Fire-fighting supervision alarm platform and method under intelligent fire-fighting ecology

Through the coordinated work of multi-sensor network layer structure and multi-level sensors, the problems of insufficient sensitivity and high false alarm rate of traditional fire monitoring systems are solved, and high-accurate fire monitoring and early warning are achieved.

CN119028110AInactive Publication Date: 2024-11-26JIANGSU NITE SMART SECURITY TECHNOLOGY CO LTD

Patent Information

Application Number
CN202411468038.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-21
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional fire monitoring systems rely on a single sensor, have limited sensitivity and are easily disturbed by external environmental factors, resulting in frequent false alarms or missed fires, affecting the reliability and accuracy of the system.

Method used

Adopt a multi-sensor network layer structure, by analyzing the parameters of the sensing equipment, setting up a multi-level sensing network, activation conditions and early warning conditions, combining first-level and second-level sensor data for fire identification, and awakening the second when the early warning conditions are not met. Level sensing network layer, conducts multi-level and multi-dimensional fire monitoring and early warning.

Benefits of technology

It improves the accuracy and reliability of fire warnings, reduces the false alarm rate, and ensures multi-level and accurate fire monitoring and early warning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119028110A_ABST
    Figure CN119028110A_ABST
Patent Text Reader

Abstract

The invention discloses a fire-fighting supervision alarm platform and method under intelligent fire-fighting ecology, and relates to the technical field of fire-fighting supervision alarm, and the method comprises a multi-sensor network layer setting module which is used for analyzing sensing equipment parameters and setting a network layer according to sensitivity; the primary sensing monitoring data acquisition module is responsible for initializing a primary sensing network and acquiring monitoring data; the primary fire identification module is used for carrying out fire identification according to the monitoring data and sending early warning information; and the second-level fire identification module is used for activating a second-level sensing network when the first-level fire identification module does not meet the early warning condition, and sending early warning information through secondary identification. The technical problems that in the prior art, traditional fire-fighting monitoring usually depends on a single sensor, sensitivity is limited, interference of external environment factors is likely to happen, and reliability and accuracy are poor are solved, and the purposes that multi-level early warning is carried out by integrating multiple sensing devices, and the monitoring accuracy is improved are achieved. And the accuracy and reliability of fire early warning are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of fire monitoring and alarm technology, and in particular to a fire monitoring and alarm platform and method under a smart fire ecology. Background Art

[0002] Under the general trend of modern urban construction and intelligent development, fire safety, as an important part of urban public safety, is gradually developing in the direction of intelligence and automation. Traditional fire monitoring usually relies on a single sensor, such as a smoke sensor or a temperature sensor, to monitor signs of fire in the environment. However, the sensitivity of a single sensor is limited and is easily interfered by external environmental factors, resulting in frequent false alarms or missed fire reports, affecting the reliability and accuracy of the system. Therefore, how to improve the responsiveness of the fire monitoring system, reduce the false alarm rate, and improve the accuracy of fire identification has become an urgent problem to be solved in the field of fire monitoring. Summary of the invention

[0003] This application provides a fire supervision and alarm platform and method under the smart fire ecology, which is used to solve the technical problems in the prior art that traditional fire monitoring usually relies on a single sensor, has limited sensitivity, is easily interfered by external environmental factors, and has poor reliability and accuracy.

[0004] The first aspect of the present application provides a fire supervision and alarm method under a smart fire ecology, the method comprising: performing parameter analysis on the sensor equipment under the smart fire ecology, setting a multi-sensor network layer according to the parameter sensitivity, wherein each network layer has an activation condition and a warning condition; initializing a first-level sensor network layer, obtaining a first-level sensor monitoring data, the first-level sensor monitoring data comprising fire parameter monitoring data and joint parameter monitoring data; using the fire parameter monitoring data and the joint parameter monitoring data as input variables, performing fire identification on the monitoring data through a warning identification module, and sending a fire warning message when the warning condition is met; when the warning condition is not met, waking up the second-level sensor network layer according to the activation condition based on the multi-sensor network layer, extracting the second-level sensor monitoring data, identifying the second-level sensor monitoring data through a warning identification module, and sending the fire warning message when the second-level warning condition is met.

