Agricultural planting operation and maintenance management system and method based on cloud platform

Abstract

The invention discloses an agricultural planting operation and maintenance management system and method based on a cloud platform, and relates to the technical field of intelligent operation and maintenance management, and the method comprises the steps: obtaining agricultural planting operation and maintenance data and a farming task set, and carrying out the preprocessing of the agricultural planting operation and maintenance data, and obtaining a standardized agricultural operation and maintenance data set; based on the standardized agricultural operation and maintenance data set, constructing a priority relation matrix between farming tasks, and performing sequence optimization on the farming task set by adopting an improved genetic algorithm to obtain an optimal operation and maintenance operation sequence; determining an operation and maintenance parameter search space of each farming task based on the optimal operation and maintenance operation sequence, and performing configuration optimization on operation and maintenance parameters by adopting an improved Bayesian algorithm to obtain an optimal operation and maintenance parameter configuration result; and performing fusion processing on the optimal operation and maintenance operation sequence and the optimal operation and maintenance parameter configuration result to obtain an agricultural planting operation and maintenance management scheme, and issuing the agricultural planting operation and maintenance management scheme to an execution terminal. According to the method, multi-task cooperative scheduling and dynamic parameter optimization are realized, and the agricultural operation efficiency, the resource utilization rate and the system adaptability are improved.

Description

Technical Field

[0001] This invention relates to the technical field of intelligent operation and maintenance management, and in particular to an agricultural planting operation and maintenance management system and method based on a cloud platform. Background Technology

[0002] In recent years, with the continuous development of agricultural informatization and cloud computing technologies, agricultural planting and operation management has gradually shifted from manual experience-based management to data-driven management. This involves collecting basic data such as temperature, humidity, and soil moisture through environmental monitoring equipment, and scheduling and controlling agricultural operations such as sowing, irrigation, and fertilization. However, existing technologies mostly manage individual agricultural operations independently, lacking a systematic mechanism for handling the collaborative relationships between multiple tasks and the impact of environmental data changes on scheduling.

[0003] Currently, Chinese invention patent CN116050861A discloses an agricultural IoT strawberry industry service operation and maintenance system and method. This method involves nano-grinding strawberry fertilizer and then pressurizing it into an ultra-micro nano-permeable containment structure to form ultra-micro permeable fertilizer nanoparticles. These nanoparticles are then delivered to the concentrated root areas of the strawberry plants via a water-fertilizer integrated machine. Large-area in-situ detection and ultrasonic detection are used to monitor the local root distribution in the root and stem growth area, tracking the concentrated root areas. An industry analysis and prediction IoT platform is used to predict and analyze strawberry growth fertilizer application and strawberry yield industry trends. Based on these predictions, intelligent management of strawberry industry fertilizer application and remote intelligent diagnosis and maintenance of strawberry equipment are implemented, accelerating the improvement of agricultural IoT strawberry industry service operation and maintenance. However, this related technology lacks sensitivity to the same data for different tasks, fails to consider the chain reaction of scheduling effects caused by data changes, and does not fully consider weather changes, soil condition fluctuations, temporary resource constraints, and changes in agricultural machinery availability. Summary of the Invention

[0004] The technical problem solved by this invention is that existing technologies, while convenient for monitoring and controlling independent irrigation, sowing, and fertilization in agricultural planting and operation management, lack sensitivity to the same data across different tasks, fail to consider the chain reaction of scheduling effects caused by data changes, and struggle to characterize the comprehensive impact of environmental data changes on multiple related tasks. Furthermore, they fail to fully consider the need for task execution order adjustments and parameter adaptation due to the time-varying nature of environmental monitoring data, and do not adequately account for weather changes, soil condition fluctuations, temporary resource constraints, and changes in agricultural machinery availability.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: Firstly, a cloud platform-based method for agricultural planting operation and maintenance management, comprising the following steps: Step S1: Obtain agricultural planting operation and maintenance data and agricultural task set, and preprocess the agricultural planting operation and maintenance data to obtain a standardized agricultural operation and maintenance dataset; Step S2: Construct a priority relationship matrix among agricultural tasks based on the standardized agricultural operation and maintenance dataset, and use an improved genetic algorithm to optimize the sequence of agricultural tasks to obtain the optimal operation and maintenance sequence; Step S3: Based on the optimal operation and maintenance operation sequence, determine the search space of operation and maintenance parameters for each agricultural task, and use an improved Bayesian algorithm to optimize the configuration of operation and maintenance parameters to obtain the optimal operation and maintenance parameter configuration result; Step S4: The optimal operation and maintenance operation sequence and the optimal operation and maintenance parameter configuration result are merged and processed to obtain the agricultural planting operation and maintenance management plan and then distributed to the execution terminal.

[0006] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method of the present invention, the agricultural planting operation and maintenance data includes environmental monitoring data, crop growth data, operation execution data, and resource consumption data. The environmental monitoring data includes temperature data, humidity data, light intensity data, and soil moisture data; The crop growth data includes crop growth period information, growth status indicators, and yield-related data. The yield-related data includes crop yield per unit area, crop growth stage yield indicators, and historical yield statistics. The operation execution data includes sowing operation data, fertilization operation data, irrigation operation data, and pest and disease control operation data; The resource consumption data includes water consumption data, fertilizer consumption data, energy consumption data, and operating cost data; The agricultural task set includes sowing tasks, fertilization tasks, irrigation tasks, pest and disease control tasks, and the attributes corresponding to each agricultural task. The attributes corresponding to each agricultural task include task priority attribute, task execution time limit, and logical constraints on the order of agricultural tasks.

[0007] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method described in this invention, the processing logic for constructing a priority relationship matrix among agricultural tasks based on a standardized agricultural operation and maintenance dataset includes: Extract the sequential logical constraints between agricultural tasks in the standardized agricultural operation and maintenance data. The sequential logical constraints between agricultural tasks include the irreversible constraint of crop physiological growth, the conflict constraint of agricultural machinery and equipment allocation, and the timing constraint of water, fertilizer and pesticide application. Build Priority relation matrix of dimension ; in, This indicates the total number of agricultural tasks. Represents the first in the priority relation matrix Line 1 Column matrix elements; If agricultural tasks Must be in agricultural tasks Execute after completion, then... Otherwise ; The Kahn algorithm is used to perform topological sorting on the priority relation matrix to obtain the initial sequence. The in-degree of each agricultural task is counted. Agricultural tasks with an in-degree of 0 are selected in a loop and added to the initial sequence while the remaining in-degrees are updated synchronously until an initial task sequence individual that meets the priority logical constraints between agricultural tasks is generated.

[0008] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method described in this invention, an improved genetic algorithm is used to optimize the agricultural task set to obtain the optimal operation and maintenance operation sequence. The processing logic includes: agricultural task set Each agricultural task in the process is mapped to a unique integer gene bit, resulting in a sequence containing... The arrangement of 1 gene loci serves as the initial chromosome population; Each arrangement represents a single task sequence. Crossover and mutation operations are performed on the initial chromosome population to obtain temporary candidate sequence individuals. ; The sequence repair operator is invoked to perform logical validity checks and corrections on temporary candidate sequence individuals, and matrix elements in the priority relation matrix are retrieved. If any two gene loci combinations are found in an individual with a temporary candidate sequence satisfy and ; in, And corresponding matrix elements If so, the individual is deemed to have violated logical constraints; The hierarchical topological order of agricultural tasks is determined based on the priority relation matrix. Gene positions that are determined to violate logical constraints are forcibly rearranged. By adjusting the arrangement of agricultural tasks, the repaired temporary candidate sequence individuals are made to satisfy all... The corresponding agricultural tasks are sequentially constrained to generate a valid sequence of operations; A fitness function is constructed with the objectives of maximizing the total duration of agricultural operations, the total cost of operation and maintenance, and the utilization rate of agricultural machinery resources. The fitness function is then used to select the legal operation sequence to obtain the optimal operation and maintenance sequence.

