A method and system for generating a crop variable seeding prescription map based on long-term dynamics and habitat resilience matching

Abstract

The application relates to a crop variable seeding prescription map generation method and system based on long-term dynamics and habitat resilience matching, relates to the field of agricultural intelligent equipment and precision agriculture technology, and is used for solving the problems that the existing variable seeding relies on a prescription map, resulting in poor adaptability, lagging adjustment response, insufficient control precision and difficult guarantee of seeding uniformity, generating variable seeding decision parameters corresponding to a space position through multi-source data of a target operation plot, generating continuous control instructions based on the variable seeding decision parameters and carrying out smooth adjustment processing, coupling and correcting the continuous control instructions and operation state parameters to form control signals, driving a seeding device to implement variable seeding and obtaining seeding state information according to the control signals, forming feedback information based on the seeding state information and carrying out closed-loop correction on the variable seeding decision parameters and the continuous control instructions, and being applicable to farmland precision seeding operation and intelligent agricultural equipment control scenes.

Description

Technical Field

[0001] This invention relates to the field of precision agriculture (smart agriculture) technology, specifically to the field of crop growth dynamics simulation. Background Technology

[0002] Under the current trend of deep integration of agricultural mechanization and informatization, precision seeding technology is gradually evolving from traditional quantitative seeding to variable seeding. This aims to optimize seed input based on different plot conditions, thereby increasing crop yield and reducing seed and resource waste. In existing technological systems, variable seeding mainly relies on a "prescription map + actuator" control model. This involves using information such as satellite remote sensing, historical yield data, and soil nutrient testing results to grid or partition farmland, generating corresponding seeding prescription maps. The seeder then executes the seeding operation by calling different seeding parameters based on location changes. For example, some systems use GPS positioning and GIS geographic information systems to achieve real-time plot location identification and adjust the seed metering motor speed through an electronic control unit, thereby changing the seeding rate per unit area. Other solutions incorporate soil sensors or crop growth monitoring equipment to attempt dynamic adjustment of the seeding rate based on real-time environmental information.

[0003] Furthermore, at the execution level, existing variable seeding devices typically achieve variations in seeding density by controlling the rotational speed, opening and closing frequency, or seeding wheel structure of the seed metering device. Some high-end equipment also incorporates electric drive technology and closed-loop control algorithms to achieve a certain degree of fine-tuning of the seeding process. Simultaneously, to improve operational efficiency, some solutions combine variable seeding systems with agricultural machinery automatic driving systems to achieve integrated operation of path planning and seeding control. In addition, some technologies attempt to achieve differentiated seeding between different rows through independent control of multiple rows to adapt to complex plot conditions.

[0004] While the aforementioned technologies have improved the accuracy and automation of sowing to some extent, they still have significant shortcomings in practical applications. First, most existing technologies rely on pre-generated prescription maps, which have long data update cycles and struggle to reflect dynamic changes in the field environment in a timely manner. When soil conditions, climate conditions, or crop requirements change, sowing strategies are adjusted lag behind, leading to discrepancies between the sown quantity and actual needs. Second, although some real-time adjustment schemes incorporate sensor data, their control logic is mostly single-variable or simple rule-based, lacking the ability to fuse and process multi-source data. This makes it difficult to comprehensively consider the impact of various factors such as soil, environment, and operational status on sowing effectiveness, thus limiting the precision of variable-rate sowing.

[0005] Meanwhile, in practical implementation, existing seeding control structures suffer from response delays and control instability during dynamic adjustments. When seeding parameters change rapidly, the seeding device struggles to adjust promptly and accurately, easily leading to uneven seeding rates in transition zones. Furthermore, insufficient coordination between the seeding device and control system can result in duplicate seeding, missed seeding, or inconsistent seeding spacing under complex conditions, affecting seedling uniformity and crop growth consistency. On the other hand, existing systems primarily focus on adjusting the seeding rate during variable control, neglecting comprehensive optimization of seeding accuracy, seed distribution uniformity, and operational continuity, leaving room for improvement in overall seeding quality.

[0006] Furthermore, existing variable seeding systems also have certain limitations in terms of hardware and software coordination. For example, the lack of tight coupling between the control algorithm and the actuator makes it difficult to efficiently convert control commands into stable mechanical actions; the system has limited adaptability to different crop types and different plot conditions, and its versatility and scalability are insufficient; at the same time, its anti-interference ability in complex environments is weak, affecting the reliability and stability of variable seeding control.