[0005] The second aspect of the present application provides a fire supervision and alarm platform under a smart fire protection ecology, and the platform includes: a multi-sensor network layer setting module, which is used to perform parameter analysis on the sensor equipment under the smart fire protection ecology, and set the multi-sensor network layer according to the parameter sensitivity, wherein each network layer has an activation condition and an early warning condition; a first-level sensor monitoring data acquisition module, which is used to initialize the first-level sensor network layer and obtain the first-level sensor monitoring data, wherein the first-level sensor monitoring data includes fire parameter monitoring data and joint parameter monitoring data; a first-level fire identification module, which is used to use the fire parameter monitoring data and the joint parameter monitoring data as input variables, and perform fire identification on the monitoring data through an early warning identification module, and send fire warning information when the early warning condition is met; a second-level fire identification module, which is used to wake up the second-level sensor network layer according to the activation condition based on the multi-sensor network layer when the early warning condition is not met, extract the second-level sensor monitoring data, identify the second-level sensor monitoring data through the early warning identification module, and send the fire warning information when the second-level early warning condition is met.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: The fire supervision and alarm platform and method under the smart fire protection ecology provided by the present application relate to the field of fire supervision and alarm technology. By performing parameter analysis on the sensor equipment, a multi-sensor network layer is set according to the sensitivity, fire parameters and joint parameter monitoring data are obtained, and fire identification is performed through the early warning identification module. When the early warning conditions are met, an early warning message is sent. If the early warning conditions are not met, the secondary sensor network layer is awakened, and the secondary sensor data is extracted to perform fire identification again, thereby ensuring multi-level and accurate fire monitoring and early warning. The technical problems in the prior art that traditional fire monitoring usually relies on a single sensor, has limited sensitivity, is easily interfered by external environmental factors, and has poor reliability and accuracy are solved, and the technical effect of improving the accuracy and reliability of fire early warnings by integrating multiple sensor devices for multi-level early warnings is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1 A flowchart of a fire monitoring and alarm method under a smart fire ecology provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of a fire monitoring and alarm platform under a smart fire protection ecology provided in an embodiment of the present application.

[0009] Explanation of the reference numerals: multi-sensor network layer setting module 11, primary sensor monitoring data acquisition module 12, primary fire identification module 13, secondary fire identification module 14. DETAILED DESCRIPTION

[0010] This application provides a fire supervision and alarm platform and method under the smart fire ecology, which is used to solve the technical problems in the prior art that traditional fire monitoring usually relies on a single sensor, has limited sensitivity, is easily interfered by external environmental factors, and has poor reliability and accuracy.

[0011] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0012] It should be noted that the terms "first", "second", etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices.

[0013] Embodiment 1, as Figure 1 As shown, the present application provides a fire monitoring alarm method under a smart fire ecology, the method comprising: P10: Analyze the parameters of the sensor equipment in the smart fire protection ecosystem, and set up multiple sensor network layers according to the parameter sensitivity, where each network layer has activation conditions and warning conditions.

[0014] Furthermore, step P10 of the embodiment of the present application further includes: P11: Obtain the sensor type, sensor distribution, and sensor model parameters in the smart fire protection ecology, and build a sensor deployment network; P12: Use the preset multi-level network layer target to traverse and match the sensors in the sensor deployment network, and establish a network hierarchical relationship, wherein the first level in the preset multi-level network layer is a smoke sensor, the second level is a flame sensor, and the third level is an infrared sensor, and the unmatched sensor is set as a joint sensor; P13: Using the fire warning fault case set, with the first-level smoke sensor, the second-level flame sensor, and the third-level infrared sensor as the target, fit the activation conditions and warning conditions, obtain the activation conditions and warning conditions of each level and add them to the network hierarchical relationship; P14: Through the fire warning fault case set, establish a collaborative warning relationship between the joint sensor and sensors at each level, and fit it to the corresponding network hierarchical relationship according to the collaborative warning relationship, wherein the first-level sensor network layer contains at least one collaborative warning relationship of the joint sensor.

[0015] It should be understood that the parameters of the sensor equipment in the smart fire protection ecology are analyzed, and a multi-level sensor network structure is established according to the sensitivity of the sensor. Specifically, the sensors are divided into different network levels according to their sensitivity, and corresponding activation conditions and warning conditions are set for each level to ensure the accuracy and comprehensiveness of fire monitoring. This process can significantly improve the system's ability to respond to fires while reducing the false alarm rate.

[0016] Specifically, in the smart fire protection ecosystem, sensor types may include smoke detectors, gas detection sensors (such as CO, CH4 detectors), temperature sensors, etc. By collecting the distribution and model parameters of these sensors, an accurate sensor deployment network can be constructed. This network can effectively cover the areas that need to be monitored, ensure that different types of sensors can work together, and improve the overall monitoring capabilities.

[0017] Using the preset multi-level network layer targets, each sensor in the sensor deployment network is traversed and matched to form a clear network hierarchy. Specifically, the first-level sensor (such as smoke detectors) is responsible for early fire signal detection, the second-level sensor (such as flame detectors) is used to verify the existence of the fire, and the third-level sensor (such as infrared thermal imaging equipment) is used as the final confirmation layer to help locate the fire source in complex situations. Unmatched sensors will be set as joint sensors, and their role is to cooperate with basic sensors to conduct multi-dimensional early warning analysis to improve the reliability of the system.

[0018] Using the fire warning failure case set, we fit the activation conditions and warning conditions for different sensor levels. The case set records the fire monitoring failure and false alarm data in different scenarios. Based on these data, we fit the activation conditions for each level. For example, the activation condition of a primary smoke sensor may be triggered when the smoke concentration reaches a certain critical value; the warning condition of a secondary flame detector may be the detection of a continuous flame signal. These conditions will be added to the network hierarchical relationship to ensure that each sensor can be activated and participate in monitoring at the right time.

