Intelligent wood plant security monitoring system based on Internet of Things

Abstract

The invention discloses an intelligent wood plant security monitoring system based on the Internet of Things, and relates to the technical field of signal monitoring, and the system comprises a visual monitoring module, a sensing monitoring module, a signal transmission module, a cloud processor and a security management module. An AI security monitoring model is constructed through a cloud processor, three-dimensional structure information of stacking density and void ratio of a wood pile is analyzed, collapse risk and fire risk of the wood pile are evaluated, a fire source estimation coordinate is further positioned, a security management scheme is generated through integration, automatic monitoring and intelligent decision making are performed, security monitoring early warning is realized, and the safety of the wood pile is improved. The wood plant security monitoring system has the advantages that manual inspection frequency is reduced, plant space layout and resource allocation are optimized, management efficiency is improved, alarm response timeliness is guaranteed, and intelligence, real-time performance and precision of wood plant security monitoring are realized through combination of the internet of things, three-dimensional modeling and an AI algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of signal monitoring, and particularly to an intelligent security monitoring system for a wood factory building based on the Internet of Things. Background Art

[0002] As a typical industrial site, a wood factory has characteristics such as being flammable, prone to collapse, and having a complex environment. There are situations of high dust and high humidity. It is necessary to regularly inspect each area of the wood factory to discover potential safety hazards such as equipment failures, hidden dangers in wood stacking (such as being too high or tilted), and aging of wires, so as to make timely rectifications. The security monitoring of traditional wood factory buildings mainly relies on single sensors or manual inspections. There are limitations such as single monitoring dimensions, lag in emergency response, and insufficient emergency linkage. Since traditional methods mostly use single sensors or cameras, it is difficult to comprehensively obtain the three-dimensional structure information of wood stacks, resulting in insufficient accuracy in assessing the collapse risk and fire risk of wood stacking. Moreover, traditional monitoring systems rely on local servers or manual data processing, unable to efficiently integrate multi-source heterogeneous data, difficult to detect dynamic changes in data in a timely manner, lacking intelligent analysis capabilities, unable to achieve real-time early warning, and lacking intelligent linkage with the factory building's fire protection facilities and personnel evacuation systems, resulting in low emergency response efficiency and being error-prone, and difficult to meet the security monitoring requirements of wood factories. In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0003] The purpose of the present invention is to solve the limitations of single monitoring dimensions, lag in emergency response, and insufficient emergency linkage existing in the security monitoring of traditional wood factory buildings, and through the combination of the Internet of Things, three-dimensional modeling, and AI algorithms, achieve the intelligentization, real-time, and precision of the security monitoring of wood factory buildings.

[0004] To achieve the above purpose, the present invention adopts the following technical solutions: An intelligent security monitoring system for a wood factory building based on the Internet of Things, comprising a visual monitoring module, a sensing monitoring module, a signal transmission module, a cloud processor, and a security management module; The visual monitoring module is used to collect wood stacking images and extract three-dimensional point cloud coordinate data; The sensing monitoring module is used to collect factory building environment data; The signal transmission module is used to deploy a LoRA signal network for data transmission: transmit the three-dimensional point cloud coordinate data and the factory building environment data to the cloud processor; The cloud processor constructs an AI security monitoring model to conduct intelligent monitoring of the wood factory building: evaluate the collapse risk of the wood stack through the three-dimensional point cloud coordinate data, evaluate the fire risk of the wood stack through the factory building environment data, and then through the temperature field image of the wood factory building, locate and obtain the estimated coordinates of the fire source, and integrate and generate a security management plan. The security management module is used to receive the security management plan and perform corresponding processing.

[0005] Furthermore, the specific processes of the visual monitoring module and the sensing monitoring module are as follows: The visual monitoring module collects the images of the wood stack through the lidar device, and extracts the three-dimensional point cloud coordinate data of the wood stack through the visual fusion technology; collects the temperature field image of the wood workshop through the infrared thermal imager; The sensing monitoring module sets a data acquisition period to regularly collect the workshop environment data; the workshop environment data includes the concentration of combustion gas, the moisture content of wood, and the air humidity in the workshop; among them, the concentration of combustion gas is collected through the gas sensor; the moisture content of wood is calculated by monitoring the attenuation of the LoRA signal; the air humidity in the workshop is collected through the humidity sensor.