[0007] In summary, existing technologies suffer from several drawbacks, including insufficient adaptability due to variable seeding relying on prescription maps, limited real-time adjustment capabilities, weak multi-source information fusion capabilities, delayed response of seeding control, insufficient dynamic adjustment precision, and difficulty in ensuring seeding uniformity. Summary of the Invention

[0008] To address the shortcomings of existing technologies, such as insufficient adaptability due to variable seeding dependence on prescription maps, limited real-time adjustment capabilities, weak multi-source information fusion capabilities, delayed response of seeding control, insufficient dynamic adjustment accuracy, and difficulty in guaranteeing seeding uniformity, the technical solution provided by this invention is as follows: A method for generating crop seed prescription maps based on crop growth dynamics and habitat resilience matching includes: The steps to acquire multi-source data of the target work site and construct an initial dataset include: performing multi-source heterogeneous data fusion and spatiotemporal alignment processing on geographic location information, soil attribute information and historical planting data, and outputting a structured initial dataset. The steps of performing fusion analysis on the initial dataset and generating variable sowing decision parameters are as follows: determining sowing requirements based on soil conditions and historical performance at each location, and outputting variable sowing decision parameters corresponding to spatial locations. The steps of generating continuous control commands and performing smooth adjustment based on the variable seeding decision parameters are as follows: continuously mapping the decision parameters at each location according to the operation path and limiting the rate of parameter change, and outputting continuous control commands. The step of coupling and correcting the continuous control command with the operation status parameters to form a control signal involves synchronously adjusting the control command according to the equipment's travel status to match the sowing requirements per unit area, and then outputting the control signal. The control signal drives the seeding device to implement variable seeding steps, and the seeding density in different areas is dynamically changed by adjusting the seeding execution rhythm and the seeding status information is output. The steps of acquiring the sowing status information and generating feedback information include monitoring the sowing execution results and outputting feedback information. The step of performing closed-loop correction of the variable seeding decision parameters and the continuous control command based on the feedback information involves dynamically adjusting the subsequent control process according to the actual seeding deviation and outputting the corrected variable seeding decision parameters.

[0009] Furthermore, in a preferred embodiment, the multi-source data includes real-time geographic location information obtained through the positioning unit, soil moisture and nutrient information obtained through the soil detection unit, and historical planting data, and the various types of data are processed for temporal and spatial alignment.

[0010] Furthermore, in a preferred embodiment, when performing fusion analysis on the initial dataset, the data from different sources are normalized and weighted according to their degree of influence.

[0011] Furthermore, in a preferred embodiment, the continuous control command is generated by interpolating the seeding decision parameters of the variables in adjacent spatial locations and constraining the rate of change of the parameters.

[0012] Furthermore, in a preferred embodiment, the control signal is coupled and corrected based on continuous control commands and the travel speed of the working equipment.

[0013] Furthermore, in a preferred embodiment, the sowing status information is collected by a detection unit installed at the sowing device. The detection content includes sowing frequency and seed distribution status, and feedback information is generated.

[0014] A prescription map generation system for crop seeding control based on growth dynamics and habitat resilience matching includes: The module acquires multi-source data of the target operation site and constructs an initial dataset. It performs multi-source heterogeneous data fusion and spatiotemporal alignment processing on geographic location information, soil attribute information and historical planting data, and outputs a structured initial dataset. The module performs fusion analysis on the initial dataset and generates variable sowing decision parameters. It determines the sowing requirements based on the soil conditions and historical performance of each location and outputs variable sowing decision parameters corresponding to the spatial location. The module generates continuous control commands and performs smooth adjustment based on the variable seeding decision parameters. It continuously maps the decision parameters at each location according to the operation path and limits the rate of parameter change, and outputs continuous control commands. The module that couples and modifies the continuous control commands with the operation status parameters to form control signals, synchronously adjusts the control commands according to the equipment's travel status to match the sowing requirements per unit area, and outputs control signals. The module that drives the seeding device to perform variable seeding according to the control signal achieves dynamic changes in seeding density in different areas by adjusting the seeding execution rhythm and outputs seeding status information. The module that acquires the sowing status information and generates feedback information monitors the sowing execution results and outputs feedback information. The module that performs closed-loop correction of the variable seeding decision parameters and the continuous control command based on the feedback information dynamically adjusts the subsequent control process according to the actual seeding deviation and outputs the corrected variable seeding decision parameters.

[0015] A computer storage medium for storing a computer program, which, when read by the computer, is executed by the computer using the method described thereon.

[0016] A computer, including a processor and a storage medium, executes the method when the processor reads a computer program stored in the storage medium.

[0017] A computer program product, which, as a computer program, implements the method when the computer program is executed.