[0019] Through the fire warning fault case collection, the collaborative warning relationship between joint sensors and sensors at all levels is further established. This means that joint sensors can assist the work of primary, secondary, and tertiary sensors. Especially in complex scenarios, these sensors can provide additional detection dimensions to ensure that signs of fire are not missed. For example, when the primary sensor detects smoke but cannot determine the source of the fire, joint sensors (such as gas sensors) can help further confirm whether there are fire factors.

[0020] Ultimately, these collaborative relationships will be fitted into the network hierarchy, where the first-level sensor network layer contains at least one collaborative early warning relationship of joint sensors, thereby improving the collaborative working ability and response speed of the entire system.

[0021] P20: Initialize the first-level sensor network layer and obtain the first-level sensor monitoring data, which includes fire parameter monitoring data and joint parameter monitoring data.

[0022] Optionally, the first-level sensor network layer is initialized and first-level sensor monitoring data from the first-level sensor is acquired in real time. The monitoring data includes two main types: fire parameter monitoring data and joint parameter monitoring data.

[0023] Among them, fire parameter monitoring data refers to sensor data directly related to fire, such as smoke concentration, temperature changes, flame appearance and other information. In the first-level sensor network layer, highly sensitive sensors such as smoke detectors and temperature sensors are usually equipped. These sensors are designed to respond quickly to early signs of fire, especially subtle environmental changes. The first-level sensors are highly sensitive and can detect potential signs of fire in a timely manner, but they may also cause false alarms due to environmental factors (such as kitchen smoke or steam). Therefore, although the data of the first-level sensors is important, it usually does not directly trigger an alarm, but is used to trigger an early warning.

[0024] Meanwhile, joint parameter monitoring data is data from joint sensors. These joint sensors are devices used to assist primary sensors in analysis and may include gas detectors (such as CO or CH4 sensors) or humidity sensors. Their role is to enhance the accuracy of fire identification through multi-dimensional data. For example, when a smoke sensor detects abnormal smoke, a joint sensor will detect gas or other environmental parameters to further confirm the possibility of a fire.

[0025] In this process, the deployment of joint sensors can be adjusted according to the specific situation of the smart fire ecosystem. If the ecosystem is more complex, more types of joint sensors may be added to deal with different fire risks. Relatively speaking, in a simple ecosystem, only a few joint sensors may be used as auxiliary equipment.

[0026] P30: The fire parameter monitoring data and the joint parameter monitoring data are used as input variables, and the monitoring data are used for fire identification through the early warning identification module. When the early warning conditions are met, fire warning information is sent.

[0027] Specifically, the fire parameter monitoring data and the joint parameter monitoring data are used as input variables and input into the early warning identification module for fire identification. At this time, the highly sensitive data of the primary sensor (such as smoke and temperature sensors) and the auxiliary data of the joint sensor (such as gas sensor) enter the early warning identification system together. The system determines whether the early warning conditions are met through a comprehensive analysis of these multi-dimensional data. If the early warning conditions are met, the system will immediately send a fire warning message.

[0028] Fire parameter monitoring data usually includes key data such as smoke concentration and temperature changes that can directly indicate the occurrence of fire. The joint parameter monitoring data is auxiliary information obtained from other sensors (such as gas, humidity or current monitoring equipment). These data are fused and processed in the early warning recognition module. The early warning recognition module is the core part of the system and is responsible for making real-time judgments on fires based on the data fed back by the sensors.

[0029] Exemplarily, the early warning identification module adopts multi-source data fusion technology, that is, a more accurate judgment is made by comprehensively analyzing the data of multiple sensors. For example, in some cases, feedback from collaborative devices (such as current monitors) can further enhance the judgment ability of the system. If the current monitor detects that the equipment is overloaded or short-circuited, and the primary sensor such as the smoke sensor also detects smoke or temperature rise, the system can quickly determine that it is a potential fire without waiting for confirmation from the secondary sensor. This combination enables the system to issue an early warning in a shorter time, thereby quickly responding to potential fires. When the collaborative device detects an abnormal increase in temperature, especially when accompanied by a signal from the smoke sensor, the system can immediately determine that the situation is a serious fire risk and issue a high-level early warning signal. This situation usually occurs in fires caused by electrical fires or equipment failures. The processing method based on multi-source data fusion can effectively reduce false alarms and improve the accuracy of early warnings, thereby ensuring that the fire protection system can respond to fire risks in a timely and accurate manner.

[0030] P40: When the warning condition is not met, the secondary sensor network layer is awakened according to the activation condition based on the multi-sensor network layer, the secondary sensor monitoring data is extracted, and the secondary sensor monitoring data is identified through the warning identification module. When the secondary warning condition is met, the fire warning information is sent.

[0031] It should be understood that when the monitoring data of the primary sensor fails to meet the warning conditions, the system will not stop monitoring immediately, but will wake up the secondary sensor network layer according to the preset activation conditions, thereby improving the accuracy of fire monitoring. The sensors in the secondary sensor network layer usually include flame detectors or medium-sensitivity thermal sensors, and their main task is to further verify and determine the existence of fire when the primary sensor fails to confirm the fire.

