A method for constructing a winter wheat frost damage risk prediction model based on big data

Abstract

The present invention relates to the technical field of winter wheat frost damage prediction, and in particular to a method for constructing a winter wheat frost damage risk prediction model based on big data, the method comprising: obtaining prediction data; calculating an ideal prediction value; determining an ideal frost damage grid; adjusting a weight combination; determining an actual frost damage grid; and generating a frost damage prediction map. The present invention effectively integrates multiple factors such as predicted temperature, historical frost damage frequency, and crop development period through a multi-level dynamic adjustment mechanism. By gradually adjusting the weight combination and grid spacing, the model can be optimized for climate change, crop growth status, and frost damage history in different regions and time periods. The model can obtain meteorological change information in real time and make flexible adjustments, thereby improving the predictability and response speed of risk warnings, and effectively solving the problem of insufficient prediction accuracy when facing different climate conditions and emergencies due to reliance on fixed vulnerability coefficients and temperature thresholds.

Description

Technical Field

[0001] The present invention relates to the technical field of winter wheat frost damage prediction, and in particular to a method for constructing a winter wheat frost damage risk prediction model based on big data. Background Art

[0002] As global climate change intensifies, extreme weather events, such as cold snaps and frost, occur frequently, causing serious impacts on agricultural production, especially crops such as winter wheat. Low temperatures in winter and cold snaps in spring often cause crops to freeze, which in turn affects food production and threatens food security. Predicting and responding to frost damage is particularly important in areas with changeable climate conditions.

[0003] The patent document with publication number CN113592201A discloses an early prediction method for yield loss of winter wheat after jointing due to freezing. The method comprises: determining the jointing start date dj of winter wheat; observing the hourly grass surface temperature Tt of the wheat field starting from the jointing start date of winter wheat; constructing a vulnerability coefficient VC curve according to the impact of frost damage on different development stages of winter wheat, wherein the vulnerability coefficient VC shows an overall upward trend with the development process of winter wheat; preselecting the frost damage temperature threshold Th based on previous observation data; constructing a comprehensive frost stress index IFS: wherein FDt is the hourly frost degree day, FDDi is the i-th temperature drop The all-day average value of frost degree days FDt hourly during the process, Tt is the lowest grass surface temperature in the tth hour during the i-th cooling process, Th is the freezing temperature threshold, VCi is the vulnerability coefficient of winter wheat during the i-th cooling process, dj and da represent the start date of jointing and flowering period of winter wheat respectively; the IFS index is calculated by setting different temperature thresholds Th, the IFS index and the yield reduction rate of frozen winter wheat are used for fitting, and the appropriate freezing temperature threshold Th is determined based on the determination coefficient and root mean square error of the model between the two; the yield reduction rate predicted by the IFS index is compared with the measured yield reduction rate for accuracy verification.

[0004] It can be seen that the early prediction method for yield loss of winter wheat frozen after jointing has the following problems: the method assumes that the vulnerability coefficient (VC) of winter wheat shows an overall upward trend with the development process. This assumption is not applicable to crops in all climatic conditions or different regions. The change of VC is affected by environmental factors and crop types, and a fixed trend cannot cover all scenarios. The method uses a fixed temperature range when determining the frost damage temperature threshold, lacks a real-time update mechanism, and cannot adjust the model parameters according to different climate years or emergencies, affecting the accuracy and flexibility of the prediction. Summary of the invention

[0005] To this end, the present invention provides a method for constructing a winter wheat frost damage risk prediction model based on big data, which is used to overcome the problem of insufficient prediction accuracy in the prior art when facing different climatic conditions and emergencies due to reliance on fixed vulnerability coefficients and temperature thresholds through big data analysis and dynamic adjustment mechanisms.

[0006] To achieve the above object, the present invention provides a method for constructing a winter wheat frost damage risk prediction model based on big data, comprising:

[0007] Real-time acquisition of the predicted temperature of each to-be-determined grid in the predicted area constructed based on the preset grid spacing within the preset future time period, the historical frequency of frost damage within the preset historical time period, and the predicted crop development period;

[0008] Calculate an ideal estimated value based on the predicted temperature, the historical frequency of frost damage, the predicted crop growth period, and a preset weight combination to form a plurality of ideal estimated values;

[0009] Determine a plurality of ideal freezing damage grids according to the ideal estimated value and a preset estimated value threshold;

[0010] Adjust the preset weight combination according to each of the ideal frost-damaged grids and all grids to form a first adjusted weight combination;

[0011] Adjusting the first adjustment weight combination according to a first adjustment estimated value determined based on the first adjustment weight combination and the preset grid spacing to form a second adjustment weight combination;

[0012] Adjusting the second adjustment weight combination according to the second adjustment estimated value determined based on the second adjustment weight combination and the preset future duration to form a third adjustment weight combination;

[0013] Calculating an actual estimated value based on the third adjustment weight combination, the predicted temperature, the historical frequency of frost damage, and the predicted crop growth period to form a plurality of actual estimated values;

[0014] Determine a number of actual frost damage grids according to each of the actual estimated values ​​and the preset estimated value threshold;

[0015] Adjusting the preset grid spacing according to the actual frost damage grid and the ideal frost damage grid;

[0016] A freezing damage prediction map is generated according to a plurality of modified freezing damage grids determined based on the preset grid spacing and the third adjustment weight combination.

[0017] Furthermore, the preset weight combination includes an exposure weight corresponding to the predicted growth period of the crop, a hazard weight corresponding to the historical frequency of frost damage, and a vulnerability weight corresponding to the predicted temperature.

[0018] Furthermore, adjusting the preset weight combination according to each of the ideal freezing damage grids and all grids includes:

[0019] Calculating the ratio of the ideal freezing damage grid to all grids to form an ideal ratio;

[0020] When the ideal proportion is greater than a preset first proportion threshold, the exposure weight is increased according to the relative deviation between the ideal proportion and the preset first proportion threshold and a preset first adjustment coefficient to form a first exposure weight, the vulnerability weight is reduced according to the relative deviation between the ideal proportion and the preset first proportion threshold and a preset second adjustment coefficient to form a first vulnerability weight, and the adjusted risk weight is determined according to the exposure weight and the vulnerability weight to form a first risk weight;

[0021] The first exposure weight, the first vulnerability weight, and the first hazard weight are combined to form the first adjusted weight combination.

