Intelligent peak shifting seeding and harvesting decision-making method and system for fresh corn

Abstract

The invention discloses a fresh corn intelligent peak shifting sowing and harvesting decision-making method and system, and relates to the technical field of intelligent agriculture. The method comprises the following steps: acquiring multi-source data of a target area, and generating standardized state codes of each plot through preprocessing; daily step length simulation is carried out based on the code-driven growth quality model, the filament activity degree is obtained by analyzing an ear image, the quality forming process is corrected, and an optimal harvesting time window of the quality of each plot is output; generating a global collaborative broadcasting and receiving plan through a multi-objective optimization algorithm by taking the harvesting time window with the optimal quality as input and the daily processing capacity of a processing link as a constraint; the plan is mapped into a resource scheduling scheme, and executable structured operation instructions are generated to drive automated equipment to perform seeding and harvesting operations. Through fusion of filament physiological state perception, multi-target collaborative decision and automatic generation of operation instructions, full-chain intelligent off-peak production of fresh corn from sowing to harvesting is realized.

Description

Technical Field

[0001] This invention relates to the field of smart agriculture technology, specifically to a method and system for intelligent staggered planting and harvesting decisions for fresh corn. Background Technology

[0002] The market value of sweet corn is closely related to its post-harvest freshness, as its optimal edible quality window is short and post-harvest quality tends to decline over time. Achieving a continuous, stable, and high-quality supply of raw materials is a crucial goal for processing enterprises and large-scale planting bases. Current technologies typically plan planting dates based on historical meteorological data and crop growth models, or preliminarily schedule harvesting times based on processing capacity. These methods have provided effective support for identifying basic environmental factors and matching production capacity. However, with the increasing demand for precision agriculture and intelligent decision-making, there is still room for further refinement and improvement in areas such as dynamically responding to meteorological changes during actual growth, global optimization through multi-plot collaboration, and decision adjustments based on real-time feedback.

[0003] Specifically, fresh corn production involves a complex scenario with multiple varieties, multiple plots, and multiple constraints. How to systematically integrate planting, harvesting, resource allocation, and processing capacity to achieve collaborative decision-making across the entire chain from planting to harvesting is one of the directions for the continuous development of related technologies. In this process, dynamic monitoring and prediction of the formation and changes of key quality indicators for fresh corn have also become an important link in improving the collaborative efficiency of the industrial chain. Summary of the Invention

[0004] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide an intelligent staggered planting and harvesting decision-making method and system for fresh corn to solve the above-mentioned technical problems.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an intelligent staggered planting and harvesting decision-making method for fresh corn, comprising:

[0006] S1: Acquire multi-source data of the target area, preprocess the multi-source data, and generate a standardized status code for each plot;

[0007] S2: Based on standardized state coding, the growth quality model of sweet corn is driven to simulate the daily step length. The silk activity is obtained by analyzing the collected ear images, and the silk activity is used to correct the quality formation process, outputting the optimal harvest time window for each plot.

[0008] S3: Using the optimal harvest time window for all plots as input and the daily processing capacity of the processing stage as a constraint, a global collaborative harvesting plan is generated through collaborative calculation using a multi-objective optimization algorithm.

[0009] S4: Map the global collaborative sowing and harvesting plan into a resource scheduling scheme and generate executable structured operation instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

[0010] The present invention is further configured such that S1 includes:

[0011] The multi-source data includes plot boundary data, soil attribute data, meteorological data, crop data, and available resource data;

[0012] The multi-source data is preprocessed, including spatial interpolation of spatial data to unify it to local parcel units, and time alignment and missing value imputation of temporal data to form a complete time series.

[0013] The preprocessed multi-source data is standardized to construct a standardized state code with land parcels as the basic spatial unit and days as the time unit. The standardized state code is used to characterize the comprehensive state of each land parcel on a specific date.

[0014] The present invention is further configured such that S2 includes:

[0015] Based on standardized state coding, the growth quality model of fresh corn is driven to perform daily step-by-step simulation calculations, wherein the simulation calculations include phenological period progression, photosynthetic production simulation, and dry matter accumulation and distribution processes.

[0016] The phenological period is advanced by accumulating the effective accumulated temperature each day and completing the stage transition when the accumulated value reaches the preset threshold for the variety.

[0017] The photosynthetic production simulation is based on the light response relationship and calculates the net photosynthetic rate according to the daily photosynthetically active radiation.

[0018] The process of dry matter accumulation and distribution involves transporting the total amount of photosynthetic products to different organs according to a pre-defined distribution pattern for each stage of growth.

[0019] After the silking stage, the activity of the silks is calculated by collecting images of the ear and extracting the morphological characteristics of the silks. The morphological characteristics of the silks include the silk attachment rate, the silk browning index and the silk elongation uniformity.

