Plant source negative oxygen ion intelligent controlled release method

Abstract

The application discloses a plant source negative oxygen ion intelligent control release method, and belongs to the technical field of intelligent regulation and control of agricultural ecological environment, and comprises the following steps of remote sensing data acquisition, vigor parameter extraction, growth state evaluation, control release strategy matching, intelligent release execution, feedback and optimization, and the like. The application does not need a ground control point, a spatial reference is directly provided by POS, image point clouds are directly generated, and the LAI increment can be increased through 5 automatic iterations, manual test parameters are not needed, a strategy is self-evolved, a vigor evaluation model is used to realize intelligent control and release of plant source negative oxygen ions, and remote sensing control release integration is realized. The plant source negative oxygen ion control release process is fully automated, a feedback mechanism and strategy updating are used, and manual intervention is not needed.

Description

Technical Field

[0001] This invention discloses an intelligent regulation technology for agricultural ecological environment, and in particular relates to an intelligent controlled release method for plant-derived negative oxygen ions. Background Technology

[0002] Negative oxygen ions improve air quality, promote plant photosynthesis, and enhance crop resistance. In recent years, plant-derived negative oxygen ions have become a research hotspot in agricultural environmental regulation due to their natural, safe, and environmentally friendly characteristics. However, existing methods for releasing negative oxygen ions are mostly static, timed, or artificially controlled, which presents several problems: they cannot achieve regionally differentiated regulation based on crop growth status; they lack real-time feedback mechanisms, making it difficult to dynamically optimize release strategies; and they have low integration with modern agricultural equipment, hindering large-scale application. With the development of UAV remote sensing technology, crop growth information can be acquired at high resolution and periodically, providing a foundation for intelligent regulation of negative oxygen ion release. Therefore, there is an urgent need for an intelligent controlled-release method that integrates remote sensing monitoring and plant-derived negative oxygen ion release to achieve precise regulation of the crop growth environment. Summary of the Invention

[0003] The purpose of this invention is to provide a smart controlled-release method for plant-derived negative oxygen ions in order to solve the above-mentioned problems.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a plant-derived negative oxygen ion intelligent controlled release method, comprising the following steps:

[0005] K1 Remote Sensing Data Acquisition: Equipped with a remote sensing system and a drone carrying a multispectral imaging system and a high-precision POS positioning module, the drone periodically flies at low altitude over the target farmland area according to a preset route to collect crop multispectral images containing multiple bands and the geographic coordinate information corresponding to each image.

[0006] K2 growth parameter extraction: Based on multispectral images, multiple vegetation indices are calculated, and key growth parameters such as crop height, leaf area index, and vegetation cover are retrieved by combining digital elevation models.

[0007] K3 Growth Status Assessment: Construct a crop growth assessment model, input growth parameters into the model, and output crop growth level. The model is trained based on historical data and implemented using machine learning algorithms.

[0008] K4 Controlled Release Strategy Matching: Based on the crop growth level, a preset plant-derived negative oxygen ion release strategy is matched. The strategy includes release concentration, release frequency, duration and area of ​​action.

[0009] K5 Intelligent Release Execution: Through plant-derived negative oxygen ion release devices deployed in the field, it receives control commands from drones or ground control stations to execute release tasks; the device includes a plant extract atomization module, a negative ion excitation module, a control module and a wireless communication module, supporting PWM or wireless remote control;

[0010] K6 Feedback and Optimization: After the release mission is completed, the UAV remote sensing flight is performed again to collect crop response data, evaluate the release effect, and optimize the controlled release strategy based on the feedback data to form a closed-loop control.

[0011] Preferably, in step K2, the vegetation index calculation includes, but is not limited to, the Normalized Difference Vegetation Index (NDVI), and the growth parameter inversion includes, but is not limited to, vegetation cover (FVC) and plant height. The parameter calculation is based on high-resolution remote sensing images after radiometric calibration, atmospheric correction, image stitching, and georegistration.

[0012] Preferably, in step K3, the crop growth assessment model is a classification or regression model built based on machine learning algorithms. The training data includes historical remote sensing images and ground-measured growth data including leaf area index (LAI), plant height, and vegetation coverage (FVC). The output of the crop growth assessment model is a crop growth level or growth score. The model supports adaptation to multiple crops and multiple growth stages, and supports online updates and transfer learning.

[0013] As a preferred option, the plant-derived negative oxygen ion release strategy is in the form of a preset strategy library. The strategy library is constructed based on crop type, growth stage, growth level, and environmental conditions including multi-dimensional parameters such as temperature, humidity, and light. It supports dynamic calling and combination, and the strategy content is stored in the form of rules or models at ground control stations or edge computing devices.

