Integrated comprehensive monitoring station based on multi-dimensional fire factor perception

Abstract

The invention provides an integrated comprehensive monitoring station based on multi-dimensional fire factor perception, and relates to the technical field of forest and grassland fire control, and the integrated comprehensive monitoring station comprises a front-end perception unit and a system management unit. The front-end sensing unit comprises at least two of a soil environment monitoring module, a surface litter monitoring module, a vegetation phenology monitoring module and an atmospheric environment monitoring module; the system management unit comprises an edge computing gateway; and the edge computing gateway is connected with each monitoring module in the front-end sensing unit and is used for carrying out fusion evaluation based on monitoring data output by each monitoring module. According to the integrated comprehensive monitoring station provided by the invention, multi-dimensional fire factor sensing and intelligent fusion evaluation capabilities are integrated and deployed in a monitoring site, so that the fire monitoring accuracy is improved; data fusion and intelligent evaluation are completed locally through the edge computing gateway, the time delay from data acquisition to early warning release is greatly shortened, and the timeliness and accuracy of fire early warning are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of forest and grassland fire protection technology, and in particular to an integrated monitoring station based on multi-dimensional fire risk factor perception. Background Technology

[0002] Forest and grassland fires are major global natural disasters, posing a serious threat to ecosystems, biodiversity, and the safety of people's lives and property. Establishing a precise and timely fire monitoring and early warning system is crucial for effectively preventing and controlling the occurrence and spread of fires.

[0003] Forest and grassland fire monitoring technologies mainly include satellite remote sensing and ground station monitoring, but both have significant limitations. Satellite remote sensing achieves fire monitoring over large areas through multi-satellite networking. However, its data processing is complex; it typically takes 20 to 40 minutes from satellite imaging to ground analysis of the fire location, resulting in a serious early warning lag and failing to meet the timeliness requirements of early detection and suppression in the early stages of a fire. Furthermore, this system is susceptible to interference from meteorological conditions such as clouds and smoke, and cannot acquire key parameters such as soil moisture. Ground station monitoring integrates micro-meteorological sensors and combustible material moisture content sensors, enabling the collection of some ground factors. However, its monitoring dimensions are still insufficient, lacking synchronous sensing of key factors such as soil profile moisture and physiological moisture content of living vegetation (through multispectral inversion), leading to inaccurate fire risk assessment models. In addition, its design for long-term power supply, multi-mode communication, and comprehensive protection in harsh field environments is inadequate.

[0004] Therefore, improving the accuracy of fire monitoring and the timeliness of fire early warning has become a pressing technical problem for the industry. Summary of the Invention

[0005] The integrated monitoring station based on multi-dimensional fire hazard factor perception provided in this application is used to solve the technical problem of how to improve the accuracy of fire monitoring and the timeliness of fire early warning.

[0006] This application provides an integrated monitoring station based on multi-dimensional fire hazard factor perception, including a front-end perception unit and a system management unit; The front-end sensing unit includes at least two of the following: a soil environment monitoring module, a surface litter monitoring module, a vegetation phenology monitoring module, and an atmospheric environment monitoring module. The soil environment monitoring module is used to monitor the soil moisture content of the soil moisture profile; The surface litter monitoring module is used to monitor the moisture content of surface litter. The vegetation phenology monitoring module is used to monitor the phenological and moisture data of living vegetation; The atmospheric environment monitoring module is used to monitor meteorological data of the atmospheric environment; The system management unit includes an edge computing gateway; the edge computing gateway is connected to each monitoring module in the front-end sensing unit, and is used to perform fusion evaluation based on the monitoring data output by each monitoring module to generate fire risk warning level and fire risk change trend.

[0007] In some embodiments, the edge computing gateway includes a data preprocessing module, a fusion evaluation module, a weight adjustment module, and a trend correction module; The data preprocessing module is used to preprocess various monitoring data; the preprocessing includes data standardization; the monitoring data includes at least one of the meteorological data, soil moisture content, surface litter moisture content, and phenological moisture data. The fusion evaluation module is used to perform fusion calculation on the preprocessed monitoring data based on the calculated weights of each monitoring data to obtain a comprehensive fire risk index. The weight adjustment module is used to match a preset weight configuration file based on at least one of the phenological moisture data, the current month and historical fire data, and to determine the calculated weight of each monitoring data and the calculated weight of each influencing factor in the meteorological data based on the matching results. The trend correction module is used to smooth the comprehensive fire risk index at the current moment and the comprehensive fire risk index at historical moments and identify trends to obtain the smoothed value of the comprehensive fire risk index and the rate of change of the comprehensive fire risk index. The smoothed value of the comprehensive fire risk index is used to determine the fire risk warning level; the rate of change of the comprehensive fire risk index is used to determine the fire risk trend.

[0008] In some embodiments, the fusion evaluation module includes a linear weighted calculation submodule and a nonlinear correction submodule; The linear weighted calculation submodule is used to perform fusion calculation on the preprocessed monitoring data based on the calculation weight of each monitoring data to obtain the basic fire risk index. The nonlinear correction submodule is used to determine a coordination effect coefficient based on at least two meteorological factors, including air temperature and wind speed, and a comprehensive aridity determined by at least one of the soil moisture content, the surface litter moisture content, and the phenological moisture data; and to perform nonlinear correction on the basic fire risk index based on the coordination effect coefficient to obtain the comprehensive fire risk index.

[0009] In some embodiments, the trend correction module includes a smoothing submodule and a trend recognition submodule; The smoothing submodule is used to perform exponential moving average smoothing on the current comprehensive fire risk index and the comprehensive fire risk index at historical times to obtain the smoothed value of the comprehensive fire risk index. The trend recognition submodule is used to perform linear fitting based on the comprehensive fire risk index at the current moment and the comprehensive fire risk index at historical moments, and obtain the rate of change of the comprehensive fire risk index based on the fitting result.

[0010] In some embodiments, a neural network model is deployed in the edge computing gateway; The edge computing gateway is used to input the monitoring data output by each monitoring module into the neural network model to obtain the fire risk warning level and fire risk change trend output by the neural network model.

[0011] In some embodiments, the neural network model is trained by a cloud server based on monitoring data samples, as well as the actual values ​​of the fire risk warning level and the actual values ​​of the fire risk change trend corresponding to the monitoring data samples. The edge computing gateway updates the neural network model in real time based on the model parameters issued by the cloud server.