[0020] During the operation of the fresh corn growth quality model, the silk activity level is used to dynamically correct the quality formation process within the fresh corn growth quality model. The correction includes a positive correlation between the silk activity level and the correction magnitude of the sugar accumulation rate, and a positive correlation between the silk activity level and the duration of the determined optimal harvest window.

[0021] Based on the corrected quality formation process, the curve of sugar content change over time is predicted, and the time interval in which the sugar content is greater than the preset sugar threshold and the duration reaches the preset duration is determined as the optimal harvest time window.

[0022] The present invention is further configured such that the filament adhesion rate is determined based on the ratio of the number of attached filaments to the total number of filaments;

[0023] The filament browning index is determined based on the ratio of the area of ​​the browned filament region to the total area of ​​the filament region;

[0024] The uniformity of filament elongation is calculated based on the statistical standard deviation and arithmetic mean of the filament length.

[0025] The present invention is further configured such that S3 includes:

[0026] Using the optimal harvest time window and corresponding expected yield of all plots as input, and combined with the daily processing capacity constraint of the processing link, a multi-objective optimization model is established. The expected yield is obtained by simulation calculation of the fresh corn growth quality model in S2.

[0027] The optimization objectives of the multi-objective optimization model include minimizing the distribution variance of the daily processing load during the planning period and maximizing the expected total quality value of all plots.

[0028] The constraints of the multi-objective optimization model include that the harvest date of each plot must fall within its own optimal quality harvest time window.

[0029] An evolutionary algorithm was used to solve the multi-objective optimization model to obtain the Pareto optimal solution set;

[0030] Based on preset decision preferences, the final solution is selected from the Pareto optimal solution set to generate a global collaborative broadcasting plan.

[0031] The present invention is further configured such that S4 includes:

[0032] The global collaborative broadcasting and harvesting plan is mapped to a resource scheduling scheme, which includes an agricultural machinery and equipment scheduling scheme and a human resource allocation scheme.

[0033] When formulating resource allocation plans, the priority of operational tasks in each plot is determined based on the optimal harvesting time window for filament activity and quality.

[0034] Based on the resource scheduling scheme, executable structured operation instructions are generated. These operation instructions include machine-readable control instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

[0035] The present invention is further configured to collect actual operation data during the execution of the structured operation instructions, the actual operation data including the actual sowing date, the actual harvesting date and the actual operation duration;

[0036] The actual operational data collected is compared with the corresponding predicted data in the global collaborative broadcasting and reception plan, and the deviation of each key indicator is calculated.

[0037] If the deviation of any key indicator exceeds the corresponding preset deviation threshold, an abnormal alarm will be automatically triggered.

[0038] The present invention is further configured to generate corresponding natural language agricultural guidance information while generating the structured operation instructions.

[0039] The present invention is further configured such that the method also includes synchronously displaying the global collaborative broadcasting plan and structured operation instructions in a graphical manner in the user interface.

[0040] This invention also provides an intelligent staggered planting and harvesting decision-making system for fresh corn, the system comprising:

[0041] Data governance module: Acquires multi-source data of the target area, preprocesses the multi-source data, and generates a standardized status code for each plot;

[0042] Quality prediction module: Based on standardized state coding, it drives the growth quality model of fresh corn to simulate the daily step length, and obtains the silk activity by analyzing the collected ear images. It then uses the silk activity to correct the quality formation process and outputs the optimal harvest time window for each plot.

[0043] Collaborative decision-making module: Taking the optimal harvest time window for all plots as input and the daily processing capacity of the processing stage as constraints, it performs collaborative calculations through a multi-objective optimization algorithm to generate a global collaborative planting and harvesting plan;

[0044] Job scheduling module: Maps the global collaborative sowing and harvesting plan into a resource scheduling scheme and generates executable structured job instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

[0045] This invention provides an intelligent staggered planting and harvesting decision-making method and system for sweet corn. The method comprises: S1: acquiring multi-source data of the target area and preprocessing the data to generate a standardized state code for each plot; S2: based on the standardized state code, driving a sweet corn growth quality model to simulate daily growth steps, and obtaining silk activity by analyzing collected ear images, using silk activity to correct the quality formation process, and outputting the optimal harvest time window for each plot; S3: using the optimal harvest time window for all plots as input, and the daily processing capacity of the processing stage as a constraint, performing collaborative calculations through a multi-objective optimization algorithm to generate a global collaborative planting and harvesting plan; S4: mapping the global collaborative planting and harvesting plan into a resource scheduling scheme and generating executable structured operation instructions to drive automated agricultural equipment to perform corresponding planting and harvesting operations. The beneficial effects include:

[0046] 1. By introducing image recognition-based silk activity, the sugar accumulation process in the fresh corn growth quality model is physiologically corrected, enabling accurate prediction of the optimal harvest time window and effectively avoiding quality deterioration caused by improper harvesting timing.

