A multi-parameter linkage automatic control system for forest fire monitoring

Abstract

This invention belongs to the field of forest fire monitoring technology, specifically a multi-parameter linkage automatic control system for forest fire monitoring, comprising: a first sensor for monitoring the rate of change of litter moisture content; a second sensor for monitoring the rate of change of soil electrical conductivity; a third sensor for monitoring the shift value of soil pH; a positioning module for acquiring the geographical coordinates of the monitoring point; and a storage module for recording event data. This invention, by monitoring the rate of change of litter moisture content, the rate of change of soil electrical conductivity, and the shift value of soil pH, and employing a dual-locking mechanism consisting of time-series window verification and chemical verification, solves the technical problems of existing forest fire monitoring systems that struggle to simultaneously achieve very early warning and low false alarm rates, and that cannot eliminate the reliability of judgments due to transient interference.

Description

Technical Field

[0001] This invention belongs to the field of forest fire monitoring technology, specifically a multi-parameter linkage automatic control system for forest fire monitoring. Background Technology

[0002] Forest fires are characterized by their suddenness and rapid spread, making early warning crucial for minimizing losses. Human activities such as ancestor worship and burning paper money at graves are common in forested areas, and the initial fires caused by these activities often begin before any open flame is visible, making them difficult to detect in time using traditional monitoring methods.

[0003] Currently, various technical solutions have been proposed in the field of forest fire monitoring; For example, Chinese patent CN108335451A discloses an automatic alarm linkage system for computer room fires, which senses temperature and smoke through fire sensor components and triggers an alarm when the temperature and smoke exceed a threshold. Chinese patent CN224163996U discloses a forest fire monitoring system. This system achieves all-weather, full-coverage perception of fire conditions by constructing a three-dimensional monitoring network that includes satellites, checkpoints, video, drones, and environmental monitoring. Chinese patent CN117576848A discloses a forest fire early warning method, which screens the credibility of multi-parameter data by determining monitoring default decision factors, and then judges the fire risk coefficient. In addition, Chinese patent CN117152893B discloses a forest disaster prevention method and system, which uses deep learning algorithms and remote sensing technology to capture key parameters such as heat sources, temperature, and humidity, and identifies abnormal temperature distributions through convolutional neural networks.

[0004] However, the aforementioned existing technologies still have the following problems: 1. Existing technologies mostly rely on parameters such as smoke, temperature, and heat sources, which can only be effectively detected when the fire develops to the open flame stage. In the very early stage of a fire (smoldering stage), surface debris has already begun to lose water at an accelerated rate, but existing technologies, which use absolute humidity values ​​or comprehensive indices as the basis for judgment, cannot capture this early signal, resulting in delayed warnings.

[0005] 2. In forest environments, fertilization can increase soil electrical conductivity, acid rain can cause pH fluctuations, and dew evaporation can cause humidity changes. Existing technologies use multi-parameter weighting or confidence factor screening, which essentially still rely on whether each parameter exceeds the threshold simultaneously. However, in real fire situations, there is a time lag between the changes of different parameters (such as the increase in electrical conductivity lagging behind the decrease in water content). Simply judging whether the threshold is exceeded simultaneously cannot distinguish between real fire situations and disturbances based on time-series characteristics.

[0006] 3. Existing technologies directly output alarms or perform confidence calculations after parameters exceed thresholds, without recording the duration of the parameter exceeding the threshold; the system cannot distinguish between instantaneous fluctuations (such as a momentary drop in pH caused by brief acid rain) and continuous changes (such as a continuous shift in pH caused by combustion), and is prone to false alarms due to instantaneous interference.

[0007] Therefore, the present invention provides a multi-parameter linkage automatic control system for forest fire monitoring. Summary of the Invention

[0008] In order to make up for the shortcomings of existing technologies and solve the technical problems that existing forest fire monitoring systems cannot achieve both very early warning and low false alarm rate, and cannot eliminate the problem of insufficient reliability of judgment due to transient interference.

