Peanut flowering heat damage early warning method based on meteorological and soil factor data processing

Abstract

The application belongs to the technical field of data processing and agricultural heat damage monitoring, and particularly relates to a peanut flowering heat damage early warning method based on meteorological and soil factor data processing, meteorological data and soil data are collected; the meteorological data and the soil data are respectively preprocessed; dynamic heat tolerance threshold value calculation is performed based on the regulating effect of soil factors on peanut heat tolerance performance; real-time atmospheric temperature and dynamic heat tolerance threshold value comparison are performed, and heat damage risk index is calculated; four risk levels of low, medium, high and extremely high are divided according to the risk index, corresponding early warning signals are generated and output. Through the above technical scheme, the meteorological factors and the soil factors can be analyzed and fused, the dynamic heat tolerance threshold value can be adaptively adjusted according to the actual soil condition, the problems of early warning deficiency or excessive early warning caused by the fixed threshold value can be avoided, and the peanut variety adaptation, early warning strategy self-learning, multi-region collaborative early warning and early warning effect evaluation functions are realized.

Description

Technical Field

[0001] This invention belongs to the field of data processing and agricultural heat damage monitoring technology, and particularly relates to a method for early warning of heat damage during peanut flowering based on meteorological and soil factor data processing. Background Technology

[0002] In recent years, with the continuous expansion of planting area and the increasing demands for yield per unit area, the prevention and control of meteorological disasters in peanut production has become a key link in ensuring increased agricultural production and income. Among them, heat damage during flowering, as one of the important meteorological disasters affecting peanut yield, is characterized by its strong concealment, dynamic changes in the damage threshold, and difficulty in prediction, which seriously restricts the improvement of peanut planting benefits.

[0003] Traditional early warning systems for heat damage during flowering primarily rely on temperature monitoring data from meteorological stations and establish fixed temperature thresholds based on historical experience to assess the risk of heat damage. Specifically, when the ambient temperature exceeds a preset critical value, the system issues a heat damage warning signal, reminding farmers to take remedial measures such as irrigation and shading. This method is relatively simple in technical implementation, depending on the deployment of temperature sensors and the operation of data acquisition modules. Its core logic lies in using temperature as the sole indicator to measure the severity of heat damage.

[0004] However, with the continuous expansion of peanut planting areas and the increasing demand for precision agriculture, the aforementioned technical solutions have gradually revealed deep-seated limitations at the principle level. The occurrence of heat damage during peanut flowering is not solely determined by atmospheric temperature; the regulatory role of soil environmental factors is equally significant. Parameters such as soil moisture content, soil temperature, soil bulk density, and soil organic matter content all affect plant transpiration, root water absorption capacity, and internal water balance, thereby altering the peanut's heat tolerance threshold and the degree of damage. More importantly, there is a complex interactive coupling relationship between soil and meteorological factors: when soil moisture is good, plants can effectively alleviate high-temperature stress through transpiration; however, under drought conditions, heat damage may occur prematurely even if the air temperature does not reach the traditional threshold. This multi-factor nonlinear interaction presents a fundamental theoretical bottleneck in the accuracy and timeliness of early warning methods based on single meteorological elements.

[0005] Meanwhile, significant differences in heat tolerance exist among different soil types, planting regions, and peanut varieties, further complicating heat damage early warning systems. Fixed threshold methods struggle to adapt to these dynamically changing real-world conditions, frequently resulting in either insufficient or excessive warnings in practice. Insufficient warnings lead to remedial measures being taken only after heat damage has occurred, missing the optimal window for defense; while excessive warnings cause unnecessary waste of water resources and increased production costs, impacting the level of refined management in agricultural production.

[0006] Furthermore, existing technologies suffer from significant structural deficiencies in data acquisition. Meteorological data typically originates from discretely distributed ground monitoring stations, resulting in limited spatial resolution and making it difficult to accurately characterize micro-meteorological environmental differences at the field scale. Soil data, on the other hand, relies primarily on manual soil sampling and testing, leading to poor timeliness and failing to meet the technical requirements for real-time early warning. The monitoring systems for meteorological and soil factors are independent, and the data fusion and collaborative analysis mechanisms are still underdeveloped, which, at the technical architecture level, restricts further improvements in the accuracy of heat hazard early warning.

[0007] Therefore, this invention aims to construct a dynamic heat damage early warning method that can comprehensively consider meteorological and soil factors, fully take into account the interaction of multiple factors, and adapt to different soil conditions and planting conditions, so as to solve the current technical problems. Summary of the Invention

[0008] The purpose of this invention is to provide a method for early warning of heat damage during peanut flowering based on meteorological and soil factor data processing, in order to solve the technical problem that the early warning method based on a single meteorological element in the prior art cannot accurately characterize the impact of soil environmental factors on the occurrence of heat damage.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A method for early warning of heat damage during peanut flowering based on meteorological and soil factor data processing includes the following steps: S1: Collect data on atmospheric temperature, relative humidity, solar radiation intensity, and wind speed in the peanut planting area through the meteorological monitoring module; and simultaneously collect data on soil temperature, soil volumetric water content, soil bulk density, and soil electrical conductivity in the peanut root zone through the soil monitoring module.

[0010] The meteorological monitoring module includes an air temperature sensor, an air humidity sensor, a solar radiation sensor, and a wind speed sensor. The air temperature sensor is a platinum resistance temperature sensor with a measurement range of -40℃ to 80℃; the air humidity sensor is a capacitive humidity sensor with a measurement accuracy of ±2%RH and a measurement range of 0%RH to 100%RH; the solar radiation sensor is a silicon photodiode sensor with a measurement range of 0W / m. 2 Up to 1500W / m 2 The sensitivity is 0.5mV / (W / m 2 The wind speed sensor is a three-cup wind speed sensor with a measurement accuracy of ±0.3m / s and a measurement range of 0m / s to 40m / s.

[0011] The soil monitoring module includes a soil temperature sensor, a soil moisture content sensor, a soil bulk density sensor, and a soil conductivity sensor. The soil temperature sensor is a thermistor sensor with a measurement accuracy of ±0.2℃ and a measurement range of -20℃ to 60℃. The soil moisture content sensor is a frequency domain reflectance sensor with a measurement accuracy of ±2%Vol and a measurement range of 0%Vol to 100%Vol. The soil bulk density sensor is a gamma-ray transmission sensor with a measurement range of 0.8 g / cm³. 3 Up to 2.0 g / cm 3 The soil conductivity sensor is an AC impedance sensor with a measurement range of 0 dS / m to 10 dS / m.

[0012] S2: The collected meteorological data is denoised using a moving average filtering algorithm with a filter window width of 5 minutes. The collected soil data is outlier-removing using a median filtering algorithm. The preprocessed meteorological and soil data are precisely aligned according to their timestamps. The preprocessing also includes data integrity checks. When any sensor data is missing, linear interpolation between adjacent time points is used to fill in the missing data. When consecutive missing data exceeds 15 minutes, this time period is marked as an outlier interval and normalized using a maximum-min normalization method.

