Soybean breeding method based on intelligent agricultural application

Abstract

The invention relates to the technical field of soybean breeding, in particular to a soybean breeding method based on intelligent agricultural application, and aims to solve the technical problems that factors such as field environmental climate and soil fertility are different every year, and environmental factors and growth indexes remarkably related to excellent characters cannot be accurately found during breeding in the prior art. Comprising the following steps: delimiting a measuring area, setting an identifier, configuring a sensor, collecting land parcel data, collecting growth data, summarizing the data, establishing a data analysis model, monitoring soybean growth environment data in real time through the sensor by virtue of an intelligent agricultural technology, and performing big data analysis by combining a data integration platform. According to the method, the growth performance of different soybean plants in a specific environment can be rapidly and accurately evaluated, plants with excellent characters and various growth parameters suitable for the plants can be screened out in a short time, the plants with the excellent characters grow rapidly by maintaining the various growth parameters, and the yield of the soybean plants is improved. And the time required from hybridization to acquisition of stable and excellent varieties is greatly shortened.

Description

Technical Field

[0001] The present invention relates to the field of soybean breeding technology, and specifically to a soybean breeding method based on smart agriculture applications. Background Art

[0002] In the global agricultural development process, soybean breeding, as an important crop with dual properties as a food and oilseed, has always been a key research direction in the agricultural field. While traditional soybean breeding has achieved certain results over a long period of development, it faces many challenges that need to be addressed in the current era.

[0003] Traditional soybean breeding relies on manual field observation and hybridization, an extremely cumbersome process. From parent selection and hybridization to multi-generational screening, cultivating new varieties with stable and superior traits often requires 8-10 generations. Based on one growing season per year, this process can take several years to over a decade. This lengthy cycle results in significant delays in the release of new varieties, making it difficult to keep pace with market demand for updated soybean varieties.

[0004] The field environment is complex and changeable, and factors such as climate and soil fertility vary from year to year, seriously interfering with soybean growth performance. During the breeding process, it is extremely difficult to distinguish between trait differences caused by genetic and environmental factors. In drought years, high-yield potential varieties may perform poorly due to environmental factors, misleading breeders in their genetic trait assessments, causing deviations in the screening of superior varieties and affecting screening efficiency. Moreover, the accurate and rapid identification of soybean disease and pest resistance and quality-related traits in the field is difficult. For example, resistance to root rot requires long-term observation in a specific disease environment, and the degree of disease varies from plot to plot, making screening more difficult.

[0005] The prior art CN117958133 discloses a soybean breeding method and system based on smart agriculture applications. However, the system only performs simple monitoring and is located in a greenhouse, which cannot reflect the characteristics of soybeans grown naturally in real scenarios. In addition, the selected samples are not representative, and the number of samples is large and complex, resulting in a large amount of data information processing, which affects the efficiency and difficulty of breeding. Summary of the Invention

[0006] In order to solve the technical problem that the main problem is that the selected samples are not representative, the samples are numerous and complex, and the amount of data information processing obtained is large, thereby affecting the efficiency and difficulty of breeding, the present invention provides a soybean breeding method based on smart agriculture applications.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a soybean breeding method based on smart agriculture application, comprising: S1: Delineate the survey area Using geographic information system (GIS) technology, combined with satellite images or high-precision maps, the soybean planting area is divided into multiple differentiated plots based on soil type, topography, and irrigation conditions.

[0008] S2: Set the logo Clear signboards are set up at the boundaries of each divided plot, marking the plot number, area, and soil type information. At the same time, a corresponding plot information database is established in the data integration platform, and detailed information of each plot is entered.

[0009] S3: Sensor Configuration According to the characteristics of different plots, corresponding sensors are deployed. Soil moisture sensors and nutrient sensors are evenly installed at different locations of each plot according to soil texture and terrain, and meteorological monitoring devices are set up in different plots.

[0010] S4: Plot data collection The data collection frequency of different growth stages is set to collect data from each sensor in real time.

[0011] S5: Growth data collection A dedicated drone flight plan is developed for each plot of land. The appropriate flight route and altitude are planned based on the shape, size and location of the plot. The drone photographs each plot according to the predetermined flight plan to obtain multispectral images. After shooting is completed, the image data is promptly transmitted to the data integration platform via wireless transmission. In the platform, the image data is classified and stored, and folders are created according to the plot number and shooting time.

[0012] S6: Data Summary The soil moisture, fertility, and meteorological data collected by sensors and the multispectral image data taken by drones are summarized in the data integration platform. Through the data association algorithm, the data from different sources are matched with the corresponding plot information.