[0006] Furthermore, the specific process of constructing the AI security monitoring model is as follows: Analyze the density and voids of the wood stack through the three-dimensional point cloud coordinate data, so as to evaluate the collapse risk of the wood stack; Conduct time series analysis through the workshop environment data to evaluate the fire risk of the wood stack; Through the temperature field image of the wood workshop, locate and obtain the estimated coordinates of the fire source; Integrate the collapse risk result, fire risk result and estimated fire source coordinates of the wood stack to generate a security management plan.

[0007] Furthermore, analyze the density and voids of the wood stack through the three-dimensional point cloud coordinate data, so as to evaluate the collapse risk of the wood stack. The specific process is as follows: Perform grayscale processing on the wood stack image, and divide it into N0 stacking areas. Mark any stacking area as i, set the volume of voxel a as Va, and perform voxelization processing on the three-dimensional point cloud data of stacking area i. When the point cloud coordinate P of the wood stack falls into voxel a, then count the total number of voxels containing the three-dimensional point cloud data of the wood stack and mark it as n0, then the approximate volume of the wood stack is Vi; collect the total mass Mi of the wood stack through the weighbridge to obtain the wood stacking density of stacking area i ; Then, count the total number of voxels not occupied by the three-dimensional point cloud data of the wood stack and mark it as n1, then the void volume of the wood stack is Vk, so as to calculate and obtain the wood stacking void ratio of stacking area i ; Through the wood stacking density of stacking area i and the wood stacking void ratio Combined to obtain the collapse risk coefficient RISKc of stacking area i; Set the risk threshold U1 of the collapse risk coefficient RISKc. When the collapse risk coefficient RISKc of the stacking area i exceeds the risk threshold U1, a stacking management signal for the factory building wood is generated.

[0008] Furthermore, perform time series analysis on the factory building environment data to evaluate the fire risk of the wood pile. The specific process is as follows: Mark the combustion gas concentration, wood moisture content, and factory building air humidity as Xqn, Xms, and Xqs respectively; Respectively set the standard intervals of the combustion gas concentration Xqn, wood moisture content Xms, and factory building air humidity Xqs, and obtain the fire risk coefficient RISKf of the stacking area i through interval comparison; Set the risk threshold U2 of the fire risk coefficient RISKf. When the fire risk coefficient RISKf of the stacking area i exceeds the risk threshold U2, a fire warning management signal is generated.

[0009] Furthermore, through the temperature field image of the wood factory building, the estimated coordinates of the fire source are located. The specific process is as follows: Build an infrared thermal imaging temperature field matrix V(x, y, t) through the temperature field image of the wood factory building, where x is the horizontal vector, y is the vertical vector, and t is the time series; Set the median filtering function MFilter to suppress the noise of the infrared thermal imaging temperature field matrix V, and obtain the denoised infrared thermal imaging temperature field matrix ; Then normalize the denoised infrared thermal imaging temperature field matrix to obtain the normalized infrared thermal imaging temperature field matrix ; Obtain the confidence weight factor of the thermal imaging through the temperature gradient ; Furthermore, obtain the fire source probability matrix of the normalized infrared thermal imaging temperature field matrix ; ; Mark the fire source probability corresponding to any point (x, y) of the temperature field image of the wood factory building at the time node t as p(x, y, z); Set the risk threshold Up of the fire source probability p(x, y, z). When the fire source probability p(x, y, z) exceeds the risk threshold Up, mark the point (x, y) as the estimated fire source coordinate point, so as to locate and obtain the estimated fire source coordinates of the temperature field image of the wood factory building.

[0010] Furthermore, measure the wood moisture content by monitoring the attenuation of the LoRA signal. The specific process is as follows: Mark the initial signal strength as RSSI0. After transmitting a signal to the wood pile and receiving feedback, mark the attenuated signal strength RSSI(d) corresponding to the path distance d, and fit the data before and after signal attenuation as well as the path distance d to obtain a correlation function. Through the attenuation coefficient of the wood medium Fit with the wood moisture content Wi, obtain the scatter plot of the two and generate the correlation function of the fitted curve, so as to analyze and measure the wood moisture content Wi through the attenuation coefficient of the wood medium

[0011] Furthermore, the security management plan includes the stacking management of the wood in the factory building, fire warning management, and fire source location and extinguishing operations. Among them, corresponding security management is carried out by receiving the stacking management signal of the wood in the factory building, the fire warning management signal, and the estimated coordinates of the fire source in the temperature field image of the wood factory building.