[0012] In some embodiments, the system management unit further includes a power supply module; the power supply module includes a battery and a power control submodule; The battery is used to provide power to the front-end sensing unit and the edge computing gateway. The power control submodule is used to monitor the remaining power of the battery and dynamically adjust the data acquisition frequency of the front-end sensing unit and the result reporting frequency of the edge computing gateway based on the remaining power.

[0013] In some embodiments, the power supply module further includes a mains power access submodule and a solar panel; The solar panel is used to provide a first power source; The mains power access submodule is used to provide a second power source.

[0014] In some embodiments, the equipment column and protective enclosure are also included; Each monitoring module in the front-end sensing unit is installed from bottom to top along the equipment column; The system management unit is located inside the protective box.

[0015] In some embodiments, a lightning rod is installed on the top of the equipment column; a metal grounding grid is laid at the bottom of the equipment column; and surge protectors are installed on the power lines and signal lines connecting the various modules in the front-end sensing unit and the system management unit.

[0016] The integrated monitoring station based on multi-dimensional fire hazard factor perception provided in this application integrates the perception and intelligent fusion assessment capabilities of multi-dimensional fire hazard factors and deploys it on the monitoring site. It realizes the synchronous and common-source collection of four major categories of key fire hazard factors: atmosphere, soil, surface combustibles, and living vegetation. This solves the fragmentation problem caused by the scattered sources and inconsistent standards of related technical data, and improves the accuracy of fire monitoring. At the same time, it completes data fusion and intelligent assessment locally through an edge computing gateway, directly generating high-level early warning information. This greatly shortens the time delay from data collection to early warning issuance, significantly improves the timeliness and accuracy of fire warnings, and provides strong technical support for achieving early detection and early suppression of fires. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0018] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is one of the structural schematic diagrams of the integrated monitoring station provided in this application.

[0020] Figure 2 This is the second structural schematic diagram of the integrated monitoring station provided in this application.

[0021] Figure 3 This is a schematic diagram of the intelligent adaptive working mode provided in this application.

[0022] Figure 4 This is a schematic diagram of the power module provided in this application.

[0023] Figure 5 This is a schematic diagram of the hardware composition and three-dimensional structure of the integrated monitoring station provided in this application.

[0024] Figure 6 This is a schematic diagram of the equipment architecture of the integrated monitoring station provided in this application. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first," "second," etc., used in this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps, units, or modules is not necessarily limited to those explicitly listed, but may include other steps, units, or modules not explicitly listed or inherent to such processes, methods, products, or devices.

[0027] The large number of satellite-taken "photos" translates to a massive amount of data, leading to problems such as time-consuming network transmission and complex processing. From the moment the satellite takes the photo to the precise location of a forest fire, it still takes 20 to 40 minutes. This is clearly too slow for forest fire fighting where every second counts. Currently, satellite data processing involves transmitting the data to the ground first, and then processing it using computers. The time-consuming data transmission and high processing complexity mean that current satellite remote sensing technology is still some distance from achieving the goals of "early detection and early suppression" in forest fire prevention.

[0028] Current forest fire monitoring primarily relies on independent equipment monitoring single factors (such as atmospheric temperature and humidity or soil moisture), leading to issues like data fragmentation and delayed early warnings. Traditional monitoring stations struggle to achieve collaborative analysis of multi-dimensional fire risk factors, resulting in insufficient accuracy in fire risk assessments and an inability to meet the disaster prevention and control needs in complex environments. While some existing technologies integrate meteorological parameters, they lack real-time monitoring of key factors such as soil moisture and vegetation flammability, and data transmission efficiency is low, hindering rapid decision-making.

[0029] In order to address the shortcomings of related technologies, Figure 1 This is one of the structural schematic diagrams of the integrated monitoring station provided in this application, such as... Figure 1 As shown, the integrated monitoring station includes a front-end sensing unit 100 and a system management unit 200.

[0030] The front-end sensing unit 100 includes at least two of the following: a soil environment monitoring module 110, a surface litter monitoring module 120, a vegetation phenology monitoring module 130, and an atmospheric environment monitoring module 140.

[0031] The soil environment monitoring module is used to monitor the soil moisture content of the soil moisture profile; The surface litter monitoring module is used to monitor the moisture content of surface litter. The vegetation phenology monitoring module is used to monitor the phenological and moisture data of living vegetation; The atmospheric environment monitoring module is used to monitor meteorological data of the atmospheric environment; The system management unit 200 includes an edge computing gateway 210. The edge computing gateway is connected to each monitoring module in the front-end sensing unit and is used to perform fusion evaluation based on the monitoring data output by each monitoring module to generate fire risk warning levels and fire risk change trends.

[0032] Specifically, the integrated monitoring station mainly includes a front-end sensing unit and a system management unit.

[0033] The core function of the front-end sensing unit is to achieve real-time monitoring of various environmental factors affecting the occurrence and spread of fire. In the embodiments of this application, the front-end sensing unit can integrate various types of monitoring modules to construct a three-dimensional sensing system from underground to the surface and then to near-ground space.

[0034] The front-end sensing unit includes at least two of the following modules: soil environment monitoring module, surface litter monitoring module, vegetation phenology monitoring module, and atmospheric environment monitoring module. This "at least two" configuration means that, in specific applications, these modules can be flexibly combined based on the typical characteristics of the monitoring area (for example, areas with sparse forest undergrowth may not require a surface litter monitoring module, and areas without living vegetation may not require a vegetation phenology monitoring module), cost budget, and requirements for early warning accuracy.

[0035] In one specific embodiment, the front-end sensing unit may include all four monitoring modules to achieve the most comprehensive collection of fire hazard factor data. The following is a detailed description of each monitoring module.

[0036] The soil environment monitoring module is used to monitor soil moisture content in soil moisture profiles. Soil moisture profiles refer to the distribution of soil moisture at different depths, directly affecting the dryness of deep combustibles (such as roots) and the sustained combustion capacity of a fire. To achieve profile monitoring, this module can include multiple soil moisture sensors, which are buried at different depths around the monitoring station. For example, one sensor can be installed at depths of 5cm, 10cm, 20cm, and 30cm to obtain a set of soil moisture data reflecting the vertical distribution of soil moisture.