[0009] The technical solution adopted by this invention to solve its technical problem is: a multi-parameter linkage automatic control system for forest fire monitoring, comprising: The first sensor is used to monitor the rate of change in the moisture content of litter; The second sensor is used to monitor the rate of change in soil electrical conductivity; The third sensor is used to monitor the shift in soil pH. The positioning module is used to obtain the geographical coordinates of the monitoring points; The storage module is used to record event data; The processor is connected to the first sensor, the second sensor, the third sensor, the positioning module, and the storage module, respectively, and the processor is configured to: (1) When the rate of change of the moisture content of the litter exceeds the first threshold, the first trigger time is recorded and the high-frequency monitoring mode is entered; (2) Within a preset time window after the first trigger time, determine whether the rate of change of the soil electrical conductivity exceeds a second threshold: If the condition is exceeded, the timing verification condition is determined to be met, and the process proceeds to step (3). If the time limit is not exceeded, the process will terminate. (3) After the time-series verification conditions are met, determine whether the soil pH offset exceeds the third threshold: If the threshold is exceeded, the chemical verification conditions are deemed met, and the combustion event is recorded in the storage module. The combustion event includes the geographic location coordinates, the current timestamp, the offset value of the soil pH, and the duration for which the offset value exceeds the third threshold. (4) When the chemical verification conditions are met, output automatic control instructions.

[0010] Furthermore, in the high-frequency monitoring mode, the processor increases the sampling frequency of the second sensor.

[0011] Preferably, the preset time window is 3 to 8 minutes.

[0012] Preferably, the processor is further configured to establish historical baselines for the rate of change of litter moisture content, the rate of change of soil electrical conductivity, and the offset of soil pH, wherein the first threshold, the second threshold, and the third threshold are adaptively adjusted according to the dynamic changes of the historical baselines.

[0013] Preferably, the processor is further configured to: continuously monitor the offset value of the soil pH after the chemical verification conditions are met; when the offset value exceeds the third threshold for more than 72 hours and the rate of change of the soil electrical conductivity does not return to the historical baseline, determine the reignition risk state and output a reignition warning command.

[0014] Preferably, the processor is further configured to: If the rate of change of soil electrical conductivity in step (2) does not exceed the second threshold, a dryness warning signal is output; If the soil pH offset value in step (3) does not exceed the third threshold, output a confirmation signal; Based on the dryness warning signal or the pending confirmation signal, the drone is automatically dispatched to conduct on-site verification.

[0015] Preferably, the automatic control commands include at least one of the following: GIS map marking, alarm information transmission, automatic drone dispatch, and reporting to the fire command center.

[0016] Preferably, it also includes a barcode scanning and tracing module for identifying the identification information of any component in the system; the storage module is also used to associate and store the combustion event with the identification information to form a traceable combustion event report.

[0017] Preferably, the processor is further configured to: The first and second sensors are controlled to operate at a first sampling frequency in normal monitoring mode, and the third sensor operates at a frequency lower than the first sampling frequency in normal monitoring mode. In the high-frequency monitoring mode, the sampling frequency of the second sensor is increased from the first sampling frequency to the third sampling frequency; When the combustion event is stored and the soil pH offset returns to the historical baseline, the sampling frequency of the second sensor is automatically restored to the first sampling frequency.

[0018] The beneficial effects of this invention are as follows: 1. This invention, by monitoring the rate of change of moisture content in litter rather than its absolute value, can trigger the first-level judgment at the smoldering stage of a fire. Compared with traditional solutions that rely on the absolute values ​​of smoke, temperature, or humidity, it can issue an early warning several minutes to tens of minutes in advance, buying valuable time for firefighting.

[0019] 2. This invention employs a dual-locking mechanism of time-series verification and chemical verification: first, the rate of change in litter moisture content is required to exceed a threshold; then, the rate of change in soil electrical conductivity is verified within a preset time window; finally, the soil pH shift value is confirmed to exceed a threshold and its duration is recorded. This mechanism strictly follows the physical-chemical causal chain of fire occurrence and can effectively eliminate parameter fluctuations caused by non-fire factors such as dew evaporation, animal activity, fertilization, and acid rain, significantly reducing the false alarm rate.