[0018] Compared with the prior art, the advantages of the technical solution provided by the present invention are as follows: By introducing a variable-based seeding decision-making mechanism based on multi-source data fusion, soil information, geographic location data, and operational status parameters are processed in a unified manner to form a basis for seeding control. Compared with relying solely on a single prescription map or a single sensor, this approach achieves comprehensive perception of spatial differences and temporal changes in plots, thereby improving the accuracy and adaptability of seeding quantity decisions. This effect stems from the multi-source information acquisition and fusion processing component of the solution.

[0019] By constructing a real-time dynamic adjustment control process, the sowing parameters can be continuously updated with the operation process and directly drive the sowing device to execute. Compared with the traditional preset parameters or delayed update method, it effectively shortens the response time from environmental changes to sowing adjustments and reduces the accumulation of sowing errors. This effect comes from the real-time control and dynamic decision output part of the solution.

[0020] By adopting a refined control method for the seeding actuator, the control signal and seeding action are matched with high precision, and the seeding amount and seeding rhythm are adjusted synchronously. Compared with the method of simply adjusting the motor speed, the accuracy and stability of the seeding amount per unit area are improved. This effect comes from the drive and control execution part of the seeding device in the solution.

[0021] By introducing a continuous transition control strategy during variable adjustment, the switching process between different sowing parameters is kept smooth, avoiding uneven sowing caused by abrupt adjustments. Compared with segmented or step-like adjustment methods, this effectively reduces reseeding and missed sowing in the transition area. This effect comes from the variable adjustment smoothing part of the scheme.

[0022] By establishing a closed-loop feedback mechanism for the sowing process, the sowing status information is fed back to the control system and participates in subsequent adjustments, enabling the control strategy to continuously correct deviations. Compared with open-loop control, this improves the system's stability and anti-interference capability in complex environments. This effect comes from the status feedback and error correction components in the scheme.

[0023] By independently controlling the multi-row or multi-unit seeding structure, the seeding parameters can be adjusted separately for different areas or between different rows according to actual needs. Compared with the overall unified control method, this enhances the system's adaptability to complex plot conditions and improves the precision of seeding spatial distribution. This effect comes from the multi-unit independent control structure in the scheme.

[0024] By optimizing the structural design and control coordination of the seeding device, the seed release process is made more uniform and controllable. Even under dynamic adjustment conditions, it can maintain a stable seed spacing distribution. Compared with the problem of uneven distribution that is easy to occur under variable speed conditions in traditional structures, it significantly improves the seeding uniformity. This effect comes from the coordinated design of the seeding structure and control in the solution.

[0025] By comprehensively considering the impact of changes in operating speed on the seeding rate during the control process and synchronously compensating for the seeding parameters, the seeding density deviation caused by speed fluctuations is effectively avoided compared to control methods that ignore changes in travel speed. This effect comes from the coupled control part of the operating state parameters in the scheme.

[0026] By constructing an integrated hardware and software control system, the algorithm output can be efficiently transformed into stable actions of the actuator. Compared with systems with a high degree of separation between control and execution, this reduces control delay and execution error, and improves the overall system reliability. This effect comes from the collaborative design of the control system and actuator in the solution.

[0027] By enhancing the system's adaptability to different plot conditions and crop requirements, the same equipment can achieve stable variable seeding control in various application scenarios. Compared with traditional systems that are more specialized, this improves the system's versatility and promotion value. This effect comes from the parameter adaptation and strategy adjustment mechanism in the solution.

[0028] It is suitable for intelligent sowing operation scenarios in agricultural production that achieve precise variable sowing control based on plot differences and operational status. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating a method for generating crop seed prescription maps based on matching growth dynamics and habitat resilience. Detailed Implementation

[0030] To make the advantages and benefits of the technical solution provided by the present invention clearer, the technical solution provided by the present invention will now be described in further detail with reference to the accompanying drawings, specifically: Implementation Method 1: This implementation method provides a method for generating crop variable sowing prescription maps based on growth dynamics and habitat resilience matching, including: The steps to acquire multi-source data of the target work site and construct an initial dataset include: performing multi-source heterogeneous data fusion and spatiotemporal alignment processing on geographic location information, soil attribute information and historical planting data, and outputting a structured initial dataset. The steps of performing fusion analysis on the initial dataset and generating variable sowing decision parameters are as follows: determining sowing requirements based on soil conditions and historical performance at each location, and outputting variable sowing decision parameters corresponding to spatial locations. The steps of generating continuous control commands and performing smooth adjustment based on the variable seeding decision parameters are as follows: continuously mapping the decision parameters at each location according to the operation path and limiting the rate of parameter change, and outputting continuous control commands. The step of coupling and correcting the continuous control command with the operation status parameters to form a control signal involves synchronously adjusting the control command according to the equipment's travel status to match the sowing requirements per unit area, and then outputting the control signal. The control signal drives the seeding device to implement variable seeding steps, and the seeding density in different areas is dynamically changed by adjusting the seeding execution rhythm and the seeding status information is output. The steps of acquiring the sowing status information and generating feedback information include monitoring the sowing execution results and outputting feedback information. The step of performing closed-loop correction of the variable seeding decision parameters and the continuous control command based on the feedback information involves dynamically adjusting the subsequent control process according to the actual seeding deviation and outputting the corrected variable seeding decision parameters.

