Fire prevention and early warning method and system based on internet of things

Abstract

This invention discloses a forestry fire prevention and early warning method and system based on the Internet of Things (IoT), relating to the fields of IoT and forestry fire monitoring technology. The method includes the following steps: S1, real-time acquisition of environmental parameters collected by each sensor node in the IoT forestry fire prevention system, including temperature, humidity, smoke concentration, and flame intensity; S2, sensor nodes preliminarily assessing fire risk based on environmental parameters and generating potential fire alarm data, which includes the source node, type, time, and original environmental parameters. This invention acquires environmental parameters such as temperature, humidity, smoke concentration, and flame intensity collected by sensor nodes in real time, performs a preliminary fire risk assessment by the sensors to generate alarm data, and then combines a preset fire discrimination model with geographical information to confirm the fire alarm level and allocate transmission priority. Simultaneously, it monitors network congestion parameters in real time to establish an evaluation model and dynamically generates a transmission scheduling strategy based on the alarm level, priority, and congestion index.

Description

Technical Field

[0001] This invention relates to the fields of Internet of Things (IoT) and forestry fire monitoring technology, specifically to a forestry fire early warning method and system based on IoT. Background Technology

[0002] Forest fires are a major ecological security issue facing the world, characterized by their suddenness, destructiveness, and difficulty in extinguishing. Traditional forestry fire prevention relies on manual patrols and fixed-point monitoring, which is not only inefficient but also makes it difficult to detect fires in a timely manner. With the development of Internet of Things (IoT) technology, deploying sensor nodes to collect environmental data such as temperature, humidity, and smoke concentration in real time, combined with wireless transmission technology, to achieve early fire warnings, has become an important research direction in forestry fire prevention. The application of IoT technology has significantly improved the timeliness and coverage of monitoring, providing new technical support for forest fire prevention and control.

[0003] However, existing IoT-based forestry fire prevention systems suffer from the following problems in data transmission: When a fire occurs, numerous sensor nodes simultaneously upload alarm data, leading to network congestion or even paralysis. Due to limited wireless communication bandwidth, data packets from multiple nodes may interfere with or be lost during transmission, thus delaying the delivery of critical fire information. Although existing systems employ simple priority scheduling mechanisms, they cannot dynamically adjust data transmission strategies, especially under high-load scenarios during sudden fires, failing to guarantee the reliability and real-time performance of data from critical nodes. This technical deficiency directly impacts the accuracy and response speed of fire warnings and urgently requires targeted solutions. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a forestry fire prevention and early warning method and system based on the Internet of Things (IoT). This solves the problems of network congestion and paralysis caused by a large amount of alarm data when a fire occurs, data loss due to interference, and the inability to dynamically adjust transmission strategies to ensure real-time and reliable transmission of critical data, compared to existing technologies.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a forestry fire early warning method based on the Internet of Things, comprising: S1. Real-time acquisition of environmental parameters collected by each sensor node in the IoT forestry fire prevention system, including temperature, humidity, smoke concentration and flame intensity; S2. The sensor node makes a preliminary assessment of the fire risk based on environmental parameters and generates potential fire alarm data. This alarm data includes the source node, type, time and original environmental parameters. S3. Based on the alarm data, combined with the preset fire identification model and geographical information, the system assesses and confirms the fire alarm level, and assigns it transmission priority and critical identifier. S4. Monitor the network congestion parameters of the wireless communication network in real time, including bandwidth utilization, packet loss rate and latency, and establish a congestion assessment model. S5. Based on the parameters obtained from the fire alarm level, transmission priority, key identifiers and congestion assessment model, dynamically generate the data transmission scheduling strategy for each sensor node. S6. Based on the scheduling strategy, dynamically adjust the transmission rate, retransmission count, transmission path and transmission timing of each sensor node to ensure that high-priority alarm data is transmitted in real time and reliably when the network is congested, and suppress or delay low-priority data.

[0006] Furthermore, the assessment and confirmation of the fire alarm level includes: Receive and denoise potential fire alarm data; The denoised data is input into the fire situation discrimination model, which integrates historical fire situation, meteorological and geographical information, and outputs an initial fire risk index. Based on risk index and geographic information, combined with the alarm status of adjacent nodes, the spread trend is predicted, and the final fire alarm level is determined, including high risk, medium risk and low risk.

[0007] Furthermore, the allocation of transmission priority and critical identifier includes: High-risk fire alarm data is assigned the highest priority and critical identifiers to ensure zero-latency and highly reliable transmission; Moderate fire alarm data is assigned medium priority and secondary criticality labels, allowing for moderate delays to ensure highly reliable transmission; Low-level fire alarm data is assigned the lowest priority and is not critical, allowing for greater latency and can be transmitted or aggregated when the network is idle.

[0008] Furthermore, the monitoring of network congestion parameters of the wireless communication network and the establishment of a network congestion assessment model include: Periodically collect data forwarding volume, buffer occupancy rate, and link throughput of relay nodes and aggregation nodes in the wireless communication network; The collected parameters are input into a deep learning model, which predicts short-term congestion trends and levels based on historical network data. Based on the prediction results, a numerical network congestion index is generated to drive dynamic adjustments to the scheduling strategy.

