Fire disaster real-time monitoring method and system based on intelligent sensing

Abstract

The application relates to the technical field of sensors and discloses a fire disaster real-time monitoring method and system based on intelligent sensing, which comprises the following steps: deploying a fire disaster monitoring device in a target area to be monitored, wherein the fire disaster monitoring device comprises a cavity optical force sensor, an acoustic surface wave resonator and a gas sensor; collecting sensor data in the target area through the fire disaster monitoring device, synchronizing the oscillation rhythm of the sensor data, obtaining an oscillation phase matrix; mapping the oscillation phase matrix into a three-dimensional ring surface flow form, topologically reconstructing the three-dimensional ring surface flow form, and obtaining a risk field flow form in the target area, wherein the risk field flow form comprises a curvature distribution, a homology group base vector and a risk intensity; and analyzing a fire disaster abnormal area, a fire disaster spreading path and a fire disaster spreading duration in the target area by using the risk field flow form. The application can synchronously generate a spatial path topology and a time evolution quantitative index of fire disaster spreading.

Description

Technical Field

[0001] This invention relates to a method and system for real-time monitoring of fire conditions based on intelligent sensing, belonging to the field of sensor technology. Background Technology

[0002] Real-time fire monitoring technology based on intelligent sensing refers to the deployment of a multi-type sensor network to collect physical parameters of fire-risk areas in real time. Combined with data fusion and intelligent analysis algorithms, it enables dynamic perception and early warning of the occurrence, spread, and risk level of fires. This technology aims to overcome the lag in traditional fire monitoring and provide high-precision, multi-dimensional decision support for fire prevention and control.

[0003] Although intelligent sensing technology has been applied in the field of fire protection, existing methods still have some core shortcomings. First, the deployment of existing sensors mostly relies on experience or the principle of uniform distribution, without taking into account the topological characteristics of the thermodynamic risk field for differentiated deployment. This leads to monitoring blind spots in key risk areas, directly causing incomplete spatial coverage of data sources. Second, in the data fusion stage, existing technologies usually use simple weighted averages or statistical models, failing to capture the dynamic oscillatory coupling relationship between temperature field, gas concentration field, and mechanical vibration field. As a result, the data fusion results cannot truly reflect the physical coordination mechanism of fire evolution. Third, existing models only focus on scalar parameter thresholds and do not map monitoring data to a high-dimensional manifold space to extract its inherent topological structure. This makes it impossible for the model to quantify the spatial connectivity and spread channel topology of the risk field.

[0004] Due to the aforementioned limitations, existing methods can only output local anomalies or risk levels, and cannot simultaneously generate spatial path topology and temporal evolution quantitative indicators of fire spread, which greatly weakens the pertinence of evacuation and intervention strategies. Summary of the Invention

[0005] This invention provides a method and system for real-time fire monitoring based on intelligent sensing, the main purpose of which is to simultaneously generate the spatial path topology and temporal evolution quantitative indicators of fire spread.

[0006] To achieve the above objectives, the present invention provides a real-time fire monitoring method based on intelligent sensing, comprising:

[0007] Fire monitoring devices are deployed in the target area where fire conditions are to be monitored. The fire monitoring devices include a cavity optical force sensor, a surface acoustic wave resonator, and a gas sensor.

[0008] The fire monitoring device collects sensor data in the target area, and the sensor data is synchronized with an oscillation rhythm to obtain an oscillation phase matrix.

[0009] The oscillation phase matrix is ​​mapped to a three-dimensional toroidal manifold, and the three-dimensional toroidal manifold is topologically reconstructed to obtain the risk field manifold in the target region, wherein the risk field manifold includes curvature distribution, homology group basis vectors, and risk intensity;

[0010] The fire anomaly area, fire spread path, and fire spread duration in the target area are analyzed using the risk field manifold.

[0011] The abnormal fire area, the fire spread path, and the degree of fire abnormality are used as the fire monitoring results in the target area.

[0012] Optionally, deploying the fire monitoring device in the target area where the fire situation needs to be monitored includes:

[0013] Identify the core thermal risk zone, structurally complex zone, and zone affected by fire sources within the target area;

[0014] Locate the walls and ceilings in the core heat risk area, structurally complex areas, and areas affected by fire sources;

[0015] Cavity optical force sensors are deployed on the walls and the ceiling;

[0016] Identify areas in the target area where smoke tends to accumulate and the return air vents of the ventilation system.

[0017] Locate suitable wall-mounted areas in the areas where smoke tends to accumulate and in the return air vents of the ventilation system.

[0018] A surface acoustic wave resonator is deployed in the wall-mountable area;

[0019] Identify areas prone to gas leaks and densely populated spaces within the target area;

[0020] Locate ventilation ducts and independent air chambers in the areas prone to gas leakage and in the densely populated spaces.

[0021] Gas sensors are deployed in the ventilation duct and the independent air chamber to complete the process of deploying a fire monitoring device in the target area.

[0022] Optionally, the step of synchronizing the oscillation rhythm of the sensor data to obtain the oscillation phase matrix includes:

[0023] The sensor data is subjected to Fourier transform to obtain frequency domain data;

[0024] Extract the significant fluctuation frequencies from the frequency domain data;

[0025] Set the initial phase angle of the sensor data;

[0026] Based on the significant frequency of the fluctuation, the initial phase angle is updated to obtain the updated phase angle;

[0027] Based on the significant frequency of the fluctuation, a cross-device synchronization model is generated between the update phase angles of various fire monitoring devices.

