A fruit tree crown layer frost monitoring and early warning system and intervention method

Abstract

The application discloses a fruit tree crown layer frost monitoring and early warning system and an intervention method, and relates to the technical field of fruit tree frost monitoring and early warning. The system comprises a temperature acquisition module, a main control module, a power supply module, a communication module and a video monitoring module; wherein the temperature acquisition module, the main control module, the power supply module and the communication module are integrated in a quick-release vertical support machine case, and the video monitoring module is installed at the top of the support; the main control module is connected with each module, and the power supply module provides power support for the system; the main control module acquires real-time temperature data through the temperature acquisition module and transmits the real-time temperature data to a monitoring platform through the communication module; an early warning algorithm analyzes a frost occurrence period based on temperature changes in an orchard and timely issues an alarm; meanwhile, the video monitoring module can observe a tree crown frost situation in real time, and combined with intervention measures such as unmanned aerial vehicle disturbance, the video monitoring module can effectively reduce the damage of frost to fruit trees; the system can accurately identify frost risks, reduce the yield reduction and fruit quality decline of fruit trees caused by frost, thereby reducing the economic losses of fruit farmers and guaranteeing the economic benefits of the orchard.

Description

Technical Field

[0001] This invention relates to the field of agricultural plant protection technology, specifically to a fruit tree canopy frost monitoring and early warning system and intervention method. Background Technology

[0002] Against the backdrop of global climate change, extreme weather events are becoming increasingly frequent. Frost, as a common agricultural meteorological disaster, has a serious impact on orchard production. Compared with other crops, fruit trees have a long growth cycle and require large upfront investments. Once damaged by frost, not only will the yield and quality of the fruit decrease sharply in the current year, but in severe cases, it may also damage the fruit trees, affecting their growth and fruiting for many years to come, resulting in huge economic losses for fruit farmers.

[0003] Currently, orchards rely primarily on traditional, experience-based methods to predict frost occurrences. Some fruit growers, based on years of accumulated experience, estimate the likelihood of frost based on seasonal changes and weather patterns. However, this method lacks scientific basis and precision, often failing to provide timely and accurate assessments. While some orchards have installed temperature sensors, these sensors are typically sparsely distributed, only acquiring temperature information from limited locations, making it difficult to comprehensively reflect temperature changes across the entire orchard. This limitation is even more pronounced in large orchards with complex terrain, easily leading to monitoring blind spots and resulting in misjudgments or missed frost detections.

[0004] Traditional methods for frost intervention, such as fumigation and mulching, have many drawbacks. While fumigation can raise near-surface temperatures to some extent, it pollutes the environment, and the range and intensity of the fumigation are difficult to control precisely, resulting in inconsistent effectiveness. Mulching requires significant manpower and resources and is cumbersome to implement, making it difficult to carry out in large orchards. Furthermore, these traditional methods are mostly reactive measures taken after frost has occurred, failing to provide early warning or proactive prevention.

[0005] In summary, existing orchard frost monitoring and intervention methods are insufficient to meet the needs of modern, precise, and intelligent orchard management. Therefore, a system and method capable of real-time, comprehensive, and accurate monitoring of orchard frost and timely implementation of effective intervention measures is a critical issue that urgently needs to be addressed. This is essential for improving orchard disaster prevention and mitigation capabilities, protecting fruit growers' economic benefits, and promoting sustainable agricultural development. Therefore, this invention proposes a fruit tree canopy frost monitoring and early warning system and intervention method. Summary of the Invention

[0006] The purpose of this invention is to provide a fruit tree canopy frost monitoring and early warning system and intervention method to solve the problems mentioned in the background above.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a fruit tree canopy frost monitoring and early warning system, comprising the following modules:

[0008] Temperature acquisition module; main control module; power supply module; communication module; video monitoring module; monitoring module;

[0009] The main control module is connected to the temperature acquisition module, the power supply module, and the communication module, respectively. The power supply module provides power support to the temperature acquisition module, the main control module, and the communication module. The main control module collects temperature data from the temperature acquisition module and sends it to the monitoring module through the communication module. The monitoring module analyzes the frost occurrence period based on the orchard temperature values ​​and issues an alarm. The video monitoring module observes the frost situation of the orchard canopy in real time. The temperature acquisition module is suspended at different heights in the canopy, including H1: 0.5m, H2: 1.5m, and H3: 2.7m, with three temperature sensors suspended at each height level.