[0009] Furthermore, the dynamic generation of data transmission scheduling strategies for each sensor node includes: Based on the fire alarm level, priority, critical indicators, and network congestion index, calculate the real-time transmission requirements of each sensor node; Based on demand and available transmission resources, a multi-objective optimization algorithm is adopted to generate a scheduling strategy, taking into account real-time performance, reliability and load balancing. The strategy includes the recommended transmission rate, maximum number of retransmissions, preferred transmission path, and next transmission timing for each sensor node; Transmission path selection is based on the principle of shortest path or least congestion.

[0010] Furthermore, the dynamic adjustment of the transmission rate, retransmission count, transmission path, and transmission timing of each sensor node includes: For high-priority alarm data, set the rate to the maximum and set a high retransmission limit to ensure data arrival; For medium-priority alarm data, the rate is adjusted to medium, allowing a reasonable amount of retransmission to balance real-time performance and load. For low-priority data, if network congestion is high, the transmission rate will be reduced to the minimum or buffered and delayed. Choose the least congested or most bandwidth-efficient link as the transmission path, adjust the sending time, and avoid data packet contention.

[0011] Furthermore, the method of enabling high-priority alarm data to be transmitted reliably in real time during network congestion includes: For high-priority alarm data, forward error correction coding or redundant multipath transmission is used to resist link interruption and data loss. The system uses a retransmission confirmation mechanism to monitor the transmission status in real time. If the transmission fails or times out, it will immediately retransmit or switch to an alternative path. Reserve dedicated bandwidth or time slice resources for high-priority data packets so that transmission is not affected by low-priority data; Real-time transmission sets a strict latency limit, and when the predicted latency exceeds the limit, the system issues an early warning and adjusts its strategy.

[0012] Furthermore, the suppression or delay of low-priority data includes: When the network congestion index reaches the threshold, low-priority packets are marked and instructed to be cached locally by nodes until the congestion is relieved. For specific low-priority data, such as routine environmental monitoring, aggregation and periodic transmission are performed when the network is under high load to reduce channel contention and overhead; In cases of extreme congestion, based on the data discarding strategy, some of the least critical and low-priority data packets are selectively discarded to release resources; Suppression and delay strategies are dynamically adjusted based on fire alarm levels and congestion indices.

[0013] Furthermore, the method also includes: Transmit high-priority alarm data to the remote command center; Based on alarm data, combined with forest area maps and meteorological information, the remote command center generates a fire spread model and firefighting plan; After receiving alarm data, the mobile terminal automatically triggers an emergency notification, displaying the location of the fire and evacuation routes; After the data transmission is completed, the system sends a confirmation message to the administrator confirming the successful transmission.

[0014] This invention also provides an IoT-based forestry fire early warning system, applied to the aforementioned IoT-based forestry fire early warning method, comprising: The data acquisition module is used to acquire environmental parameter information and potential fire alarm data collected by each sensor node in real time; The fire assessment module is used to assess the fire alarm level based on potential fire alarm data, preset fire discrimination models and geographical information, and to assign transmission priority and key identifiers to the alarm data. The network monitoring module is used to monitor network congestion parameters of the wireless communication network in real time, including bandwidth utilization, packet loss rate and latency, and to establish a congestion assessment model. The transmission strategy generation module is used to dynamically generate data transmission scheduling strategies for each sensor node based on parameters obtained from fire alarm level, transmission priority, critical identifiers and congestion assessment model. The data transmission control module is used to dynamically adjust the transmission rate, retransmission count, transmission path and transmission timing of each sensor node according to the scheduling strategy, so that high-priority fire alarm data can be transmitted in real time and reliably when the network is congested, and low-priority data can be suppressed or delayed.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention acquires environmental parameters such as temperature, humidity, smoke concentration, and flame intensity from sensor nodes in real time. The sensors initially assess fire risk and generate alarm data. This data is then combined with a pre-set fire risk assessment model and geographic information to confirm the fire alarm level and assign transmission priorities. Simultaneously, network congestion parameters are monitored in real time to establish an evaluation model. Based on the alarm level, priority, and congestion index, a dynamic transmission scheduling strategy is generated, adjusting the sensor transmission rate, retransmission count, transmission path, and timing. This effectively solves the problems of network congestion paralysis, data loss, and lack of dynamic scheduling in existing systems during fires. By using hierarchical prioritization and dynamic strategies, it avoids interference caused by limited bandwidth. High-priority data uses forward error correction coding and reserved bandwidth to ensure real-time reliable transmission, while low-priority data is cached or aggregated to reduce channel contention. Furthermore, high-priority data can be transmitted to the command center to generate firefighting plans, improving the accuracy and response speed of fire warnings and ensuring efficient forestry fire prevention. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a system structure diagram of the present invention. Detailed Implementation

[0017] 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, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Please see Figure 1 This invention provides a forestry fire prevention and early warning method based on the Internet of Things, comprising: S1. Real-time acquisition of environmental parameters collected by each sensor node in the IoT forestry fire prevention system, including temperature, humidity, smoke concentration and flame intensity; S2. The sensor node makes a preliminary assessment of the fire risk based on environmental parameters and generates potential fire alarm data. This alarm data includes the source node, type, time and original environmental parameters. S3. Based on the alarm data, combined with the preset fire identification model and geographical information, the system assesses and confirms the fire alarm level, and assigns it transmission priority and critical identifier. S4. Monitor the network congestion parameters of the wireless communication network in real time, including bandwidth utilization, packet loss rate and latency, and establish a congestion assessment model. S5. Based on the parameters obtained from the fire alarm level, transmission priority, key identifiers and congestion assessment model, dynamically generate the data transmission scheduling strategy for each sensor node. S6. Based on the scheduling strategy, dynamically adjust the transmission rate, retransmission count, transmission path and transmission timing of each sensor node to ensure that high-priority alarm data is transmitted in real time and reliably when the network is congested, and suppress or delay low-priority data.

