An intelligent safety and fire linkage monitoring system and early warning method based on internet of things and 5G-Lora

Abstract

This invention discloses a smart security and fire protection linkage monitoring system and early warning method based on the Internet of Things (IoT) and 5G-LoRa, belonging to the field of security and fire protection technology. It includes a terminal sensing layer, a network transmission layer, a cloud platform layer, and a terminal execution layer. The terminal sensing layer deploys various security and fire protection monitoring devices to collect multi-dimensional security and fire protection data. The network transmission layer adopts a 5G and LoRa dual-mode communication architecture to achieve low-power, high-speed transmission of multi-dimensional security and fire protection data, as well as local preprocessing and cloud synchronization. The cloud platform layer integrates a multi-modal data fusion module, an adaptive threshold early warning module, a dynamic load balancing module, a hierarchical linkage response module, and a data backtracking and model optimization module. The terminal execution layer deploys emergency linkage devices to receive linkage commands issued by the system, enabling coordinated response from fire protection and security equipment. This invention, using the above method, effectively reduces false alarm and missed alarm rates, and improves the comprehensiveness and accuracy of hazard identification.

Description

Technical Field

[0001] This invention relates to the technical field of security and fire protection, and in particular to a smart security and fire protection linkage monitoring system and early warning method based on the Internet of Things and 5G-Lora. Background Technology

[0002] With the advancement of digital and intelligent transformation in the security and fire protection field, traditional security and fire monitoring models are no longer sufficient to meet the security and control needs of complex scenarios. Currently, the industry generally faces technical bottlenecks such as insufficient monitoring accuracy, unstable communication transmission, incomplete hazard identification, and inflexible joint response. Specifically, existing security and fire monitoring relies heavily on single-modal sensors (such as smoke detectors and heat detectors monitoring separately), lacking multi-dimensional data fusion analysis. This makes it difficult to identify hidden hazards (where a single modality may not exceed the standard, but multi-modal coordination is abnormal), leading to false alarms and missed alarms, and failing to achieve the transformation from passive response to proactive prevention. Furthermore, the warning thresholds of traditional monitoring systems are mostly fixed values, not dynamically adjusted according to scenario type and environmental changes, resulting in poor adaptability and difficulty in meeting the differentiated warning needs of scenarios with different risk levels.

[0003] In terms of communication transmission, a single communication link has obvious defects: 5G communication has high speed and low latency, but high power consumption, and is prone to overload in complex electromagnetic environments, leading to unstable transmission; LoRa communication has low power consumption and wide coverage, but slow speed, which cannot meet the high-speed transmission requirements of massive security and fire data. The two have not achieved effective coordinated scheduling, making it difficult to balance transmission efficiency and stability.

[0004] In addition, the existing system lacks a sound data backtracking and model optimization mechanism. Data after the handling of potential hazards is not fully archived and traceable, making it impossible to optimize the early warning algorithm by reviewing historical data. At the same time, the parameters of the early warning model are fixed and cannot be dynamically updated according to changes in the scenario, equipment aging, and other factors, making it difficult to continuously improve the accuracy of the early warning and adapt to long-term operation needs. Summary of the Invention

[0005] The purpose of this invention is to provide a smart security and fire protection linkage monitoring system and early warning method based on the Internet of Things and 5G-Lora, which solves the problems of single-modal monitoring lacking multi-dimensional fusion analysis, resulting in difficulty in identifying hidden dangers and high false alarm and missed alarm rates.

