A pollution-free green agricultural product standardized planting management method

Abstract

The application belongs to the technical field of green plant protection cultivation, and discloses a standardized planting and management method for pollution-free green agricultural products, which realizes multi-index screening of soil heavy metals, pesticide residues, microorganisms and water, soil and air, establishes a soil plant protection database, and applies passivation materials, compound beneficial microorganism groups and decomposed organic fertilizers to the soil according to pollution, continuous cropping and short soil fertility, and improves the stress resistance of seedlings by matching seed pretreatment; the disease and pest risk grades of plots are divided by weighting and quantifying five types of planting data, and a risk map is drawn, high, medium and low risk plots are differentially configured with pest-resistant varieties, intercropping, crop rotation and pest-repellent crops, and a hierarchical prepared prevention and control mechanism is matched, the planting environment is optimized from the aspects of farmland basement and field ecological layout; through the fusion of the Internet of Things, unmanned aerial vehicle remote sensing, pest trapping and meteorological data, the early latent characteristics of diseases and pests are captured through a deep learning model, and the development trend of diseases and pests is deduced in advance and divided into three levels of early warning.

Description

Technical Field

[0001] This invention belongs to the field of green plant protection and cultivation technology, specifically a standardized planting and management method for pollution-free green agricultural products. Background Technology

[0002] As green crop planting management systems become more widespread, conventional planting methods in the industry mainly rely on manual field management, regular pest and disease control, and conventional irrigation and fertilization. These methods only complete basic field management work according to general planting standards, which represents the production goal of farmland ecological protection and standardized planting management. However, existing conventional planting management technologies still have the following technical problems: Workers often rely on planting experience to spray chemical agents to control pests and diseases. Although this can quickly kill pests and diseases, long-term and high-frequency use of chemical pesticides will cause pesticide residues to accumulate in agricultural products. At the same time, chemical agents will seep into the soil and water, which will damage the ecological structure of farmland, cause soil microbial imbalance, water and soil pollution and other problems, damage the farmland planting environment and hinder the green and sustainable planting and production of crops.

[0003] Traditional planting relies on manual inspection and visual observation to judge the status of pests and diseases. Manual monitoring has obvious lag and limitations, and cannot identify the subtle disease symptoms in the early stage of pests and diseases. It can only make up for the lack of control after the large-scale outbreak of pests and diseases. Delayed control can easily lead to repeated outbreaks of pests and diseases, increased pesticide input, and exacerbation of excessive pesticide residues and water and soil pollution.

[0004] Traditional field planting operations are relatively scattered, and there are no standardized channels for storing and archiving data on soil treatment, water and fertilizer management, and pest and disease control. The standardization of planting operations varies, resulting in significant differences in the quality of agricultural products from different batches and plots. At the same time, there is no effective traceability for the entire planting process of agricultural products, making quality control and risk assessment difficult. Summary of the Invention

[0005] The purpose of this invention is to provide a standardized planting and management method for pollution-free green agricultural products, so as to solve one or more problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a standardized planting and management method for pollution-free green agricultural products, including a soil preparation stage, a layout optimization stage, a pest and disease early warning stage, a green prevention and control stage, a water and fertilizer precision control stage, a traceability management stage, and a post-harvest management stage; Furthermore, during the soil preparation stage, a portable X-ray fluorescence spectrometer was used to detect the content of heavy metals and available forms in the soil of the planting plot, a gas chromatography-mass spectrometry system was used to detect pesticide residue metabolites in the soil, and high-throughput sequencing technology was used to detect the soil microbial community structure and soil enzyme activity. At the same time, the soil pH, organic matter content, nitrogen, phosphorus and potassium available nutrient content, irrigation water quality, field atmospheric environmental quality were detected, and historical pest and disease breeding data of the plot were retrieved to establish a soil plant protection baseline database for the plot. For planting plots with mild heavy metal pollution, straw biochar, sepiolite and humic acid composite passivating agent are applied. For planting plots with severe continuous cropping obstacles, a compound beneficial microbial community of Bacillus subtilis, Trichoderma and Pseudomonas is quantitatively inoculated. For planting plots with poor soil fertility, well-rotted farmyard manure and bio-organic fertilizer are applied. Before sowing, crop seeds are pretreated by soaking in hot water and biological seed dressing.

[0007] Furthermore, in the layout optimization stage, five categories of basic rating data are collected: soil plant protection data, historical pest and disease outbreak frequency data, regional meteorological temperature and humidity data, crop susceptible pest and disease types data, and farmland ecological resistance data. The five categories of basic rating data are weighted and calculated, and quantitative scoring thresholds are set for plots with low pest and disease risk (0-40 points), medium pest and disease risk (41-70 points), and high pest and disease risk (71-100 points). Based on the calculation results, a pest and disease risk map of the plot is drawn to complete the risk level classification of the planting plot. In high-risk areas, plant highly resistant pest and disease varieties and densely plant insect-repellent crops between crop rows. In medium-risk areas, plant moderately resistant high-quality varieties and adopt different crop intercropping patterns. In medium- and low-risk areas, regularly carry out green manure rotation and water-dryland rotation. At the same time, a three-level prevention and control reserve mechanism is set up for high-risk areas, a two-level prevention and control reserve mechanism is set up for medium-risk areas, and basic field ecological maintenance is carried out for low-risk areas.