[0013] S3: Based on the regulatory effect of soil factors on peanut heat resistance, dynamic heat resistance thresholds were determined. T t The calculation of the interaction term coefficient in the dynamic heat resistance threshold calculation. β i The model was trained using historical observation data; a training dataset was constructed using records of heat damage during peanut flowering over the past five years and corresponding meteorological and soil data; the model parameters were optimized using a multiple linear regression method combined with an L2 regularization term, with the regularization coefficient set to 0.01.

[0014] S4: Acquire real-time atmospheric temperature data and compare it with dynamic heat resistance threshold. T t A comparison is performed; when the real-time atmospheric temperature exceeds the dynamic heat resistance threshold, the heat damage risk index is calculated. RI .

[0015] S5: Classify the risk level based on the heat hazard risk index calculated in step S4; the risk level is divided into four levels: the heat hazard risk index corresponding to the low risk level. RI The range is 0 to 0.5, representing the heat hazard risk index corresponding to the medium risk level. RI The range is 0.5 to 1.0, and the heat hazard risk index corresponds to the high-risk level. RIThe range is 1.0 to 1.5, representing the heat hazard risk index corresponding to the extremely high risk level. RI The range is greater than 1.5.

[0016] S6: Based on the risk level classification, generate and output corresponding warning signals. Low risk level corresponds to a green warning signal, indicating that peanut growth is normal under current conditions and no preventive measures are needed; medium risk level corresponds to a yellow warning signal, indicating that there is a risk of heat damage and it is recommended to start irrigation equipment to increase soil moisture; high risk level corresponds to an orange warning signal, indicating that the risk of heat damage is relatively high and it is recommended to start irrigation equipment and turn on shade nets; extremely high risk level corresponds to a red warning signal, indicating that the risk of heat damage is extremely high and it is recommended to take comprehensive protective measures immediately and continuously monitor.

[0017] Preferably, the specific process of step S3 is as follows: S31: Based on real-time soil temperature, calculate the adjustment correction value for peanut heat resistance, i.e., calculate the soil temperature correction factor. T s : When 25≤ T soil ≤35, T s =0.15•( T soil -30); when T soil <25, T s = -0.2; or T soil >35, T s = -0.4; in, T soil Real-time temperature of the pretreated soil; T s Value constraint: -0.4 ≤ T s ≤0.75; S32: Calculation of soil volumetric water content correction factor i : When 0.15≤ i v ≤0.35, i =0.8•( i v -0.25); when i v <0.15, θ=- 0.08; when i v >0.35, θ= 0.08; in, i v : Soil volumetric water content after pretreatment; 0.25 is the optimum soil volumetric water content during peanut flowering, and 0.15 is the critical threshold for soil drought; i Value constraint: -0.08≤ i ≤0.08; S33: Calculate the soil bulk density correction factor r b : r b =0.5•(1.3- r bulk ); in, r bulk The bulk density of the pretreated soil was 1.3 g / cm³. 3 Suitable soil bulk density for peanut growth; r b Value constraint: -0.35≤ r b ≤0.25; S34: Calculate the soil electrical conductivity correction factor EC corr : EC ≤2.0dS / m, EC corr =0; EC >2.0dS / m, EC corr =0.02·(2.0- EC ); Where EC: soil electrical conductivity after pretreatment; 2.0 dS / m is the critical value of peanut soil salt stress; EC corr Value constraint: -0.16≤ EC corr ≤0; S35: Calculate the sum of meteorological-soil interaction terms Σ β i •X i : Σ β i •X i = β 1(T a + i v )+β 2( W s + T soil})+ β 3( R s + r bulk )+ β 4( RH + EC ); in, β i The weight coefficients for the interaction term of meteorological factors are obtained through training using five years of historical data combined with multiple linear regression and L2 regularization, with a regularization coefficient of 0.01. i =1,2,3,4; X i These are the interaction terms between various meteorological factors and soil factors; S36: Calculate the basic dynamic heat resistance threshold T t0 : T t0 = T base + α 1• T s + α 2• i + α 3• r b + α 4• EC corr +Σ β i •X i ; in, T t0 Basic dynamic heat resistance threshold; T base The baseline heat tolerance threshold for peanut flowering period can be taken as 32℃. α 1. α 2. α 3. α 4 represents the weighting coefficients for soil temperature, soil volumetric water content, soil bulk density, and soil electrical conductivity.

[0018] Preferably, it also includes variety coefficient correction to calculate the final dynamic heat resistance threshold. T t : T t =T t0·K var ; in, T t Final dynamic heat resistance threshold; K var Peanut variety correction coefficient.

[0019] Preferably, the specific process of step S4 is as follows: S41: Read real-time atmospheric temperature T a and final dynamic heat resistance threshold T t ,when T a ≤ T t No risk of heat damage, RI=0; when T a > T t Proceed to step S42 for heat damage risk index calculation; S42: When T a > T t If the duration is ≤2 hours, the conventional heat damage risk index is calculated: RI = k 1•( T a -T t )+ k 2•( RH min -RH )+ k 3•( R s / R smax )+ k 4•( W smin -W s ); in, RI Indicates the risk index of heat damage; T a Real-time atmospheric temperature; T t The dynamic heat resistance threshold; RH Real-time atmospheric relative humidity; RH min This represents the lower limit of the suitable relative humidity during the peanut flowering period. R s Real-time solar radiation intensity; R smaxThis represents the maximum radiation intensity that the peanut canopy can withstand. W s Real-time wind speed; W smin This is the lower limit of the effective ventilation velocity; k 1. k 2. k 3. k 4 represents the risk weight coefficients for each factor; S43: When T a > T t If the duration is greater than 2 hours, a high-temperature time accumulation correction will be applied: C time =1+0.1·Δ t ; in, C time : Time accumulation factor; Δ t For real-time temperature exceeding T t Duration; Revised heat damage risk index RI for: RI = C time · k 1•( T a -T t )+ k 2•( RH min -RH )+ k 3•( R s / R smax )+ k 4•( W smin -W s ).

[0020] Preferably, the risk level described in step S5 is divided into four levels: The heat hazard risk index for low-risk levels ranges from 0 to 0.5; The heat hazard risk index for medium-risk levels ranges from 0.5 to 1.0; The heat hazard risk index for high-risk levels ranges from 1.0 to 1.5; The range of the heat hazard risk index corresponding to the extremely high risk level is greater than 1.5.

[0021] Preferably, after step S6, a self-learning process for the early warning strategy is also included: recording the actual damage data of peanuts after each early warning signal is issued; when peanuts actually suffer heat damage, backtracking the data of the period when peanuts actually suffer heat damage and adjusting the parameter weights in the calculation of the dynamic heat tolerance threshold; when peanuts do not suffer heat damage but the early warning signal level is high, backtracking the data and reducing the threshold parameter weights, and using an online gradient adaptive weight update method to adjust the parameter weights.