[0019] Specifically, step S1 is executed first to acquire environmental parameters in real time. The deployed sensor nodes include various types, such as the SHT30 model temperature and humidity sensor, which measures from -40℃ to 125℃ and 0 to 100%RH; the MQ-2 model smoke concentration sensor, which can detect smoke concentrations from 0.1 to 10 ppm; and the UV-TRON model flame intensity sensor, which can identify ultraviolet flame signals with wavelengths from 200 to 280 nm. These sensor nodes are distributed in the forest area at 50-meter intervals, collecting data every 2 seconds and uploading it to nearby relay nodes.

[0020] In step S2, the sensor nodes perform a preliminary fire risk assessment. Risk thresholds for various parameters are pre-set; for example, when the temperature exceeds 35℃, humidity is below 40%, smoke concentration exceeds 50ppm, or ultraviolet light is detected in the flame intensity, the sensor nodes automatically generate potential fire alarm data. The alarm data includes a unique identifier code for the source node, such as Node-086; the data type is labeled as abnormal temperature or abnormal smoke; the acquisition time is accurate to milliseconds; and the original environmental parameters fully record the specific values ​​of the currently acquired temperature, humidity, smoke concentration, and flame intensity.

[0021] Step S3: The system assesses the fire alarm level and assigns transmission priorities. The preset fire discrimination model uses the random forest algorithm, which integrates nearly 10 years of historical fire data, real-time meteorological information, and geographical information for the forest area. The geographical information includes the forest area's altitude, vegetation type, and distance from water sources. First, the received potential fire alarm data is denoised using a Kalman filter algorithm to eliminate outliers caused by environmental interference. The denoised data is then input into the fire discrimination model, which outputs an initial fire risk index, calculated using the following formula: ; in, This is the initial fire risk index, with a value ranging from 0 to 1; , , , The weights for temperature, humidity, smoke concentration, and flame intensity are respectively, and The weighting coefficients were obtained by combining the Analytic Hierarchy Process (AHP) with historical fire data. , , These are the current temperature, humidity, and smoke concentration, respectively. , These represent the lowest and highest historical temperatures in the forest area, for example, -20℃ and 40℃. , These represent the lowest and highest historical humidity levels in the forest area, for example, 10%RH and 90%RH. , These are the minimum detection value and the danger threshold for smoke concentration, such as 0 ppm and 100 ppm, respectively. The fire intensity is indicated by a value of 1 when a fire is detected and 0 when no fire is detected. The spread trend is then predicted by combining geographical information and the alarm status of adjacent nodes. For example, if a node's initial risk index is 0.8, and two of its three adjacent nodes also generate alarm data, and the area is coniferous forest with a wind speed greater than 3 meters per second, the final fire alarm level is determined to be high-risk, and the alarm data is assigned the highest transmission priority and criticality label. If the initial risk index is 0.5, and only one adjacent node alarms, and the vegetation is broadleaf forest, the level is moderate, and it is assigned medium priority and secondary criticality labels. If the initial risk index is below 0.3 and no adjacent nodes alarm, the level is low, and it is assigned the lowest priority and non-criticality labels.

[0022] Step S4 involves real-time monitoring of congestion parameters in the wireless communication network and the establishment of a congestion assessment model. A 10-second sampling period is used to collect data forwarding volume, buffer occupancy rate, and link throughput of relay nodes and aggregation nodes in the wireless communication network. Data forwarding volume is the total number of data packets forwarded by a node in each period; buffer occupancy rate is the number of data packets currently stored in the buffer divided by the total buffer capacity (e.g., if the total buffer capacity is 1000 data packets and 300 are currently stored, the occupancy rate is 0.3); link throughput is the actual amount of data transmitted by the link in each period divided by the maximum transmission capacity of the link. These collected parameters are input into an LSTM-based deep learning model, which, trained on nearly three months of historical network data, can predict the network congestion trend and severity within the next 30 seconds. A numerical network congestion index is generated based on the prediction results, calculated as follows: ; in, This is the network congestion index, with a value ranging from 0 to 1; , , These are the weighting coefficients for bandwidth utilization, packet loss rate, and latency, respectively. The weighting coefficients are obtained by combining the entropy weighting method with expert correction. As of the current bandwidth utilization, The maximum bandwidth utilization is 1; This represents the packet loss rate, with a value ranging from 0 to 1. For the current delay, The preset maximum allowable delay, for example, 500 milliseconds.