[0047] 2. Taking processing capacity as a hard constraint, the planting and harvesting plans of multiple planting plots are collaboratively calculated through a multi-objective optimization algorithm, generating a globally optimal plan that balances processing load and maximizes total quality value, thus solving the processing congestion problem caused by concentrated market launch at the system level.

[0048] 3. The optimized decision-making scheme is automatically mapped into executable structured operation instructions and resource scheduling schemes, and drives automated agricultural equipment to execute them, reducing human intervention and improving the efficiency and accuracy of decision-making and execution.

[0049] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0051] Figure 1 A flowchart illustrating an intelligent staggered planting and harvesting decision-making method for fresh corn, as shown in an exemplary embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram illustrating the structure of an intelligent staggered planting and harvesting decision system for fresh corn, as an exemplary embodiment of the present invention. Detailed Implementation

[0053] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0054] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0055] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0056] Example 1:

[0057] A smart method for staggered planting and harvesting of sweet corn, such as Figure 1 As shown, it includes:

[0058] S1: Acquire multi-source data of the target area, preprocess the multi-source data, and generate a standardized status code for each plot;

[0059] S2: Based on standardized state coding, the growth quality model of sweet corn is driven to simulate the daily step length. The silk activity is obtained by analyzing the collected ear images, and the silk activity is used to correct the quality formation process, outputting the optimal harvest time window for each plot.

[0060] S3: Using the optimal harvest time window for all plots as input and the daily processing capacity of the processing stage as a constraint, a global collaborative harvesting plan is generated through collaborative calculation using a multi-objective optimization algorithm.

[0061] S4: Map the global collaborative sowing and harvesting plan into a resource scheduling scheme and generate executable structured operation instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

[0062] The present invention is further configured such that S1 includes:

[0063] The multi-source data includes plot boundary data, soil attribute data, meteorological data, crop data, and available resource data;

[0064] The multi-source data is preprocessed, including spatial interpolation of spatial data to unify it to local parcel units, and time alignment and missing value imputation of temporal data to form a complete time series.

[0065] The preprocessed multi-source data is standardized to construct a standardized state code with plots as the basic spatial unit and days as the time unit. This standardized state code characterizes the comprehensive state of each plot on a specific date. Specifically, during plot boundary data collection, established technical specifications are followed. High-precision positioning equipment is used to collect coordinate points at fixed intervals along the plot boundary line to ensure the formation of high-precision closed polygons. Soil property data includes soil chemical and physical properties. Soil chemical property data sampling follows a specific sampling pattern, setting a predetermined number of sampling points per unit area, and representative samples are collected from soil layers at different depths. In the laboratory, standard chemical analysis methods were used to determine soil chemical characteristics, including soil organic carbon content, total nitrogen content, total phosphorus content, total potassium content, soil pH, and cation exchange capacity. Soil physical property data were obtained by measuring soil bulk density using the ring sampler method and quantifying soil particle composition using laser particle size analysis to determine the mass percentages of clay, silt, and sand. Meteorological data collection relied on automatic monitoring stations deployed in the fields. Sensors recorded air temperature, relative humidity, wind speed, precipitation, and photosynthetically active radiation data at preset fixed frequencies. The raw high-frequency data was processed and aggregated into daily numerical sequences, thus forming complete time-series data. Crop data was obtained through… This was achieved through a systematic combination of field observation and laboratory testing. Crop data included growth period data, plant characteristic data, ear trait data, quality data, and yield data. Growth period data was recorded through regular field inspections, using the date when 50% of the individuals in the population reached a specific phenological stage as the criterion. The recorded growth stages covered emergence, jointing, tasseling, silking, and harvest. Plant characteristic data were collected in pre-designed representative quadrats, and the measured indicators included plant height, stem diameter, and leaf area index (LAI), with LAI measured using a canopy analyzer. Ear trait data were obtained through sampling at physiological maturity, and the measured items included ear length, ear diameter, tip barrenness length, and number of ear rows. The number of grains per row and quality data are determined during the harvest period. After collecting representative ear samples, the soluble sugar content of the grains is measured using a handheld digital saccharimeter, and the pericarp thickness and peel-poor ratio are measured using a texture analyzer. Yield data is obtained through actual yield measurement, i.e., all ears are harvested within a designated sample plot, the fresh ear weight is weighed, and converted into yield per unit area. Available resource data is integrated and obtained through the farm management system. The available resource data includes the number of available agricultural machines, the number of laborers, and the resource utilization rate. The number of available agricultural machines is defined as the number of agricultural machines in an operational state. The number of laborers refers to the number of personnel available for attendance on a given day, as determined by the shift schedule. The resource utilization rate is defined as the ratio of the actual operating time of agricultural machines to the maximum operational time within a historical period.In the data preprocessing stage, for soil attribute data with spatial variability, spatial interpolation methods from geostatistics are used to transform discrete sampling point data into continuous spatial raster surfaces. Then, the statistical average of the internal raster values ​​is calculated based on the geometric boundaries of each plot, thereby allocating attribute values ​​to each plot unit. For time series data, the data is uniformly converted to a standard time resolution, and time series forecasting or spatial interpolation methods are used to fill in missing data according to the severity of the missing data, ensuring the temporal integrity of the data. After the above preprocessing, various types of data are standardized to eliminate the influence of different units and numerical ranges. Finally, a structured standardized status code is constructed for each plot on each specific date. This standardized status code adopts a common data exchange format and systematically integrates plot identifiers, timestamps, standardized soil attribute vectors, meteorological vectors, crop vectors, and available resource vectors. All standardized status codes are organized chronologically and undergo integrity verification. The soil attribute vector contains standardized soil attribute data, the meteorological vector contains standardized meteorological data, the crop vector contains standardized crop data, and the available resource vector contains standardized available agricultural machinery, labor force, and resource utilization rate. ;