[0012] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are as follows: The present invention collects the wood stacking image through the visual monitoring module and extracts the three-dimensional point cloud coordinate data, and collects the factory building environment data through the sensing monitoring module, realizing multi-dimensional data fusion and high-precision monitoring. The cloud processor constructs an AI security monitoring model to analyze the three-dimensional structure information of the stacking density and porosity of the wood pile, evaluate the collapse risk and fire risk of the wood pile, and then locate the estimated coordinates of the fire source, and integrate and generate a security management plan for automated monitoring and intelligent decision-making, realizing early warning of security monitoring, reducing the frequency of manual inspections, optimizing the factory building space layout and resource allocation, improving the management efficiency, ensuring the timeliness of alarm response, and realizing the intelligence, real-time and precision of the security monitoring of the wood factory building through the combination of the Internet of Things, three-dimensional modeling and AI algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 Shows the connection schematic diagram of the system modules of the present invention; Figure 2 Shows the step schematic diagram of the working process of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0015] Embodiment 1: As Figure 1-2As shown in the figure, an intelligent security monitoring system for a wood factory based on the Internet of Things includes a visual monitoring module, a sensing monitoring module, a signal transmission module, a cloud processor, and a security management module. Among them, the visual monitoring module, the sensing monitoring module, the signal transmission module, the cloud processor, and the security management module are communicatively connected; With the popularization of Internet of Things technology and the maturity of low-power wide-area network (LPWAN) technology (such as LoRa, NB-IoT), the deployment cost of sensor networks in wood factories has been significantly reduced, supporting the real-time collection and transmission of large-scale environmental data (temperature, humidity, gas concentration); The working steps are as follows: S1, the visual monitoring module collects the wood stacking image and extracts the three-dimensional point cloud coordinate data; The specific process of the visual monitoring module is: collecting the wood stacking image through a lidar device, and extracting the three-dimensional point cloud coordinate data of the wood stacking through visual fusion technology; Collecting the temperature field image of the wood factory through an infrared thermal imager; S2, the sensing monitoring module collects the factory environment data; The specific process of the sensing monitoring module is: the factory environment data includes the concentration of combustion gas, the moisture content of wood, and the air humidity in the factory; setting a data collection period to regularly collect the factory environment data; Collecting the concentration of combustion gas through a gas sensor (CO2 sensor); calculating the moisture content of wood by monitoring the attenuation of LoRA signals; collecting the air humidity in the factory through a humidity sensor; The specific process of calculating the moisture content of wood by monitoring the attenuation of LoRA signals is as follows: The LoRA (Low-Rank Adaptation) model is a plugin for fine-tuning large language models, which can be trained with a small amount of data to adjust the output results; Mark the initial signal strength as RSSI0, send a signal to the wood pile and receive the feedback, mark the attenuated signal strength RSSI(d) corresponding to the path distance d, and fit the data before and after signal attenuation and the path distance d to obtain the correlation function: ; Among them, n is the path loss exponent, which is obtained through the slope of the fitted line, is the attenuation coefficient of the wood medium, is the error coefficient of other factors; Through the attenuation coefficient of the wood medium and the moisture content of wood Wi are fitted, the scatter plot of the two is obtained and the correlation function of the fitted curve is generated, so as to analyze and calculate the moisture content of wood Wi through the attenuation coefficient of the wood medium : ; Among them, e is the natural constant and e takes 2.7, β0 is the conversion coefficient obtained through experimental fitting, and β1 is the intercept of the fitting curve; S3. The signal transmission module deploys a LoRA signal network for data transmission: transmitting the three-dimensional point cloud coordinate data and the factory building environment data to the cloud processor; S4. The cloud processor constructs an AI security monitoring model to intelligently monitor the wood factory building: evaluating the collapse risk of the wood pile through the three-dimensional point cloud coordinate data, evaluating the fire risk of the wood pile through the factory building environment data, and then through the temperature field image of the wood factory building, thereby locating and obtaining the estimated coordinates of the fire source, and integrating and generating a security management plan; The specific process of constructing the AI security monitoring model is as follows: S4-1. Analyze the density and voids of the wood stacking through the three-dimensional point cloud coordinate data, so as to evaluate the collapse risk of the wood pile; Analyze the density and voids of the wood stacking through the three-dimensional point cloud coordinate data, so as to evaluate the collapse risk of the wood pile; Perform gray processing on the wood stacking image and divide it into N0 stacking areas. Mark any stacking area as i. Set the volume of voxel a as Va. Perform voxelization processing on the three-dimensional point cloud data of stacking area i. When the point cloud coordinate P of the wood pile falls into voxel a, then count the total number of voxels containing the three-dimensional point cloud data of the wood pile and mark it as n0. Then the approximate volume of the wood pile is Vi: ; Collect the total mass Mi of the wood pile through a weighbridge, so as to obtain the wood stacking density of stacking area i : ; Then count the total number of voxels not occupied by the three-dimensional point cloud data of the wood pile and mark it as n1. Then the void volume of the wood pile is Vk: , thereby calculating and obtaining the wood stacking void ratio of stacking area i : ; Through the wood stacking density of stacking area i and the wood stacking void ratio are combined to obtain the collapse risk coefficient RISKc of stacking area i: ; Among them, Tc is the detection time length of the wood stacking image, t is the time node corresponding to any video frame, is the wood stacking density at the initial time of the wood stacking image, and α1 and α2 are the wood stacking density and the wood stacking void ratio weight factor coefficients, and the weight factor coefficients are preset after being calculated through a large amount of experimental data, and are specifically set in combination with the actual