Agricultural planting decision-making method based on simulated annealing algorithm

Abstract

The invention relates to the technical field of agricultural planting, in particular to an agricultural planting decision-making method based on a simulated annealing algorithm. Factors such as types, areas, time and land parcels of crops planted on different types of cultivated lands are modeled to obtain a mathematical model, a series of constraint conditions are added according to actual conditions, the model is solved under the constraint conditions to obtain an optimal planting decision, agricultural planting profit is maximized, and agricultural planting efficiency is improved. The model is optimized by using Monte Carlo simulation and a particle swarm optimization algorithm, so that the planting decision method calculated by the model is more practical; the method can be widely applied to the technical field of agricultural planting.

Description

Technical Field

[0001] The present invention relates to the field of agricultural planting technology, and in particular to an agricultural planting decision-making method based on a simulated annealing algorithm. Background Art

[0002] In the current global context of sustainable development, the transformation and development of the rural economy is particularly important. However, due to the low temperature throughout the year in some areas, some arable land in the area can only be planted once a year, which seriously limits the agricultural output and economic benefits of farmers' cultivated land.

[0003] In order to maintain the fertility of cultivated land, the same crop cannot be planted year after year, otherwise it will lead to soil nutrient imbalance, aggravation of pests and diseases, imbalance of soil microbial community, self-toxicity and deterioration of soil physical properties. Leguminosae plants have a unique nitrogen fixation effect that can better enhance the fertility of cultivated land, so leguminous plants should be planted at least once every three years when planting cultivated land.

[0004] On the basis of maintaining the fertility of cultivated land, the planting products and planting area of ​​cultivated land are modeled based on the simulated annealing algorithm. At the same time, the above model is optimized according to the changing trends of factors such as sales volume and price of cultivated crops, and finally a planting decision-making method that maximizes economic benefits is obtained. Summary of the invention

[0005] The purpose of the present invention is to provide an agricultural planting decision-making method based on a simulated annealing algorithm to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: an agricultural planting decision-making method based on a simulated annealing algorithm, which models the types, areas, time and plots of crops planted on different types of cultivated land to obtain the following mathematical model:

[0007]

[0008] R1 is the total profit without considering dumping, and R2 is the total profit when selling at half price considering dumping. The simulated annealing algorithm is used to solve R1 and R2 in the above model to obtain the planting decision method that maximizes economic benefits.

[0009] The above model can more accurately predict the total profit of planting crops, making it easier for people to make planting decisions based on the model.

[0010] As a preference, the mathematical model has the following constraints when used:

[0011] Flat dry land, terraced fields, and hillside land can only grow a single crop of grain (except rice) each year:

[0012] Plot j can only grow one crop i every year during all time periods t:

[0013]

[0014] And plot j cannot grow rice within any time t:

[0015]

[0016] Flat dry land, terraces and hillside land are limited by their own conditions and can only be planted in a single season. Rice planting requires a lot of water. Adding the above constraints to the model makes it more realistic.

[0017] As a preference, the mathematical model has the following constraints when used:

[0018] Irrigated land can be used to grow rice in one season or vegetables in two seasons:

[0019] In time t, irrigated land j can be planted with either rice or vegetables, but not both:

[0020]

[0021] Only specific crops i can be planted in the second season of irrigated land j and all time periods t:

[0022]

[0023] Irrigated land can be used to grow rice in a single season or vegetables in two seasons: Adding the above constraints to the model makes it more realistic.

[0024] Preferably, the mathematical model has the following constraints when used: Chinese cabbage, white radish and red radish can only be planted in the second season of irrigated land; specific vegetable i is prohibited from being planted in the first season t of irrigated land j:

[0025]

[0026] Chinese cabbage, white radish and red radish can only be planted in the second season of irrigated land: specific vegetable i is prohibited from being planted in the first season t of irrigated land j; adding the above constraints to the model makes the model more realistic.

[0027] As a preferred embodiment, the mathematical model satisfies the following constraints when used: ordinary greenhouses have two seasons a year, a variety of vegetables in the first season, and only edible fungi can be planted in the second season; in ordinary greenhouses j, one vegetable i must be planted in the first season:

[0028]

[0029] In ordinary greenhouses, edible fungi must be planted in the second season:

[0030]

[0031] Ordinary greenhouses have two planting seasons each year. In the first season, a variety of vegetables are planted, and in the second season, only edible fungi can be planted. In ordinary greenhouse j, one vegetable i must be planted in the first season. Adding the above constraints to the model makes it more realistic.

