A multi-sensor irrigation response system and method based on UAV spectral remote sensing

Abstract

This invention belongs to the field of UAV irrigation technology, specifically a multi-sensor irrigation response system and method based on UAV spectral remote sensing. The UAV acquires spectral images and infrared thermal images of tea gardens. Simultaneously, it obtains a series of growth environment information of tea trees from a meteorological station. After preprocessing and data extraction of the spectral and infrared thermal images, a tea tree water stress index is established by combining the tea tree growth information from the meteorological station. The irrigation amount for each area is determined according to the tea tree water requirement formula and the degree of tea tree water stress. Irrigation decision information is sent to each microcontroller node through a gateway, opening solenoid valves to realize irrigation. The amount of water infiltration and soil moisture status determine whether to continue irrigation. Irrigation stops when soil moisture saturation is reached. This invention facilitates accurate acquisition of soil moisture content information, facilitates calculation and decision-making on the water required for irrigation of different soil types, improves water use efficiency, and reduces water waste.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) irrigation technology, specifically a multi-sensor irrigation response system and method based on UAV spectral remote sensing. Background Technology

[0002] Chinese Patent CN201711478980.6 discloses a drone-based irrigation control system and method for mountain citrus orchards. The system includes multiple irrigation nodes and drones wirelessly connected to the irrigation nodes. Each irrigation node includes a first core processor and a soil temperature and humidity sensor module, a bistable solenoid valve module, a GPS module, and a first wireless communication module electrically connected to the first core processor. The drone includes a second core processor and a nine-axis motion sensor module, a brushless motor electronic speed controller module, a barometer module, a GPS electronic compass module, a second wireless communication module, and a wireless control module electrically connected to the second core processor. The first and second wireless communication modules are wirelessly connected. This invention provides a drone-based irrigation control system and method for mountain citrus orchards, applying drone technology to mountain citrus orchards, effectively reducing labor input in citrus cultivation and improving irrigation efficiency.

[0003] Existing drone irrigation systems and methods suffer from difficulties in accurately obtaining soil moisture information, calculating the amount of water required for irrigation of different soil types, and improving water use efficiency.

[0004] Therefore, a multi-sensor irrigation response system and method based on UAV spectral remote sensing is proposed to address the above problems. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, this paper addresses the problems of existing drone irrigation systems and methods, such as difficulty in accurately obtaining soil moisture information, difficulty in calculating and deciding the amount of water required for irrigation of different soil types, and difficulty in improving water use efficiency.

[0006] The technical solution adopted by this invention to solve its technical problem is:

[0007] Includes the following steps:

[0008] S1. The drone acquires spectral and infrared thermal images of the tea garden, and simultaneously obtains a series of environmental information about the tea trees through the meteorological station;

[0009] S2. A multi-sensor irrigation response system and method based on UAV spectral remote sensing, characterized in that: after preprocessing and extracting the spectral image and infrared thermal imaging image information, a tea tree water stress index is established by combining the tea tree growth information from the meteorological station, and the irrigation amount for each area is determined according to the tea tree water requirement formula and the degree of tea tree water stress.

[0010] S3. A multi-sensor irrigation response system and method based on UAV spectral remote sensing, characterized in that: irrigation decision information is sent to each microcontroller node through a gateway to open the solenoid valve and realize irrigation;

[0011] S4. A multi-sensor irrigation response system and method based on UAV spectral remote sensing, characterized in that: whether to continue irrigation is determined according to the real-time water infiltration rate and soil moisture status, and irrigation is stopped after the soil moisture is saturated.

[0012] Preferably, a drone equipped with a multispectral camera and infrared imaging is used to cruise and photograph the entire tea garden to obtain spectral images and infrared thermal images of the tea garden.

[0013] Preferably, a series of growth environment information of tea trees is obtained through a meteorological station and collected by multiple sensors. The growth environment information includes: maximum temperature, minimum temperature, maximum humidity, minimum humidity, light intensity, and wind speed.

[0014] Preferably, the image information acquired by the drone and the tea tree growth environment information acquired by the meteorological station are uploaded to the server in different regions. After being saved on the server, the data is transmitted to the intelligent decision-making system for data processing.

[0015] Preferably, the spectral image processing includes preprocessing and the establishment of a linear regression prediction model. Preprocessing mainly reduces image scattering and eliminates the image signal-to-noise ratio. Partial least squares regression is used to predict the moisture content of the tea tree canopy.