[0009] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method of the present invention, the processing logic for selecting legal job sequences using a fitness function includes: For each valid job sequence, calculate its fitness value. The formula for the fitness function is: ; in, Indicates the fitness value. Indicates the total duration of agricultural operations. This represents the total cost of operation and maintenance. Indicates the utilization rate of agricultural machinery resources. This represents the weighting coefficient for the total duration of agricultural operations. This represents the weighting coefficient of total operation and maintenance costs. This represents the weighting coefficient for agricultural machinery resource utilization. Based on the fitness function value, the legal job sequences are filtered, and the legal job sequence with the best fitness value is selected as the current optimal sequence, and the current optimal sequence is used as the basis for the next generation of job sequences. By performing selection, crossover, and mutation operations, new candidate job sequences are generated based on the current optimal job sequence, and the selection, crossover, and mutation operations are repeated until the preset number of iterations is reached.

[0010] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method of the present invention, the processing logic for determining the search space of operation and maintenance parameters for each agricultural task based on the optimal operation and maintenance operation sequence includes: Extract historical agronomic parameters from the standardized agricultural operation and maintenance dataset that correspond to the agricultural task set, and identify the process baseline value of each agricultural task under the preset yield target. The process reference values ​​include sowing depth, fertilizer nutrient ratio, irrigation duration, and pesticide ratio; Based on the execution time span of each agricultural task in the optimal operation and maintenance sequence, the corresponding environmental monitoring data and crop growth data are matched, and the range of physiological demand fluctuations of each agricultural task in the current growth period is calculated through crop growth status indicators. The initial value range of each operation and maintenance parameter is determined by taking the process baseline value as the central reference point and combining it with the range of physiological demand fluctuations. The system retrieves the operating cost threshold and total resource supply from the resource consumption data, performs boundary constraint pruning on the initial value range, removes parameters that exceed the cost budget for resource reserve capacity, and generates a search space for operation and maintenance parameters corresponding to each agricultural task.

[0011] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method described in this invention, an improved Bayesian algorithm is used to optimize the configuration of operation and maintenance parameters to obtain the optimal configuration result. The processing logic includes: Step S31: Generate an objective function based on the standardized agricultural operation and maintenance dataset. The calculation formula is as follows: ; in, Represents the objective function value. This indicates the total number of agricultural tasks. Indicates the task index. This represents the operational parameters to be evaluated. Indicates the first Resource costs of agricultural tasks Indicates the first The projected yield and revenue of the agricultural task. Indicates the first Risk loss coefficient of agricultural tasks This represents the first weighting coefficient. This represents the second weighting coefficient. Indicates the third weighting coefficient; Step S32: In each round of configuration updates, the potential of untested operational parameters is evaluated using the expected improvement acquisition function, calculated as follows: ; in, This indicates the potential value obtained by improving the acquisition function. This represents the currently known optimal objective function value. This represents the predicted value of the operation and maintenance parameter to be evaluated by the mapping function. This represents the expected value operation; Step S33: Introduce an adaptive weight adjustment mechanism to calculate the internal parameters of the probabilistic correlation in real time, and correct the distribution characteristics of the probabilistic correlation through the objective function value. The correction formula is as follows: ; in, This indicates the corrected length scaling parameter. This indicates the length scaling parameters before correction. Represents an exponential function. Indicates the learning rate step size. Represents the gradient operator. Represents the logarithmic marginal likelihood function. Represents the probability density function. This represents the input sample set of operation and maintenance parameters. This represents the set of output values ​​of the objective function. Indicates the length scaling parameter; Step S34: In each round of optimization, execute the random selection logic. By generating a random number and comparing it with the preset exploration probability, if the random number is less than the preset exploration probability, random sampling is performed in the operation and maintenance parameter search space. Otherwise, the operation and maintenance parameter with the highest calculated value is selected based on the expected improvement collection function for the next round of iteration. Step S35: Repeat steps S32 to S34 until the preset maximum number of iterations is met, and obtain the optimal operation and maintenance parameter configuration result; The optimal operation and maintenance parameter configuration results include the optimal sowing depth, optimal fertilizer nutrient ratio, optimal irrigation duration, optimal pesticide application ratio, optimal agricultural machinery driving speed when performing agricultural tasks, and the start timestamp of each agricultural task plan corresponding to the optimal operation and maintenance sequence.

[0012] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method of the present invention, the optimal operation and maintenance operation sequence and the optimal operation and maintenance parameter configuration result are fused and processed to obtain the agricultural planting operation and maintenance management plan, which is then distributed to the execution terminal. The processing logic includes: Based on the agricultural task index in the optimal operation and maintenance sequence, retrieve the corresponding agricultural pest task operation and maintenance parameters in the optimal operation and maintenance parameter configuration results, and construct an agricultural planting operation and maintenance management plan table. Obtain real-time environmental monitoring data and match and verify the preset operation and maintenance parameters in the agricultural planting operation and maintenance management plan table with the environmental early warning threshold at the current moment; If the environmental monitoring data exceeds the preset safe operating range, the timestamp information in the operation and maintenance management plan table is adjusted according to the priority relationship matrix, and the adjusted operation and maintenance management plan table is converted into a structured control instruction set for execution terminal identification. The structured control instruction set includes the coordinates of the agricultural machinery automatic navigation operation trajectory, the pulse width modulation control parameters of the precision fertilizer applicator, and the opening and closing duration instructions of the irrigation solenoid valve. The structured control instruction set is transmitted to the corresponding execution terminal through the cloud platform interface; Collect feedback data from the execution terminal during the operation and store the feedback data in a standardized agricultural operation and maintenance dataset; The feedback data includes job completion status, execution time, resource consumption, and environmental monitoring data.

[0013] As a preferred embodiment of the cloud-based agricultural planting operation and maintenance management method of the present invention, the processing logic for converting the adjusted operation and maintenance management scheme table into a structured control instruction set for execution terminal recognition includes: Based on the agricultural task index in the adjusted operation and maintenance management plan table, retrieve the optimal operation and maintenance parameter configuration results corresponding to each agricultural task; Based on the execution sequence and operation time boundaries of agricultural tasks, the operation and maintenance parameters of each agricultural task are converted into a control instruction format that can be recognized by the execution terminal. The coordinates of the automatic navigation operation trajectory of agricultural machinery, the pulse width modulation control parameters of precision fertilizer applicators, and the opening and closing duration instructions of irrigation solenoid valves are converted into standardized control instructions according to a preset structured data format. Perform compliance checks on standardized control commands to determine whether the control parameters meet the operating specifications of the execution terminal and the requirements of agricultural tasks.

[0014] Secondly, the cloud-based agricultural planting operation and maintenance management system includes: a data acquisition module, a matrix construction module, a sequence optimization module, and an operation and maintenance management module. The data acquisition module is used to acquire agricultural planting and operation data and agricultural task sets, and to preprocess the agricultural planting and operation data to obtain a standardized agricultural operation and operation dataset. The matrix construction module is used to construct a priority relationship matrix among agricultural tasks based on a standardized agricultural operation and maintenance dataset, and to use an improved genetic algorithm to optimize the sequence of agricultural tasks to obtain the optimal operation and maintenance sequence. The sequence optimization module is used to determine the search space of operation and maintenance parameters for each agricultural task based on the optimal operation and maintenance operation sequence, and to use an improved Bayesian algorithm to optimize the configuration of operation and maintenance parameters to obtain the optimal operation and maintenance parameter configuration result. The operation and maintenance management module is used to integrate the optimal operation and maintenance operation sequence with the optimal operation and maintenance parameter configuration result to obtain the agricultural planting operation and maintenance management plan and distribute it to the execution terminal.