[0008] Furthermore, during the pest and disease early warning stage, IoT monitoring devices and pest and disease traps are deployed in the planting area. Drones equipped with hyperspectral imagers and multispectral cameras are used to regularly conduct aerial photography in the field, simultaneously collecting real-time monitoring data on soil temperature and humidity, soil moisture, and soil microbial community. The drone aerial photography also captures remote sensing data on micro-damage to crop leaves and abnormal crop growth. The pest and disease traps collect data on the breeding of insect eggs and larvae in the field, as well as real-time meteorological data on temperature, humidity, rainfall, and light intensity in the region. Data on the incubation period and breeding patterns of pests and diseases in the field are also retrieved. The data is fused and cleaned to construct a pest and disease early warning dataset and build a deep learning model for identifying the incubation period of pests and diseases. The characteristics of the incubation period of pests and diseases are taken as the characteristics of the incubation period of crop leaves, local rot of stems, and accumulation of trace insect eggs in the soil. The model parameters are trained by using these characteristics as training samples. The model combines real-time data to infer the breeding speed, spread range and outbreak probability of pests and diseases. The model input and output data fields are predefined: Model input is divided into image input and time-series environment input. Image input fields store the original pixel data of crop leaves and stems collected by UAV hyperspectral and multispectral cameras, which are uniformly scaled to a fixed pixel size before input. Time-series environment input fields store structured values ​​of soil temperature and humidity, soil moisture, insect egg statistics, ambient temperature and humidity, sunshine duration, and total precipitation. The model output fields are fixed as four values: pest and disease incubation level, daily pest and disease reproduction rate, disease spread area, and pest and disease outbreak probability. When connecting to the field scenario, the raw data collected by field IoT devices, pest and disease traps, and UAVs are standardized and converted, and missing values ​​are filled in before being mapped and matched to the model input fields. The model output fields are directly connected to the early warning push system through the JSON data interface.

[0009] Based on the probability of pest and disease outbreaks, Level 1, Level 2, and Level 3 early warnings are established. Corresponding early warning information is pushed out and matched with corresponding field control operation plans. After the early warning results are output, the corresponding level of field control operation process is automatically matched and started.

[0010] Furthermore, the green prevention and control stage combines the pest and disease warning level with the plot risk level to carry out graded prevention and control operations. The density of prevention and control deployment and the amount of pesticides are fine-tuned according to the pest and disease risk level of the plot. Under the first-level warning state, weeds and crop diseased residues in the field are regularly removed, and insecticidal lamps, sticky insect sticks and pheromone attractants are deployed in the field. Under the Level II warning state, ladybugs and predatory mite insects are released into the field, and Bacillus thuringiensis and Beauveria bassiana biological pesticides are sprayed. At the same time, insect-proof nets are set up around the field. Under the Level III warning state, low-toxicity green pesticides such as matrine, azadirachtin, and Bordeaux mixture are applied in limited quantities. The dosage, frequency, and safe interval of pesticide application and crop harvesting are controlled. Pesticide application data are recorded and operation logs are kept.

[0011] Furthermore, the water and fertilizer precision control stage constructs a phased water and fertilizer requirement standard model that includes the crop seedling stage, growth stage, and fruit expansion stage. Combined with real-time soil moisture data, soil nutrient residue data, and the risk level of pests and diseases in the field, a single plot, single crop, and phased water and fertilizer management plan is set. Compared with the conventional fertilization standard, the phosphorus and potassium nutrient ratio is appropriately increased for plots with high pest and disease risk. The water and fertilizer mixture is delivered to the crop root zone by the integrated water and fertilizer drip irrigation equipment. The intelligent sensors on the equipment collect field data in real time and adjust the irrigation time, single irrigation volume and water and fertilizer concentration. Well-rotted organic fertilizer and biological water-soluble fertilizer are selected as base fertilizer and top dressing in the field. Calcium, magnesium and boron trace element foliar fertilizers are sprayed during the vigorous growth period of crops and the critical period of nutrient demand.

[0012] Furthermore, in the traceability and control phase, an agricultural product planting blockchain traceability system is established. Data such as plot risk classification, soil improvement operation data, pest and disease monitoring and early warning data, prevention and control operation records, water and fertilizer management operation data, and crop quality testing data are entered into the system. A unique traceability QR code is assigned to each batch of agricultural products. A portable near-infrared spectrometer is used to regularly test crop quality indicators, and all quality testing data are synchronously entered into the blockchain system. The system retrieves the on-chain stored data for the current batch of products, covering the entire process from soil improvement, pest and disease control, water and fertilizer application to finished product testing. It extracts the blockchain-specific hash code corresponding to the batch data, calls the system's built-in QR code generation program, and binds the product batch number, origin information, and blockchain code to generate a standard vector format QR code image. The generated QR code image file is transmitted to the workshop's inkjet printing control system via the local area network, where it is combined with the product's basic text information for layout. The inkjet printing equipment completes the packaging printing according to the layout, and simultaneously archives the batch's QR code information separately into the corresponding product file entry on the blockchain.

[0013] When quality abnormalities occur, the system retrieves the stored operational data to locate the corresponding abnormal operational steps, classifies and disposes of substandard agricultural products, and periodically summarizes the stored operational data to statistically analyze the pest and disease control data and crop quality data corresponding to different risk plots, different warning levels, and different control plans. Based on the statistical results, the system adjusts the plot risk classification quantification threshold, pest and disease warning model parameters, and gradient control operation standards.

