Agricultural morphological model construction method

Abstract

The invention discloses an agricultural morphological model construction method, which comprises the following steps: S1, constructing a multi-level biologic chain model, and designing a substance and energy transfer path between modules; s2, collecting multi-modal data, and generating a dynamic parameter combination in combination with a Bayesian optimization algorithm; s3, based on an ivy optimization algorithm, optimizing a cyclic utilization path of wastes; s4, dynamically adjusting a resource allocation proportion through a multi-agent reinforcement learning technology; s5, optimizing an energy and substance flow path by adopting a dynamic biologic chain modeling and metabolic flow simulation technology; s6, purifying and recycling the water body, and planning a water resource distribution path based on an ivy optimization algorithm; s7, based on the time sequence prediction model, dynamically predicting key indexes and feeding back regulation and control; and S8, constructing a modular ecological product processing area, optimizing a processing path and displaying a result. According to the method, efficient cyclic utilization and dynamic optimization regulation of agricultural resources are realized, and ecological sustainable development is promoted.

Description

Technical Field

[0001] The present invention relates to the field of agricultural technology, and in particular to a method for constructing an agricultural morphological model. Background Art

[0002] With the rapid development of global agricultural modernization, traditional agricultural production models are gradually transforming towards intelligence, precision and sustainability. As a complex material and energy flow system, the agricultural ecosystem includes multiple functional modules such as planting areas, breeding areas, and resource recycling areas. At this stage, agricultural production has increasingly higher requirements for resource utilization efficiency, especially in terms of water resources, soil nutrients, energy consumption, and waste treatment. How to achieve efficient recycling, dynamic optimization configuration, and precise regulation has become a key issue for the sustainable development of agriculture.

[0003] In the existing agricultural ecosystem, the allocation and recycling of water resources is a core link. However, traditional agricultural water resource management mainly relies on experience and timed and quantitative irrigation methods, lacking the ability to monitor and regulate water quality changes, crop demand and meteorological conditions in real time. This extensive water resource management model easily leads to problems such as water resource waste, soil salinization and water pollution, making it difficult to achieve efficient use of resources. In addition, the existing agricultural waste treatment system often regards crop residues, livestock manure and breeding waste as useless burdens, with a single treatment method, and fails to effectively resource and recycle these wastes, resulting in widespread environmental pollution and resource waste.

[0004] For the problem of dynamic resource allocation in the agricultural production process, the existing technology lacks systematic modeling and optimization methods. The material and energy flow paths in the agricultural ecosystem are complex, and the resource flow between different modules (such as mulberry planting areas, fruit planting areas, aquaculture areas, and edible fungus planting areas) shows nonlinear and time-varying characteristics. Existing resource regulation methods mainly rely on static planning, which makes it difficult to make dynamic adjustments according to actual needs and real-time changes in system status during agricultural production. This static allocation strategy not only leads to inefficient resource allocation, but also cannot adapt to the real-time dynamic needs in the agricultural ecosystem, thereby limiting the improvement of agricultural production efficiency.

[0005] In addition, in the existing agricultural system, some areas use sensor technology to monitor parameters such as soil moisture, light intensity, and meteorological data, but lack effective data fusion and optimization mechanisms. Existing technologies often rely on a single data source to adjust agricultural production parameters, ignoring the comprehensive analysis and utilization of multimodal data in the agricultural production process (such as water quality parameters, crop growth status, and resource flow data, etc.), resulting in low accuracy of system decision-making. In addition, the existing agricultural regulation system has also failed to effectively combine time series prediction models for dynamic prediction, and cannot predict crop demand and environmental changes in advance, thus affecting the precise control of agricultural production parameters (such as irrigation, fertilization, and feeding amount).

[0006] In terms of waste recycling, existing technologies have failed to establish an efficient dynamic optimization path. Although waste within agricultural ecosystems (such as fish pond mud, silkworm excrement, mushroom residue, etc.) has high resource utilization potential, traditional treatment methods lack intelligent path planning and dynamic adjustment capabilities, and fail to achieve efficient recycling of resources through algorithm optimization. In addition, existing waste resource technologies often focus on the treatment of a single module and fail to effectively consider the overall material and energy flow of the ecosystem, making it difficult to achieve the coordinated development of agricultural production and ecological protection.

[0007] Water purification and resource redistribution are another important link in modern agricultural production. Traditional water purification technologies mainly rely on physical or chemical methods, which are costly, energy-intensive, and prone to secondary pollution. In the existing water purification system, there is a lack of path optimization methods based on intelligent algorithms, and aquatic plants and biological filtration technology are not fully utilized to achieve efficient purification. In addition, the existing water resource recycling paths are mostly statically planned, and cannot be dynamically adjusted according to the real-time changes in agricultural irrigation needs and aquaculture recycling needs, resulting in low water resource utilization efficiency.

[0008] Therefore, how to provide a method for constructing an agricultural morphology model is an urgent problem that technicians in this field need to solve. Summary of the invention

[0009] One purpose of the present invention is to propose a method for constructing an agricultural morphology model. The present invention makes full use of a multimodal sensor system, an ivy optimization algorithm, and aquatic plant purification technology to construct a dynamic model of material and energy flow in an agricultural ecosystem, and combines an intelligent feedback control system to achieve dynamic optimization of resource allocation, waste recycling, and water resource purification paths. The present invention describes in detail the resource transfer path optimization and control algorithm between modules in agricultural production, and has the advantages of high resource utilization efficiency, strong dynamic adaptability, complete ecological cycle, and strong system sustainability.