[0022] Preferably, the specific process of the early warning strategy's self-learning process is as follows: S71: After each triggering of a medium-risk or higher warning, record the corresponding data, including the warning period: t0~t1; weight: α 1~ α 4, β 1~ β 4; Data: T a , T soil , i v , r bulk , EC , RH , R s , W s Warning level: Yellow, Orange, or Red; Actual heat hazard label: Y=1: Heat hazard actually occurred; Y=0: No heat hazard occurred; S72: Determine the update direction: Y=1: Increase the weight of the corresponding soil factor and interaction term, and reduce the dynamic heat resistance threshold; Y=0, warning ≥ medium risk: reduce the weight of the corresponding soil factor and interaction term, and increase the dynamic heat resistance threshold; S73: Employs an online gradient adaptive weight update method, with a learning rate... or Weight updates are achieved by controlling the magnitude of each adjustment.

[0023] Preferably, the weight update formula in step S73 is as follows: i new =θ old +η⋅δ⋅L ( i old ); in, i new Updated weights; i old Weights before update; or Learning rate; dUpdate direction coefficients; L ( i old ): The contribution of the current weight to this event is calculated from the factor maximum-min normalization result; Soil factor weight update: α i,new =α i,old +n · d · F i / F sum ; in, α i,new Updated version i Weights of soil factors; α i,old : Before the update i Weights of soil factors; F i : No. i Normalized values ​​of soil factors during the warning period; F sum The sum of the normalized values ​​of the four soil factors; Meteorological factor interaction term weight update: β i,new = β i,old + or · d · X i / X sum ; in, β i,new Updated version i Weight of each interaction item; β i,old : Before the update i Weight of each interaction item; X i : No. i Normalized values ​​of the meteorological-soil interaction term during the warning period; X sum The sum of the normalized values ​​of the four interaction items; To avoid abnormal weights, set parameter boundary constraints: α i,new ≥0.1, β i,new ≥0.01; If the updated value is lower than the lower limit, the lower limit value will be forcibly taken.

[0024] Preferably, it also includes a multi-regional collaborative early warning process: when the heat hazard risk index of a certain region is detected to enter a high-risk level, the monitoring data of adjacent regions are automatically queried; if the dynamic heat resistance threshold of the adjacent region is significantly higher than that of the current region, regional spread warning information is added to the early warning signal of the current region.

[0025] The preferred multi-regional collaborative early warning process is as follows: When the current regional heat hazard risk level is high or extremely high, the multi-regional collaborative early warning query process is automatically initiated. Based on a preset regional adjacency table, the system automatically obtains: adjacent region IDs and location information; real-time meteorological data of adjacent regions; and dynamic heat tolerance thresholds of adjacent regions. T t邻 Current heat hazard risk index in adjacent areas RI 邻 ; Calculate the difference in dynamic heat resistance threshold Δ between regions T t : Δ T t = T t当 - T t邻 ; in, T t当 : The final dynamic heat resistance threshold of the current region; T t邻 : The final dynamic heat resistance threshold of adjacent regions; Determine if the propagation warning conditions are met: Δ T t > T th ; in, T th This is the threshold value for the difference in dynamic heat resistance thresholds between regions. T th =2.0℃; Generate and add area spread warning messages: If Δ is satisfied T t If the temperature exceeds 2.0℃, the following warning will be automatically added to the current area's warning signal: "High risk of heat damage, weather conditions are likely to spread to surrounding areas, and it is recommended that neighboring areas strengthen monitoring."

[0026] The beneficial effects of this invention include: 1. By simultaneously collecting meteorological and soil factors and performing integrated calculations, the traditional method of using only air temperature as the basis for judging heat damage is broken through. Soil temperature, soil volumetric water content, soil bulk density, and soil electrical conductivity are incorporated into the dynamic threshold model, which significantly improves the accuracy and reliability of heat damage early warning and effectively avoids insufficient or excessive early warning.

[0027] 2. A dynamic adaptive heat tolerance threshold is constructed, which can automatically adjust the critical value of heat tolerance during peanut flowering period according to the real-time soil environment in the field. Compared with the fixed threshold early warning method, it is more in line with the actual heat damage occurrence pattern under different plots, different soil moisture and different soil structures, and the early warning accuracy is higher and the applicability is stronger.

[0028] 3. By introducing a meteorological-soil interaction coupling term, the influence of the synergistic effect of multiple factors on the heat resistance of peanuts is fully considered. The model has clear physical meaning and rigorous agronomic logic, which can more realistically reflect the actual damage risk of peanuts under high temperature stress and improve the scientificity and generalization ability of the early warning model.

[0029] 4. Set peanut variety correction coefficients to adaptively correct thresholds for different varieties such as heat-resistant, intermediate, and heat-sensitive varieties, so as to achieve compatibility and adaptation of the same early warning system for multiple peanut varieties and expand the application scope of the system.

[0030] 5. Introduce a cumulative factor of high temperature duration in heat hazard risk assessment to quantify the time effect of high temperature stress, making the risk index more consistent with the actual disaster development pattern, improving the matching degree between the warning level and the actual degree of harm, and reducing misjudgment and omission.

[0031] 6. A self-learning mechanism for the early warning strategy is set up. Based on online gradient adaptive weight update, the model parameters are iteratively optimized online. The weight coefficients can be continuously corrected according to the actual occurrence of heat damage, so that the early warning results become more and more accurate over time. Long-term operation can significantly reduce the false alarm rate and the missed alarm rate.

[0032] 7. Through a multi-regional collaborative early warning mechanism and regional adjacency relationships, it enables joint judgment and spread warning of heat damage risks between plots, has regional joint prevention and control capabilities, can provide early warning of risk spread trends, and provides more comprehensive disaster prevention decision support for large-scale planting bases. Attached Figure Description

[0033] Figure 1 This is a flowchart illustrating the peanut flowering heat damage early warning method based on meteorological and soil factor data processing according to the present invention.

[0034] Figure 2 This is a schematic diagram illustrating the dynamic heat resistance threshold calculation method and heat damage risk assessment principle of the present invention. Detailed Implementation

[0035] The peanut flowering period heat damage early warning method of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments are intended to enable those skilled in the art to implement the technical solution of the present invention based on the following description, but the scope of protection of the present invention is not limited to these specific embodiments.

[0036] See appendix Figure 1 As shown, the peanut flowering period heat damage early warning method based on meteorological and soil factor data processing includes the following steps: S1: Meteorological and soil data of the peanut planting area are collected through the meteorological monitoring module. The meteorological data includes atmospheric temperature, relative humidity, solar radiation intensity and wind speed data. Soil data is collected synchronously through the soil monitoring module, including soil temperature, soil volumetric water content, soil bulk density and soil electrical conductivity data of the peanut root zone soil.