[0014] Furthermore, after the agricultural products are harvested in the field during the post-harvest management stage, they are sent to the pre-cooling operation area. Vacuum pre-cooling or differential pressure pre-cooling is selected according to the type of agricultural product for cooling treatment. After pre-cooling, physical disinfection methods are used to treat the surface of the agricultural products. According to the agricultural product grading standards, the agricultural products are graded and screened based on individual size, appearance, integrity, and internal quality indicators; the screened agricultural products are then packaged in a sterile environment using biodegradable materials.

[0015] The beneficial effects of this invention are as follows: 1. The detection equipment of this invention completes the screening of multiple indicators of soil heavy metals, pesticide residues, microorganisms, and water, soil and air, establishes a soil plant protection base database, and applies passivating materials, compound beneficial microbial communities and decomposed organic fertilizers in a classified manner according to pollution, continuous cropping and fertility deficiency, and improves seedling resistance by combining seed pretreatment; it divides the disease and pest risk level of plots by weighted quantitative classification of five types of planting data and draws risk maps, and differentiates the configuration of insect-resistant varieties, intercropping, rotation and insect-repelling crops for high, medium and low risk plots, and supports a graded preparatory prevention and control mechanism, thus optimizing the planting environment from the dual aspects of arable land base and field ecological layout.

[0016] 2. This invention integrates the Internet of Things, UAV remote sensing, pest and disease trapping, and meteorological data. Through deep learning models, it captures the early latent characteristics of pests and diseases, predicts their development trends in advance, and classifies them into three levels of early warning. This enables pest and disease control to shift from post-event remediation to pre-event prediction and prevention. According to the early warning level, a tiered control plan is implemented: Level 1 uses physical control methods such as insecticidal lamps and color boards; Level 2 releases natural enemy insects in combination with biological pesticides; and Level 3 uses low-toxicity green pesticides in limited quantities and standardizes pesticide use records and safety intervals, achieving coordinated development of plant protection and farmland ecology.

[0017] 3. The blockchain technology of this invention collects data from all stages of soil treatment, pest and disease control, water and fertilizer management, and finished product testing. Each item has a unique code to achieve full life cycle traceability. When product quality is abnormal, the problematic production stage can be traced and the unqualified product can be disposed of. The system continuously accumulates planting big data, reverse-optimizes plot risk scoring, early warning model parameters and control operation standards, and is equipped with standardized post-harvest pre-cooling, physical sterilization, grading and screening and biodegradable packaging processes, and uniformly marks traceability information. Attached Figure Description

[0018] Figure 1 This is a flowchart of the standardized planting and management method for pollution-free green agricultural products according to the present invention; Figure 2 This is a flowchart of the intelligent monitoring and early warning operation for pests and diseases according to the present invention. Detailed Implementation

[0019] 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.

[0020] like Figures 1 to 2 As shown, this embodiment of the invention provides a standardized planting and management method for pollution-free green agricultural products, including soil preparation stage, layout optimization stage, pest and disease early warning stage, green prevention and control stage, water and fertilizer precision control stage, traceability management stage, and post-harvest management stage. In this embodiment of the invention, during the soil preparation stage, a portable X-ray fluorescence spectrometer is used to detect the content of heavy metals and available forms in the soil of the planting plot; a gas chromatography-mass spectrometry system is used to detect pesticide residue metabolites in the soil; high-throughput sequencing technology is used to detect the soil microbial community structure and soil enzyme activity; and soil pH, organic matter content, nitrogen, phosphorus and potassium available nutrient content, irrigation water quality, field atmospheric environmental quality are detected simultaneously, and historical pest and disease breeding data of the plot are retrieved to establish a soil plant protection baseline database for the plot. Sampling points were delineated at the test sites using the five-point sampling method. Soil samples were collected from the surface to the topsoil using a soil auger. Stones, crop residues, dead branches, and other foreign matter were removed from the soil samples. After natural air drying, the samples were ground and sieved, then divided into two equal portions for later use. The X-ray fluorescence spectrometer was preheated and its parameters stabilized. The sieved soil samples were then pressed into standard sample discs using a sample press and placed stably in the instrument's detection window for sequential testing of multiple heavy metals. The instrument's output readings were recorded. Soil samples for pesticide residue testing were weighed to a fixed weight and then subjected to a series of pretreatment processes, including organic solvent extraction, solid-phase purification, and rotary concentration. The pretreated samples were transferred to sample vials and sent in batches to a gas chromatography-mass spectrometry (GC-MS) instrument for on-machine testing. For microbial testing, the prepared soil samples underwent microbial cell lysis and genomic nucleic acid extraction at low temperatures. A dedicated sequencing library was constructed using extracted nucleic acids. After the library passed quality control, it was subjected to high-throughput sequencing. After all tests were completed, the original test data were classified and archived according to the sampling site number.

[0021] For planting plots with mild heavy metal pollution, straw biochar, sepiolite and humic acid composite passivating agent are applied. For planting plots with severe continuous cropping obstacles, a compound beneficial microbial community of Bacillus subtilis, Trichoderma and Pseudomonas is quantitatively inoculated. For planting plots with poor soil fertility, well-rotted farmyard manure and bio-organic fertilizer are applied. Before sowing, crop seeds are pretreated by soaking in hot water and biological seed dressing.