[0015] The beneficial effects of this invention are as follows: This invention innovatively constructs a priority relation matrix that includes the irreversibility of crop physiological growth, conflicts in agricultural machinery allocation, and time constraints on water, fertilizer, and pesticide application. It utilizes a sequence repair operator to forcibly rearrange gene positions that violate logical constraints. An adaptive weight adjustment mechanism based on the log-likelihood function gradient and a random selection exploration logic are introduced into the Bayesian algorithm, achieving precise positioning of high-dimensional agricultural parameters. The combination of the priority relation matrix and the sequence repair operator effectively avoids the risks of agricultural delays and machinery idleness. By using a fitness function and an improved Bayesian algorithm for optimization, the invention maximizes the approximation of the preset yield target while reducing the cost of resources such as fertilizer and water, achieving an optimal balance between resource input and output. Through spatial pruning strategies and adaptive weight adjustment mechanisms, the optimization accuracy and convergence speed of the algorithm in environments with few samples are significantly enhanced, solving the problem of difficulty in obtaining agricultural operation and maintenance parameters. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the basic process of an agricultural planting operation and maintenance management method based on a cloud platform, provided as an embodiment of the present invention. Detailed Implementation

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0018] Example, refer to Figure 1 As an embodiment of the present invention, an agricultural planting operation and maintenance management based on a cloud platform is provided, including the following steps: Step S1: Obtain agricultural planting operation and maintenance data and agricultural task set, and preprocess the agricultural planting operation and maintenance data to obtain a standardized agricultural operation and maintenance dataset; Step S2: Construct a priority relationship matrix among agricultural tasks based on the standardized agricultural operation and maintenance dataset, and use an improved genetic algorithm to optimize the sequence of agricultural tasks to obtain the optimal operation and maintenance sequence; Step S3: Based on the optimal operation and maintenance operation sequence, determine the search space of operation and maintenance parameters for each agricultural task, and use an improved Bayesian algorithm to optimize the configuration of operation and maintenance parameters to obtain the optimal operation and maintenance parameter configuration result; Step S4: The optimal operation and maintenance operation sequence and the optimal operation and maintenance parameter configuration result are merged and processed to obtain the agricultural planting operation and maintenance management plan and then distributed to the execution terminal.

[0019] Specifically, in step S1, the agricultural task set includes pre- and post-concurrent logical constraints between agricultural tasks during its construction. These constraints describe the technological sequence and dependencies of different agricultural tasks in the agricultural production process. During the standardization of agricultural operation and maintenance data, these pre- and post-concurrent logical constraints are recorded in a structured form within the standardized agricultural operation and maintenance dataset. In step S3, the extraction of pre- and post-concurrent logical constraints between agricultural tasks involves reading and calling existing constraint information from the standardized agricultural operation and maintenance dataset, rather than regenerating and deriving these constraints. Furthermore, in step S4, the processing of the optimal operation and maintenance operation sequence and the optimal operation and maintenance parameter configuration results is only used to correlate operation and maintenance parameters with the operation sequence to generate an executable agricultural planting operation and maintenance management plan; it does not involve further optimization or recalculation of the operation sequence and operation and maintenance parameters.

[0020] In practice, agricultural planting and operation data includes environmental monitoring data, crop growth data, operation execution data, and resource consumption data. Environmental monitoring data includes temperature data, humidity data, light intensity data, and soil moisture data; Crop growth data includes information on crop growth stage, growth status indicators, and yield-related data; Yield-related data include crop yield per unit area, crop growth stage yield indicators, and historical yield statistics. Operational data includes sowing data, fertilization data, irrigation data, and pest and disease control data. Resource consumption data includes water consumption data, fertilizer consumption data, energy consumption data, and operating cost data; The agricultural task set includes sowing tasks, fertilization tasks, irrigation tasks, pest and disease control tasks, and the corresponding attributes for each agricultural task; The attributes corresponding to each agricultural task include task priority, task execution time limit, and logical constraints on the order of agricultural tasks.

[0021] In specific implementation, the processing logic for constructing a priority relationship matrix among agricultural tasks based on standardized agricultural operation and maintenance datasets includes: Extract the sequential logical constraints between agricultural tasks in the standardized agricultural operation and maintenance data. The sequential logical constraints between agricultural tasks include the irreversible constraint of crop physiological growth, the conflict constraint of agricultural machinery and equipment allocation, and the timing constraint of water, fertilizer and pesticide application. Build Priority relation matrix of dimension ; in, This indicates the total number of agricultural tasks. Represents the first in the priority relation matrix Line 1 Column matrix elements; If agricultural tasks Must be in agricultural tasks Execute after completion, then... Otherwise ; The Kahn algorithm is used to perform topological sorting on the priority relation matrix to obtain the initial sequence. The in-degree of each agricultural task is counted. Agricultural tasks with an in-degree of 0 are selected in a loop and added to the initial sequence while the remaining in-degrees are updated synchronously until an initial task sequence individual that meets the priority logical constraints between agricultural tasks is generated.

[0022] Specifically, the irreversible constraints of crop physiological growth stem from the inherent laws of the plant life cycle, constituting an inviolable rigid temporal sequence. For example, in a winter wheat field, deep plowing must precede sowing, as unprepared soil cannot be effectively sown; similarly, sowing must precede all post-emergence field management tasks, determined by the biological sequence of seed germination and plant establishment. The constraints of conflict in agricultural machinery allocation reflect the competition for limited resources in the physical world. For instance, if a farm has only one high-performance sprayer, its dual tasks of integrated pest management and foliar fertilizer application cannot be performed simultaneously; the order of these tasks must be clearly defined, constituting a scheduling constraint. The timing constraints of water, fertilizer, and pesticide application involve best practices for agronomic optimization and risk avoidance. For example, to avoid pesticide washout, irrigation should only be carried out at least 24 hours after herbicide application; and to improve fertilizer utilization, irrigation should ideally follow the application of granular fertilizer to promote dissolution and absorption into the soil.

[0023] Encode these constraints into a Priority relation matrix of dimension Each element in the matrix It is a binary indicator: agricultural tasks can be determined based on either of the above types of constraints. Must be in agricultural tasks Execute after completion, then... If there is no such mandatory " Prior to "Relationship, then..." .

[0024] The Kahn algorithm is used to topologically sort the priority relation matrix, generating multiple initial feasible task sequences that satisfy all constraints. The algorithm starts by calculating the in-degree of each agricultural task, i.e., the number of "1"s in each column of the priority relation matrix, representing the number of prerequisite tasks that must be completed beforehand. The algorithm initializes an empty sequence and repeatedly performs the following core operations: identifying all tasks with an in-degree of 0 (i.e., tasks without any incomplete prerequisite tasks) and adding them to the sequence; logically "removing" these tasks from the sequence, i.e., updating the matrix state by decrementing the in-degree of all tasks with these tasks as prerequisites by 1. This process is repeated cyclically, like untangling a complex ball of yarn, removing the currently unconstrained task at the beginning and updating the dependencies between the remaining tasks, until all tasks are added to the final sequence in order. The resulting initial sequence ensures feasibility in terms of agronomic logic, equipment resources, and operational safety, providing a compliant and high-quality search starting point for intelligent optimization algorithms.

[0025] This application establishes a solid and reliable logical foundation for intelligent scheduling of agricultural planting by constructing a priority relation matrix and using a topological sorting algorithm to automatically generate an initial task sequence that meets all constraints. By providing a structured feasible solution space, it significantly improves the search efficiency and convergence speed of the improved genetic algorithm and avoids wasting resources on a large number of invalid solutions.