[0020] 3. This invention achieves automatic switching between two monitoring modes: normal and high frequency. Under normal conditions, the sensor operates at a lower frequency to reduce power consumption. When the rate of change in the moisture content of litter triggers the first threshold, it automatically enters the high frequency monitoring mode, increasing the sampling frequency of the second sensor to accurately capture subsequent changes in conductivity. This mechanism ensures both the monitoring response speed within the critical time window and avoids the energy waste caused by long-term high-frequency sampling, making it suitable for long-term field deployment. Attached Figure Description

[0021] The invention will now be further described with reference to the accompanying drawings.

[0022] Figure 1 This is a system structure block diagram of the present invention; Figure 2 This is a flowchart of the forest fire monitoring and judgment process of the present invention. Detailed Implementation

[0023] I. Definition of Key Terms To facilitate understanding of the technical solution of this invention, the following key terms are first defined: Rate of change of moisture content in litter: This refers to the change in moisture content of litter per unit time. The calculation formula is: Rate of change = (Current measurement value - Previous measurement value) / Time interval. A negative rate of change indicates a decrease in moisture content, while a positive value indicates an increase in moisture content. This invention focuses on scenarios where moisture content decreases rapidly (i.e., a large absolute value of the negative rate of change).

[0024] The rate of change of soil electrical conductivity refers to the amount of change in soil electrical conductivity per unit time. The calculation formula is: Rate of change = (Current measurement value - Previous measurement value) / Time interval. A positive rate of change indicates an increase in electrical conductivity.

[0025] Soil pH offset: This refers to the absolute difference between the current soil pH measurement and the historical baseline value. The calculation formula is: Offset = |Current pH - Historical Baseline pH|. The historical baseline pH refers to the average pH value over a past period (e.g., 7 days) at the same time.

[0026] Preset time window: refers to a fixed time interval calculated from the first trigger time, used to limit the judgment range of timing verification conditions.

[0027] II. Application Scenarios This invention is applicable to planted forests and natural forest reserves in arid and semi-arid regions. Taking a forest area in the eastern Loess Plateau as an example, this region receives approximately 400-500 mm of annual precipitation, has high evaporation rates, and its vegetation is dominated by Pinus tabuliformis and Robinia pseudoacacia, resulting in a thick layer of litter under the trees. Around the Qingming Festival and during the autumn harvest season, local customs such as burning paper money for ancestral worship, visiting graves, and using fire for agricultural purposes create high-risk periods for forest fires. Traditional monitoring methods in this region suffer from problems such as delayed response and high false alarm rates. This invention provides an optimized design for such scenarios.

[0028] III. Example 1 (Conventional Forest Area - Qingming Festival Ritual Scene) System configuration and parameter settings in Example 1 This embodiment was deployed in a state-owned forest farm in the eastern part of the Loess Plateau. The forest farm covers an area of ​​approximately 20,000 mu (about 1333 hectares), with the main tree species being Pinus tabuliformis (Chinese pine). The litter layer under the trees is 5-8 centimeters thick. Around the Qingming Festival each year, villagers in the surrounding area have the custom of burning paper money at graves, which is a high-risk period for fires.

[0029] The system in this embodiment includes: a first sensor (litter moisture content sensor), a second sensor (soil conductivity sensor), a third sensor (soil pH sensor), a GPS positioning module, a storage module, and a processor. Each sensor, positioning module, and storage module is connected to the processor.

[0030] In this embodiment, the sampling parameters of each sensor are set as shown in Table 1: Table 1 System parameter setting table for Example 1 First sensor sampling interval 60 seconds Under normal monitoring mode The second sensor's normal sampling interval 60 seconds Under normal monitoring mode Second sensor high-frequency sampling interval 10 seconds After entering high-frequency monitoring mode Third sensor sampling interval 5 minutes Always sample at a low frequency First threshold -0.3% / minute The absolute value of the rate of decrease in the moisture content of litter exceeds 0.3% / minute. Second threshold +0.15 mS / cm / minute The rate of increase in soil electrical conductivity exceeded 0.15 mS / cm / min. Third threshold 0.4 pH offset value exceeds 0.4 Preset time window 5 minutes Within 5 minutes after the first trigger time The specific execution process of the processor's judgment logic—taking the burning of paper money during the Qingming Festival as an example. Scene Description: On the morning of April 4, 2025 (Qingming Festival), local villagers were burning paper money as a form of worship at their ancestral graves in a pine forest. The burning paper money ignited nearby debris, and the fire started to develop from a smoldering stage.