[0032] The activation condition means that when the primary sensor detects a potential fire signal but does not meet the warning standard, the system decides whether to start the secondary sensor for monitoring according to the preset conditions. This mechanism ensures that the system will not issue a false warning due to the misjudgment of a single sensor. For example, the primary smoke sensor may detect smoke, but due to environmental factors (such as cooking smoke or dust), the signal cannot be confirmed as a fire. At this time, the system will activate the secondary sensor, such as a flame detector or a thermal sensor, to further confirm whether there is a fire.

[0033] After the secondary sensor network layer is awakened, the system extracts the secondary sensor monitoring data and inputs it into the early warning recognition module for further analysis. The data from the secondary sensor is usually more targeted at direct signs of fire, such as flame detection and a sharp rise in temperature. By combining the data from the primary and secondary sensors, the system can make a more accurate judgment on the possibility of fire. When the monitoring data from the secondary sensor meets the early warning conditions, the system will immediately send a fire warning message to notify relevant personnel to take emergency measures.

[0034] The sensors in the secondary sensor network layer are usually of medium sensitivity and can provide more targeted fire confirmation capabilities based on the primary sensors. For example, flame detectors can effectively distinguish between fire and no fire interference in a smoky environment. Through this multi-level monitoring mechanism, the system can filter out most false alarms and improve the accuracy of early warning.

[0035] Furthermore, the embodiment of the present application further includes step P50, and step P50 further includes: P51: When the second-level warning conditions are not met, the first-level sensor monitoring data is aligned with the second-level sensor monitoring data, and time series data is introduced to perform trend analysis on the sensor monitoring data; P52: When the trend of the first-level monitoring data and the trend of the second-level monitoring data are synchronously weakened or in the opposite trend, a fire warning termination message is sent.

[0036] In a possible embodiment of the present application, when the monitoring data of the secondary sensor also fails to meet the warning conditions, the system further performs a series of analyses to ensure the accuracy of fire identification and avoid unnecessary false alarms.

[0037] First, establish the alignment of primary sensor monitoring data and secondary sensor monitoring data. This means that the system will synchronously process sensor data from two different levels for a more in-depth analysis. To ensure the rationality of the data, time series data is introduced, and the trend of data changes is determined by trend analysis of the time series. The introduction of the time series data is to avoid false alarms caused by environmental fluctuations in a short period of time. For example, a smoke sensor may detect a rapid increase in smoke concentration in a short period of time (such as caused by cooking or other environmental changes), but these changes are usually not sustained. Therefore, through trend analysis, these short-term fluctuations can be filtered out to ensure that only persistent signs of fire will trigger further early warning mechanisms.

[0038] During the trend analysis process, the system compares the changing trends of the primary monitoring data and the secondary monitoring data. When the monitoring data of the two levels show a synchronous weakening (i.e., the signs of fire weaken), or an opposite trend (such as one sensor showing a decrease in smoke concentration, while the other shows no significant temperature change), the system determines that the possibility of fire is low and immediately sends a fire warning termination message. This trend analysis method can significantly reduce false alarms. For example, a smoke sensor may detect a short-term increase in smoke, but if the secondary sensor (such as a flame detector or temperature sensor) does not detect the corresponding change in the fire, the system will determine that this is a false alarm through trend comparison, thereby terminating the warning process. This process ensures that the system can make more intelligent judgments when facing complex environments and avoid unnecessary alarms due to single abnormal data.

[0039] For example, trend analysis can be based on a time series algorithm, by continuously monitoring sensor data at multiple time points and using mathematical models to analyze the long-term trend of data changes. This approach can effectively distinguish short-term environmental disturbances from real signs of fire. For example, the detected smoke may fluctuate for a short time due to environmental disturbances, but through trend analysis, it will be found that the fluctuation is not continuous or there is no supporting data from other sensors, so no false alarms will be issued. Through this layer-by-layer filtering and trend analysis, the system can minimize false alarms when the early signs of fire are not obvious or uncertain, while ensuring that real fire risks can be captured and responded to in a timely manner.

[0040] Furthermore, step P50 of the embodiment of the present application also includes: P53: When the trend of the first-level monitoring data and the trend of the second-level monitoring data do not meet the fire warning termination conditions, activate the third-level confirmation layer, and the sensor of the third-level confirmation layer is an infrared sensor, including a thermal imaging sensor and an infrared gas imaging sensor; P54: Arrange the infrared monitoring images in time sequence according to the time sequence relationship, identify the abnormal distribution area in the infrared monitoring image, align the features of the abnormal distribution area according to the time sequence relationship, and obtain the abnormal diffusion feature; P55: When the abnormal diffusion feature meets the warning condition, send the fire warning information.

[0041] Optionally, when the monitoring data of the primary and secondary sensors fail to meet the fire warning termination conditions (i.e., signs of fire have not been completely eliminated), the third-level confirmation layer is activated to further confirm the fire through more sophisticated sensors.

[0042] The core sensors of the three-level confirmation layer are infrared sensors, including thermal imaging sensors and infrared gas imaging sensors. The special thing about these sensors is that they can detect temperature changes and gas changes on the surface of objects and hidden areas, and can find fire sources even without visible flames or smoke. For example, in complex scenarios such as inside equipment, walls, and underground fire sources, traditional flame or smoke sensors may not be able to detect fires, while infrared sensors can detect potential fire sources through thermal radiation signals. This level of confirmation is very suitable for detecting fires inside electrical equipment or other hidden fire sources.