[0026] In specific implementation, an improved genetic algorithm is used to optimize the agricultural task set to obtain the optimal operation and maintenance sequence. The processing logic includes: agricultural task set Each agricultural task in the process is mapped to a unique integer gene bit, resulting in a sequence containing... The arrangement of 1 gene loci serves as the initial chromosome population; Each arrangement represents a single task sequence. Crossover and mutation operations are performed on the initial chromosome population to obtain temporary candidate sequence individuals. ; The sequence repair operator is invoked to perform logical validity checks and corrections on temporary candidate sequence individuals, and matrix elements in the priority relation matrix are retrieved. If any two gene loci combinations are found in an individual with a temporary candidate sequence satisfy and ; in, And corresponding matrix elements If so, the individual is deemed to have violated logical constraints; The hierarchical topological order of agricultural tasks is determined based on the priority relation matrix. Gene positions that are determined to violate logical constraints are forcibly rearranged. By adjusting the arrangement of agricultural tasks, the repaired temporary candidate sequence individuals are made to satisfy all... The corresponding agricultural tasks are sequentially constrained to generate a valid sequence of operations; A fitness function is constructed with the objectives of maximizing the total duration of agricultural operations, the total cost of operation and maintenance, and the utilization rate of agricultural machinery resources. The fitness function is then used to select the legal operation sequence to obtain the optimal operation and maintenance sequence.

[0027] Specifically, each agricultural task in the agricultural task set is mapped to a unique integer gene bit. For example, deep plowing (T1) is mapped to integer 1, basal fertilization (T2) to integer 2, sowing (T3) to integer 3, pre-winter irrigation (T4) to integer 4, topdressing during the greening period (T5) to integer 5, and harvesting (T6) to integer 6. A complete sequence of operations, such as [1,2,3,4,5,6], corresponds to the entire operational process from deep plowing to harvesting. Each integer sequence constitutes a chromosome individual, and multiple such individuals form the initial population. The initial population can be randomly generated and permuted, and feasible sequences generated by the Kahn algorithm are used as high-quality seeds to ensure population quality.

[0028] Crossover and mutation operations on the initial chromosome population are the core mechanisms of genetic algorithms for generating new individuals. Crossover uses a sequential crossover method, selecting a segment of gene sequence from each of the two parent individuals and generating new offspring sequences through exchange and adjustment. For example, if parent A is [1,2,3,4,5,6] and parent B is [2,1,4,3,5,6], exchanging the middle segment may produce a new sequence like [1,4,3,2,5,6]. Mutation operations may randomly swap the tasks at two positions in the sequence. For example, randomly swapping the tasks at positions 2 and 4 in the sequence [1,2,3,4,5,6] results in [1,4,3,2,5,6].

[0029] Calling the sequence repair operator to perform logical validity checks and corrections on temporary candidate sequences is a key innovation in improving genetic algorithms. The repair operator first needs to check whether the sequence violates the constraints defined by the priority relation matrix. The detection method involves iterating through any combination of two gene positions in the sequence. ,in Its position in the sequence is earlier than Query priority relation matrix Is the value 1? If it is 1, it means... Should be in Previously executed, but in the current sequence But in Previously, this constituted a constraint violation. For example, in the sequence [2,1,3,4,5,6], if the precedence relation matrix stipulates that T1 must precede T2 ( However, since T2 precedes T1 in the sequence, a violation was detected.

[0030] The hierarchical topological order of agricultural tasks is determined based on the priority relation matrix. Gene positions deemed to violate logical constraints are forcibly rearranged. Following a repair strategy based on topological sorting, agricultural tasks are inserted into the new sequence sequentially according to their original order among temporary candidate sequence individuals. However, each insertion ensures that all predecessor tasks of that agricultural task are added to the new sequence based on the priority relation matrix B, including all tasks completed before it. If a predecessor task of an agricultural task has not yet been inserted, that task is temporarily skipped and placed in a waiting queue, while subsequent agricultural tasks continue to be processed. After one round of processing, the agricultural tasks in the waiting queue are re-checked, and this process is repeated until all agricultural tasks are inserted into the new sequence. For example, for the constraint-violation sequence [2,1,4,3,5,6], the repair operator rearranges the sequence according to topological order: first inserting T1 (without a predecessor), then inserting T2 (with a satisfied predecessor), followed by T3, then T4 (dependent on T3), and finally T6, resulting in the valid sequence [1,2,3,4,5,6]. This repair mechanism ensures that the algorithm always searches within the feasible solution region when exploring a wide solution space, greatly improving optimization efficiency.

[0031] By combining these mechanisms, the improved genetic algorithm can systematically search for the optimal job sequence with the shortest total job time, lowest maintenance cost, and highest utilization rate of agricultural machinery resources, while satisfying complex agronomic constraints, thus providing a scientific and efficient scheduling solution for agricultural planting.

[0032] By introducing a constraint guarantee mechanism combining sequence repair operators and topological sorting, this approach fully leverages the global search capabilities of genetic algorithms while fundamentally solving the core problems of traditional optimization methods, such as the tendency to generate illegal solutions and low search efficiency when dealing with complex agricultural constraints. It maps agricultural tasks to genes for evolutionary operations, inheriting the advantages of genetic algorithms in efficient exploration within large-scale solution spaces. The key improvement lies in real-time logical validity checks and mandatory corrections based on priority relation matrices for temporary candidate sequences generated after crossover and mutation, ensuring that all individuals in each generation of the evolutionary population strictly comply with hard constraints such as crop physiology, equipment allocation, and agronomic safety. This bold exploration and real-time repair strategy allows the algorithm to search freely in a vast space of permutations and combinations without worrying about disrupting basic agronomic logic, thus concentrating computational resources entirely on optimizing economic objectives such as operation time, cost, and resource utilization. The system can stably and automatically generate agricultural operation sequences that are absolutely feasible in terms of business and close to the optimal in terms of economics. It achieves a highly efficient unity between rigid obedience to agronomic rules and flexible optimization of operational benefits, providing an intelligent solution with both innovation and extremely high practical value for the refined management of agricultural production.

[0033] In specific implementation, the processing logic for selecting a valid job sequence using the fitness function includes: For each valid job sequence, calculate its fitness value. The formula for the fitness function is: ; in, Indicates the fitness value. Indicates the total duration of agricultural operations. This represents the total cost of operation and maintenance. Indicates the utilization rate of agricultural machinery resources. This represents the weighting coefficient for the total duration of agricultural operations. This represents the weighting coefficient of total operation and maintenance costs. This represents the weighting coefficient for agricultural machinery resource utilization. Based on the fitness function value, the legal job sequences are filtered, and the legal job sequence with the best fitness value is selected as the current optimal sequence, and the current optimal sequence is used as the basis for the next generation of job sequences. By performing selection, crossover, and mutation operations, new candidate job sequences are generated based on the current optimal job sequence, and the selection, crossover, and mutation operations are repeated until the preset number of iterations is reached.

[0034] Specifically, in each generation of the population, according to the fitness function formula... Calculate a quantified fitness value for each individual in a valid job sequence. Among them, the total duration of agricultural operations The total operational cost is obtained by summing the execution time and switching time of each task. The utilization rate of agricultural machinery resources is obtained by summing the resource consumption and labor costs of each task in the sequence. This is calculated by statistically analyzing the proportion of effective agricultural machinery operation time to total operation time. Three weighting coefficients are used. , , There are no fixed values; instead, they should be dynamically configured based on priorities and objectives in agricultural production, with the sum of the three coefficients always remaining at 1. After calculating the fitness values ​​of all individuals, the algorithm employs a tournament selection strategy: that is, each time, three individuals are randomly selected from the population, and the one with the best fitness value (i.e., the one with the highest fitness value) is chosen. The individual with the smallest value (because the function treats duration and cost as positive penalties and utilization as a positive benefit) is copied into the "mating pool." This process is repeated until the number of individuals in the "mating pool" reaches a preset size. This selection method ensures that superior genes have a higher probability of being preserved and passed on, while maintaining population diversity and avoiding premature entrapment in local optima.