[0023] Step S5 generates a transmission scheduling strategy based on multiple parameters. First, the real-time transmission requirements of each sensor node are calculated using the following formula: ; in, For real-time transmission requirements, the value range is 0 to 1; , , These are the weighting coefficients for fire alarm level, transmission priority, and network congestion index, respectively. The weighting coefficients were obtained through the Delphi method combined with simulation experiments. These are the normalized values ​​for fire alarm levels: 1 for high risk, 0.5 for moderate risk, and 0 for low risk. This is the normalized value for transmission priority, with the highest being 1, medium being 0.5, and the lowest being 0. This represents the network congestion index. Then, based on transmission demand and available transmission resources, the NSGA-II multi-objective optimization algorithm is used to generate a scheduling strategy with the objectives of minimizing transmission delay, maximizing transmission reliability, and balancing network load. The strategy specifies the recommended transmission rate for each sensor node (e.g., 1 Mbps for the highest priority node); the maximum number of retransmissions (e.g., 5 for the highest priority); the preferred transmission path (e.g., selecting the shortest path based on Dijkstra's algorithm or the path with the highest link throughput); and the timing of the next transmission (e.g., avoiding periods of high network load by sending 2 seconds after the current cycle).

[0024] Step S6 dynamically adjusts transmission parameters according to the scheduling strategy. For high-priority alarm data, the sending rate is set to the maximum, and a high retransmission limit is set to ensure successful data delivery. For medium-priority alarm data, the sending rate is set to medium, allowing a moderate number of retransmissions to balance network load while ensuring real-time performance. For low-priority data, when the network congestion index exceeds 0.8, the sending rate is reduced to the minimum, such as 0.1 Mbps, or buffered to local node storage to delay transmission. The transmission path is selected from links with a utilization rate of less than 50% and a packet loss rate of less than 5%, and the sending time is adjusted, for example, staggered by 1 second from the sending time of other high-priority nodes, to avoid data packets competing for the channel during transmission.

[0025] In this embodiment, assessing and confirming the fire alarm level includes: Receive and denoise potential fire alarm data; The denoised data is input into the fire situation discrimination model, which integrates historical fire situation, meteorological and geographical information, and outputs an initial fire risk index. Based on risk index and geographic information, combined with the alarm status of adjacent nodes, the spread trend is predicted, and the final fire alarm level is determined, including high risk, medium risk and low risk.

[0026] Specifically, after receiving potential fire alarm data uploaded by sensor nodes, the Kalman filter algorithm is first used to denoise the data, eliminating abnormal values ​​caused by sensor hardware errors or environmental interference. For example, the temperature value that jumps instantaneously is corrected to a value that conforms to the normal trend of change.

[0027] The denoised data is input into a fire risk assessment model, which integrates historical fire data from the past five years, real-time meteorological information, geographic information, and other data from the forest area. The model calculates and outputs an initial fire risk index using the initial fire risk index formula.

[0028] Based on the initial fire risk index and geographical information, the fire spread trend is predicted by combining the alarm status of adjacent nodes. For example, if the initial risk index of a node is 0.7, the node is located in a coniferous forest area at an altitude of 800 meters and a slope of 25 degrees, the real-time wind speed is 4 meters per second, and 3 out of its 5 adjacent nodes also generate alarm data within 10 seconds, then the fire spread speed is judged to be relatively fast, and the final fire alarm level is determined to be high risk; if the initial risk index is 0.4, only 1 adjacent node alarms, the area is a broad-leaved forest with a slope of 10 degrees and a wind speed of 1 meter per second, then the level is determined to be moderate; if the initial risk index is 0.2, no adjacent node alarms, and the area vegetation is grassland, then the level is determined to be low.

[0029] In this embodiment, allocating transmission priority and critical identifiers includes: High-risk fire alarm data is assigned the highest priority and critical identifiers to ensure zero-latency and highly reliable transmission; Moderate fire alarm data is assigned medium priority and secondary criticality labels, allowing for moderate delays to ensure highly reliable transmission; Low-level fire alarm data is assigned the lowest priority and is not critical, allowing for greater latency and can be transmitted or aggregated when the network is idle.

[0030] Specifically, based on the confirmed fire alarm level, different transmission priorities and criticality indicators are assigned to alarm data of different levels. For high-risk fire alarm data, such as alarm data with detected flame signals and temperatures exceeding 40°C and smoke concentrations exceeding 80ppm, the highest transmission priority and criticality indicator are assigned. This type of data enjoys the highest resource allocation during transmission, ensuring zero-latency and high-reliability transmission, ensuring the command center can obtain critical fire information immediately. For moderate fire alarm data, such as alarm data with temperatures between 35°C and 40°C, smoke concentrations between 50ppm and 80ppm, and no flame signals, a medium transmission priority and secondary criticality indicator are assigned, allowing moderate delays, such as no more than 200 milliseconds, while ensuring high transmission reliability to ensure timely transmission of fire information. For low-risk fire alarm data, such as alarm data with humidity below 40% and other parameters normal, the lowest transmission priority and non-criticality indicator are assigned, allowing larger delays, such as up to 1000 milliseconds, which can be transmitted when the network is idle or aggregated with other low-priority data before transmission.

[0031] In this embodiment, monitoring network congestion parameters of the wireless communication network and establishing a network congestion assessment model includes: Periodically collect data forwarding volume, buffer occupancy rate, and link throughput of relay nodes and aggregation nodes in the wireless communication network; The collected parameters are input into a deep learning model, which predicts short-term congestion trends and levels based on historical network data. Based on the prediction results, a numerical network congestion index is generated to drive dynamic adjustments to the scheduling strategy.