[0043] When the third level confirmation layer is activated, the system collects infrared monitoring images provided by infrared sensors. These images record data on surface temperature and gas changes at different time points. The system arranges these images in time sequence and analyzes abnormal changes in temperature and gas in each frame. During image processing, the system identifies abnormal distribution areas, that is, those areas that show abnormal temperature or gas changes. By aligning and analyzing the features of these abnormal areas, the system can extract abnormal diffusion features of these areas, such as the direction and area of ​​temperature diffusion. Through these diffusion features, the system can determine the potential location and development trend of the fire.

[0044] When abnormal diffusion characteristics are identified and these characteristics meet the warning conditions, the system will immediately send a fire warning message. For example, if the infrared sensor detects that the temperature of a certain area inside the equipment is gradually rising, accompanied by gas changes, the system will determine that there is a fire source in this area and send a high-level warning signal. In this way, even if there is no visible flame or obvious smoke, the system can give early warning to avoid the spread of potential fires.

[0045] In summary, the three-level confirmation layer relies on infrared imaging technology and gas imaging technology, which can capture subtle changes in the surface of objects and the surrounding environment when traditional sensors cannot perceive them. This time series analysis technology can effectively identify the diffusion pattern of fire sources, especially in hidden areas, such as fires caused by short circuits in wires and fires inside walls, where the role of infrared sensors is particularly important.

[0046] Furthermore, step P54 of the embodiment of the present application also includes: P54-1: Set a sliding time window, which is the interval time for collecting a frame of the picture, and set the timestamp of the infrared monitoring image according to the collection time; P54-2: Perform gradient recognition on each frame of the infrared monitoring image, mark the image according to the gradient recognition result, and visualize the gradient change area; P54-3: Locate the alignment edge of the infrared monitoring image, use the alignment edge to align each frame of the image, perform time series change analysis on the image mark, and obtain the abnormal diffusion characteristics, including the diffusion direction and diffusion area.

[0047] Specifically, in order to ensure the accuracy of image processing, a sliding time window is first set, which defines the interval time for collecting each frame of infrared monitoring image. The role of the sliding time window is to filter out irrelevant environmental changes in the short term, while providing the system with enough data points for trend analysis. For example, by capturing one frame of image per second within 5 seconds, the trend of temperature or gas changes can be tracked through this time window. In practical applications, infrared sensors capture images at fixed time intervals, that is, frame acquisition. In order to ensure that each frame of image can be correctly identified and processed, a timestamp is set for each frame of image according to the acquisition time to ensure that subsequent timing analysis can be performed in the correct time sequence.

[0048] Furthermore, gradient recognition is performed on each frame of infrared monitoring images. The purpose of gradient recognition is to capture changes in temperature or gas distribution through changes in pixel intensity of the image. Using gradient recognition algorithms (such as the Sobel operator) to calculate horizontal and vertical gradients in the image, the system can identify areas in the image where the temperature changes dramatically. Based on the results of these gradient changes, each frame of the image is marked, and areas with obvious gradient changes are marked as gradient change areas, and these areas are visualized. This marking method enables the system to clearly see changes in temperature or gas distribution, thereby better identifying potential locations of fires.

[0049] When processing multiple frames of images, the system also needs to ensure that the aligned edges between each frame are accurate. The image alignment process refers to aligning the edges of each frame of image so that these images can form a continuous time series. Using the aligned images, the temperature changes between different frames can be analyzed. By analyzing the time series changes of the logo of each frame of image, the system can extract abnormal diffusion features. These features include diffusion direction (i.e., the propagation direction of temperature changes) and diffusion area (i.e., the range covered by temperature changes). By analyzing these features, the location of the fire source and the development trend of the fire can be determined.

[0050] Furthermore, step P54-2 of the embodiment of the present application also includes: P54-21: Calculate the horizontal gradient and vertical gradient of the infrared monitoring image using the Sobel operator using the horizontal gradient convolution kernel and the vertical gradient convolution kernel; P54-22: Synthesize the horizontal gradient and the vertical gradient to obtain the gradient change , where G X is the horizontal gradient, G y is the vertical gradient, and the horizontal gradient convolution kernel is: , the vertical gradient convolution kernel is .

[0051] Optionally, the processing flow of the infrared monitoring image is further refined, and the gradient change of the infrared image is calculated by using the Sobel operator to further identify the area of ​​temperature or gas change.

[0052] First, the Sobel operator is used to calculate the horizontal gradient (Gx) and vertical gradient (Gy) of the infrared monitoring image. The Sobel operator is a commonly used edge detection algorithm that detects areas with drastic brightness changes in the image through convolution operations. Through calculations in both horizontal and vertical directions, the edges with the most significant changes in temperature or gas concentration in the image are identified. The horizontal gradient convolution kernel is used to detect horizontal changes in the image. The following convolution kernel is applied to each pixel of the image: , the vertical gradient convolution kernel is used to detect vertical changes in the image, specifically: The infrared monitoring image is convolved by these convolution kernels to calculate the horizontal and vertical gradient values ​​of each pixel, which are marked as Gx and Gy respectively.