situation; When the wood stacking density and the void ratio of the wood stacking When there is a time series change, it indicates that there is a risk of collapse in the stacking area i; Set the risk threshold U1 of the collapse risk coefficient RISKc. When the collapse risk coefficient RISKc of the stacking area i exceeds the risk threshold U1, a stacking management signal for the factory building wood is generated, thereby triggering the wood stacking alarm mechanism; S4-2, conduct time series analysis through the factory building environment data to evaluate the fire risk of the wood pile; Conduct time series analysis through the factory building environment data to evaluate the fire risk of the wood pile; Mark the combustion gas concentration, the moisture content of the wood, and the air humidity in the factory building as Xqn, Xms, and Xqs respectively; The combustion gas concentration refers to the concentration of gases generated by combustion, such as CO2 concentration; Respectively set the standard intervals of the combustion gas concentration Xqn, the moisture content of the wood Xms, and the air humidity in the factory building Xqs, and obtain the fire risk coefficient RISKf of the stacking area i through interval comparison; ; Among them, g1, g2, and g3 are the interval comparison formulas for the combustion gas concentration Xqn, the moisture content of the wood Xms, and the air humidity in the factory building Xqs respectively; μ1, μ2, and μ3 are the weight factors of the combustion gas concentration Xqn, the moisture content of the wood Xms, and the air humidity in the factory building Xqs respectively, and μ1, μ2, and μ3 are all greater than 0; Regard the combustion gas concentration Xqn, the moisture content of the wood Xms, and the air humidity in the factory building Xqs as the parameters of the factory building environment data; mark any parameter of the factory building environment data as r, mark the parameter value of the parameter r as Xr, and mark the standard interval of the parameter r as [Qr1, Qr2]; Then the interval comparison formula for the parameter r is gr: ; When the combustion gas concentration Xqn, the moisture content of the wood Xms, and the air humidity in the factory building Xqs are respectively within the corresponding standard intervals, it is determined that the combustion gas concentration Xqn, the moisture content of the wood Xms, and the air humidity in the factory building Xqs are normal parameters; otherwise, it is determined that the parameter is abnormal; Set the risk threshold U2 of the fire risk coefficient RISKf. When the fire risk coefficient RISKf of the stacking area i exceeds the risk threshold U2, a fire warning management signal is generated, thereby triggering the wood fire warning mechanism; S4-3, through the temperature field image of the wood factory building, locate and obtain the estimated coordinates of the fire source; Build an infrared thermal imaging temperature field matrix V(x, y, t) through the temperature field image of the wood factory building, where x is the horizontal vector, y is the vertical vector, and t is the time series; Set the median filtering function MFilter to suppress noise in the infrared thermal imaging temperature field matrix V, and obtain the denoised infrared thermal imaging temperature field matrix : ; Then, perform normalization on the denoised infrared thermal imaging temperature field matrix to obtain the normalized infrared thermal imaging temperature field matrix : ; wherein, is the average value of the temperature, is the standard deviation of the temperature; Obtain the confidence weight factor of the thermal imaging through the temperature gradient : ; wherein, γ is the temperature sensitivity coefficient, and the temperature sensitivity coefficient is obtained through presetting. For example, calibrate γ = 50 °C, which is significantly higher than the daily room temperature; Thus, obtain the fire source probability matrix of the normalized infrared thermal imaging temperature field matrix : : ; Mark the fire source probability corresponding to any point (x, y) of the temperature field image of the wood factory building at the time node t as p(x, y, z); Set the risk threshold Up of the fire source probability p(x, y, z). When the fire source probability p(x, y, z) exceeds the risk threshold Up, mark the point (x, y) as the fire source estimation coordinate point, so as to locate and obtain the fire source estimation coordinates of the temperature field image of the wood factory building; S4-4, Integrate and generate a security management plan through the collapse risk result, fire risk result and fire source estimation coordinates of the wood pile; S5, The security management module receives the security management plan and performs corresponding processing; The security management plan includes the stacking management of the wood in the factory building, fire warning management and fire source positioning and extinguishing operations; The stacking management of the wood in the factory building is to manage the wood stacking areas with collapse risks, including planning the layout, standardizing the stacking, and limiting the load and quantity; Among them, planning and layout means dividing a special wood stacking area according to the plant area, structure and fire protection requirements, and separating it from other operating areas, such as setting up fire walls, fireproof roller shutters, etc. for fire separation; standardized stacking means controlling the height and spacing of wood piles, and stacking them according to wood types and specifications to avoid mixing, especially wood that is prone to spontaneous combustion or decomposition when exposed to water, which needs to be stored in a place with low temperature, good ventilation and dry air; load and quantity limits refer to determining the load capacity of the stacking area and the amount of wood stored, preventing overweight and overquantity, and avoiding compression damage and decreased stability of the bottom wood; Fire warning management is to provide early warning for wood storage areas with fire risks, including establishing a graded warning mechanism and setting up emergency plans; Among them, according to the degree of danger, different levels are divided, such as low, medium and high, and different response measures are taken accordingly. At the same time, the early warning system is linked with fire-fighting facilities, personnel evacuation indication systems, etc., and the corresponding actions are automatically executed when the early warning is triggered. For example, the sprinkler system is linked to prevent spontaneous combustion; The fire source location and fire extinguishing operation is to activate the fire fighting system in the fire source location coordinate area and contact professional rescue; Among them, starting the fire-fighting system means that when the fire is large, the automatic alarm will prompt professionals to start the wood factory's own fire-fighting system, such as the sprinkler system, fire hydrant system, and fire extinguisher system to spray water to extinguish the fire. At the same time, they will immediately call the police to contact professional rescue, cooperate with fire rescue personnel to extinguish the fire, and provide information such as factory layout and stored materials.