[0031] The multi-source data includes real-time geographic location information obtained through the positioning unit, soil moisture and nutrient information obtained through the soil detection unit, and historical planting data, and performs time and space alignment processing on various types of data.

[0032] When performing fusion analysis on the initial dataset, data from different sources are normalized and weighted according to their degree of influence.

[0033] Continuous control commands are generated by interpolating the decision parameters of the variable seeding at adjacent spatial locations and constraining the rate of parameter change.

[0034] The control signal is coupled and corrected based on continuous control commands and the travel speed of the working equipment.

[0035] Sowing status information is collected by a detection unit installed at the sowing device. The detection content includes sowing frequency and seed distribution status, and feedback information is generated.

[0036] A prescription map generation system for crop seeding control based on growth dynamics and habitat resilience matching includes: The module acquires multi-source data of the target operation site and constructs an initial dataset. It performs multi-source heterogeneous data fusion and spatiotemporal alignment processing on geographic location information, soil attribute information and historical planting data, and outputs a structured initial dataset. The module performs fusion analysis on the initial dataset and generates variable sowing decision parameters. It determines the sowing requirements based on the soil conditions and historical performance of each location and outputs variable sowing decision parameters corresponding to the spatial location. The module generates continuous control commands and performs smooth adjustment based on the variable seeding decision parameters. It continuously maps the decision parameters at each location according to the operation path and limits the rate of parameter change, and outputs continuous control commands. The module that couples and modifies the continuous control commands with the operation status parameters to form control signals, synchronously adjusts the control commands according to the equipment's travel status to match the sowing requirements per unit area, and outputs control signals. The module that drives the seeding device to perform variable seeding according to the control signal achieves dynamic changes in seeding density in different areas by adjusting the seeding execution rhythm and outputs seeding status information. The module that acquires the sowing status information and generates feedback information monitors the sowing execution results and outputs feedback information. The module that performs closed-loop correction of the variable seeding decision parameters and the continuous control command based on the feedback information dynamically adjusts the subsequent control process according to the actual seeding deviation and outputs the corrected variable seeding decision parameters.

[0037] A computer storage medium for storing a computer program, which, when read by the computer, is executed by the computer using the method described thereon.

[0038] A computer, including a processor and a storage medium, executes the method when the processor reads a computer program stored in the storage medium.

[0039] A computer program product, which, as a computer program, implements the method when the computer program is executed.

[0040] Implementation Method Two: This implementation method is a further detailed description of the technical solution provided in Implementation Method One, specifically: The steps to collect multi-source data from the target work site and construct an initial dataset By spatially locating the work area and simultaneously collecting soil and environmental information, the basic data input for variable seeding is formed.

[0041] After the operating equipment enters the target plot, the positioning unit continuously obtains the current position of the equipment and subdivides the plot into a grid, so that each position unit has a unique spatial identifier. At the same time, the soil detection unit obtains information such as soil moisture, nutrient content and structural characteristics at the corresponding position, and combines the environmental sensing unit to obtain temperature, humidity and other environmental parameters that affect sowing. Then, the historical planting data and historical yield distribution data of the plot are retrieved. The above data are matched and aligned according to the time dimension and spatial location, outliers are removed and uniformly formatted, and finally a structured initial dataset containing location, soil properties and historical performance is formed as input for subsequent processing.

[0042] The steps to merge the initial dataset and generate variable seeding decision parameters The collected multi-source data is comprehensively analyzed and the corresponding sowing control parameters for each location are output.

[0043] The initial dataset is input into the data processing unit, where data from different sources are normalized and weighted according to their impact on the sowing results. By comprehensively analyzing the soil fertility level, water content, and historical yield performance at each location, the appropriate sowing density for the corresponding area is determined. The analysis results are then transformed into specific sowing control parameters, including the number of seeds per unit area or the sowing rhythm parameters, so that each spatial location corresponds to a set of clear variable sowing parameters, which serve as the direct basis for subsequent control execution.