[0022] Furthermore, when the ideal proportion is less than or equal to a preset second proportion threshold, the exposure weight is reduced according to the relative deviation between the preset second proportion threshold and the ideal proportion and the preset first adjustment coefficient to form a first exposure weight; the vulnerability weight is increased according to the relative deviation between the ideal proportion and the preset second proportion threshold and the preset second adjustment coefficient to form a first vulnerability weight; the adjusted risk weight is determined according to the exposure weight and the vulnerability weight to form a first risk weight.

[0023] Further, adjusting the first adjustment weight combination according to the first adjustment estimated value determined based on the first adjustment weight combination and the preset grid spacing to form a second adjustment weight combination includes:

[0024] Calculating the sum of the first adjusted estimated values ​​of all ideal freezing damage grids to form a total estimated value;

[0025] Calculating the sum of the squares of the grid spacing to form a total grid area;

[0026] Calculating the ratio of the total estimated value to the total grid area to form a frost damage density;

[0027] When the frost damage density is greater than a preset density threshold, the first exposure weight is increased according to the relative deviation between the frost damage density and the preset density threshold and a preset third adjustment coefficient to form a second exposure weight, the first vulnerability weight is reduced according to the relative deviation between the frost damage density and the preset density threshold and a preset fourth adjustment coefficient to form a second vulnerability weight, and the adjusted first hazard weight is determined according to the second exposure weight and the second vulnerability weight to form a second hazard weight;

[0028] The second exposure weight, the second vulnerability weight, and the second hazard weight are combined to form the second adjusted weight combination.

[0029] Further, adjusting the second adjustment weight combination according to the second adjustment estimated value determined based on the second adjustment weight combination and the preset future duration to form a third adjustment weight combination includes:

[0030] Calculating the standard deviation of all second adjusted estimated values ​​within the preset future time period to form an estimated fluctuation value;

[0031] When the estimated fluctuation value is greater than the preset fluctuation threshold, the second risk weight is increased according to the relative deviation between the estimated fluctuation value and the preset fluctuation threshold and the preset fifth adjustment coefficient to form a third risk weight, the second exposure weight is reduced according to the relative deviation between the estimated fluctuation value and the preset fluctuation threshold and the preset sixth adjustment coefficient to form a third exposure weight, and the adjusted second vulnerability weight is determined according to the third risk weight and the third exposure weight to form a third vulnerability weight;

[0032] The third exposure weight, the third vulnerability weight, and the third hazard weight are combined to form the third adjusted weight combination.

[0033] Furthermore, the calculation of the ideal estimated value according to the predicted temperature, the historical frequency of frost damage, the predicted crop growth period and a preset weight combination includes:

[0034] The exposure weight, the hazard weight, the vulnerability weight, the relative deviation between the predicted crop development period and the preset development period threshold, the relative deviation between the historical frost damage frequency and the preset frequency threshold, and the relative deviation between the predicted temperature and the preset temperature threshold are weighted and summed to form an ideal estimated value.

[0035] Furthermore, the determining of a plurality of ideal freezing damage grids according to the ideal estimated value and the preset estimated value threshold comprises:

[0036] When the ideal estimated value is greater than the preset estimated value threshold, the grid to be determined is determined to be an ideal frost-damaged grid.

[0037] Further, the determining of a plurality of actual freezing damage grids according to each of the actual estimated values ​​and the preset estimated value threshold comprises:

[0038] When the actual estimated value is greater than the preset estimated value threshold, the ideal frost damage grid is determined to be an actual frost damage grid, and a plurality of actual frost damage grids are formed.

[0039] Further, adjusting the preset grid spacing according to the actual frost damage grid and the ideal frost damage grid includes:

[0040] Calculating the absolute value of the relative deviation between the number of the actual frost-damaged grids and the number of the ideal frost-damaged grids to form a grid number deviation;

[0041] When the grid quantity deviation is greater than a preset quantity deviation threshold, the preset grid spacing is adjusted according to a relative deviation between the grid quantity deviation and the preset quantity deviation threshold and a preset grid coefficient.

[0042] Compared with the prior art, the beneficial effect of the present invention is that, through a multi-level dynamic adjustment mechanism, multiple factors such as predicted temperature, historical frost damage frequency and crop development period are effectively integrated to achieve accurate prediction of frost damage risk. By gradually adjusting the weight combination and grid spacing, the model can optimize climate change, crop growth status and frost damage history in different regions and time periods, making the prediction results more timely and accurate. Especially with the support of big data, the model can obtain meteorological change information in real time and make flexible adjustments, which improves the predictability and response speed of risk warning. In addition, the frost damage prediction map of the model also provides timely warning information for agricultural production.

[0043] Furthermore, by reasonably setting and comprehensively using exposure weights, hazard weights and vulnerability weights, the present invention can comprehensively and accurately assess the risk of winter wheat frost damage. These weights take into account factors such as the crop's development period, historical frost damage frequency and predicted temperature, and can dynamically reflect changes in risk under different environmental conditions. The exposure weight ensures that the sensitivity of crops to frost damage at different growth stages can be quantified, the hazard weight provides an estimate of the probability of frost damage in combination with historical data, and the vulnerability weight assesses the threat of frost damage based on temperature fluctuations. Through the comprehensive use of these three weights, the model can more accurately predict the possibility of frost damage, effectively solving the problem of insufficient prediction accuracy when facing different climatic conditions and emergencies due to reliance on fixed vulnerability coefficients and temperature thresholds.

[0044] Furthermore, by combining multiple factors such as the crop growth stage, historical frost damage frequency, and predicted temperature, and calculating the ideal estimate using a weighted summation method, it can not only accurately reflect the relative importance of each factor to the frost damage risk, but also effectively eliminate the impact of individual factor deviations on the overall assessment.

[0045] Furthermore, by comparing the ideal estimated value with the preset threshold, high-risk areas that may be affected by frost damage can be accurately identified.

[0046] Furthermore, when the probability of frost damage is high, increasing the exposure weight can help strengthen the focus on risk areas, while reducing the vulnerability weight can mitigate the impact of over-predicting certain vulnerable areas. By dynamically adjusting the hazard weight, the entire model can better adapt to changes in different weather conditions and crop growth cycles, thereby improving the accuracy and practicality of frost damage risk estimation.