[0016] The steps for establishing the spectral image prediction model are as follows:

[0017] S1. First, perform multivariate scattering correction preprocessing on the spectral image.

[0018]

[0019]

[0020]

[0021] In the formula:

[0022] A: n×p dimensional calibration spectral data matrix; n: number of calibration samples; Average spectral vector; A i: Spectral vector of a single sample; m i : Relative offset coefficient; b i Relative offset;

[0023] S2. After the correction is completed, partial least squares regression (PLSR) is used to linearly model the spectral image.

[0024] First, decompose matrix S and matrix M.

[0025] S = ZK Z +A

[0026] M = RD Z +B

[0027] In the formula: S: spectral matrix; Z: score matrix of X; K: loading matrix of S; A: residual matrix of S; M: concentration matrix; R: score matrix of M; D: loading matrix of M; B: residual matrix of M;

[0028] S3. Next, regression analysis was performed on Z and R.

[0029] R = ZL

[0030] L=(Z z Z) -1 Z Z M;

[0031] S4. When predicting the spectrum of an unknown sample, first obtain the unknown sample's spectral matrix S_unknown and the unknown Z_unknown score matrix from K, and then calculate M_unknown.

[0032] M 未知 =Z 未知 LD

[0033] Preferably, the infrared image is processed as follows: first, the image is preprocessed to reduce noise and remove interfering backgrounds, then the canopy region is extracted, and the canopy temperature is finally obtained by fitting a curve using a normal distribution.

[0034] Preferably, a tea tree water stress index is established based on UAV canopy temperature data and meteorological station tea tree growth information, combined with tea canopy moisture content.

[0035] The formula for the water stress index is as follows:

[0036]

[0037] In the formula: CWSI is the crop water stress index; T c and T a These are the surface temperature of the canopy and the temperature of the air above it, respectively.

[0038] (T c -Ta ) II The canopy temperature difference is the temperature difference between the crop and the canopy under potential evaporation conditions, and it is the lower limit of the canopy temperature difference: °C; The canopy temperature difference under non-transpiration conditions is the upper limit of the canopy temperature difference: ℃

[0039] Among them, (T) c -T a ) II =A+Bp vpd A and B are the linear regression coefficients, p vpd The saturated water vapor pressure difference of air is expressed in Pa.

[0040]

[0041] Where Ta is the air temperature and RH is the air relative humidity, which can be obtained from the weather station's sensors.

[0042] The AB coefficient can be obtained by collecting multiple sets of Tc and Ta values.

[0043] p vpg The temperature is T a The air saturation vapor pressure and temperature at that time are T a Air saturated water vapor pressure difference at +A: Pa.

[0044] By combining the tea canopy temperature obtained by drones, the air temperature above the canopy obtained by meteorological stations, and the relative humidity information, the temperature difference between the tea canopy and the tea tree is calculated using a formula, and finally the water stress index of the tea tree is obtained.

[0045] Preferably, the steps for establishing the tea tree water requirement formula in S3 are as follows:

[0046] Crop evapotranspiration

[0047] In the formula, R n Net radiation to crop surface: MJ / (m2·d); G: soil heat flux: MJ / (m2·d); T: daily average air temperature at 2m altitude: ℃; u2: wind speed at 2m altitude: m / s; e s Saturated vapor pressure: kPa; e a Actual water vapor pressure: kPa;

[0048] Δ is the slope of the curve relating saturated water vapor pressure and air temperature: kPa / ℃; γ is the hygrometer constant: kPa / ℃.

[0049] Tea crop water requirement ET c =K c ET0

[0050] In the formula, ET cActual water consumption of crops: mm / d; K c ET0 is the crop coefficient; ET0 is the reference crop evapotranspiration: mm / d.

[0051] Preferably, the irrigation system of S4 includes: a host computer, a GPRS gateway, a Zigbee coordinator, irrigation nodes, moisture sensors, solenoid valves, etc. The host computer receives irrigation amount and duration information for the tea plantation area from the intelligent decision-making system, and sends soil moisture infiltration information and soil moisture status from the GPRS gateway to the intelligent decision-making system.

[0052] Preferably, when the host computer is deployed in the cloud or far from the irrigation system, the GPRS gateway is used to receive information and send control information to the Zigbee coordinator. The Zigbee coordinator then sends the control information to each irrigation node through Zigbee wireless communication technology.