[0037] During the data collection process, the meteorological monitoring module was positioned 1.5m to 2.5m above the peanut canopy in the peanut planting area, preferably at a height of 2.0m. This height was chosen because meteorological data at 2.0m above the peanut canopy accurately reflects the atmospheric environment of the peanut growing area. Too high a height would result in significant discrepancies between the collected data and the actual peanut growing environment, while too low a height would lead to measurement blind spots due to canopy obstruction. The meteorological monitoring module comprises four sub-units: an air temperature sensor, an air humidity sensor, a solar radiation sensor, and a wind speed sensor. The air temperature sensor uses a platinum resistance temperature sensor with a measurement accuracy of ±0.1℃ and a measurement range of -40℃ to 80℃. This accuracy meets the requirements for monitoring peanut flowering temperature, as the variation in the peanut flowering heat tolerance threshold is typically within the range of 1 to 2℃. Higher measurement accuracy ensures the accuracy of dynamic heat tolerance threshold calculation. The air humidity sensor uses a capacitive humidity sensor with a measurement accuracy of ±2%RH and a measurement range of 0%RH to 100%RH. Capacitive humidity sensors are characterized by fast response and good stability, and can reflect changes in atmospheric humidity in real time. The solar radiation sensor uses a silicon photodiode sensor with a measurement range of 0W / m². 2 Up to 1500W / m 2 The sensitivity is 0.5mV / (W / m 2 Silicon photodiode sensors have the advantages of wide spectral response range and good linearity, enabling accurate measurement of the solar radiation intensity received by the peanut canopy. Three-cup anemometers are used for wind speed measurement, with a measurement accuracy of ±0.3 m / s and a measurement range of 0 m / s to 40 m / s. The three-cup anemometer has a simple structure, high reliability, and can operate stably for long periods in the complex environment of peanut growing areas.

[0038] The sensor probes of the soil monitoring module are buried at depths of 5cm to 25cm below the soil surface. Specifically, the soil temperature sensor is buried at a depth of 10cm, the soil moisture sensor at 15cm, the soil bulk density sensor at 20cm, and the soil conductivity sensor at 25cm. The selection of these depths is based on the fact that the main distribution area of ​​peanut roots is concentrated within the 5cm to 25cm depth range, with 10cm to 15cm being the dense root zone and 20cm to 25cm being the deep root distribution zone. Deploying sensors at these depths accurately reflects the environmental conditions of the soil in the peanut root zone. The soil temperature sensor uses a thermistor sensor with a measurement accuracy of ±0.2℃ and a measurement range of -20℃ to 60℃. Thermistor sensors are characterized by their small size and fast response, enabling rapid response to changes in soil temperature. The soil moisture sensor employs a frequency domain reflectance (FDR) sensor, with a measurement accuracy of ±2%Vol and a measurement range of 0%Vol to 100%Vol. The FDR sensor infers soil moisture content by measuring the soil's dielectric constant, offering advantages such as high accuracy and insensitivity to soil salinity. The soil bulk density sensor uses a gamma-ray transmission sensor, with a measurement accuracy of ±0.02 g / cm³. 3 The measurement range is 0.8 g / cm³. 3 Up to 2.0 g / cm 3 The gamma-ray transmission sensor calculates soil bulk density by measuring the attenuation of gamma rays in the soil, enabling in-situ online monitoring of soil bulk density. The soil conductivity sensor uses an AC impedance sensor with a measurement accuracy of ±5%EC and a measurement range of 0 dS / m to 10 dS / m. The AC impedance sensor calculates soil conductivity by measuring the AC impedance of the soil, reflecting the soil's salinity.

[0039] In the specific implementation of data acquisition, the sampling frequency of the meteorological monitoring module and the soil monitoring module was set to once per minute, that is, to collect readings from each sensor every 60 seconds. This sampling frequency setting comprehensively considers the requirements of data volume and early warning timeliness: a sampling frequency that is too low will lead to early warning delays, failing to meet the needs of real-time early warning; a sampling frequency that is too high will generate a large amount of redundant data, increasing the burden of data storage and processing. A sampling frequency of once per minute ensures timely capture of temperature change trends during the development of heat damage in peanuts during the flowering period, while also providing a moderate amount of data for subsequent data processing and analysis.

[0040] S2: Data preprocessing stage: The collected meteorological data is denoised using a moving average filtering algorithm; the collected soil data is outlier removed using a median filtering algorithm; the preprocessed meteorological data and soil data are aligned according to timestamps.

[0041] For meteorological data, a moving average filtering algorithm is used for noise reduction. The filtering window width is set to 5 minutes, meaning that each filtering process takes the arithmetic mean of five data points (the current time plus two minutes before and after) as the filtered output. The moving average filtering algorithm can effectively remove high-frequency noise components from meteorological data while preserving the overall trend of the data. The selection of a 5-minute window width is based on the characteristics of meteorological element changes: changes in meteorological elements such as temperature, humidity, and wind speed have a certain inertia, and changes within 5 minutes are usually not too drastic, while meaningful trends may appear on time scales exceeding 5 minutes. Therefore, a 5-minute window width achieves a good balance between noise reduction and preservation of effective information.

[0042] For soil data, a median filtering algorithm is used to remove outliers. The specific implementation of the median filtering algorithm is as follows: For each soil data point, 11 data points (5 before and 5 after the original point) are sorted, and the median value is taken as the filtered output for that point. If the difference between the original data point and the filtered output exceeds three standard deviations, the data point is considered an outlier and is removed. The removal threshold of three standard deviations is determined based on the statistical characteristics of the normal distribution. Under normal distribution conditions, approximately 99.7% of the data points will fall within three standard deviations. Therefore, data points exceeding three standard deviations can be considered low-probability events, likely due to sensor malfunctions or data transmission errors.

[0043] The data preprocessing stage also includes data integrity verification. When any sensor data is missing, linear interpolation between adjacent time points is used to fill in the missing data. For example, if soil temperature data is missing at a certain moment, the arithmetic mean of the soil temperature data measured before and after that moment is used as the filling value. Linear interpolation is suitable for cases where the missing data time is short, and it can approximate the true value as closely as possible while ensuring data integrity. When consecutive missing data exceeds 15 minutes, this time period is marked as a data anomaly interval, and a conservative estimation strategy is adopted in subsequent risk assessments. The specific implementation of the conservative estimation strategy is as follows: within the data anomaly interval, the dynamic heat resistance threshold is reduced by 0.5℃, and the calculated result of the heat damage risk index is adjusted upward by 0.2 to ensure that the early warning system does not miss potential heat damage risks due to missing data.

[0044] The preprocessed meteorological and soil data are precisely aligned according to their timestamps, with the time alignment error controlled within 1 second. The time alignment is implemented as follows: both the meteorological and soil monitoring modules use a unified time server for synchronization, achieving a time synchronization accuracy better than 1 second. During the data preprocessing stage, the system sorts and aligns the data according to the timestamps of each sensor, ensuring accurate matching of meteorological and soil data at the same time point, thus providing a data foundation for subsequent dynamic heat resistance threshold calculations.