[0010] A method for constructing an agricultural morphology model according to an embodiment of the present invention comprises the following steps:

[0011] S1. Construct a multi-level food chain model, divide the agricultural system into mulberry planting area, fruit tree planting area, aquaculture area, bee breeding area, edible fungus planting area and diversified product processing area, design the material and energy transfer paths between modules, and form the initial closed-loop structure of the agricultural ecosystem;

[0012] S2. Use multimodal sensors to collect soil moisture, light intensity, water quality parameters and meteorological conditions data, and combine with Bayesian optimization algorithm to generate dynamic parameter combinations for agricultural production, including irrigation time, fertilizer application amount and feeding frequency;

[0013] S3. Based on the dynamic parameter combination of agricultural production and combined with the ivy optimization algorithm, the recycling path of waste is optimized to convert fish pond mud, silkworm excrement and fungus residue into organic fertilizer and biogas;

[0014] S4. Based on the optimized waste recycling path and dynamic parameter combination of agricultural production, the ecological module is modeled as an intelligent agent through multi-agent reinforcement learning technology, and the resource allocation ratio is dynamically adjusted using a collaborative learning algorithm;

[0015] S5. Based on the resource allocation results, dynamic food chain modeling and metabolic flow simulation technology are used to analyze the flow paths of substances and energy in the agricultural ecosystem and dynamically adjust the energy utilization ratio;

[0016] S6. Use the adjusted energy utilization ratio and resource transfer parameters for water purification and recycling equipment, use aquatic plants to absorb nitrogen and phosphorus pollutants discharged from the aquaculture area, combine the ivy optimization algorithm to plan the redistribution path of purified water, and generate dynamic control parameters for water resource recycling;

[0017] S7. Based on the generated dynamic control parameters and multimodal data of water resource recycling, the key indicators of the agricultural system are dynamically predicted based on the time series prediction model, and the agricultural production parameters are optimized in real time through the feedback control system;

[0018] S8. Build a modular ecological product processing area, combine dynamic scheduling algorithms to optimize the processing paths of crops and breeding products, produce high value-added ecological products, and use augmented reality technology to display them in the ecological leisure area.

[0019] Optionally, the mulberry planting area provides mulberry leaves to silkworms and absorbs nitrogen and phosphorus discharged from the aquaculture area, and the mulberry branches are used for edible fungus cultivation.

[0020] Optionally, the data collected by multimodal sensors in the mulberry planting area and the crop planting area, including soil moisture, light intensity and water quality parameters, are combined with dynamic parameters generated by a Bayesian optimization algorithm to optimize irrigation time, fertilizer amount and feeding frequency, and intercropping leguminous plants is used during the irrigation and fertilization process.

[0021] Optionally, the S3 specifically includes:

[0022] S31. Establish a waste cycle path optimization model, and model the waste paths in fish pond mud, silkworm excrement and fungus residue and agricultural ecosystem as a multi-objective optimization problem;

[0023] S32. Collect attribute data of waste, including moisture content, nutrient content and pollutant composition, and classify the waste based on a multimodal classification algorithm into three categories: suitable for direct conversion into organic fertilizer, used for biogas fermentation and requiring further treatment;

[0024] S33, construct a waste transfer path planning matrix, and calculate the total path cost based on the matrix:

[0025]

[0026] Where n is the total number of ecological modules, C ij represents the unit transfer cost of waste from module i to module j, P ij is the waste transfer path planning matrix, C total is the total cost function for calculating the path;

[0027] S34. Use the Ivy optimization algorithm to dynamically optimize the waste recycling path:

[0028]

[0029] Where minZ is the objective function, n represents the total number of modules involved in waste transfer in the agricultural ecosystem, i and j represent the index numbers of the modules in the agricultural ecosystem, and T ij represents the transfer time of waste from module i to module j, C ij represents unit cost, E ij represents the unit environmental impact, β 1 , β 2 , β 3 To optimize the weight coefficient, satisfy β 1 +β 2 +β 3 =1;

[0030] S35. Process the optimization path results in a resource-based manner. The fish pond sludge is directly converted into organic fertilizer and applied to mulberry and fruit tree planting areas. The silkworm excrement is mixed with the fungus residue and then fermented to generate biogas for energy supply in the agricultural system. The biogas residue is further used as organic fertilizer and applied to crop planting areas and aquaculture areas.

[0031] S36, updating waste attribute parameters based on real-time sensor data, and reallocating the flow direction and proportion of waste resource utilization by dynamically adjusting the path planning matrix and the optimization weight of the objective function;

[0032] S37. In the edible fungus growing area, the environmental conditions of the edible fungus growing area, including temperature, humidity and water supply, are automatically adjusted in combination with real-time sensor data.