[0066] The present invention is further configured such that S2 includes:

[0067] Based on standardized state coding, a daily-step simulation calculation is performed to drive the growth quality model of sweet corn. This simulation includes phenological progression, photosynthetic production simulation, and dry matter accumulation and distribution. Phenological progression involves accumulating daily effective accumulated temperature and completing a stage transition when the accumulated value reaches a preset threshold for the variety. Photosynthetic production simulation is based on light response relationships, calculating the net photosynthetic rate according to daily effective photosynthetic radiation. The dry matter accumulation and distribution process involves transporting the total amount of photosynthetic products to different organs according to a preset distribution pattern for each growth stage. Specifically, the realization of the phenological progression process relies on the cumulative calculation of heat resources. The sweet corn growth quality model extracts the highest temperature and lowest humidity from the standardized state coding daily. Low-temperature data is used to calculate the daily average temperature and compare it with the biological zero-degree threshold for crop growth and development. When the daily average temperature is greater than the biological zero-degree threshold, the difference is identified as the effective accumulated temperature for that day, which is added to the total effective accumulated temperature accumulated during the current growth period. The model pre-stores the effective accumulated temperature thresholds required for the transition of each key phenological stage for a specific maize variety. After each day's calculation, the system compares the current accumulated effective accumulated temperature with the effective accumulated temperature threshold required for the next phenological stage. When the accumulated value is greater than or equal to the threshold, the model automatically updates the value in the crop growth stage status register, advancing it to the next phenological stage, thus completing the phenological stage transition. Photosynthetic production simulation. The core of the process lies in determining the daily net assimilation produced by canopy photosynthesis. This process is based on the rectangular hyperbolic light response relationship. The fresh corn growth quality model obtains photosynthetically active radiation data from standardized state codes as the main input daily. Based on the incident photosynthetically active radiation intensity and pre-calibrated photosynthetic response characteristic parameters reflecting the photosynthetic capacity of leaves, the potential maximum photosynthetic rate is calculated. This process considers the saturation effect between light intensity and photosynthetic rate, i.e., as light intensity continues to increase, the growth of the photosynthetic rate tends to level off. Subsequently, the amount of respiration necessary for the plant to maintain its life activities is subtracted from the total photosynthetic output to obtain the net photosynthetic yield available for plant growth. The dry matter accumulation and distribution process is responsible for converting the daily net photosynthetic yield into a single unit. The growth quality model of fresh maize transforms into biomass growth in various organs of the plant. It uses the calculated daily net photosynthetic yield as the sole input source of assimilated products. Based on the specific phenological stage of the crop, it invokes a preset distribution law, which is defined by a set of dynamic distribution coefficients. This distribution coefficient set specifies the fixed proportion of total assimilated products transported to different organs such as roots, stems, leaves, and ears at a specific growth stage. Its dynamic characteristics are reflected in the fact that the distribution center tends to be on leaves and roots in the early stage of vegetative growth, while it shifts to ears in the reproductive growth stage. The model queries the corresponding distribution coefficient set according to the current phenological stage every day and distributes the net photosynthetic yield to each organ according to this proportion, thereby updating its biomass. The entire simulation process is iterated in daily increments until the crop reaches physiological maturity.