[0032] Specifically, the system periodically collects data forwarding volume, buffer occupancy rate, and link throughput data from relay nodes and aggregation nodes in the wireless communication network, with a 5-second sampling cycle. Data forwarding volume is obtained through the node's packet counter, recording the total number of packets forwarded by the node in each cycle. Buffer occupancy rate is obtained by reading the number of packets currently stored in the node's buffer and dividing it by the total buffer capacity. For example, if the total buffer capacity is 2000 packets and 600 packets are currently stored, the buffer occupancy rate is 0.3. Link throughput is calculated by testing the actual amount of data transmitted by the link in each cycle and dividing it by the link's maximum transmission capacity. For example, if the maximum transmission capacity is 2 Mbps and the actual transmission is 1 Mbps, the throughput is 0.5.

[0033] The collected parameters are input into an LSTM-based deep learning model. This model is trained on nearly three months of historical network data, including network parameter changes under different time periods and fire conditions. It can learn the patterns of network congestion changes and predict the trend and degree of network congestion in the next 30 seconds.

[0034] Based on the model's predictions, a numerical network congestion index is generated. The formula for calculating the network congestion index is as described above and will not be repeated here. The network congestion index can intuitively reflect the current network congestion status, providing an accurate quantitative basis for subsequent dynamic adjustments to data transmission scheduling strategies. This enables the system to proactively address potential network congestion and ensure the stability of data transmission.

[0035] In this embodiment, the data transmission scheduling strategy for each sensor node is dynamically generated, including: Based on the fire alarm level, priority, critical indicators, and network congestion index, calculate the real-time transmission requirements of each sensor node; Based on demand and available transmission resources, a multi-objective optimization algorithm is adopted to generate a scheduling strategy, taking into account real-time performance, reliability and load balancing. The strategy includes the recommended transmission rate, maximum number of retransmissions, preferred transmission path, and next transmission timing for each sensor node; Transmission path selection is based on the principle of shortest path or least congestion.

[0036] Specifically, the real-time transmission requirements of each sensor node are first calculated based on the fire alarm level, transmission priority, critical identifier, and network congestion index. The calculation process uses the formula in step S5. Based on the calculated real-time transmission requirements and currently available transmission resources, such as remaining network bandwidth and the number of idle links, the NSGA-II multi-objective optimization algorithm is adopted. This algorithm can find the optimal solution among multiple conflicting objectives, including minimizing transmission delay, maximizing transmission reliability, and balancing network load, and generate a data transmission scheduling strategy for each sensor node. The NSGA-II multi-objective optimization algorithm is a mature existing algorithm and will not be described in detail here.

[0037] The generated scheduling strategy includes recommended transmission rates for each sensor node (e.g., 1 Mbps for high-priority nodes, 0.5 Mbps for medium-priority nodes, and 0.2 Mbps for low-priority nodes); maximum retransmission count (e.g., 5 retransmissions for high-priority nodes, 3 for medium-priority nodes, and 1 for low-priority nodes); preferred transmission path (based on the shortest path or minimum congestion principle, such as using Dijkstra's algorithm to calculate the shortest path or selecting a path with a link congestion index below 0.3); and next transmission timing (e.g., selecting the time with the lowest network load within the next 10 seconds based on network congestion prediction results). This dynamically generated scheduling strategy allows the transmission behavior of each sensor node to match network conditions and data importance, improving the efficiency of network resource utilization.

[0038] In this embodiment, dynamically adjusting the transmission rate, retransmission count, transmission path, and transmission timing of each sensor node includes: For high-priority alarm data, set the rate to the maximum and set a high retransmission limit to ensure data arrival; For medium-priority alarm data, the rate is adjusted to medium, allowing a reasonable amount of retransmission to balance real-time performance and load. For low-priority data, if network congestion is high, the transmission rate will be reduced to the minimum or buffered and delayed. Choose the least congested or most bandwidth-efficient link as the transmission path, adjust the sending time, and avoid data packet contention.

[0039] Specifically, during actual data transmission, the transmission parameters of each sensor node are dynamically adjusted according to the generated scheduling strategy. For high-priority alarm data, the transmission rate is set to the maximum, for example, reaching the maximum transmission rate allowed by the link of 1Mbps, while a high retransmission limit is set, such as 5 times, to ensure that the data can successfully reach the receiving end. For medium-priority alarm data, the transmission rate is set to medium, for example, 0.5Mbps, allowing a reasonable number of retransmissions, such as 3 times, to balance the network load while ensuring real-time data transmission. For low-priority data, when the network congestion index exceeds 0.8, the transmission rate is reduced to the minimum, for example, 0.1Mbps, or temporarily cached in the node's local storage module, and sent after the network congestion is relieved.

[0040] Regarding transmission path selection, the status of each link is monitored in real time, and the least congested or most bandwidth-efficient link is selected as the transmission path. For example, links with a utilization rate of less than 40% and a packet loss rate of less than 3% are selected. Simultaneously, data transmission time is adjusted. For instance, when multiple nodes are detected preparing to send data within a certain time period, the transmission time of high-priority nodes is advanced, and the transmission time of low-priority nodes is delayed, avoiding channel contention during transmission and reducing data loss and latency.

[0041] In this embodiment, enabling high-priority alarm data to be transmitted reliably in real time during network congestion includes: For high-priority alarm data, forward error correction coding or redundant multipath transmission is used to resist link interruption and data loss. The system uses a retransmission confirmation mechanism to monitor the transmission status in real time. If the transmission fails or times out, it will immediately retransmit or switch to an alternative path. Reserve dedicated bandwidth or time slice resources for high-priority data packets so that transmission is not affected by low-priority data; Real-time transmission sets a strict latency limit, and when the predicted latency exceeds the limit, the system issues an early warning and adjusts its strategy.