[0047] Furthermore, by adjusting the weights when the ideal proportion is lower than the preset threshold, the exposure, vulnerability and hazard weights can be dynamically adjusted to optimize the frost damage risk estimate. Reducing the exposure weight can help reduce excessive alarms in low-risk areas, while increasing the vulnerability weight can better identify areas susceptible to frost damage.

[0048] Furthermore, by dynamically adjusting the ratio of frost damage density to grid area, the corresponding weights can be adjusted according to the density of frost damage in different regions, avoiding blind adjustment of weights and ensuring more accurate prediction results in high-risk areas. By increasing the exposure weight and reducing the vulnerability weight, the model pays more attention to the risks in areas with high frost damage density.

[0049] Furthermore, by further adjusting the weight combination, the weights can be flexibly adjusted according to the estimated volatility in the future, so that the system can more accurately reflect risk changes when facing situations with large fluctuations and high uncertainties.

[0050] Furthermore, by accurately screening out the actual frost damage grids, it helps to improve the accuracy of risk assessment, avoid misjudgments or omissions, and ensure that the estimated results are more realistic and reliable.

[0051] Furthermore, by dynamically adjusting the grid spacing according to the number deviation between the actual and ideal grids, it is possible to effectively cope with complex factors such as meteorological changes and uneven distribution of frost damage. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a structural schematic diagram of the present invention;

[0053] Figure 2 This is a logic diagram for determining an ideal freezing damage grid in this embodiment;

[0054] Figure 3 A decision logic diagram for adjusting the preset weight combination for this embodiment;

[0055] Figure 4 This is a decision logic diagram for adjusting the first adjustment weight combination in this embodiment. DETAILED DESCRIPTION

[0056] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0057] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0058] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0059] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0060] See also Figure 1 As shown, it is a flow chart of the method for constructing a winter wheat frost damage risk prediction model based on big data in this embodiment;

[0061] This embodiment provides a method for constructing a winter wheat frost damage risk prediction model based on big data, including:

[0062] Real-time acquisition of the predicted temperature of each grid to be determined within the predicted area constructed based on the preset grid spacing within the preset future time period, the historical frequency of frost damage within the preset historical time period, and the predicted crop development period;

[0063] Calculate an ideal estimated value based on the predicted temperature, the historical frequency of frost damage, the predicted crop growth period, and a preset weight combination to form a plurality of ideal estimated values;

[0064] Determine a plurality of ideal freezing damage grids according to the ideal estimated value and a preset estimated value threshold;

[0065] Adjust the preset weight combination according to each of the ideal frost-damaged grids and all grids to form a first adjusted weight combination;

[0066] Adjusting the first adjustment weight combination according to a first adjustment estimated value determined based on the first adjustment weight combination and the preset grid spacing to form a second adjustment weight combination;

[0067] Adjusting the second adjustment weight combination according to the second adjustment estimated value determined based on the second adjustment weight combination and the preset future duration to form a third adjustment weight combination;

[0068] Calculating an actual estimated value based on the third adjustment weight combination, the predicted temperature, the historical frequency of frost damage, and the predicted crop growth period to form a plurality of actual estimated values;

[0069] Determine a number of actual frost damage grids according to each of the actual estimated values ​​and the preset estimated value threshold;

[0070] Adjusting the preset grid spacing according to the actual frost damage grid and the ideal frost damage grid;

[0071] A freezing damage prediction map is generated according to a plurality of modified freezing damage grids determined based on the preset grid spacing and the third adjustment weight combination.

[0072] The preset grid spacing refers to the spatial range represented by each grid when gridding is performed in the prediction area. It usually depends on the size of the prediction area, the data accuracy requirements and the computing resources. It is usually set between 500m and 5000m. In this embodiment, it is set to 1000m, which can ensure sufficient spatial resolution and computational efficiency.

[0073] The preset future duration refers to the time period used to predict future frost damage risks, which usually depends on the periodicity of climate change and the time span of crop development. It is set between 1 day and 30 days. In this embodiment, it is set to 10 days, which can balance the accuracy of short-term predictions and the stability of long-term trends.

[0074] The preset historical duration refers to the time range used to refer to historical frost damage data, which usually depends on the historical record period of frost damage in the region and is set between 10 and 30 years. In this embodiment, it is set to 30 years, which can better cover the changing patterns of historical frost damage and provide a stable basis for risk assessment.

[0075] The preset estimated value threshold refers to the critical value for judging the risk of frost damage, which usually depends on the cold resistance of crops and the sensitivity of meteorological changes. It is set at a value that can significantly reflect the probability of frost damage. In this embodiment, it is set to 0.6, which can accurately distinguish between high-risk areas and relatively safe areas.

[0076] Real-time acquisition of predicted temperatures, historical frost damage frequency, and predicted crop development period for each grid to be determined in the prediction area mainly relies on big data technology and the integration of multi-source data. The predicted temperature is obtained through channels such as meteorological satellites, ground meteorological stations, and meteorological models, and the prediction model is used to provide information on temperature changes in the future. The historical frequency of frost damage is based on historical predicted temperatures and frost damage records, obtained through long-term accumulation and analysis of frost damage event data, and calculated in combination with the correlation between historical climate trends and frost damage occurrence. The predicted crop development period is predicted based on real-time crop growth conditions, meteorological factors, and growth stage related data with the help of agricultural Internet of Things sensors, remote sensing technology, and crop growth models. All data are integrated and processed through a unified real-time data platform to ensure that the information obtained is accurate and timely to support subsequent risk assessment and decision-making.

[0077] The process of generating a frost damage prediction map is based on the spatial distribution of the modified frost damage grids and the corresponding risk estimation values. First, the modified frost damage value of each grid point is determined by combining the preset grid spacing and the third adjustment weight combination. This value combines factors such as meteorological conditions, historical frost damage frequency, and crop development period. After step-by-step adjustment, it more accurately reflects the frost damage risk of each grid. Next, these modified frost damage grid data are visualized, and the blank areas between each grid are smoothed by a spatial interpolation algorithm (such as inverse distance weighting) to obtain a continuous frost damage risk distribution map. Finally, the frost damage prediction map presents the frost damage risk level of each plot in the entire region, providing decision makers with an intuitive and clear risk assessment view, which helps to formulate targeted prevention and control measures.