[0028] The oscillation phase matrix is ​​obtained by synchronizing the updated phase angle oscillation rhythm through the cross-device synchronization model.

[0029] Optionally, before mapping the oscillation phase matrix to a three-dimensional toroidal manifold, the method further includes:

[0030] Acquire the frequency domain data and significant fluctuation frequencies corresponding to the sensor data;

[0031] Extract the phase amplitude corresponding to the significant frequency of the fluctuation in the frequency domain data.

[0032] Optionally, mapping the oscillation phase matrix to a three-dimensional toroidal manifold includes:

[0033] Each row of data in the oscillation phase matrix is ​​used as a three-dimensional spatial oscillation source;

[0034] Obtain the phase amplitude;

[0035] Based on the three-dimensional spatial oscillation source and the phase amplitude, calculate the three-dimensional toroidal position of each different fire monitoring device;

[0036] Generate independent topological loops for each different fire monitoring device;

[0037] Obtain the coupling strength coefficient between different fire monitoring devices;

[0038] Based on the three-dimensional toroidal position and the coupling strength coefficient, the independent topological loops are connected into a three-dimensional toroidal manifold.

[0039] Optionally, the topological reconstruction of the three-dimensional toroidal manifold to obtain the risk field manifold in the target region includes:

[0040] The three-dimensional toroidal mesh is discretized into triangular patches;

[0041] Identify the connection relationships between adjacent faces in the triangular facet;

[0042] Based on the triangular facets and their connection relationships, generate the face-edge relationship matrix of the three-dimensional toroidal manifold;

[0043] Calculate the first-order homology group corresponding to the face-edge relation matrix;

[0044] Obtain the homology group basis vector of the first-order homology group;

[0045] The face-edge relation matrix is ​​decomposed into a diagonal matrix using Smith canonical form;

[0046] Extract the number of non-zero elements in the diagonal matrix;

[0047] The number of non-zero elements is taken as the number of holes in the target area;

[0048] Calculate the Ricci curvature value of each vertex in the triangular facet;

[0049] Calculate the curvature gradient corresponding to the Ricci curvature value;

[0050] The curvature distribution of the triangular facet is constructed using the curvature gradient;

[0051] The risk intensity in the target region is determined by the number of holes and the curvature gradient.

[0052] The curvature distribution, the homology group basis vector, and the risk intensity are used as the risk field manifold in the target region.

[0053] Optionally, calculating the Ricci curvature value of each vertex in the triangular patch includes:

[0054] Obtain the neighborhood triangles of the triangular facet;

[0055] Calculate the angular deficit between the vertices of the triangular facet and the neighboring triangles;

[0056] The angular deficit is used as the Ricci curvature value.

[0057] Optionally, calculating the curvature gradient corresponding to the Ricci curvature value includes:

[0058] Obtain the adjacent vertices of the vertex corresponding to the Ricci curvature value;

[0059] Calculate the curvature difference between each adjacent vertex;

[0060] Based on the curvature difference, calculate the curvature gradient corresponding to the Ricci curvature value.

[0061] Optionally, the step of using the risk field manifold analysis to identify abnormal fire areas, fire spread paths, and fire spread duration within the target area includes:

[0062] Obtain the curvature distribution, homology group basis vectors, and risk intensity in the risk field manifold;

[0063] By comparing the curvature distribution with a preset threshold, abnormal fire areas in the target area are identified.

[0064] Extract the vertices of the circular path from the homology group basis vectors;

[0065] Calculate the center point between the vertices of each circular path;

[0066] The sequence of hole locations formed by the center points is used as the fire spread path;

[0067] The fire spread rate in the target area is calculated based on the risk intensity.

[0068] The fire spread time in the target area is calculated using the fire spread rate and the safety boundary in the target area.

[0069] To address the aforementioned problems, the present invention also provides a real-time fire monitoring system based on intelligent sensing, the system comprising:

[0070] The device deployment module is used to deploy a fire monitoring device in a target area where a fire is to be monitored. The fire monitoring device includes a cavity optical force sensor, a surface acoustic wave resonator, and a gas sensor.

[0071] The rhythm synchronization module is used to collect sensor data in the target area through the fire monitoring device, synchronize the oscillation rhythm of the sensor data, and obtain the oscillation phase matrix.

[0072] The topology reconstruction module is used to map the oscillation phase matrix into a three-dimensional torus manifold, and to perform topology reconstruction on the three-dimensional torus manifold to obtain the risk field manifold in the target region, wherein the risk field manifold includes curvature distribution, homology group basis vectors and risk intensity;

[0073] The fire analysis module is used to analyze the fire anomaly areas, fire spread paths, and fire spread duration in the target area using the risk field manifold.

[0074] The result monitoring module is used to take the abnormal fire area, the fire spread path and the degree of fire abnormality as the fire monitoring results in the target area.