[0010] Preferably, the video surveillance module includes a monitoring camera pan-tilt unit, which is mounted on the top of a fixed bracket. A solar panel and a chassis are also fixedly mounted on the surface of the fixed bracket. The temperature acquisition module, main control module, power supply module, and communication module are located inside the chassis.

[0011] Preferably, the power supply module includes a battery and a step-down module, wherein the step-down module includes a transformer and an adapter port; the battery is connected to the transformer, the adapter port and the solar panel; the temperature acquisition module, the video monitoring module and the communication module are all connected to the adapter port; and the step-down module is connected to the main control module.

[0012] Preferably, the main control module includes a main control MCU, a first communication conversion module, and a second communication conversion module. The main control MCU is connected to the first communication conversion module and the second communication conversion module, and the first communication conversion module and the second communication conversion module are connected to the temperature monitoring module. It accesses the cloud server through a serial communication module, which is connected to the main control MCU.

[0013] Preferably, the system's early warning application process includes the following steps:

[0014] S1. Use the Z-score algorithm to monitor outliers and calculate the Z-score for the data values ​​monitored by the temperature acquisition module;

[0015] S2. Perform linear interpolation on the removed outliers to complete the removed outliers and ensure data continuity;

[0016] S3. Use the extended Kalman filter algorithm to improve the accuracy of temperature measurement, and perform fusion calculation on the data values ​​after monitoring outliers by Z-score and after linear interpolation.

[0017] S4. Use the Kriging interpolation algorithm to predict the temperature of unknown points in the orchard. Perform spatial structure analysis and modeling on the data after Z-score monitoring outliers, linear interpolation and extended Kalman filtering fusion calculation, so as to use the temperature data of known points and their spatial correlation to predict the temperature information of unknown points in the orchard.

[0018] S5. Calculate the average of the measured data to obtain the representative temperature of this layer in the entire orchard; select temperature values ​​at different heights at the same time, where the temperature of the upper layer > the temperature of the middle layer > the temperature of the lower layer, and the temperature reaches the frost critical temperature of different fruit trees.

[0019] S6. When monitoring indicates that frost has occurred, intervention should be carried out based on the orchard monitoring results.

[0020] Preferably, in S1, the input data is processed using Z-score calculation, and the calculation formula is as follows:

[0021]

[0022] In the formula, x i The data is monitored by the temperature monitoring module;

[0023] μ represents the average temperature value at the same altitude.

[0024] σ is the standard deviation

[0025] Preferably, the specific steps of S2 are as follows:

[0026] S21. Determine the known point, that is, the temperature value corresponding to the left-hand data x0 of the discarded value is y0.

[0027] The temperature value corresponding to the data x1 on the right is y1.

[0028] S22. Calculate the slope m, which is the rate of temperature change between two known points. The formula is:

[0029]

[0030] S23. The formula for applying linear interpolation is:

[0031] y = y0 + m·(x - x0),

[0032] In the formula, y0 is the temperature value of the known point x0;

[0033] m is the slope, representing the rate of change of temperature with respect to x;

[0034] (x-x0) is the offset between the point to be removed x and the known point x0;

[0035] y is the calculated rejection point temperature value.

[0036] Preferably, the specific steps of S3 are as follows:

[0037] S31. Construct the state equation x k =f(x) k-1 ,u k )+w k ;

[0038] In the formula, x k Let f(x) be the temperature at time k. k-1 ,uk) is the state transition function;

[0039] u k External disturbances or system dynamics;

[0040] w k This is process noise;

[0041] S32: Construct the measurement equations, where each sensor measures the value z. k The sensor output, z, is described using a nonlinear measurement model. k =h(x k )+v k ;

[0042] S33: Initial state estimation, setting the initial state estimate Given the initial error covariance P0, use the state equation to predict the state at the current time. and error covariance

[0043] S34: Update step, using sensor data to update state estimate and error covariance, and calculate Kalman gain;

[0044] S35: Update the temperature estimate and error covariance based on the Kalman gain.