[0013] S7: Build data analysis model: Using data analysis software and statistical methods, a soybean growth model was established. Soil moisture, fertility, meteorological environment data, and plant growth data obtained from drone image analysis were used as input variables, and soybean yield, quality, disease and pest resistance and other excellent traits were used as output variables. The correlation between each input variable and output variable was analyzed through multiple linear regression and principal component analysis statistical methods. Through data analysis, environmental factors and growth indicators that were significantly correlated with excellent traits were found to achieve efficient breeding.

[0014] Preferably, the method for demarcating the survey area includes demarcating plots with differences in soil quality, altitude, and coverage of irrigation water sources according to different soil quality, altitude, and coverage of irrigation water sources.

[0015] Preferably, the sensors are placed in the following locations in different soil types: Sandy soil: Insert sensors vertically in the four corners and the center of the field.

[0016] Loam soil: Divide the land into a grid and insert sensors at each grid intersection.

[0017] Clay: Select representative low-lying and high-altitude locations and set up sensors in each location.

[0018] Preferably, the sensors are arranged at the following locations at different altitudes: Low-altitude plots: Insert the sensor at the edge of the plot, close to a water source or river.

[0019] High-altitude areas: sensors are deployed in different terrain locations.

[0020] Preferably, the sensors and the coverage of different irrigation water sources are arranged at the following locations: Flood irrigation plots: Set up sensors at the inlet, middle and outlet of the water flow.

[0021] Drip irrigation plots: Sensors should be arranged around the drippers.

[0022] Sprinkler irrigation plots: Sensors are evenly distributed throughout the plot.

[0023] Preferably, the flight plan of the drone in the growth data collection includes: for regularly shaped plots, a rectangular route can be used for flight; for irregular plots, a route that can fully cover the plots is designed, and the flight altitude is controlled at 100-150 meters.

[0024] Preferably, data aggregation also includes: processing the aggregated data, removing outliers and erroneous data, checking whether there are values ​​in the sensor data that are obviously beyond a reasonable range, denoising the drone image data, and then standardizing the data to convert different types of data into a unified format and unit.

[0025] Preferably, the soybean breeding method further comprises S8: abnormal warning A real-time data monitoring module is built into the data integration platform. This module continuously acquires data uploaded by sensors and drone images and compares it with a preset normal data range. If an anomaly is detected, the platform alerts the operator on the operation interface.

[0026] Preferably, the data integration platform uses the AWS IoT Core Internet of Things platform.

[0027] Compared with the prior art, the present invention has the following beneficial effects: 1. With the help of smart agricultural technology, real-time monitoring of soybean growth environment data through sensors and combined with big data analysis on the data integration platform can quickly and accurately evaluate the growth performance of different soybean plants in specific environments. Plants with excellent traits and various growth parameters suitable for the plants can be screened out in a relatively short period of time. Subsequently, by maintaining various growth parameters, plants with excellent traits can grow rapidly, greatly shortening the time required from hybridization to obtaining stable and excellent varieties, and reducing the difficulty of breeding.

[0028] 2. By arranging sensors based on soil quality, altitude, and irrigation water sources, the problem of inaccurate and unrepresentative data obtained by existing breeding methods is solved. This allows for the detection of representative data for the current plot, which has high reference value and improves breeding efficiency. 3. Process the aggregated data to remove outliers and erroneous data, improve data accuracy, further enhance data representativeness, and reduce breeding difficulty. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0030] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0031] The following describes a soybean breeding method based on smart agriculture applications. Figure 1 Shown, including: S1: Delineate the survey area Using geographic information system (GIS) technology, combined with satellite images or high-precision maps, the soybean planting area is divided into multiple differentiated plots based on soil type, topography, and irrigation conditions.

[0032] The land parcels classified by differences include: low altitude land parcels, high altitude land parcels, flood irrigation land parcels, drip irrigation land parcels, sprinkler irrigation land parcels, etc. This embodiment is not limited to the above-mentioned land parcels, and the classification is specifically based on actual soil conditions.

[0033] S2: Set the logo Clear signboards are set up at the boundaries of each divided plot, marking the plot number, area, and soil type information. At the same time, a corresponding plot information database is established in the data integration platform, and detailed information of each plot is entered.

[0034] S3: Sensor Configuration According to the characteristics of different plots of land, corresponding sensors are deployed. Soil moisture sensors and nutrient sensors are evenly installed at different locations of each plot according to soil texture and terrain to ensure accurate monitoring of soil moisture and fertility.

[0035] S4: Plot data collection The data collection frequency for different growth stages is set to collect data from each sensor in real time. During the soybean germination and seedling stages, the soil moisture and nutrient sensors collect data every 30 minutes. After entering the flowering and podding stages, the soil-related sensors collect data every 1 hour. The collected data is sent to the data integration platform in real time via the wireless transmission module.