[0045] S3: Dynamic heat resistance threshold calculation: Based on the regulatory effect of soil factors on peanut heat resistance, a dynamic heat resistance threshold is calculated. T t For calculations, see Figure 2 As shown, the specific process is as follows: S31: Based on real-time soil temperature, calculate the adjustment correction value for peanut heat resistance, i.e., calculate the soil temperature correction factor. T s : When 25≤ T soil ≤35, T s =0.15•( T soil -30); when T soil <25, T s = -0.2; or T soil >35, T s = -0.4; in, T soil : Real-time soil temperature after pretreatment; 25, 35: Upper and lower limits of suitable soil temperature for peanut roots; 30: Midpoint of optimal soil temperature for peanut roots; T s Value constraint: -0.4 ≤ T s ≤0.75; S32: Based on real-time soil volumetric water content, calculate the correction value of soil moisture to the heat tolerance threshold, i.e., calculate the soil volumetric water content correction factor. i : When 0.15≤ i v ≤0.35, i =0.8•( i v -0.25); when iv <0.15, θ=- 0.08; when i v >0.35, θ= 0.08; in, i v : Soil volumetric moisture content after pretreatment; 0.25: Optimal soil volumetric moisture content during peanut flowering; 0.15: Lower limit threshold for soil drought; i Value constraint: -0.08≤ i ≤0.08; S33: Based on real-time soil bulk density, calculate the correction value of soil compaction to heat resistance, i.e., calculate the soil bulk density correction factor. r b : r b =0.5•(1.3- r bulk ); in, r bulk 1.3: Soil bulk density after pretreatment; Suitable soil bulk density for peanut growth; r b Value constraint: -0.35≤ r b ≤0.25; S34: Based on real-time soil electrical conductivity, calculate the correction value of salinity on heat resistance, i.e., calculate the soil electrical conductivity correction factor. EC corr : EC ≤2.0dS / m, EC corr =0; EC >2.0dS / m, EC corr =0.02·(2.0- EC ); Where EC: soil electrical conductivity after pretreatment; 2.0 dS / m is the critical value of peanut soil salt stress; EC corr Value constraint: -0.16≤ EC corr ≤0; S35: Taking into account the coupling effect of meteorological and soil factors, calculate the sum Σ of meteorological-soil interaction terms. β i •X i : Σ βi •X i = β 1(T a + i v )+ β 2( W s + T soil})+ β 3( R s + r bulk )+ β 4( RH + EC ); in, β i The weight coefficients for the interaction term of meteorological factors are obtained through training using five years of historical data combined with multiple linear regression and L2 regularization, with a regularization coefficient of 0.01. i =1,2,3,4; X i These are the interaction terms between various meteorological factors and soil factors; S36: By integrating the baseline value, four soil correction terms, and interaction terms, the uncorrected baseline threshold is obtained, which is used to calculate the dynamic heat resistance threshold of the baseline. T t0 : T t0 = T base + α 1• T s + α 2• i + α 3• r b + α 4• EC corr +Σ β i •X i .

[0046] in, T t0 Basic dynamic heat resistance threshold; T base : Baseline heat tolerance threshold for peanut flowering period; α 1. α 2. α 3. α 4 represents the weighting coefficients for soil temperature, soil volumetric water content, soil bulk density, and soil electrical conductivity, which can be set as follows:α 1 = 0.8 α 2 = 1.2, α 3 = 0.5; α 4 = 0.3. β i These are the weighting coefficients of the meteorological factor interaction term. i =1,2,3,4; X i The interaction terms between various meteorological and soil factors include the interaction terms between temperature and soil moisture content, wind speed and soil temperature, radiation intensity and soil bulk density, and humidity and soil electrical conductivity.

[0047] The principle of the soil temperature correction factor in the above formula is as follows: When the soil temperature is in the range of 25℃ to 35℃, the dynamic heat resistance threshold is adjusted by 0.15℃ for every 1℃ deviation of the soil temperature from 30℃. This is because moderate soil warming is beneficial to enhancing the physiological activity of peanuts and improving their heat resistance. When the soil temperature is below 25℃, the activity of peanut roots decreases, the ability to absorb water and fertilizer decreases, and the heat resistance is correspondingly weakened. Therefore, the correction factor is set to a fixed value of -0.2. When the soil temperature is above 35℃, the high soil temperature will cause stress to the roots, which will also reduce the heat resistance of peanuts. Therefore, the correction factor is set to a fixed value of -0.4. Since the damage to peanut roots from high soil temperature is much greater than that from low temperature, the correction intensity is different.

[0048] The principle behind the soil moisture content correction factor is as follows: Soil moisture content is one of the key factors affecting the heat resistance of peanuts. Suitable soil moisture content can regulate plant temperature through transpiration, thereby improving heat resistance. When the soil moisture content is 0.25% (i.e., 25% volumetric water content), it represents the optimal soil moisture state for peanut flowering, at which point the correction factor is 0. When the soil moisture content is higher than 0.25%, the correction factor is positive, increasing the dynamic heat resistance threshold. When the soil moisture content is lower than 0.25%, the correction factor is negative, decreasing the dynamic heat resistance threshold. (Setting...) i The minimum value is 0.15, which is to prevent the dynamic heat resistance threshold calculation results from dropping unreasonably due to an excessively low correction factor when the soil is excessively dry.

[0049] The principle behind soil bulk density correction factors is as follows: soil bulk density reflects the compaction of the soil; a suitable soil bulk density (approximately 1.3 g / cm³) is... 3 This promotes root growth and water absorption, thereby improving heat resistance. When the soil bulk density is less than 1.3 g / cm³... 3 When the correction factor is positive, it can increase the dynamic heat resistance threshold; when the soil bulk density is greater than 1.3 g / cm³, it can also increase the dynamic heat resistance threshold. 3 When the correction factor is negative, the dynamic heat resistance threshold is reduced.

[0050] The principle behind the soil conductivity correction factor is as follows: Soil conductivity reflects the soil salinity. Excessively high soil conductivity makes it difficult for roots to absorb water, reducing heat resistance. The correction factor is directly proportional to soil conductivity. Soil conductivity exceeding 2.0 dS / m results in salt stress. The higher the conductivity, the larger the negative value of the correction factor, and the dynamic heat resistance threshold gradually decreases. However, since salt has almost no effect on peanuts in a low-salt environment, no correction is needed.

[0051] Weighting coefficients of various soil factors α 1. α 2. α 3. α The basis for setting 4 is that soil moisture content has the most significant impact on the heat resistance of peanuts, therefore its weighting coefficient is [missing information]. α 2 is set to the maximum value of 1.2; the influence of soil temperature is secondary, with a weighting coefficient of 1.2. α 1 is set to 0.8; the influence of soil bulk density is relatively small, and the weighting coefficient is... α 3 is set to 0.5; the influence of soil electrical conductivity is minimal, and the weighting coefficient is [missing value]. α 4 is set to 0.3.