[0053] Preferably, the steps for establishing the water infiltration model in S5 are as follows:

[0054] S1. The Kostiakov three-parameter infiltration model is selected, which is a purely empirical formula. The Kostiakov three-parameter infiltration formula is: i(t)=Kt -a +f0

[0055] In the formula: I(t) is the cumulative infiltration amount at time t: cm; t is the infiltration time: min; K is the infiltration coefficient, which refers to the cumulative infiltration amount after deducting the relatively stable infiltration rate at the end of the first unit time after the start of infiltration: cm; a is the infiltration index, which reflects the decay rate of soil infiltration capacity; both are empirical constants obtained by fitting measured data, determined by the basic physical and chemical properties of the soil and the initial infiltration conditions. f0 is the relatively stable soil infiltration rate, that is, the infiltration rate of saturated soil or the infiltration rate of unsaturated soil reaching the relatively stable stage under a unit soil potential gradient: cm / min. These formulas are widely used in practical work, especially the three-parameter model, which is widely applied to the simulation of surface irrigation infiltration processes.

[0056] S2. Obtaining Kostiakov's coefficients: First, various tests were conducted on the tea plantation soil, such as soil bulk density, soil moisture content, and soil organic matter composition. These factors all affect the coefficients; some with minor influences can be discarded. A T-test was established to obtain the structural characteristics of each coefficient. This process is not the core content of this invention; the parameter acquisition process is briefly described. Finally, the linear prediction model structure for each coefficient was determined.

[0057] K=β0+β1r0+β2θ1+β3ω1+β4ω2+β2G

[0058] α=β0+β1r0+β2r2+β3θ2+β4ω2+β5ω3+β6G

[0059] f0=β0+β1r0+β2ω2+β3ω3+β4G

[0060] In the formula: γ0 is the soil bulk density at 0–10 cm (g / cm³); γ1 is the soil bulk density at 0–20 cm (g / cm³); γ2 is the soil bulk density at 20–40 cm (g / cm³); θ1 is the soil volumetric moisture content at 0–20 cm; θ2 is the soil volumetric moisture content at 20–40 cm; ω1 is the mass fraction of sand at 0–20 cm; ω2 is the mass fraction of silt at 0–20 cm; ω3 is the mass fraction of clay at 0–20 cm; G is the organic matter content at 0–20 cm.

[0061] S3. Then, using the infiltration formula: i(t) = Kt -a +f0, calculates the cumulative soil infiltration at any given time.

[0062] The above experiments showed that soil moisture content has a strong correlation with K and α. Therefore, real-time monitoring of soil moisture content in different layers is of great significance for soil water infiltration models.

[0063] Preferably, the soil moisture state determination step in S5 is as follows:

[0064] The amount of water infiltration at different time periods is obtained by fitting the curve of the infiltration model. After each irrigation, the data fusion value of multiple sensors is monitored in real time to determine whether irrigation needs to continue. The entire irrigation infiltration process is set to 1 hour / cycle.

[0065] like Figure 3 As shown, the soil to be measured is divided into i groups vertically in one dimension according to the unit area, with each group spaced j cm apart. The soil depth is measured at i*j cm. It is also divided into k groups horizontally, each located at a different position. The specific implementation can be determined according to the actual soil environment.

[0066] This invention provides a method for obtaining water permeability through multi-sensor data fusion:

[0067] The steps for obtaining moisture infiltration data through multi-sensor data fusion are as follows:

[0068] S1. Because the humidity changes rapidly in the early stage of soil water infiltration, in order to make the input variable (humidity) simulate the current humidity state of soil water infiltration, it is fuzzified so that each input variable corresponds to a humidity state.

[0069] S2. Based on expert experience, establish fuzzy sets "100% moisture content", "80% moisture content", "60% moisture content", "40% moisture content", and "20% moisture content";

[0070] S3. Fuzzy set representation is:

[0071] F={μ F (u1) / u1, μ F (u2) / u2,……,μ F (u n ) / u n};

[0072] S4. Based on expert trials and experience, set the humidity range corresponding to each percentage of moisture content, and then use MATLAB to establish the membership function u;

[0073] S5. Establish a fuzzy comprehensive evaluation model. There are three elements to the comprehensive evaluation:

[0074] Factor set U = {u1, u2, u3, ..., u n}

[0075] It is a set of factors constituting the object of evaluation, which here represents n different areas at different depths in the soil where the humidity sensor is located.