[0006] To achieve the above objectives, the present invention provides a smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa, comprising a terminal sensing layer, a network transmission layer, a cloud platform layer, and a terminal execution layer; The terminal sensing layer is equipped with a variety of safety and fire monitoring devices to collect multi-dimensional safety and fire data, including smoke concentration, ambient temperature, and combustible gas concentration. The network transport layer adopts a 5G and LoRa dual-mode communication architecture, including a LoRa gateway, a 5G router, and an edge computing gateway, to achieve low-power, high-speed transmission of multi-dimensional security and fire data, as well as local preprocessing and cloud synchronization. The cloud platform layer is used for data aggregation, storage, query, statistical analysis, real-time monitoring of equipment status, and historical data backtracking. The cloud platform layer integrates a multimodal data fusion module, an adaptive threshold early warning module, a dynamic load balancing module, a hierarchical linkage handling module, and a data backtracking and model optimization module. The multimodal data fusion module is used to achieve deep fusion of multi-dimensional security and fire data; the adaptive threshold early warning module is used to achieve dynamic early warning threshold generation and hidden danger identification; the dynamic load balancing module is used to achieve dynamic scheduling of network link load; and the hierarchical linkage response module is used to achieve hierarchical linkage closed-loop response to security and fire incidents. Emergency response equipment is deployed at the terminal execution layer to receive linkage commands issued by the system and enable coordinated response from fire protection and security equipment.

[0007] Preferably, the deep fusion of multi-dimensional security and fire protection data adopts the following method: a weighted fusion algorithm is used to perform deep fusion of multi-dimensional security and fire protection data, and the fusion formula is: ; in, This is the comprehensive monitoring value after multimodal data fusion. The number of modal data types participating in the fusion. For the first Weighting coefficients for modal data The value range of is [0,1], and satisfies , For the first The standardized values ​​of the modal data are unitless and range from [0, 10]. The standardization formula is: ; in, For the first The raw acquired values ​​of the modal data, For the first The minimum reasonable value of modal data, For the first The maximum reasonable value of the modal data.

[0008] Preferably, the dynamic warning threshold is generated based on a clustering analysis algorithm. Machine learning is used to train the algorithm on historical normal data from different scenarios to dynamically generate a warning threshold adapted to that scenario. The calculation formula is as follows: ; in, For dynamic early warning thresholds, This is the average of the comprehensive monitoring values ​​after integrating historical normal data. The standard deviation of the comprehensive monitoring value after integrating historical normal data. This is the threshold adjustment coefficient, with a value range of [1.5, 2.5].

[0009] Preferably, the identification of hidden hazards adopts the following method: using correlation coefficients to determine the degree of correlation between different modalities of data to assist in the identification of hidden hazards. The correlation coefficient formula is as follows: ; in, The correlation coefficient between the two modalities is... The covariance of the two modal data The standard deviation of the first modality data. The standard deviation of the second modality data; when And neither modal data exceeded the standard, but At that time, it was determined to be a hidden danger.

[0010] Preferably, dynamic load balancing in 5G-Lora heterogeneous networks includes the following: real-time detection of the load status of 5G links and LoRa gateways, calculation of the load using a load rate formula, and dynamic allocation of data transmission links based on scheduling conditions. The 5G link load rate calculation formula is as follows: ; in, This refers to the real-time load rate of the 5G link. This represents the current bandwidth usage of the 5G link. This refers to the total bandwidth of the 5G link; The formula for calculating the load rate of a Lora gateway is: ; in, This refers to the real-time load rate of the LoRa gateway. This represents the number of terminal devices currently connected to the LoRa gateway. This represents the maximum number of connected devices that a LoRa gateway can support. The link scheduling decision conditions are: ; in, The 5G link load threshold is set to 80%. Set the LoRa gateway load threshold to 70%.

[0011] Preferably, the tiered, coordinated, closed-loop response mechanism includes the following: based on the difference between the integrated monitoring value and the dynamic early warning threshold... Classify warning levels, when At that time, it was classified as a general hidden danger, among which ;when At that time, it was classified as a major hidden danger, among which, ;when When the warning level is reached, it is classified as a major hidden danger; based on the warning level, the corresponding linkage strategy is automatically triggered.

[0012] Preferably, the linkage strategy includes the following: general hazards trigger audible and visual alarms; major hazards trigger emergency broadcasts, access control unlocking, and smoke exhaust fan startup; major hazards trigger fire sprinklers, emergency power cut-off, and 119 automatic alarm, while simultaneously linking cameras to focus on the hazard area and push the image.