[0068] After the silking stage, images of the panicle are collected and morphological features of the filaments are extracted to calculate filament activity. These morphological features include filament attachment rate, filament browning index, and filament elongation uniformity. The invention further specifies that the filament attachment rate is determined based on the ratio of the number of attached filaments to the total number of filaments; the filament browning index is determined based on the ratio of the area of ​​the browned filament region to the area of ​​the total filament region; and the filament elongation uniformity is calculated based on the statistical standard deviation and arithmetic mean of the filament length. Specifically, after the crop population in the field enters the silking stage, the criterion is that more than 50% of the plants… With the silks fully emerged, images were collected at several specific time points when key changes in silk vigor occurred, such as midday on the 3rd, 5th, and 7th days after silking. A drone equipped with a high-definition camera was used for image acquisition. The drone hovered at a preset fixed flight altitude to ensure consistent resolution and scale in each image. Within each selected plot, a predetermined number of representative ears of fruit were photographed to obtain a representative sample. The collected image data was processed and analyzed using a dedicated image recognition algorithm. First, the fruit... The number of filaments still attached to the rachis of the panicle was counted, and traces of filaments that had fallen off or withered due to successful pollination were identified. The filament attachment rate was calculated by determining the proportion of filaments still attached to the total number of filaments. Based on a pre-established color model, regions of filaments in a fresh state were distinguished from regions of filaments that had undergone browning and aging. The percentage of the pixel area of ​​browned regions to the total pixel area of ​​filament regions was calculated to obtain the filament browning index. The length values ​​of all visible filaments in the image were measured, and the ratio of the statistical standard deviation to the arithmetic mean of the length values ​​was calculated. This ratio reflects... To assess the uniformity of filament elongation, a ratio of 1 minus this ratio was used as a quantitative indicator of filament elongation uniformity to more intuitively characterize the uniformity of filament elongation. Based on the daily filament attachment rate, filament browning index, and filament elongation uniformity, a daily comprehensive score was calculated. This score comprehensively characterizes the attachment status, aging degree, and elongation consistency of the filaments on that day. The comprehensive scores obtained in each key observation day were arithmetically averaged to obtain a comprehensive value characterizing the overall activity of the filaments in the plot. This comprehensive value was then mapped to the interval between 0 and 1 using a preset S-shaped mathematical function to obtain the filament activity level.

[0069] During the operation of the sweet corn growth quality model, the silk activity level is used to dynamically correct the quality formation process within the model. This correction includes a positive correlation between the silk activity level and the sugar accumulation rate, and a positive correlation between the silk activity level and the duration of the determined optimal harvest window. Based on the corrected quality formation process, the sugar content change curve over time is predicted, and the time interval where the sugar content exceeds a preset sugar threshold and remains above the preset duration is defined as the optimal harvest time window. Specifically, in the dynamic correction of the quality formation process, the calculated silk activity level is integrated into the quality formation sub-model within the sweet corn growth quality model. This correction mechanism specifically manifests as a positive correlation between the sugar accumulation rate and the sugar accumulation rate. Synchronous correction of the optimal harvest window: Silk activity is positively correlated with sugar accumulation rate; higher silk activity reflects better pollination quality and more synchronized and vigorous grain filling process. The model will adjust the sugar accumulation rate parameter upwards according to a preset ratio, and vice versa. Simultaneously, silk activity is positively correlated with harvestable duration; high silk activity indicates good grain development synchronization, and the optimal harvest window is correspondingly widened due to the extended sugar peak plateau period, while low silk activity indicates a narrowing harvest window. The corrected quality formation sub-model is re-run to simulate the dynamic change curve of grain sugar after pollination. The optimal harvest time window for a specific plot is defined by identifying the time interval from when the sugar concentration first exceeds the preset sugar threshold to when it falls below that threshold.

[0070] The present invention is further configured such that S3 includes:

[0071] Using the optimal harvest time window and corresponding expected yield of all plots as input, and combined with the daily processing capacity constraint of the processing link, a multi-objective optimization model is established. The expected yield is obtained by simulation calculation of the fresh corn growth quality model in S2.

[0072] The optimization objectives of the multi-objective optimization model include minimizing the distribution variance of the daily processing load during the planning period and maximizing the expected total quality value of all plots.

[0073] The constraints of the multi-objective optimization model include that the harvest date of each plot must fall within its own optimal quality harvest time window.

[0074] An evolutionary algorithm was used to solve the multi-objective optimization model to obtain the Pareto optimal solution set;