[0032] As a preference, the mathematical model satisfies the following constraints when used: the smart greenhouse has two seasons of vegetables each year, excluding specific vegetables: in the smart greenhouse j, one vegetable i must be planted in each season:

[0033]

[0034] In the smart greenhouse j, certain crops cannot be planted i:

[0035]

[0036] The smart greenhouse has two seasons of vegetables every year, excluding specific vegetables: In smart greenhouse j, one vegetable i must be planted every season. Adding the above constraints to the model makes the model more realistic.

[0037] As a preference, the mathematical model satisfies the following constraints when used: the actual sales volume cannot exceed the planting area multiplied by the output or the maximum sales volume:

[0038] Sales volume z of crop i at time t, season s and plot j i,t,s,j The maximum sales volume S of the crop cannot be exceeded i :

[0039]

[0040] Sales volume z of crop i at time t, season s and plot j i,t,s,j Cannot exceed the planting area x of the crop on plot j i,t,s,j and unit area yield y i The product of:

[0041]

[0042] Sales volume z of crop i at time t, season s and plot j i,t,s,j The maximum sales volume S of the crop cannot be exceeded i ; Sales volume z of crop i at time t, season s and plot j i,t,s,j Cannot exceed the planting area x of the crop on plot j i,t,s,j and unit area yield y i Adding the above constraints to the model makes it more realistic.

[0043] As a preference, the mathematical model satisfies the following constraints when used: the planting area x of crop i at time t, season s and plot j i,t,s,j Cannot exceed the total area A of plot j j Multiply by the decision variable b i,t,s,j , where the decision variable b i,t,s,j Indicates whether to choose to plant crops at this plot and time:

[0044]

[0045] Planting decision variable b of crop i in plot j at time t and t+1 i,t,s,j and b i,t+1,s,j The sum of cannot exceed 1, that is, the same crop cannot be planted on the same plot for two consecutive years:

[0046]

[0047] As a preferred method, the total profit of cultivated land planting in the next T years is optimized and modeled to obtain:

[0048]

[0049] The model complies with the following constraints:

[0050] y i，t =y i ·(1+δ i，t ), δ i，t ∈[-0.1, 0.1]

[0051] S i，t =S i ·(1+∈ i，t ),∈ i，t ∈[-0.05, 0.05]

[0052] c i，t =c i (1+0.05) t-2023

[0053] z i,t,s,j ≤x i,t,s,j ·y i,t

[0054] The Monte Carlo simulation model is used to optimize and solve the above model under the above constraints.

[0055] By using the Monte Carlo simulation model to optimize the maximum planting profit in the next few years under the above constraints, the agricultural planting profit is maximized.

[0056] As a preference, the following constraints are added to the mathematical model:

[0057]

[0058] The particle swarm optimization algorithm is used to optimize and solve the model under the above constraints.

[0059] The price fluctuations that may occur during the planting process are converted into the above constraint model, and the particle swarm optimization algorithm is used to solve the model under the above constraint model to make the model more suitable for practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 A symbol description table for the symbols used in the model;

[0061] Figure 2 The table of the existing crops planted in the village and the plots and categories suitable for the crops in Example 1;

[0062] Figure 3 The planting area table of some crops in the village in Example 1;

[0063] Figure 4 The planting season and corresponding data table of the village in Example 1;

[0064] Figure 5 It is a summary table of some crop planting areas in the village in Example 1;

[0065] Figure 6 This is a flow chart of using the simulated annealing algorithm to solve the model in Example 1;

[0066] Figure 7 The table of the number of acres of some crops planted on each piece of land obtained by using the simulated annealing algorithm in Example 1;

[0067] Figure 8 The figure is a flow chart for optimizing and solving the model in Example 2 using the Monte Carlo simulation model. DETAILED DESCRIPTION

[0068] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0069] Example 1

[0070] Villages in the mountainous areas of North China face the dual challenges of climate conditions and arable land resources, and are in urgent need of exploring an agricultural development model that suits their own characteristics. The village has 1,201 acres of arable land, divided into 34 different types of plots, including flat dry land, terraces, hillside land and irrigated land. Due to the low temperature all year round, most arable land can only grow one crop per year, which limits the diversity and economic benefits of agricultural output. The model is optimized based on the situation where overproduction crops cannot be sold and lead to waste; the other is that in the case of dumping, the overproduction part can be sold at half price.

