection paths, and related safety exit locations.

[0038] For the convenience of positioning and visualization, define the two-dimensional coordinate system of the family floor plan and project the actual three-dimensional space (width, height and depth) of the family into the two-dimensional plane. For example, you can set the lower left corner of the family as the origin and set the x and y axes as the horizontal and vertical directions on the floor plan.

[0039] The fire source location coordinates predicted by the CNN-LSTM combined model are mapped to the two-dimensional coordinate system of the home floor plan, the height z in the three-dimensional coordinates is omitted, and only x and y are retained as the positions on the floor plan. The fire source location at each future time step is obtained based on the output of the CNN-LSTM combined model. These predicted locations are mapped to the floor plan one by one, and high-risk areas are distinguished by color or shading in the floor plan according to the diffusion range and spread speed of the fire source.

[0040] Using color gradients or risk markers, the fire spreading area is displayed as a dynamic risk area on the floor plan. At the same time, relatively safe areas are marked on the floor plan, and recommended escape routes are generated based on the current fire source location and spread trend. As the fire data is updated, the future location of the fire source is continuously predicted and refreshed on the home floor plan, allowing users to view the fire spread path and the best escape route in real time.

[0041] The machine learning algorithm trains and predicts the fire status data set output by the data preprocessing unit as follows: Based on the data in the preprocessed structured fire state dataset, a multidimensional feature matrix is ​​constructed. The feature matrix shape is The multidimensional feature matrix of is the number of time steps, is the number of features for each time step. Each row of the matrix contains the features of temperature, smoke concentration, humidity, oxygen content and fire source distance. The convolutional neural network (CNN) extracts spatial features from the feature matrix and models the spatial relationship of the fire data through the convolution layer. The output Z of the convolution layer is expressed as: ; in, The output feature map is at position The value at is the input feature matrix at position The value at is the convolution kernel at position The weight of is the bias term of the convolutional layer, is the activation function.

[0042] A long short-term memory network (LSTM) is used to model the time series to capture the temporal dependency of the fire spread. , the calculation of LSTM includes input gate, forget gate and output gate: ; ; ; ; ; in, They are input gate, forget gate and output gate respectively; is the input feature vector, the fire state input data at time step t, including temperature, smoke density, humidity, oxygen content, and fire source distance, is the weight matrix, is the bias term; is the hidden state, is the memory cell state, In hidden state.

[0043] : The weight matrix of the input gate, : The weight matrix of the forget gate, : The weight matrix of the output gate, : The weight matrix used to update the memory cell state; : Bias term of input gate, : Bias term of forget gate, : Bias term of output gate, : Bias term used to update the state of the memory unit; : Sigmoid activation function, used to calculate the input gate, forget gate and output gate; tanh: Hyperbolic tangent activation function, used to calculate candidate values ​​for memory cells.

[0044] Regional Risk Index The distance from each area to the fire source and the severity of the fire in each area Get, for map display: ,in, represents the regional risk index at the sth moment, is the predicted distance from the fire source to each area at the sth moment (and the distance from the personnel to the fire source The calculation is the same. The home floor plan can be divided into multiple node areas (including doorways, corridors, exits, etc.), and different detection points are set in each node area to represent the representative position of the area). is the minimum value that prevents the denominator from reaching zero.

[0045] The adaptive escape path optimization module receives the escape time window Wescape and the dynamic risk map, and generates multiple initial escape paths of different priorities based on the current fire source location and its spread path, combined with the safe exit location in the home floor plan. , Indicates The generated escape paths contain continuous routes connected by multiple nodes (doors, corridors, exits, etc.); path priority Time required by route , Wescape, and the regional risk index on the path, giving priority to low-risk paths that can be completed within the escape time window; Path time (i.e. through the path The time required) is calculated as: ,in, For path The number of nodes, including doorways, corridors, and exits, For path Midpoint To Node The distance Through the node The speed of travel at Path priority Calculated as: ,in, represents the path at time s The sum of the regional risk indexes of the nodes above (by node Distance from area to fire source and fire severity in node areas calculate).

[0046] Based on the output of path generation, the path with the highest priority and meeting the escape time window is recommended to the user. When the fire changes or the escape window is reduced, the recommended path is updated in real time to provide the most appropriate escape guidance.

[0047] 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.

[0048] 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. Intelligent household fire automatic alarm and escape guidance system, characterized by: include: Time-sensitive escape window assessment module: Based on the fire source temperature, smoke concentration, humidity and oxygen content collected by the fire detection sensor network, combined with the fire source location, it conducts real-time assessment of the severity of the current fire and its impact on the escape window, generates a dynamic change model of the escape time window size, and provides preliminary escape time warning; Fire spread trend prediction module: Based on the fire detection sensor network data and machine learning algorithms, it predicts the fire spread path and generates a dynamic risk map; Adaptive escape path optimization module: Based on the output results of the time-sensitive escape window assessment module and combined with the dynamic risk map, it generates multiple escape paths with different priorities and optimizes the paths in real time according to fire changes to ensure that the escape paths are always in low-risk areas.