[0075] Compared to the problems described in the background technology, the embodiments of the present invention eliminate monitoring blind spots by deploying cavity optical sensors in the thermal core area and surface acoustic wave resonators in the ventilation openings through differentiated deployment driven by the risk field. The embodiments of the present invention synchronize the oscillation rhythms of various devices through a cross-device synchronization model, thereby utilizing the synchronization phenomenon between the oscillation rhythms of different physical fields to construct a phase matrix to restore and describe the dynamic coupling relationship between these physical fields. Furthermore, the embodiments of the present invention quantify the geometric structure of the risk field, its spatial connectivity, and the topology of the spread path through three-dimensional toroidal manifold mapping and topology reconstruction. The embodiments of the present invention analyze the fire anomaly areas, fire spread paths, and fire spread duration in the target area using the risk field manifold analysis to synchronously generate spatial path topology and temporal evolution quantitative indicators of fire spread. Therefore, the real-time fire monitoring method and system based on intelligent sensing provided by the embodiments of the present invention can synchronously generate spatial path topology and temporal evolution quantitative indicators of fire spread. Attached Figure Description

[0076] Figure 1 This is a flowchart illustrating a real-time fire monitoring method based on intelligent sensing, according to an embodiment of the present invention.

[0077] Figure 2 This is a schematic diagram of the oscillation phase matrix of a real-time fire monitoring method based on intelligent sensing provided in an embodiment of the present invention.

[0078] Figure 3 A schematic diagram of the face-edge relationship matrix of a real-time fire monitoring method based on intelligent sensing provided in an embodiment of the present invention;

[0079] Figure 4 This is a schematic diagram of a module for implementing the real-time fire monitoring system based on intelligent sensing, according to an embodiment of the present invention.

[0080] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0081] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0082] This application provides a real-time fire monitoring method based on intelligent sensing. The executing entity of the real-time fire monitoring method based on intelligent sensing includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the real-time fire monitoring method based on intelligent sensing can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0083] Example 1:

[0084] Reference Figure 1 The diagram shown is a flowchart illustrating a real-time fire monitoring method based on intelligent sensing according to an embodiment of the present invention. In this embodiment, the real-time fire monitoring method based on intelligent sensing includes:

[0085] S1. Deploy a fire monitoring device in the target area where the fire situation needs to be monitored, wherein the fire monitoring device includes a cavity optical force sensor, a surface acoustic wave resonator and a gas sensor.

[0086] In this embodiment of the invention, cavity optical sensors are deployed in the thermal core area and surface acoustic wave resonators are deployed in the ventilation openings to eliminate monitoring blind spots through differentiated deployment driven by risk fields.

[0087] In one embodiment of the present invention, the deployment of a fire monitoring device in a target area to be monitored includes: identifying the core thermal risk zone, structurally complex zone, and zone affected by fire sources within the target area; locating walls and ceilings within the core thermal risk zone, structurally complex zone, and zone affected by fire sources; deploying cavity optical sensors on the walls and ceilings; identifying areas prone to smoke accumulation and ventilation system return air vents within the target area; locating wall-mountable areas within the areas prone to smoke accumulation and ventilation system return air vents; deploying surface acoustic wave resonators in the wall-mountable areas; identifying areas prone to gas leakage and densely populated spaces within the target area; locating ventilation ducts and independent air chambers within the areas prone to gas leakage and densely populated spaces; and deploying gas sensors in the ventilation ducts and independent air chambers, thereby completing the process of deploying the fire monitoring device in the target area.

[0088] The thermal risk core area refers to the area within the target area with a high probability of fire occurrence and rapid heat accumulation. This area typically contains a large amount of flammable materials or high-temperature equipment, such as power distribution rooms, equipment rooms, and the top of kitchen stoves. The structurally complex area refers to areas with complex building structures, numerous obstacles, or narrow spaces, such as factory shelving areas or densely piped areas in industrial plants. The fire-affected area refers to areas easily affected directly or indirectly by fire sources, such as areas near the thermal risk core area or areas with flammable material transmission channels. The cavity photoelectric sensor refers to a microcavity photoelectric oscillation sensor. Utilizing the cavity enhancement effect and photoelectric interaction, this sensor has high sensitivity to minute physical quantities (such as thermal radiation and pressure changes) and can monitor photon resonant frequency drift caused by flame thermal disturbance. In fire monitoring, the microcavity photoelectric oscillation sensor can monitor changes in flame thermal radiation in real time and capture weak thermal radiation signals in the early stages of a fire. The smoke accumulation area refers to areas where smoke easily accumulates during a fire, such as corridor corners and the upper part of elevator shafts. The gaps between warehouse shelves; the return air inlet of the ventilation system refers to the inlet in which air is drawn in and returned to the ventilation equipment. Smoke is easily drawn into the return air inlet, making it an important location for monitoring smoke; the surface acoustic wave resonator refers to a sensor that works by utilizing the resonance characteristics of surface acoustic waves (SAW). The SAW resonator is highly sensitive to acoustic signals (such as the acoustic spectrum generated by the impact of smoke particles) and can be used to analyze the acoustic fingerprint of smoke particles; the gas leakage area refers to areas where gas leaks are likely to occur, such as gas pipeline interfaces, laboratory fume hoods, and battery storage compartments; the densely populated space refers to areas with frequent and high-density personnel activity, such as conference rooms and data center server rooms; the gas sensor refers to a MOF-based gas sensor that can monitor the concentration of gases such as CO and carbon dioxide in real time. MOF materials have high specific surface area, porosity, and adjustable pore size, and have excellent adsorption and separation performance for gas molecules. The MOF-based gas sensor uses the interaction between MOF materials and gas molecules to detect the presence and concentration of gases.