[0022] When preparing the compound passivating agent, first mix the three raw materials in a fixed mass ratio. After mixing, spread the mixture evenly on the surface of the field using a field fertilizer spreader. After all the materials have been spread, till the soil to a depth of 20-30cm using a rotary tiller to bury the passivating material in the main tillage layer of the crop. Before using the compound beneficial microbial flora, dilute it with sterile water according to the predetermined dilution ratio. Use a field furrow application process, opening shallow furrows along the crop sowing rows, and then spray the bacterial solution evenly into the furrows, covering it with soil. After mixing well-rotted farmyard manure with bio-organic fertilizer, apply it in two applications. The base fertilizer is spread throughout the field during land preparation, and the remaining fertilizer is shallowly buried in the soil before sowing. Before the warm water soaking operation, the soaking water temperature is determined according to the categories of melons and fruits, grains and oils, and leafy vegetables. Seeds are added in batches and the water is stirred regularly. After soaking, the seeds are removed and the surface water is drained. The biological seed coating operation is completed inside a closed seed coating equipment. Crop seeds and solid bacterial powder are added in sequence, and the equipment continuously and evenly tumbles and stirs until the bacterial powder is evenly attached to the seed surface.

[0023] In this embodiment of the invention, the layout optimization stage collects five categories of basic rating data: soil plant protection data, historical pest and disease outbreak frequency data, regional meteorological temperature and humidity data, crop susceptible pest and disease types data, and farmland ecological resistance data. The five categories of basic rating data are weighted and calculated, and quantitative scoring thresholds are set for low pest and disease risk plots (0-40 points), medium pest and disease risk plots (41-70 points), and high pest and disease risk plots (71-100 points). Based on the calculation results, a pest and disease risk map of the plot is drawn to complete the risk level classification of the planting plot. A five-member expert review panel, comprised of technicians specializing in soil testing, crop cultivation, and plant protection monitoring, finalized the individual weight values ​​for the five categories of indicators through multiple rounds of item-by-item discussions, and compiled a paper archive table of weight parameters. Measured values ​​were extracted from each of the five categories of raw monitoring data, and extreme value normalization was uniformly applied to eliminate differences in measurement units between different indicators, converting all raw data to a score range of 0-100. The converted score for each category of indicator was multiplied by the predetermined weight value for that category, and the results of the five sub-items were summed sequentially to obtain the comprehensive risk score for each plot of farmland. The risk level of the plot was then marked according to the predetermined score range, and the corresponding score data was simultaneously marked on the electronic map layer.

[0024] In high-risk areas, plant highly resistant pest and disease varieties and densely plant insect-repellent crops between crop rows. In medium-risk areas, plant moderately resistant high-quality varieties and adopt different crop intercropping patterns. In medium- and low-risk areas, regularly carry out green manure rotation and water-dryland rotation. At the same time, a three-level prevention and control reserve mechanism is set up for high-risk areas, a two-level prevention and control reserve mechanism is set up for medium-risk areas, and basic field ecological maintenance is carried out for low-risk areas.

[0025] Dedicated storage warehouses are set up near the fields, and the internal storage areas of the warehouses are divided according to the risk level of the plots. The zones are marked with signs to distinguish the types of materials. The three-level prevention and control reserve area stores insecticidal lamps, trapping accessories, natural enemy breeding and storage equipment, various biological agents, registered low-toxicity green pesticides and protective labor protection products. The two-level prevention and control reserve area centrally stores physical control equipment and biological pesticide raw materials, and does not store chemical agents. The warehouses set up for low-risk plots only store green manure seeds, straw, insect nets, seed dressing agents and other ecological maintenance materials. On a fixed date each month, the quantity of materials in each area is inventoried, the inventory ledger is checked, and any shortages are registered and replenished.

[0026] Based on the plot area and soil type, several independent rotation zones are divided, and a written rotation implementation plan is prepared on a calendar year basis. The plan indicates the rotation type and start and end time of the corresponding zone. Under the green manure rotation mode, green manure is sown after the main crop is harvested. When the green manure grows to the specified height, it is pulverized by rotary tillage. The pulverized green manure material is then turned into the soil layer through deep plowing. When switching from paddy field to dryland rotation, the field surface is leveled, irrigation and drainage ditches are dug and repaired before the conversion from dryland to paddy field. After the canal system is inspected for seepage prevention, the field is flooded in stages. When converting from paddy field to dryland, the water in the field is drained, the soil is dried, and the whole field is deep plowed and dried. The field furrow structure is then uniformly repaired.

[0027] In this embodiment of the invention, during the pest and disease early warning stage, IoT monitoring devices and pest and disease traps are deployed in the planting area. Drones equipped with hyperspectral imagers and multispectral cameras are used to regularly conduct aerial photography in the field, simultaneously collecting real-time monitoring data on soil temperature and humidity, soil moisture, and soil microbial community. The drone aerial photography also collects remote sensing data on micro-damage to crop leaves and abnormal crop growth. The pest and disease traps collect data on the breeding of insect eggs and larvae in the field, as well as real-time meteorological data on temperature, humidity, rainfall, and light intensity in the region. Data on the incubation period and breeding patterns of pests and diseases in the field are also retrieved. The field IoT monitoring terminal automatically stores raw monitoring data according to the preset collection frequency. During fixed working hours each day, data is downloaded remotely in batches from the terminal memory via wired or wireless transmission. Staff regularly go to the field to dismantle the pest traps, sort and distinguish between samples of insect eggs, larvae, and adults, count the number of insects in each plot, and manually enter the statistical results into the field data ledger. After the drone returns, the onboard storage hardware is removed, and all raw multispectral and hyperspectral aerial images are exported. Independent folders are created according to the plots corresponding to the aerial images for classification and storage. All image data is manually screened, and invalid images caused by cloud cover, large debris, or flight deviation are deleted. The sorted structured data is then imported in batches into the project's local database.