[0014] Preferably, the negative oxygen ion releasing device includes the following modules:

[0015] Plant extract atomization module: used to atomize plant-derived extracts into micron-sized droplets;

[0016] Negative ion excitation module: Employs high-voltage corona discharge or plasma excitation to enable atomized droplets to carry negative charges, forming negative oxygen ions;

[0017] Control module: Uses STM32 or ESP32 embedded chip, supports PWM control, timed tasks, and remote command parsing;

[0018] Communication module: Supports Wi-Fi, LoRa, 4G / 5G, and NB-IoT wireless communication methods for connecting to the drone ground station;

[0019] Power module: Supports solar power, lithium battery power or AC power, and has low power management and power monitoring functions.

[0020] As a preferred approach, feedback and optimization mechanisms include:

[0021] Response assessment indicators include plant height growth rate and vegetation cover rate (FVC) estimation.

[0022] Policy optimization algorithm: Reinforcement learning, Bayesian optimization or genetic algorithm are used to automatically optimize the concentration, frequency and region of release policy parameters;

[0023] Closed-loop control cycle: Remote sensing re-flight and evaluation are completed within 7 days after each mission release, and the strategy update cycle is once per reproductive period or once per month.

[0024] Compared with the prior art, the beneficial effects of the present invention are:

[0025] This invention requires no ground control points; the spatial reference is directly provided by the POS; image point clouds are directly generated; the LAI increment can be improved in 5 automatic iterations; no manual parameter testing is required; the strategy is self-evolving; and a growth assessment model is used to achieve intelligent control and release of plant-derived negative oxygen ions, realizing the integration of remote sensing control and release.

[0026] The controlled release process of plant-derived negative oxygen ions is fully automated, employing feedback mechanisms and strategy updates, requiring no human intervention. Attached Figure Description

[0027] Figure 1 This is a flowchart of the overall model of the present invention;

[0028] Figure 2 A flowchart illustrating the process of acquiring high-resolution remote sensing images after image stitching and georegistration through single-pixel back projection;

[0029] Figure 3 A flowchart of the crop growth assessment model;

[0030] Figure 4 A flowchart illustrating the process of obtaining LAI through experimentally trained random forest regression and pixel-by-pixel prediction;

[0031] Figure 5 The flowchart for updating and iterating a Gaussian process. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. In this description, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0033] A smart controlled release method for plant-derived negative oxygen ions, such as Figure 1 As shown, it includes the following steps:

[0034] K1 Remote Sensing Data Acquisition: Equipped with a remote sensing system and a drone carrying a multispectral imaging system and a high-precision POS positioning module, the drone periodically flies at low altitude over the target farmland area according to a preset route to collect crop multispectral images containing multiple bands and the geographic coordinate information corresponding to each image.

[0035] The route generation model is as follows:

[0036] Lateral spacing (distance between adjacent flight paths) dw = w²·(1-Sp)

[0037] Where w2 is the image frame ground width, which refers to the coverage width of the image on the ground, and Sp is the lateral overlap rate, which refers to the overlap ratio of adjacent images in the lateral direction;

[0038] The heading-based photo interval (the distance between adjacent photo points on the same flight path) is dh = h²·(1-Op).

[0039] Where h2 is the image frame ground height, which refers to the coverage height of the image on the ground, and Op is the heading overlap rate, which refers to the overlap ratio of adjacent images on the same flight path in the heading direction;

[0040] Image width at ground level w2 = 2·a·tan(FOV / 2)

[0041] Where 'a' is the flight altitude, referring to the drone's flight altitude relative to the ground, and 'FOV' is the field of view, referring to the size of the camera lens's field of view.

[0042] Used to calculate the spacing of the drone flight path and the shooting spacing to ensure that the acquired images have appropriate overlap and meet the requirements of remote sensing monitoring and image processing;

[0043] The data acquisition hardware includes a five-panel wide-angle camera combined with an airborne GNSS / INS integrated navigation system PPOI-A13POS (Lide Space High-Precision Positioning and Attitude Determination System). During data acquisition, the five lenses are simultaneously triggered to provide centimeter-level exterior orientation elements for each image from the five-panel camera: geographic coordinates X, Y, Z and attitude angles ω, φ, κ.

[0044] K2 growth parameter extraction: Based on multispectral images, multiple vegetation indices are calculated, and key growth parameters such as crop height, leaf area index, and vegetation cover are retrieved by combining digital elevation models.

[0045] In step K2, vegetation index calculation includes, but is not limited to, Normalized Difference Vegetation Index (NDVI), and growth parameter inversion includes, but is not limited to, vegetation cover (FVC) and plant height. The parameter calculation is based on high-resolution remote sensing images after radiometric calibration, atmospheric correction, image stitching, and georegistration.