[0089] For example, the process of deploying the cavity optical force sensor on the wall and the ceiling is as follows: the cavity optical force sensor is embedded in the ceiling or side wall, with its axis aligned with the direction of the heat source, and the distance between it and the target area (heat risk core area, structurally complex area, area affected by fire source) is ≤8m (8m for high-risk areas) or ≤15m (15m for ordinary areas). Furthermore, the process of deploying the surface acoustic wave resonator in the wall-mountable area is as follows: the surface acoustic wave resonator is wall-mounted at a distance of 0.3m-0.5m from the ceiling, at a 30° angle to the airflow direction to enhance particle capture efficiency. Furthermore, the process of deploying the gas sensor in the ventilation duct and the independent air chamber is as follows: the gas sensor is integrated into the inner wall of the ventilation duct or the independent air chamber, and a dustproof filter membrane is installed at the air inlet of the gas sensor.

[0090] S2. Collect sensor data in the target area through the fire monitoring device, synchronize the oscillation rhythm of the sensor data, and obtain the oscillation phase matrix.

[0091] This invention embodiment synchronizes the oscillation rhythms between various devices through a cross-device synchronization model, thereby utilizing the synchronization phenomenon between the oscillation rhythms of different physical fields to construct a phase matrix to restore and describe the dynamic coupling relationship between these physical fields.

[0092] The sensor data includes infrared radiation variation data in a specific wavelength band (3.5–4.0 μm) captured by a microcavity photoelectric sensor, characteristic acoustic spectra generated by smoke particles impacting the resonator (typically in the 20–100 kHz frequency band) resolved by a surface acoustic wave resonator, and adsorption trajectory data of CO and carbon dioxide molecules within the pores recorded by a MOF gas sensor. It should be noted that the MOF gas sensor contains many tiny pores through which gas molecules (such as carbon monoxide and carbon dioxide) enter and are adsorbed. The sensor can record the movement paths of these molecules within the pores, forming a trajectory map. This trajectory map helps identify the type and concentration of the gas. For example, as the CO concentration increases, the adsorption path encoding value changes, such as from 0.15 to 0.38.

[0093] In one embodiment of the present invention, the step of synchronizing the oscillation rhythm of the sensor data to obtain an oscillation phase matrix includes: performing a Fourier transform on the sensor data to obtain frequency domain data; extracting significant fluctuation frequencies from the frequency domain data; setting an initial phase angle for the sensor data; updating the initial phase angle according to the significant fluctuation frequencies to obtain an updated phase angle; and generating a cross-device synchronization model between the updated phase angles of different fire monitoring devices using the following formula based on the significant fluctuation frequencies:

[0094]

[0095] in, This indicates a cross-device synchronization model. This represents the update phase angle of the a-th fire monitoring device at time t. This indicates that the b-th fire monitoring device is in The phase angle is updated at each moment. Indicates the pre-calibrated equipment delay duration. This represents the preset coupling strength coefficient. This represents the significant frequency of fluctuations corresponding to the a-th fire monitoring device;

[0096] The oscillation phase matrix is ​​obtained by synchronizing the updated phase angle oscillation rhythm through the cross-device synchronization model.

[0097] Wherein, the initial phase angle is 0, that is The formula for calculating the updated phase angle is: That is, the initial phase angle is increased over time, for example, Then after 2 seconds It should be noted that the coupling strength coefficient here can be calibrated manually based on experience, or it can be calibrated based on optimization algorithms such as the least squares method. Generally, the value is in the range of 0.1 to 1.0.

[0098] For example, the process of extracting significant frequencies of fluctuation in the frequency domain data is as follows: when the thermal radiation data fluctuates at 0.5 Hz, ω = 0.5, where ω represents the significant frequency of fluctuation. Furthermore, the process of synchronizing the updated phase angle oscillation rhythm through the cross-device synchronization model to obtain the oscillation phase matrix is ​​as follows: through iteration... Approaching 0, i.e., phase difference Approaching 0, ( (Faster) The formula result is (positive number), then The rate of change increases. Phase acceleration catch-up Otherwise it is Phase deceleration waiting ,For example, , , That is, the phase speed increase increased from 0.5 to Based on this phase increase rate, a new updated phase angle is calculated, thereby realizing the [function / response] of each sensor. Approaching 0.

[0099] See Figure 2The diagram shown is a schematic representation of the oscillation phase matrix of a real-time fire monitoring method based on intelligent sensing provided in an embodiment of the present invention. Figure 2 middle, , , This is the phase angle.

[0100] S3. Map the oscillation phase matrix to a three-dimensional torus manifold, and perform topological reconstruction on the three-dimensional torus manifold to obtain the risk field manifold in the target region, wherein the risk field manifold includes curvature distribution, homology group basis vectors and risk intensity.

[0101] In one embodiment of the present invention, before mapping the oscillation phase matrix to a three-dimensional toroidal manifold, the method further includes: acquiring frequency domain data and significant fluctuation frequencies corresponding to the sensor data; and extracting the phase amplitude corresponding to the significant fluctuation frequencies in the frequency domain data.

[0102] In one embodiment of the present invention, mapping the oscillation phase matrix to a three-dimensional toroidal manifold includes: using each row of data in the oscillation phase matrix as a three-dimensional spatial oscillation source; obtaining the phase amplitude; and calculating the three-dimensional toroidal position of each different fire monitoring device using the following formula based on the three-dimensional spatial oscillation source and the phase amplitude:

[0103]

[0104]

[0105]

[0106] in,( , , () indicates the position of the three-dimensional torus. This represents the radius of the torus, determined by the building's dimensions. This represents the updated phase angle in a three-dimensional spatial oscillation source. Indicates the azimuth angle of the fire monitoring device. Indicates phase amplitude;

[0107] Generate independent topological loops for each different fire monitoring device; obtain the coupling strength coefficient between each different fire monitoring device; and connect the independent topological loops into a three-dimensional toroidal manifold based on the three-dimensional toroidal position and the coupling strength coefficient.