[0031] Step (1): The processor continuously receives litter moisture content data collected by the first sensor. The normal moisture content of litter in this forest area is 15%-20%. At 10:00:00 AM, the first sensor measured a moisture content of 18.0%; 60 seconds later, the moisture content was measured as 17.5%, with a change rate of (17.5% - 18.0%) / 60 seconds = -0.5% / minute. Since -0.5% / minute < -0.3% / minute (first threshold), the processor determines that the first-level trigger condition is met, records 10:00:00 as the first trigger time, and enters the high-frequency monitoring mode—adjusting the sampling interval of the second sensor from 60 seconds to 10 seconds.

[0032] Step (2): Within a 5-minute preset time window after 10:00:00, the processor monitors the data from the second sensor at a frequency of 10 seconds per time. The burning of paper money during the sacrificial ceremony causes the surface temperature of the soil to rise, accelerates water evaporation, and increases the concentration of soluble salts in the soil. At 2 minutes and 30 seconds after 10:00:00, the conductivity measured by the second sensor is 0.12 mS / cm; 10 seconds later, it is measured as 0.125 mS / cm. The rate of change is calculated as (0.125 - 0.12) / 10 seconds = 0.0005 mS / cm / second = 0.03 mS / cm / minute. Since +0.03 mS / cm / minute < +0.15 mS / cm / minute (the second threshold), the processor determines that the conditions for entering step (3) are not met and continues to maintain high-frequency monitoring until the 5-minute window ends. As a control, if the timing verification conditions are met, the system enters step (3) for chemical verification, and the specific process is as follows: Step (3): Assuming the timing verification conditions are met, the processor reads the data from the third sensor. The historical baseline pH of the soil in this forest area is 7.2 (loess is slightly alkaline). Ash and organic acids from combustion seep into the soil, and the current pH is 6.6, with an offset of |6.6 - 7.2| = 0.6. Since 0.6 > 0.4 (the third threshold), the processor determines that the chemical verification conditions are met. The processor also records the duration for which the pH offset exceeds 0.4: the timer starts from when the pH first drops below 6.8 (offset of 0.4) and ends when the pH rises back above 6.8, with this duration being 43 minutes.

[0033] The processor records the combustion event in the storage module, including: geographic coordinates (112.3°E, 37.8°N), current timestamp (April 4, 2025, 10:02:30), pH offset (0.6), and duration exceeding the third threshold (43 minutes).

[0034] Step (4): The processor outputs automatic control instructions, including: marking the coordinate location in the geographic information system (GIS), sending alarm information to the forest fire prevention command center, and automatically dispatching drones to the site for verification.

[0035] Actual results: The drone arrived at the scene at 10:08 AM and confirmed that the fire was in the smoldering stage with a burned area of ​​approximately 2 square meters. Forest fire prevention team members arrived at the scene at 10:15 AM and successfully extinguished the fire, preventing a forest fire. The entire process, from triggering the warning to extinguishing the fire, took only 15 minutes.

[0036] Experimental data verification During peak periods of ancestor worship such as Qingming Festival, Cold Clothes Festival, and Spring Festival, the system of Example 1 was subjected to a year-long follow-up test, and the results are shown in Table 2: Table 2 Statistical Table of Tracking Test Results in Example 1 Fire caused by burning paper offerings during ancestral worship 23 times 22 times 95.7% Non-fire interference elimination (dew evaporation) 18 times 17 times (Only 1 false alarm) 94.4% Elimination of non-fire-related disturbances (animal activity) 12 times 12 times (0 false alarms) 100% Compared with the single-threshold humidity alarm system previously used locally, Example 1 provides an average warning time that is about 7 minutes earlier and reduces the false alarm rate by about 75%.