[0036] S5: Growth data collection A dedicated drone flight plan is developed for each plot of land. The appropriate flight route and altitude are planned based on the shape, size and location of the plot. The drone photographs each plot according to the predetermined flight plan to obtain multispectral images. After shooting is completed, the image data is promptly transmitted to the data integration platform via wireless transmission. In the platform, the image data is classified and stored, and folders are created according to the plot number and shooting time.

[0037] S6: Data Summary The soil moisture, fertility, and meteorological data collected by sensors and the multispectral image data taken by drones are summarized in the data integration platform. Through the data association algorithm, data from different sources are matched with the corresponding plot information to ensure the accuracy and completeness of the data.

[0038] S7: Build data analysis model: Using data analysis software and statistical methods, a soybean growth model was developed. Soil moisture, fertility, and meteorological data, along with plant growth data obtained from drone imagery analysis, were used as input variables. Soybean yield, quality, and pest and disease resistance—all desirable traits—were used as output variables. Multiple linear regression and principal component analysis were used to analyze the correlations between these input and output variables. This data analysis identified environmental factors and growth indicators that were significantly correlated with desirable traits. In principal component analysis, principal components corresponding to eigenvalues ​​with cumulative contributions exceeding 80% were selected.

[0039] Leveraging smart agricultural technology, sensors monitor soybean growth environment data in real time. Combined with big data analysis on data integration platforms, this allows for rapid and accurate assessment of the growth performance of different soybean plants under specific conditions, enabling the selection of plants with superior traits in a relatively short period of time. This enables breeders to more quickly identify plants with superior traits, reducing unnecessary breeding generations and significantly shortening the time required from hybridization to obtaining stable, high-quality varieties. This has the potential to reduce the breeding cycle, which originally took several to more than a decade, to a few years or even less.

[0040] The different plots in the survey area are delineated according to the different soil types, altitudes, and coverage of irrigation water sources. Multiple survey areas comprehensively test the growth status of soybeans in different plots, and the data obtained is more accurate.

[0041] The sensors are placed in the following locations for different soil types: Sandy soil: Sandy soil has large particles, rapid water penetration, and poor water retention. Soil moisture varies greatly both vertically and horizontally. In this type of soil, sensors should be placed at different depths and horizontal positions, with sensors inserted vertically at the four corners and the center of the soil.

[0042] Loam: Loam has moderate particle size, good water retention and air permeability, and relatively uniform soil moisture. In loam plots, a more uniform sensor layout can be adopted. The plots are divided into grids, and sensors are inserted at the intersection of each grid.

[0043] Clay: Clay particles are small and have strong water retention, but poor air permeability. Water moves relatively slowly in the soil. For clay plots, the layout of sensors should focus more on monitoring the long-term changes in soil moisture and drainage conditions. Within the plot, select representative low-lying and high locations and set sensors separately.

[0044] The sensors are placed at the following locations at different altitudes: Low-altitude plots: Low-altitude areas usually have higher temperatures, greater water evaporation, and may have higher groundwater levels. In low-altitude plots, sensor placement should consider both soil moisture evaporation and the impact of groundwater levels on soil moisture. Sensors should be placed on the edge of the plot, close to water sources or rivers. High-altitude areas: The temperature in high-altitude areas is lower, precipitation may mainly be snowfall, and the soil freezing period is longer. In high-altitude areas, sensors must first be deployed in different terrain locations.

[0045] The sensors and coverage of different irrigation water sources are configured at the following locations: Flooded plots: During flood irrigation, the distribution of water within the plot is greatly affected by the terrain and the direction of water flow. In flooded plots, sensors must first be installed at the inlet, middle, and outlet of the water flow.

[0046] Drip irrigation plots: The drip irrigation system can control the water supply more accurately. Soil moisture mainly forms a wet area around the drippers. In drip irrigation plots, sensors should be arranged around the drippers.

[0047] Sprinkler irrigation plots: During sprinkler irrigation, water falls from the air in the form of raindrops, and the soil moisture distribution is relatively even. However, there may be insufficient moisture in the edge areas. In sprinkler irrigation plots, sensors are evenly arranged throughout the plot.

[0048] The setting of the sensor directly determines whether the collected data is representative.

[0049] The drone flight plan for growth data collection includes: for regularly shaped plots, a rectangular route can be used; for irregular plots, a route that can provide full coverage is designed. The flight altitude is generally controlled at 100-150 meters to ensure that the captured images can clearly show the details of the soybean plants and cover the entire plot.

[0050] Data aggregation also includes: processing the aggregated data to remove outliers and erroneous data, checking whether there are values ​​in the sensor data that are significantly outside of a reasonable range, denoising the drone image data, and then standardizing the data to convert different types of data into a unified format and units.