[0078] This method uses a multi-level adjustment strategy to construct a winter wheat frost damage risk estimation model based on big data and real-time predicted temperature. First, the predicted temperature, historical frost damage frequency and predicted crop development period in the prediction area are obtained, the ideal estimated value is calculated and the ideal frost damage grid is determined according to the preset estimated value threshold. Then, the estimation results are optimized by adjusting the weight combination and making gradual adjustments based on the grid number deviation and frost damage density. Finally, the adjusted grid spacing and weight combination are combined to generate a corrected frost damage grid and draw a frost damage estimation map to provide visual prediction results for risk warning.

[0079] Through a multi-level dynamic adjustment mechanism, multiple factors such as predicted temperature, historical frost damage frequency, and crop development period are effectively integrated to achieve accurate prediction of frost damage risks. By gradually adjusting the weight combination and grid spacing, the model can optimize climate change, crop growth status, and frost damage history in different regions and time periods, making the prediction results more timely and accurate. Especially with the support of big data, the model can obtain meteorological change information in real time and make flexible adjustments, which improves the predictability and response speed of risk warning. In addition, the frost damage prediction map of the model also provides timely warning information for agricultural production, effectively solving the problem of insufficient prediction accuracy when facing different climate conditions and emergencies due to reliance on fixed vulnerability coefficients and temperature thresholds.

[0080] Specifically, the preset weight combination includes an exposure weight corresponding to the predicted growth period of the crop, a hazard weight corresponding to the historical frequency of frost damage, and a vulnerability weight corresponding to the predicted temperature.

[0081] The exposure weight is a measure of the sensitivity of crops to frost damage at different developmental stages, which usually depends on the growth stage of the crop, especially during key growth stages (such as booting, flowering, and grain filling), when the risk of exposure is higher. The value of the exposure weight is usually set between 0 and 1. In this embodiment, it is set to 0.3, which can highlight the key impact of the crop development stage on frost damage, so that the more sensitive stages of the development period receive more attention in the estimation, thereby providing an accurate reference for risk estimation.

[0082] The risk weight reflects the frequency of historical frost damage and its threat to crops. It is usually based on the historical frost damage data of the region. Regions with frequent frost damage are given higher risk weights. The value is usually between 0 and 1. In this embodiment, it is set to 0.2, which can help the model comprehensively consider the historical frost damage frequency and probability of occurrence, so that historical data from different regions can effectively affect the current risk assessment.

[0083] The vulnerability weight measures the impact of temperature fluctuations on crops. It is usually determined based on the difference between the forecast temperature and the cold tolerance of the crops. Regions with drastic temperature changes have higher vulnerability weights. The common value range of the vulnerability weight is 0 to 1. In this example, it is set to 0.5, which helps to fine-tune the assessment of temperature sensitivity.

[0084] By reasonably setting and comprehensively using exposure weights, hazard weights, and vulnerability weights, the present invention can comprehensively and accurately assess the risk of winter wheat frost damage. These weights take into account factors such as the crop's development period, historical frost damage frequency, and predicted temperature, and can dynamically reflect changes in risk under different environmental conditions. The exposure weight ensures that the sensitivity of crops to frost damage at different growth stages can be quantified, the hazard weight provides an estimate of the probability of frost damage in combination with historical data, and the vulnerability weight assesses the threat of frost damage based on temperature fluctuations. Through the combined use of these three weights, the model can more accurately predict the possibility of frost damage.

[0085] Specifically, the calculation of the ideal estimated value according to the predicted temperature, the historical frequency of frost damage, the predicted crop growth period and a preset weight combination includes:

[0086] The exposure weight, the hazard weight, the vulnerability weight, the relative deviation between the predicted crop development period and the preset development period threshold, the relative deviation between the historical frost damage frequency and the preset frequency threshold, and the relative deviation between the predicted temperature and the preset temperature threshold are weighted and summed to form an ideal estimated value.

[0087] The value of the preset developmental period threshold usually depends on the specific growth cycle of the crop and its sensitive period to frost damage. For winter wheat, it is usually set in the middle and late stages of its growth, that is, from the spring budding and greening to the heading stage, especially from the jointing to the heading stage, and is more sensitive to frost damage. In this embodiment, the preset developmental period threshold is set to the greening period to the heading period of winter wheat, which can help the model accurately judge the risk of frost damage based on the growth pattern and frost damage sensitive period of winter wheat, avoid risk assessment at irrelevant growth stages, and improve the accuracy of the estimate.

[0088] The preset frequency threshold is a reference value set according to the frequency of historical frost damage, which is usually determined based on the statistical data of past frost damage and is usually set within the average frequency range of historical frost damage in a specific area. In this embodiment, the value is set to the frequency of frost damage in the past thirty years, which can better scientifically predict the risk of future frost damage based on historical data.

[0089] The preset temperature threshold refers to the temperature limit for specific crops and frost damage risks, which usually depends on the tolerance of different crops to low temperatures and the temperature range of frost damage, and is usually set at the lowest temperature that the crop can withstand or the temperature threshold for frost damage. According to the setting of the winter wheat development period, in this embodiment, the greening period of winter wheat is -4°C, the rising period is -1°C, the early jointing period is 2°C, the late jointing period is 4°C, and the heading period is 6°C. Because the temperatures corresponding to the above development periods are known to be the risk temperatures for frost damage to winter wheat, the frost damage risk areas under low temperature conditions can be accurately identified, thereby improving the accuracy of the estimation.

[0090] By performing a weighted summation calculation on multiple key factors, the relative deviation is calculated based on the exposure weight, hazard weight, and vulnerability weight, combined with the predicted crop growth period, historical frost damage frequency, and temperature data, and compared with the preset threshold. Then, these deviations are weighted summed to form an ideal estimate by comprehensively considering the impact of various factors on the frost damage risk.

[0091] By combining multiple factors such as the crop growth stage, historical frost damage frequency, and predicted temperature, and using a weighted summation method to calculate the ideal estimate, it can not only accurately reflect the relative importance of each factor to the frost damage risk, but also effectively eliminate the impact of individual factor deviations on the overall assessment.

[0092] Please continue reading Figure 2 As shown, it is a determination logic diagram for determining an ideal freezing damage grid in this embodiment;

[0093] Specifically, determining a number of ideal freezing damage grids according to the ideal estimated value and the preset estimated value threshold comprises:

[0094] When the ideal estimated value is greater than the preset estimated value threshold, the grid to be determined is determined to be an ideal frost-damaged grid.