[0037] IV. Example 2 (High Fire Risk Area - Spring Burning Scenario) System configuration and parameter settings in Example 2 This embodiment was deployed in a hilly agricultural area in the eastern Loess Plateau. The terrain in this area is fragmented, with farmland and forest interspersed. Every spring (March to April), local farmers have a tradition of burning straw and stubble to prepare for planting, which poses an extremely high risk of fire. The climate in this area is dry, with an average relative humidity of only 40%-50% in spring, and the moisture content of surface litter is usually below 10% year-round.

[0038] The difference between this embodiment and Embodiment 1 is that, in response to the high fire risk and rapid response requirements of this area, some parameters have been adjusted (see Table 3 for a comparison of specific parameters) to improve the sensitivity of the system.

[0039] Table 3 Comparison of parameters between Example 1 and Example 2

[0040] First sensor sampling interval 60 seconds 30 seconds Increase data density to capture changes faster The second sensor's normal sampling interval 60 seconds 60 seconds Consistent with Example 1 Second sensor high-frequency sampling interval 10 seconds 5 seconds Improve capture accuracy within the time window First threshold -0.3% / minute -0.2% / minute Lowering the trigger threshold and providing earlier warnings Second threshold +0.15 mS / cm / minute +0.10 mS / cm / minute Adapting to the characteristics of conductivity variation in dry areas Third threshold 0.4 0.3 Increase sensitivity to pH changes Preset time window 5 minutes 4 minutes Dry areas respond faster and have shorter windows. Taking the specific execution process of the processor's judgment logic as an example, which involves the spread of spring burning into forest areas... Scene Description: On the afternoon of March 15, 2025, a villager burned straw in his farmland. The fire spread to the nearby locust forest due to strong winds.

[0041] Step (1): The processor monitors the moisture content of litter at a frequency of 30 seconds per measurement. The normal moisture content of litter in this forest area is only 8%-10%. At 14:00:00, the moisture content was 9.2%; 30 seconds later, it was measured at 8.8%, with a change rate of -0.8% / minute (exceeding the first threshold of -0.2% / minute). The processor records the first trigger time as 14:00:00, enters high-frequency monitoring mode, and adjusts the sampling interval of the second sensor from the normal 60 seconds per measurement to 5 seconds per measurement.

[0042] Step (2): Within a 4-minute time window after 14:00:00, the processor monitors the soil electrical conductivity at a frequency of 5 seconds per time. At 1 minute and 15 seconds after 14:00:00, the rate of change in electrical conductivity first exceeds +0.10 mS / cm / minute, and the processor determines that the timing verification condition is met.

[0043] Step (3): The processor reads data from the third sensor. The historical baseline pH is 7.0, the current pH is 6.5, and the offset is 0.5 (exceeding the third threshold of 0.3). The processor records the pH offset for 28 minutes. After recording the combustion event, an automatic control command is output.

[0044] Step (4): The processor outputs a reignition warning command: After the fire is extinguished, the processor continuously monitors the pH offset value. In this embodiment, the pH offset value remained above 0.3 for 78 hours, and the soil conductivity change rate did not return to the historical baseline. The processor determined that the fire was at risk of reignition and output a reignition warning command. The fire prevention team returned to the scene according to the warning command, found smoldering points underground, and dealt with them in a timely manner, thus preventing a reignition accident.

[0045] Comparison of experimental data The early warning performance of Example 1 and Example 2 was compared under the same fire scenario, and the results are shown in Table 4: Table 4 Comparison of Early Warning Performance between Example 1 and Example 2 Average warning time (calculated from the start of smoldering) 2 minutes and 15 seconds 1 minute 20 seconds Example 2: 55 seconds in advance False alarm rate (10 tests without fire interference) 1 false alarm 2 false alarms Example 2: High sensitivity but slightly increased false alarm rate Reignition identification accuracy 88% (7 out of 8 tests were successful) 93% (14 out of 15 tests were successful) Example 2 is better V. Example 3 (Grave Visit and Ancestral Worship in High-Interference Areas + Areas with Dense Human Activity) System configuration and parameter settings in Example 3 This embodiment is deployed in a coal-rich area in the eastern Loess Plateau. Besides traditional sacrificial activities such as burning paper money and visiting ancestral graves, this area also experiences human interference from mining and transportation. The area contains numerous ancient tombs dating back to the Ming and Qing dynasties, and local villagers frequently engage in sacrificial activities, with peak periods occurring around the first day of the tenth lunar month (Cold Clothes Festival), the Spring Festival, and Qingming Festival. Furthermore, the soil in this area exhibits some degree of pH abnormality due to coal mining activities, and there are numerous non-fire-related disturbance factors.