[0052] Interaction term coefficient β i The model was trained using historical observation data. Specifically, a training dataset was constructed using records of heat damage during peanut flowering periods over the past five years, along with corresponding meteorological and soil data. This dataset contained at least 500 valid samples, each including meteorological and soil factor data at the time of heat damage, as well as the severity level of the heat damage. Multiple linear regression combined with an L2 regularization term was used to optimize the model parameters, with the regularization coefficient set to 0.01. The L2 regularization term was introduced to prevent overfitting and improve the model's generalization ability on unseen data. Through the above training process, the interaction term coefficient... β 1. β 2. β 3. β The value of 4 can be 0.12, 0.08, 0.15 and 0.06.

[0053] In another specific embodiment, the method also includes variety coefficient correction to calculate the final dynamic heat resistance threshold. T t : T t =T t0 ·K var ; in, T t Final dynamic heat resistance threshold; K varPeanut variety correction coefficients: Heat-resistant type: 1.1; Intermediate type: 1.0; Heat-sensitive type: 0.9; 0.9≤ K var ≤1.1.

[0054] The heat resistance type of peanut varieties is determined based on the performance data of the variety under historical high temperature stress conditions: if the variety has experienced more than 20 days with a cumulative maximum daily temperature exceeding 38℃ in the past five years without showing obvious heat damage symptoms, it is determined to be a heat-resistant variety; if the frequency of heat damage symptoms under similar conditions exceeds 50%, it is determined to be a heat-sensitive variety; the rest are determined to be intermediate varieties.

[0055] S4: Heat Hazard Risk Assessment Phase: Obtain real-time atmospheric temperature data and compare it with the dynamic heat resistance threshold calculated in step S3. T t A comparison is performed; when the real-time atmospheric temperature exceeds the dynamic heat resistance threshold, the heat damage risk index is calculated. RI .

[0056] In another specific embodiment, see Figure 2 As shown, the specific process of step S4 is as follows: S41: Read real-time atmospheric temperature T a and final dynamic heat resistance threshold T t ,when T a ≤ T t No risk of heat damage, RI=0; when T a > T t Proceed to step S42 for heat damage risk index calculation; S42: When T a > T t If the duration is ≤2 hours, then the conventional heat damage risk index calculation is performed. RI = k 1•( T a -T t )+ k 2•( RH min -RH )+ k 3•( R s / R smax )+ k 4•( Wsmin -W s ); in, RI Indicates the risk index of heat damage; T a Real-time atmospheric temperature; T t The dynamic heat resistance threshold; RH Real-time atmospheric relative humidity; RH min The lower limit of the suitable relative humidity during the peanut flowering period is set at 65%. R s Real-time solar radiation intensity; R smax The maximum radiation intensity that the peanut canopy can withstand is taken as 800 W / m². W s Real-time wind speed; W smin The lower limit for effective ventilation velocity is set at 2 m / s; k 1. k 2. k 3. k 4 represents the risk weight coefficients for each factor, which are set to 1.5, 0.8, 0.6, and 0.4, respectively. S43: When T a > T t If the duration is greater than 2 hours, a time accumulation factor is introduced. C time Amplifying risk, i.e., performing cumulative high-temperature time correction: C time =1+0.1·Δ t ; in, C time Time accumulation factor C time The upper limit is set to 3.0; Δ t For real-time temperature exceeding T t Duration; Revised heat damage risk index RI for: RI = C time · k 1•( T a -T t )+ k 2•( RHmin -RH )+ k 3•( R s / R smax )+ k 4•( W smin -W s ); Risk index RI constraint: 0≤RI≤2.0.

[0057] The time accumulation factor is introduced to account for the cumulative effect of heat stress over time. The degree of damage to peanuts under high-temperature stress depends not only on the current temperature level but also on the duration of the stress. High-temperature stress lasting more than 2 hours can cause irreversible damage to peanuts. Therefore, when the duration of high temperature exceeds 2 hours, the time accumulation effect needs to be incorporated into the risk index calculation to raise the warning level.

[0058] S5: Risk Level Classification Stage: Based on the heat hazard risk index calculated in step S4, risk levels are classified; the risk levels are divided into four levels: The heat hazard risk index range corresponding to the low risk level is [0, 0.5]. The heat hazard risk index range corresponding to the medium risk level is (0.5, 1.0]. The heat hazard risk index ranges from 1.0 to 1.5 for high-risk levels; The range of the heat hazard risk index corresponding to the extremely high risk level is greater than 1.5.

[0059] S6: Warning signal generation stage: Based on the risk level classification results in step S5, generate the corresponding warning signal and output it to the warning terminal; low risk level corresponds to green warning signal, medium risk level corresponds to yellow warning signal, high risk level corresponds to orange warning signal, and extremely high risk level corresponds to red warning signal.

[0060] A low-risk level corresponds to a green warning signal, indicating that peanuts are growing normally under current conditions and no preventative measures are needed; a medium-risk level corresponds to a yellow warning signal, indicating a risk of heat damage, and it is recommended to start irrigation equipment to increase soil moisture; a high-risk level corresponds to an orange warning signal, indicating a relatively high risk of heat damage, and it is recommended to start irrigation equipment and turn on shade nets; a very high-risk level corresponds to a red warning signal, indicating an extremely high risk of heat damage, and it is recommended to immediately take comprehensive protective measures and continuously monitor the situation.

[0061] Step S6 is followed by a self-learning process for the early warning strategy: recording the actual damage data of peanuts after each early warning signal is issued; when peanuts actually suffer heat damage, backtracking the data of the period when peanuts actually suffer heat damage and adjusting the parameter weights in the calculation of the dynamic heat tolerance threshold; when peanuts do not suffer heat damage but the early warning signal level is high, backtracking the data and reducing the threshold parameter weights, and using an online gradient adaptive weight update method to adjust the parameter weights.

[0062] In another specific embodiment, the self-learning process of the early warning strategy is as follows: S71: Data Recording and Sample Construction: After each triggering of a medium-risk or higher alert, record: Warning period: t0~ t1; Weight: α 1~ α 4, β 1~ β 4; data: T a , T soil , i v , r bulk , EC , RH , R s , W s ; Warning level: Yellow, Orange, or Red; Actual heat damage label: Y=1: Heat damage actually occurred; Y=0: No heat damage occurred.

[0063] The actual damage labels for peanuts are obtained through two methods: manual inspection and image recognition. Manual inspection involves technicians conducting on-site investigations of peanut fields twice a week to record the occurrence of heat damage symptoms such as yellowing, curling, and shedding of leaves. Image recognition uses multispectral cameras deployed in peanut fields to collect field images and automatically identifies heat damage symptoms through image analysis algorithms.