[0071] The simulated annealing algorithm was used to model the decision-making method of planting the village's arable land in the next year.

[0072] The model was established with the following assumptions

[0073] 1. It is assumed that during the study period, climate conditions remain stable and there are no extreme weather events that affect crop growth.

[0074] 2. Assume that the crop growth cycle, yield and demand are relatively stable and consistent with historical data and trends.

[0075] 3. Assume that the market demand, prices and sales channels of agricultural products remain relatively stable in the next few years.

[0076] 4. Assume that all historical data and statistical information used in model construction are accurate and reliable.

[0077] 5. Assume that the crop output is our expected sales volume.

[0078] 6. Suppose a piece of land can grow multiple crops.

[0079] like Figure 2The figure shows the existing crops planted in the village and the plots and categories suitable for the crops. We have conducted a detailed classification and analysis of the cultivated land for various crops. The study found that the selection of cultivated land for crops is closely related to their growth characteristics and needs. For example, grain crops such as soybeans, black beans, red beans, mung beans, climbing beans, wheat, corn, millet, sorghum, millet, buckwheat, pumpkin, sweet potato, oats and barley are usually planted on traditional cultivated land such as flat dry land, terraces and hillsides, most of which can only be harvested once a year. In contrast, rice, as a special case among grain crops, is exclusively planted in irrigated land with abundant water resources due to its high dependence on humid environment, which reflects the consideration of specific ecological adaptability in crop planting planning. The distribution of cultivated land for vegetable crops is more diverse. Legumes such as cowpeas, sword beans, kidney beans, etc., as well as potatoes, tomatoes, eggplants, spinach, green peppers, cauliflower, cabbage, lettuce, pakchoi, cucumbers, lettuce, peppers, water spinach, yellow cabbage, celery, Chinese cabbage, etc., are widely planted in irrigated fields, ordinary greenhouses and smart greenhouses. Smart greenhouses use solar energy to automatically adjust the temperature, ensuring continuous planting of vegetables in winter and realizing an efficient two-season planting model. In addition, edible fungi such as elm yellow mushrooms, shiitake mushrooms, white mushrooms and morels are particularly focused on greenhouse cultivation to maintain an ideal growth environment. For example, elm yellow mushrooms are adapted to cultivation in the subsequent planting season of ordinary greenhouses, emphasizing the specific needs of edible fungi species for unique cultivation periods and environments.

[0080] In summary, the data table of this study shows that the type of cultivated land has a significant impact on the growth cycle, yield and quality of crops. Scientific planning and utilization of cultivated land can significantly improve crop production potential and enhance agricultural efficiency and sustainability.

[0081] like Figure 3 The data shows the current status of some crops, including the specific varieties of grains, vegetables and edible fungi and their total planting area. The data reveals that wheat and corn dominate grain crops, while potatoes and white radishes are the main vegetables. Although the area of ​​edible fungi is relatively small, the planting of various varieties such as elm yellow mushrooms and morels is also clearly presented. These data provide us with valuable insights into the distribution of different crop plantings and their proportion in the overall area.

[0082] like Figure 4 As shown in the figure, it reveals that single-season crops occupy the most planting plots, with a total area of ​​1,046 mu. Vegetable crops are planted in different seasons, with the planting areas of 22.9 mu and 17.7 mu in the first and second seasons respectively. The cultivation of edible fungi is also divided into two seasons, with a total area of ​​6.0 mu. These data reveal the distribution of crops in different planting seasons, providing a basis for our in-depth analysis of the current status of crop planting in each season.

[0083] like Figure 5As shown in the table, the total planting area data of various crops are presented. Among them, wheat has the largest planting area, reaching 380 mu. The total planting area of ​​major crops such as corn and soybeans is also listed in the table. For vegetable crops, the planting area of ​​potatoes and white radish is also reflected. In addition, although the planting area of ​​edible fungi is relatively small, the table still shows the planting situation of different edible fungi varieties. These summary data help us evaluate the planting scale of different crops and the overall situation of agricultural production.