[0035] New candidate job sequences are generated based on the current optimal job sequence. Two parent sequences are randomly selected from the "crossover pool," one of which is a copy of the current optimal sequence. The crossover operation employs a priority-preserving crossover method. By analyzing the two parent sequences and the priority relation matrix, task subsequence blocks that are sequentially identical in both parents and conform to the matrix constraints are inherited by the offspring. The remaining tasks are then filled in, thus minimizing the risk of generating illegal sequences while mixing parent genes. Mutation operations are performed on the offspring sequences generated by crossover with a certain probability, such as randomly swapping the positions of two tasks in the sequence or inserting a task into a new position. Both crossover and mutation may unintentionally violate priority constraints; therefore, each newly generated temporary candidate sequence must immediately call the sequence repair operator for inspection and correction. The repair operator traverses the sequence, and once it detects any pair of tasks violating the order defined in the priority relation matrix (i.e., the agricultural task that appears earlier in the sequence), it checks and corrects the sequence. However, it was a later agricultural task. Prerequisite tasks, satisfy If a task violates the constraint, a forced rearrangement is initiated. The rearrangement is based on the previously determined hierarchical topology of agricultural tasks. The repair algorithm repositions tasks that violate the constraints according to their correct topological hierarchy; for example, moving them to a legal position after all their predecessor tasks and before all their successor tasks, thus ensuring a legal job sequence is generated. By repeatedly executing the iterative cycle of "selection > crossover > mutation > repair," the population continuously evolves until a preset number of iterations is reached. The preset number of iterations is 500, which ensures that the algorithm fully converges to obtain the globally optimal operation and maintenance sequence while also considering the computational efficiency of the cloud platform and avoiding redundant consumption of computational resources due to excessive iteration, thus achieving a balance between optimality and real-time responsiveness. The optimal sequence obtained at this point is the optimal operation and maintenance job sequence that balances duration, cost, and utilization. This method tightly integrates heuristic search with hard constraint guarantees, making the optimization process both efficient and reliable.

[0036] Through competitive selection based on fitness values ​​and continuous crossover and mutation operations, this mechanism not only identifies better scheduling sequences but also explores unknown combinations of solutions by simulating an evolutionary process. This effectively avoids the limitations of traditional planning methods, which are prone to getting trapped in local optima or relying too heavily on a single rule. The final output of this process is an operational plan that achieves an optimal balance among multiple key indicators under given preferences, thus substantially improving the scientific, economical, and refined management level of agricultural planting and maintenance at the algorithmic level. Furthermore, by explicitly defining the optimization objective and weight coefficients in formulaic form, the technical solution is fully disclosed, and a flexible adjustment interface is provided to adapt to specific management strategies for different farms and seasons.

[0037] In specific implementation, the processing logic for determining the search space of operation and maintenance parameters for each agricultural task based on the optimal operation and maintenance operation sequence includes: Extract historical agronomic parameters from the standardized agricultural operation and maintenance dataset that correspond to the agricultural task set, and identify the process baseline value of each agricultural task under the preset yield target. The process reference values ​​include sowing depth, fertilizer nutrient ratio, irrigation duration, and pesticide ratio; Based on the execution time span of each agricultural task in the optimal operation and maintenance sequence, the corresponding environmental monitoring data and crop growth data are matched, and the range of physiological demand fluctuations of each agricultural task in the current growth period is calculated through crop growth status indicators. The initial value range of each operation and maintenance parameter is determined by taking the process baseline value as the central reference point and combining it with the range of physiological demand fluctuations. The system retrieves the operating cost threshold and total resource supply from the resource consumption data, performs boundary constraint pruning on the initial value range, removes parameters that exceed the cost budget for resource reserve capacity, and generates a search space for operation and maintenance parameters corresponding to each agricultural task.

[0038] Specifically, historical agronomic parameters refer to the set of operational parameter settings and actual effect data corresponding to each agricultural task in past planting cycles. For example, the specific duration and amount of irrigation, detailed fertilizer formulas, and crop growth response records corresponding to these operations.

[0039] Based on this historical data, data mining and agronomic model analysis are used to identify the technological baseline values ​​for each agricultural task under the preset yield target. By screening historical planting records that have reached or approached the current preset yield target, the mode and mean of the operational parameters such as sowing depth and fertilizer ratio are calculated to determine the set of baseline parameters.

[0040] After obtaining the baseline value, the system enters the dynamic adjustment phase. It matches the corresponding environmental monitoring data and crop growth data based on the specific execution time window planned for each agricultural task in the optimal operation and maintenance sequence. The system extracts real-time environmental data within that time window from the database, as well as observed or simulated crop growth data during that growth period.

[0041] The system calculates the range of physiological demand fluctuations for each agricultural task during the current growth stage using crop growth status indicators. This relies on embedded crop growth models or expert knowledge rules. For example, based on the difference between the currently monitored crop growth and the standard growth model, the system calculates the crop's additional demand or tolerance range for water or nutrients at the current stage, thereby quantifying the positive and negative percentage fluctuations allowed for adjustments to various operational parameters based on baseline values.

[0042] Using the process baseline as the central reference point, and combining it with the fluctuation range of physiological needs, the initial value range of each operation and maintenance parameter is determined. This is a simple mathematical application, applying the calculated fluctuation range of ±20% to the baseline value to form a theoretical initial optimization range. For example, based on the baseline value of 4 hours, the irrigation duration can be explored between [3.2 hours, 4.8 hours].

[0043] By retrieving the operating cost threshold and total resource supply from resource consumption data, the initial value range is pruned to meet boundary constraints. The system simulates and calculates the total water consumption and total fertilizer cost corresponding to various parameter values ​​within the initial range, comparing them with preset budget limits and physical inventory levels. Any parameter value that would lead to resource overruns or cost overruns is removed from the range. For example, if the initial irrigation duration range causes the total water consumption to exceed the reservoir's current quota, the upper limit of the range will be forcibly lowered to a value that meets the quota requirements. Through this series of layer-by-layer screening and pruning based on agronomy, environment, and resources, a feasible operation and maintenance parameter space that can be directly searched by subsequent optimization algorithms is generated.

[0044] In practice, an improved Bayesian algorithm is used to optimize the configuration of operation and maintenance parameters to obtain the optimal configuration result. The processing logic includes: Step S31: Generate an objective function based on the standardized agricultural operation and maintenance dataset. The calculation formula is as follows: ; in, Represents the objective function value. This indicates the total number of agricultural tasks. Indicates the task index. This represents the operational parameters to be evaluated. Indicates the first Resource costs of agricultural tasks Indicates the first The projected yield and revenue of the agricultural task. Indicates the first Risk loss coefficient of agricultural tasks This represents the first weighting coefficient. This represents the second weighting coefficient. Indicates the third weighting coefficient; Step S32: In each round of configuration updates, the potential of untested operational parameters is evaluated using the expected improvement acquisition function, calculated as follows: ; in, This indicates the potential value obtained by improving the acquisition function. This represents the currently known optimal objective function value. This represents the predicted value of the operation and maintenance parameter to be evaluated by the mapping function. This represents the expected value operation; Step S33: Introduce an adaptive weight adjustment mechanism to calculate the internal parameters of the probabilistic correlation in real time, and correct the distribution characteristics of the probabilistic correlation through the objective function value. The correction formula is as follows: ; in, This indicates the corrected length scaling parameter. This indicates the length scaling parameters before correction. Represents an exponential function. Indicates the learning rate step size. Represents the gradient operator. Represents the logarithmic marginal likelihood function. Represents the probability density function. This represents the input sample set of operation and maintenance parameters. This represents the set of output values ​​of the objective function. Indicates the length scaling parameter; Step S34: In each round of optimization, execute the random selection logic. By generating a random number and comparing it with the preset exploration probability, if the random number is less than the preset exploration probability, random sampling is performed in the operation and maintenance parameter search space. Otherwise, the operation and maintenance parameter with the highest calculated value is selected based on the expected improvement collection function for the next round of iteration. Step S35: Repeat steps S32 to S34 until the preset maximum number of iterations is met, and obtain the optimal operation and maintenance parameter configuration result; The optimal operation and maintenance parameter configuration results include the optimal sowing depth, optimal fertilizer nutrient ratio, optimal irrigation duration, optimal pesticide application ratio, optimal agricultural machinery driving speed when performing agricultural tasks, and the start timestamp of each agricultural task plan corresponding to the optimal operation and maintenance sequence.