[0044] The steps of generating continuous control commands based on variable seeding decision parameters and performing transition adjustment. The discrete seeding decision results are transformed into continuously changing control commands and smooth adjustment is achieved.

[0045] As the equipment moves along the working path, the seeding control parameters at the corresponding locations are called point by point based on real-time location information. The parameter changes between adjacent locations are interpolated based on the continuity of the path, so that the control commands form a continuous change curve in time and space. At the same time, the rate of change of control parameters is limited by adjustment strategies to avoid abrupt changes, thereby ensuring that the seeding device operates smoothly during execution and transmitting the smoothed control commands to the execution control unit.

[0046] The steps of coupling continuous control commands with work status parameters to form the final control signal. Control commands are adjusted in real time based on the equipment's operating status to ensure sowing accuracy.

[0047] After the continuous control command is generated, the system collects the status parameters such as the equipment's travel speed and work rhythm in real time. The status parameters are then correlated with the control command. When the equipment speed changes, the seeding control parameters are adjusted synchronously to ensure that the seeding rate per unit time is consistent with the target seeding rate per unit area. This avoids deviations in seeding density caused by speed fluctuations, and the corrected control results are then used to form the final control signal.

[0048] The steps of driving the seeding device to achieve variable seeding according to the final control signal The seeding device is adjusted according to the control signal to complete the seeding operation in different areas.

[0049] The final control signal is transmitted to the seeding drive unit. By adjusting the rotation speed or opening and closing frequency of the seeding device, the seeds are released according to the set seeding parameters. During the forward movement of the equipment, each position automatically adjusts the seeding status according to the corresponding control parameters, thereby realizing the dynamic change of seeding density in different areas and ensuring the continuity and stability of seed release.

[0050] The steps to monitor the status of the sowing process and generate feedback information The actual sowing situation is monitored and feedback data is generated for subsequent corrections.

[0051] A detection unit is set up at or near the sowing device to collect data in real time on the actual sowing frequency, seed passage, and seed distribution. The detection results are converted into feedback data that can be used for analysis and stored in correspondence with the current control parameters to reflect the difference between the actual performance and the expected results.

[0052] The steps to perform closed-loop correction of the control process based on feedback information Feedback data is used to dynamically adjust seeding control to improve accuracy.

[0053] Feedback data is input into the control unit and compared with the original sowing control parameters. When seeding deviation or uneven distribution is detected, the control parameters of subsequent positions are corrected and the control commands are fine-tuned, so that the system gradually eliminates errors during continuous operation and achieves adaptive adjustment and precision improvement in the sowing process.

[0054] The steps to independently control multi-row or multi-unit seeding structures to achieve regionally differentiated sowing. Different seeding units are controlled separately to adapt to the differences in complex plots.

[0055] In multi-row seeding equipment, each seeding unit is matched with its corresponding plot location. Independent seeding control parameters and control instructions are generated for each seeding unit. During execution, the synchronous and coordinated operation between the units is maintained, so that different areas can be seeded according to their own conditions. This achieves refined control in the overall operation and improves seeding uniformity and adaptability.

[0056] Implementation Method 3, in conjunction with Appendix Figure 1 This embodiment describes the technical solution provided above in further detail through specific examples. Specifically: like Figure 1 As shown, the overall structure of the variable seeding decision-making process of this scheme is illustrated. It starts from the acquisition of multi-source data, and after the extraction of environmental and crop information, feature quantification and modeling analysis, it finally forms the variable seeding strategy output for different regions. The whole presents a logical relationship of top-down layer-by-layer processing and step-by-step evolution from data to decision.

[0057] At the top layer, the multi-source data acquisition module is used to obtain basic input information. This part includes obtaining topographic and surface information through L1 point cloud and M3M imagery, and obtaining soil attribute data through soil sampling, thus forming the original data foundation. The above data is passed down to the environmental covariate extraction module. In this module, topographic factors, including slope, topographic humidity index and light conditions, are extracted based on the digital elevation model. At the same time, the bare soil spectral information is combined to form key variables reflecting the differences in the plot environment. Meanwhile, another branch obtains crop growth information through the crop dynamic growth monitoring module. This part models and represents the crop growth process based on the normalized difference red edge index and the double logistic growth curve fitting method, thereby obtaining the dynamic change characteristics of crops.