[0045] Preferably, the specific steps of S4 are as follows:

[0046] S41: Calculation of the variogram:

[0047] S42: Select a spherical model to fit the variogram;

[0048] S43: Solution Formula The temperature estimate of the unknown point x0 is obtained.

[0049] An intervention method for fruit tree canopy frost uses a drone airflow disturbance strategy: the high-speed rotation of the propeller generates a strong airflow, which promotes full mixing of the upper and lower layers of air; the warm air in the upper layer accelerates downward and forms convection with the cold air in the near-ground frost-threatened area, increasing the near-ground temperature. The intervention height is 7-12m, and the intervention frequency is 2 minutes / time, which can alleviate the damage of frost to crops and reduce agricultural losses.

[0050] Smoke generation method: After a frost warning, smoke piles are reasonably arranged and ignited in the field to reduce heat loss from the ground, increase atmospheric back radiation, and raise the near-ground temperature by using smoke.

[0051] High-rise anti-frost fan disturbance method: Install fans at high altitudes around farmland and turn them on when signs of frost appear to blow warm air from the upper layer to the ground, accelerate vertical air exchange, increase near-ground temperature, and take multiple measures to protect the crop growth environment and reduce losses.

[0052] Beneficial effects:

[0053] This invention uses an early warning algorithm to analyze the timing of frost occurrence based on orchard temperature changes and issue timely alerts. At the same time, the video monitoring module can observe the frost situation on the tree canopy in real time. Combined with intervention measures such as drone disturbance, it can effectively reduce the damage of frost to fruit trees. The system can accurately identify frost risks, reduce fruit yield reduction and fruit quality decline caused by frost, thereby reducing economic losses for fruit farmers and ensuring the economic benefits of orchards. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of a fruit tree canopy frost monitoring system.

[0055] Figure 2 A schematic diagram of the internal structure of a canopy frost monitoring system for fruit trees.

[0056] Figure 3 A flowchart of an early warning method for an orchard frost monitoring system;

[0057] Figure 4 A schematic diagram illustrating the spatial structure analysis and modeling of outlier data monitored by Z-scores;

[0058] Figure 5 A flowchart for the decision-making process triggered by drone intervention operations;

[0059] In the diagram: 1. Surveillance camera pan-tilt unit; 2. Solar panel; 3. Chassis; 4. Mounting bracket; 5. Temperature sensor; 6. Main control MCU; 7. First communication conversion module; 8. Second communication conversion module; 9. Adapter port; 10. Serial communication module; 11. Transformer; 12. Battery. Detailed Implementation

[0060] The present solution will now be described in conjunction with the accompanying drawings and specific embodiments.

[0061] A fruit tree canopy frost monitoring and early warning system includes a temperature acquisition module, a main control module, a power supply module, a communication module, a video monitoring module, and a monitoring module.

[0062] The main control module is connected to the temperature acquisition module, power supply module, and communication module respectively. The power supply module provides power support for the temperature acquisition module, main control module, and communication module. The main control module collects temperature data from the temperature acquisition module and sends it to the monitoring module through the communication module. The monitoring module analyzes the frost occurrence period based on the orchard temperature values ​​and issues an alarm. The video monitoring module observes the frost situation of the orchard canopy in real time. The temperature acquisition module is suspended at different heights in the canopy, including H1: 0.5m, H2: 1.5m, and H3: 2.7m, with three temperature sensors suspended at each height level.

[0063] like Figure 1 As shown, the video surveillance module includes a monitoring camera pan-tilt unit 1, which is mounted on the top of a fixed bracket 4. A solar panel 2 and a chassis 3 are also fixedly mounted on the surface of the fixed bracket 4. The temperature acquisition module, main control module, power supply module, and communication module are located inside the chassis 3.