[0016] In summary, the present invention collects wood stacking images and extracts three-dimensional point cloud coordinate data through the visual monitoring module, and collects plant environment data through the sensor monitoring module, thereby realizing multi-dimensional data fusion and high-precision monitoring, and performing automatic monitoring and intelligent decision-making through the cloud processor, thereby realizing early warning of security monitoring, reducing the frequency of manual inspections, optimizing plant space layout and resource allocation, and improving management efficiency. Through the combination of the Internet of Things, three-dimensional modeling and AI algorithms, the intelligent, real-time and precise security monitoring of wood plants is realized; The visual monitoring module uses laser radar and infrared thermal imager to obtain the three-dimensional point cloud coordinates and temperature field images of the wood pile in real time, and combines voxel analysis to accurately calculate the density, void ratio and collapse risk of the wood pile; the sensor monitoring module uses LoRA signal attenuation to calculate the moisture content of the wood, and combines gas and humidity sensor data to dynamically evaluate the fire risk and cover comprehensive monitoring of environmental parameters; The signal transmission module deploys the LoRA signal network for data transmission: the three-dimensional point cloud coordinate data and plant environment data are transmitted to the cloud processor. The LoRa low-power wide area network reduces the data transmission cost. At the same time, the moisture content of the wood is inverted through signal attenuation, reducing the use of expensive sensors such as microwave moisture meters. The cloud processor constructs an AI security monitoring model to intelligently monitor the wood workshop, evaluate the collapse risk and fire risk of the wood piles, then locate the estimated coordinates of the fire source, and integrate and generate a security management plan to reduce the incidence of wood pile collapse and fire accidents, reduce property losses and casualties, and ensure safe production; The security management module receives the security management plan and conducts corresponding processing, integrates the collapse, fire risk and fire source coordinates, generates a comprehensive plan including standardized stacking, hierarchical early warning, and fire extinguishing operations, realizes the closed-loop management of "monitoring - analysis - response", and links the fire protection system with the emergency plan, greatly shortening the response time, making up for the deficiencies of traditional methods in data dimension, analysis ability and emergency response, and providing an innovative solution for the safety management of the wood processing industry.