[0064] S72: Determine the update direction, that is, determine the direction of parameter adjustment based on the matching relationship between the warning level and the actual heat hazard: Actual heat damage occurred (Y=1): This indicates that the model threshold is too high or the weight is too low, and the early warning is too conservative. In this case, the weights of the corresponding soil factors and interaction terms should be increased, the dynamic heat tolerance threshold should be reduced, and the sensitivity should be improved. No heat damage occurred but the warning level is high (Y=0, warning ≥ medium risk): This indicates that the model threshold is too low or the weight is too high, and the warning is excessive. In this case, reduce the weight of the corresponding soil factor and interaction term, increase the dynamic heat tolerance threshold, and reduce false alarms. S73: Employs an online gradient adaptive weight update method, with a learning rate... or Weight updates are achieved by controlling the magnitude of each adjustment, thus avoiding sudden parameter changes.

[0065] The weight update formula in step S73 is as follows: i new =θ old +η⋅δ⋅L ( i old ); in, i new Updated weights; i old Weights before and after the update; the weights before and after the update include the weight coefficients of soil factors. α 1~ α 4 or the weighting coefficient of the interaction term of meteorological factors β 1~ β 4; or The learning rate is fixed at 0.05. d Update directional coefficient: Actual thermal damage Y=1: d =+1 (weighting); No heat damage occurred but a high warning level was issued: Y=0: d = -1 (weight reduction); L ( i old ): The contribution of the current weight to this event (0~1), calculated from the factor maximum-minimum normalization result; Soil factor weight update: α i,new =α i,old +n · d · F i / F sum ; in, α i,new Updated version i Weights of soil factors; α i,old : Before the update i Weights of soil factors; F i : No. i Normalized values ​​of soil factors during the warning period; F sum The sum of the normalized values ​​of the four soil factors; Meteorological factor interaction term weight update: β i,new = β i,old + or · d · X i / X sum ; in, β i,new Updated version i Weight of each interaction item; β i,old : Before the update i Weight of each interaction item; X i : No. i Normalized values ​​of the meteorological-soil interaction term during the warning period; X sum The sum of the normalized values ​​of the four interaction items; To avoid abnormal weights, set parameter boundary constraints: α i,new ≥0.1, β i,new ≥0.01; If the updated value is lower than the lower limit, the lower limit value will be forcibly taken.

[0066] It also includes a multi-regional collaborative early warning process: when the heat hazard risk index of a certain region is detected to enter a high-risk level, the monitoring data of adjacent regions are automatically queried; if the dynamic heat resistance threshold of the adjacent region is significantly higher than that of the current region, regional spread warning information is added to the early warning signal of the current region.

[0067] The multi-regional collaborative early warning process is as follows: When the current regional heat hazard risk level is high or extremely high, the multi-regional collaborative early warning query process is automatically initiated. Based on a preset regional adjacency table, the system automatically obtains: adjacent region IDs and location information; real-time meteorological data of adjacent regions; and dynamic heat resistance thresholds calculated from S3 for adjacent regions. T t邻 Current heat hazard risk index in adjacent areas RI 邻 ; Calculate the difference in dynamic heat resistance threshold Δ between regions T t : Δ T t = T t当 - T t邻 ; in, T t当 : The final dynamic heat resistance threshold of the current region; T t邻 : The final dynamic heat resistance threshold of adjacent regions; Determine if the propagation warning conditions are met: Δ T t > T th ; in, T th This is the threshold value for the difference in dynamic heat resistance thresholds between regions. T th =2.0℃; Generate and add area spread warning messages: If Δ is satisfied T t If the temperature exceeds 2.0℃, the following warning will be automatically added to the current area's warning signal: "High risk of heat damage, weather conditions are likely to spread to surrounding areas, and it is recommended that neighboring areas strengthen monitoring."

[0068] When the heat hazard risk index of a certain area reaches a high-risk level, the system automatically queries the monitoring data of adjacent areas. The determination of adjacent areas is based on a pre-set regional adjacency table, with the number of adjacent areas set to 3 to 5. If the dynamic heat tolerance threshold of an adjacent area is significantly higher than that of the current area (threshold difference exceeding 2℃), a regional spread warning message is added to the warning signal of the current area, indicating the possibility that the heat hazard risk in that area may spread to adjacent areas. The generation of the regional spread warning message is based on the fact that when a certain area experiences a high heat hazard risk, its high-temperature air mass may spread to neighboring areas, leading to an increase in the heat hazard risk in those areas. The difference in dynamic heat tolerance threshold reflects the moderating effect of soil conditions on heat tolerance in different areas. If the heat tolerance threshold of an adjacent area is significantly higher than that of the current area, it indicates that the soil conditions in the adjacent area are more conducive to resisting high-temperature stress, but the source of its high-temperature air mass may increase the heat hazard risk in the current area.

Claims

1. A method for early warning of heat damage during peanut flowering period based on meteorological and soil factor data processing, characterized in that, Includes the following steps: S1: Collect meteorological and soil data of the peanut planting area. The meteorological data includes atmospheric temperature, relative humidity, solar radiation intensity and wind speed data. Soil data is collected synchronously through the soil monitoring module, including soil temperature, soil volumetric water content, soil bulk density and soil electrical conductivity data of the peanut root zone soil. S2: Noise reduction processing for meteorological data: outlier removal from collected soil data; The preprocessed meteorological and soil data were aligned according to timestamps and then normalized. S3: Based on the regulatory effect of soil factors on peanut heat resistance, dynamic heat resistance thresholds were determined. T t Calculation; S4: Acquire real-time atmospheric temperature data and compare it with dynamic heat resistance threshold. T t A comparison is performed; when the real-time atmospheric temperature exceeds the dynamic heat resistance threshold, the heat damage risk index is calculated. RI ; S5: Classify the risk level based on the heat damage risk index calculated in step S4; S6: Based on the risk level classification results, generate and output corresponding warning signals, including green warning signals for low risk level, yellow warning signals for medium risk level, orange warning signals for high risk level, and red warning signals for extremely high risk level.

2. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 1, characterized in that, The specific process of step S3 is as follows: S31: Based on real-time soil temperature, calculate the adjustment correction value for peanut heat resistance, i.e., calculate the soil temperature correction factor. T s : When 25≤ T soil ≤35, T s =0.15•( T soil -30); when T soil <25, T s = -0.2; or T soil >35, T s = -0.4; in, T soil Real-time temperature of the pretreated soil; T s Value constraint: -0.4 ≤ T s ≤0.75; S32: Calculation of soil volumetric water content correction factor θ : When 0.15≤ θ v ≤0.35, θ =0.8•( θ v -0.25); when θ v <0.15, θ=- 0.08; when θ v >0.35, θ= 0.08; in, θ v : Soil volumetric water content after pretreatment; 0.25 is the optimum soil volumetric water content during peanut flowering, and 0.15 is the critical threshold for soil drought; θ Value constraint: -0.08≤ θ ≤0.08; S33: Calculate the soil bulk density correction factor ρ b : ρ b =0.5•(1.3- ρ bulk )； in, ρ bulk The bulk density of the pretreated soil was 1.3 g / cm³. 3 Suitable soil bulk density for peanut growth; ρ b Value constraint: -0.35≤ ρ b ≤0.25; S34: Calculate the soil electrical conductivity correction factor EC corr : EC ≤2.0dS / m, EC corr =0; EC >2.0dS / m, EC corr =0.02·(2.0- EC ); Where EC: soil electrical conductivity after pretreatment; 2.0 dS / m is the critical value of peanut soil salt stress; EC corr Value constraint: -0.16≤ EC corr ≤0; S35: Calculate the sum of meteorological-soil interaction terms Σ β i •X i : S β i •X i = β 1(T a + θ v )+ β 2( W s + T soil })+ β 3( R s + ρ bulk )+ β 4( RH + EC ); in, β i The weight coefficients for the interaction term of meteorological factors are obtained through training using five years of historical data combined with multiple linear regression and L2 regularization, with a regularization coefficient of 0.