[0108] The three-dimensional toroidal position refers to the coordinates of each different fire monitoring device within the larger toroidal surface.

[0109] Optionally, the process of using each row of data in the oscillation phase matrix as a three-dimensional spatial oscillation source means treating each sensor data point (time t, coordinates x, y, z, phase θ1 / θ2 / θ3) in the phase matrix as an oscillation source in three-dimensional space. For example, in a power distribution room (1.2, 3.4, 2.0), at t = 0.1s, the thermal radiation phase θ1 = 0.52. The power distribution room is then mapped as a point on the toroidal surface. Further, the process of generating independent topological rings for each different fire monitoring device refers to: the u-direction corresponding to the spatial distribution of sensors (e.g., floor location), and the v-direction corresponding to phase changes (the composite value of θ1 / θ2 / θ3). Each sensor generates an independent small ring (independent oscillating ring), which reflects the oscillation characteristics of the device on the three-dimensional toroidal surface. Further, connecting the independent topological rings into a three-dimensional toroidal manifold based on the three-dimensional toroidal surface position and the coupling strength coefficient means connecting adjacent small rings through coupling strength k, prioritizing the connection of the two closest small rings based on the minimum spanning tree principle to form a global large toroidal surface. That is, according to the three-dimensional toroidal surface position and the coupling strength coefficient, each independent oscillating ring (the position of each independent oscillating ring is the position of the three-dimensional toroidal surface) is connected to form a complete three-dimensional toroidal manifold. Regarding... The dimensions are predetermined by the building size; for example, for a chemical plant, R=20 meters. This means the larger the target area, the better. The larger the value, the more it needs to be determined manually based on the actual scenario.

[0110] Furthermore, embodiments of the present invention quantify the geometric structure of the risk field, the spatial connectivity of the risk field, and the topology of the propagation channels through three-dimensional toroidal manifold mapping and topological reconstruction.

[0111] In one embodiment of the present invention, the topological reconstruction of the three-dimensional toroidal manifold to obtain the risk field manifold in the target region includes: discretizing the three-dimensional toroidal manifold mesh into triangular patches; identifying the connectivity relationships between adjacent patches in the triangular patches; generating the face-edge relation matrix of the three-dimensional toroidal manifold based on the triangular patches and the connectivity relationships; and calculating the first-order homology group corresponding to the face-edge relation matrix using the following formula:

[0112]

[0113] in, Denotes a first-order homology group. Indicates satisfaction vector , Represents the face-edge relationship matrix. Indicates a closed region that can be filled;

[0114] Obtain the homology group basis vector of the first-order homology group; decompose the face-edge relation matrix into a diagonal matrix using Smith canonical form; extract the number of non-zero elements in the diagonal matrix; use the number of non-zero elements as the number of holes in the target region; calculate the Ricci curvature value of each vertex in the triangular facet; calculate the curvature gradient corresponding to the Ricci curvature value; construct the curvature distribution of the triangular facet using the curvature gradient; determine the risk intensity in the target region using the number of holes and the curvature gradient; use the curvature distribution, the homology group basis vector, and the risk intensity as the risk field manifold in the target region.

[0115] The first-order homology group is a concept in algebraic topology, used to describe the number and structure of one-dimensional "holes" in a topological space. In the risk field manifold, it is used to quantify possible fire spread channels (such as ventilation shafts, corridors, etc.) in the target area. The three-dimensional torus manifold refers to a three-dimensional topological structure similar to a donut. For example, the data from 32 sensors in a chemical plant are woven into a three-dimensional torus, with ventilation shaft positions corresponding to the holes in the torus. The Smith canonical form refers to the method of transforming a matrix into a diagonal matrix, which is often used to calculate homology groups. The risk intensity is the product of the number of holes and the curvature gradient.

[0116] Optionally, the process of discretizing the three-dimensional toroidal manifold mesh into triangular patches can be implemented using an algorithm based on Delaunay triangulation. The process of identifying the connection relationship between adjacent patches in the triangular patches is as follows: for each triangular patch, find the patches adjacent to it. Two patches are adjacent if they share an edge. Further, the process of generating the face-edge relationship matrix of the three-dimensional toroidal manifold based on the triangular patches and the connection relationship is as follows: create a matrix where rows represent edges, columns represent patches, and matrix elements represent the relationship between edges and patches. If an edge belongs to a patch, fill in 1 or -1 in the corresponding position (use -1 if the direction is opposite), otherwise fill in 0.

[0117] See Figure 3 The diagram shown is a schematic representation of the surface-edge relationship matrix of a real-time fire monitoring method based on intelligent sensing provided in an embodiment of the present invention. Figure 3 In the middle, the edge of patch 1 is , The direction is , The direction is , The direction is The edge of patch 2 is , The direction is (Same as patch 1), that is, the connection relationship between adjacent patches. The direction is , The direction is .

[0118] In another embodiment of the present invention, calculating the Ricci curvature value of each vertex in the triangular facet includes: obtaining the neighboring triangles of the triangular facet; and calculating the angular deficit between the vertices of the triangular facet and the neighboring triangles using the following formula:

[0119]

[0120] in, Indicates an angular deficit. Represents the vertices of a triangular face. This indicates that the neighborhood triangle is at the vertex The interior angle at that point, Indicates all neighborhood triangles at the vertex The sum of the interior angles at point , This represents the position index of a vertex in a triangular facet;

[0121] The angular deficit is used as the Ricci curvature value.