[0017] The setting of the size of the interval and threshold is for the convenience of comparison. Regarding the size of the threshold, it depends on the amount of sample data and the number of base numbers set by those skilled in the art for each group of sample data; as long as it does not affect the proportional relationship between the parameters and the quantified values.

[0018] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulations to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation; As described above, only the specific preferred embodiments of the present invention are provided, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent replacements or changes, and should be covered by the protection scope of the present invention.

Claims

1. An intelligent security monitoring system for a wood workshop based on the Internet of Things, characterized in that: It includes a visual monitoring module, a sensing monitoring module, a signal transmission module, a cloud processor, and a security management module; The visual monitoring module is used to collect wood stacking images and extract three-dimensional point cloud coordinate data; The sensing monitoring module is used to collect plant environment data; The signal transmission module is used to deploy a LoRA signal network for data transmission: transmitting the three-dimensional point cloud coordinate data and the plant environment data to the cloud processor; The cloud processor constructs an AI security monitoring model to intelligently monitor the wood plant: evaluating the collapse risk of the wood pile through the three-dimensional point cloud coordinate data, evaluating the fire risk of the wood pile through the plant environment data, and then through the temperature field image of the wood plant to locate and obtain the estimated coordinates of the fire source, and integrating to generate a security management plan; The security management module is used to receive the security management plan and perform corresponding processing.

2. The intelligent security monitoring system for a wood factory based on the Internet of Things according to claim 1, wherein: The specific processes of the visual monitoring module and the sensing monitoring module are as follows: The visual monitoring module collects wood stacking images through lidar equipment and extracts the three-dimensional point cloud coordinate data of the wood stacking through visual fusion technology; collects the temperature field image of the wood plant through an infrared thermal imager; The sensing monitoring module sets a data collection period to regularly collect plant environment data; the plant environment data includes the concentration of combustible gases, the moisture content of wood, and the air humidity in the plant; among them, the concentration of combustible gases is collected through a gas sensor; the moisture content of wood is measured by monitoring the attenuation of LoRA signals; the air humidity in the plant is collected through a humidity sensor.

3. The intelligent security monitoring system for a wood factory based on the Internet of Things according to claim 2, wherein: The specific process of constructing the AI security monitoring model is as follows: Analyze the density and voids of the wood stacking through the three-dimensional point cloud coordinate data to evaluate the collapse risk of the wood pile; Conduct time series analysis through the plant environment data to evaluate the fire risk of the wood pile; Locate and obtain the estimated coordinates of the fire source through the temperature field image of the wood plant; Integrate the collapse risk result, fire risk result, and estimated fire source coordinates of the wood pile to generate a security management plan.