[0046] Radiometric calibration converts DN values ​​to radiance. The calibration method involves placing five 1m×1m grayscale boards (with reflectances of 7%, 17%, 33%, 36%, and 58%) on the test ground. After establishing DN, a system of linear equations for ρ is calculated to obtain the radiometric calibration of the entire image. Simultaneously captured grayscale boards are used as reflectance calibration boards to establish a linear relationship between radiance L and surface reflectance ρ. The calibration coefficients are then verified on-site.

[0047] ρ=π(L-Lp) / (τv·τs·Es·cosθ)

[0048] Lp is the path radiation, estimated by dark pixel DN or inverted by calibration plate; τv (direction: surface → sensor) and τs (direction: sun → surface) are atmospheric transmittance; Es is solar irradiance; θ is solar zenith angle, obtained by combining POS record UTC with image time astronomical calendar.

[0049] The high-resolution remote sensing image, after image stitching and georegistration, is obtained through single-pixel backprojection, as shown in the code below. Figure 2 As shown.

[0050] K3 Growth Status Assessment: Construct a crop growth assessment model, input growth parameters into the model, and output crop growth level. The model is trained based on historical data and implemented using machine learning algorithms.

[0051] By backprojecting each ground grid (X,Y,Z_DSM) on the DSM to all visible images through single-pixel backprojection, the weighted average reflectance is calculated. For the same ground pixel, only the image with the smallest Xc[2] is retained, which is the image closest to the sensor and participates in the weighting. Building / tree canopy occlusion is automatically eliminated for the calculation of Normalized Difference Vegetation Index (NDVI), plant height, and vegetation coverage (FVC). In step K3, as shown in the figure... Figure 3As shown, the crop growth assessment model is a classification or regression model built based on machine learning algorithms. The training data includes historical remote sensing images, and the ground-measured growth data includes leaf area index (LAI), plant height, and vegetation coverage (FVC). The crop growth assessment model outputs crop growth level (0, 1, 2, 3) or growth score (poor, medium, good, excellent). The model supports adaptation to multiple crops and multiple growth stages, and supports online updates and transfer learning.

[0052] K4 Controlled Release Strategy Matching: Based on the crop growth level, a preset plant-derived negative oxygen ion release strategy is matched. The strategy includes release concentration, release frequency, duration and area of ​​action.

[0053] K5 Intelligent Release Execution: Through plant-derived negative oxygen ion release devices deployed in the field, it receives control commands from drones or ground control stations to execute release tasks; the device includes a plant extract atomization module, a negative ion excitation module, a control module and a wireless communication module, supporting PWM or wireless remote control;

[0054] K6 Feedback and Optimization: After the release mission is completed, the UAV remote sensing flight is performed again to collect crop response data, evaluate the release effect, and optimize the controlled release strategy based on the feedback data to form a closed-loop control.

[0055] The plant-derived negative oxygen ion release strategy is in the form of a preset strategy library. The strategy library is constructed based on crop type, growth period, growth level, and environmental conditions including temperature, humidity, and light multi-dimensional parameters. It supports dynamic calling and combination. The strategy content is stored in the form of rules or models in ground control stations or edge computing devices.

[0056] The negative oxygen ion releasing device includes the following modules:

[0057] Plant extract atomization module: used to atomize plant-derived extracts into micron-sized droplets;

[0058] Negative ion excitation module: Employs high-voltage corona discharge or plasma excitation to enable atomized droplets to carry negative charges, forming negative oxygen ions;

[0059] Control module: Uses STM32 or ESP32 embedded chip, supports PWM control, timed tasks, and remote command parsing;

[0060] Communication module: Supports Wi-Fi, LoRa, 4G / 5G, and NB-IoT wireless communication methods for connecting to the drone ground station;

[0061] Power module: Supports solar power, lithium battery power or AC power, and has low power management and power monitoring functions;

[0062] Pine needles and mugwort (volume ratio 7:3) were distilled on water at a material-to-water ratio of 1:4. After distillation at 100°C and atmospheric pressure for 4 hours, the aqueous phase was collected using a separatory funnel to obtain the plant distillate. A negative ion excitation module used an ultrasonic transducer (Φ20mm, 113kHz) to break the plant distillate into 1–5μm droplets, forming a cold mist. A 6kV DC high voltage and a current <5mA were applied to the needle-ring electrode to form a corona layer at the needle tip. When the droplets passed through the corona region, they were ionized to generate O2-(H2O)n (n=3-6), i.e., plant-derived negative oxygen ions. These ions were then generated by a 12V axial flow fan for 0.8ms. -1 Upward airflow lifts and horizontally diffuses the cloud of negative oxygen ions from plants, thus releasing them.

[0063] Feedback and optimization mechanisms include:

[0064] Response assessment indicators include plant height growth rate and vegetation cover rate (FVC) estimation.