[0028] The site was divided into equally spaced grids of monitoring points using a topographic map. Shallow pits were excavated at each point to install fixed bases for soil temperature, humidity, and nutrient sensors. Galvanized poles were used to erect IoT data collection terminals for air temperature, humidity, and light. Each device's number and geographic coordinates were entered to create a device management ledger. Pest traps were installed in designated areas according to the grid points, with traps for moths and larvae categorized and placed at specific locations. Installation information for the traps was regularly recorded. Before drone operations, the imaging lens was cleaned and calibrated, the onboard storage device was formatted, and the onboard battery power was checked. Pre-defined flight path coordinates were imported, and low-altitude aerial photography was conducted sequentially according to the block order. After each block's aerial photography mission, the original onboard images and spectral data were exported on-site and stored in designated hard drives.

[0029] The data is fused and cleaned to construct a pest and disease early warning dataset and build a deep learning model for identifying the incubation period of pests and diseases. The characteristics of the incubation period of pests and diseases are small yellow spots on crop leaves, local rot of stems, and accumulation of trace insect eggs in the soil. These characteristics are used as training samples for the model to complete the training of model parameters. The model combines real-time data to infer the breeding speed, spread range and outbreak probability of pests and diseases. The model training process is performed in four steps: First, the preprocessed warning dataset is randomly divided into training, validation, and test sets in a 7:2:1 ratio. Before partitioning, all samples are one-hot encoded with labels, and the sample storage format is standardized. Second, the Adam optimizer is configured as the model parameter optimization operator, with an initial learning rate of 0.0005, a learning rate decay coefficient of 0.95, a batch size of 32 samples per training run, a maximum number of epochs of 120, and a composite loss function combining mean squared error (MSE) and cross-entropy. Third, an early stopping mechanism is enabled; if the validation set loss value does not decrease after 15 consecutive epochs, the iteration automatically stops, and the model weights corresponding to each epoch are saved simultaneously. Fourth, the test set is called to iterate through and verify the model weights, compare the error indices of each group of weights, and retain the model weight file with the best verification results for deployment.

[0030] The deep learning model adopts a CNN+LSTM cascaded composite network architecture. The network is divided into four independent functional modules from top to bottom: image feature extraction module, environmental temporal feature encoding module, feature fusion and concatenation layer, and fully connected classification and regression layer. The image feature extraction module consists of 5 convolutional layers and 3 max pooling layers connected sequentially. Each convolutional layer is followed by a batch normalization layer and a ReLU activation function. The environmental temporal feature encoding module has 3 independent LSTM recurrent units. The LSTM units are fully interconnected through input gates, forget gates, and output gates. Each LSTM layer is followed by a Dropout random deactivation layer. The feature tensor output by the image feature extraction module and the feature vector output by the environmental temporal feature encoding module are connected to the feature fusion and concatenation layer to complete data dimension alignment and feature concatenation. The concatenation result is then directed into the fully connected classification and regression layer, which has two layers of neurons arranged in a progressive hierarchy.

[0031] Based on the probability of pest and disease outbreaks, Level 1, Level 2, and Level 3 early warnings are established. Corresponding early warning information is pushed through mobile apps and SMS platforms, and corresponding field control operation plans are matched. After the early warning results are output, the corresponding level of field control operation process is automatically matched and initiated.

[0032] In this embodiment of the invention, the green prevention and control stage combines the pest and disease warning level with the plot risk level to carry out graded prevention and control operations. The density of prevention and control deployment and the amount of pesticides are fine-tuned according to the pest and disease risk level of the plot. Under the first-level warning state, weeds and crop diseased residues in the field are regularly removed, and insecticidal lamps, sticky insect sticks and pheromone attractants are deployed in the field to complete physical prevention and control. The farmland is divided into several independent work areas using the field ridges as boundaries. Workers carry out field cleaning work in each area, manually removing weeds and cutting off crop branches and leaves carrying disease spots and insect eggs. All cleaned materials are collected in special storage baskets outside the field and transported to a centralized storage area outside the field at regular intervals. The length and width of the plots are measured on-site with a tape measure, and the spots are positioned according to fixed horizontal and vertical spacing. Steel rods are driven into the soil to form a fixed base, and the installation height of the insecticidal lamps is adjusted and the fixing accessories are tightened. The sticky insect sticks are unpacked and tied to the crop supports with nylon rope, and the hanging position is uniformly 12cm above the crop canopy. The pheromone lures are disassembled and individually sealed, inserted into the matching plastic trap slots, and the traps are fixed on the pre-erected poles in the field, completing the uniform deployment of the entire field.

[0033] Under the Level II early warning status, ladybugs and predatory mite natural enemies are released in the field, and Bacillus thuringiensis and Beauveria bassiana biological pesticides are sprayed. At the same time, insect-proof nets are set up around the field. The total area of ​​the plot was divided into multiple natural enemy release zones based on actual measurements. Natural enemy packaging was disassembled according to the zone area, and natural enemy insects were evenly released during two low-temperature periods, before sunrise and before sunset. The biological pesticide raw material was weighed according to the mixing ratio, and clean water was added to the mixing tank in several batches. The stirring rod was turned on and stirred continuously until the pesticide was completely dispersed and dissolved. The mixture was then filled into a pressurized sprayer, and the operator held the nozzle and moved at a constant speed along the crop rows to evenly spray the pesticide solution on both sides of the crop leaves. A rigid pole was pre-buried every 3 meters along the perimeter of the farmland, and the entire roll of insect-proof netting was unfolded. The sides of the netting were tied and fixed to the poles, and the bottom edge of the netting was sealed with layers of fine soil compacted and sealed.