4. The intelligent security monitoring system for a wood factory based on the Internet of Things according to claim 3, wherein: Analyze the density and voids of the wood stacking through the three-dimensional point cloud coordinate data to evaluate the collapse risk of the wood pile. The specific process is as follows: Perform grayscale processing on the wood stacking image, and divide it into N0 stacking areas. Mark any stacking area as i, set the volume of voxel a as Va, and perform voxelization on the three-dimensional point cloud data of stacking area i. When the point cloud coordinate P of the wood pile falls into voxel a, count and mark as n0 the total number of voxels containing the three-dimensional point cloud data of the wood pile, then the approximate volume of the wood pile is Vi; collect the total mass Mi of the wood pile through a weighbridge to obtain the wood stacking density of stacking area i ; Then, by counting all the voxels not occupied by the three-dimensional point cloud data of the wood pile and marking it as n1, the void volume of the wood pile is Vk, and thus the void ratio of the wood stacking in the stacking area i is calculated and obtained. ; The wood stacking density in stacking area i and the wood stacking void ratio are combined to obtain the collapse risk coefficient RISKc of stacking area i; Set the risk threshold U1 of the collapse risk coefficient RISKc. When the collapse risk coefficient RISKc of the stacking area i exceeds the risk threshold U1, a stacking management signal for the plant wood is generated.

5. The intelligent security monitoring system for wood factories based on the Internet of Things according to claim 4, wherein: Conduct time series analysis through the plant environment data to evaluate the fire risk of the wood pile. The specific process is as follows: Mark the concentration of combustible gases, the moisture content of wood, and the air humidity in the plant as Xqn, Xms, and Xqs respectively; Respectively set the standard intervals of the concentration of combustible gases Xqn, the moisture content of wood Xms, and the air humidity Xqs in the plant, and obtain the fire risk coefficient RISKf of the stacking area i through interval comparison; Set the risk threshold U2 of the fire risk coefficient RISKf. When the fire risk coefficient RISKf of the stacking area i exceeds the risk threshold U2, a fire warning management signal is generated.

6. The intelligent wood factory security monitoring system based on the Internet of Things according to claim 5, characterized in that: Locate and obtain the estimated coordinates of the fire source through the temperature field image of the wood plant. The specific process is as follows: Build an infrared thermal imaging temperature field matrix V(x, y, t) through the temperature field image of the wood plant, where x is the horizontal vector, y is the vertical vector, and t is the time series; Set the median filtering function MFilter to suppress the noise of the infrared thermal imaging temperature field matrix V, and obtain the denoised infrared thermal imaging temperature field matrix ; Then, normalize the denoised infrared thermal imaging temperature field matrix to obtain a normalized infrared thermal imaging temperature field matrix ; Obtaining a confidence weight factor for thermal imaging through a temperature gradient ; Furthermore, a standardized infrared thermal imaging temperature field matrix is obtained of the fire source probability matrix ; Mark the fire source probability corresponding to any point (x, y) of the temperature field image of the wood workshop at time node t as p(x, y, z). Set the risk threshold Up of the fire source probability p(x, y, z). When the fire source probability p(x, y, z) exceeds the risk threshold Up, mark the point (x, y) as the fire source estimation coordinate point, so as to locate and obtain the fire source estimation coordinates of the temperature field image of the wood workshop.

7. An intelligent security monitoring system for a timber factory based on the Internet of Things according to claim 6, characterized in that: Measure the moisture content of wood by monitoring the attenuation of LoRA signals. The specific process is as follows: Mark the initial signal strength as RSSI0. After sending a signal to the wood pile and receiving the feedback, mark the attenuated signal strength RSSI(d) corresponding to the path distance d. Fit the data before and after signal attenuation and the path distance d to obtain the correlation function. Attenuation coefficient of wood medium is fitted with the wood moisture content Wi, a scatter plot of the two is obtained, and a correlation function of the fitted curve is generated, so that through the attenuation coefficient of the wood medium , the wood moisture content Wi is analyzed and measured.

8. The intelligent security monitoring system for a wood factory based on the Internet of Things according to claim 7, characterized in that: The security management plan includes the stacking management of the wood in the workshop, fire warning management, and fire source location and extinguishing operations. Among them, corresponding security management is carried out by receiving the stacking management signal of the wood in the workshop, the fire warning management signal, and the fire source estimation coordinates of the temperature field image of the wood workshop.

Images