[0046] This embodiment focuses on optimizing the false alarm rejection capability and traceability function. The optimized parameter settings are shown in Table 5.

[0047] Table 5 System Parameter Optimization Settings for Example 3

[0048] First threshold -0.3% / minute -0.5% / minute Raise the trigger threshold to avoid false alarms caused by human intervention. Second threshold +0.15 mS / cm / minute +0.20 mS / cm / minute Raising the threshold for changes in conductivity Third threshold 0.4 0.6 Increase pH offset threshold Preset time window 5 minutes 7 minutes The regional soil response is slow, so the window needs to be extended. pH duration recording accuracy Minutes Seconds More accurately distinguish between instantaneous and continuous disturbances Scan to trace Optional Standard configuration It facilitates tracking of device performance degradation in complex environments. The specific execution process of the processor's judgment logic—with a focus on false alarm elimination. In this embodiment, the processor enhances its false alarm rejection capabilities. The following compares four scenarios: Scene A: Real Fire - Burning Paper Money at Gravesites on the Cold Clothes Festival On October 29, 2025 (the first day of the tenth lunar month, the Cold Clothes Festival), local villagers burned paper money as a form of worship in front of an ancient Ming Dynasty tomb. The burning paper money and offerings ignited surrounding debris.

[0049] Step (1): The change rate of litter moisture content is -0.7% / minute (exceeding the -0.5% threshold), triggering the process.

[0050] Step (2): Within the 7-minute window, the rate of change of conductivity reached +0.25 mS / cm / minute at 2 minutes and 30 seconds (exceeding the +0.20 threshold), and the timing verification was passed.

[0051] Step (3): The pH offset value was 0.9 (exceeding the 0.6 threshold) and the duration was recorded as 52 minutes, thus the chemical verification was successful.

[0052] The processor outputs a combustion event and automatic control commands. Firefighters arrived at the scene 12 minutes later and extinguished the initial fire, which covered an area of ​​approximately 5 square meters.

[0053] Scenario B: Non-fire interference – Dust from mining transport vehicles Coal transport vehicles travel through the area, and the dust they generate covers the surface of fallen debris, which may affect sensor readings.

[0054] Step (1): Dust causes a short-term change in the surface moisture content of fallen materials, but the change rate is -0.1% / minute (not exceeding the -0.5% threshold), so it does not trigger the system and the system continues normal monitoring.

[0055] The dust disturbance was successfully eliminated.

[0056] Scenario C: Non-fire interference - the paper offerings were burned but did not ignite (only the paper was burned, and the fallen debris did not ignite). Villagers burned paper offerings at the sacrificial site, but the area where the paper was burned was a hardened ground, and the fire source did not come into contact with the fallen paper offerings.

[0057] Step (1): The rate of change of the moisture content of litter is -0.6% / minute (exceeding the -0.5% threshold), triggering the entry into step (2).

[0058] Step (2): Since the litter was not ignited, the soil temperature did not rise significantly, and the rate of change in electrical conductivity was only +0.05 mS / cm / minute (not exceeding the +0.20 threshold). The timing verification failed, the process was terminated, and no alarm was triggered.

[0059] The system correctly ruled out the safety issue of "burning paper but not igniting it," thus avoiding unnecessary alarms.

[0060] Scenario D: Non-fire disturbance - animal activity + localized fertilization Wild animals digging or fertilizing the soil can disturb the topsoil.

[0061] Step (1): Animals digging in the soil may cause changes in the moisture content of litter. In this embodiment, the change rate is -0.4% / minute (not exceeding the -0.5% threshold), so it is not triggered.

[0062] If the rate of change exceeds the threshold, after entering step (2), the rate of change of conductivity is usually positive but the magnitude is small and it is difficult to exceed the +0.20 threshold. The timing verification fails and the process terminates.