[0013] An early warning method for a smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-Lora includes the following steps: S1. Various monitoring devices in the terminal sensing layer collect multi-dimensional security and fire safety data in real time. The collected data is transmitted in encrypted form and processed in a standardized manner. S2. Standardized security and fire data is transmitted through a 5G-Lora dual-mode network. The transmission link is automatically allocated by a dynamic load balancing strategy. Low-speed data is transmitted through the Lora link, while high-speed, high-priority data is transmitted through the 5G link. The edge computing gateway performs preliminary screening and cleaning of the data, and after removing invalid data, it is synchronized to the cloud platform and local cache. S3, the cloud platform adopts an adaptive threshold early warning algorithm for multimodal data fusion. First, it calculates the integrated monitoring value after fusion through a weighted fusion formula, and then combines it with the dynamic early warning threshold. It uses a difference formula to determine whether there are any safety hazards, and at the same time uses a correlation coefficient formula to identify hidden hazards, generate early warning information and determine the early warning level. S4. The system automatically triggers the corresponding linkage strategy according to the warning level and scenario type, controls the terminal execution layer devices to work together, collects and handles data in real time and updates the warning status. S5. After the hazard is dealt with, the early warning data, handling data, and linkage data are stored in the cloud database to form a historical archive. The adaptive threshold early warning algorithm continuously learns from the historical data to optimize the early warning accuracy.

[0014] Therefore, the present invention employs the above-mentioned intelligent security and fire protection linkage monitoring system and early warning method based on the Internet of Things and 5G-LoRa, which has the following advantages: (1) In this invention, by constructing a multimodal data fusion system and combining it with an adaptive threshold early warning algorithm, the multi-dimensional safety and fire data such as smoke, heat, and combustible gas are standardized and weighted and fused. At the same time, by analyzing the correlation coefficient of the multimodal data, the hidden dangers that are not exceeded in a single mode but are abnormal in multimodal coordination can be accurately identified, effectively reducing the false alarm and missed alarm rates, realizing the transformation of safety and fire hazards from passive response to active prevention, and improving the comprehensiveness and accuracy of hazard identification.

[0015] (2) In this invention, a 5G-Lora dual-mode heterogeneous network architecture is adopted, and a dynamic load balancing strategy is designed to monitor the load status of the 5G link and the Lora link in real time. The data transmission link is dynamically allocated according to the load rate, which takes into account the advantages of high speed and low latency of 5G communication, while retaining the low power consumption and wide coverage of Lora communication. This ensures stable and efficient data transmission in complex scenarios and avoids data loss or transmission delay caused by a single link failure.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the framework of a smart security and fire protection linkage monitoring system based on the Internet of Things and 5G-Lora according to the present invention; Figure 2 This is a schematic diagram of the early warning method of a smart security and fire protection linkage monitoring system based on the Internet of Things and 5G-Lora according to the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Specific model specifications need to be selected and determined according to the actual specifications of the device, etc. The specific selection calculation method adopts existing technology in the art, and therefore will not be described in detail.

[0019] Example like Figure 1 As shown, the present invention provides a smart security and fire protection linkage monitoring system based on the Internet of Things and 5G-LoRa, including a terminal sensing layer, a network transmission layer, a cloud platform layer and a terminal execution layer; The terminal sensing layer deploys various safety and fire monitoring devices to collect multi-dimensional safety and fire data such as smoke concentration, ambient temperature, and combustible gas concentration. The monitoring devices in the terminal sensing layer include high-definition network cameras (supporting flame and smoke feature recognition, relying on the camera's built-in recognition algorithm to extract the core features of flame and smoke, and then quantifying and normalizing them to obtain video image feature values ​​that can be directly used for standardized processing), smoke detectors, heat detectors, combustible gas detectors, electrical fire monitoring detectors, infrared intrusion detectors, emergency broadcast terminals, etc. All devices support the Internet of Things protocol, enabling real-time data acquisition and simple local alarms. The network transport layer adopts a 5G and LoRa dual-mode communication architecture, including a LoRa gateway, a 5G router, and an edge computing gateway, to achieve low-power, high-speed transmission of multi-dimensional security and fire data, as well as local preprocessing and cloud synchronization. The cloud platform layer is used for data aggregation, storage, query, statistical analysis, real-time monitoring of equipment status, and historical data backtracking. The cloud platform layer integrates a multimodal data fusion module, an adaptive threshold early warning module, a dynamic load balancing module, a hierarchical linkage handling module, and a data backtracking and model optimization module. The multimodal data fusion module is used to achieve deep fusion of multi-dimensional security and fire data; the adaptive threshold early warning module is used to achieve dynamic early warning threshold generation and hidden danger identification; the dynamic load balancing module is used to achieve dynamic scheduling of network link load; and the hierarchical linkage response module is used to achieve hierarchical linkage closed-loop response to security and fire incidents. The terminal execution layer is equipped with emergency linkage devices to receive linkage commands issued by the system and realize coordinated response of fire protection and security equipment. The emergency linkage devices in the terminal execution layer include fire sprinkler systems, smoke exhaust fans, emergency lighting, access control systems, audible and visual alarms, etc. All devices support remote control and local manual control.