2. The intelligent household fire automatic alarm and escape guidance system according to claim 1 is characterized in that: The time-sensitive escape window assessment module specifically includes: Data receiving unit: used to receive the fire source temperature collected by the fire detection sensor network , smoke concentration ,humidity and oxygen content ,The fire detection sensor network includes fire source temperature sensor, smoke sensor, and oxygen sensor; Fire trend analysis unit: based on the output results of the data receiving unit and combined with the location of the fire source , the speed of fire development, by analyzing the severity of the current fire, the spatial location of the fire source and the spread range, the severity of the current fire and fire development speed Conduct analysis; Dynamic escape window calculation unit: receives the output of the fire trend analysis unit, generates a dynamic escape time window size model based on the current fire situation, and calculates the escape time window and provide preliminary escape time warning .

3. The intelligent household fire automatic alarm and escape guidance system according to claim 2 is characterized in that: Described fire severity The calculation is expressed as: ； in, is the maximum threshold of fire source temperature; is the maximum threshold of smoke concentration; is the maximum threshold of humidity; is the maximum value of oxygen content, is the weight coefficient of different parameters to ensure that the severity of the fire can accurately reflect the impact of each parameter, and ; Indicates the location of the fire With personnel location spatial distance.

4. The intelligent household fire automatic alarm and escape guidance system according to claim 3 is characterized in that: The fire development speed The calculation is expressed as: ； in, is the weight coefficient of different parameters, satisfying ; is the temperature change rate, which indicates the rate of change of temperature over time; is the smoke concentration change rate, which indicates the rate of change of smoke concentration over time; is the humidity change rate, which indicates the rate of change of humidity over time; is the rate of change of oxygen content, which indicates the rate of change of oxygen content over time; It is the rate of change of fire source distance, which indicates the rate of change of the distance between the fire source and personnel over time.

5. The intelligent household fire automatic alarm and escape guidance system according to claim 4 is characterized in that: The escape time window It indicates the available time for personnel to escape from the fire source safely from the current time. It is calculated by combining the distance from the fire source to the escape area and the speed of fire development. The calculation is expressed as: ,in, is the initial safety distance, that is, the distance between the personnel position and the initial boundary of the fire source position, Indicates the current fire source distance, that is, the distance from the fire source to the personnel at the current time.

6. The intelligent household fire automatic alarm and escape guidance system according to claim 5 is characterized in that: The escape time warning :Determine whether to trigger the escape time warning based on the escape time window Wescape and the set minimum safety threshold Wthreshold.

7. The intelligent household fire automatic alarm and escape guidance system according to claim 6 is characterized in that: The minimum safety threshold Wthreshold is dynamically adjusted based on the fire severity R. When the fire severity R is too high, the minimum safety threshold Wthreshold is lowered so that an early warning is triggered within a short escape time window Wescape.

8. The intelligent household fire automatic alarm and escape guidance system according to claim 4 is characterized in that: The fire spread trend prediction module specifically includes: Data preprocessing unit: Receives data collected by the fire detection sensor network, including temperature, smoke concentration, humidity, oxygen content and fire source location, and also receives the distance between the fire source location and the personnel , the data is normalized to form a structured fire status data set; Fire spread path prediction unit: Based on the machine learning algorithm, the fire status data set output by the data preprocessing unit is trained and predicted. By analyzing the time series changes of fire characteristics, a prediction model of the fire spread path is generated. The machine learning algorithm includes a CNN-LSTM combination model. CNN extracts the spatial characteristics of the fire status data. The time dependency of the modeling features is used by LSTM to predict the fire spread path. CNN processes the input multi-dimensional fire status data and extracts the spatial characteristics of the data. Each convolution layer performs convolution operations on the collected temperature, humidity, smoke concentration, and oxygen content to generate a feature map to capture the spatial distribution information of the fire status. The feature map extracted by CNN is input into LSTM. LSTM processes the time series information of the feature data to capture the dynamic changes in the fire spread path. Combined with the output of the CNN-LSTM combination model, the prediction of the future location of the fire source is obtained. ,in is the number of time steps for prediction; Dynamic risk map generation unit: receives the diffusion area output by the fire spread path prediction unit, maps the fire spread path to the home floor plan, generates a dynamic risk map, and Differentiate and display high-risk areas and relatively safe areas.

9. The intelligent household fire automatic alarm and escape guidance system according to claim 8, characterized in that: The regional risk index The distance from each area to the fire source and the severity of the fire in each area Get, for map display: ,in, represents the regional risk index at the sth moment, is the distance from the fire source to each area at the sth moment, is the minimum value that prevents the denominator from being zero.

10. The intelligent household fire automatic alarm and escape guidance system according to claim 8, characterized in that: The adaptive escape path optimization module receives the escape time window Wescape and the dynamic risk map, and generates multiple initial escape paths of different priorities according to the current fire source location and its spread path, combined with the safe exit location in the home floor plan. , Indicates The generated escape paths contain multiple nodes connected into a continuous route; path priority Time required by route , Wescape, and the regional risk index on the path, giving priority to low-risk paths that can be completed within the escape time window; Path time Calculated as: ,in, For path The number of nodes, including doorways, corridors, and exits, For path Midpoint To Node The distance Through the node The speed of travel at Path priority Calculated as: ,in, represents the path at time s The sum of the regional risk indexes of all nodes above.

Images