[0053] After obtaining the horizontal gradient (Gx) and vertical gradient (Gy) of each pixel, the system will synthesize these two gradients to obtain the overall gradient change G. The formula for the synthesized gradient is: , which means that the gradient change G is the square root of the sum of the squares of the horizontal gradient and the vertical gradient. Through this synthesis method, the total amplitude of the temperature or gas change of each pixel in the image is calculated. This synthetic gradient result will be used to further analyze the abnormal distribution area of ​​temperature or gas to help identify potential fire hazards.

[0054] Furthermore, step P54-3 of the embodiment of the present application also includes: P54-31: Setting image identification features according to the amount of gradient change, including one or more of color features, gradient quantization, and level marking; P54-32: Performing image identification diffusion annotation of adjacent aligned image frames in sequence according to the time sequence relationship to obtain the diffusion direction and diffusion area; P54-33: Connecting the diffusion direction and diffusion area according to the time sequence relationship to obtain the abnormal diffusion feature.

[0055] It should be understood that the gradient variation of the infrared monitoring image is further processed to identify abnormal diffusion features. This process mainly includes three parts: setting image identification features, diffusion annotation, and connection and analysis of diffusion features.

[0056] First, after obtaining the gradient change of each pixel, the identification features of the image are set according to these changes. Image identification features can include color features, gradient quantization, and level marking. Among them, color features can be used to visualize changes in temperature or gas concentration. For example, different colors are used to represent different gradient areas with temperatures from low to high, which can intuitively show the diffusion trend of temperature. Gradient quantization refers to discretizing continuous gradient changes, dividing the gradient into several levels, and assigning a specific value to each gradient interval. This method can simplify complex gradient changes and facilitate subsequent processing and analysis. Level marking divides gradient changes into different levels, such as low, medium, and high, through pre-set thresholds to distinguish the severity of diffusion areas.

[0057] Based on the image identification features set previously, diffusion annotations are performed on adjacent frames according to the time sequence relationship. Adjacent image frames collected at different time points are aligned, and the diffusion direction and diffusion area are determined by comparing the image identification features between frames (such as color changes or gradient quantization results). For example, if the temperature of a certain area in a certain frame rises and expands to a larger area in the next frame, the system can determine that the temperature of the area is diffusing.

[0058] After determining the diffusion direction and area of ​​each frame, these diffusion features are connected according to the time series relationship. By connecting the diffusion directions and areas in different frames, a complete abnormal diffusion feature is formed, which describes the spatial and temporal distribution characteristics of temperature or gas changes. These abnormal diffusion features can help the system determine the specific location and diffusion trend of the fire source, thereby making accurate fire warnings.

[0059] In summary, the embodiments of the present application have at least the following technical effects: This application analyzes the parameters of the sensor equipment, sets up multiple sensor network layers according to the sensitivity, obtains fire protection parameters and joint parameter monitoring data, and identifies fires through the early warning identification module. When the early warning conditions are met, an early warning message is sent. If the early warning conditions are not met, the secondary sensor network layer is awakened, and the secondary sensor data is extracted to identify the fire again, ensuring multi-level and accurate fire monitoring and early warning.

[0060] The technical effect of integrating multiple sensor devices to provide multi-level warnings and improving the accuracy and reliability of fire warnings has been achieved.

[0061] Embodiment 2 is based on the same inventive concept as the fire monitoring alarm method under the smart fire ecology in the previous embodiment. Figure 2 As shown, this application provides a fire monitoring and alarm platform under the smart fire ecology. The platform and method embodiments in the embodiments of this application are based on the same inventive concept. Among them, the platform includes: The multi-sensor network layer setting module 11 is used to perform parameter analysis on the sensor equipment under the smart fire protection ecology and set the multi-sensor network layer according to the parameter sensitivity, wherein each network layer has activation conditions and warning conditions.

[0062] The first-level sensor monitoring data acquisition module 12 is used to initialize the first-level sensor network layer and obtain the first-level sensor monitoring data, wherein the first-level sensor monitoring data includes fire protection parameter monitoring data and joint parameter monitoring data.

[0063] The first-level fire identification module 13 is used to use the fire parameter monitoring data and the joint parameter monitoring data as input variables, perform fire identification on the monitoring data through the early warning identification module, and send fire warning information when the early warning conditions are met.

[0064] The secondary fire identification module 14 is used to wake up the secondary sensor network layer according to the activation condition based on the multi-sensor network layer when the warning condition is not met, extract the secondary sensor monitoring data, identify the secondary sensor monitoring data through the warning identification module, and send the fire warning information when the secondary warning condition is met.