[0037] The surface litter monitoring module is used to monitor the moisture content of surface litter. Surface litter, such as dead branches and fallen leaves, is the primary initial ignition source and fuel for fire spread in forest fires, and its moisture content is the most critical factor determining its flammability. This module can be a multispectral phenological sensor; it can also be a dedicated litter moisture content sensor, whose probe can be directly inserted into or placed over the litter layer on the ground surface to measure the moisture content of the litter in real time using resistance, capacitance, or other principles.

[0038] The vegetation phenology monitoring module is used to monitor the phenological moisture data of living vegetation. The health status and water content of living vegetation are also important factors affecting fire risk, especially during dry seasons when living vegetation can become combustible material. Phenological moisture data is a comprehensive indicator used to characterize the growth status and water stress of living vegetation. Specifically, this module can be a multispectral combustible phenological sensor, for example, capable of acquiring spectral data in specific wavelength bands (such as blue, green, red, red-edged, near-infrared, etc.). By calculating from these spectral data, the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) can be derived. These indices are phenological moisture data that can quantitatively assess vegetation biomass, growth stage, and real-time water content.

[0039] The atmospheric environment monitoring module is used to monitor meteorological data of the atmospheric environment. This module can be an integrated multi-element meteorological sensor, usually installed on the top of the monitoring station column to avoid obstruction, and can simultaneously monitor multiple meteorological parameters, including but not limited to: air temperature, air humidity, wind speed, wind direction, rainfall, atmospheric pressure, and ambient light intensity.

[0040] In addition, the front-end sensing unit may also include high-precision temperature and humidity sensors. These sensors are integrated on top of the combustible soil moisture sensor and use a non-contact measurement method to directly measure air temperature and humidity within 5 cm above the ground surface.

[0041] The system management unit is responsible for data processing, communication, and device management. The system management unit may include a communication and management unit, a power supply system, etc. In this embodiment, its core component is an edge computing gateway.

[0042] An edge computing gateway can be an industrial-grade computing device with an embedded high-performance processor, integrated into the intelligent management and protection box of a monitoring station. It connects to various monitoring modules in the front-end sensing unit via wired or wireless means (e.g., through RS485 bus, analog signal interface, etc.) to receive the monitoring data collected by them.

[0043] The key function of an edge computing gateway lies in its ability to fuse and evaluate multi-source, heterogeneous monitoring data at the data source (i.e., locally at the monitoring station). This monitoring data includes meteorological data, soil moisture content, surface litter moisture content, and phenological moisture data collected by one or more of the aforementioned modules. Fusion evaluation refers to the specific algorithms or models running within the edge computing gateway. These algorithms or models can comprehensively analyze these data from different dimensions and with different physical meanings to determine the overall fire risk level formed by their combination. For example, even if the temperature is not high, if the wind speed is extremely high, and the moisture content of the soil, litter, and vegetation has been at extremely low levels for several consecutive days, the edge computing gateway can still determine that the current situation is a high fire risk through fusion evaluation.

[0044] Ultimately, based on the results of the fusion assessment, the edge computing gateway generates two key outputs: fire risk warning level and fire risk change trend.

[0045] Fire risk warning levels are static assessments of the current fire risk status, and can be divided into several discrete levels such as "low fire risk", "medium fire risk", "relatively high fire risk" and "high fire risk".

[0046] Fire risk trend describes the dynamic changes in fire risk status and is used to warn of the accumulation or sudden changes in risk. For example, it can output trend information such as "stable", "slowly rising", or "risk increasing sharply".

[0047] Edge computing gateways can also transmit monitoring data output by various monitoring modules back to cloud servers to achieve the same fire risk analysis technology.

[0048] The integrated monitoring station provided in this application combines the perception and intelligent fusion assessment capabilities of multi-dimensional fire hazard factors and deploys them at the monitoring site. It achieves synchronous and homogeneous collection of four major categories of key fire hazard factors: atmosphere, soil, surface combustibles, and living vegetation. This solves the fragmentation problem caused by the dispersed sources and inconsistent standards of related technical data, and improves the accuracy of fire monitoring. At the same time, it completes data fusion and intelligent assessment locally through an edge computing gateway, directly generating high-level early warning information. This greatly shortens the time delay from data collection to early warning issuance, significantly improves the timeliness and accuracy of fire warnings, and provides strong technical support for achieving early detection and early suppression of fires.

[0049] In some embodiments, the edge computing gateway includes a data preprocessing module, a fusion evaluation module, a weight adjustment module, and a trend correction module; The data preprocessing module is used to preprocess various monitoring data; the preprocessing includes data standardization; the monitoring data includes at least one of meteorological data, soil moisture content, surface litter moisture content, and phenological moisture data; The fusion assessment module is used to perform fusion calculations on the preprocessed monitoring data based on the calculated weights of each monitoring data to obtain a comprehensive fire risk index. The weight adjustment module is used to match at least one of the following: phenological moisture data, current month and historical fire data, with a preset weight configuration file. Based on the matching results, the module determines the calculated weight of each monitoring data and the calculated weight of each influencing factor in the meteorological data. The trend correction module is used to smooth the current comprehensive fire risk index and the comprehensive fire risk index at historical times and identify trends to obtain the smoothed value of the comprehensive fire risk index and the rate of change of the comprehensive fire risk index. The smoothed value of the composite fire risk index is used to determine the fire risk warning level; the rate of change of the composite fire risk index is used to determine the trend of fire risk change.

[0050] Specifically, the data preprocessing module, fusion evaluation module, weight adjustment module, and trend correction module can be implemented by independent hardware modules or by corresponding software programs. These four modules work together to form a complete intelligent processing engine from raw data to final early warning information, making fire risk assessment not only multi-dimensional but also adaptive and dynamically forward-looking.

[0051] The main purpose of the data preprocessing module is to address the issues of heterogeneous sources and inconsistent dimensions of various monitoring data collected by the front-end sensing unit. For example, in meteorological data, temperature is measured in °C (degrees Celsius), humidity in % (percentage), and wind speed in m / s (meters per second); while soil moisture content is expressed as a volume percentage; and phenological moisture data (such as NDVI) is a dimensionless index. If these data are directly fused and calculated without processing, their physical meaning will be unclear, and variables with large numerical ranges will disproportionately dominate the calculation results, leading to distorted assessments.