[0122] It should be noted that 'i' represents the vertical index of a grid vertex (e.g., the floor plan of a building), corresponding to the 'u' direction of the parameterized torus manifold, and 'j' represents the horizontal index of a grid vertex (e.g., the planar coordinates within a floor), corresponding to the 'v' direction of the parameterized torus manifold. Each (i,j) corresponds to the coordinates (x,y,z) of a vertex in the parameterized grid. The original spatial coordinates (x,y,z) need to be converted into (u,v) parametric coordinates using the torus formula, and then quantized into integers (i,j), i.e., u = azimuth angle. →Mapped to index j, v = phase angle θ →Mapped to index i.

[0123] In another embodiment of the present invention, calculating the curvature gradient corresponding to the Ricci curvature value includes: obtaining the neighboring vertices of the vertex corresponding to the Ricci curvature value; and calculating the curvature difference between each neighboring vertex using the following formula:

[0124]

[0125]

[0126] in, Represents the curvature difference on the X-axis. Represents the curvature difference on the Y-axis. Indicates the position index is curvature distribution, Indicates the position index is curvature distribution, Indicates the position index is curvature distribution, Indicates the position index is curvature distribution, Indicates the x-axis and Spatial step size between Indicates the Y-axis and Spatial step size between;

[0127] Based on the curvature difference, the curvature gradient corresponding to the Ricci curvature value is calculated using the following formula:

[0128]

[0129] in, Represents the curvature gradient. Represents the curvature difference on the X-axis. This represents the curvature difference on the Y-axis.

[0130] in, , , , for Each of the adjacent vertices.

[0131] S4. Utilize the risk field manifold analysis to identify abnormal fire areas, fire spread paths, and fire spread duration within the target area.

[0132] This invention utilizes the risk field manifold analysis to identify fire anomaly areas, fire spread paths, and fire spread duration within the target area, thereby simultaneously generating spatial path topology and temporal evolution quantitative indicators for fire spread.

[0133] In one embodiment of the present invention, the step of analyzing the fire anomaly region, fire spread path, and fire spread duration in the target region using the risk field manifold includes: obtaining the curvature distribution, homology group basis vector, and risk intensity in the risk field manifold; identifying the fire anomaly region in the target region by comparing the curvature distribution with a preset threshold; extracting the vertices of the circular path from the homology group basis vector; calculating the center point between each circular path vertex; using the sequence of hole positions formed by the center points as the fire spread path; calculating the fire spread velocity in the target region using the risk intensity; and calculating the fire spread duration in the target region using the fire spread velocity and the safety boundary in the target region.

[0134] For example, the process of identifying abnormal fire areas in the target area by comparing the curvature distribution with a preset threshold is as follows: when the curvature distribution is greater than 0.1, it indicates the presence of an abnormal fire. Further, the process of extracting the vertices of the circular path from the homology group basis vectors is as follows: basis vector 1 is [vertices A, B, C], representing the circular boundary path of the ventilation shaft; basis vector 2 is [vertices D, E, F], representing the circular path of the cable channel. The basis vectors are the vertex sequence of the loop, i.e., the edge lines of the holes. Further, the process of calculating the center point between each circular path vertex is as follows: for example, if the basis vector vertices are (2.0, 3.0, 1.0), (4.0, 3.0, 1.0), (3.0, 7.0, 1.0), then the mean of (2.0, 3.0, 1.0), (4.0, 3.0, 1.0), and (3.0, 7.0, 1.0) is calculated to obtain the center point (3.0, 4.3, ...). 1.0) refers to the location of the ventilation shaft (i.e., the opening, the location where anomalies occur). The channels formed by the locations where anomalies occur are considered as fire spread paths. For example, ventilation shafts A and B are both locations where anomalies occur, and these two constitute a fire spread path. Furthermore, the fire spread rate in the target area is calculated based on the risk intensity, for example: v = k' Ψ and v represent the fire spread rate, and k' is a correction factor (determined by the type of combustible material, such as 1.2 for chemical warehouses and 0.8 for ordinary residential buildings). Furthermore, the fire spread time in the target area is calculated using the fire spread rate and the safety boundary in the target area. For example, if the distance from the fire source to the safety boundary is d (unit: meters), then T = d / v, where T is the fire spread time.

[0135] S5. The abnormal fire area, the fire spread path, and the degree of fire abnormality are taken as the fire monitoring results in the target area.

[0136] Compared to the problems described in the background technology, the embodiments of the present invention eliminate monitoring blind spots by deploying cavity optical sensors in the thermal core area and surface acoustic wave resonators in the ventilation openings through differentiated deployment driven by the risk field. The embodiments of the present invention synchronize the oscillation rhythms of various devices through a cross-device synchronization model, thereby utilizing the synchronization phenomenon between the oscillation rhythms of different physical fields to construct a phase matrix to restore and describe the dynamic coupling relationship between these physical fields. Furthermore, the embodiments of the present invention quantify the geometric structure of the risk field, its spatial connectivity, and the topology of the spread path through three-dimensional toroidal manifold mapping and topology reconstruction. The embodiments of the present invention analyze the fire anomaly areas, fire spread paths, and fire spread duration in the target area using the risk field manifold analysis to synchronously generate spatial path topology and temporal evolution quantitative indicators of fire spread. Therefore, the real-time fire monitoring method and system based on intelligent sensing provided by the embodiments of the present invention can synchronously generate spatial path topology and temporal evolution quantitative indicators of fire spread.