[0064] like Figure 2 As shown, the power supply module includes a battery 12 and a step-down module, wherein the step-down module includes a transformer 11 and an adapter port 9; the battery 12 is connected to the transformer 11, the adapter port 9 and the solar panel 2; the temperature acquisition module, the video monitoring module and the communication module are all connected to the adapter port 9; the step-down module is connected to the main control module; the main control module includes a main control MCU 6, a first communication conversion module 7 and a second communication conversion module 8; the main control MCU 6 is connected to the first communication conversion module 7 and the second communication conversion module 8; the first communication conversion module 7 and the second communication conversion module 8 are connected to the temperature monitoring module; and it is connected to the cloud server through a serial communication module 10, which is connected to the main control MCU 6.

[0065] The monitoring module uses data-level fusion algorithms and decision-level fusion algorithms to monitor changes in canopy temperature. The dataset comes from the temperature values ​​collected by the temperature acquisition module of the main control unit. The data-level fusion algorithm is mainly used to accurately calculate temperature values, remove outliers, and calculate the temperature at different locations at the same height in the orchard.

[0066] like Figure 3 As shown, an early warning method for an orchard frost monitoring system includes the following steps:

[0067] S1. Use the Z-score algorithm to monitor outliers and calculate the Z-score for the data values ​​monitored by the temperature acquisition module;

[0068] Furthermore, the data-level fusion algorithm removes outliers using Z-score monitoring. Z-score calculations are performed on the data values ​​monitored by the temperature monitoring module, primarily including:

[0069]

[0070] Where, x i The data is monitored by the temperature monitoring module;

[0071] μ represents the average temperature value at the same altitude.

[0072] σ is the standard deviation

[0073] The Z-score of each data point in the dataset is calculated and compared precisely with a preset threshold. If the Z-score is higher than the upper limit of the threshold or lower than the lower limit of the threshold, the data point is determined to be an outlier.

[0074] S2. Perform linear interpolation on the removed outliers to complete the removed outliers and ensure data continuity;

[0075] Furthermore, the data-level fusion algorithm ensures data continuity, and the completion of data for removed points mainly includes:

[0076] Determine the known points, namely, the temperature value corresponding to the left-hand data x0 of the discarded value is y0, and the temperature value corresponding to the right-hand data x1 is y1.

[0077] The slope m is the rate of temperature change between two known points, calculated using the following formula:

[0078]

[0079] The formula for applying linear interpolation is:

[0080] y = y0 + m·(x - x0),

[0081] In the formula, y0 is the temperature value of the known point x0;

[0082] m is the slope, representing the rate of change of temperature with respect to x;

[0083] (x-x0) is the offset between the point to be removed x and the known point x0;

[0084] y represents the calculated rejection point temperature.

[0085] S3. Use the extended Kalman filter algorithm to improve the accuracy of temperature measurement, and perform fusion calculation on the data values ​​after monitoring outliers by Z-score and linear interpolation;

[0086] Furthermore, the data-level fusion algorithm accurately calculates temperature values ​​using extended Kalman filtering. The extended Kalman filtering calculation is performed on the data values ​​monitored by the temperature monitoring module, mainly including:

[0087] Construct the state equations:

[0088] x k =f(x) k-1 ,u k )+w k ,

[0089] In the formula, x k Let f(x) be the temperature at time k. k-1 (uk) is the state transition function;

[0090] u k External disturbances or system dynamics;

[0091] w k This is process noise;

[0092] Construct a measurement equation and use a nonlinear measurement model to describe the sensor output:

[0093] z k =h(x k )+v k ,

[0094] In the formula, z k For the measurements from all sensors, z k [z1………z i ];

[0095] h(x k This represents the relationship between temperature and sensor measurements; v k It measures noise.

[0096] The steps of extended Kalman filtering are as follows:

[0097] Initial state estimation, setting the initial state estimate Given the initial error covariance P0, use the state equation to predict the state at the current time. and error covariance

[0098]

[0099] In the formula, F k-1 Let be the Jacobian matrix of the state transition function with respect to the state;

[0100] Q k The process noise covariance matrix;

[0101] The update steps utilize sensor data to update the state estimate and error covariance, and calculate the Kalman gain:

[0102]

[0103] In the formula, H k R is the Jacobian matrix of the measurement equation with respect to the state. k To measure the noise covariance matrix;

[0104] Update the temperature estimate and error covariance based on the Kalman gain:

[0105]

[0106] S4. Use the Kriging interpolation algorithm to predict the temperature at unknown points in the orchard. Perform spatial structure analysis and modeling on the data after Z-score outlier monitoring, linear interpolation, and extended Kalman filtering fusion calculations. Figure 4 As shown; thus, by utilizing the temperature data of known points and their spatial correlation, the temperature information of unknown points in the orchard can be predicted.