[0095] First, the ideal estimated value of each grid is calculated, and then compared with the preset estimated value threshold. When the ideal estimated value of a grid is greater than the preset threshold, the grid is judged as an ideal frost damage grid. This process effectively screens out areas with higher risks through threshold judgment for subsequent adjustments and optimization.

[0096] By comparing the ideal estimate with the preset threshold, high-risk areas that may be affected by frost damage can be accurately identified.

[0097] Please continue reading Figure 3 As shown, it is a decision logic diagram for adjusting the preset weight combination in this embodiment;

[0098] Specifically, adjusting the preset weight combination according to each of the ideal freezing damage grids and all grids includes:

[0099] Calculating the ratio of the ideal freezing damage grid to all grids to form an ideal ratio;

[0100] When the ideal proportion is greater than a preset first proportion threshold, the exposure weight is increased according to the relative deviation between the ideal proportion and the preset first proportion threshold and a preset first adjustment coefficient to form a first exposure weight, the vulnerability weight is reduced according to the relative deviation between the ideal proportion and the preset first proportion threshold and a preset second adjustment coefficient to form a first vulnerability weight, and the adjusted risk weight is determined according to the exposure weight and the vulnerability weight to form a first risk weight;

[0101] The first exposure weight, the first vulnerability weight, and the first hazard weight are combined to form the first adjusted weight combination.

[0102] The relative deviation between the ideal proportion and the preset first proportion threshold is positively correlated with the exposure weight, and the relative deviation between the ideal proportion and the preset first proportion threshold is positively correlated with the vulnerability weight.

[0103] The preset first ratio threshold refers to the ratio threshold of the ideal frost damage grid to all grids used when adjusting the exposure weight and the vulnerability weight. It depends on the common occurrence mode of frost damage in the target area and the expected prediction sensitivity. It is usually set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which can trigger appropriate adjustments when areas with higher frost damage risks appear, thereby ensuring that the model responds more sensitively and accurately to frost damage in key areas.

[0104] The preset first adjustment coefficient is a coefficient used to adjust the exposure weight and vulnerability weight according to the relative deviation between the ideal proportion and the preset first proportion threshold. It is determined by the historical data of actual frost damage events and the sensitivity requirements of the model. It is usually set between 0.5 and 1.5. In this embodiment, it is set to 1.2. Through the adjustment of this coefficient, the model can appropriately strengthen the calculation of the exposure weight when encountering a large proportion deviation, thereby improving the accuracy of risk assessment.

[0105] The preset second adjustment coefficient is a coefficient used to adjust the vulnerability weight and the hazard weight according to the relative deviation between the ideal ratio and the preset second ratio threshold. It is set according to the actual environment and the model's requirements for the sensitivity of the frost risk warning. It is usually between 0.5 and 1.5. In this embodiment, it is set to 0.8, which can appropriately reduce the vulnerability weight when the ideal ratio is low, thereby reducing the risk of over-prediction.

[0106] According to the ratio of the ideal frost damage grid and all grids, the ideal ratio is calculated and then compared with the preset first ratio threshold. If the ideal ratio is greater than the threshold, firstly, by calculating its relative deviation from the threshold and combining it with the preset adjustment coefficient, the exposure weight is increased to form the first exposure weight. At the same time, the vulnerability weight is reduced to form the first vulnerability weight. Then, the hazard weight is recalculated by the adjusted exposure weight and vulnerability weight to form the first hazard weight. Finally, these three adjusted weights are combined to form the first adjusted weight combination, which provides an accurate risk assessment model for subsequent estimation and analysis.

[0107] When the probability of frost damage is high, increasing the exposure weight can help strengthen the focus on risk areas, while reducing the vulnerability weight can mitigate the impact of over-forecasting certain vulnerable areas. By dynamically adjusting the hazard weight, the entire model can better adapt to changes in different weather conditions and crop growth cycles, thereby improving the accuracy and practicality of frost damage risk estimation.

[0108] Specifically, when the ideal proportion is less than or equal to the preset second proportion threshold, the exposure weight is reduced according to the relative deviation between the preset second proportion threshold and the ideal proportion and the preset first adjustment coefficient to form a first exposure weight; the vulnerability weight is increased according to the relative deviation between the ideal proportion and the preset second proportion threshold and the preset second adjustment coefficient to form a first vulnerability weight; the adjusted risk weight is determined according to the exposure weight and the vulnerability weight to form a first risk weight.

[0109] The preset second proportion threshold refers to the critical value for adjusting the exposure and vulnerability weights when the proportion of the ideal frost damage grid is less than or equal to this value. It usually depends on the distribution characteristics of the frost damage risk in the prediction area and is generally set between 0.2 and 0.4 to avoid excessive adjustment due to too small a proportion. In this embodiment, it is set to 0.3, which can avoid excessive adjustments while ensuring that the estimated results have good accuracy and sensitivity, and adapt to the spatial distribution of actual frost damage risks.

[0110] The preset second adjustment coefficient is a coefficient used to adjust the vulnerability weight when the ideal proportion is less than or equal to the preset second proportion threshold. It is usually set based on the actual frequency of frost damage in the area and the sensitivity of the impact. It is generally set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which can effectively adjust the increase in the vulnerability weight, so that the model can better identify potential high-risk areas for frost damage, while avoiding excessive increase in vulnerability, ensuring the balance and stability of the model.

[0111] When the ideal proportion is less than or equal to the preset second proportion threshold, the relative deviation between the ideal proportion and the preset second proportion threshold is first calculated. Then, the exposure weight is reduced according to the relative deviation and the preset first adjustment coefficient to form a new first exposure weight; at the same time, the vulnerability weight is increased according to the relative deviation and the preset second adjustment coefficient to form a new first vulnerability weight. Subsequently, the adjusted risk weight is calculated by the adjusted exposure weight and vulnerability weight, and finally the first risk weight is formed.

[0112] By adjusting the weights when the ideal ratio is lower than the preset threshold, the exposure, vulnerability and hazard weights can be dynamically adjusted to optimize the frost damage risk estimate. Reducing the exposure weight can help reduce excessive alarms in low-risk areas, while increasing the vulnerability weight can better identify areas susceptible to frost damage.