[0065] Policy optimization algorithm: Reinforcement learning, Bayesian optimization or genetic algorithm are used to automatically optimize the concentration, frequency and region of release policy parameters;

[0066] Closed-loop control cycle: Remote sensing re-flight and evaluation are completed within 7 days after each mission release. The strategy update cycle is once per reproductive period or once per month, using a Bayesian optimization strategy.

[0067] Negative growth potential response f(x) = -ΔLAI(x), x = [concentration, frequency, duration]

[0068] ΔLAI(x) is the LAI increment after the release strategy x takes effect.

[0069] LAI is obtained through 25-band reflectance, measured training of random forest regression, and pixel-by-pixel prediction. It can be directly used to optimize the closed loop of ΔLAI. The acquisition method is as follows: Figure 4 As shown.

[0070] Expected improvement EI(x) = σ(x)[γΦ(γ) + ϕ(γ)];

[0071] σ(x) is the standard deviation of ΔLAI(x), γ is the standardized improved value, Φ(γ) is the standard normal CDF, ϕ(γ) is the standard normal PDF, the iteration is 5 times per fertility period, and the Gaussian process is updated as follows: Figure 5 As shown.

[0072] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0073] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for intelligent controlled release of plant-derived negative oxygen ions, characterized in that, Includes the following steps: K1 Remote Sensing Data Acquisition: A drone equipped with a remote sensing system and a multispectral imaging system and a high-precision POS positioning module will periodically fly at low altitude over the target farmland area according to a preset route to collect crop multispectral images containing multiple bands and the geographic coordinate information corresponding to each image. K2 growth parameter extraction: Based on multispectral images after radiometric calibration, atmospheric correction, image stitching and georegistration, the normalized vegetation index (NDVI) is calculated, and key growth parameters of crops such as plant height, leaf area index (LAI) and vegetation cover (FVC) are retrieved by combining digital elevation model. K3 Growth Status Assessment: A crop growth assessment model based on machine learning algorithms is constructed. The training data includes historical remote sensing images, ground-measured leaf area index (LAI), plant height, and vegetation cover (FVC) data. The model outputs crop growth level or growth score. K4 Controlled Release Strategy Matching: Based on the crop growth level, a preset library of plant-derived negative oxygen ion release strategies is matched. The strategy library is built based on crop type, growth period, growth level, and environmental condition parameters, and supports dynamic calling and combination. The strategy content is stored in the form of rules or models at ground control stations or edge computing devices. K5 Intelligent Release Execution: Executes release tasks by receiving control commands from drones or ground control stations through plant-derived negative oxygen ion release devices deployed in the field; the device includes: The plant extract atomization module is used to atomize plant-derived extracts into micron-sized droplets of 1-5μm. The plant-derived extract is a mixed distillate of pine needles and artemisia at a volume ratio of 7:3, prepared by a process of material-to-water ratio of 1:4 and atmospheric distillation at 100°C for 4 hours. The negative ion excitation module adopts a needle-ring electrode structure with a 6kV DC high voltage and a discharge current of <5mA, so that the atomized droplets carry negative charges to form O2-(H2O)n negative oxygen ions, and diffuse them by generating an upward airflow of 0.8m / s through a 12V axial flow fan. The control module uses an STM32 or ESP32 embedded chip and supports PWM control and remote command parsing. The wireless communication module supports Wi-Fi, LoRa, or 4G / 5G access to the drone ground station; K6 Feedback and Optimization: Within 7 days after the release mission is completed, the UAV remote sensing flight is performed again to collect crop response data. The plant height growth rate and vegetation coverage (FVC) change are used as evaluation indicators. The release strategy parameters (concentration, frequency, and region) are automatically optimized using a Bayesian optimization algorithm to form a closed-loop control. Remote sensing re-fly and evaluation are completed within 7 days after each release mission. The strategy update cycle is once per growth period or once per month.

2. The intelligent controlled release method for plant-derived negative oxygen ions according to claim 1, characterized in that: In step K2, the growth parameter inversion uses single-pixel backprojection, which backprojects each ground grid on the digital surface model (DSM) onto all visible images, retaining only the image pixels closest to the sensor for weighted calculation, and automatically eliminating building / tree canopy occlusion.

3. The intelligent controlled release method for plant-derived negative oxygen ions according to claim 2, characterized in that: In step K3, the crop growth assessment model is a regression model built based on the random forest regression algorithm, which supports adaptation to multiple crops and multiple growth stages, and achieves online updates and transfer learning through incremental learning.

4. The intelligent controlled release method for plant-derived negative oxygen ions according to claim 3, characterized in that: In step K6, the objective function of the Bayesian optimization algorithm is the negative growth response function f(x) = -ΔLAI(x), where x = [concentration, frequency, duration], and ΔLAI(x) is the leaf area index increment after the release strategy is applied. The leaf area index LAI is obtained by predicting pixel by pixel through experimental training of the random forest regression model.

Images