[0020] The deep fusion of multi-dimensional security and fire protection data is specifically carried out in the following way: a weighted fusion algorithm is used to perform deep fusion of multi-dimensional security and fire protection data, and the fusion formula is as follows: ; in, This is the comprehensive monitoring value after multimodal data fusion. The number of modal data types participating in the fusion can be adjusted according to the scenario. For the first Weighting coefficients for modal data The value range of is [0,1], and satisfies , For the first The standardized values ​​of the modal data are unitless and range from [0, 10]. The standardization formula is: ; in, For the first The raw acquired values ​​of the modal data, For the first The minimum reasonable value for each modal data is set according to the sensor parameters. For the first The maximum reasonable value of each modal data is set according to safety and fire protection regulations and sensor range settings.

[0021] The dynamic warning threshold is generated based on a clustering analysis algorithm. Machine learning is used to train the algorithm on historical normal data from different scenarios, dynamically generating warning thresholds adapted to the specific scenario. The calculation formula is as follows: ; in, For dynamic early warning thresholds, This is the average of the comprehensive monitoring values ​​after integrating historical normal data. The standard deviation of the comprehensive monitoring value after integrating historical normal data. This is a threshold adjustment coefficient, with a value range of [1.5, 2.5], such as in densely populated areas or flammable and explosive areas. k =1.5, standard office area k =2.0, underground parking garage k =2.5.

[0022] The identification of hidden hazards is carried out in the following way: The correlation coefficient is used to determine the degree of correlation between different modalities of data to assist in the identification of hidden hazards. The formula for the correlation coefficient is: ; in, The correlation coefficient between the two modalities is... The covariance of the two modal data The standard deviation of the first modality data. The standard deviation of the second modality data; when And neither modal data exceeded the standard, but When this occurs, it is identified as a hidden danger; if the correlation coefficient between a core mode and at least two related modes is... This indicates that the multi-dimensional data shows a strong synergistic correlation, pointing to the same hidden risk trend.

[0023] 5G-Lora heterogeneous network dynamic load balancing includes the following: real-time detection of the load status of 5G links and LoRa gateways, calculation of load status using a load rate formula, and dynamic allocation of data transmission links based on scheduling conditions. The 5G link load rate calculation formula is as follows: ; in, This refers to the real-time load rate of the 5G link. This represents the current bandwidth usage of the 5G link. This refers to the total bandwidth of the 5G link; The formula for calculating the load rate of a Lora gateway is: ; in, This refers to the real-time load rate of the LoRa gateway. This represents the number of terminal devices currently connected to the LoRa gateway. This represents the maximum number of connected devices that a LoRa gateway can support. The link scheduling decision conditions are: ; in, The 5G link load threshold is set to 80%. Set the LoRa gateway load threshold to 70%.

[0024] The tiered, coordinated, and closed-loop response mechanism includes the following: based on the difference between the integrated monitoring value and the dynamic early warning threshold... Classify warning levels, when At that time, it was classified as a general hidden danger, among which ;when At that time, it was classified as a major hidden danger, among which, ;when When the warning level is reached, it is classified as a major hidden danger; based on the warning level, the corresponding linkage strategy is automatically triggered.

[0025] The linkage strategy includes the following: general hazards trigger audible and visual alarms; major hazards trigger emergency broadcasts, access control unlocking, and smoke exhaust fan startup; major hazards trigger fire sprinklers, emergency power cut-off, and 119 automatic alarm, while simultaneously linking cameras to focus on the hazard area and push the image.