[0075] Based on preset decision preferences, a final solution is selected from the Pareto optimal solution set to generate a global collaborative harvesting plan. Specifically, this embodiment details the optimization process of generating the global collaborative harvesting plan. This process uses the optimal harvesting time window and corresponding expected yield obtained by all plots through the aforementioned steps as core input parameters. At the same time, the daily processing capacity limit of the processing stage is set as a rigid constraint, and a multi-objective optimization model is constructed accordingly. First, resource conflict detection is performed, and the expected total processing load for each day within the planning period is calculated. This expected total processing load is the sum of the expected yields of all plots planned for harvesting on that day. Dates where the expected total processing load exceeds the rated daily processing capacity of the processing plant are identified and marked. Such dates are defined as resource conflict days. The multi-objective optimization model sets two optimization objectives. The primary objective is to minimize the distribution variance of the daily processing load during the planning period, aiming to achieve a uniform distribution of the processing load over time, thereby ensuring the smooth operation of the production line and the efficient utilization of resources. The secondary objective is to maximize the expected total quality value of all plots. This expected total quality value is quantified by summing the products of the expected yield of each plot and its corresponding expected quality value. The expected quality value is calculated from the sugar content predicted by the fresh corn growth quality model using preset rules. This expected quality value is based on the predicted sugar content at a specific time point and is further enhanced by introducing silk activity. The fitness parameters are corrected to obtain the model. The key constraint of the multi-objective optimization model is that the actual harvest date of each plot must be strictly limited to the optimal quality harvest time window determined after model correction, so as to ensure that the product quality meets the preset standards during harvesting. An evolutionary algorithm is used to solve the multi-objective optimization model, specifically including: encoding the entire production plan, i.e., the set of planting dates of all plots, into a complete scheme code, where each code unit corresponds to the planting date of a specific plot; constructing a fitness function based on the weighted sum of the normalized processing variance and the total quality value, which can evaluate the uniformity of processing load and the total quality value, and assign these two sub-objectives to the model. Adjustable weighting coefficients are assigned to reflect different decision preferences. During the algorithm's iterative optimization process, the following processing rules are preset: for plots with low filament activity, their sowing dates are protected during the optimization process to avoid drastic date adjustments to meet the global objective; for plots with extremely low filament activity, they are given the highest optimization priority to ensure that they can be harvested first within the optimal time window; the convergence condition of the evolutionary algorithm is set to the following: the quality improvement of the optimal solution over several consecutive generations is less than a preset threshold, or the total number of iterations reaches the upper limit. After the convergence condition is met, a set of Pareto optimal solutions is output, where each solution represents a feasible sowing and harvesting plan that achieves a balance among multiple objectives.Finally, based on the preset decision preference strategy, a final solution is selected from the solution set to generate a global collaborative planting and harvesting plan. This plan is output in tabular form, specifying the expected planting date, predicted harvest date, expected operation time, and expected quality value for each plot, while ensuring that the maximum daily processing load does not exceed the system's processing capacity limit.

[0076] The present invention is further configured such that S4 includes:

[0077] The global collaborative broadcasting and harvesting plan is mapped to a resource scheduling scheme, which includes an agricultural machinery and equipment scheduling scheme and a human resource allocation scheme.

[0078] When formulating resource allocation plans, the priority of operational tasks in each plot is determined based on the optimal harvesting time window for filament activity and quality.

[0079] According to the resource scheduling plan, executable structured operation instructions are generated. These instructions include machine-readable control commands to drive automated agricultural equipment to perform corresponding sowing and harvesting operations. Specifically, the global collaborative sowing and harvesting plan is parsed to extract the daily sowing and harvesting tasks and their corresponding plot areas. Based on pre-set agricultural machinery operation efficiency parameters, such as the daily area that a seeder can complete and the daily area that a harvester can process, the theoretical number of various types of agricultural machinery required each day is calculated. Subsequently, resource conflict detection and scheduling optimization are performed. The calculated resource requirements are compared with the actual number of agricultural machines available on the farm to identify time periods with resource usage conflicts, i.e., situations where the demand exceeds the available quantity on the same day. For the detected conflicts, a priority-based scheduling algorithm is used to resolve them. This algorithm prioritizes plots with lower filament activity or shorter harvesting time windows for optimal quality when allocating resources, ensuring that these plots sensitive to harvesting timing can obtain resources first. The specific implementation of the scheduling algorithm is as follows: on days with resource conflicts, all tasks to be scheduled that day are sorted according to a predetermined priority rule, in descending order of priority. Resource allocation for tasks; for tasks that cannot be scheduled within the optimal window due to insufficient resources, a local adjustment procedure will be initiated to find other available resources or fine-tune their operation dates within the optimal harvest time window, and reassess the impact on the overall plan; when formulating specific schedules, the following operation rules shall be followed: the continuous operation time of a single agricultural machine shall not exceed the rated working time per day, necessary rest time shall be arranged during continuous operation, and the same agricultural machine shall not be assigned tasks with overlapping time and space on the same day; in addition, the scheduling plan shall also meet the time cost constraint of agricultural machine transfer, that is, when an agricultural machine needs to perform multiple plot tasks continuously on the same day, the transfer time calculated by the geographical distance between the task locations and the transfer speed of the agricultural machine must be included in the operation time calculation; according to the determined resource scheduling plan, control instruction files that can be directly issued to intelligent agricultural equipment such as tractors, seeders, and harvesters shall be generated in accordance with the internationally accepted agricultural machinery bus communication standard. The control instruction files shall adopt structured data formats such as XML or JSON, and include the boundary coordinates of the operation plots and the process parameters such as the preset operation path planning, as well as the start and end timestamps of the tasks.

[0080] The present invention is further configured to collect actual operation data during the execution of the structured operation instructions, the actual operation data including the actual sowing date, the actual harvesting date and the actual operation duration;

[0081] The actual operational data collected is compared with the corresponding predicted data in the global collaborative broadcasting and reception plan, and the deviation of each key indicator is calculated.