[0137] Example 2:

[0138] like Figure 4 The diagram shown is a functional block diagram of a real-time fire monitoring system based on intelligent sensing according to the present invention.

[0139] The intelligent sensing-based real-time fire monitoring system 400 described in this invention can be installed in an electronic device. Depending on the functions implemented, the intelligent sensing-based real-time fire monitoring system may include a device deployment module 401, a rhythm synchronization module 402, a topology reconstruction module 403, a fire analysis module 404, and a result monitoring module 405. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0140] In this embodiment of the invention, the functions of each module / unit are as follows:

[0141] The device deployment module 401 is used to deploy a fire monitoring device in the target area where the fire situation is to be monitored. The fire monitoring device includes a cavity optical force sensor, a surface acoustic wave resonator, and a gas sensor.

[0142] The rhythm synchronization module 402 is used to collect sensor data in the target area through the fire monitoring device, synchronize the oscillation rhythm of the sensor data, and obtain an oscillation phase matrix.

[0143] The topology reconstruction module 403 is used to map the oscillation phase matrix into a three-dimensional toroidal manifold, and to perform topology reconstruction on the three-dimensional toroidal manifold to obtain the risk field manifold in the target region, wherein the risk field manifold includes curvature distribution, homology group basis vectors and risk intensity;

[0144] The fire analysis module 404 is used to analyze the fire anomaly area, fire spread path and fire spread duration in the target area using the risk field manifold analysis.

[0145] The result monitoring module 405 is used to take the abnormal fire area, the fire spread path and the degree of fire abnormality as the fire monitoring results in the target area.

[0146] In detail, the modules in the intelligent sensing-based real-time fire monitoring system 400 described in this embodiment of the invention employ the same methods as described above. Figure 1 The method uses the same technical means as the real-time fire monitoring method based on intelligent sensing described in the article, and can produce the same technical effect, so it will not be repeated here.

[0147] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A fire disaster real-time monitoring method based on intelligent sensing, characterized in that, The method includes: Fire monitoring devices are deployed in the target area where fire conditions are to be monitored. The fire monitoring devices include a cavity optical force sensor, a surface acoustic wave resonator, and a gas sensor. The fire monitoring device collects sensor data from the target area, and synchronizes the sensor data with an oscillation rhythm to obtain an oscillation phase matrix, specifically including: The sensor data is subjected to Fourier transform to obtain frequency domain data; Extract the significant fluctuation frequencies from the frequency domain data; Set the initial phase angle of the sensor data; Based on the significant frequency of the fluctuation, the initial phase angle is updated to obtain the updated phase angle; Based on the significant frequency of the fluctuation, a cross-device synchronization model is generated between the update phase angles of various fire monitoring devices. The oscillation phase matrix is ​​obtained by synchronizing the updated phase angle oscillation rhythm using the cross-device synchronization model. Mapping the oscillation phase matrix to a three-dimensional toroidal manifold specifically includes: using each row of data in the oscillation phase matrix as a three-dimensional spatial oscillation source; Obtain the phase amplitude; Based on the three-dimensional spatial oscillation source and the phase amplitude, calculate the three-dimensional toroidal position of each different fire monitoring device; Generate independent topological loops for each different fire monitoring device; Obtain the coupling strength coefficient between different fire monitoring devices; Based on the three-dimensional toroidal position and the coupling strength coefficient, the independent topological loops are connected into a three-dimensional toroidal manifold; Topological reconstruction of the three-dimensional toroidal manifold to obtain the risk field manifold in the target region specifically includes: discretizing the three-dimensional toroidal manifold mesh into triangular patches; Identify the connection relationships between adjacent faces in the triangular facet; Based on the triangular facets and their connection relationships, generate the face-edge relationship matrix of the three-dimensional toroidal manifold; Calculate the first-order homology group corresponding to the face-edge relation matrix; Obtain the homology group basis vector of the first-order homology group; The face-edge relation matrix is ​​decomposed into a diagonal matrix using Smith canonical form; Extract the number of non-zero elements in the diagonal matrix; The number of non-zero elements is taken as the number of holes in the target area; Calculate the Ricci curvature value of each vertex in the triangular facet; Calculate the curvature gradient corresponding to the Ricci curvature value; The curvature distribution of the triangular facet is constructed using the curvature gradient; The risk intensity in the target region is determined by the number of holes and the curvature gradient. The curvature distribution, the homology group basis vector, and the risk intensity are used as the risk field manifold in the target region; The risk field manifold includes curvature distribution, homology group basis vectors, and risk intensity. The risk field manifold analysis is used to identify abnormal fire areas, fire spread paths, and fire spread duration within the target area, specifically including: Obtain the curvature distribution, homology group basis vectors, and risk intensity in the risk field manifold; By comparing the curvature distribution with a preset threshold, abnormal fire areas in the target area are identified. Extract the vertices of the circular path from the homology group basis vectors; Calculate the center point between the vertices of each circular path; The sequence of hole locations formed by the center points is used as the fire spread path; The fire spread rate in the target area is calculated based on the risk intensity. The fire spread time in the target area is calculated using the fire spread rate and the safety boundary in the target area; The abnormal fire area, the fire spread path, and the degree of fire abnormality are used as the fire monitoring results in the target area.