[0107] Furthermore, the data-level fusion algorithm calculates the temperature at different locations within the same temperature layer in the orchard using the Kriging interpolation algorithm. This involves applying Kriging interpolation to the remaining data after outliers have been removed to calculate the temperature at different locations within that temperature layer. This mainly includes:

[0108] Calculation of variogram:

[0109]

[0110] In the formula, r* represents the experimental variability function value when the lag distance is h, which calculates the average temperature difference between sample point pairs at a distance of h.

[0111] h: lag distance, i.e., the distance between pairs of sample points;

[0112] N(h): The number of sample point pairs with a distance of h. A sample point pair is a combination of two sample points that are spatially separated by a distance of h.

[0113] Z(x i ):x i Temperature value at location;

[0114] A spherical model was chosen to fit the variogram:

[0115]

[0116] In the formula, r(h): the theoretical value of the variogram when the lag distance is h;

[0117] h: Similar to the lag distance in the experimental variogram, representing the distance between sample points;

[0118] C0: Gold nugget value, reflecting measurement error and random variation smaller than the sampling scale;

[0119] C: Sill value, representing the total degree of variation of a variable when there is no longer spatial autocorrelation between sample points outside the range;

[0120] a: Variable range, indicating that there is spatial autocorrelation between sample points within this distance range;

[0121] Solving for temperature at unknown locations:

[0122] Suppose we want to estimate the temperature at an unknown point x0, the Kriging estimate is... It is a linear combination of the known temperature values ​​Z(x0) of the sample points:

[0123]

[0124] in, Kriging estimate of temperature at unknown point x0;

[0125] n: The number of known sample points;

[0126] λ i Weighting coefficients determine the estimated value of each known sample point Z(x0) for the unknown point. Contribution;

[0127] Z(x0): Temperature value of the i-th known sample point;

[0128] λ i The weights are calculated by solving the Kriging equations to ensure the unbiasedness and minimum variance of the estimates.

[0129]

[0130] Where, λ j Weighting coefficients: Similar to λ in the Kriging estimate formula. i They have the same meaning and are used to determine the contribution of the i-th known sample point to the estimate of the unknown point.

[0131] r(x i -x0): The variogram value between sample points xi and xj, reflecting the degree of spatial autocorrelation between the sample points. The closer the distance, the higher the variogram value. i The smaller the -x0) value, the stronger the correlation;

[0132] u: Lagrange multiplier, used to solve for the weighting coefficients under the condition of unbiasedness;

[0133] r(x i-x0): The variogram value between sample point xi and unknown point x0, which reflects the degree of spatial autocorrelation between the known sample point and the unknown point, and is used to determine the magnitude of the influence of the known sample point on the estimation of the unknown point;

[0134] Unbiasedness condition: The Kriging estimate is unbiased, that is, the expected value of the estimate is equal to the true value. This condition ensures that the estimation process will not systematically overestimate or underestimate the attribute values ​​of unknown points.

[0135] Solving the above system of linear equations yields the weighting coefficients λ. i And the Lagrange multiplier u, the weight system λ obtained by solving i Substituting into the Kriging estimation formula The temperature estimate of the unknown point x0 is obtained.

[0136] S4. Calculate the average of the measured data to obtain the representative temperature of this layer in the entire orchard; select temperature values ​​at different heights at the same time, where the temperature of the upper layer > the temperature of the middle layer > the temperature of the lower layer, and the temperature reaches the frost critical temperature of different fruit trees.