[0033] Optionally, the S4 specifically includes:

[0034] S41. Model each module in the agricultural ecosystem as an intelligent agent and define the state, behavior and reward value of each intelligent agent;

[0035] S42. Use multi-agent reinforcement learning technology to establish a collaborative optimization mechanism for the state and behavior of the agents, build the system's resource allocation objective function, and generate the system's preliminary resource allocation strategy with the goal of maximizing overall resource utilization efficiency:

[0036]

[0037] Among them, the reward value R i is the benefit after resource allocation of the module, γ is the discount factor, G is the overall resource allocation benefit of the agricultural ecosystem, and n is the number of agents in the agricultural ecosystem;

[0038] S43. According to the generated preliminary resource allocation strategy, the state and behavior of the agent are optimized, the resource allocation strategy of each agent is dynamically adjusted, and the resource allocation ratio between modules is updated:

[0039]

[0040] Where E is the expected value, is the optimal strategy of module i in the current state, S i is the current state of module i, A i is the current behavior of module i, argmax represents the optimal strategy

[0041] S44, dynamically adjust the resource allocation ratio between modules according to the optimal behavior strategy of the intelligent agent, and update the state and behavior of the intelligent agent in real time, and re-optimize the resource allocation strategy in combination with the resource demand changes between modules and the real-time data collected by sensors;

[0042] S45. Based on the optimized resource allocation strategy, dynamically adjust the resource flow paths of each module in the ecosystem;

[0043] S46. Based on the optimization results of resource flow paths, the resource allocation in aquaculture areas is further optimized by adopting a multi-layer farming structure, including bottom fish farming, middle-layer shrimp farming and upper-layer plankton farming, and the optimization results are fed back to the updating process of resource allocation strategy.

[0044] Optionally, the S5 specifically includes:

[0045] S51. Based on the dynamic resource flow paths and distribution ratios of the agricultural ecosystem, define the material and energy flow states of each module in the agricultural ecosystem, where the bee breeding area is adjacent to the mulberry planting area and the crop planting area, and the yield of crops and mulberry trees is increased through bee pollination, and honey and beeswax are collected;

[0046] S52. Construct a dynamic food chain model of the agricultural ecosystem. Based on the resource flow status between modules, analyze the resource transfer path between different modules and clarify the complete flow link of resources from input module to output module.

[0047] S53. Use metabolic flow simulation technology to analyze the material and energy flow paths between modules:

[0048]

[0049] Where E is the global energy allocation efficiency of the system, F ij is the resource flow between modules, η ij is the energy utilization efficiency from module i to module j, L i is the output resource amount of module i, and n is the total number of modules;

[0050] S54, dynamically adjusting the resource allocation ratio between modules according to the optimization result of the resource flow path between modules, generating a new resource flow path, and optimizing the transfer efficiency and global utilization rate of resources;

[0051] S55, combining the adjusted resource flow path, generating an optimized material flow network between modules, and updating the resource flow status of the module;

[0052] S56. Output the optimized dynamic food chain model, including the optimized resource flow path between modules and the updated resource flow status of each module.

[0053] Optionally, the S6 specifically includes:

[0054] S61. Based on the optimization results of the resource flow paths between modules, water quality parameters of the aquaculture area, including nitrogen content, phosphorus content and dissolved oxygen concentration, are collected, and water purification demand data is generated using sensors;

[0055] S62. Construct a water purification model, combine water quality parameters with agricultural irrigation needs and aquaculture recycling needs, and define water purification targets, including pollutant removal efficiency, purification time, and purified water volume;

[0056] S63, using aquatic plant purification system, selecting duckweed to absorb and purify nitrogen and phosphorus pollutants, while combining biological filtration technology to remove organic pollutants in the water to generate preliminary purified water;

[0057] S64. Based on the Ivy optimization algorithm, the water purification path is dynamically optimized. Combined with water quality parameters, purification goals and resource allocation ratios, the redistribution path of purified water is optimized to generate a water resource transfer matrix between modules:

[0058]

[0059] Among them, P ij represents the water allocation ratio from module i to module j, U represents the overall benefit of water resources after allocation among modules, and Q ij represents the corresponding amount of water, n is the total number of ecological modules, and n is the total number of modules in the agricultural ecosystem;

[0060] S65. Redistribute the purified water resources according to the optimized path, with a portion used for agricultural irrigation modules and the other portion returned to the aquaculture area, so as to achieve efficient recycling of water resources;

[0061] S66. Update the water purification model in real time, dynamically adjust the purification target and resource allocation path according to the changes in water quality parameters collected by the sensor and the module demand data, and generate an updated purification path and water resource allocation plan.

[0062] Optionally, the S7 specifically includes:

[0063] S71. Based on the collected multimodal data, the time series prediction model is used to dynamically predict the key indicators in the agricultural system;

[0064] S72. Based on dynamic prediction results and combined with feedback control system, key parameters in agricultural production are optimized in real time;

[0065] S73. Through the feedback control system, dynamically adjust production factors according to the predicted results to optimize irrigation, fertilization and feeding amount:

[0066]

[0067] Among them, Y i represents the actual value of the i-th key indicator, is the predicted value of the i-th key indicator, w i is the weight coefficient of the corresponding indicator, Z is the objective function of minimizing the overall error, n represents the total number of key indicators predicted in the agricultural system, and i represents the index number of the key indicator in the agricultural system;

[0068] S74. Based on prediction and regulation, dynamic adjustment instructions are generated, and optimization operations are performed through control devices.

[0069] The beneficial effects of the present invention are:

[0070] The present invention realizes efficient dynamic configuration and recycling of resources in agricultural ecosystems through advanced technologies such as multi-agent reinforcement learning, dynamic food chain modeling, time series prediction and ivy optimization algorithm. Compared with the prior art, the present invention can comprehensively collect multimodal data in agricultural production, including key indicators such as soil moisture, temperature, light intensity, water quality parameters and crop growth status, and accurately predict future agricultural production needs through time series prediction models, and provide scientific basis for the dynamic regulation of agricultural parameters such as irrigation time, fertilizer application and breeding feeding frequency. At the same time, using multi-agent reinforcement learning technology, each module in the ecosystem (such as mulberry planting area, fruit tree planting area, aquaculture area, bee breeding area and edible fungus planting area) is modeled as an intelligent body, through collaborative learning and dynamic resource allocation strategy, optimize resource transfer path, improve the utilization efficiency of material and energy in the system, and form a closed loop. In addition, based on the ivy optimization algorithm, the present invention realizes the dynamic optimization of water purification path, purifies nitrogen and phosphorus pollutants discharged in the breeding area through aquatic plant purification system and biological filtration technology, ensures efficient recycling of water resources, and reduces water pollution.