[0065] Furthermore, the multi-sensor network layer setting module 11 is also used to perform the following steps: The sensor type, sensor distribution, and sensor model parameters in the smart fire protection ecology are obtained to construct a sensor deployment network; the sensors in the sensor deployment network are traversed and matched using the preset multi-level network layer targets to establish a network hierarchical relationship, wherein the first level in the preset multi-level network layer is a smoke sensor, the second level is a flame sensor, and the third level is an infrared sensor, and the unmatched sensors are set as joint sensors; using the fire warning fault case set, with the first level smoke sensor, the second level flame sensor, and the third level infrared sensor as the target, activation conditions and warning conditions are fitted to obtain activation conditions and warning conditions of each level and add them to the network hierarchical relationship; through the fire warning fault case set, a collaborative warning relationship between the joint sensor and sensors at each level is established, and the collaborative warning relationship is fitted to the corresponding network hierarchical relationship, wherein the first level sensor network layer contains at least one collaborative warning relationship of a joint sensor.

[0066] Furthermore, the system also includes: A trend analysis module, which is used to align the first-level sensor monitoring data with the second-level sensor monitoring data when the second-level warning conditions are not met, and introduce time series data to perform trend analysis on the sensor monitoring data; an early warning termination information sending module, which is used to send fire early warning termination information when the trend of the first-level monitoring data and the trend of the second-level monitoring data are synchronously weakened or in the opposite trend.

[0067] Furthermore, the system also includes: A three-level confirmation layer activation module, the three-level confirmation layer activation module is used to activate the three-level confirmation layer when the trend of the first-level monitoring data and the trend of the second-level monitoring data do not meet the fire warning termination condition, the sensor of the three-level confirmation layer is an infrared sensor, including a thermal imaging sensor and an infrared gas imaging sensor; an abnormal diffusion feature acquisition module, the abnormal diffusion feature acquisition module is used to arrange the infrared monitoring image in time sequence according to the time sequence relationship, identify the abnormal distribution area in the infrared monitoring image, align the features of the abnormal distribution area according to the time sequence relationship, and obtain the abnormal diffusion feature; a fire warning information sending module, the fire warning information sending module is used to send the fire warning information when the abnormal diffusion feature meets the warning condition.

[0068] Furthermore, the abnormal diffusion feature acquisition module is also used to perform the following steps: A sliding time window is set, where the sliding time window is the interval time for collecting a frame of picture, and the timestamp of the infrared monitoring image is set according to the collection time; gradient recognition is performed on each frame of the infrared monitoring image, and image identification is performed according to the gradient recognition result, and the gradient change area is visualized; the alignment edge of the infrared monitoring image is located, and each frame of the image is aligned using the alignment edge, and the image identification is analyzed for time series changes to obtain the abnormal diffusion characteristics, including the diffusion direction and the diffusion area.

[0069] Furthermore, the abnormal diffusion feature acquisition module is also used to perform the following steps: The image identification feature is set according to the gradient change, including one or more of color feature, gradient quantization, and level mark; the image identification diffusion annotation of adjacent aligned image frames is performed in sequence according to the time sequence relationship to obtain the diffusion direction and diffusion area; the diffusion direction and diffusion area are connected according to the time sequence relationship to obtain the abnormal diffusion feature.

[0070] Furthermore, the abnormal diffusion feature acquisition module is also used to perform the following steps: The horizontal gradient and vertical gradient of the infrared monitoring image are calculated by Sobel operator using horizontal gradient convolution kernel and vertical gradient convolution kernel respectively; the horizontal gradient and vertical gradient are synthesized to obtain the gradient change amount , where G X is the horizontal gradient, G y is the vertical gradient, and the horizontal gradient convolution kernel is: , the vertical gradient convolution kernel is .

[0071] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0072] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0073] This specification and the drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. Fire supervision and alarm platform under the smart fire protection ecology, It is characterized in that The platform includes: The multi-sensor network layer setting module is used to analyze the parameters of the sensor equipment in the smart fire protection ecology and set the multi-sensor network layer according to the parameter sensitivity, where each network layer has activation conditions and warning conditions; A first-level sensor monitoring data acquisition module is used to initialize the first-level sensor network layer and obtain first-level sensor monitoring data, wherein the first-level sensor monitoring data includes fire protection parameter monitoring data and joint parameter monitoring data; A primary fire identification module, used to use the fire parameter monitoring data and the joint parameter monitoring data as input variables, perform fire identification on the monitoring data through the early warning identification module, and send fire warning information when the early warning condition is met; The secondary fire identification module is used to wake up the secondary sensor network layer according to the activation condition based on the multi-sensor network layer when the warning condition is not met, extract the secondary sensor monitoring data, identify the secondary sensor monitoring data through the warning identification module, and send the fire warning information when the secondary warning condition is met.

2. The fire monitoring and alarm platform under the smart fire ecology as claimed in claim 1, It is characterized in that The secondary sensor monitoring data is identified by the early warning identification module, and then it also includes: When the second-level warning conditions are not met, the first-level sensor monitoring data is aligned with the second-level sensor monitoring data, and the time series data is introduced to perform trend analysis on the sensor monitoring data; When the trend of the primary monitoring data and the trend of the secondary monitoring data are weakening synchronously or in the opposite direction, a fire warning termination message is sent.