[0137] The average of the measured data is calculated to obtain the representative temperature for that layer height. Temperature values ​​at different heights at the same time are selected. When the temperature of the upper layer > the middle layer > the bottom layer, and the temperature reaches the frost critical temperature for different fruit trees, for example, the frost critical temperature for apple buds is -2.8 to -3.8℃, the frost critical temperature for flowering is -1.6 to -2.2℃, and the frost critical temperature for young fruit is -1.1 to -2.2℃; the frost critical temperature for peach buds is -2.8 to -3.9℃, and the frost critical temperature for flowering is below 0℃; the frost critical temperature for pear buds is -3.9 to -8℃, the frost critical temperature for flowering is -2.5 to -3.4℃, the frost critical temperature for fully open flowers is -1.5 to -2℃, and the frost critical temperature for young fruit is -0.7 to -2.5℃, then the probability of frost occurrence is determined.

[0138] S5. When monitoring determines that frost has occurred, intervention should be carried out based on the orchard monitoring results;

[0139] A strategy involving airflow disturbance using drones can be employed. High-speed propeller rotation generates a strong airflow, promoting thorough mixing of upper and lower air layers. The warm upper air accelerates downwards, forming convection currents with cold air in frost-threatened areas near the ground, thus increasing near-surface temperatures. An intervention height of 7-12 meters and a frequency of 2 minutes per intervention are recommended to mitigate frost damage to crops and reduce agricultural losses. Alternatively, a smoke-generating method can be used. After a frost warning, smoke-generating piles can be strategically placed and ignited in the field to reduce heat loss from the ground, increase atmospheric back radiation, and raise near-surface temperatures. Another approach is to use high-altitude frost-prevention fan disturbance. Fans installed at high points around farmland are activated when frost signs appear, blowing warm upper air towards the ground, accelerating vertical air exchange, and increasing near-surface temperatures. These multiple measures aim to protect the crop growing environment and minimize losses.

[0140] The embodiments described above merely illustrate implementation methods of the present invention and should not be construed as limiting the scope of the invention patent, nor as imposing any form of limitation on the structure of the present invention. It should be noted that those skilled in the art can make various changes and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention.

Claims

1. A fruit tree canopy frost monitoring and early warning system, characterized in that, The system includes the following modules: a temperature acquisition module, a main control module, a power supply module, a communication module, a video monitoring module, and a monitoring module. The main control module is connected to the temperature acquisition module, the power supply module, and the communication module, respectively. The power supply module provides power to the temperature acquisition module, the main control module, and the communication module. The main control module acquires temperature data from the temperature acquisition module and sends it to the monitoring module through the communication module. The monitoring module analyzes the frost occurrence period based on the orchard's temperature values ​​and issues an alarm. The video monitoring module observes the frost situation in the orchard's tree canopy in real time. The temperature acquisition module is suspended at different heights in the tree canopy, including H1: 0.5m, H2: 1.5m, and H3: 2.7m, with three temperature sensors suspended at each height level.

2. The fruit tree canopy frost monitoring and early warning system according to claim 1, characterized in that: The video monitoring module includes a monitoring camera pan-tilt unit (1), which is installed on the top of a fixed bracket (4). A solar panel (2) and a chassis (3) are also fixedly installed on the surface of the fixed bracket (4). The temperature acquisition module, main control module, power supply module, and communication module are located inside the chassis (3).

3. The fruit tree canopy frost monitoring and early warning system according to claim 2, characterized in that: The power supply module includes a battery (12) and a step-down module. The step-down module includes a transformer (11) and an adapter port (9). The battery (12) is connected to the transformer (11), the adapter port (9), and the solar panel (2). The temperature acquisition module, the video monitoring module, and the communication module are all connected to the adapter port (9). The step-down module is connected to the main control module.

4. The fruit tree canopy frost monitoring and early warning system according to claim 3, characterized in that: The main control module includes a main control MCU (6), a first communication conversion module (7) and a second communication conversion module (8). The main control MCU (6) is connected to the first communication conversion module (7) and the second communication conversion module (8). The first communication conversion module (7) and the second communication conversion module (8) are connected to the temperature monitoring module. The module is connected to the cloud server through a serial communication module (10). The serial communication module (10) is connected to the main control MCU (6).