[0045] Specifically, the objective function takes the overall operation and maintenance cost as the core of the evaluation. By positively penalizing the task-level cost and risk items and negatively offsetting the benefit items, it achieves a unified quantitative comparison of different operation and maintenance parameter configuration schemes. Therefore, the smaller the function value, the better the overall performance of the corresponding scheme in terms of economy and security.

[0046] In the improved Bayesian optimization algorithm, the introduction of an adaptive weight adjustment mechanism aims to dynamically optimize key parameters within the probabilistic model, thereby enhancing accuracy. Specifically, after each round of trials, the algorithm adjusts the length scaling parameter of the kernel function in the Gaussian process model in real time based on the newly obtained actual objective function value. This adjustment is achieved through gradient descent: the gradient of the log-marginal likelihood function with respect to the length scaling parameter is calculated, reflecting the direction of the model's fit to the data under the current parameter settings; the length scaling parameter is updated along the gradient direction according to a preset learning rate step size, where the use of an exponential function ensures that the updated parameter always remains positive. Through this continuous adaptive correction, the model can better capture the true distribution characteristics of the objective function, reduce prediction bias, and thus more reliably guide subsequent optimization searches.

[0047] In each round of optimization, a simple random comparison mechanism balances exploration and utilization. Specifically, a uniformly distributed random number between 0 and 1 is generated and compared to a pre-set exploration probability threshold. If the generated random number is less than this exploration probability, an exploration operation is performed, randomly selecting a configuration point within the operational parameter search space for experimentation. This helps discover potential new optimal regions. Conversely, a utilization operation is performed, calculating the potential value of all candidate configuration points based on the expected improvement collection function, and selecting the point with the highest potential value for the next iteration. This helps deepen the search for the current optimal region. This random comparison mechanism encourages exploration with a controllable probability, effectively preventing the algorithm from prematurely getting trapped in local optima and ensuring that the optimization process has both breadth and depth.

[0048] This method transforms the complex problem of multi-objective parameter optimization—a high-dimensional, nonlinear, and uncertain problem—into an efficient, stepwise approximation process, thereby precisely identifying the most efficient operational parameters within the agronomically feasible domain. The core advantage of this method lies in its systematic nature: it defines a quantitative objective that balances cost, benefit, and risk loss, aligning the optimization direction with the actual concerns of farm operations; it employs a Bayesian optimization framework, significantly improving optimization efficiency through Gaussian processes and expectation-based improvements to the acquisition function, overcoming the blindness and high energy consumption of traditional grid search or random search; crucially, the introduced adaptive weight adjustment mechanism allows the algorithm to continuously correct its internal probability model based on historical experimental feedback, dynamically adjusting the prediction confidence for unknown regions, thereby continuously enhancing the targeting and accuracy of its search and effectively addressing common noise and nonlinear relationships in agricultural data; simultaneously, the embedded ε-greedy strategy cleverly balances the contradiction between utilizing known optimal regions and exploring potentially better regions, preventing the algorithm from prematurely converging to local optima. Ultimately, this method not only automatically outputs the optimal set of parameters covering all aspects of sowing, fertilization, irrigation, pesticide application, and agricultural machinery operation, but also provides a solid and reliable technical core for coping with changing environments and crop growth dynamics and achieving truly adaptive precision farming management by building the optimization process itself into a continuously learning and self-correcting intelligent system.

[0049] In practice, the optimal operation and maintenance sequence and the optimal operation and maintenance parameter configuration results are fused together to obtain the agricultural planting operation and maintenance management plan, which is then distributed to the execution terminal. The processing logic includes: Based on the agricultural task index in the optimal operation and maintenance sequence, retrieve the corresponding agricultural pest task operation and maintenance parameters in the optimal operation and maintenance parameter configuration results, and construct an agricultural planting operation and maintenance management plan table. Obtain real-time environmental monitoring data and match and verify the preset operation and maintenance parameters in the agricultural planting operation and maintenance management plan table with the environmental early warning threshold at the current moment; If the environmental monitoring data exceeds the preset safe operating range, the timestamp information in the operation and maintenance management plan table is adjusted according to the priority relationship matrix, and the adjusted operation and maintenance management plan table is converted into a structured control instruction set for execution terminal identification. The structured control instruction set includes the coordinates of the agricultural machinery automatic navigation operation trajectory, the pulse width modulation control parameters of the precision fertilizer applicator, and the opening and closing duration instructions of the irrigation solenoid valve; The structured control instruction set is transmitted to the corresponding execution terminal through the cloud platform interface; Collect feedback data from the execution terminal during the operation and store the feedback data in a standardized agricultural operation and maintenance dataset; Feedback data includes job completion status, execution time, resource consumption, and environmental monitoring data.

[0050] Specifically, based on the agricultural task index in the optimal operation and maintenance operation sequence, the refined parameters corresponding to each task are extracted from the optimal operation and maintenance parameter configuration results to construct an agricultural planting operation and maintenance management plan table; The agricultural planting operation and maintenance management plan is shown in Table 1. Table 1 ; After generating the plan table, it is not immediately issued. Instead, a dynamic safety verification process is initiated. Real-time field environmental monitoring data, such as soil moisture, temperature, and wind speed, is acquired, and the preset environmental safety range conditions for each task in the plan table are matched and verified against the current actual data. For example, for sowing task 1, the system will verify whether the current soil moisture is between 15% and 25%, whether the wind speed is ≤3, and query weather warning information. If all real-time data are within the preset safe operating range, the verification passes; if any indicator exceeds the standard, it is determined that the environmental monitoring data exceeds the preset safe operating range.

[0051] When the verification fails, the scheme adjustment mechanism is automatically triggered. The core basis of the adjustment logic is the constructed priority relationship matrix. According to the task dependencies defined in the priority relationship matrix, the affected tasks are rescheduled. For example, if the sowing task cannot be executed as originally planned due to strong winds, the system will postpone its planned timestamp and automatically calculate and postpone the planned times of all subsequent dependent tasks according to the priority relationship matrix, while ensuring that the logical constraints between all tasks still hold, generating an adjusted and secure agricultural planting operation and maintenance management scheme table.

[0052] If environmental monitoring data exceeds the preset safe operating range, the timestamp information in the operation and maintenance management plan table is adjusted according to the priority relationship matrix. During the execution phase, the optimal operation and maintenance sequence determined in step S2 is used to constrain the relative order between various agricultural tasks, and the optimal operation and maintenance parameter configuration result obtained in step S3 is used to limit the executable range of each agricultural task under different environmental conditions. The task timestamps in the operation and maintenance management plan table are not generated independently, but are planned while maintaining the established operation sequence structure and operation and maintenance parameter configuration. When the real-time environmental monitoring data is within the preset safe operating range, each agricultural task is executed according to the optimal operation and maintenance sequence and the corresponding optimal operation and maintenance parameters. When the environmental monitoring data temporarily exceeds the safe operating range corresponding to the optimal operation and maintenance parameters, the system does not recalculate the operation sequence or operation and maintenance parameters, but rearranges the task timestamps in the operation and maintenance management plan table according to the priority relationship matrix, so that the task execution time falls back into the applicable environmental range of the operation and maintenance parameters. In this way, the execution plan can be dynamically adjusted without destroying the global optimization result.