[0071] The present invention effectively solves the problems of static resource allocation, single waste treatment path, insufficient dynamic feedback and waste of water resources in the prior art, and improves the overall efficiency and ecological sustainability of agricultural production. In terms of waste recycling, the present invention optimizes the resource utilization path of agricultural wastes such as fish pond sludge, silkworm excrement and fungus residue, and rationally converts waste into organic fertilizer and biogas through metabolic flow simulation and path planning, and makes resource circulation more efficient by dynamically adjusting the path allocation results. In terms of ecological module collaboration, the present invention introduces a bee pollination mechanism, and increases the yield of crops and mulberry trees through the adjacent layout of bee breeding areas, mulberry planting areas and crop planting areas, while collecting honey and beeswax, further improving the economic benefits of the agricultural system. In terms of water resource management, the present invention realizes efficient recycling of water resources by dynamically optimizing the water resource allocation path through real-time collection of water quality parameters and combining purification needs with agricultural irrigation goals.

[0072] The present invention has the beneficial effects of high resource utilization efficiency, strong dynamic adaptability, complete ecological cycle, environmental friendliness and significant economic benefits. Through multi-level resource regulation and dynamic feedback optimization, the present invention constructs a highly intelligent and sustainable agricultural ecosystem, providing a new solution for efficient production and ecological protection of modern agriculture. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0074] Figure 1 A flowchart of a method for constructing an agricultural morphology model proposed by the present invention;

[0075] Figure 2 This is a schematic diagram of water resource recycling path optimization for an agricultural morphology model construction method proposed in the present invention. DETAILED DESCRIPTION

[0076] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0077] refer to Figure 1-2 , a method for constructing an agricultural morphology model, comprising the following steps:

[0078] S1. Construct a multi-level food chain model, divide the agricultural system into mulberry planting area, fruit tree planting area, aquaculture area, bee breeding area, edible fungus planting area and diversified product processing area, design the material and energy transfer paths between modules, and form the initial closed-loop structure of the agricultural ecosystem;

[0079] S2. Use multimodal sensors to collect soil moisture, light intensity, water quality parameters and meteorological conditions data, and combine with Bayesian optimization algorithm to generate dynamic parameter combinations for agricultural production, including irrigation time, fertilizer application amount and feeding frequency;

[0080] S3. Based on the dynamic parameter combination of agricultural production and combined with the ivy optimization algorithm, the recycling path of waste is optimized to convert fish pond mud, silkworm excrement and fungus residue into organic fertilizer and biogas;

[0081] S4. Based on the optimized waste recycling path and dynamic parameter combination of agricultural production, the ecological module is modeled as an intelligent agent through multi-agent reinforcement learning technology, and the resource allocation ratio is dynamically adjusted using a collaborative learning algorithm;

[0082] S5. Based on the resource allocation results, dynamic food chain modeling and metabolic flow simulation technology are used to analyze the flow paths of substances and energy in the agricultural ecosystem and dynamically adjust the energy utilization ratio;

[0083] S6. Use the adjusted energy utilization ratio and resource transfer parameters for water purification and recycling equipment, use aquatic plants to absorb nitrogen and phosphorus pollutants discharged from the aquaculture area, combine the ivy optimization algorithm to plan the redistribution path of purified water, and generate dynamic control parameters for water resource recycling;

[0084] S7. Based on the generated dynamic control parameters and multimodal data of water resource recycling, the key indicators of the agricultural system are dynamically predicted based on the time series prediction model, and the agricultural production parameters are optimized in real time through the feedback control system;

[0085] S8. Build a modular ecological product processing area, combine dynamic scheduling algorithms to optimize the processing paths of crops and breeding products, produce high value-added ecological products, and use augmented reality technology to display them in the ecological leisure area.

[0086] In this embodiment, the mulberry planting area provides mulberry leaves to silkworms and absorbs nitrogen and phosphorus discharged from the aquaculture area, and mulberry branches are used for edible fungus cultivation.

[0087] In this embodiment, the data collected by multimodal sensors in the mulberry planting area and the crop planting area, including soil moisture, light intensity and water quality parameters, are combined with dynamic parameters generated by the Bayesian optimization algorithm to optimize irrigation time, fertilizer amount and feeding frequency, and intercropping leguminous plants is used during the irrigation and fertilization process.