[0034] Under the Level 3 early warning status, the application of low-toxicity green pesticides such as matrine, azadirachtin, and Bordeaux mixture is limited, and the application dosage, application frequency, and safe interval for crop harvesting are controlled. The application data is recorded and the operation log is kept.

[0035] Retrieve the planting records of the plots, and calculate the theoretical total dosage of pesticides for a single application based on the plot area and the planting density of plants in the field. Divide the calculated total dosage into 2 to 3 applications, and record the specific date of each application and the number of days between applications in a paper record sheet. Mark the time interval between the last application and the harvest of the agricultural products. The paper ledgers are classified and bound according to the plot numbers, and the specific type of pesticide, the actual weight of each application, the name of the on-site operator, and the specific area number of the application are filled in item by item. All ledger information is simultaneously entered into the local industrial control storage terminal.

[0036] In this embodiment of the invention, the water and fertilizer precision control stage constructs a standard model of water and fertilizer requirements for crop seedling stage, growth stage and fruit expansion stage. Combined with real-time soil moisture data, soil nutrient residue data and plot pest and disease risk level, water and fertilizer management schemes for single plot, single crop and stage are set up. For plots with high pest and disease risk, the phosphorus and potassium nutrient ratio is appropriately increased. Retrieve the recent soil nutrient test report for this plot, extract the measured data of existing nitrogen, phosphorus, and potassium residues in the soil, match it with the water and fertilizer standard model corresponding to the current growth stage of the crop, first calculate the basic nitrogen, phosphorus, and potassium addition base of the formula, adjust the weight of nitrogen fertilizer raw materials in the formula downward, and simultaneously increase the weight of phosphorus and potassium fertilizer raw materials; add all solid fertilizer raw materials to the fertilizer mixing tank according to the new ratio, add irrigation water, start the stirring device in the tank and stir continuously for 40 minutes, and after the fertilizer is completely dissolved, prepare the water and fertilizer stock solution.

[0037] The integrated water and fertilizer drip irrigation system delivers the prepared water and fertilizer mixture to the area where the crop roots are concentrated. The intelligent sensors on the equipment collect field data in real time to regulate the irrigation duration, the amount of water and fertilizer per irrigation, and the water and fertilizer concentration. Well-rotted organic fertilizer and bio-water-soluble fertilizer are used as base fertilizer and top dressing. During the vigorous growth period of the crop and the critical period of nutrient demand, calcium, magnesium, and boron micronutrient foliar fertilizers are sprayed to supplement the micronutrients required by the crop.

[0038] Weigh the calcium, magnesium, and boron fertilizer raw materials according to the corresponding ratio based on the application area in the field. Add clean water at room temperature to the mixing tank in several batches and stir continuously with a stirring device until the solid raw materials are completely dissolved. Adjust the spraying equipment nozzle to fine mist mode and select a time before sunrise when there is no strong sunlight for field spraying. The operator should move at a constant speed along the crop rows to ensure that both the upper and lower leaves of the crop are covered with the solution. Any remaining fertilizer solution that is not used on the same day should be put into a sealed plastic container, and the preparation date and raw material ratio information should be marked on the container. Store it in a cool, dark warehouse.

[0039] Before starting the water and fertilizer delivery process, the main supply pipeline, field branch pipelines, and drip irrigation emitters are inspected one by one to clear any blockages in the pipelines and water outlets. After the pipeline inspection is completed, the unit's booster pump is turned on, and the zone-controlled solenoid valves are remotely opened in sequence according to the plot zone number. The equipment's sensors continuously collect field soil moisture data and transmit it back to the control cabinet in real time. The operator dynamically adjusts the opening duration and supply flow of individual zone solenoid valves based on the transmitted data. After the entire water and fertilizer delivery process is completed, the clean water pipeline is switched to flush the main pipeline and branch drip irrigation pipelines with water to drain any residual fertilizer solution inside the pipelines. The equipment is then shut down and the pipelines on site are sealed.

[0040] In this embodiment of the invention, the traceability and control stage establishes a blockchain traceability system for agricultural products planting. The system records and encrypts the risk classification data of the plot, soil improvement operation data, pest and disease monitoring and early warning data, prevention and control operation records, water and fertilizer management operation data, and crop quality testing data. A unique traceability QR code is assigned to each batch of agricultural products. A portable near-infrared spectrometer is used to regularly test crop quality indicators, and all quality testing data is synchronously recorded into the blockchain system. Samples of the crops to be tested were collected from multiple points in the corresponding planting plots using the five-point sampling method. External impurities such as soil and residual leaves adhering to the sample surface were removed, and larger samples were cut into smaller pieces. The near-infrared spectrometer was preheated, and the built-in detection model for the corresponding crop was selected. The processed sample was then smoothly placed against the instrument's detection probe, and the scanning button was pressed to complete the spectral acquisition. The instrument automatically saved the raw data for each test. After all sample testing for the day was completed, all test files stored on the instrument were compiled, and data tables were organized according to plot number and harvest batch. The corresponding data was then uploaded item by item to the data block bound to the product via the blockchain backend.

[0041] Independent service nodes of the consortium blockchain were deployed in three locations: the central control room of the planting base, the mobile workstation in the field, and the agricultural product sorting workshop, and the multi-node network debugging was completed. Data collected in the field is hashed and encrypted at the front end of the upload link to generate a data digest. The original business data and the encrypted digest are packaged and integrated into a single block data packet. The data packet is written into the blockchain storage space after passing the consensus verification process of all online nodes. When a traceability QR code is generated for each batch of agricultural products, the QR code is bound to the unique hash code of the corresponding block.