[0026] like Figure 2 As shown, an early warning method for a smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa includes the following steps: S1. Various monitoring devices in the terminal sensing layer collect multi-dimensional security and fire safety data in real time. The collected data is transmitted in encrypted form and processed in a standardized manner. S2. Standardized security and fire data is transmitted through a 5G-Lora dual-mode network. The transmission link is automatically allocated by a dynamic load balancing strategy. Low-speed data (data with low real-time requirements, such as historical data backup) is transmitted through the LoRa link, while high-speed, high-priority data is transmitted through the 5G link. The edge computing gateway performs preliminary screening and cleaning of the data, and after removing invalid data, it is synchronized to the cloud platform and local cache. S3, the cloud platform adopts an adaptive threshold early warning algorithm for multimodal data fusion. First, it calculates the integrated monitoring value after fusion through a weighted fusion formula, and then combines it with the dynamic early warning threshold. It uses a difference formula to determine whether there are any safety hazards, and at the same time uses a correlation coefficient formula to identify hidden hazards, generate early warning information and determine the early warning level. S4. The system automatically triggers the corresponding linkage strategy according to the warning level and scenario type, controls the terminal execution layer devices to work together, collects and handles data in real time and updates the warning status. S5. After the hazard is dealt with, the early warning data, handling data, and linkage data are stored in the cloud database to form a historical archive. The adaptive threshold early warning algorithm continuously learns from the historical data to optimize the early warning accuracy.

[0027] Therefore, this invention adopts a smart safety and fire protection linkage monitoring system and early warning method based on the Internet of Things and 5G-Lora to ensure stable and efficient data transmission of safety and fire protection data in complex scenarios, avoid data loss or transmission delay caused by single link failure, effectively reduce false alarm and missed alarm rates, realize the transformation of safety and fire protection hazards from passive response to active prevention, and improve the comprehensiveness and accuracy of hazard identification.

[0028] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa, characterized in that: It includes the terminal perception layer, network transmission layer, cloud platform layer, and terminal execution layer; The terminal sensing layer is equipped with a variety of safety and fire monitoring devices to collect multi-dimensional safety and fire data, including smoke concentration, ambient temperature, and combustible gas concentration. The network transport layer adopts a 5G and LoRa dual-mode communication architecture, including a LoRa gateway, a 5G router, and an edge computing gateway, to achieve low-power, high-speed transmission of multi-dimensional security and fire data, as well as local preprocessing and cloud synchronization. The cloud platform layer is used for data aggregation, storage, query, statistical analysis, real-time monitoring of equipment status, and historical data backtracking. The cloud platform layer integrates a multimodal data fusion module, an adaptive threshold early warning module, a dynamic load balancing module, a hierarchical linkage handling module, and a data backtracking and model optimization module. The multimodal data fusion module is used to achieve deep fusion of multi-dimensional security and fire data; the adaptive threshold early warning module is used to achieve dynamic early warning threshold generation and hidden danger identification; the dynamic load balancing module is used to achieve dynamic scheduling of network link load; and the hierarchical linkage response module is used to achieve hierarchical linkage closed-loop response to security and fire incidents. Emergency response equipment is deployed at the terminal execution layer to receive linkage commands issued by the system and enable coordinated response from fire protection and security equipment.

2. The intelligent security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa as described in claim 1, characterized in that: The deep fusion of multi-dimensional security and fire protection data is specifically carried out in the following way: a weighted fusion algorithm is used to perform deep fusion of multi-dimensional security and fire protection data, and the fusion formula is as follows: ； in, This is the comprehensive monitoring value after multimodal data fusion. The number of modal data types participating in the fusion. For the first Weighting coefficients for modal data The value range of is [0,1], and satisfies , For the first The standardized values ​​of the modal data are unitless and range from [0, 10]. The standardization formula is: ； in, For the first The raw acquired values ​​of the modal data, For the first The minimum reasonable value of modal data, For the first The maximum reasonable value of the modal data.