[0052] Therefore, the data preprocessing module is used to preprocess the received monitoring data. The monitoring data here may include at least one or any combination of meteorological data, soil moisture content, surface litter moisture content, and phenological moisture data mentioned in the foregoing embodiments.

[0053] In this embodiment, the core step of preprocessing is data standardization. The purpose is to map all monitoring data values ​​from different sources to a unified, dimensionless interval, such as [0, 1]. Generally, within this interval, a larger value indicates a higher contribution to fire risk. A feasible standardization method is maximum-minimum normalization. For example, for positively correlated factors such as temperature and wind speed, their standardized values ​​can be calculated using the formula (current value - period minimum) / (period maximum - period minimum). For negatively correlated factors such as humidity, rainfall, and moisture content, they can be calculated using formula 1 - (current value - period minimum) / (period maximum - period minimum) to ensure that their standardized values ​​also conform to the rule that "the larger the value, the higher the contribution to fire risk." The output of this module is a set of standardized data that can be compared and calculated on the same scale.

[0054] The standardization of meteorological data can be expressed as: ; ; ; .

[0055] in, Air temperature (°C); This represents the minimum value of the air temperature cycle. This represents the maximum value of the air temperature cycle. This is a periodic standardized value for air temperature; Air humidity (%) This represents the minimum value of the air humidity cycle. This represents the maximum value of the air humidity cycle. This is a periodic standardized value for air humidity. Wind speed (m / s); This represents the minimum value of the wind speed period; This represents the maximum value of the wind speed cycle; This is the standardized value for the wind speed cycle; Rainfall (mm); This represents the minimum value of the rainfall cycle; This represents the maximum value of the rainfall cycle; This is a standardized value for the rainfall cycle.

[0056] The maximum and minimum values ​​of the above factors can be set based on historical climate data or seasonal dynamics at the location of the monitoring station.

[0057] Meteorological factor coefficient It is a weighted sum of the standardized values ​​of air temperature, humidity, wind speed, and rainfall, and its calculation formula is: .

[0058] The default weighting ratio for meteorological factor coefficients is as follows: , , , .

[0059] The standardized form of soil water content (SWC) can be expressed as: 。

[0060] in, Soil volumetric moisture content (%) This represents the periodic minimum value of soil volumetric moisture content; This represents the periodic maximum value of soil volumetric water content; This is the periodic standardized value of soil volumetric moisture content.

[0061] The standardized value of litter water content (LWC) can be expressed as: 。

[0062] in, Moisture content of surface litter (%) This represents the periodic minimum moisture content of surface litter. This represents the periodic maximum moisture content of surface litter. This is the periodic standardized value of the moisture content of surface litter.

[0063] The standardized form of vegetation phenological water index (VWI) data can be represented as: 。

[0064] in, The vegetation phenological moisture index value; This represents the periodic minimum value of the vegetation phenological moisture index. VWI represents the periodic maximum value of the vegetation phenological moisture index; VWI represents the periodic standardized value of the vegetation phenological moisture index.

[0065] The weight adjustment module reflects the adaptability of this application's embodiments. The importance of various fire hazard factors changes dynamically depending on the season, vegetation growth stage, and even weather conditions. For example, during the peak vegetation growth season, the moisture content of living vegetation is crucial in determining regional fire risk; while during the dormant period when vegetation is withered, the dryness of surface litter becomes more important.

[0066] The function of the weight adjustment module is to dynamically determine the calculation weight of each monitoring data in the subsequent fusion calculation based on the current environmental background. It takes at least one of the following as input: phenological moisture data (such as NDVI index, which reflects vegetation growth status), current month (which reflects seasonal climate characteristics), and historical fire data (which reflects the historical high-risk pattern of the area), and uses this input information to match the preset weight configuration file.

[0067] The preset weight configuration file can be a lookup table or rule base stored within the edge computing gateway, which predefines various environmental states (such as "growing season", "dormant season", "drought state", "strong wind state", "historical high-risk period", etc.) and their corresponding weight sets. For example, when the module identifies that the current NDVI value is high and the month is the growing season, it will match the weight set corresponding to "growing state" from the configuration file. This set may assign a higher calculation weight to "phenological moisture data" and a relatively lower weight to "soil moisture content".

[0068] The output of this module consists of two key sets of calculated weights: one set is the macroscopic calculated weights between various monitoring data (meteorology, soil, litter, vegetation); the other set is the calculated weights of various influencing factors (such as temperature, humidity, wind speed, and rainfall) within the meteorological data.

[0069] In one specific embodiment, the input parameters of the weight adjustment module include: the current normalized vegetation index (NDVI), the current month, the current vegetation moisture index (VWI), the recent average wind speed (Wind_Avg), and the historical high-frequency fire risk flag (History_Risk_Flag).

[0070] The input parameters for the weight adjustment module include: the updated calculated weights ( , , , These correspond to meteorological data, soil moisture content, surface litter moisture content, and phenological moisture data, respectively, as well as the updated default weights within the meteorological data. , , , These correspond to air temperature, humidity, wind speed, and rainfall, respectively.

[0071] The simplified algorithm flow for the weight adjustment module is as follows: Step 1: Status recognition.

[0072] The weight adjustment module identifies the current main environmental state based on the input parameters. State types include, but are not limited to: dry state (when VWI or LMC is extremely low), windy state (when Wind_Avg is consistently high), growing state (when NDVI is high and it is the growing season), dormant state (when NDVI is low and it is the non-growing season), and historically high-risk state (when History_Risk_Flag is true).

[0073] Step 2: Weight matching.

[0074] The system has a pre-stored weight preset table corresponding to each environmental state. After identifying the current dominant state, the corresponding basic weight set is directly called from the table. For example: (1) Growth state: reduce soil moisture weight Increase the weight of vegetation moisture status Because living vegetation is the main combustible material. (2) Drought conditions: significantly increase the weight of surface litter moisture content. and soil weight (3) Strong wind conditions: Increase the weight of wind speed among meteorological factors. .

[0075] Step 3: Weight fusion and normalization.

[0076] If multiple states are matched simultaneously, a weighted average of the multiple sets of basic weights is calculated to obtain the final weight. Finally, the weights are normalized to ensure that their sum is 1.