[0042] When quality abnormalities occur, the system retrieves the stored operational data to locate the corresponding abnormal operational steps, classifies and disposes of substandard agricultural products, and periodically summarizes the stored operational data to statistically analyze the pest and disease control data and crop quality data corresponding to different risk plots, different warning levels, and different control plans. Based on the statistical results, the system adjusts the plot risk classification quantification threshold, pest and disease warning model parameters, and gradient control operation standards.

[0043] In this embodiment of the invention, after the agricultural products are harvested in the field during the post-harvest management stage, they are sent to the pre-cooling operation area. Vacuum pre-cooling or differential pressure pre-cooling is selected according to the type of agricultural product for cooling treatment. After pre-cooling, physical disinfection methods such as ultraviolet sterilization, ozone sterilization, and low-temperature plasma sterilization are used to treat the surface of the agricultural products. According to product specifications, the products are graded and stacked onto the perforated plastic pre-cooling trays. The thickness of each layer of agricultural products is controlled. The trays are neatly and orderly pushed into the sealed pre-cooling chamber. The chamber door is closed and the sealing buckle is locked. When using the vacuum pre-cooling process, the water circulation unit and vacuum pump equipment are started in sequence to reduce the ambient air pressure in the chamber in multiple stages. The temperature readings in the chamber are collected at regular intervals throughout the process. When using the differential pressure pre-cooling process, the non-operational ventilation holes in the chamber are sealed, the position of the air duct guide baffle is adjusted, the circulating air supply equipment is started, and the fan operation level is set in stages. After the center temperature of the product reaches the preset index, all equipment is shut down, and the chamber door is slowly opened to release pressure and discharge the material.

[0044] Harvested products are categorized into leafy vegetables, melons and fruits, and root vegetables. A single disinfection process is selected based on the category. The products are neatly stacked into a sealed disinfection turnover frame and then sent into a sealed disinfection chamber. The chamber door is closed to seal the chamber. The preset running time and equipment power parameters are entered into the equipment control panel, and the disinfection equipment is started for standardized operation. After the disinfection timer ends, the exhaust pipe of the chamber is turned on to continuously exhaust and replace the stagnant gas in the chamber. After the air pressure in the chamber returns to normal, the chamber door is opened and the finished fruits and vegetables are taken out.

[0045] According to the agricultural product grading standards, the agricultural products are graded and screened based on individual size, appearance, integrity, and internal quality indicators. The screened agricultural products are packaged in a sterile environment using biodegradable materials, and the packaging surface is uniformly labeled with the product name, place of origin, harvest and production date, shelf life, pollution-free product logo, and a unique blockchain traceability QR code.

[0046] Before the packaging operation begins, all equipment in the sterile workshop, including the operating table, automatic sealing machine, and cutting tools, is wiped down with compliant disinfection reagents. After standing for a specified period, the workshop air purification device is turned on. Biodegradable packaging films and containers of the corresponding sizes are cut according to the specifications of the agricultural products. The sorted agricultural products are then transported to the packaging station at a uniform speed, where they are filled and sealed by automated equipment. The parameters of the inkjet printing equipment are adjusted, and text information and traceability QR codes are sequentially printed on the preset packaging surface. After the inkjet printing is completed, the finished products are grouped according to the production batch, and the corresponding batch number and classification code are marked on the outside of the outer turnover box before being placed in the warehouse.

[0047] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0048] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A standardized planting and management method for pollution-free green agricultural products, characterized in that, The specific steps include the following: During the soil preparation stage, core indicators of soil, irrigation water and field atmosphere in the planting plot are tested, data on the occurrence of pests and diseases in the plot are collected and a soil plant protection base database is established, and soil improvement, microbial community optimization and fertility cultivation pretreatment are carried out according to the plot conditions. In the layout optimization stage, multiple types of basic data are integrated for weighted quantitative calculation, the risk level of pests and diseases in the plots is divided and the risk map of pests and diseases in the plots is drawn, the corresponding resistant crop varieties and ecological planting layout are matched, and a graded field prevention and control preparation mechanism is set up. During the pest and disease early warning stage, a warning dataset is constructed by integrating multi-source field data. The characteristics of the incubation period of pests and diseases are identified through a deep learning model. The breeding and spread of pests and diseases are deduced and the pest and disease warning is divided into three levels: Level 1, Level 2, and Level 3. The warning information is intelligently pushed and the corresponding level of control operation process is initiated. During the green prevention and control phase, a tiered prevention and control system is established. Physical, biological, and low-toxicity green pesticides are used for tiered prevention and control operations based on the pest and disease early warning level and the risk level of the plot. Pesticide application parameters are standardized and operation log data is kept. During the precise water and fertilizer control stage, water and fertilizer management plans are set up based on the crop's water and fertilizer requirements, combined with the field soil moisture, real-time nutrient residue status, and pest and disease risk level, and water and fertilizer are supplied through integrated water and fertilizer drip irrigation equipment. During the traceability and control phase, a blockchain-based agricultural product traceability and control system is constructed. All operational data is encrypted and stored, and a unique blockchain traceability QR code is assigned to each agricultural product. The unique traceability QR code is generated through block hash encoding. When quality abnormalities occur, data is retrieved to locate the cause and non-compliant agricultural products are classified and disposed of. The data is summarized and the planting control parameters are iteratively optimized. During the post-harvest management stage, the harvested agricultural products are pre-cooled and sterilized, and after grading and screening, they are aseptically packaged using biodegradable materials.