[0042] Specifically, to ensure the real-time and reliable transmission of high-priority alarm data during network congestion, multiple safeguards are employed. Regarding data encoding, forward error correction encoding, such as RS(255,223) encoding, is used for high-priority alarm data. This encoding method automatically corrects some errors during data transmission. Simultaneously, redundant multipath transmission is employed, selecting two different transmission paths for high-priority data. If one path fails or becomes congested, the data can be transmitted via the other path, mitigating link interruptions and data loss.

[0043] In terms of transmission control, the transmission status of high-priority data is monitored in real time through an acknowledgment and retransmission mechanism. For example, using the ARQ protocol, the receiving end promptly sends acknowledgment information to the sending end after receiving the data. When the sending end does not receive acknowledgment information within the specified time or detects data transmission failure, it immediately initiates the retransmission mechanism or switches to an alternative transmission path to continue transmission.

[0044] In terms of resource reservation, dedicated bandwidth or time slice resources are reserved for high-priority data packets. For example, 30% of the total network bandwidth is reserved as dedicated bandwidth for high-priority data, or 30% of the time slice is allocated to high-priority data transmission in each transmission cycle, so that the transmission of high-priority data is not affected by low-priority data.

[0045] At the same time, a strict delay limit is set for high-priority data, such as 500 milliseconds. By monitoring the data transmission delay in real time, when the predicted delay exceeds the set limit, the system immediately issues an early warning and adjusts the transmission strategy to ensure that high-priority data can arrive within the specified time, thus ensuring the timeliness of fire warnings.

[0046] In this embodiment, suppressing or delaying low-priority data includes: When the network congestion index reaches the threshold, low-priority packets are marked and instructed to be cached locally by nodes until the congestion is relieved. For specific low-priority data, such as routine environmental monitoring, aggregation and periodic transmission are performed when the network is under high load to reduce channel contention and overhead; In cases of extreme congestion, based on the data discarding strategy, some of the least critical and low-priority data packets are selectively discarded to release resources; Suppression and delay strategies are dynamically adjusted based on fire alarm levels and congestion indices.

[0047] Specifically, when the network congestion index reaches a preset threshold, such as 0.8, the system automatically marks low-priority data packets and sends instructions to the corresponding sensor nodes, instructing the nodes to cache low-priority data packets in the local storage module until the network congestion index drops below the threshold and the network congestion is relieved before transmission.

[0048] For specific types of low-priority data, such as routine environmental monitoring data, when the network is under high load, data aggregation is used. For example, the collected routine environmental monitoring data is aggregated every hour to calculate key information such as the average, maximum, and minimum values ​​within that time period. The aggregated results are sent only periodically, such as once per hour, to reduce the amount of data transmitted, thereby reducing channel contention and network overhead.

[0049] In extreme congestion situations, such as when the network congestion index reaches 0.95, based on the preset data discarding policy, some of the least critical low-priority data packets are selectively discarded, such as historical environmental monitoring data from 3 days ago or repeatedly collected routine data, to free up network resources and ensure the transmission of high-priority data.

[0050] The strategy of suppressing and delaying low-priority data will be dynamically adjusted according to the current fire alarm level and network congestion index. For example, when the fire alarm level is high and the network congestion index is high, the suppression of low-priority data will be further strengthened. When the fire alarm level is reduced and network congestion is relieved, the restrictions on low-priority data will be appropriately relaxed to ensure that network resources can be reasonably allocated.

[0051] In this embodiment, the method further includes: Transmit high-priority alarm data to the remote command center; Based on alarm data, combined with forest area maps and meteorological information, the remote command center generates a fire spread model and firefighting plan; After receiving alarm data, the mobile terminal automatically triggers an emergency notification, displaying the location of the fire and evacuation routes; After the data transmission is completed, the system sends a confirmation message to the administrator confirming the successful transmission.

[0052] Specifically, high-priority alarm data is transmitted to the remote command center in real time after transmission is completed. The remote command center's system combines the alarm data with electronic maps of the forest area and real-time meteorological information. For example, it loads map information such as terrain and vegetation distribution of the forest area through a GIS system, and combines it with real-time meteorological data such as wind speed and direction to generate a fire spread model. For example, it uses a cellular automata model to simulate the spread speed and range of fire under different terrain and meteorological conditions. At the same time, based on the fire spread model and the distribution of fire-fighting resources in the forest area, it automatically generates a preliminary fire-fighting plan, including recommended fire-fighting routes, the quantity and type of fire-fighting resources to be allocated, etc.

[0053] After receiving alarm data, the mobile terminal will automatically trigger an emergency notification, alerting the user through sound, vibration, and pop-up windows. The notification will display the specific location of the fire, such as latitude and longitude coordinates and the corresponding forest area name, as well as evacuation routes generated based on the terrain and the direction of fire spread.

[0054] Once high-priority alarm data is successfully transmitted to the remote command center and related information is synchronized to the mobile terminal, the system will automatically send a transmission success confirmation message to the administrator. The confirmation message can be sent via SMS, system message, or other means to inform the administrator that the data has been successfully transmitted, facilitating the administrator to promptly carry out subsequent fire response work and improve the efficiency and safety of fire fighting.