[0077] The fusion assessment module is the core of the fire risk index calculation. It receives standardized data from the data preprocessing module and dynamically calculated weights from the weight adjustment module. Based on these dynamic weights, the fusion assessment module performs a fusion calculation on the preprocessed monitoring data to obtain a comprehensive value (comprehensive fire risk index) that fully reflects the combined impact of multiple factors. A specific implementation method is weighted summation, which involves multiplying the standardized values ​​of various monitoring data by their corresponding weights and then summing all the results to obtain a preliminary fire risk index. This comprehensive fire risk index is an instantaneous value over a time series, reflecting the comprehensive fire risk level at the current moment.

[0078] The trend correction module aims to address the issue of early warning of dynamic changes in fire risk, rather than simply providing a static level. The urgency of danger is entirely different between a process that slowly rises to "high fire risk" and a process that rapidly surges from "medium fire risk" to "high fire risk" in a short period of time.

[0079] The trend correction module performs smoothing and trend identification by analyzing the current comprehensive fire risk index and a series of historical comprehensive fire risk indices (i.e., a time series).

[0080] Smoothing is used to filter out exponential noise caused by instantaneous sensor fluctuations or minor environmental variations, resulting in a smoothed composite fire risk index value that better reflects the true baseline fire risk level. Trend identification is used to calculate the recent rate of change of the fire risk index, i.e., the composite fire risk index change rate.

[0081] Ultimately, these two output values ​​are used to generate the final warning information. The smoothed value of the composite fire risk index is compared with preset thresholds (e.g., 0.3, 0.5, 0.7) to determine the specific fire risk warning level (e.g., low, medium, high, high fire risk). If the rate of change of the composite fire risk index exceeds a certain threshold, it can be used to determine the trend of fire risk changes and trigger an emergency alarm such as "rapid risk escalation".

[0082] The integrated monitoring station provided in this application embodiment designs the edge computing gateway as an algorithm flow containing four functional modules. It can not only integrate multi-dimensional data, but also realize the model's environmental adaptation (through weight adjustment) and dynamic risk early warning (through trend correction). As a result, it can generate fire risk early warning information that is far more accurate, forward-looking and of greater decision-making reference value than traditional fixed models, which greatly improves the precision prevention and control capabilities of forest and grassland fires.

[0083] In some embodiments, the fusion evaluation module includes a linear weighted calculation submodule and a nonlinear correction submodule; The linear weighted calculation submodule is used to perform fusion calculation on the preprocessed monitoring data based on the calculation weight of each monitoring data to obtain the basic fire risk index. The nonlinear correction submodule is used to determine the coordination effect coefficient based on at least two meteorological factors, including air temperature and wind speed, and the comprehensive aridity determined by at least one of soil moisture content, surface litter moisture content, and phenological moisture data; and to perform nonlinear correction on the basic fire risk index based on the coordination effect coefficient to obtain the comprehensive fire risk index.

[0084] Specifically, the fusion assessment module adopts a fusion architecture of "linear weighted basic model + nonlinear correction factor" to calculate the basic fire risk index. With the comprehensive fire risk index .

[0085] The linear weighted calculation submodule performs linear fusion calculations on the preprocessed monitoring data based on the calculation weights to obtain the basic fire risk index. This can be expressed as a formula: .

[0086] in, , , , Meteorological data ( ), soil moisture content ( ), surface litter moisture content ( ) and phenological moisture data ( The weights are calculated as follows; the default weighting is: , , , =0.2.

[0087] The nonlinear correction submodule, based on seasonal climate conditions or regional vegetation characteristics (such as high temperatures and strong winds combined with extremely dry vegetation and soil), adds a nonlinear synergistic enhancement effect between meteorological factors and combustible dryness, introducing a correction factor based on a product term. This can be expressed as a formula: 。

[0088] in, The coefficient of synergistic effect. =0.5, which can be adjusted according to the regional vegetation type. The product of the three terms in parentheses reflects the coupling effect of "high temperature," "strong wind," and "overall dryness." When all three are high, This will amplify and correct the basic fire risk index.

[0089] Comprehensive fire risk index The calculation formula can be expressed as: 。

[0090] Among them, a nonlinear correction is introduced by using a product method, and the upper limit of the exponent is clamped at 1.0 by the min function.

[0091] The integrated monitoring station provided in this application not only considers the independent contributions of various fire hazard factors, but also creatively quantifies the synergistic enhancement effect when extreme conditions such as high temperature, strong wind, and dryness are superimposed; it can more accurately capture the nonlinear abrupt change process of fire risk, thereby issuing more timely and stronger early warning signals under the most dangerous critical conditions, significantly improving the scientificity and reliability of fire monitoring.

[0092] In some embodiments, the trend correction module includes a smoothing submodule and a trend recognition submodule; The smoothing submodule is used to perform exponential moving average smoothing on the current comprehensive fire risk index and the comprehensive fire risk index at historical times to obtain the smoothed value of the comprehensive fire risk index. The trend recognition submodule is used to perform linear fitting based on the current comprehensive fire risk index and the comprehensive fire risk index at historical times, and to obtain the rate of change of the comprehensive fire risk index based on the fitting results.

[0093] Specifically, the trend correction module considers the cumulative effect of fire risk and early warning of sudden changes, and performs time-series-based smoothing and trend reinforcement on the final index.

[0094] The smoothing submodule for the current time Comprehensive fire risk index and historical moments Comprehensive fire risk index After performing exponential moving average smoothing, the smoothed value of the comprehensive fire risk index is obtained, expressed by the formula: 。

[0095] in, As a smoothing factor, =0.3, used to filter out short-term noise and obtain a more stable fire hazard baseline. .

[0096] The trend identification submodule is used to perform linear fitting based on the current comprehensive fire risk index and the comprehensive fire risk index at historical times to obtain the rate of change of the comprehensive fire risk index, expressed by the formula: 。

[0097] in, This is the rate of change (slope) of the comprehensive fire risk index. The number of historical moments involved in the fitting calculation (which can be the most recent hour, for example) ); This represents a historical moment (the time scale can be measured in hours). For the first A historic moment; express The average of historical moments; for The average of the comprehensive fire risk index at a historical moment; For the first The comprehensive fire risk index at a historical moment.