[0076] U = {soil depth 3cm, soil depth 6m, ..., soil depth 30m}

[0077] Judgment set V = {v1, v2, v3, ..., v n}

[0078] It is a set of comments, representing the fuzzy state corresponding to the sensor, such as:

[0079] V = {20% moisture content, 40% moisture content, 60% moisture content, 80% moisture content, 100% moisture content} is used to collect soil moisture status corresponding to multi-point sensor data.

[0080] Single-factor judgment, that is, for a single factor u i The evaluation of (i = 1, 2, 3, ..., n) yields the fuzzy set (r) on V. i1 ,r i2 ,r i3 ,…,r im It is a fuzzy mapping from U to V, f:U→F(V); u i →(r i1 ,r i2 ,r i3 ,…,r im );

[0081] S6. A fuzzy mapping f can determine a fuzzy relation, called the evaluation matrix R:

[0082]

[0083] The evaluation matrix R is composed of the set F of all evaluations of a single factor. Since the factors are not necessarily of equal importance, they need to be weighted. For example, different moisture sensors at different depths will have different influence weights.

[0084] S7. Using the set F on U, A = (a1, a2, a3, ..., a... n The weights of each factor are represented by ), and their combination with the evaluation matrix R serves as a comprehensive evaluation of each factor, resulting in a fusion-based comprehensive evaluation model.

[0085]

[0086] Where A = (a1, a2, a3, ... a n );

[0087] R = (r ij ) n*m ,

[0088]

[0089] In addition, b j It is r ij The set B of the functions (i = 1, 2, 3, ..., n) is called the evaluation function;

[0090] S8. Data fusion from moisture sensors at various depths: Moisture and humidity data collected from multiple points are processed through fuzzification to transform them into the soil moisture and humidity status corresponding to a specific area. This ensures that the obtained data accurately reflects the soil moisture and humidity status under the current environment.

[0091] A multi-sensor irrigation response system based on UAV spectral remote sensing includes a meteorological station set up in a tea garden, a UAV equipped with a spectral camera and infrared imaging, a server, an intelligent decision-making system, a host computer that controls the irrigation system, and a microcontroller that executes irrigation decisions, or irrigation nodes.

[0092] Meteorological station: Obtain basic information about the tea tree's growing environment;

[0093] UAV: Acquire infrared thermal imaging information of spectral images of tea plantations;

[0094] Server: Receives and stores data from drones and weather stations, and sends the processed information to the intelligent decision-making system;

[0095] Intelligent decision-making system: It makes the best decisions based on the tea tree growth information using an expert experience database, and determines the best irrigation time based on data experience;

[0096] Host computer: Sends irrigation commands to irrigation nodes via GPRS network.

[0097] Preferably, the use of GPRS network enhances the adaptability and robustness of the network, expands the deployment range of the irrigation system, and, combined with Zigbee wireless network, facilitates the achievement of comprehensive network coverage.

[0098] The advantages of this invention are:

[0099] 1. As can be seen from the above technical solution, compared with the existing technology, the use of advanced UAV remote sensing technology can more accurately obtain information on the moisture content of tea leaves and the temperature of the tea leaf canopy. Compared with traditional instrument collection, it greatly improves the detection efficiency and reduces manual labor.

[0100] 2. Establishing a water stress index for tea tree irrigation aligns with the growth and development requirements of tea trees. Furthermore, monitoring tea trees in conjunction with soil conditions, which are crucial for their growth, represents a future research direction for precision irrigation. By establishing a soil moisture infiltration model and fusion of multi-sensor data to monitor soil moisture status, we can obtain the water requirements of tea trees for irrigation and monitor their water absorption through water infiltration. This approach not only maximizes the utilization of water resources but also allows for better control and monitoring of tea tree water information, providing a new research direction for precision irrigation.

[0101] 3. This invention rapidly acquires information on the moisture content of tea leaves and canopy temperature using drones, enhancing the accuracy of irrigation decisions. Utilizing a GPRS gateway, data transmission and irrigation can still be completed even in scenarios with poor network conditions. A soil moisture infiltration model is used to monitor soil moisture infiltration, and water and humidity status is obtained through multi-sensor data fusion, achieving the unification of irrigation and soil moisture and humidity status monitoring, thus saving water resources. Attached Figure Description

[0102] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0103] Figure 1 This is a flowchart of the steps of the multi-sensor irrigation response system and method using UAV spectral remote sensing according to the present invention;

[0104] Figure 2 This is a flowchart illustrating the workflow of the multi-sensor irrigation response system based on UAV spectral remote sensing of the present invention.