3. The fire monitoring and alarm platform under the smart fire ecology as claimed in claim 2, It is characterized in that Also includes: When the trend of the primary monitoring data and the trend of the secondary monitoring data do not meet the fire warning termination condition, the third-level confirmation layer is activated, and the sensor of the third-level confirmation layer is an infrared sensor, including a thermal imaging sensor and an infrared gas imaging sensor; Arrange the infrared monitoring images in time sequence according to the time sequence relationship, identify the abnormal distribution area in the infrared monitoring images, align the features of the abnormal distribution area according to the time sequence relationship, and obtain the abnormal diffusion features; When the abnormal diffusion characteristics meet the warning conditions, the fire warning information is sent.

4. The fire monitoring and alarm platform under the smart fire ecology as claimed in claim 3, It is characterized in that The infrared monitoring images are arranged in time sequence according to the time sequence relationship, the abnormal distribution areas in the infrared monitoring images are identified, and the features of the abnormal distribution areas are aligned according to the time sequence relationship to obtain the abnormal diffusion features, including: Setting a sliding time window, wherein the sliding time window is the interval time for collecting a frame of picture, and setting the timestamp of the infrared monitoring image according to the collection time; Perform gradient recognition on each frame of infrared monitoring image, mark the image according to the gradient recognition result, and visualize the gradient change area; The alignment edge of the infrared monitoring image is located, each frame of the image is aligned using the alignment edge, and the image identification is analyzed for time series changes to obtain the abnormal diffusion characteristics, including the diffusion direction and the diffusion area.

5. The fire monitoring and alarm platform under the smart fire ecology as claimed in claim 4, It is characterized in that Performing time series change analysis on the image identifier to obtain the abnormal diffusion characteristics includes: Setting image identification features according to the gradient change amount, including one or more of color features, gradient quantization, and level marking; Perform image identification diffusion annotation of adjacent aligned image frames in sequence according to the time sequence relationship to obtain diffusion direction and diffusion area; The diffusion direction and the diffusion area are connected according to a time sequence relationship to obtain the abnormal diffusion characteristics.

6. The fire monitoring and alarm platform under the smart fire ecology as claimed in claim 5, It is characterized in that The step of performing gradient recognition on each frame of infrared monitoring image includes: Through the Sobel operator, the horizontal gradient convolution kernel and the vertical gradient convolution kernel are used to calculate the horizontal gradient and vertical gradient of the infrared monitoring image respectively; The horizontal gradient and the vertical gradient are synthesized to obtain the gradient change amount , where G X is the horizontal gradient, G y is the vertical gradient, and the horizontal gradient convolution kernel is: , the vertical gradient convolution kernel is .

7. The fire monitoring and alarm platform under the smart fire ecology as claimed in claim 1, It is characterized in that The sensor equipment under the smart fire protection ecology is analyzed for parameters, and a multi-sensor network layer is set according to the parameter sensitivity, wherein each network layer has activation conditions and warning conditions, including: Obtain sensor types, sensor distribution, and sensor model parameters in the smart firefighting ecosystem, and build a sensor deployment network; Using a preset multi-level network layer target to traverse and match the sensors in the sensor deployment network, and establish a network hierarchical relationship, wherein the first level of the preset multi-level network layer is a smoke sensor, the second level is a flame sensor, the third level is an infrared sensor, and the unmatched sensor is set as a joint sensor; Using the fire warning fault case set, with the first-level smoke sensor, the second-level flame sensor, and the third-level infrared sensor as the target, the activation conditions and warning conditions are fitted, and the activation conditions and warning conditions of each level are obtained and added to the network hierarchical relationship; Through the fire warning fault case set, a collaborative warning relationship between the joint sensor and sensors at each level is established, and the collaborative warning relationship is fitted into the corresponding network level relationship, wherein the first-level sensor network layer contains at least one collaborative warning relationship of the joint sensor.

8. Fire supervision and alarm methods under the smart fire ecology, It is characterized in that The method comprises: Analyze the parameters of sensor equipment in the smart fire protection ecosystem, and set up multiple sensor network layers according to parameter sensitivity, where each network layer has activation conditions and warning conditions; Initializing a first-level sensor network layer to obtain first-level sensor monitoring data, wherein the first-level sensor monitoring data includes fire protection parameter monitoring data and joint parameter monitoring data; The fire parameter monitoring data and the joint parameter monitoring data are used as input variables, and the monitoring data are used for fire identification through the early warning identification module, and when the early warning conditions are met, fire warning information is sent; When the warning condition is not met, the secondary sensor network layer is awakened according to the activation condition based on the multi-sensor network layer, the secondary sensor monitoring data is extracted, and the secondary sensor monitoring data is identified through the warning identification module. When the secondary warning condition is met, the fire warning information is sent.

Citation Information

Patent Citations

  • Fire intelligent detection system and control method thereof

    CN109887221A

  • Intelligent fire protection graded early warning method and system based on deep learning

    CN110070690A

  • Security inspection early warning method and system

    CN112286190A

  • Cable fire monitoring device

    CN115641686A

  • Intelligent fire early warning method and system for electric power system integrated with multi-source data

    CN117765481A

Cited By

  • Fire-fighting host information processing method and device, equipment and storage medium

    CN120390023A