[0113] Please continue reading Figure 4 As shown, it is a decision logic diagram for adjusting the first adjustment weight combination in this embodiment;

[0114] Specifically, adjusting the first adjustment weight combination according to the first adjustment estimated value determined based on the first adjustment weight combination and the preset grid spacing to form a second adjustment weight combination includes:

[0115] Calculating the sum of the first adjusted estimated values ​​of all ideal freezing damage grids to form a total estimated value;

[0116] Calculating the sum of the squares of the grid spacing to form a total grid area;

[0117] Calculating the ratio of the total estimated value to the total grid area to form a frost damage density;

[0118] When the frost damage density is greater than a preset density threshold, the first exposure weight is increased according to the relative deviation between the frost damage density and the preset density threshold and a preset third adjustment coefficient to form a second exposure weight, the first vulnerability weight is reduced according to the relative deviation between the frost damage density and the preset density threshold and a preset fourth adjustment coefficient to form a second vulnerability weight, and the adjusted first hazard weight is determined according to the second exposure weight and the second vulnerability weight to form a second hazard weight;

[0119] The second exposure weight, the second vulnerability weight, and the second hazard weight are combined to form the second adjusted weight combination.

[0120] The calculation of the first adjustment estimate includes:

[0121] The first exposure weight, the first hazard weight, the first vulnerability weight, the relative deviation between the predicted crop development period and a preset development period threshold, the relative deviation between the historical frost damage frequency and a preset frequency threshold, and the relative deviation between the predicted temperature and a preset temperature threshold are weighted and summed to form a first adjusted estimated value.

[0122] The preset third adjustment coefficient is used to adjust the coefficient of the second risk weight according to the relative deviation between the estimated volatility value and the preset volatility threshold, which depends on the model's sensitivity requirements to risk fluctuations, and the setting range is usually between 0.1 and 0.5. The third adjustment coefficient is used to control how the risk weight is adjusted when the estimated volatility is large, to ensure that the model can reflect the risk changes during large fluctuations. In this embodiment, the third adjustment coefficient is set to 0.3, so as to reasonably increase the risk weight when dealing with large fluctuations and enhance the model's ability to respond to risks.

[0123] The preset fourth adjustment coefficient is used to adjust the coefficient of the second exposure weight according to the relative deviation between the estimated volatility value and the preset volatility threshold, which depends on the model's response requirements to changes in exposure, and the setting range is generally between 0.1 and 0.3. The role of the fourth adjustment coefficient is to balance the changes in exposure weights during large fluctuations to avoid excessive changes in exposure, which may lead to model instability. In this embodiment, the preset fourth adjustment coefficient is set to 0.2, which helps to maintain a relatively stable exposure weight adjustment in actual applications, making the overall estimate of the model more stable and reasonable.

[0124] The relative deviation between the frost damage density and the preset density threshold is positively correlated with the first exposure weight, and the relative deviation between the frost damage density and the preset density threshold is positively correlated with the first vulnerability weight.

[0125] The preset density threshold refers to the critical value used to determine whether the frost damage density exceeds the preset standard. It depends on the historical frost damage pattern, crop type and regional characteristics. It is usually set between 0.2 and 0.5. In this embodiment, it is set to 0.3, which can effectively identify high-risk areas and ensure appropriate weight adjustments in areas with higher frost damage density, thereby improving the prediction accuracy of the model.

[0126] The preset third adjustment coefficient is a proportional factor for adjusting the exposure weight, which depends on the sensitivity of climate change and the growth characteristics of different crops. It is usually set between 0.05 and 0.2. In this embodiment, it is set to 0.1, which can ensure that in areas with higher frost damage density, the exposure weight is moderately increased, which is conducive to improving the accuracy of frost damage risk estimation in the area.

[0127] The preset fourth adjustment coefficient is a proportional factor for adjusting the vulnerability weight, which depends on the frost resistance of the crop and the impact of meteorological conditions on the crop. It is usually set between 0.05 and 0.2. In this embodiment, it is set to 0.15, which can ensure that the vulnerability weight is moderately reduced in areas with high frost damage density, effectively avoiding excessive amplification of the impact of vulnerability on frost damage estimation, thereby making the weight adjustment of the model in high-risk areas more reasonable.

[0128] According to the first adjustment weight combination and the ideal estimated value and the preset grid spacing, the sum of the estimated values ​​of all ideal frost damage grids is first calculated to form a total estimated value. Next, the sum of the squares of the grid spacing is calculated to obtain the total grid area, and the ratio of the total estimated value to the total grid area is further calculated to obtain the frost damage density. When the frost damage density exceeds the preset density threshold, the exposure weight and the vulnerability weight are adjusted using the preset third and fourth adjustment coefficients according to the relative deviation between the frost damage density and the threshold to form a new weight combination. During this adjustment process, the hazard weight is also adjusted to ensure that the final second adjustment weight combination can more accurately reflect the frost damage risk in the predicted area, and improve the adaptability and accuracy of the model, especially for optimizing high-density frost damage areas.

[0129] By dynamically adjusting the ratio of frost damage density to grid area, the corresponding weights can be adjusted according to the density of frost damage in different regions, avoiding blind adjustment of weights and ensuring more accurate prediction results in high-risk areas. By increasing the exposure weight and reducing the vulnerability weight, the model pays more attention to the risks in areas with high frost damage density.

[0130] Specifically, adjusting the second adjustment weight combination according to the second adjustment estimated value determined based on the second adjustment weight combination and the preset future duration to form a third adjustment weight combination includes:

[0131] Calculating the standard deviation of all second adjusted estimated values ​​within the preset future time period to form an estimated fluctuation value;

[0132] When the estimated fluctuation value is greater than the preset fluctuation threshold, the second risk weight is increased according to the relative deviation between the estimated fluctuation value and the preset fluctuation threshold and the preset fifth adjustment coefficient to form a third risk weight, the second exposure weight is reduced according to the relative deviation between the estimated fluctuation value and the preset fluctuation threshold and the preset sixth adjustment coefficient to form a third exposure weight, and the adjusted second vulnerability weight is determined according to the third risk weight and the third exposure weight to form a third vulnerability weight;

[0133] The third exposure weight, the third vulnerability weight, and the third hazard weight are combined to form the third adjusted weight combination.