[0105] Figure 3 This is a one-dimensional diagram of soil per unit area according to the present invention.

[0106] The image shows: 1. Infrared thermal imaging; 2. Tea canopy image; 3. Drone; 4. Wireless network / GPRS gateway; 5. Weather station; 6. Information such as temperature, humidity, light intensity, and wind speed; 7. Server; 8. Intelligent decision-making system; 9. Host computer; 10. GPRS gateway; 11. Zigbee coordinator; 12. Irrigation node; 13. Solenoid valve; 14. Multiple moisture and humidity sensors. Detailed Implementation

[0107] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0108] Example 1

[0109] Please see Figures 1 to 3 As shown, a multi-sensor irrigation response system and method based on UAV spectral remote sensing, such as Figure 1 The multi-sensor irrigation response system and method based on UAV 3D spectral remote sensing shown includes the following steps:

[0110] First, a drone (3) is used to cruise and acquire overall spectral images of the tea garden. During the round trip, the drone lowers its altitude to a certain level to capture infrared thermal images of the tea garden (1), which are then transmitted to a server (7) via a wireless network. After image processing, the images are sent to the intelligent decision-making system (8). Additionally, a weather station (5) is deployed in the tea garden. The weather station (5) primarily acquires basic information about the tea tree growing environment, such as maximum and minimum temperatures, maximum and minimum humidity, light intensity, and wind speed. This information is also transmitted to the intelligent decision-making system (8).

[0111] After the spectral images captured by UAV 3 are preprocessed with multivariate scattering correction to improve the spectral signal-to-noise ratio, the partial least squares regression method is used to model the spectral images to predict the moisture content of tea leaves.

[0112] Processing of infrared thermal images: First, after preprocessing, interfering backgrounds are removed, the canopy region is extracted, and then the canopy temperature is extracted.

[0113] The canopy temperature of tea leaves is obtained by processing the infrared thermal image 1. Then, the degree of water stress of tea trees in a certain area can be calculated by the water stress index.

[0114] The moisture stress index model was established as follows: the canopy surface temperature was obtained by processing infrared thermal imaging 1. For a certain area, the canopy temperature was obtained from the mode of the normal distribution. The air temperature above the canopy was obtained by air temperature and humidity sensors arranged in meteorological station 5. The relative humidity and air temperature required for the saturated water vapor pressure difference were obtained by air temperature and humidity sensors arranged in meteorological station 5. Then, based on multiple sets of canopy surface temperature and water vapor pressure difference data, the relationship between the canopy air temperature difference and the air saturation difference was fitted. After that, the moisture stress index of a certain area could be calculated.

[0115] The intelligent decision-making system 8 calculates the water stress of tea trees in each region based on the crop stress index (CWSI) and the combined information on tea leaf moisture content, and predicts the required irrigation amount and irrigation time for each region.

[0116] The irrigation amount is calculated using the formulas for crop evapotranspiration and crop water requirement, and the irrigation duration is calculated based on the irrigation amount per unit time of the irrigation device.

[0117] The intelligent decision-making system 8 sends the required irrigation amount and irrigation time for each area to the irrigation host computer 9. The host computer 9 issues irrigation commands to each microcontroller, and the microcontroller controls the opening and closing of the solenoid valve 13 to complete the irrigation.

[0118] The intelligent decision-making system 8 transmits the required irrigation amount and irrigation time for each area to the host computer 9 via wireless or GPRS network. The host computer 9 sends the irrigation information to the GPRS gateway 10. The GPRS gateway 10 sends the data packet to the Zigbee coordinator 11 node. The Zigbee coordinator 11 sends the data to each microcontroller module. The microcontroller starts the solenoid valve 13 by changing the level of the corresponding port of the solenoid valve 13, and irrigates each area.

[0119] The aforementioned GPRS gateway 10 uses a SIM900A module to ensure normal network data transmission even in areas with poor network coverage, thus maintaining irrigation operations as intended, such as in rural or mountainous regions. Simultaneously, the GPRS network utilizes weather data from the meteorological station 5 to provide weather warnings and sends SMS messages to users via the GPRS module.