[0084] The establishment of the model, the objective function: maximize the total profit, that is, the sum of the product of the planting area of ​​all crops and their unit profit (sales price p minus cost), where R1 is the total profit without considering dumping, and R2 is the total profit when selling at half price considering dumping:

[0085]

[0086] We design the following constraints in the above function to ensure that the planting arrangements of various plots and crops follow scientific planning principles.

[0087] Constraint 1: Flat dry land, terraced fields, and hillside land can only grow a single crop of food crops (except rice) each year:

[0088] Plot j can only grow one crop i every year during all time periods t:

[0089]

[0090] And plot j cannot grow rice within any time t:

[0091]

[0092] Constraint 2: Irrigated land can be used to grow rice in one season or vegetables in two seasons:

[0093] In time t, irrigated land j can be planted with either rice or vegetables, but not both:

[0094]

[0095] Only specific crops i can be planted in the second season of irrigated land j and all time periods t:

[0096]

[0097] Constraint 3: Chinese cabbage, white radish and carrot can only be planted in the second season of irrigated land: Specific vegetable i is prohibited from being planted in the first season t of irrigated land j:

[0098]

[0099] Constraint 4: Ordinary greenhouses have two seasons each year. In the first season, a variety of vegetables are grown, and in the second season, only edible fungi can be grown. In ordinary greenhouse j, one vegetable i must be planted in the first season:

[0100]

[0101] In ordinary greenhouses, edible fungi must be planted in the second season:

[0102]

[0103] Constraint 5: Two seasons of vegetables in the smart greenhouse each year, excluding specific vegetables: In smart greenhouse j, one vegetable i must be planted in each season:

[0104]

[0105] In the smart greenhouse j, certain crops cannot be planted i:

[0106]

[0107] Constraint 6: The actual sales volume cannot exceed the planting area multiplied by the yield or the maximum sales volume:

[0108] Sales volume z of crop i at time t, season s and plot j i,t,s,j The maximum sales volume S of the crop cannot be exceeded i :

[0109]

[0110] Sales volume z of crop i at time t, season s and plot j i,t,s,j Cannot exceed the planting area x of the crop on plot j i,t,s,j and unit area yield y i The product of:

[0111]

[0112] Constraint 7: Relationship between planted area and binary decision variables:

[0113] The planting area x of crop i at time t, season s and plot j i,t,s,j Cannot exceed the total area A of plot j j Multiply by the decision variable b i,t,s,j , where the decision variable b i,t,s,j Indicates whether to choose to plant crops at this plot and time:

[0114]

[0115] Constraint 8: A plot cannot be planted with the same crop for two consecutive years:

[0116] Planting decision variable b of crop i in plot j at time t and t+1 i,t,s,j and b i,t+1,s,j The sum of cannot exceed 1, that is, the same crop cannot be planted on the same plot for two consecutive years:

[0117]

[0118] In summary, we have arrived at the following mathematical model of maxR1orR2:

[0119]

[0120] We use the simulated annealing algorithm to optimize the above function model. The core mechanism of the algorithm is to randomly perturb the current solution and decide whether to accept the perturbed solution based on the current "annealing temperature". As the temperature decreases, the algorithm's tolerance for poor solutions decreases, gradually guiding the search process to the global optimal solution, and finally achieving an approximate global optimal solution.

[0121] like Figure 6 As shown, the solution using the simulated annealing algorithm includes the following steps:

[0122] 1. Generation of initial solution: First, according to the constraints of the problem, we generate an initial planting plan Z i,t,s,j , as the starting point of the algorithm.

[0123] 2. Definition of objective function: We define the objective function as total profit, and its calculation formula is:

[0124]

[0125]

[0126] Where pi represents the unit profit of crop i, c i represents the unit cost of crop i. The goal is to maximize the total profit.

[0127] 3. Generation of neighborhood solutions: By analyzing the current planting plan Z i,t,s,j Perform random perturbations to generate neighborhood solutions. The perturbation operation can include randomly changing the planting area or crop type of a certain plot of land.

[0128] 4. Acceptance criteria: If the objective function value of the new solution is better than the current solution, the new solution is accepted. If the new solution is worse, the new solution is accepted with a certain probability, which is determined by the annealing temperature T and the calculation formula is:

[0129]

[0130] Among them, ΔProfit represents the profit difference between the new solution and the current solution.

[0131] 5. Temperature update: After each iteration, the annealing temperature T will decrease. Usually the geometric cooling method is used, and the update formula is:

[0132] T new =α·T current

[0133] Among them, α is the temperature reduction coefficient, and its value range is usually 0.8≤α≤0.99.