[0134] The calculation of the second adjusted estimate includes:

[0135] The second exposure weight, the second hazard weight, the second vulnerability weight, the relative deviation between the predicted crop development period and the preset development period threshold, the relative deviation between the historical frost damage frequency and the preset frequency threshold, and the relative deviation between the predicted temperature and the preset temperature threshold are weighted and summed to form a second adjusted estimate.

[0136] The preset fluctuation threshold is a critical value used to determine whether the estimated fluctuation is significant. It is usually set according to the fluctuation characteristics of the predicted data, depending on the sensitivity and fluctuation range of the model. It is usually set between 0.1 and 0.5 to determine when the estimated fluctuation is large and trigger the weight adjustment. In this embodiment, the preset fluctuation threshold is set to 0.2, which is intended to reasonably screen out large fluctuations to ensure that the model can be effectively adjusted when the fluctuation is large, and avoid unnecessary weight changes due to too small fluctuations.

[0137] The preset fifth adjustment coefficient is used to adjust the third risk weight according to the relative deviation between the estimated volatility value and the preset volatility threshold, and is usually set between 0.1 and 0.3, depending on the sensitivity requirements of the model for risk adjustment. A larger coefficient causes a more significant change in the risk weight when the volatility is large. In this embodiment, the preset fifth adjustment coefficient is set to 0.25, which is intended to ensure that the risk weight can be fully enhanced in the case of large fluctuations, thereby improving the model's adaptability to fluctuations and the accuracy of risk prediction.

[0138] The preset sixth adjustment coefficient is used to adjust the third exposure weight according to the relative deviation between the estimated volatility value and the preset volatility threshold, and is usually set between 0.1 and 0.3. It is set according to the flexibility requirements of the model for exposure adjustment. A smaller coefficient can avoid excessive exposure weight adjustment. In this embodiment, the preset sixth adjustment coefficient is set to 0.15, which helps to adjust the exposure weight appropriately when the estimated volatility is large, without causing excessive changes in exposure, thereby maintaining the stability of the model prediction.

[0139] By calculating the standard deviation of the ideal estimated value within a preset future time period, the estimated fluctuation value is obtained and compared with the preset fluctuation threshold. If the estimated fluctuation value is greater than the preset fluctuation threshold, the various weights are adjusted according to the relative deviation between the two and the preset adjustment coefficient. Specifically, the third hazard weight is formed by increasing the second hazard weight, the third exposure weight is formed by reducing the second exposure weight, and the third vulnerability weight is determined based on these two adjusted weights. Finally, the three adjusted weights are combined to obtain the third adjusted weight combination for more accurate frost damage risk estimation.

[0140] By further adjusting the weight combination, the weights can be flexibly adjusted according to the estimated volatility in the future, so that the system can more accurately reflect risk changes when facing situations with large fluctuations and high uncertainty.

[0141] Specifically, determining a number of actual freezing damage grids according to the actual estimated values ​​and the preset estimated value thresholds includes:

[0142] When the actual estimated value is greater than the preset estimated value threshold, the ideal frost damage grid is determined to be an actual frost damage grid, and a plurality of actual frost damage grids are formed.

[0143] By comparing the actual estimated value with the preset estimated value threshold, it is determined whether frost damage has occurred. If the actual estimated value is higher than the preset threshold, the grid is determined to be an actual frost damage grid, forming a group of actual frost damage grids. This judgment basis ensures that only areas that reach the preset threshold are identified as having frost damage risks, thereby accurately identifying high-risk areas.

[0144] By accurately screening out the actual frost damage grids, it helps to improve the accuracy of risk assessment, avoid misjudgments or omissions, and ensure that the estimated results are more realistic and reliable.

[0145] Specifically, adjusting the preset grid spacing according to the actual frost damage grid and the ideal frost damage grid includes:

[0146] Calculating the absolute value of the relative deviation between the number of the actual frost-damaged grids and the number of the ideal frost-damaged grids to form a grid number deviation;

[0147] When the grid number deviation is greater than a preset number deviation threshold, adjusting the preset grid spacing according to a relative deviation between the grid number deviation and the preset number deviation threshold and a preset grid coefficient;

[0148] When the grid quantity deviation is a positive number, reducing the preset grid spacing according to the relative deviation between the grid quantity deviation and the preset quantity deviation threshold and the preset grid coefficient;

[0149] When the grid quantity deviation is a negative number, the preset grid spacing is increased according to the relative deviation between the grid quantity deviation and the preset quantity deviation threshold and the preset grid coefficient.

[0150] The relative deviation between the grid quantity deviation and the preset quantity deviation threshold is positively correlated with the preset grid spacing.

[0151] The preset quantity deviation threshold is a tolerance range for evaluating the quantity difference between the actual frost damage grid and the ideal frost damage grid. It depends on the accuracy requirements of the evaluation and the fault tolerance in practical applications. It is usually set between 5% and 20%. In this embodiment, it is set to 10%, which can tolerate small errors in the estimation process to a certain extent, while avoiding over-correction of the number of grids, thereby ensuring the stability and operability of the model calculation.

[0152] The preset grid coefficient is a coefficient used to adjust the preset grid spacing, which usually depends on the specific environmental characteristics of the frost damage assessment area and the requirements for the result accuracy. Generally speaking, the grid coefficient is set between 1.1 and 1.5. In this embodiment, it is set to 1.2, which helps to reasonably adjust the grid spacing according to the grid number deviation, thereby optimizing the calculation accuracy, while avoiding over-calculation or inaccurate prediction caused by too large or too small grid adjustment, and improving the practical applicability of the estimated results.

[0153] First, the relative deviation between the actual number of frost damage grids and the ideal number of frost damage grids is calculated to obtain the grid number deviation. When the grid number deviation exceeds the preset number deviation threshold, the system adjusts the preset grid spacing through the preset grid coefficient according to the deviation value and the relative deviation of the preset threshold. This adjustment process helps to refine the grid spacing and ensure the accuracy of frost damage risk assessment.

[0154] By dynamically adjusting the grid spacing according to the number deviation between the actual and ideal grids, it is possible to effectively cope with complex factors such as meteorological changes and uneven distribution of frost damage.