[0098] like > (Slope threshold, default 0.05 / hour), and currently... If the risk level is already at or above the medium level, the "rapidly rising risk" indicator will be triggered, and this will be specifically highlighted in the warning message.

[0099] The fire hazard level ultimately used for early warning is based on a smoothed index. And combine trend indicators to make a judgment: In 0 < If the value is ≤0.3, the fire risk is low. In 0.3 < If the value is ≤0.5, it is considered a medium fire hazard. In 0.5 < If the value is ≤0.7, it is considered a high fire risk; In 0.7 < If the value is ≤1.0, it is considered a high fire hazard. When the fire is at a "higher fire risk" or "higher fire risk" level and the system detects a "rapidly rising risk" indicator, the warning message will be upgraded to "high risk and the risk is increasing sharply".

[0100] The integrated monitoring station provided in this application embodiment can effectively extract stable and reliable fire risk levels from noisy time series data and can keenly capture the dynamic changing trend of risks.

[0101] The following is a specific example. Assume the following set of environmental data has been collected: Standardized results of meteorological factor variables: =0.75, =0.625, =0.25, =1.0, then the meteorological factor coefficient (MET) = 0.6625. Soil moisture content coefficient (SWC) = 0.667. Surface litter moisture content coefficient (LMC) = 0.8. Vegetation phenology-moisture content coefficient (VWI) = 0.5.

[0102] Calculate the basic fire risk index: =0.4 0.6625 + 0.2 0.667 + 0.2 0.8 + 0.2 0.5 = 0.6584.

[0103] Calculate the nonlinear synergistic correction: Overall dryness = (0.667 + 0.8 + 0.5) / 3 = 0.657, dryness factor = 1 - 0.6557 = 0.3443.

[0104] =1+0.5 (0.75) 0.25 (0.3443) = 1 + 0.5 0.06456 = 1.03228.

[0105] Calculate the final composite fire risk index: =min(0.6584) 1.03228, 1.0) = 0.6796.

[0106] Trend Analysis: Assuming the previous 6-hour index slope K = 0.08 > 0.05 (threshold), and =0.65.

[0107] Judgment result: =0.65, which falls under the "higher fire risk" category, and is also marked with a "rapidly rising risk" indicator. The system will issue a more urgent warning than the original algorithm (which only judged it as a higher fire risk), such as "The fire risk level is high, and the risk is accumulating rapidly, requiring strengthened patrols and vigilance."

[0108] In some embodiments, a neural network model is deployed in the edge computing gateway; The edge computing gateway is used to input the monitoring data output by each monitoring module into the neural network model to obtain the fire risk warning level and fire risk change trend output by the neural network model.

[0109] Specifically, in this embodiment, a neural network model (a lightweight artificial intelligence model or a small neural network) can be directly deployed in the edge computing gateway. This means that the core logic of fire risk assessment is no longer composed of a series of pre-written mathematical formulas, but is carried by a trained neural network that can autonomously learn complex patterns in the data.

[0110] The edge computing gateway preprocesses the monitoring data output from each monitoring module and then directly inputs it as the input vector into the deployed neural network model. The neural network model then performs forward inference calculations and outputs the final fire risk warning level and fire risk trend.

[0111] The integrated monitoring station provided in this application provides an end-to-end intelligent fire risk assessment solution. When dealing with fire risk assessments under extreme weather conditions or special geographical environments, it demonstrates higher accuracy and robustness, thereby providing more intelligent and powerful decision support for forest and grassland fire prevention work.

[0112] In some embodiments, the neural network model is trained by the cloud server based on monitoring data samples, as well as the actual values ​​of the fire risk warning level and the actual values ​​of the fire risk change trend corresponding to the monitoring data samples; The edge computing gateway updates the neural network model in real time based on the model parameters issued by the cloud server.

[0113] Specifically, the neural network model deployed in the edge computing gateway, both in its initial and subsequent updated versions, is trained by a cloud server using supervised learning on a large number of monitoring data samples, along with the actual values ​​of fire risk warning levels and fire risk trends corresponding to these data samples. Furthermore, the edge computing gateway is capable of updating the locally deployed neural network model in real time based on the model parameters distributed from the cloud server.

[0114] The integrated monitoring station provided in this application can continuously learn from newly collected data and fire cases, automatically adapting to climate change and environmental evolution, ensuring that the accuracy and sophistication of its fire risk assessment remain at a high level. This "cloud-edge" architecture perfectly combines the powerful computing capabilities of the cloud with the low-latency response capabilities of the edge, which is key to realizing a large-scale, high-precision, adaptive intelligent fire risk early warning system.

[0115] In some embodiments, Figure 2 This is the second structural schematic diagram of the integrated monitoring station provided in this application, as shown below. Figure 2 As shown, the system management unit 200 also includes a power supply module 220. The power supply module 220 includes a battery 221 and a power control submodule 222; The battery is used to provide power to the front-end sensing unit and the edge computing gateway. The power control submodule is used to monitor the remaining power of the battery and dynamically adjust the data acquisition frequency of the front-end sensing unit and the result reporting frequency of the edge computing gateway based on the remaining power.

[0116] Specifically, the battery is the core of the monitoring station's energy storage. It is used to store electrical energy and provide a stable and continuous power supply for all power-consuming equipment in the entire monitoring station, including various sensors in the front-end sensing unit and edge computing gateways.

[0117] The power control submodule represents a leap from passive "power supply" to active "energy management." This submodule is typically a circuit board integrating a microcontroller, tightly connected to the battery, power input, and the power-consuming equipment. Its core functions can be divided into two aspects: (1) Remaining power monitoring: The power control submodule continuously and accurately monitors the remaining power of the battery. This is usually done by measuring the battery's terminal voltage and combining it with parameters such as current and temperature to calculate the current state of charge (SOC) of the battery, and expressing it as a percentage.

[0118] (2) Dynamic Frequency Adjustment: This is the most critical intelligent function of this submodule. Based on the monitored remaining power (SOC value), it dynamically adjusts the working mode of the monitoring station to "conserve power" when power is scarce and "increase power" when power is abundant. Specifically, the adjustment targets are the data acquisition frequency of the front-end sensing unit and the result reporting frequency of the edge computing gateway. These two frequencies are one of the main sources of system power consumption; the higher the frequency, the greater the power consumption.