[0134] 6. Termination condition: The algorithm terminates under any of the following conditions: the temperature drops below the preset threshold, or the solution does not improve significantly in multiple consecutive iterations.

[0135] In the simulated annealing algorithm, the handling of model constraints is the key to ensuring the feasibility of the solution. The following are the handling strategies for specific constraints:

[0136] 1. Handling of hard constraints: For those constraints that must be strictly followed, such as plot j can only grow one type of food crop (except rice) every year in all time t, we enforce these constraints in the process of generating neighborhood solutions to ensure that each generated solution is feasible. The specific constraint expression is:

[0137]

[0138] 2. Handling of soft constraints: For those constraints that can be moderately relaxed, such as a plot of land cannot be planted with the same crop for two consecutive years, we use the penalty function method to handle them. If a solution fails to meet a specific constraint, we will impose a penalty by introducing a penalty term in the objective function to guide the algorithm's search process to move closer to the feasible area that meets the constraint. The specific constraint expression is:

[0139]

[0140] Through the above methods, we can effectively deal with various constraints in the model, thereby ensuring the feasibility of the solution and the realization of the optimization goal while maintaining the flexibility of the algorithm.

[0141] The model parameters of this embodiment are set as follows:

[0142] To ensure that the simulated annealing algorithm can effectively traverse the solution space and converge to the global optimal solution, we set the key parameters in detail:

[0143] Initial temperature T0: set to 1000 to provide enough randomness in the early stage of the algorithm, so as to accept a larger range of de-perturbations at the beginning of the search.

[0144] Cooling coefficient α: set to 0.95 to regulate the rate of temperature reduction and achieve a balance between exploration and development.

[0145] Termination temperature T min : Set to 1×10 -5 , in order to clearly define the temperature limit for algorithm termination and prevent endless iterations.

[0146] Maximum number of iterations: Set to 10,000 to ensure that the algorithm has enough iterations before reaching a stable state.

[0147] like Figure 7 As shown, the above model is used to calculate the number of acres of some crops planted on each piece of land.

[0148] Example 2

[0149] In actual production, factors such as the sales volume and price of crops are uncertain and will change over time. It is necessary to use optimization algorithms, such as genetic algorithms or Monte Carlo simulations, to explore the optimal planting combination under different market conditions. This embodiment optimizes the mathematical model of Example 1 and obtains a crop planting strategy model from 2024 to 2030. The specific modeling process is as follows:

[0150] 1. Decision variables

[0151] Define the planting area of ​​crop i on plot j in year t, season s as X i,t,s,j , and the binary decision variable b for crop planting i,t,s,j At the same time, the actual sales volume of crop i in year t, season s, plot j is defined as z i,t,s,j .

[0152] 2. Objective function The goal of this embodiment is to maximize the annual net planting income from 2024 to 2030. The objective function expression is:

[0153]

[0154] Among them, p i,t represents the sales price of crop i in year t, c i,t represents the planting cost, z i,t,s,j is the actual sales volume of the crop.

[0155] 3. Constraints

[0156] 1. Per-acre yield fluctuation: Considering that crop yield fluctuates every year due to climate change, the yield fluctuation range is defined as ±10%:

[0157] y i,t =y i ·(1+δi,t ),δ i,t ∈[-0.1,0.1]

[0158] 2. Expected sales volume fluctuations: The expected sales volume of various crops may fluctuate by ±5% each year. For wheat and corn, the expected annual sales growth rate is between 5%-10%:

[0159] S i,t =S i ·(1+ε i,t ),ε i,t ∈[-0.05,0.05]

[0160] 3. Planting cost growth: The planting cost increases by 5% on average each year, and the calculation formula is:

[0161] c i,t =c i (1+0.05) t-2023

[0162] 4. Sales price fluctuations: The prices of grain crops remain stable; the prices of vegetable crops increase by 5% on average each year; the prices of edible fungi decrease by 1%-5% each year:

[0163]

[0164] 5. Actual sales volume restrictions: The actual sales volume of crops is subject to the restrictions on their planting area and yield. The constraints are:

[0165] z i,t,s,j ≤x i,t,s,j ·y i,t

[0166] By modeling the above decision variables, objective functions and constraints, the model in Example 1 is optimized, and the obtained model can be solved by random simulation or robust optimization methods.