[0088] In this implementation, S3 specifically includes:

[0089] S31. Establish a waste cycle path optimization model, and model the waste paths in fish pond mud, silkworm excrement and fungus residue and agricultural ecosystem as a multi-objective optimization problem;

[0090] S32. Collect attribute data of waste, including moisture content, nutrient content and pollutant composition, and classify the waste based on a multimodal classification algorithm into three categories: suitable for direct conversion into organic fertilizer, used for biogas fermentation and requiring further treatment;

[0091] S33, construct a waste transfer path planning matrix, and calculate the total path cost based on the matrix:

[0092]

[0093] Where n is the total number of ecological modules, C ij represents the unit transfer cost of waste from module i to module j, P ijis the waste transfer path planning matrix, C total is the total cost function for calculating the path;

[0094] S34. Use the Ivy optimization algorithm to dynamically optimize the waste recycling path:

[0095]

[0096] Where minZ is the objective function, n represents the total number of modules involved in waste transfer in the agricultural ecosystem, i and j represent the index numbers of the modules in the agricultural ecosystem, and T ij represents the transfer time of waste from module i to module j, C ij represents unit cost, E ij represents the unit environmental impact, β 1 , β 2 , β 3 To optimize the weight coefficient, satisfy β 1 +β 2 +β 3 =1;

[0097] S35. Process the optimization path results in a resource-based manner. The fish pond sludge is directly converted into organic fertilizer and applied to mulberry and fruit tree planting areas. The silkworm excrement is mixed with the fungus residue and then fermented to generate biogas for energy supply in the agricultural system. The biogas residue is further used as organic fertilizer and applied to crop planting areas and aquaculture areas.

[0098] S36, updating waste attribute parameters based on real-time sensor data, and reallocating the flow direction and proportion of waste resource utilization by dynamically adjusting the path planning matrix and the optimization weight of the objective function;

[0099] S37. In the edible fungus growing area, the environmental conditions of the edible fungus growing area, including temperature, humidity and water supply, are automatically adjusted in combination with real-time sensor data.

[0100] In this implementation, S4 specifically includes:

[0101] S41. Model each module in the agricultural ecosystem as an intelligent agent and define the state, behavior and reward value of each intelligent agent;

[0102] S42. Use multi-agent reinforcement learning technology to establish a collaborative optimization mechanism for the state and behavior of the agents, build the system's resource allocation objective function, and generate the system's preliminary resource allocation strategy with the goal of maximizing overall resource utilization efficiency:

[0103]

[0104] Among them, the reward value R iis the benefit after resource allocation of the module, γ is the discount factor, G is the overall resource allocation benefit of the agricultural ecosystem, and n is the number of agents in the agricultural ecosystem;

[0105] S43. According to the generated preliminary resource allocation strategy, the state and behavior of the agent are optimized, the resource allocation strategy of each agent is dynamically adjusted, and the resource allocation ratio between modules is updated:

[0106]

[0107] Where E is the expected value, is the optimal strategy of module i in the current state, S i is the current state of module i, A i is the current behavior of module i, argmax represents the optimal strategy

[0108] S44, dynamically adjust the resource allocation ratio between modules according to the optimal behavior strategy of the intelligent agent, and update the state and behavior of the intelligent agent in real time, and re-optimize the resource allocation strategy in combination with the resource demand changes between modules and the real-time data collected by sensors;

[0109] S45. Based on the optimized resource allocation strategy, dynamically adjust the resource flow paths of each module in the ecosystem;

[0110] S46. Based on the optimization results of resource flow paths, the resource allocation in aquaculture areas is further optimized by adopting a multi-layer farming structure, including bottom fish farming, middle-layer shrimp farming and upper-layer plankton farming, and the optimization results are fed back to the updating process of resource allocation strategy.

[0111] In this implementation manner, S5 specifically includes:

[0112] S51. Based on the dynamic resource flow paths and distribution ratios of the agricultural ecosystem, define the material and energy flow states of each module in the agricultural ecosystem, where the bee breeding area is adjacent to the mulberry planting area and the crop planting area, and the yield of crops and mulberry trees is increased through bee pollination, and honey and beeswax are collected;

[0113] S52. Construct a dynamic food chain model of the agricultural ecosystem. Based on the resource flow status between modules, analyze the resource transfer path between different modules and clarify the complete flow link of resources from input module to output module.

[0114] S53. Use metabolic flow simulation technology to analyze the material and energy flow paths between modules:

[0115]

[0116] Where E is the global energy allocation efficiency of the system, F ij is the resource flow between modules, η ij is the energy utilization efficiency from module i to module j, L i is the output resource amount of module i, and n is the total number of modules;

[0117] S54, dynamically adjusting the resource allocation ratio between modules according to the optimization result of the resource flow path between modules, generating a new resource flow path, and optimizing the transfer efficiency and global utilization rate of resources;

[0118] S55, combining the adjusted resource flow path, generating an optimized material flow network between modules, and updating the resource flow status of the module;

[0119] S56. Output the optimized dynamic food chain model, including the optimized resource flow path between modules and the updated resource flow status of each module.

[0120] In this implementation manner, S6 specifically includes:

[0121] S61. Based on the optimization results of the resource flow paths between modules, water quality parameters of the aquaculture area, including nitrogen content, phosphorus content and dissolved oxygen concentration, are collected, and water purification demand data is generated using sensors;

[0122] S62. Construct a water purification model, combine water quality parameters with agricultural irrigation needs and aquaculture recycling needs, and define water purification targets, including pollutant removal efficiency, purification time, and purified water volume;

[0123] S63, using aquatic plant purification system, selecting duckweed to absorb and purify nitrogen and phosphorus pollutants, while combining biological filtration technology to remove organic pollutants in the water to generate preliminary purified water;

[0124] S64. Based on the Ivy optimization algorithm, the water purification path is dynamically optimized. Combined with water quality parameters, purification goals and resource allocation ratios, the redistribution path of purified water is optimized to generate a water resource transfer matrix between modules:

[0125]

[0126] Among them, P ij represents the water allocation ratio from module i to module j, U represents the overall benefit of water resources after allocation among modules, and Q ij represents the corresponding amount of water, n is the total number of ecological modules, and n is the total number of modules in the agricultural ecosystem;

[0127] S65. Redistribute the purified water resources according to the optimized path, with a portion used for agricultural irrigation modules and the other portion returned to the aquaculture area, so as to achieve efficient recycling of water resources;

[0128] S66. Update the water purification model in real time, dynamically adjust the purification target and resource allocation path according to the changes in water quality parameters collected by the sensor and the module demand data, and generate an updated purification path and water resource allocation plan.