[0063] False alarm elimination effect data The results of a 60-day field test conducted in this coal-rich area are shown in Table 6. Table 6 Statistical Table of False Alarm Elimination Effect in Example 3 Paper offerings burned during the sacrifice (but not ignited) 15 times 3 times 0 times 100% Dust from mining transportation 20 times 2 times 0 times 100% Animal activities 10 times 2 times 1 time 50% Local fertilization 8 times 1 time 0 times 100% Acid rain / dust deposition 12 times 2 times 1 time 50% total 65 times 10 times 2 times 80% Example 3 reduced the total false alarm rate from 15.4% to 3.1% by increasing the thresholds and extending the time window. The false alarm elimination effect was significant and it is particularly suitable for forest areas with dense human activities and complex interference factors.

[0064] Specific applications of QR code traceability function In this embodiment, the QR code traceability module is standardly included in the system. Taking a real fire incident as an example: Following a fire caused by the Cold Clothes Festival rituals in 2025, fire prevention personnel used a handheld terminal to scan the QR code on the casing of the first sensor at the monitoring point to obtain the sensor's identification information (serial number: SEN-2024-00888, batch number: BATCH-2408, production date: August 2024). The fire event stored in the storage module was linked to this identification information to generate a traceable report, the content of which is as follows: Time of the fire: October 29, 2025, 09:15:20 Geographical location: 112.8°E, 37.5°N (within a protected area of ​​a Ming Dynasty ancient tomb complex) Cause of the fire: Burning paper money during a sacrificial ceremony pH offset: 0.9 Duration exceeding the third threshold: 52 minutes Monitoring equipment serial number: SEN-2024-00888 Monitoring equipment batch number: BATCH-2408 Subsequent analysis revealed that the sensor with batch number BATCH-2408 exhibited pH measurement drift in humid environments. Using traceability capabilities, the management recalibrated this batch of sensors and provided performance data to the manufacturer, thus informing subsequent equipment procurement.

[0065] VI. Summary and Comparison of Parameters in Examples 1-3 (See Table 7) Table 7 Summary Comparison of Key Parameters and Effects in Examples 1-3 First sensor sampling interval 60 seconds 30 seconds 60 seconds Second sensor high-frequency sampling interval 10 seconds 5 seconds 10 seconds First threshold -0.3% / minute -0.2% / minute -0.5% / minute Second threshold +0.15 mS / cm / minute +0.10 mS / cm / minute +0.20 mS / cm / minute Third threshold 0.4 0.3 0.6 Preset time window 5 minutes 4 minutes 7 minutes pH duration recording accuracy Minutes Minutes Seconds Scan to trace Optional Optional Standard configuration Applicable Scenarios Regular forest areas / Qingming Festival Dry / burnt high-incidence areas densely populated / complexly disturbed areas Core advantages Balancing early warnings and false alarms Very early warning Low false alarm rate + traceability Average advance warning time Approximately 7 minutes Approximately 8 minutes Approximately 5 minutes False alarm rate Approximately 6-8% Approximately 8-10% Approximately 3-5% VII. Summary of Technical Results The above three embodiments were verified in conventional forest areas (Qingming Festival ancestor worship scenario), high fire risk agricultural areas (spring burning scenario), and high disturbance coal mining areas in the eastern Loess Plateau (grave-visiting and ancestor worship + human activities scenario), respectively, proving the technical effect of the present invention: Very early warning: By using the rate of change in the moisture content of fallen debris instead of its absolute value, the first-level judgment can be triggered as early as the smoldering stage of a fire. Tests in Example 2 on a scenario of stubble spreading by burning showed that the warning time can be more than 8 minutes earlier on average than traditional methods.

[0066] Low false alarm rate: Through a dual-locking mechanism of time-series verification and chemical verification, non-fire-related interferences such as unignited paper offerings, mine dust, animal activity, and fertilization are effectively eliminated. Tests in a complex interference area in Example 3 show that the total false alarm rate can be controlled below 3.1%, a reduction of approximately 80% compared to traditional methods.