[0167] like Figure 8 As shown, this embodiment uses a Monte Carlo simulation model to optimize and solve the above model.

[0168] Example 3

[0169] Different crops have certain substitutability and complementarity, as well as their correlation with sales price and planting cost. It is possible to use methods such as multi-objective optimization to balance yield, cost and market adaptability; at the same time, appropriate correlation analysis is used to identify the optimal strategy for different crop combinations. In addition, we can further optimize the model in Example 2 by simulating planting scenarios under different market and climate conditions to ensure the feasibility and flexibility of long-term planting plans.

[0170] Decision variables:

[0171] X i,t,s,j : The planting area of ​​the crop in the tth year, the sth season, and the plot j (mu);

[0172] b i,t,s,j : a binary variable (valued as 0 or 1) indicating whether the crop was planted in year t, season s, and plot j;

[0173] z i,t,s,j : The actual sales volume of the crop in year t, season s, and plot j (jin).

[0174] Objective function:

[0175] The goal is to maximize the total net profit during the entire planting cycle (2024-2030), the formula is as follows:

[0176]

[0177] where p i，t is the sales price of the crop in year t, which may be affected by the yield; c i，t is the planting cost of crop i in year t.

[0178] Constraints:

[0179] 1. Substitution constraints: Some crops cannot be grown on the same land at the same time due to conflicts in resources or planting cycles. The constraints in the model are:

[0180]

[0181] Among them, i and k are crops that cannot coexist, indicating that these crops cannot be planted at the same time in the same season on the same plot of land.

[0182] 2. Complementarity constraint: Some crops (such as beans) have the effect of improving soil, so each plot of land needs to be planted with complementary crops at least once in three years. The constraint is:

[0183]

[0184] 3. Sales volume-price correlation constraint: When the planting volume exceeds the market demand, the crop price will fall. Therefore, the sales price of crop i in year t should be related to the actual sales volume:

[0185]

[0186] Among them, α is the coefficient of price reduction, D i,t is the expected sales volume of crop i in year t.

[0187] 4. Actual sales volume restrictions: Actual sales volume is subject to the restrictions of planting area and per-acre yield, expressed as:

[0188]

[0189] In order to further optimize the planting plan, we introduced the particle swarm optimization algorithm (PSO) based on the Monte Carlo simulation. PSO is a heuristic optimization algorithm that gradually approaches the optimal solution by simulating the dynamic behavior of particles in the solution space. Each particle represents a planting plan, and its fitness is determined by the planting benefit. The position and speed update rules of PSO are as follows:

[0190]

[0191] vi,t,s,j is the velocity of the particle, xi,t,s,j is the position of the particle (i.e., the planting area);

[0192] ω is the inertia coefficient, c1, c2 are learning factors, r1, r2 are random numbers;

[0193] p best is the historical optimal solution of the particle, g best It is the global optimal solution.

[0194] In this embodiment, the randomness of yield and price is processed by Monte Carlo simulation, and particle swarm optimization is used to find the optimal planting plan. The specific process is as follows:

[0195] 1. Initialize the particle swarm, where each particle represents a randomly generated planting plan;

[0196] 2. Perform Monte Carlo simulation on each particle and calculate its fitness (net benefit);

[0197] 3. Update the speed and position of particles, gradually approaching the optimal planting plan;

[0198] 4. Repeat the iteration until the convergence condition is reached (such as the number of iterations or fitness convergence);

[0199] 5. Output the optimal solution, that is, the planting plan with the greatest profit.

[0200] By combining Monte Carlo simulation and particle swarm optimization, we are able to effectively deal with uncertainty and find the best solution in complex planting decisions.

[0201] The algorithm model provided by the present invention can be applied to different regions and crop types, and has the potential to be extended to a wider agricultural environment, providing decision support for policy makers in resource allocation and agricultural planning, and providing scientific planting guidance for local farmers; the model provided by the present invention has the following advantages:

[0202] (1) Advanced Algorithm: It adopts a variety of objective optimization hybrid algorithms such as simulated annealing algorithm to solve the problem. The algorithm has excellent global search capability, can effectively avoid local optimal traps, and improves the quality of the solution.

[0203] (2) Strong adaptability: The model takes into account future uncertainties, such as fluctuations in sales volume, output, cost, and price. By dynamically adjusting the planting strategy, the model’s adaptability and advantages in chaotic systems are enhanced.