[0129] In this implementation manner, the S7 specifically includes:

[0130] S71. Based on the collected multimodal data, the time series prediction model is used to dynamically predict the key indicators in the agricultural system;

[0131] S72. Based on dynamic prediction results and combined with feedback control system, key parameters in agricultural production are optimized in real time;

[0132] S73. Through the feedback control system, dynamically adjust production factors according to the predicted results to optimize irrigation, fertilization and feeding amount:

[0133]

[0134] Among them, Y i represents the actual value of the i-th key indicator, is the predicted value of the i-th key indicator, w i is the weight coefficient of the corresponding indicator, Z is the objective function of minimizing the overall error, n represents the total number of key indicators predicted in the agricultural system, and i represents the index number of the key indicator in the agricultural system;

[0135] S74. Based on prediction and regulation, dynamic adjustment instructions are generated, and optimization operations are performed through control devices.

[0136] Embodiment 1:

[0137] In order to verify the feasibility of the present invention in implementation, the present invention is applied to a certain agricultural demonstration park, covering an area of ​​about 300 mu. The park has a mild and humid climate and fertile soil, which is suitable for the growth of a variety of crops and cash crops. At present, the park faces problems such as low resource utilization efficiency, insufficient agricultural waste treatment, and unreasonable water resource allocation. It is urgent to achieve efficient recycling and dynamic optimization of agricultural resources through intelligent technical means.

[0138] The present invention models and optimizes the transfer paths of matter and energy in agricultural ecosystems based on technologies such as multi-agent reinforcement learning, time series prediction models, and dynamic food chain modeling. The experimental park has deployed a multimodal sensor system to collect data such as soil moisture, light intensity, temperature, meteorological conditions, water quality parameters, and crop growth status of the agricultural system in real time. Sensor data showed that soil moisture varies significantly in different regions. In some areas, excessive humidity leads to waste of water resources, while in some areas, excessive humidity affects crop growth. Pollutants such as nitrogen and phosphorus in the water discharged from aquaculture areas exceed the standard.

[0139] In the implementation process of the present invention, the multimodal sensor system is first used to collect various data of the agricultural system. For example, in the mulberry planting area and the crop planting area, the sensor monitors that the average soil moisture is 21.5%, which is lower than the optimum value of 28%; in the aquaculture area, the nitrogen content in the water reaches 6.3mg / L and the phosphorus content reaches 2.1mg / L, which affects the growth of aquatic organisms.

[0140] Based on these data, the time series prediction model is used to dynamically predict the key indicators of the agricultural system. The prediction results show that soil moisture will continue to drop by 3%-5% in the next week. If irrigation is not carried out in time, it will lead to a reduction in crop yields. In addition, the concentration of water pollutants will further increase by 10%. If it is not purified, it will affect the recycling of water resources.

[0141] Combined with the forecast data, the system optimizes agricultural production parameters in real time through the feedback control system. In the mulberry planting area and the crop planting area, the irrigation system was dynamically adjusted, the irrigation time was increased by 20 minutes, and the humidity was controlled within the most suitable range of 28%-30%. In the aquaculture area, the water purification system was started, and duckweed and biological filtration technology were used to absorb and purify nitrogen and phosphorus pollutants in the water. The purified water was redistributed to the agricultural irrigation system to achieve the recycling of water resources. At the same time, combined with the Ivy optimization algorithm, the distribution path of water resources was optimized, 65% of the purified water was distributed to the crop planting area, and 35% was returned to the aquaculture area, realizing the efficient use of water resources.

[0142] Table 1 Comparison table before and after dynamic optimization of water resources

[0143]

[0144] Table 2 Comparison of waste resource utilization

[0145]

[0146] Table 1 clearly shows the changes in key water resource indicators in the agricultural system before and after optimization. Before optimization, the soil moisture in the crop planting area was only 21.5%, far below the most suitable growth humidity for crops, which seriously affected the yield of crops. However, after dynamically optimizing the irrigation time and water volume through the intelligent monitoring and feedback control system, the soil moisture increased to 28.5%, meeting the growth needs of crops, and the humidity increased by 7.0%. At the same time, in the aquaculture area, the nitrogen content and phosphorus content of the water body dropped from 6.3mg / L and 2.1mg / L to 2.5mg / L and 0.8mg / L respectively, and the pollutant removal rate exceeded 60%, which significantly improved the water quality and reduced pollution to the environment. In addition, the water resource utilization efficiency increased from 70% to 92%, an increase of 22%, which fully demonstrated the efficiency and accuracy of the present invention in the dynamic regulation and recycling of water resources.