01. i =1,2,3,4; X i These are the interaction terms between various meteorological factors and soil factors; S36: Calculate the basic dynamic heat resistance threshold T t0 : T t0 = T base + α 1• T s + α 2• θ + α 3• ρ b + α 4• EC corr +S β i •X i ; in, T t0 Basic dynamic heat resistance threshold; T base The baseline heat tolerance threshold for peanut flowering period can be taken as 32℃. α 1. α 2. α 3. α 4 represents the weighting coefficients for soil temperature, soil volumetric water content, soil bulk density, and soil electrical conductivity.

3. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 2, characterized in that, It also includes variety coefficient correction and calculation of the final dynamic heat resistance threshold. T t : T t =T t0 ·K var ； in, T t Final dynamic heat resistance threshold; K var Peanut variety correction coefficient.

4. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 3, characterized in that, The specific process of step S4 is as follows: S41: Read real-time atmospheric temperature T a and final dynamic heat resistance threshold T t ,when T a ≤ T t No risk of heat damage, RI=0; when T a > T t Proceed to step S42 for heat damage risk index calculation; S42: When T a > T t If the duration is ≤2 hours, the conventional heat damage risk index is calculated: RI = k 1•( T a -T t )+ k 2•( RH min -RH )+ k 3•( R s / R smax )+ k 4•( W smin -W s )； in, RI Indicates the risk index of heat damage; T a Real-time atmospheric temperature; T t The dynamic heat resistance threshold; RH Real-time atmospheric relative humidity; RH min This represents the lower limit of the suitable relative humidity during the peanut flowering period. R s Real-time solar radiation intensity; R smax This represents the maximum radiation intensity that the peanut canopy can withstand. W s Real-time wind speed; W smin This is the lower limit of the effective ventilation velocity; k 1. k 2. k 3. k 4 represents the risk weight coefficients for each factor; S43: When T a > T t If the duration is greater than 2 hours, a high-temperature time accumulation correction will be applied: C time =1+0.1·D t ; in, C time : Time accumulation factor; Δ t For real-time temperature exceeding T t Duration; Revised heat damage risk index RI for: RI = C time · k 1•( T a -T t )+ k 2•( RH min -RH )+ k 3•( R s / R smax )+ k 4•( W smin -W s )。 5. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 4, characterized in that, The risk levels described in step S5 are divided into four levels: The heat hazard risk index for low-risk levels ranges from 0 to 0.5; The heat hazard risk index for medium-risk levels ranges from 0.5 to 1.0; The heat hazard risk index for high-risk levels ranges from 1.0 to 1.5; The range of the heat hazard risk index corresponding to the extremely high risk level is greater than 1.

5.

6. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 5, characterized in that, Step S6 is followed by a self-learning process for the early warning strategy: recording the actual damage data of peanuts after each early warning signal is issued; When peanuts actually suffer heat damage, backtrack the data of the period when the heat damage actually occurred and adjust the parameter weights in the dynamic heat tolerance threshold calculation. When peanuts have not suffered heat damage but the warning signal level is high, the data is also backtracked and the threshold parameter weight is reduced. The online gradient adaptive weight update method is used to adjust the parameter weight.

7. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 6, characterized in that, The specific process of the early warning strategy's self-learning process is as follows: S71: After each triggering of a medium-risk or higher warning, record the corresponding data, including the warning period: t0~t1; Weight : α 1~ α 4, β 1~ β 4; Data: T a , T soil , θ v , ρ bulk , EC , RH , R s , W s ; Warning Alert level: Yellow, Orange, or Red; actual Heat damage labels: Y=1: Heat damage actually occurred; Y=0: No heat damage occurred; S72: Determine the update direction: Y=1: Increase the weight of the corresponding soil factor and interaction term, and reduce the dynamic heat resistance threshold; Y=0, warning ≥ medium risk: reduce the weight of the corresponding soil factor and interaction term, and increase the dynamic heat resistance threshold; S73: Employs an online gradient adaptive weight update method, with a learning rate... η Weight updates are achieved by controlling the magnitude of each adjustment.

8. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 7, characterized in that, The weight update formula in step S73 is as follows: θ new =θ old +η⋅δ⋅L ( θ old )； in, θ new Updated weights; θ old Weights before update; η Learning rate, acceptable η= 0.05; δ Update direction coefficients; L ( θ old ): The contribution of the current weight to this event, calculated from the factor maximum-minimum normalization result.

9. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 1, characterized in that, It also includes a multi-regional collaborative early warning process: when the heat hazard risk index of a certain region is detected to enter a high-risk level, the monitoring data of adjacent regions are automatically queried; if the dynamic heat resistance threshold of the adjacent region is significantly higher than that of the current region, regional spread warning information is added to the early warning signal of the current region.

10. The peanut flowering period heat damage early warning method based on meteorological and soil factor data processing according to claim 9, characterized in that, The multi-regional collaborative early warning process is as follows: When the current regional heat hazard risk level is high or extremely high, the multi-regional collaborative early warning query process is automatically initiated. Based on the preset regional adjacency table, the adjacent region numbers and location information are automatically obtained; Real-time meteorological data of adjacent areas; dynamic heat tolerance thresholds of adjacent areas T t邻 ; Current heat hazard risk index of adjacent areas RI 邻 ; Calculate the difference in dynamic heat resistance threshold Δ between regions T t : D T t = T t当 - T t邻 ; in, T t当 : The final dynamic heat resistance threshold of the current region; T t邻 : The final dynamic heat resistance threshold of adjacent regions; Determine if the propagation warning conditions are met: D T t > T th ; in, T th This is the threshold value for the difference in dynamic heat resistance thresholds between regions. T th =2.0℃; Generate and add area spread warning messages: If Δ is satisfied T t If the temperature exceeds 2.0℃, the following warning will be automatically added to the current area's warning signal: "High risk of heat damage, weather conditions are likely to spread to surrounding areas, and it is recommended that neighboring areas strengthen monitoring."

Images