[0055] Please see Figure 2 The present invention also provides an Internet of Things (IoT)-based forestry fire early warning system, applied to the aforementioned IoT-based forestry fire early warning method, comprising: The data acquisition module is used to acquire environmental parameter information and potential fire alarm data collected by each sensor node in real time; The fire assessment module is used to assess the fire alarm level based on potential fire alarm data, preset fire discrimination models and geographical information, and to assign transmission priority and key identifiers to the alarm data. The network monitoring module is used to monitor network congestion parameters of the wireless communication network in real time, including bandwidth utilization, packet loss rate and latency, and to establish a congestion assessment model. The transmission strategy generation module is used to dynamically generate data transmission scheduling strategies for each sensor node based on parameters obtained from fire alarm level, transmission priority, critical identifiers and congestion assessment model. The data transmission control module is used to dynamically adjust the transmission rate, retransmission count, transmission path and transmission timing of each sensor node according to the scheduling strategy, so that high-priority fire alarm data can be transmitted in real time and reliably when the network is congested, and low-priority data can be suppressed or delayed.

[0056] Specifically, the data acquisition module consists of sensor nodes and a data receiving unit. The sensor nodes include various types such as temperature and humidity sensors, smoke concentration sensors, and flame intensity sensors, which can collect environmental parameter information in real time. The data receiving unit is deployed at the relay node and receives environmental parameter information and potential fire alarm data uploaded by the sensor nodes through a wireless communication protocol. After preliminary processing of the data, it is transmitted to subsequent modules.

[0057] The fire assessment module is deployed on an edge gateway or cloud server, such as using a Raspberry Pi 4 as the edge gateway. This module receives potential fire alarm data transmitted by the data acquisition module, combines it with a preset fire discrimination model and geographical information, assesses the fire alarm level, and assigns corresponding transmission priorities and key identifiers to the alarm data according to the assessment results. The algorithm and formula in step S3 are used in the assessment process.

[0058] The network monitoring module is mainly deployed at relay nodes and aggregation nodes. Through the built-in monitoring program, it collects network congestion parameters such as bandwidth utilization, packet loss rate and latency of the wireless communication network in real time. The collection period is, for example, 10 seconds. At the same time, it establishes a congestion assessment model based on the collected parameters and generates a network congestion index. The index is calculated using the formula in step S4.

[0059] The transmission strategy generation module is deployed on a cloud server. This module receives the fire alarm level, transmission priority and criticality identifier sent by the fire assessment module, as well as the network congestion index sent by the network monitoring module. Based on these parameters, it calculates the real-time transmission requirements of each sensor node, uses the formula in step S5, and generates the data transmission scheduling strategy for each sensor node using a multi-objective optimization algorithm. The strategy is then sent to the data transmission control module.

[0060] The data transmission control module is embedded in the MCU of the sensor node. This module receives the scheduling strategy sent by the transmission strategy generation module and dynamically adjusts the transmission rate, retransmission number, transmission path and transmission timing of the sensor node according to the strategy. This enables real-time and reliable transmission of high-priority fire alarm data when the network is congested, while suppressing or delaying low-priority data to ensure the stable operation of the entire system.

[0061] In summary, this invention acquires environmental parameters such as temperature, humidity, smoke concentration, and flame intensity collected by sensor nodes in real time. The sensors initially assess the fire risk and generate alarm data. This data is then combined with a pre-set fire risk assessment model and geographic information to confirm the fire alarm level and assign transmission priorities. Simultaneously, network congestion parameters are monitored in real time to establish an evaluation model. Based on the alarm level, priority, and congestion index, a dynamic transmission scheduling strategy is generated, adjusting the sensor transmission rate, retransmission count, transmission path, and timing. This effectively solves the problems of network congestion paralysis, data loss, and lack of dynamic scheduling in existing systems during fires. By using hierarchical prioritization and dynamic strategies, interference caused by limited bandwidth is avoided. High-priority data uses forward error correction coding and reserved bandwidth to ensure real-time reliable transmission, while low-priority data is cached or aggregated to reduce channel contention. Furthermore, high-priority data can be transmitted to the command center to generate firefighting plans, improving the accuracy and response speed of fire warnings and ensuring efficient forestry fire prevention.