[0058] In the intermediate layer, environmental and crop information is further transformed into quantitative features that can be used for decision-making. On the one hand, the static habitat resilience quantification module characterizes long-term stable environmental attributes, including comprehensive terrain score, terrain risk factor, and light potential parameters, thereby reflecting the basic productivity and risk level of different regions. On the other hand, the dynamic growth feature extraction module extracts the stage-based and trend-based features of crop growth, including sensitive parameters of key growth stages, growth potential persistence indicators, and cumulative photosynthetic trends, thereby reflecting the crop's response to current environmental conditions. The above static and dynamic features characterize plot differences from two dimensions: environmental foundation and crop performance, and together serve as inputs to the subsequent decision-making model.

[0059] At the core processing layer, the above-mentioned multidimensional features are fused and analyzed through a variable seeding decision model. This model is constructed based on a combination of physical mechanisms and data-driven approaches, and introduces a risk decoupling mechanism to distinguish between environmental risks and growth potential, thereby enabling refined decision-making on seeding strategies for different regions. The model outputs variable seeding decision results for different spatial units.

[0060] At the output layer, differentiated application strategies are formed based on the model results. For high-potential areas, a denser sowing strategy is adopted to make full use of high fertility and low-risk conditions to increase the yield ceiling. For high-risk areas, a reduced-density sowing strategy is adopted to reduce resource waste and ensure crop survival rate. The overall sowing efficiency is optimized through differentiated control of different regions.

[0061] The core of this implementation lies in extracting the temporal dynamic characteristics of crop canopy from multi-source remote sensing data and integrating multi-dimensional static data such as topography and soil to construct a source-sink matching index (SSMI) that couples crop dynamic growth (CVI) and static habitat resilience (SLRI), automatically generating a high-precision gridded variable seed prescription map. Specific steps and control parameters are as follows: Step 1: Acquisition and Preprocessing of Multi-Source Heterogeneous Agricultural Data Acquire UAV RGB and multispectral image sequences of the target plot covering the entire growth period of the crop (e.g., 11 stages from emergence to maturity), and extract time-series data of Normalized Red Edge Differential Vegetation Index (NDRE). Acquire five UAV LiDAR data points for the late reproductive growth stage (R1-R6), and extract high-precision digital elevation model (DEM), slope, aspect, and topographic moisture index (TWI) from these data. Acquire high-resolution (e.g., 3-meter) ground-measured soil physicochemical data, including pH, soil organic matter (SOM), available nitrogen (AN), available phosphorus (AP), and available potassium (AK).

[0062] Step 2: Construct the dynamic canopy vitality index (CVI) This step aims to quantify the temporal dynamics of crop growth: Time series reconstruction and denoising: The Savitzky-Golay (SG) filtering algorithm was used to smooth the original 11-period NDRE time series. A second-order polynomial (Order=2) was selected, and the sliding window size was set to 3 (Window Size=3). This step effectively removed background noise such as cloud shadows and preserved the peak characteristics of the curves to the greatest extent.

[0063] Dynamic mechanism modeling: A double logistic model is used to characterize crop growth trajectory and deconstruct the processes of vegetative growth and reproductive senescence. The mathematical model is as follows:

[0064] Physical meaning of parameters: t is the number of days; V base :NDRE min This refers to the basic spectral values ​​of bare soil or before seedling emergence; V amp :Amplitude,NDRE max -NDRE min m1 represents the maximum span amplitude (representing the maximum potential biomass); m2 represents the vegetative growth rate (controlling the speed of canopy closure). The greenness peak day; m2 represents the reproductive aging rate; This marks the start of the aging process.

[0065] Parameter optimization and feature extraction: The Levenberg-Marquardt (LM) algorithm is introduced for nonlinear iterative solution. Two core indicators are extracted: maximum growth rate (Vmax, representing seedling explosive growth) and efficient photosynthetic period (EPW, defined as the number of days that NDRE remains above 80% of its peak value, reflecting greenness retention during grain filling).

[0066] Calculate CVI: (in, This index is used to screen high-yielding plant types that combine "fast start-up, long green period, and slow aging" to determine aging rate.

[0067] Step 3: Construct the Static Habitat Resilience Index (SLRI) This step aims to quantify the baseline capacity of farmland ecosystems to maintain stable yields in the face of topographic heterogeneity: 1. Comprehensive Soil Fertility Score (SFS) Calculation: First, the ranges of each soil physicochemical index are normalized to the [0, 1] interval. The formula is as follows: Principal component analysis (PCA) was then used to determine the weights:

[0068] 2. Integrated Environmental Risk (IER) Physical Coupling Modeling: Slope risk (non-linear threshold penalty): Set the hydrological safety physical critical shear force threshold.