[0119] Figure 3 This is a schematic diagram of the intelligent adaptive working mode provided in this application, such as... Figure 3 As shown, the power control submodule has a built-in low-power operating mode strategy. When the program reads the battery SOC value of the power controller through the RS485 port, it automatically configures the following operating mode based on the read SOC value parameter. (1) When the battery SOC value is ≥70%, the program maintains the default normal working mode of collecting and reporting data once every 1 hour.

[0120] (2) When 40% < battery SOC value < 70%, the program continues to maintain normal working mode.

[0121] (3) When the battery SOC value is ≤40%, the program automatically modifies the collection and reporting frequency of the gateway device, switching from the normal working mode to a low-power working mode that collects and reports once every 4 hours.

[0122] (4) After entering the low power mode, the program will automatically switch from "low power mode" to "normal mode" when the battery SOC value is ≥70%, so as to maintain efficient and timely data collection and reporting.

[0123] The integrated monitoring station provided in this application adopts an adaptive working mode strategy, offering highly flexible working mode configuration to cope with diverse business scenarios and environmental challenges. It can intelligently adjust power consumption according to its own energy status, achieving a dynamic balance between performance and battery life. This effectively solves the pain point of field monitoring equipment failing due to energy depletion, ensuring long-term stable and uninterrupted operation of monitoring and early warning services under extreme weather conditions such as continuous rain. It can remotely and in real-time adjust the SOC threshold and data reporting frequency, enabling the equipment's working strategy to change as needed. This represents a leap from "fixed strategy" to "scenario-driven," fully guaranteeing the system's adaptability and optimal performance in different application environments.

[0124] In some embodiments, Figure 4 This is a schematic diagram of the power module provided in this application, as shown below. Figure 4 As shown, the power supply module 220 also includes a mains power access submodule 223 and a solar panel 224.

[0125] Specifically, the power supply module can be configured with a "solar panel-battery-intelligent controller" off-grid power supply system consisting of 2×200W solar panels, 2×150Ah batteries, and a 20A charge / discharge energy controller, providing the primary power source for the monitoring station. Through precise power consumption calculations and in conjunction with an adaptive operating mode, the system can ensure normal operation even under 30 consecutive cloudy or rainy days.

[0126] The power supply module also includes a mains power access submodule, which is compatible with mains power access, and is equipped with a voltage regulator module to provide a second power source for the monitoring station.

[0127] The integrated monitoring station provided in this application uses clean and universal solar energy as its main energy source, ensuring its applicability in vast areas without electricity. At the same time, it is compatible with grid power access, providing a higher level of energy security for scenarios where grid power can be used locally. This enables the monitoring station to cope with various complex deployment environments and energy conditions, greatly expanding its application scope and further improving the long-term operational reliability of the entire system.

[0128] In some embodiments, the integrated monitoring station also includes an equipment column and a protective box; Each monitoring module in the front-end sensing unit is installed from bottom to top along the equipment column; The system management unit is housed in a protective enclosure.

[0129] Specifically, to adapt to outdoor environments, this embodiment of the application includes an equipment support column and a protective enclosure. The equipment support column provides mechanical support for the front-end sensing unit and the system management unit.

[0130] Each monitoring module is installed along the equipment column from bottom to top.

[0131] The soil environment monitoring module is buried at different depths around the column; the surface litter monitoring module and the vegetation phenology monitoring module are integrated in the middle of the column; the atmospheric environment monitoring module is set at the top of the column.

[0132] The protective enclosure is used to provide physical protection for the system management unit.

[0133] In some embodiments, a lightning rod is installed on the top of the equipment column; a metal grounding grid is laid at the bottom of the equipment column; and surge protectors are installed on the power lines and signal lines connecting the various modules in the front-end sensing unit and the system management unit.

[0134] Specifically, in addition to providing physical and structural protection using equipment pillars and protective enclosures, the integrated monitoring station also provides: (1) Direct lightning protection: A lightning rod is installed on the top of the column, with a protection angle ≤ 45°.

[0135] (2) Grounding system design: A metal grounding grid with a depth of ≥2 meters is laid at the bottom of the equipment column, and a resistance reducing agent is used. The grounding resistance is ≤4Ω (ohms).

[0136] (3) Surge protection: AC power surge protectors, DC power surge protectors, signal surge protectors, and power and video combined surge protectors are installed at all interfaces of power lines (AC / DC), signal lines, video lines, etc., forming a multi-level protection.

[0137] Through the above-mentioned multi-level protection design, the integrated monitoring station is able to operate stably for a long time in harsh environments such as no mains power and weak signal.

[0138] In some embodiments, the integrated monitoring station also provides data security protection.

[0139] The edge computing gateway communicates with the outside world using elliptic curve public key cryptography (SM2) for encryption and authentication. The device firmware has a protection mechanism to prevent data tampering and malicious flashing.

[0140] In some embodiments, the edge computing gateway periodically checks the system status (including sensor connection status, battery level, network signal, storage space, etc.) and supports remote parameter configuration, over-the-air (OTA) upgrades, restarts, log viewing, and red, green and blue (RGB) monitoring.

[0141] The following is a specific design scheme for an integrated monitoring station.

[0142] Figure 5 This is a schematic diagram of the hardware composition and three-dimensional structure of the integrated monitoring station provided in this application, as shown below. Figure 5 As shown, this integrated monitoring station can be used for monitoring and early warning of forest and grassland fire risks.

[0143] Its hardware components include: 1. Iron lightning rod; 2. Insulator terminal block; 3. Multi-element meteorological sensor; 4. Multispectral ambient light sensor; 5. Two 200W monocrystalline silicon solar panels; 6. Carbon steel column 3 meters high and 140mm in diameter (the monitoring station is composed of two 3-meter columns spliced ​​together); 7. Multispectral combustible phenological sensor; 8. Intelligent management and protection box; 9. Soil moisture sensor; 10. Ten-meter lightning protection metal grounding grid.

[0144] The intelligent management and protection box includes a system management unit and a power supply system. For areas where wireless communication is not possible, a BeiDou fusion gateway can be optionally installed. The system management unit includes solar panels, batteries, lightning protection modules, a power supply module, and an IoT management module, primarily providing sustainable energy for the integrated monitoring station and managing the equipment.