[0053] The revised agricultural planting operation and maintenance management plan is converted into a set of structured control instructions that can be directly recognized and executed by each execution terminal, such as the coordinates of the automatic navigation operation trajectory of agricultural machinery, the pulse width modulation (PWM) control parameters of the precision fertilizer applicator, and the opening and closing duration instructions of the irrigation solenoid valve. These instructions are then sent to the corresponding field execution terminals through the cloud platform interface to drive the equipment to complete automated and precise agricultural operations.

[0054] In practice, the processing logic for converting the adjusted operation and maintenance management scheme table into a structured control instruction set for execution terminal identification includes: Based on the agricultural task index in the adjusted operation and maintenance management plan table, retrieve the optimal operation and maintenance parameter configuration results corresponding to each agricultural task; Based on the execution sequence and operation time boundaries of agricultural tasks, the operation and maintenance parameters of each agricultural task are converted into a control instruction format that can be recognized by the execution terminal. The coordinates of the automatic navigation operation trajectory of agricultural machinery, the pulse width modulation control parameters of precision fertilizer applicators, and the opening and closing duration instructions of irrigation solenoid valves are converted into standardized control instructions according to a preset structured data format. Perform compliance checks on standardized control commands to determine whether the control parameters meet the operating specifications of the execution terminal and the requirements of agricultural tasks.

[0055] Specifically, the pre-defined structured data format is essentially a pre-defined and negotiated unified machine-readable data encapsulation and exchange specification among various agricultural execution terminals. Its core technology is a framework containing a fixed sequence of fields, strict data type definitions, and explicit communication semantic rules. Specifically, each complete standardized control command data packet follows this framework, typically including a protocol header and checksum to identify the start and end of the command and ensure data integrity; a device addressing identifier to precisely specify the target device; an instruction opcode to define the specific operation type; a parameter payload area carrying all operational parameter values ​​in a predetermined order and unit; and metadata indicating execution time or priority. The specific implementation of this format can be based on the common JSON data serialization standard or adapted to specific bus protocols in the agricultural field, such as ISO 11783.

[0056] When processing a seeding task, parameters such as operation speed, seeding depth, and navigation trajectory coordinates are extracted from the operation and maintenance management plan table. Based on a preset structured data format, these parameters are mapped and encapsulated into a structured data packet. For example, a JSON-compliant instruction object is generated, containing complete header information, device identifier, timestamp, specific seeding instruction code, and payload data organized as key-value pairs of speed, depth, and coordinate point arrays. Before issuing the instruction, the data packet undergoes compliance checks based on this format, such as verifying whether the parameter values ​​are within the safe operating range of the terminal.

[0057] The primary purpose of this standardized architecture is to achieve broad device interoperability. It serves as a reliable common language between the cloud platform and diverse, multi-brand agricultural equipment, effectively solving the fragmentation problem in agricultural IoT. Through its fixed structure and built-in verification mechanisms, the reliability and security of control commands during transmission and parsing are fundamentally guaranteed, meeting the stringent requirements of precision agriculture for high reliability of operational commands. This design endows the system with excellent scalability and maintainability. When new agricultural machinery needs to be integrated, only the corresponding instruction codes and payload formats need to be extended, without changing the core communication architecture, greatly reducing the complexity of system integration and upgrades.

[0058] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0059] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the protection scope of the present invention.

Claims

1. A cloud-based agricultural planting operation and maintenance management method, characterized in that, Includes the following steps: Step S1: Obtain agricultural planting operation and maintenance data and agricultural task set, and preprocess the agricultural planting operation and maintenance data to obtain a standardized agricultural operation and maintenance dataset; Step S2: Construct a priority relationship matrix among agricultural tasks based on the standardized agricultural operation and maintenance dataset, and use an improved genetic algorithm to optimize the sequence of agricultural tasks to obtain the optimal operation and maintenance sequence; The processing logic for constructing a priority relationship matrix among agricultural tasks based on standardized agricultural operation and maintenance datasets includes: Extract the sequential logical constraints between agricultural tasks in the standardized agricultural operation and maintenance data. The sequential logical constraints between agricultural tasks include the irreversible constraint of crop physiological growth, the conflict constraint of agricultural machinery and equipment allocation, and the timing constraint of water, fertilizer and pesticide application. Build Priority relation matrix of dimension ; in, This indicates the total number of agricultural tasks. Represents the first in the priority relation matrix Line 1 Column matrix elements; If agricultural tasks Must be in agricultural tasks Execute after completion, then... Otherwise ; The Kahn algorithm is used to perform topological sorting on the priority relation matrix to obtain the initial sequence. The in-degree of each agricultural task is counted. Agricultural tasks with an in-degree of 0 are selected in a loop and added to the initial sequence while the remaining in-degrees are updated synchronously until an initial task sequence individual that meets the priority logical constraints between agricultural tasks is generated. Step S3: Based on the optimal operation and maintenance operation sequence, determine the search space of operation and maintenance parameters for each agricultural task, and use an improved Bayesian algorithm to optimize the configuration of operation and maintenance parameters to obtain the optimal operation and maintenance parameter configuration result; Step S4: The optimal operation and maintenance operation sequence and the optimal operation and maintenance parameter configuration result are merged and processed to obtain the agricultural planting operation and maintenance management plan and then distributed to the execution terminal.

2. The cloud-based agricultural planting operation and maintenance management method as described in claim 1, characterized in that, The agricultural planting and operation data includes environmental monitoring data, crop growth data, operation execution data, and resource consumption data; The environmental monitoring data includes temperature data, humidity data, light intensity data, and soil moisture data; The crop growth data includes crop growth period information, growth status indicators, and yield-related data. The yield-related data includes crop yield per unit area, crop growth stage yield indicators, and historical yield statistics. The operation execution data includes sowing operation data, fertilization operation data, irrigation operation data, and pest and disease control operation data; The resource consumption data includes water consumption data, fertilizer consumption data, energy consumption data, and operating cost data; The agricultural task set includes sowing tasks, fertilization tasks, irrigation tasks, pest and disease control tasks, and the attributes corresponding to each agricultural task. The attributes corresponding to each agricultural task include task priority attribute, task execution time limit, and logical constraints on the order of agricultural tasks.

3. The cloud-based agricultural planting operation and maintenance management method as described in claim 1, characterized in that, An improved genetic algorithm is used to optimize the sequence of agricultural tasks to obtain the optimal operation and maintenance job sequence. The processing logic includes: agricultural task set Each agricultural task in the process is mapped to a unique integer gene bit, resulting in a sequence containing... The arrangement of 1 gene loci serves as the initial chromosome population; Each arrangement represents a single task sequence. Crossover and mutation operations are performed on the initial chromosome population to obtain temporary candidate sequence individuals. ; The sequence repair operator is invoked to perform logical validity checks and corrections on temporary candidate sequence individuals, and matrix elements in the priority relation matrix are retrieved. If any two gene loci combinations are found in an individual with a temporary candidate sequence satisfy and ; in, And corresponding matrix elements If so, the individual is deemed to have violated logical constraints; The hierarchical topological order of agricultural tasks is determined based on the priority relation matrix. Gene positions that are determined to violate logical constraints are forcibly rearranged. By adjusting the arrangement of agricultural tasks, the repaired temporary candidate sequence individuals are made to satisfy all... The corresponding agricultural tasks are sequentially constrained to generate a valid sequence of operations; A fitness function is constructed with the objectives of maximizing the total duration of agricultural operations, the total cost of operation and maintenance, and the utilization rate of agricultural machinery resources. The fitness function is then used to select the legal operation sequence to obtain the optimal operation and maintenance sequence.