[0204] (3) Flexibility: The model allows simulation of different market scenarios and provides the ability to make timely adjustments in an uncertain environment to respond to changes in the market and environment.

Claims

1. An agricultural planting decision-making method based on simulated annealing algorithm, characterized in that: The types, areas, time and plots of crops planted on different types of cultivated land are modeled to obtain the following mathematical model: R1 is the total profit without considering dumping, and R2 is the total profit when selling at half price considering dumping. The simulated annealing algorithm is used to solve R1 and R2 in the above model to obtain the planting decision method that maximizes economic benefits.

2. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1, characterized in that: The mathematical model is subject to the following constraints when used: Flat dry land, terraced fields, and hillside land can only grow a single crop of grain (except rice) each year: Plot j can only grow one crop i every year during all time periods t: And plot j cannot grow rice within any time t: b Rice,t,1,j ＝0, 3. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1, characterized in that: The mathematical model is subject to the following constraints when used: Irrigated land can be used to grow rice in one season or vegetables in two seasons: In time t, irrigated land j can be planted with either rice or vegetables, but not both: Only specific crops i can be planted in the second season of irrigated land j and all time periods t:

4. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1, characterized in that: The mathematical model is used with the following constraints: Chinese cabbage, white radish and red radish can only be planted in the second season of irrigated land; specific vegetable i is prohibited from being planted in the first season t of irrigated land j: b i,t,1，j ＝0, 5. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1 is characterized in that: The mathematical model meets the following constraints when used: ordinary greenhouses have two seasons a year, with multiple vegetables in the first season and only edible fungi in the second season; in ordinary greenhouse j, one vegetable i must be planted in the first season: In ordinary greenhouses, edible fungi must be planted in the second season: b Mushroom,t,2,j ＝1, 6. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1, characterized in that: The mathematical model meets the following constraints when used: There are two vegetable seasons in the smart greenhouse every year, excluding specific vegetables: In the smart greenhouse j, one vegetable i must be planted in each season: In the smart greenhouse j, certain crops cannot be planted i: b i，t,s，j ＝0, 7. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1 is characterized in that: The mathematical model is used to meet the following constraints: the actual sales volume cannot exceed the planting area multiplied by the yield or the maximum sales volume: Sales volume z of crop i at time t, season s and plot j i，t，s，j The maximum sales volume S of the crop cannot be exceeded i : With i，t，s，j ≤S i , Sales volume z of crop i at time t, season s and plot j i，t，s，j Cannot exceed the planting area x of the crop on plot j i，t，s，j and unit area yield y i The product of: z i，t，s，j ≤x i，t，s，j ·y i , 8. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 1, characterized in that: The mathematical model meets the following constraints when used: the planting area x of crop i at time t, season s and plot j i，t，s，j Cannot exceed the total area A of plot j j Multiply by the decision variable b i，t，s，j , where the decision variable b i，t，s，j Indicates whether to choose to plant crops at this plot and time: x i，t,s，j ≤A j ·b i，t，s，j , Planting decision variable b of crop i in plot j at time t and t+1 i，t，s，j and b i，t+1，s，j The sum of cannot exceed 1, that is, the same crop cannot be planted on the same plot for two consecutive years: b i，t，s，j +b i，t+1,s，j ≤1, 9. An agricultural planting decision-making method based on simulated annealing algorithm according to any one of claims 1 to 8, characterized in that: The total profit of cultivated land planting in the next T years is optimized and modeled: The model complies with the following constraints: y i，t =y i ·(1+d i，t ),d i，t ∈[-0.1, 0.1] S i，t =S i ·(1+∈ i，t ),∈ i，t ∈[-0.05,0.05] c i，t =c i ·(1+0.05) t-2023 ξ i，t ∈[0.01,0.05]z i，t，s，j ≤x i，t，s，j ·y i，t The Monte Carlo simulation model is used to optimize and solve the above model under the above constraints.

10. The agricultural planting decision-making method based on simulated annealing algorithm according to claim 9, characterized in that: The following constraints are added to the mathematical model: b i，t,s，j +b k，t，s，j ≤1, Every three years, z i，t，s，j ≤x i，t,s，j ·y i，t , The particle swarm optimization algorithm is used to optimize and solve the model under the above constraints.

Images