[0145] Figure 6 This is a schematic diagram of the equipment architecture of the integrated monitoring station provided in this application, such as... Figure 6 As shown, multi-element meteorological sensors, multi-spectral ambient light sensors, multi-spectral combustible phenology sensors, and soil moisture sensors constitute the front-end sensing equipment, realizing three-dimensional and multi-dimensional sensing of fire risk factors such as air temperature, air humidity, ambient light intensity, wind speed, wind direction, rainfall, soil moisture content, surface litter moisture content, and phenology.

[0146] The system management unit mainly includes: a converged computing gateway, a communication management unit, the equipment's own anti-theft alarm system, power supply system, lightning protection and grounding system, etc. It mainly realizes the collection and intelligent identification of various monitoring data at the front end of the monitoring system of this comprehensive monitoring station. The built-in intelligent monitoring and management algorithm analyzes and processes various monitoring data and equipment operating status parameters, and intelligently controls the equipment. At the same time, it reports the front-end algorithm processing results to the command center and enables staff to conduct remote centralized monitoring at the command center on a daily basis. It realizes the functions of intelligent monitoring data collection, analysis, automatic alarm, command and dispatch through wireless digital transmission network.

[0147] The installation accessories mainly include intelligent management protection boxes, columns, brackets, fences, etc.; the installation accessories can help fix the sensors to obtain accurate data, as well as protect the equipment and prevent damage from environmental or human influence.

[0148] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0149] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. An integrated monitoring station based on multi-dimensional fire hazard factor perception, characterized in that, Includes a front-end sensing unit and a system management unit; The front-end sensing unit includes at least two of the following: a soil environment monitoring module, a surface litter monitoring module, a vegetation phenology monitoring module, and an atmospheric environment monitoring module. The soil environment monitoring module is used to monitor the soil moisture content of the soil moisture profile; The surface litter monitoring module is used to monitor the moisture content of surface litter. The vegetation phenology monitoring module is used to monitor the phenological and moisture data of living vegetation; The atmospheric environment monitoring module is used to monitor meteorological data of the atmospheric environment; The system management unit includes an edge computing gateway; the edge computing gateway is connected to each monitoring module in the front-end sensing unit, and is used to perform fusion evaluation based on the monitoring data output by each monitoring module to generate fire risk warning level and fire risk change trend.

2. The integrated monitoring station based on multi-dimensional fire hazard factor perception as described in claim 1, characterized in that, The edge computing gateway includes a data preprocessing module, a fusion evaluation module, a weight adjustment module, and a trend correction module; The data preprocessing module is used to preprocess various monitoring data; the preprocessing includes data standardization; the monitoring data includes at least one of the meteorological data, soil moisture content, surface litter moisture content, and phenological moisture data. The fusion evaluation module is used to perform fusion calculation on the preprocessed monitoring data based on the calculated weights of each monitoring data to obtain a comprehensive fire risk index. The weight adjustment module is used to match a preset weight configuration file based on at least one of the phenological moisture data, the current month and historical fire data, and to determine the calculated weight of each monitoring data and the calculated weight of each influencing factor in the meteorological data based on the matching results. The trend correction module is used to smooth the comprehensive fire risk index at the current moment and the comprehensive fire risk index at historical moments and identify trends to obtain the smoothed value of the comprehensive fire risk index and the rate of change of the comprehensive fire risk index. The smoothed value of the comprehensive fire risk index is used to determine the fire risk warning level; The rate of change of the comprehensive fire risk index is used to determine the trend of fire risk change.

3. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to claim 2, characterized in that, The fusion evaluation module includes a linear weighted calculation submodule and a nonlinear correction submodule; The linear weighted calculation submodule is used to perform fusion calculation on the preprocessed monitoring data based on the calculation weight of each monitoring data to obtain the basic fire risk index. The nonlinear correction submodule is used to determine the coordination effect coefficient based on at least two meteorological factors, including air temperature and wind speed, and the comprehensive aridity determined by at least one of the soil moisture content, the surface litter moisture content, and the phenological moisture data. The comprehensive fire risk index is obtained by nonlinearly correcting the basic fire risk index based on the coordination effect coefficient.

4. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to claim 2, characterized in that, The trend correction module includes a smoothing submodule and a trend recognition submodule; The smoothing submodule is used to perform exponential moving average smoothing on the current comprehensive fire risk index and the comprehensive fire risk index at historical times to obtain the smoothed value of the comprehensive fire risk index. The trend recognition submodule is used to perform linear fitting based on the comprehensive fire risk index at the current moment and the comprehensive fire risk index at historical moments, and obtain the rate of change of the comprehensive fire risk index based on the fitting result.

5. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to claim 1, characterized in that, The edge computing gateway is equipped with a neural network model. The edge computing gateway is used to input the monitoring data output by each monitoring module into the neural network model to obtain the fire risk warning level and fire risk change trend output by the neural network model.

6. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to claim 5, characterized in that, The neural network model was trained by the cloud server based on monitoring data samples, as well as the actual values ​​of the fire risk warning level and the actual values ​​of the fire risk change trend corresponding to the monitoring data samples. The edge computing gateway updates the neural network model in real time based on the model parameters issued by the cloud server.

7. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to any one of claims 1 to 6, characterized in that, The system management unit also includes a power supply module; the power supply module includes a battery and a power control submodule. The battery is used to provide power to the front-end sensing unit and the edge computing gateway. The power control submodule is used to monitor the remaining power of the battery and dynamically adjust the data acquisition frequency of the front-end sensing unit and the result reporting frequency of the edge computing gateway based on the remaining power.

8. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to claim 7, characterized in that, The power supply module also includes a mains power access submodule and a solar panel; The solar panel is used to provide a first power source; The mains power access submodule is used to provide a second power source.

9. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to any one of claims 1 to 6, characterized in that, It also includes equipment columns and protective boxes; Each monitoring module in the front-end sensing unit is installed from bottom to top along the equipment column; The system management unit is located inside the protective box.

10. The integrated monitoring station based on multi-dimensional fire hazard factor perception according to claim 9, characterized in that, A lightning rod is installed on the top of the equipment column; a metal grounding grid is laid at the bottom of the equipment column; and surge protectors are installed on the power lines and signal lines connecting the various modules in the front-end sensing unit and the system management unit.

Images