2. The intelligent sensor based fire emergency real time monitoring method as claimed in claim 1 wherein, The deployment of fire monitoring devices in the target area where fire conditions are to be monitored includes: Identify the core thermal risk zone, structurally complex zone, and zone affected by fire sources within the target area; Locate the walls and ceilings in the core heat risk area, structurally complex areas, and areas affected by fire sources; Cavity optical force sensors are deployed on the walls and the ceiling; Identify areas in the target area where smoke tends to accumulate and the return air vents of the ventilation system. Locate suitable wall-mounted areas in the areas where smoke tends to accumulate and in the return air vents of the ventilation system. A surface acoustic wave resonator is deployed in the wall-mountable area; Identify areas prone to gas leaks and densely populated spaces within the target area; Locate ventilation ducts and independent air chambers in the areas prone to gas leakage and in the densely populated spaces. Gas sensors are deployed in the ventilation duct and the independent air chamber to complete the process of deploying a fire monitoring device in the target area.

3. The intelligent sensor based fire incident real time monitoring method as claimed in claim 1 wherein, Before mapping the oscillation phase matrix to a three-dimensional toroidal manifold, the method further includes: Acquire the frequency domain data and significant fluctuation frequencies corresponding to the sensor data; Extract the phase amplitude corresponding to the significant frequency of the fluctuation in the frequency domain data.

4. The intelligent sensor based fire incident real time monitoring method as claimed in claim 1 wherein, The calculation of the Ricci curvature value of each vertex in the triangular facet includes: Obtain the neighborhood triangles of the triangular facet; Calculate the angular deficit between the vertices of the triangular facet and the neighboring triangles; The angular deficit is used as the Ricci curvature value.

5. The real-time fire monitoring method based on intelligent sensing as described in claim 1, characterized in that, The calculation of the curvature gradient corresponding to the Ricci curvature value includes: Obtain the adjacent vertices of the vertex corresponding to the Ricci curvature value; Calculate the curvature difference between each adjacent vertex; Based on the curvature difference, calculate the curvature gradient corresponding to the Ricci curvature value.

6. A fire disaster real-time monitoring system based on intelligent sensing, characterized in that, The system includes: The device deployment module is used to deploy a fire monitoring device in a target area where a fire is to be monitored. The fire monitoring device includes a cavity optical force sensor, a surface acoustic wave resonator, and a gas sensor. The rhythm synchronization module is used to collect sensor data in the target area through the fire monitoring device, synchronize the oscillation rhythm of the sensor data, and obtain an oscillation phase matrix. Specifically, it includes: The sensor data is subjected to Fourier transform to obtain frequency domain data; Extract the significant fluctuation frequencies from the frequency domain data; Set the initial phase angle of the sensor data; Based on the significant frequency of the fluctuation, the initial phase angle is updated to obtain the updated phase angle; Based on the significant frequency of the fluctuation, a cross-device synchronization model is generated between the update phase angles of various fire monitoring devices. The oscillation phase matrix is ​​obtained by synchronizing the updated phase angle oscillation rhythm using the cross-device synchronization model. The topology reconstruction module is used to map the oscillation phase matrix into a three-dimensional toroidal manifold, specifically including: using each row of data in the oscillation phase matrix as a three-dimensional spatial oscillation source; Obtain the phase amplitude; Based on the three-dimensional spatial oscillation source and the phase amplitude, calculate the three-dimensional toroidal position of each different fire monitoring device; Generate independent topological loops for each different fire monitoring device; Obtain the coupling strength coefficient between different fire monitoring devices; Based on the three-dimensional toroidal position and the coupling strength coefficient, the independent topological loops are connected into a three-dimensional toroidal manifold; Topological reconstruction of the three-dimensional toroidal manifold to obtain the risk field manifold in the target region specifically includes: discretizing the three-dimensional toroidal manifold mesh into triangular patches; Identify the connection relationships between adjacent faces in the triangular facet; Based on the triangular facets and their connection relationships, generate the face-edge relationship matrix of the three-dimensional toroidal manifold; Calculate the first-order homology group corresponding to the face-edge relation matrix; Obtain the homology group basis vector of the first-order homology group; The face-edge relation matrix is ​​decomposed into a diagonal matrix using Smith canonical form; Extract the number of non-zero elements in the diagonal matrix; The number of non-zero elements is taken as the number of holes in the target area; Calculate the Ricci curvature value of each vertex in the triangular facet; Calculate the curvature gradient corresponding to the Ricci curvature value; The curvature distribution of the triangular facet is constructed using the curvature gradient; The risk intensity in the target region is determined by the number of holes and the curvature gradient. The curvature distribution, the homology group basis vector, and the risk intensity are used as the risk field manifold in the target region; The risk field manifold includes curvature distribution, homology group basis vectors, and risk intensity. The fire analysis module is used to analyze the fire anomaly areas, fire spread paths, and fire spread duration in the target area using the risk field manifold analysis, specifically including: Obtain the curvature distribution, homology group basis vectors, and risk intensity in the risk field manifold; By comparing the curvature distribution with a preset threshold, abnormal fire areas in the target area are identified. Extract the vertices of the circular path from the homology group basis vectors; Calculate the center point between the vertices of each circular path; The sequence of hole locations formed by the center points is used as the fire spread path; The fire spread rate in the target area is calculated based on the risk intensity. The fire spread time in the target area is calculated using the fire spread rate and the safety boundary in the target area; The result monitoring module is used to take the abnormal fire area, the fire spread path and the degree of fire abnormality as the fire monitoring results in the target area.

Images