4. The cloud-based agricultural planting operation and maintenance management method as described in claim 3, characterized in that, The processing logic for selecting a valid job sequence using a fitness function includes: For each valid job sequence, calculate its fitness value. The formula for the fitness function is: ； in, Indicates the fitness value. Indicates the total duration of agricultural operations. This represents the total cost of operation and maintenance. Indicates the utilization rate of agricultural machinery resources. This represents the weighting coefficient for the total duration of agricultural operations. This represents the weighting coefficient of total operation and maintenance costs. This represents the weighting coefficient for agricultural machinery resource utilization. Based on the fitness function value, the legal job sequences are filtered, and the legal job sequence with the best fitness value is selected as the current optimal sequence, and the current optimal sequence is used as the basis for the next generation of job sequences. By performing selection, crossover, and mutation operations, new candidate job sequences are generated based on the current optimal job sequence, and the selection, crossover, and mutation operations are repeated until the preset number of iterations is reached.

5. The cloud-based agricultural planting operation and maintenance management method as described in claim 1, characterized in that, The processing logic for determining the search space of operation and maintenance parameters for each agricultural task based on the optimal operation and maintenance job sequence includes: Extract historical agronomic parameters from the standardized agricultural operation and maintenance dataset that correspond to the agricultural task set, and identify the process baseline value of each agricultural task under the preset yield target. The process reference values ​​include sowing depth, fertilizer nutrient ratio, irrigation duration, and pesticide ratio; Based on the execution time span of each agricultural task in the optimal operation and maintenance sequence, the corresponding environmental monitoring data and crop growth data are matched, and the range of physiological demand fluctuations of each agricultural task in the current growth period is calculated through crop growth status indicators. The initial value range of each operation and maintenance parameter is determined by taking the process baseline value as the central reference point and combining it with the range of physiological demand fluctuations. The system retrieves the operating cost threshold and total resource supply from the resource consumption data, performs boundary constraint pruning on the initial value range, removes parameters that exceed the cost budget for resource reserve capacity, and generates a search space for operation and maintenance parameters corresponding to each agricultural task.

6. The agricultural planting operation and maintenance management method based on a cloud platform as described in claim 1, characterized in that, An improved Bayesian algorithm is used to optimize the configuration of operation and maintenance parameters, resulting in the optimal configuration. The processing logic includes: Step S31: Generate an objective function based on the standardized agricultural operation and maintenance dataset. The calculation formula is as follows: ； in, Represents the objective function value. This indicates the total number of agricultural tasks. Indicates the task index. This represents the operational parameters to be evaluated. Indicates the first Resource costs of agricultural tasks Indicates the first The projected yield and revenue of the agricultural task. Indicates the first Risk loss coefficient of agricultural tasks This represents the first weighting coefficient. This represents the second weighting coefficient. Indicates the third weighting coefficient; Step S32: In each round of configuration updates, the potential of untested operational parameters is evaluated using the expected improvement acquisition function, calculated as follows: ； in, This indicates the potential value obtained by improving the acquisition function. This represents the currently known optimal objective function value. This represents the predicted value of the operation and maintenance parameter to be evaluated by the mapping function. This represents the expected value operation; Step S33: Introduce an adaptive weight adjustment mechanism to calculate the internal parameters of the probabilistic correlation in real time, and correct the distribution characteristics of the probabilistic correlation through the objective function value. The correction formula is as follows: ； in, This indicates the corrected length scaling parameter. This indicates the length scaling parameters before correction. Represents an exponential function. Indicates the learning rate step size. Represents the gradient operator. Represents the logarithmic marginal likelihood function. Represents the probability density function. This represents the input sample set of operation and maintenance parameters. This represents the set of output values ​​of the objective function. Indicates the length scaling parameter; Step S34: In each round of optimization, execute the random selection logic. By generating a random number and comparing it with the preset exploration probability, if the random number is less than the preset exploration probability, random sampling is performed in the operation and maintenance parameter search space. Otherwise, the operation and maintenance parameter with the highest calculated value is selected based on the expected improvement collection function for the next round of iteration. Step S35: Repeat steps S32 to S34 until the preset maximum number of iterations is met, and obtain the optimal operation and maintenance parameter configuration result; The optimal operation and maintenance parameter configuration results include the optimal sowing depth, optimal fertilizer nutrient ratio, optimal irrigation duration, optimal pesticide application ratio, optimal agricultural machinery driving speed when performing agricultural tasks, and the start timestamp of each agricultural task plan corresponding to the optimal operation and maintenance sequence.

7. The agricultural planting operation and maintenance management method based on a cloud platform as described in claim 1, characterized in that, The optimal operation and maintenance work sequence and the optimal operation and maintenance parameter configuration results are fused together to obtain an agricultural planting operation and maintenance management plan, which is then distributed to the execution terminal. The processing logic includes: Based on the agricultural task index in the optimal operation and maintenance sequence, retrieve the corresponding agricultural pest task operation and maintenance parameters in the optimal operation and maintenance parameter configuration results, and construct an agricultural planting operation and maintenance management plan table. Obtain real-time environmental monitoring data and match and verify the preset operation and maintenance parameters in the agricultural planting operation and maintenance management plan table with the environmental early warning threshold at the current moment; If the environmental monitoring data exceeds the preset safe operating range, the timestamp information in the operation and maintenance management plan table is adjusted according to the priority relationship matrix, and the adjusted operation and maintenance management plan table is converted into a structured control instruction set for execution terminal identification. The structured control instruction set includes the coordinates of the agricultural machinery automatic navigation operation trajectory, the pulse width modulation control parameters of the precision fertilizer applicator, and the opening and closing duration instructions of the irrigation solenoid valve. The structured control instruction set is transmitted to the corresponding execution terminal through the cloud platform interface; Collect feedback data from the execution terminal during the operation and store the feedback data in a standardized agricultural operation and maintenance dataset; The feedback data includes job completion status, execution time, resource consumption, and environmental monitoring data.

8. The cloud-based agricultural planting operation and maintenance management method as described in claim 7, characterized in that, The processing logic for converting the adjusted operation and maintenance management plan table into a structured control instruction set for execution terminal recognition includes: Based on the agricultural task index in the adjusted operation and maintenance management plan table, retrieve the optimal operation and maintenance parameter configuration results corresponding to each agricultural task; Based on the execution sequence and operation time boundaries of agricultural tasks, the operation and maintenance parameters of each agricultural task are converted into a control instruction format that can be recognized by the execution terminal. The coordinates of the automatic navigation operation trajectory of agricultural machinery, the pulse width modulation control parameters of precision fertilizer applicators, and the opening and closing duration instructions of irrigation solenoid valves are converted into standardized control instructions according to a preset structured data format. Perform compliance checks on standardized control commands to determine whether the control parameters meet the operating specifications of the execution terminal and the requirements of agricultural tasks.

9. A cloud-based agricultural planting operation and maintenance management system, which is applied in the cloud-based agricultural planting operation and maintenance management method as described in any one of claims 1-8, characterized in that, include: Data acquisition module, matrix construction module, sequence optimization module, operation and maintenance management module; The data acquisition module is used to acquire agricultural planting and operation data and agricultural task sets, and to preprocess the agricultural planting and operation data to obtain a standardized agricultural operation and operation dataset. The matrix construction module is used to construct a priority relationship matrix among agricultural tasks based on a standardized agricultural operation and maintenance dataset, and to use an improved genetic algorithm to optimize the sequence of agricultural tasks to obtain the optimal operation and maintenance sequence. The sequence optimization module is used to determine the search space of operation and maintenance parameters for each agricultural task based on the optimal operation and maintenance operation sequence, and to use an improved Bayesian algorithm to optimize the configuration of operation and maintenance parameters to obtain the optimal operation and maintenance parameter configuration result. The operation and maintenance management module is used to integrate the optimal operation and maintenance operation sequence with the optimal operation and maintenance parameter configuration result to obtain the agricultural planting operation and maintenance management plan and distribute it to the execution terminal.

Images