[0069] (Exponential 1.5 simulates the nonlinear acceleration of water erosion with increasing slope).

[0070] Water Risk (Physiological Stress Decoupling): Calculating the Two-Way Penalty for Deviation from the Median after Normalizing TWI (Capture low-lying waterlogged areas and high-altitude drought).

[0071]

[0072] Light and heat risk (reverse measurement of local habitat):

[0073] Comprehensive IER calculation:

[0074] 3. Calculate SLRI:

[0075] Step 4: Construction of Crop Growth Index (SSMI) and Generation of Prescription Map 1. Data centralization: Mean centralization is applied to both CVI and SLRI. ), so that the value 1.0 represents the average level of the whole field.

[0076] 2. Calculate the Source-Library Matching Index (SSMI):

[0077] 3. Grid-based prescription generation: Divide the target plot into a 6m × 6m grid matching the operating width of mainstream seeders. Assign prescription rules: Resource surplus type (SSMI<0.8): The soil fertility is stable, but the current growth has not fully utilized the potential. Decision: Increase seeding density to fill the ecological niche.

[0078] Resource-depleted type (SSMI>1.2): Plants exhibit artificially high growth but have fragile habitats (e.g., low-lying areas with high water risk), making them highly susceptible to lodging or photosynthetic collapse. Decision: Reduce sowing density to restore balance.

[0079] 4. Output job file: Export the grid containing density decisions as a Shapefile format file and import it directly into the agricultural machinery for execution.

[0080] Innovation and Benefits Innovation Point 1 (Mechanism Breakthrough): The "Water Physiological Stress Decoupling Model" was proposed, which for the first time separated the contradiction between "false high greenness" caused by nutrient enrichment in low-lying areas in the early stage and the actual "root hypoxia-induced senescence" at the algorithm level, avoiding the fatal defect of traditional algorithms that erroneously increase the sowing density in low-lying areas.

[0081] Innovation Point 2 (Model Innovation): The Double Logistic function is used to characterize the growth of the entire growth period, which overcomes the "Runge phenomenon" (fitting distortion) that is prone to occur at the beginning and end of the traditional polynomial model. The whole field fitting determination coefficient R2 is as high as 0.91.

[0082] Innovation Point 3 (Engineering Implementation): It breaks through the barrier of misalignment between traditional remote sensing pixels and agricultural machinery operations, and uniformly sets a 6m * 6m grid that strictly matches the physical width of mainstream agricultural machinery in VRS, realizing "plug and play" control files.

[0083] In one specific embodiment: Taking the variable planting operation of corn in plot 1-1-4 of Heshan Farm, Heilongjiang Province (double-row planting mode) as an example.

[0084] First, the system generated 1823 rectangular grid layers covering the entire field at a physical size of 6m × 6m. The system automatically retrieved UAV NDRE imagery sets covering 11 key growth stages from the previous growing season for this plot, and ran the built-in Savitzky-Golay filtering and LM optimization algorithm to calculate the maximum growth rate (V) for each grid. max Parameters such as ) are used to generate a dynamic CVI layer.

[0085] Subsequently, the system loaded the 3-meter resolution DEM and soil physicochemical properties (AN, SOM, etc.) of the site. During the system calculations, a certain micro-topography within the site was identified as a low-lying waterlogged area: although its Comprehensive Soil Fertility Score (SFS) was extremely high, the slope assessment and two-way moisture risk assessment based on a θ=2.64° threshold showed that this grid had an extremely high waterlogging penalty value. This led to a significant decrease in its static habitat resilience index (SLRI).

[0086] Final matching calculations revealed that the SSMI of this depression grid reached 1.35 (>1.2, indicating resource overdraft). Based on this, the system inferred that the historically high greenness of this grid was a "false appearance," actually facing an extremely high risk of lodging and root hypoxia. Therefore, the system automatically generated a "reduce seeding density" operation instruction in the prescription attributes of this grid. Conversely, in the gentle slope resource-surplus grid with SSMI <0.8, the system issued an instruction to increase seeding density.

[0087] The system exported these 1823 grids with clearly defined plant spacing and sowing density as a standard Shapefile file. The operator imported this file via USB drive into a tractor equipped with a variable seeding control terminal, successfully completing the precision sowing operation.