5. The early warning method of the fruit tree canopy frost monitoring and early warning system according to claim 4, characterized in that, The early warning method includes the following steps: S1. Use the Z-score algorithm to monitor outliers and calculate the Z-score for the data values ​​monitored by the temperature acquisition module; S2. Perform linear interpolation on the removed outliers to complete the removed outliers and ensure data continuity; S3. Use the extended Kalman filter algorithm to improve the accuracy of temperature measurement, and perform fusion calculation on the data values ​​after monitoring outliers by Z-score and after linear interpolation. S4. Use the Kriging interpolation algorithm to predict the temperature of unknown points in the orchard. Perform spatial structure analysis and modeling on the data after Z-score monitoring outliers, linear interpolation and extended Kalman filtering fusion calculation, so as to use the temperature data of known points and their spatial correlation to predict the temperature information of unknown points in the orchard. S5. Calculate the average of the measured data to obtain the representative temperature of this layer in the entire orchard; S6. When monitoring indicates that frost has occurred, intervention should be carried out based on the orchard monitoring results.

6. The method for monitoring and early warning of frost damage to fruit tree canopies according to claim 5, characterized in that: In S1, the input data is processed using Z-score calculation, and the calculation formula is as follows: Where, x i The data is monitored by the temperature monitoring module; μ represents the average temperature value at the same altitude. σ is the standard deviation 7. The method for monitoring and early warning of frost damage to fruit tree canopies according to claim 5, characterized in that: The specific steps for S2 are as follows: S21. Determine the known points, that is, the temperature value corresponding to the left end data x0 of the discarded value is y0, and the temperature value corresponding to the right end data x1 is y1. S22. Calculate the slope m, which is the rate of temperature change between two known points. The formula is: S23. The formula for applying linear interpolation is: y = y0 + m·(x - x0), In the formula, y0 is the temperature value of the known point x0; m is the slope, representing the rate of change of temperature with respect to x; (x-x0) is the offset between the point to be removed x and the known point x0; y represents the calculated rejection point temperature.

8. The method for monitoring and early warning of frost damage to fruit tree canopies according to claim 5, characterized in that: The specific steps for S3 are as follows: S31. Constructing the state equations: x k =f(x k-1 ,u k )+w k , In the formula, x k Let f(x) be the temperature at time k. k-1 u k ) is the state transition function, u k For external disturbances or system dynamics, w k This is process noise; S32: Construct the measurement equation, using a nonlinear measurement model to describe the sensor output, i.e., z. k =h(x k )+v k In the formula, z k For the measurements from all sensors, z k [z1………z i ], h(x k ) represents the relationship between temperature and sensor measurements, v k It measures noise; S33: Initial state estimation, setting the initial state estimate Given the initial error covariance P0, use the state equation to predict the state at the current time. and error covariance S34: Update step, using sensor data to update state estimate and error covariance, and calculate Kalman gain; S35: Update the temperature estimate and error covariance based on the Kalman gain.

9. A method for monitoring and early warning of frost damage to fruit tree canopies according to claim 5, characterized in that, The specific steps for S4 are as follows: S41: Calculation of the variogram; S42: Select a spherical model to fit the variogram; S43: Solution Formula The temperature estimate of the unknown point x0 is obtained, where, Let x0 be the Kriging estimate of the temperature at the unknown point, n be the number of known sample points, and λ be the Kriging value. i The weighting coefficients determine the estimated value of each known sample point Z(x0) for the unknown point. Contribution, Z(x0) is the temperature value of the i-th known sample point.

10. A method for intervening in frost damage to the canopy of fruit trees, characterized in that, The airflow disturbance strategy using drones: the high-speed rotation of the propellers generates a strong airflow, which promotes full mixing of the upper and lower layers of air; the warm air in the upper layer accelerates downward and forms convection with the cold air in the near-ground frost-threatened area, which increases the near-ground temperature. The intervention height is 7-12m and the intervention frequency is 2 minutes / time, which can alleviate the damage of frost to crops and reduce agricultural losses. Smoke generation method: After a frost warning, smoke-generating piles are reasonably arranged and ignited in the field to reduce heat loss from the ground, increase atmospheric back radiation, and raise the near-ground temperature by using smoke. High-rise anti-frost fan disturbance method: Install fans at high altitudes around farmland and turn them on when signs of frost appear to blow warm air from the upper layer to the ground, accelerate vertical air exchange, increase near-ground temperature, and take multiple measures to protect the crop growth environment and reduce losses.

Images