2. The standardized planting and management method for pollution-free green agricultural products according to claim 1, characterized in that, During the soil preparation stage, a portable X-ray fluorescence spectrometer was used to detect the content of heavy metals and available forms in the soil. Gas chromatography-mass spectrometry was used to detect pesticide residue metabolites in the soil. High-throughput sequencing technology was used to detect the structure of soil microbial communities and enzyme activity. Soil pH, organic matter content, available nitrogen, phosphorus and potassium content, irrigation water quality and field atmospheric environmental quality were detected simultaneously. Historical pest and disease data of the plots were retrieved to establish a soil plant protection baseline database.

3. The standardized planting and management method for pollution-free green agricultural products according to claim 2, characterized in that, During the soil preparation stage, straw biochar, sepiolite and humic acid composite passivating agent are applied to plots with light heavy metal pollution. For plots with severe continuous cropping obstacles, a fixed ratio of Bacillus subtilis, Trichoderma and Pseudomonas compound beneficial microbial community is quantitatively inoculated. For plots with fertility deficiency, well-rotted farmyard manure and bio-organic fertilizer are applied. Before sowing, crop seeds are pretreated by soaking in hot water and biological seed dressing.

4. The standardized planting and management method for pollution-free green agricultural products according to claim 3, characterized in that, The layout optimization stage involves collecting five categories of basic data: soil plant protection baseline database, historical pest and disease outbreak frequency, regional meteorological temperature and humidity, types of crop susceptible pests and diseases, and farmland ecological resistance. These data are then weighted and calculated. Quantitative scoring thresholds are set for plots with low pest and disease risk (0-40 points), medium pest and disease risk (41-70 points), and high pest and disease risk (71-100 points). Plot pest and disease risk maps are then drawn.

5. The standardized planting and management method for pollution-free green agricultural products according to claim 4, characterized in that, The layout optimization phase involves planting highly resistant pest and disease varieties and densely planting insect-repellent crops in high-risk plots, planting moderately resistant high-quality varieties and adopting intercropping patterns in medium-risk plots, and regularly carrying out green manure rotation and water-dryland rotation operations in medium- and low-risk plots; a three-level prevention and control preparedness mechanism is preset for high-risk plots, a two-level prevention and control preparedness mechanism is preset for medium-risk plots, and a basic ecological maintenance and prevention and control mechanism is provided for low-risk plots.

6. The standardized planting and management method for pollution-free green agricultural products according to claim 5, characterized in that, During the disease and pest early warning stage, IoT monitoring devices and disease and pest traps are deployed in the planting plots. Drones equipped with hyperspectral imagers and multispectral cameras are used to conduct regular aerial remote sensing inspections in the fields. Simultaneously, field soil environment monitoring data, crop growth data, disease and pest breeding data, and regional meteorological time series data are collected, and data on the incubation period and breeding patterns of diseases and pests in the plots are retrieved.

7. The standardized planting and management method for pollution-free green agricultural products according to claim 6, characterized in that, The disease and pest early warning stage involves fusing and cleaning the collected data to construct an early warning dataset, building a deep learning model for identifying the incubation period of diseases and pests, and using the phenomena of slight yellow spots on crop leaves, localized stem rot, and the accumulation of trace insect eggs in the soil as characteristics of the incubation period of diseases and pests. The model is trained using these characteristics, and real-time data is combined to infer the breeding speed, spread range, and outbreak probability of diseases and pests. Based on the outbreak probability threshold of diseases and pests, the early warning is divided into Level 1, Level 2, and Level 3.

8. The standardized planting and management method for pollution-free green agricultural products according to claim 7, characterized in that, The green prevention and control phase combines the pest and disease warning level with the plot risk level to carry out graded prevention and control operations. Under the first-level warning state, weeds and diseased plant debris in the field are removed and insecticidal lamps, sticky insect boards and pheromone attractants are set up. Under the second-level warning state, natural enemy insects are released and biological pesticides are sprayed. Under the third-level warning state, low-toxicity green pesticides are applied in limited quantities and application parameters are controlled.

9. The standardized planting and management method for pollution-free green agricultural products according to claim 8, characterized in that, The water and fertilizer precision control stage constructs a phased water and fertilizer requirement standard model that includes the seedling stage, growth stage, and fruit expansion stage of crops. Combined with real-time soil moisture, nutrient residue data, and pest and disease risk levels, a phased water and fertilizer management plan for a single plot and a single crop is set. The water and fertilizer mixture is delivered to the concentrated area of ​​crop roots through integrated water and fertilizer equipment. In plots with high pest and disease risk, the phosphorus and potassium ratio is increased. During the key growth stages of crops, foliar fertilizers of calcium, magnesium, and boron micronutrients are sprayed. Based on the field nutrient residue and crop growth data, the phased water and fertilizer ratio benchmark of the plot is optimized in reverse.

10. The standardized planting and management method for pollution-free green agricultural products according to claim 9, characterized in that, During the traceability and control phase, a portable near-infrared spectrometer is used to regularly detect crop quality indicators and simultaneously record them into the blockchain system. When quality is abnormal, the operation data is retrieved to locate the abnormal link. The system regularly summarizes statistical data and adjusts the plot's pest and disease risk quantification scoring threshold, pest and disease early warning model parameters, and gradient control operation standards accordingly.

Images