[0120] During irrigation, the real-time irrigation infiltration rate and the water and humidity status obtained by fusing data from multiple sensors are sent to the Zigbee coordinator 11 via Zigbee nodes. The Zigbee coordinator 11 then sends the data to the GPRS gateway 10. The GPRS gateway 10 sends the data to the intelligent decision-making system 8. Once the preset water saturation value is reached, irrigation is stopped. After the accumulated water infiltrates, the tea tree growth information is re-detected to ensure minimal water loss and optimal soil moisture content.

[0121] like Figure 2As shown, a multi-sensor irrigation response system based on UAV 3-spectral remote sensing includes: UAV 3, used to capture images 2 of the tea canopy and 1 of infrared thermal imaging in a tea garden; meteorological station 5, mainly collecting growth information related to tea trees; wireless network / GPRS gateway 4, mainly used to transmit data from UAV 3 and meteorological station 5 to server 7; server 7, used to acquire data from UAV 3 and meteorological station 5 in real time and save the data for a certain period of time; intelligent decision-making system 8, used to process multiple data and obtain irrigation decision information, composed of expert user experience; host computer 9, which receives data from intelligent decision-making system 8 and controls irrigation nodes 12; and GPRS gateway 10, used to send decision information down and upload nodes. Point sensor data; Zigbee coordinator 11 collects sensor information from nodes on one hand, and sends irrigation information transmitted by GPRS gateway 10 on the other hand; Irrigation node 12 is responsible for receiving irrigation information sent by the coordinator, and at the same time sending out the collected multi-sensor moisture and humidity information. The drone 3 quickly obtains the moisture content of tea leaves and canopy temperature information, which enhances the accuracy of irrigation decisions. Using GPRS gateway 10, even in scenarios with poor network conditions, data transmission can still be completed and irrigation can be implemented. A soil moisture infiltration model is used to monitor the amount of soil moisture infiltration, and the moisture and humidity status is obtained through multi-sensor data fusion, realizing the unification of irrigation and soil moisture and humidity status monitoring, saving water resources.

[0122] Working principle:

[0123] First, a drone 3 is used to cruise and acquire an overall spectral image of the tea garden. During the round trip, the drone lowers its flight altitude to a certain level and captures an infrared thermal image 1 of the tea garden. This image is then transmitted to a server 7 via a wireless network. After image processing, the image is transmitted to an intelligent decision-making system 8. Next, a weather station 5 is set up in the tea garden. The weather station 5 mainly acquires basic information about the tea tree growth environment, such as the highest and lowest temperatures, highest and lowest humidity, light intensity, and wind speed. This information is also transmitted to the intelligent decision-making system 8. The intelligent decision-making system 8 calculates the water stress of the tea trees in each area based on the crop stress index (CWSI) and the combined information on the moisture content of the tea leaves. It predicts the required irrigation amount and irrigation time for each area, calculates the irrigation amount using the crop evapotranspiration and crop water requirement formulas, and calculates the irrigation duration based on the irrigation amount per unit time of the irrigation device.

[0124] Next, the intelligent decision-making system 8 transmits the required irrigation amount and irrigation time for each area to the host computer 9 via wireless or GPRS network. The host computer 9 sends the irrigation information to the GPRS gateway 10, and the GPRS gateway 10 sends the data packet to the Zigbee coordinator 11 node. The Zigbee coordinator 11 sends the data to each microcontroller module. The microcontroller starts the solenoid valve 13 by changing the level of the corresponding port of the solenoid valve 13, and irrigates each area.

[0125] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes fall within the scope of the claimed invention.