[0155] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0156] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing a winter wheat frost damage risk prediction model based on big data, characterized in that: include: Real-time acquisition of the predicted temperature of each grid to be determined within the predicted area constructed based on the preset grid spacing within the preset future time period, the historical frequency of frost damage within the preset historical time period, and the predicted crop development period; Calculate an ideal estimated value based on the predicted meteorological data, the historical frequency of frost damage, the predicted crop growth period and a preset weight combination to form a plurality of ideal estimated values; Determine a plurality of ideal freezing damage grids according to the ideal estimated value and a preset estimated value threshold; Adjust the preset weight combination according to each of the ideal frost-damaged grids and all grids to form a first adjusted weight combination; Adjusting the first adjustment weight combination according to a first adjustment estimated value determined based on the first adjustment weight combination and the preset grid spacing to form a second adjustment weight combination; Adjusting the second adjustment weight combination according to the second adjustment estimated value determined based on the second adjustment weight combination and the preset future duration to form a third adjustment weight combination; Calculating an actual estimated value based on the third adjustment weight combination, the predicted temperature, the historical frequency of frost damage, and the predicted crop growth period to form a plurality of actual estimated values; Determine a number of actual frost damage grids according to each of the actual estimated values ​​and the preset estimated value threshold; Adjusting the preset grid spacing according to the actual frost damage grid and the ideal frost damage grid; Generate a frost damage prediction map according to a plurality of modified frost damage grids determined based on the preset grid spacing and the third adjustment weight combination; The preset weight combination includes an exposure weight corresponding to the predicted growth period of the crop, a hazard weight corresponding to the historical frequency of frost damage, and a vulnerability weight corresponding to the predicted temperature.

2. The method for constructing a winter wheat frost injury risk prediction model based on big data according to claim 1, characterized in that: The adjusting of the preset weight combination according to each of the ideal freezing damage grids and all grids comprises: Calculating the ratio of the ideal freezing damage grid to all grids to form an ideal ratio; When the ideal proportion is greater than a preset first proportion threshold, the exposure weight is increased according to the relative deviation between the ideal proportion and the preset first proportion threshold and a preset first adjustment coefficient to form a first exposure weight, the vulnerability weight is reduced according to the relative deviation between the ideal proportion and the preset first proportion threshold and a preset second adjustment coefficient to form a first vulnerability weight, and the adjusted risk weight is determined according to the exposure weight and the vulnerability weight to form a first risk weight; The first exposure weight, the first vulnerability weight, and the first hazard weight are combined to form the first adjusted weight combination.

3. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 2, characterized in that: When the ideal proportion is less than or equal to the preset second proportion threshold, the exposure weight is reduced according to the relative deviation between the preset second proportion threshold and the ideal proportion and the preset first adjustment coefficient to form a first exposure weight; the vulnerability weight is increased according to the relative deviation between the ideal proportion and the preset second proportion threshold and the preset second adjustment coefficient to form a first vulnerability weight; the adjusted risk weight is determined according to the exposure weight and the vulnerability weight to form a first risk weight.

4. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 3, characterized in that: The adjusting the first adjustment weight combination according to the first adjustment estimated value determined based on the first adjustment weight combination and the preset grid spacing to form a second adjustment weight combination comprises: Calculating the sum of the first adjusted estimated values ​​of all ideal freezing damage grids to form a total estimated value; Calculating the sum of the squares of the grid spacing to form a total grid area; Calculating the ratio of the total estimated value to the total grid area to form a frost damage density; When the frost damage density is greater than a preset density threshold, the first exposure weight is increased according to the relative deviation between the frost damage density and the preset density threshold and a preset third adjustment coefficient to form a second exposure weight, the first vulnerability weight is reduced according to the relative deviation between the frost damage density and the preset density threshold and a preset fourth adjustment coefficient to form a second vulnerability weight, and the adjusted first hazard weight is determined according to the second exposure weight and the second vulnerability weight to form a second hazard weight; The second exposure weight, the second vulnerability weight, and the second hazard weight are combined to form the second adjusted weight combination.

5. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 4, characterized in that: The step of adjusting the second adjustment weight combination according to the second adjustment estimated value determined based on the second adjustment weight combination and the preset future duration to form a third adjustment weight combination includes: Calculating the standard deviation of all second adjusted estimated values ​​within the preset future time period to form an estimated fluctuation value; When the estimated fluctuation value is greater than the preset fluctuation threshold, the second risk weight is increased according to the relative deviation between the estimated fluctuation value and the preset fluctuation threshold and the preset fifth adjustment coefficient to form a third risk weight, the second exposure weight is reduced according to the relative deviation between the estimated fluctuation value and the preset fluctuation threshold and the preset sixth adjustment coefficient to form a third exposure weight, and the adjusted second vulnerability weight is determined according to the third risk weight and the third exposure weight to form a third vulnerability weight; The third exposure weight, the third vulnerability weight, and the third hazard weight are combined to form the third adjusted weight combination.

6. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 5, characterized in that: The calculation of the ideal estimated value based on the predicted meteorological data, the historical frequency of frost damage, the predicted crop growth period and the preset weight combination includes: The exposure weight, the hazard weight, the vulnerability weight, the relative deviation between the predicted crop development period and the preset development period threshold, the relative deviation between the historical frost damage frequency and the preset frequency threshold, and the relative deviation between the predicted temperature and the preset temperature threshold are weighted and summed to form an ideal estimated value.

7. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 6, characterized in that: Determining a plurality of ideal freezing damage grids according to the ideal estimated value and the preset estimated value threshold comprises: When the ideal estimated value is greater than the preset estimated value threshold, the grid to be determined is determined to be an ideal frost-damaged grid.

8. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 7, characterized in that: Determining a plurality of actual freezing damage grids according to the actual estimated values ​​and the preset estimated value thresholds comprises: When the actual estimated value is greater than the preset estimated value threshold, the ideal frost damage grid is determined to be an actual frost damage grid, and a plurality of actual frost damage grids are formed.

9. The method for constructing a winter wheat frost damage risk prediction model based on big data according to claim 8, characterized in that: The adjusting the preset grid spacing according to the actual frost damage grid and the ideal frost damage grid comprises: Calculating the absolute value of the relative deviation between the number of the actual frost-damaged grids and the number of the ideal frost-damaged grids to form a grid number deviation; When the grid quantity deviation is greater than a preset quantity deviation threshold, the preset grid spacing is adjusted according to a relative deviation between the grid quantity deviation and the preset quantity deviation threshold and a preset grid coefficient.

Images