[0062] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0063] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A forest fire prevention and early warning method based on the Internet of Things, characterized in that, include: S1. Real-time acquisition of environmental parameters collected by each sensor node in the IoT forestry fire prevention system, including temperature, humidity, smoke concentration and flame intensity; S2. The sensor node makes a preliminary assessment of the fire risk based on environmental parameters and generates potential fire alarm data. This alarm data includes the source node, type, time and original environmental parameters. S3. Based on the alarm data, combined with the preset fire identification model and geographical information, the system assesses and confirms the fire alarm level, and assigns it transmission priority and critical identifier. S4. Monitor the network congestion parameters of the wireless communication network in real time, including bandwidth utilization, packet loss rate and latency, and establish a congestion assessment model. S5. Based on the parameters obtained from the fire alarm level, transmission priority, key identifiers and congestion assessment model, dynamically generate the data transmission scheduling strategy for each sensor node. S6. Based on the scheduling strategy, dynamically adjust the transmission rate, retransmission count, transmission path and transmission timing of each sensor node to ensure that high-priority alarm data is transmitted in real time and reliably when the network is congested, and suppress or delay low-priority data. 2.The forest fire prevention and early warning method based on the Internet of Things according to claim 1, characterized in that, The assessment and confirmation of the fire alarm level includes: Receive and denoise potential fire alarm data; The denoised data is input into the fire situation discrimination model, which integrates historical fire situation, meteorological and geographical information, and outputs an initial fire risk index. Based on risk index and geographic information, combined with the alarm status of adjacent nodes, the spread trend is predicted, and the final fire alarm level is determined, including high risk, medium risk and low risk. 3.The forest fire prevention and early warning method based on the Internet of Things according to claim 1, characterized in that, The allocation of transmission priority and key identifiers includes: High-risk fire alarm data is assigned the highest priority and critical identifiers to ensure zero-latency and highly reliable transmission; Moderate fire alarm data is assigned medium priority and secondary criticality labels, allowing for moderate delays to ensure highly reliable transmission; Low-level fire alarm data is assigned the lowest priority and is not critical, allowing for greater latency and can be transmitted or aggregated when the network is idle. 4.The forest fire prevention and early warning method based on the Internet of Things according to claim 1, characterized in that, The monitoring of network congestion parameters of the wireless communication network and the establishment of a network congestion assessment model include: Periodically collect data forwarding volume, buffer occupancy rate, and link throughput of relay nodes and aggregation nodes in the wireless communication network; The collected parameters are input into a deep learning model, which predicts short-term congestion trends and levels based on historical network data. Based on the prediction results, a numerical network congestion index is generated to drive dynamic adjustments to the scheduling strategy. 5.The forest fire prevention and early warning method based on Internet of Things according to claim 1, characterized in that, The dynamically generated data transmission scheduling strategy for each sensor node includes: Based on the fire alarm level, priority, critical indicators, and network congestion index, calculate the real-time transmission requirements of each sensor node; Based on demand and available transmission resources, a multi-objective optimization algorithm is adopted to generate a scheduling strategy, taking into account real-time performance, reliability and load balancing. The strategy includes the recommended transmission rate, maximum number of retransmissions, preferred transmission path, and next transmission timing for each sensor node; Transmission path selection is based on the principle of shortest path or least congestion. 6.The forest fire prevention and early warning method based on the Internet of Things according to claim 1, characterized in that, The dynamic adjustment of the transmission rate, retransmission count, transmission path, and transmission timing of each sensor node includes: For high-priority alarm data, set the rate to the maximum and set a high retransmission limit to ensure data arrival; For medium-priority alarm data, the rate is adjusted to medium, allowing a reasonable amount of retransmission to balance real-time performance and load. For low-priority data, if network congestion is high, the transmission rate will be reduced to the minimum or buffered and delayed. Choose the least congested or most bandwidth-efficient link as the transmission path, adjust the sending time, and avoid data packet contention. 7.The forest fire prevention and warning method based on Internet of Things according to claim 1, characterized in that, The method of ensuring real-time and reliable transmission of high-priority alarm data during network congestion includes: For high-priority alarm data, forward error correction coding or redundant multipath transmission is used to resist link interruption and data loss. The system uses a retransmission confirmation mechanism to monitor the transmission status in real time. If the transmission fails or times out, it will immediately retransmit or switch to an alternative path. Reserve dedicated bandwidth or time slices for high-priority data packets so that transmission is not affected by low-priority data; Real-time transmission sets a strict latency limit, and when the predicted latency exceeds the limit, the system issues an early warning and adjusts its strategy. 8.The forest fire prevention and warning method based on the Internet of Things according to claim 1, characterized in that, The suppression or delay of low-priority data includes: When the network congestion index reaches the threshold, low-priority packets are marked and instructed to be cached locally by nodes until the congestion is relieved. For specific low-priority data, such as routine environmental monitoring, aggregation and periodic transmission are performed when the network is under high load to reduce channel contention and overhead; In cases of extreme congestion, based on the data discarding strategy, some of the least critical and low-priority data packets are selectively discarded to release resources; Suppression and delay strategies are dynamically adjusted based on fire alarm levels and congestion indices. 9.The forest fire prevention and warning method based on the Internet of Things according to claim 1, characterized in that, The method further includes: Transmit high-priority alarm data to the remote command center; Based on alarm data, combined with forest area maps and meteorological information, the remote command center generates a fire spread model and firefighting plan; After receiving alarm data, the mobile terminal automatically triggers an emergency notification, displaying the location of the fire and evacuation routes; After the data transmission is completed, the system sends a confirmation message to the administrator confirming the successful transmission.

10. The forest fire prevention and warning system based on the Internet of Things, applied to the forest fire prevention and warning method based on the Internet of Things in claims 1-9, characterized in that, include: The data acquisition module is used to acquire environmental parameter information and potential fire alarm data collected by each sensor node in real time; The fire assessment module is used to assess the fire alarm level based on potential fire alarm data, preset fire discrimination models and geographical information, and to assign transmission priority and key identifiers to the alarm data. The network monitoring module is used to monitor network congestion parameters of the wireless communication network in real time, including bandwidth utilization, packet loss rate and latency, and to establish a congestion assessment model. The transmission strategy generation module is used to dynamically generate data transmission scheduling strategies for each sensor node based on parameters obtained from fire alarm level, transmission priority, critical identifiers and congestion assessment model. The data transmission control module is used to dynamically adjust the transmission rate, retransmission count, transmission path and transmission timing of each sensor node according to the scheduling strategy, so that high-priority fire alarm data can be transmitted in real time and reliably when the network is congested, and low-priority data can be suppressed or delayed.

Images