[0147] Table 2 shows in detail the resource utilization of various wastes in the agricultural ecosystem, including the treatment effects of aquaculture sludge, silkworm excrement and edible fungus residue. Before optimization, the utilization rate of these wastes was generally low, with only 30% of the sludge being recycled, 40% of the silkworm excrement, and 50% of the residue. Through the dynamic optimization of the waste recycling path of the present invention, combined with metabolic flow simulation technology and ivy optimization algorithm, these wastes have achieved efficient resource utilization. After optimization, the utilization rate of aquaculture sludge increased to 95%, and was converted into organic fertilizer for mulberry and fruit tree planting areas; the utilization rate of silkworm excrement reached 98%, and biogas was generated through biogas fermentation for agricultural energy supply, and biogas residue was further converted into organic fertilizer; the utilization rate of edible fungus residue increased to 92%, and was processed into breeding feed and organic fertilizer. This process not only reduces the accumulation and pollution of agricultural waste, but also realizes the efficient recycling of resources, and improves the ecological and economic benefits of the agricultural system.

[0148] In summary, the present invention realizes efficient utilization and circulation of agricultural ecosystem resources through dynamic regulation of water resources, resource utilization of waste and optimization of bee pollination mechanism. The efficiency of water resource utilization is significantly improved, and soil moisture and water pollution are effectively improved; agricultural waste is efficiently converted into resources, reducing environmental pollution; bee pollination effectively increases the yield of crops and mulberry trees, and brings additional economic benefits. These data fully demonstrate the feasibility and superiority of the present invention in agricultural resource management, which is efficient, eco-friendly and economically sustainable, and provides a scientific, reasonable and intelligent solution for modern agriculture.

[0149] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for constructing an agricultural morphology model, characterized in that: The steps include: S1. Construct a multi-level food chain model, divide the agricultural system into mulberry planting area, fruit tree planting area, aquaculture area, bee breeding area, edible fungus planting area and diversified product processing area, design the material and energy transfer paths between modules, and form the initial closed-loop structure of the agricultural ecosystem; S2. Use multimodal sensors to collect soil moisture, light intensity, water quality parameters and meteorological conditions data, and combine with Bayesian optimization algorithm to generate dynamic parameter combinations for agricultural production, including irrigation time, fertilizer application amount and feeding frequency; S3. Based on the dynamic parameter combination of agricultural production and combined with the ivy optimization algorithm, the recycling path of waste is optimized to convert fish pond mud, silkworm excrement and fungus residue into organic fertilizer and biogas; S4. Based on the optimized waste recycling path and dynamic parameter combination of agricultural production, the ecological module is modeled as an intelligent agent through multi-agent reinforcement learning technology, and the resource allocation ratio is dynamically adjusted using a collaborative learning algorithm; S5. Based on the resource allocation results, dynamic food chain modeling and metabolic flow simulation technology are used to analyze the flow paths of substances and energy in the agricultural ecosystem and dynamically adjust the energy utilization ratio; S6. Use the adjusted energy utilization ratio and resource transfer parameters for water purification and recycling equipment, use aquatic plants to absorb nitrogen and phosphorus pollutants discharged from the aquaculture area, combine the ivy optimization algorithm to plan the redistribution path of purified water, and generate dynamic control parameters for water resource recycling; S7. Based on the generated dynamic control parameters and multimodal data of water resource recycling, the key indicators of the agricultural system are dynamically predicted based on the time series prediction model, and the agricultural production parameters are optimized in real time through the feedback control system; S8. Build a modular ecological product processing area, combine dynamic scheduling algorithms to optimize the processing paths of crops and breeding products, produce high value-added ecological products, and use augmented reality technology to display them in the ecological leisure area.

2. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The mulberry planting area provides mulberry leaves for silkworms and absorbs nitrogen and phosphorus discharged from the aquaculture area, and the mulberry branches are used for edible fungus planting.

3. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The data collected by multimodal sensors in the mulberry planting area and the crop planting area, including soil moisture, light intensity and water quality parameters, are combined with dynamic parameters generated by a Bayesian optimization algorithm to optimize irrigation time, fertilizer amount and feeding frequency, and intercrop leguminous plants are used during irrigation and fertilization.

4. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The S3 specifically includes: S31. Establish a waste cycle path optimization model, and model the waste paths in fish pond mud, silkworm excrement and fungus residue and agricultural ecosystem as a multi-objective optimization problem; S32. Collect attribute data of waste, including moisture content, nutrient content and pollutant composition, and classify the waste based on a multimodal classification algorithm into three categories: suitable for direct conversion into organic fertilizer, used for biogas fermentation and requiring further treatment; S33, construct a waste transfer path planning matrix, and calculate the total path cost based on the matrix: Where n is the total number of ecological modules, C ij represents the unit transfer cost of waste from module i to module j, P ij is the waste transfer path planning matrix, C total is the total cost function for calculating the path; S34. Use the Ivy optimization algorithm to dynamically optimize the waste recycling path: Where minZ is the objective function, n represents the total number of modules involved in waste transfer in the agricultural ecosystem, i and j represent the index numbers of the modules in the agricultural ecosystem, and T ij represents the transfer time of waste from module i to module j, C ij represents unit cost, E ij represents the unit environmental impact, β1, β2, β3 are the optimization weight coefficients, satisfying β1+β2+β3=1; S35. Process the optimization path results in a resource-based manner. The fish pond sludge is directly converted into organic fertilizer and applied to mulberry and fruit tree planting areas. The silkworm excrement is mixed with the fungus residue and then fermented to generate biogas for energy supply in the agricultural system. The biogas residue is further used as organic fertilizer and applied to crop planting areas and aquaculture areas. S36, updating waste attribute parameters based on real-time sensor data, and reallocating the flow direction and proportion of waste resource utilization by dynamically adjusting the path planning matrix and the optimization weight of the objective function; S37. In the edible fungus growing area, the environmental conditions of the edible fungus growing area, including temperature, humidity and water supply, are automatically adjusted in combination with real-time sensor data.

5. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The S4 specifically includes: S41. Model each module in the agricultural ecosystem as an intelligent agent and define the state, behavior and reward value of each intelligent agent; S42. Use multi-agent reinforcement learning technology to establish a collaborative optimization mechanism for the state and behavior of the agents, build the system's resource allocation objective function, and generate the system's preliminary resource allocation strategy with the goal of maximizing overall resource utilization efficiency: Among them, the reward value R i is the benefit after resource allocation of the module, γ is the discount factor, G is the overall resource allocation benefit of the agricultural ecosystem, and n is the number of agents in the agricultural ecosystem; S43. According to the generated preliminary resource allocation strategy, the state and behavior of the agent are optimized, the resource allocation strategy of each agent is dynamically adjusted, and the resource allocation ratio between modules is updated: Where E is the expected value, is the optimal strategy of module i in the current state, S i is the current state of module i, A i is the current behavior of module i, argmax represents the optimal strategy S44, dynamically adjust the resource allocation ratio between modules according to the optimal behavior strategy of the intelligent agent, and update the state and behavior of the intelligent agent in real time, and re-optimize the resource allocation strategy in combination with the resource demand changes between modules and the real-time data collected by sensors; S45. Based on the optimized resource allocation strategy, dynamically adjust the resource flow paths of each module in the ecosystem; S46. Based on the optimization results of resource flow paths, the resource allocation in aquaculture areas is further optimized by adopting a multi-layer farming structure, including bottom fish farming, middle-layer shrimp farming and upper-layer plankton farming, and the optimization results are fed back to the updating process of resource allocation strategy.

6. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The S5 specifically includes: S51. Based on the dynamic resource flow paths and distribution ratios of the agricultural ecosystem, define the material and energy flow states of each module in the agricultural ecosystem, where the bee breeding area is adjacent to the mulberry planting area and the crop planting area, and the yield of crops and mulberry trees is increased through bee pollination, and honey and beeswax are collected; S52. Construct a dynamic food chain model of the agricultural ecosystem. Based on the resource flow status between modules, analyze the resource transfer path between different modules and clarify the complete flow link of resources from input module to output module. S53. Use metabolic flow simulation technology to analyze the material and energy flow paths between modules: Where E is the global energy allocation efficiency of the system, F ij is the resource flow between modules, η ij is the energy utilization efficiency from module i to module j, L i is the output resource amount of module i, and n is the total number of modules; S54, dynamically adjusting the resource allocation ratio between modules according to the optimization result of the resource flow path between modules, generating a new resource flow path, and optimizing the transfer efficiency and global utilization rate of resources; S55, combining the adjusted resource flow path, generating an optimized material flow network between modules, and updating the resource flow status of the module; S56. Output the optimized dynamic food chain model, including the optimized resource flow path between modules and the updated resource flow status of each module.

7. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The S6 specifically includes: S61. Based on the optimization results of the resource flow paths between modules, water quality parameters of the aquaculture area, including nitrogen content, phosphorus content and dissolved oxygen concentration, are collected, and water purification demand data is generated using sensors; S62. Construct a water purification model, combine water quality parameters with agricultural irrigation needs and aquaculture recycling needs, and define water purification targets, including pollutant removal efficiency, purification time, and purified water volume; S63, using aquatic plant purification system, selecting duckweed to absorb and purify nitrogen and phosphorus pollutants, while combining biological filtration technology to remove organic pollutants in the water to generate preliminary purified water; S64. Based on the Ivy optimization algorithm, the water purification path is dynamically optimized. Combined with water quality parameters, purification goals and resource allocation ratios, the redistribution path of purified water is optimized to generate a water resource transfer matrix between modules: Among them, P ij represents the water allocation ratio from module i to module j, U represents the overall benefit of water resources after allocation among modules, and Q ij represents the corresponding amount of water, n is the total number of ecological modules, and n is the total number of modules in the agricultural ecosystem; S65. Redistribute the purified water resources according to the optimized path, with a portion used for agricultural irrigation modules and the other portion returned to the aquaculture area, so as to achieve efficient recycling of water resources; S66. Update the water purification model in real time, dynamically adjust the purification target and resource allocation path according to the changes in water quality parameters collected by the sensor and the module demand data, and generate an updated purification path and water resource allocation plan.

8. The method for constructing an agricultural morphology model according to claim 1, characterized in that: The S7 specifically includes: S71. Based on the collected multimodal data, the time series prediction model is used to dynamically predict the key indicators in the agricultural system; S72. Based on dynamic prediction results and combined with feedback control system, key parameters in agricultural production are optimized in real time; S73. Through the feedback control system, dynamically adjust production factors according to the predicted results to optimize irrigation, fertilization and feeding amount: Among them, Y i represents the actual value of the i-th key indicator, is the predicted value of the i-th key indicator, w i is the weight coefficient of the corresponding indicator, Z is the objective function of minimizing the overall error, n represents the total number of key indicators predicted in the agricultural system, and i represents the index number of the key indicator in the agricultural system; S74. Based on prediction and regulation, dynamic adjustment instructions are generated, and optimization operations are performed through control devices.

Images