[0051] The soybean breeding method also includes S8: abnormal warning.

[0052] A real-time data monitoring module has been established within the data integration platform. This module continuously acquires data uploaded by sensors and drone images and compares it with a preset normal data range. If an anomaly is detected, the platform immediately alerts the operator on the operation interface.

[0053] Once a data anomaly is detected, the platform immediately displays a prominent alert window on the user interface. Highlighted with a red background, the alert window displays the sensor ID, the plot to which it belongs, the specific monitoring indicator (e.g., excessive soil temperature, zero light intensity, and other abnormalities), and the time the anomaly occurred. To help operators quickly locate the problem, clicking on the relevant information in the pop-up window directly redirects them to the sensor's map location and historical data charts, allowing them to analyze anomaly trends. Alerts can also be sent via SMS and email notifications, as well as audio and visual alarms.

[0054] The data integration platform uses AWS IoT Core Internet of Things platform.

[0055] AWS IoT Core can process sensor data. Its scalability allows it to easily handle large-scale IoT device connections and analyze sensor data for soybean breeding, building models. It is well-suited as a data integration platform for this breeding method.

[0056] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A soybean breeding method based on smart agriculture application, characterized by: include: S1: Delineate the survey area Using geographic information system technology, combined with satellite images or high-precision maps, the soybean planting area is divided into multiple plots with different characteristics according to soil type, topography, and irrigation conditions; S2: Set the logo Set up clear signboards at the boundaries of each divided plot, marking the plot number, area, and soil type information. At the same time, establish a plot information database corresponding to the plot in the data integration platform and enter detailed information for each plot; S3: Sensor Configuration Deploy corresponding sensors based on the characteristics of different plots. Install soil moisture sensors and nutrient sensors evenly at different locations on each plot according to soil texture and topography. Set up meteorological monitoring devices in different plots. S4: Plot data collection Set the data collection frequency for soybeans at different growth stages and collect data from each sensor in real time; S5: Growth data collection A dedicated drone flight plan is developed for each plot of land. The appropriate flight route and altitude are planned based on the shape, size, and location of the plot. The drone photographs each plot according to the predetermined flight plan to obtain multispectral images. After the shooting is completed, the image data is promptly transmitted to the data integration platform via wireless transmission. In the platform, the image data is classified and stored, and folders are created according to the plot number and shooting time. S6: Data Summary The soil moisture, fertility, and meteorological data collected by sensors and multispectral image data captured by drones are aggregated in a data integration platform. Using data association algorithms, data from different sources are matched with corresponding plot information. S7: Build data analysis model Using data analysis software and statistical methods, a soybean growth model was established. Soil moisture, fertility, meteorological environment data, and plant growth data obtained from drone image analysis were used as input variables, and soybean yield, quality, disease and pest resistance and other excellent traits were used as output variables. The correlation between each input variable and output variable was analyzed through multiple linear regression and principal component analysis statistical methods. Through data analysis, the environmental factors and growth indicators that were significantly correlated with excellent traits were found.

2. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The different plots in the demarcated survey area are demarcated according to different soil types, altitudes, and coverage of irrigation water sources.

3. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The sensors are placed in the following locations in different soil types: Sandy soil: Insert sensors vertically into the four corners and the center of the plot. Loam: Divide the land into grids and insert sensors at the intersection of each grid. Clay: Select representative low-lying and high-altitude locations and set up sensors in each location.

4. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The sensors are located at the following locations at different altitudes: Low-altitude plots: Insert the sensor at the edge of the plot, close to the water source or river; High-altitude areas: sensors are deployed in different terrain locations.

5. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The sensors and coverage of different irrigation water sources are configured at the following locations: Flood irrigation plots: Set sensors at the inlet, middle and outlet of the water flow; Drip irrigation plots: sensors should be arranged around the drippers; Sprinkler irrigation plots: Sensors are evenly distributed throughout the plot.

6. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The UAV flight plan for the growth data collection includes: for regularly shaped plots, a rectangular route can be used for flight; for irregular plots, a route that can fully cover the plots is designed, and the flight altitude is controlled at 100-150 meters.

7. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The data aggregation also includes: processing the aggregated data, removing outliers and erroneous data, denoising the drone image data, and then standardizing the data to convert different types of data into a unified format and unit.

8. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The soybean breeding method also includes S8: abnormal warning. In the data integration platform, a real-time data monitoring module is established. The module continuously obtains data uploaded by sensors and drone images, and compares it with the preset normal data range. When data abnormality is detected, the platform alerts the operator on the operation interface.

9. The soybean breeding method based on smart agriculture application according to claim 1, characterized in that: The data integration platform uses AWS IoT Core Internet of Things platform.

Images