Claims

1. A multi-sensor irrigation response method based on unmanned aerial vehicle spectral remote sensing, characterized in that, It comprises the following steps: S1. The unmanned aerial vehicle obtains the tea garden spectral image and infrared thermal image information, and simultaneously obtains a series of tea tree growth environment information through the weather station; S2. After the spectral image and infrared thermal imaging image information are preprocessed and data is extracted, the tea tree water stress index is established in combination with the tea tree growth information of the weather station, the irrigation amount of each region is determined according to the tea tree water requirement formula and the tea tree water stress degree; S3. The irrigation decision information is sent to each microcontroller node through the gateway to open the electromagnetic valve and realize irrigation; S4. Whether to continue irrigation is determined according to the real-time water infiltration amount and soil moisture state, and irrigation is stopped after the soil moisture humidity saturation is reached; The data is sent and uploaded through the GPRS network; the GPRS gateway adopts a SIM900A module, which is used to realize normal transmission of network data under the condition that the network coverage is poor, so as to achieve the purpose of normal operation of irrigation, such as in rural areas, mountainous areas and the like; meanwhile, the GPRS network realizes weather warning through the weather data of the weather station, sends a mobile phone short message to the user through the GPRS module, and the user monitors the tea garden environment information according to the short message, which is simple and convenient to operate without complicated interface operation, and is suitable for popularization to the general public; A soil water infiltration model is established to calculate the real-time infiltration amount of soil water, and the soil water infiltration model is established as follows: S1:Penetration model selected Kostiakov three-parameter infiltration model, which is a purely empirical formula, Kostiakov three-parameter infiltration formula is: i(t) = Kt -a +f0 In the formula, i(t) is the cumulative infiltration amount at t time: cm; t is the infiltration time: min; K is the infiltration coefficient, which refers to the cumulative infiltration amount at the end of the first unit time after the start of infiltration minus the relative stable infiltration rate: cm; a is the infiltration index, which can reflect the attenuation speed of the soil infiltration capacity; both are empirical constants fitted from the measured data, which are determined by the basic physicochemical properties of the soil and the initial conditions of the infiltration; f0 is the relative stable infiltration rate of the soil, that is, the infiltration speed of the saturated soil under the unit soil potential gradient or the infiltration speed of the unsaturated soil when the infiltration reaches the relative stable stage: cm / min; the application in the actual work is more common, especially the three-parameter model which is widely used in the simulation of the ground irrigation infiltration process; S2: Obtain the coefficients of Kostiakov First, carry out various tests on the tea garden soil, such as: Soil bulk density, soil moisture content, soil organic matter composition and the like, which will all have an impact on the coefficients, of course, some of the smaller impacts can be eliminated, the structure characteristics of the coefficients are obtained through the establishment of T test; finally, the linear prediction model structure of the coefficients is determined: K=β0+β1r0+β2θ1+β3W1+β4W2+β2G α=β0+β1r0+β2r2+β3θ2+β4W2+β5W3+β6G F0=β0+β1r0+β2W2+β3W3+β4G θ2 is the soil volume moisture content of 20-40 cm; w1 is the mass fraction of sand particles of 0-20 cm; w2 is the mass fraction of silt particles of 0-20 cm; w3 is the mass fraction of clay particles of 0-20 cm; G is the organic matter content of 0-20 cm; In the formula, γ0 is the soil bulk density at 0-10 cm: g / cm 3 ; γ1 is the soil bulk density at 0-20 cm: g / cm 3 ; γ2 is the soil bulk density at 20-40 cm: g / cm 3 ; θ1 is the soil volumetric water content at 0-20 cm; Tea tree water stress condition acquisition S3: Through the infiltration formula: i(t) = Kt -a +f0, the cumulative soil infiltration at any time is obtained. 2.The multi-sensor irrigation response method based on UAV spectral remote sensing of claim 1, wherein: The tea tree water stress condition acquisition step comprises: ​ S1: The water stress index formula is as follows: In the formula, CWST is a crop water stress index; Tc and Ta are a surface temperature of a canopy and an air temperature above the canopy, respectively, (Tc-Ta) II (Tc-Ta)u1 is a canopy-air temperature difference of the crop under a non-transpiration condition, which is an upper limit of the canopy-air temperature difference: ℃; S2: where (Tc - Ta) II = A + Bp vpd , A, B are linear regression coefficients, p vpd Difference in saturated water vapor pressure for air: Pa; Wherein, Ta is air temperature, RH is air relative humidity, which can be collected by weather station sensor; S4: By collecting multiple Tc, Ta values, the AB coefficient value can be obtained; S5: (Tc - Ta)u1 = A + Bp vpg ,p vpg Pa is the difference between the water vapor pressure of air saturated at temperature Ta and the water vapor pressure of air saturated at temperature Ta + A; S6: The tea leaf canopy temperature obtained by the unmanned aerial vehicle, the canopy air temperature above the weather station, and the air relative humidity information are substituted into the formula to obtain the tea leaf canopy air temperature difference, and finally the water stress index of tea tree is obtained. 3.The multi-sensor irrigation response method based on UAV spectral remote sensing of claim 1, wherein: The calculation of the water requirement, S1 : crop evapotranspiration In the formula, Rnis the net radiation on the crop surface, MJ / (m 2 ·d); G is the soil heat flux: MJ / (m 2 ·d); T is the daily average air temperature at a height of 2 m: °C, u2is the wind speed at a height of 2 m: m / s; e s is the saturated water vapor pressure: kPa; e a is the actual water vapor pressure: kPa. △ is the slope of the saturated water vapor pressure and air temperature relationship curve: kPa / ℃; γ is the humidity table constant: kPa / ℃; S2: Tea tree crop water requirement ETc=KcET0 In the formula, ETc is the actual water consumption of crops: mm / d; Kc is the crop coefficient; ET0 is the reference crop evapotranspiration: mm / d.