[0067] High reliability of judgment: By recording the duration of pH shift values, the system can distinguish between transient disturbances (lasting for several minutes) and continuous combustion (lasting for tens of minutes to several hours). In Example 1, the fire caused by the sacrificial ceremony lasted for 43 minutes, while disturbances such as acid rain usually only last for a few minutes. The system accurately distinguishes between them by the duration.

[0068] Strong traceability: The burning event is associated with the equipment identification information and stored through the QR code traceability module, which facilitates the analysis of the cause of the fire after the event (such as tracing back to the specific sacrificial behavior and time in Example 3) and the evaluation of equipment performance (such as discovering batch calibration drift).

[0069] Adaptive balance between power consumption and response speed: By switching between normal and high-frequency monitoring modes, response speed is maintained within critical time windows while reducing long-term operating power consumption. Calculations show that in all three embodiments, the system can operate continuously for more than 6 months in the field powered by batteries.

Claims

1. A multi-parameter linkage automatic control system for forest fire monitoring, characterized in that, include: The first sensor is used to monitor the rate of change in the moisture content of litter; The second sensor is used to monitor the rate of change in soil electrical conductivity; The third sensor is used to monitor the shift in soil pH. The positioning module is used to obtain the geographical coordinates of the monitoring points; The storage module is used to record event data; The processor is connected to the first sensor, the second sensor, the third sensor, the positioning module, and the storage module, respectively, and the processor is configured to: (1) When the absolute value of the rate of change of the moisture content of the litter exceeds the first threshold, the first trigger time is recorded and the high-frequency monitoring mode is entered; (2) Within a preset time window after the first trigger time, determine whether the rate of change of the soil electrical conductivity exceeds a second threshold: If the condition is exceeded, the timing verification condition is determined to be met, and the process proceeds to step (3). If the time limit is not exceeded, the process will terminate. (3) After the time-series verification conditions are met, determine whether the soil pH offset exceeds the third threshold: If the threshold is exceeded, the chemical verification conditions are deemed met, and the combustion event is recorded in the storage module. The combustion event includes the geographic location coordinates, the current timestamp, the offset value of the soil pH, and the duration for which the offset value exceeds the third threshold. (4) When the chemical verification conditions are met, output automatic control instructions.

2. The system according to claim 1, characterized in that, The preset time window is 3 to 8 minutes.

3. The system according to claim 1, characterized in that, The processor is also used to establish historical baselines for the rate of change of litter moisture content, the rate of change of soil electrical conductivity, and the offset of soil pH, wherein the first threshold, the second threshold, and the third threshold are adaptively adjusted according to the dynamic changes of the historical baseline.

4. The system according to claim 1, characterized in that, The processor is also configured to: continuously monitor the pH deviation of the soil after the chemical verification conditions are met; when the deviation exceeds the third threshold for more than 72 hours and the rate of change of the soil conductivity does not return to the historical baseline, determine the reignition risk state and output a reignition warning command.

5. The system according to claim 1, characterized in that: The processor is also configured to: If the rate of change of soil electrical conductivity in step (2) does not exceed the second threshold, a dryness warning signal is output; If the soil pH offset value in step (3) does not exceed the third threshold, output a confirmation signal; Based on the dryness warning signal or the pending confirmation signal, the drone is automatically dispatched to conduct on-site verification.

6. The system according to claim 1, characterized in that, The automatic control commands include at least one of the following: GIS map marking, alarm information transmission, automatic drone dispatch, and reporting to the fire command center.

7. The system according to claim 1, characterized in that, It also includes a QR code traceability module for identifying the identification information of any component in the system; the storage module is also used to associate and store the combustion event with the identification information to form a traceable combustion event report.

8. The system according to claim 1, characterized in that, The processor is also configured to: The first and second sensors are controlled to operate at a first sampling frequency in normal monitoring mode, and the third sensor operates at a frequency lower than the first sampling frequency in normal monitoring mode. In the high-frequency monitoring mode, the sampling frequency of the second sensor is increased from the first sampling frequency to the third sampling frequency; When the combustion event is stored and the soil pH offset returns to the historical baseline, the sampling frequency of the second sensor is automatically restored to the first sampling frequency.

Images