[0088] The above description of several specific embodiments further details the technical solution provided by the present invention in order to highlight the advantages and benefits of the technical solution provided by the present invention. However, the above-described specific embodiments are not intended to limit the present invention. Any reasonable modifications and improvements to the present invention, combinations of embodiments, and equivalent substitutions based on the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for generating crop seed prescription maps based on growth dynamics and habitat resilience matching, characterized in that, include: The steps to acquire multi-source data of the target work site and construct an initial dataset include: performing multi-source heterogeneous data fusion and spatiotemporal alignment processing on geographic location information, soil attribute information and historical planting data, and outputting a structured initial dataset. The steps of performing fusion analysis on the initial dataset and generating variable sowing decision parameters are as follows: determining sowing requirements based on soil conditions and historical performance at each location, and outputting variable sowing decision parameters corresponding to spatial locations. The steps of generating continuous control commands and performing smooth adjustment based on the variable seeding decision parameters are as follows: continuously mapping the decision parameters at each location according to the operation path and limiting the rate of parameter change, and outputting continuous control commands. The step of coupling and correcting the continuous control command with the operation status parameters to form a control signal involves synchronously adjusting the control command according to the equipment's travel status to match the sowing requirements per unit area, and then outputting the control signal. The control signal drives the seeding device to implement variable seeding steps, and the seeding density in different areas is dynamically changed by adjusting the seeding execution rhythm and the seeding status information is output. The steps of acquiring the sowing status information and generating feedback information include monitoring the sowing execution results and outputting feedback information. The step of performing closed-loop correction of the variable seeding decision parameters and the continuous control command based on the feedback information involves dynamically adjusting the subsequent control process according to the actual seeding deviation and outputting the corrected variable seeding decision parameters.

2. The method for generating crop variable sowing prescription maps based on growth dynamics and habitat resilience matching according to claim 1, characterized in that, The multi-source data includes real-time geographic location information obtained through the positioning unit, soil moisture and nutrient information obtained through the soil detection unit, and historical planting data, and performs time and space alignment processing on various types of data.

3. The method for generating crop variable sowing prescription maps based on growth dynamics and habitat resilience matching according to claim 1, characterized in that, When performing fusion analysis on the initial dataset, data from different sources are normalized and weighted according to their degree of influence.

4. The method for generating crop variable sowing prescription maps based on growth dynamics and habitat resilience matching according to claim 1, characterized in that, Continuous control commands are generated by interpolating the decision parameters of the variable seeding at adjacent spatial locations and constraining the rate of parameter change.

5. The method for generating crop variable sowing prescription maps based on growth dynamics and habitat resilience matching according to claim 1, characterized in that, The control signal is coupled and corrected based on continuous control commands and the travel speed of the working equipment.

6. The method for generating crop variable sowing prescription maps based on growth dynamics and habitat resilience matching according to claim 1, characterized in that, Sowing status information is collected by a detection unit installed at the sowing device. The detection content includes sowing frequency and seed distribution status, and feedback information is generated.

7. A prescription map generation system for crop variable sowing control based on growth dynamics and habitat resilience matching, characterized in that, include: The module acquires multi-source data of the target operation site and constructs an initial dataset. It performs multi-source heterogeneous data fusion and spatiotemporal alignment processing on geographic location information, soil attribute information and historical planting data, and outputs a structured initial dataset. The module performs fusion analysis on the initial dataset and generates variable sowing decision parameters. It determines the sowing requirements based on the soil conditions and historical performance of each location and outputs variable sowing decision parameters corresponding to the spatial location. The module generates continuous control commands and performs smooth adjustment based on the variable seeding decision parameters. It continuously maps the decision parameters at each location according to the operation path and limits the rate of parameter change, and outputs continuous control commands. The module that couples and modifies the continuous control commands with the operation status parameters to form control signals, synchronously adjusts the control commands according to the equipment's travel status to match the sowing requirements per unit area, and outputs control signals. The module that drives the seeding device to perform variable seeding according to the control signal achieves dynamic changes in seeding density in different areas by adjusting the seeding execution rhythm and outputs seeding status information. The module that acquires the sowing status information and generates feedback information monitors the sowing execution results and outputs feedback information. The module that performs closed-loop correction of the variable seeding decision parameters and the continuous control command based on the feedback information dynamically adjusts the subsequent control process according to the actual seeding deviation and outputs the corrected variable seeding decision parameters.

8. A computer storage medium for storing computer programs, characterized in that, When the computer program is read by the computer, the computer executes the method of claim 1.

9. A computer, comprising a processor and a storage medium, characterized in that, When the processor reads the computer program stored in the storage medium, the computer executes the method of claim 1.

10. A computer program product, as a computer program, is characterized by: When the computer program is executed, it implements the method of claim 1.

Images