3. The intelligent security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa as described in claim 2, characterized in that: The dynamic warning threshold is generated based on a clustering analysis algorithm. Machine learning is used to train the algorithm on historical normal data from different scenarios, dynamically generating warning thresholds adapted to the specific scenario. The calculation formula is as follows: ； in, For dynamic early warning thresholds, This is the average of the comprehensive monitoring values ​​after integrating historical normal data. The standard deviation of the comprehensive monitoring value after integrating historical normal data. This is the threshold adjustment coefficient, with a value range of [1.5, 2.5].

4. The intelligent security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa as described in claim 3, characterized in that: The identification of hidden hazards is carried out in the following way: The correlation coefficient is used to determine the degree of correlation between different modalities of data to assist in the identification of hidden hazards. The formula for the correlation coefficient is: ； in, The correlation coefficient between the two modalities is... The covariance of the two modal data The standard deviation of the first modality data. The standard deviation of the second modality data; when And neither modal data exceeded the standard, but At that time, it was determined to be a hidden danger.

5. A smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa as described in claim 4, characterized in that: 5G-Lora heterogeneous network dynamic load balancing includes the following: real-time detection of the load status of 5G links and LoRa gateways, calculation of load status using a load rate formula, and dynamic allocation of data transmission links based on scheduling conditions. The 5G link load rate calculation formula is as follows: ； in, This refers to the real-time load rate of the 5G link. This represents the current bandwidth usage of the 5G link. This refers to the total bandwidth of the 5G link; The formula for calculating the load rate of a Lora gateway is: ； in, This refers to the real-time load rate of the LoRa gateway. This represents the number of terminal devices currently connected to the LoRa gateway. This represents the maximum number of connected devices that a LoRa gateway can support. The link scheduling decision conditions are: ； in, The 5G link load threshold is set to 80%. Set the LoRa gateway load threshold to 70%.

6. The intelligent security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa as described in claim 5, characterized in that: The tiered, coordinated, and closed-loop response mechanism includes the following: based on the difference between the integrated monitoring value and the dynamic early warning threshold... Classify warning levels, when At that time, it was classified as a general hidden danger, among which ;when At that time, it was classified as a major hidden danger, among which, ;when When the warning level is reached, it is classified as a major hidden danger; based on the warning level, the corresponding linkage strategy is automatically triggered.

7. A smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa as described in claim 6, characterized in that: The linkage strategy includes the following: general hazards trigger audible and visual alarms; major hazards trigger emergency broadcasts, access control unlocking, and smoke exhaust fan startup; major hazards trigger fire sprinklers, emergency power cut-off, and 119 automatic alarm, while simultaneously linking cameras to focus on the hazard area and push the image.

8. An early warning method for a smart security and fire prevention linkage monitoring system based on the Internet of Things and 5G-LoRa, as described in any one of claims 1-7, characterized in that: Includes the following steps: S1. Various monitoring devices in the terminal sensing layer collect multi-dimensional security and fire safety data in real time. The collected data is transmitted in encrypted form and processed in a standardized manner. S2. Standardized security and fire data is transmitted through a 5G-Lora dual-mode network. The transmission link is automatically allocated by a dynamic load balancing strategy. Low-speed data is transmitted through the Lora link, while high-speed, high-priority data is transmitted through the 5G link. The edge computing gateway performs preliminary screening and cleaning of the data, and after removing invalid data, it is synchronized to the cloud platform and local cache. S3, the cloud platform adopts an adaptive threshold early warning algorithm for multimodal data fusion. First, it calculates the integrated monitoring value after fusion through a weighted fusion formula, and then combines it with the dynamic early warning threshold. It uses a difference formula to determine whether there are any safety hazards, and at the same time uses a correlation coefficient formula to identify hidden hazards, generate early warning information and determine the early warning level. S4. The system automatically triggers the corresponding linkage strategy according to the warning level and scenario type, controls the terminal execution layer devices to work together, collects and handles data in real time and updates the warning status. S5. After the hazard is dealt with, the early warning data, handling data, and linkage data are stored in the cloud database to form a historical archive. The adaptive threshold early warning algorithm continuously learns from the historical data to optimize the early warning accuracy.

Images