[0082] If the deviation of any key indicator exceeds the corresponding preset deviation threshold, an anomaly alarm will be automatically triggered. Specifically, while the automated agricultural equipment executes structured planting and harvesting operation instructions, it will automatically record actual operation data through the equipment's IoT terminal. The main fields collected include: the actual planting date, the actual harvesting date, and the actual operation time consumed to complete the corresponding operation for each plot. The collected actual operation data will be compared item by item with the corresponding predicted data in the global collaborative planting and harvesting plan driving this operation. The comparison items include: the time deviation between the actual planting date and the expected planting date, the time deviation between the actual harvesting date and the predicted harvesting period, and the difference between the actual operation time and the expected operation time. The actual values ​​and predicted values ​​of each key indicator will be calculated. The deviation is measured and compared with the corresponding preset deviation threshold. These thresholds are preset based on the statistical analysis of historical operation data and agronomic requirements. For example, the sowing time deviation threshold is limited to a specific number of days plus or minus the predicted sowing date, the harvesting time deviation threshold is limited to a specific number of days plus or minus the predicted harvesting date, and the operation duration deviation threshold is set to a certain percentage upper limit of the expected operation duration. When the deviation of any key indicator exceeds its corresponding preset deviation threshold, an abnormal alarm is immediately triggered, and the alarm information is pushed to relevant management personnel in real time through monitoring dashboards, mobile application notifications, etc., to ensure that abnormal situations are detected and handled in a timely manner. The alarm information must include a specific description of the abnormality and its corresponding deviation value.

[0083] The present invention is further configured to generate corresponding natural language agricultural guidance information simultaneously with the structured operation instructions. Specifically, key data from the global collaborative planting and harvesting plan is automatically populated into a predefined standardized statement template to generate basic operation instruction text. The key data includes plot number, operation date, and operation type, etc. Then, key decision logic is input as prompt information into the large language model to generate easily understandable explanatory text, providing a detailed explanation of the reasons for the operation arrangement. The decision logic includes, for example, explanations of planting date adjustments made to avoid processing conflicts based on the filament activity and harvesting window conditions of a specific plot.

[0084] The invention is further configured such that the method also includes synchronously displaying the global collaborative planting and harvesting plan and structured operation instructions in a graphical manner on the user interface; specifically, the user interface adopts a multi-view collaborative display method, with the main view being a geographic information map, overlaying the boundaries and current status of all plots, such as pending planting, already planted, pending harvesting, already harvested, and the planned task timeline; through the timeline control, users can browse the distribution of planned operation tasks for any date; the resource scheduling scheme is presented in the form of a Gantt chart, clearly showing the daily task schedule of each agricultural machine and human resources; when the user selects a specific plot or operation task, the sidebar of the interface synchronously displays its detailed forecast data, structured operation instruction parameters, and real-time status.

[0085] Example 2:

[0086] Please see Figure 2 This exemplary intelligent staggered planting and harvesting decision-making system for fresh corn includes:

[0087] Data governance module: Acquires multi-source data of the target area, preprocesses the multi-source data, and generates a standardized status code for each plot;

[0088] Quality prediction module: Based on standardized state coding, it drives the growth quality model of fresh corn to simulate the daily step length, and obtains the silk activity by analyzing the collected ear images. It then uses the silk activity to correct the quality formation process and outputs the optimal harvest time window for each plot.

[0089] Collaborative decision-making module: Taking the optimal harvest time window for all plots as input and the daily processing capacity of the processing stage as constraints, it performs collaborative calculations through a multi-objective optimization algorithm to generate a global collaborative planting and harvesting plan;

[0090] Job scheduling module: Maps the global collaborative sowing and harvesting plan into a resource scheduling scheme and generates executable structured job instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

[0091] It should be noted that the intelligent staggered planting and harvesting decision-making system for fresh corn provided in the above embodiments and the intelligent staggered planting and harvesting decision-making method for fresh corn provided in the above embodiments belong to the same concept. The specific methods of execution of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the intelligent staggered planting and harvesting decision-making system for fresh corn provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the system can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0092] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A smart method for staggered planting and harvesting decision-making for fresh corn, characterized in that, include: S1: Acquire multi-source data of the target area, preprocess the multi-source data, and generate a standardized status code for each plot; S2: Based on standardized state coding, the growth quality model of sweet corn is driven to simulate the daily step length. The silk activity is obtained by analyzing the collected ear images, and the silk activity is used to correct the quality formation process, outputting the optimal harvest time window for each plot. S3: Using the optimal harvest time window for all plots as input and the daily processing capacity of the processing stage as a constraint, a global collaborative harvesting plan is generated through collaborative calculation using a multi-objective optimization algorithm. S4: Map the global collaborative sowing and harvesting plan into a resource scheduling scheme and generate executable structured operation instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

2. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 1, characterized in that, S1 includes: The multi-source data includes plot boundary data, soil attribute data, meteorological data, crop data, and available resource data; The multi-source data is preprocessed, including spatial interpolation of spatial data to unify it to local parcel units, and time alignment and missing value imputation of temporal data to form a complete time series. The preprocessed multi-source data is standardized to construct a standardized state code with land parcels as the basic spatial unit and days as the time unit. The standardized state code is used to characterize the comprehensive state of each land parcel on a specific date.

3. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 1, characterized in that, S2 includes: Based on standardized state coding, the growth quality model of fresh corn is driven to perform daily step-by-step simulation calculations, wherein the simulation calculations include phenological period progression, photosynthetic production simulation, and dry matter accumulation and distribution processes. The phenological period is advanced by accumulating the effective accumulated temperature each day and completing the stage transition when the accumulated value reaches the preset threshold for the variety. The photosynthetic production simulation is based on the light response relationship and calculates the net photosynthetic rate according to the daily photosynthetically active radiation. The process of dry matter accumulation and distribution involves transporting the total amount of photosynthetic products to different organs according to a pre-defined distribution pattern for each stage of growth. After the silking stage, the activity of the silks is calculated by collecting images of the ear and extracting the morphological characteristics of the silks. The morphological characteristics of the silks include the silk attachment rate, the silk browning index and the silk elongation uniformity. During the operation of the fresh corn growth quality model, the silk activity level is used to dynamically correct the quality formation process within the fresh corn growth quality model. The correction includes a positive correlation between the silk activity level and the correction magnitude of the sugar accumulation rate, and a positive correlation between the silk activity level and the duration of the determined optimal harvest window. Based on the corrected quality formation process, the curve of sugar content change over time is predicted, and the time interval in which the sugar content is greater than the preset sugar threshold and the duration reaches the preset duration is determined as the optimal harvest time window.

4. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 3, characterized in that, The filament adhesion rate is determined based on the ratio of the number of attached filaments to the total number of filaments; The filament browning index is determined based on the ratio of the area of ​​the browned filament region to the total area of ​​the filament region; The uniformity of filament elongation is calculated based on the statistical standard deviation and arithmetic mean of the filament length.

5. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 1, characterized in that, S3 includes: Using the optimal harvest time window and corresponding expected yield of all plots as input, and combined with the daily processing capacity constraint of the processing link, a multi-objective optimization model is established. The expected yield is obtained by simulation calculation of the fresh corn growth quality model in S2. The optimization objectives of the multi-objective optimization model include minimizing the distribution variance of the daily processing load during the planning period and maximizing the expected total quality value of all plots. The constraints of the multi-objective optimization model include that the harvest date of each plot must fall within its own optimal quality harvest time window. An evolutionary algorithm was used to solve the multi-objective optimization model to obtain the Pareto optimal solution set; Based on preset decision preferences, the final solution is selected from the Pareto optimal solution set to generate a global collaborative broadcasting plan.

6. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 1, characterized in that, S4 includes: The global collaborative broadcasting and harvesting plan is mapped to a resource scheduling scheme, which includes an agricultural machinery and equipment scheduling scheme and a human resource allocation scheme. When formulating resource allocation plans, the priority of operational tasks in each plot is determined based on the optimal harvesting time window for filament activity and quality. Based on the resource scheduling scheme, executable structured operation instructions are generated. These operation instructions include machine-readable control instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

7. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 6, characterized in that, During the execution of the structured operation instructions, actual operation data is collected, including the actual sowing date, the actual harvesting date, and the actual operation duration. The actual operational data collected is compared with the corresponding predicted data in the global collaborative broadcasting and reception plan, and the deviation of each key indicator is calculated. If the deviation of any key indicator exceeds the corresponding preset deviation threshold, an abnormal alarm will be automatically triggered.

8. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 6, characterized in that, While generating the structured operation instructions, corresponding natural language agricultural guidance information is also generated.

9. The intelligent staggered planting and harvesting decision-making method for fresh corn according to claim 1, characterized in that, The method also includes synchronously displaying the global collaborative broadcasting plan and structured job instructions in a graphical manner in the user interface.

10. An intelligent staggered planting and harvesting decision-making system for fresh sweet corn, used to implement the intelligent staggered planting and harvesting decision-making method for fresh sweet corn as described in any one of claims 1-9, characterized in that, include: Data governance module: Acquires multi-source data of the target area, preprocesses the multi-source data, and generates a standardized status code for each plot; Quality prediction module: Based on standardized state coding, it drives the growth quality model of fresh corn to simulate the daily step length, and obtains the silk activity by analyzing the collected ear images. It then uses the silk activity to correct the quality formation process and outputs the optimal harvest time window for each plot. Collaborative decision-making module: Taking the optimal harvest time window for all plots as input and the daily processing capacity of the processing stage as constraints, it performs collaborative calculations through a multi-objective optimization algorithm to generate a global collaborative planting and harvesting plan; Job scheduling module: Maps the global collaborative sowing and harvesting plan into a resource scheduling scheme and generates executable structured job instructions to drive automated agricultural equipment to perform corresponding sowing and harvesting operations.

Images