4. A multi-sensor irrigation response system based on unmanned aerial vehicle spectral remote sensing, characterized by, The method for performing the multi-sensor irrigation response system based on unmanned aerial vehicle spectral remote sensing according to any one of claims 1-3, comprising: the unmanned aerial vehicle cruising to obtain tea garden spectral image and infrared thermal image information; the weather station obtaining tea tree growth environment information, such as: maximum temperature, minimum temperature, maximum humidity, minimum humidity, illuminance, wind speed; obtaining soil moisture state by multi-sensor fuzzy fusion; The soil moisture state establishment step is as follows: S1: According to the fitting curve of the infiltration model, the water infiltration amount in different time periods is obtained, and after each irrigation, the data fusion value of the multi-sensor is monitored in real time, and then it is judged again whether irrigation needs to continue, and the whole irrigation infiltration process is set to 1h / period; S2: The soil to be measured is divided into i groups by unit area using one-dimensional vertical method, each group is j cm apart, the soil depth is i*j cm, and the horizontal division is k groups, which are respectively located at different positions. The specific implementation can be determined according to the actual soil environment; S3: Because the humidity changes rapidly in the early stage of soil water infiltration, in order to make the input variable (humidity) simulate the current soil water infiltration humidity state, the fuzzy processing is carried out, so that each input variable corresponds to a humidity state; S4: According to expert experience, the fuzzy sets "100% water content", "80% water content", "60% water content", "40% water content", "20% water content" are established; S5: The fuzzy set is represented as: F = { μ F (u1) / u1,μ F (u2) / u2,……μ F (u n ) / u n}; S6: According to expert experience, set the humidity interval corresponding to each percentage water content, and then use MATLAB to establish the membership function u; S7: A fuzzy comprehensive evaluation model is established, and there are three elements for comprehensive evaluation: The set of factors U = {u1, u2, u3,..., u n} It is a set composed of factors of the evaluation object, which represents n different areas of different depths of the humidity sensor in the soil here, U={soil depth 3cm, soil depth 6m, …, soil depth 30m} judgment set V={V1, V2, V3, …, Vn} It is a set composed of evaluation groups, which represents the fuzzy state corresponding to the sensor here, such as: V={20% water content, 40% water content, 60% water content, 80% water content, 100% water content}, which is used to collect the soil humidity state corresponding to the multi-point sensor data; Single factor judgment, namely the judgment of single factor u(i = 1, 2, 3, …, n) to get the fuzzy set (r i1 , i2 , i3 , im ) on V, which is a fuzzy mapping from U to V, f: U→F(V(u i →(r i1 , i2 , i3 , im ); S8: The fuzzy mapping f can determine a fuzzy relationship, which is called evaluation matrix R: The evaluation matrix R is composed of all the F sets of single-factor evaluation. Since the positions of various factors are not necessarily equal, weighting is needed for various factors, for example, the influence weight corresponding to each moisture sensor is different due to different depths. S9: The weight distribution of each factor is represented by the F set A=(a1, a2, a3, …an) on U, which is combined with the evaluation matrix R to obtain a comprehensive evaluation model after data fusion: AoR=B=(b1, b2, b3, …, bm) Where, A=(a1, a 2， a3, ...a n ); R＝(r ij ) n*m ， Furthermore, b j is a function of r ij ; i = 1, 2, 3,..., n, and call the set B the evaluation function. S10: The data fusion of moisture sensors at various depths is performed on the moisture and humidity data collected from multiple points, which is converted into the soil moisture and humidity state corresponding to a certain area through fuzzy processing, so that the obtained data can accurately reflect the soil moisture and humidity state under the current environment.

Images