Intelligent irrigation method and system based on digital country

Abstract

The invention discloses an intelligent irrigation method and system based on a digital country, and relates to the technical field of intelligent agriculture, and the method comprises the steps: collecting ecological data of a farmland region, carrying out the modeling, dividing the farmland into a plurality of sub-regions according to an ecological constraint condition, and distributing an intelligent agent; each agent independently formulates a sub-region irrigation strategy and shares strategy data; constructing a multi-agent optimization model based on a time cooperative game, and carrying out global collaborative optimization on the irrigation strategy of the whole farmland area; and executing irrigation of the sub-regions according to the irrigation strategy after global collaborative optimization. According to the method, the global water resource optimal distribution of the farmland can be realized, the problem that local optimization and global optimization are not matched in the prior art is avoided, the irrigation plan is more accurate, the irrigation strategy can be dynamically adjusted according to the real-time environmental data, and thus the water resource waste is reduced. Future irrigation requirements are predicted, and it is ensured that efficient decisions can be made under uncertain conditions.

Description

Technical Field

[0001] The present invention relates to the field of smart agricultural technology, and in particular to a smart irrigation method and system based on digital villages. Background Art

[0002] With the advancement of global agricultural modernization, intelligent agriculture has gradually become a core direction for improving agricultural production efficiency and water resource utilization. Traditional agricultural irrigation methods often rely on manual adjustments and lack real-time monitoring and precise management of factors such as soil, moisture, and climate, resulting in water waste and low irrigation efficiency. In recent years, with the rapid development of the Internet of Things (IoT), big data, artificial intelligence (AI), and sensor technology, agricultural irrigation solutions based on intelligent systems have gradually emerged. These systems use sensor networks to collect data such as soil moisture, temperature, and climate, and use data analysis and optimization algorithms to adjust irrigation plans in real time. This enables precise irrigation and significantly improves water resource efficiency. The development of intelligent irrigation systems mainly focuses on how to automate and intelligentize the farmland irrigation process through data collection, analysis, and optimization models.

[0003] However, existing intelligent irrigation systems still have certain shortcomings. Most current intelligent irrigation systems rely on a single agent to make irrigation decisions within a region, lacking system-level global optimization. While some systems are capable of automatically irrigating local areas based on sensor data, these systems often lack flexibility and precision when faced with challenges such as resource allocation between different regions, long-term climate change prediction, and global coordination of farmland. Furthermore, existing technologies still have significant room for improvement in optimizing global resource allocation on farmland, ensuring coordination between different regions, and avoiding resource conflicts. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing technical solutions usually adopt a single optimization method, resulting in local optimization failing to match global optimization, unbalanced distribution of water resources, and differences in irrigation effects.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: an intelligent irrigation method based on digital villages, comprising: Collect ecological data of farmland areas and model them, divide the farmland into multiple sub-areas based on ecological constraints, and assign agents to them; Each agent independently formulates a sub-region irrigation strategy and shares strategy data; Construct a multi-agent optimization model based on time-cooperative game to perform global collaborative optimization of irrigation strategies for the entire farmland area; The irrigation of the sub-areas is performed according to the irrigation strategy after global collaborative optimization.

[0007] As a preferred embodiment of the digital village-based smart irrigation method of the present invention, the ecological constraint conditions include collecting ecological data of the farmland area through soil, meteorological, and temperature sensors, and constructing an ecological model of the farmland area based on the ecological data; the ecological data specifically includes soil type, geographical location, climate data, crop type, and planting distribution; Farmland ecological constraints are constructed based on soil type, climate conditions and topography in the ecological model, and farmland sub-areas are divided according to the ecological constraints.

[0008] As a preferred embodiment of the digital village-based intelligent irrigation method of the present invention, the independent formulation of sub-region irrigation strategies includes configuring an intelligent agent for each sub-region, the intelligent agent being responsible for controlling the irrigation equipment in the sub-region and making independent decisions; The agent collects environmental data within the sub-regions in real time, including soil moisture, air temperature, and crop growth status. It analyzes this data and calculates the sub-region's irrigation needs. The irrigation needs are a function of soil moisture, temperature, and crop status parameters, representing the required irrigation amount and duration for each sub-region. Each agent By analyzing the environmental data of this area and irrigation needs , calculate the local irrigation strategy of the sub-area of ​​the agroecological model ; Local strategy The goal is to meet the irrigation needs of the region.

[0009] As a preferred solution of the digital village-based intelligent irrigation method described in the present invention, the shared strategy data includes: the intelligent agent shares the environmental data and irrigation needs of the sub-area through the communication protocol; the intelligent agent updates the irrigation strategy through collaborative optimization based on the current state and the state of the adjacent intelligent agent, and updates the irrigation strategy according to the state of other intelligent agents. Irrigation needs and environmental data Make adjustments; Among them, the collaborative optimization goal is to make the global irrigation strategy of all agents Achieve optimal resource allocation for the entire farmland.

[0010] As a preferred solution of the digital village-based intelligent irrigation method described in the present invention, the multi-agent optimization model includes introducing a local collaborative reinforcement learning algorithm to define the state space , represents each agent Environmental information of the area and status information of the neighboring areas, including soil moisture, temperature, crop growth status and status information of the neighboring areas of the neighboring agents; Define the action space, each agent The action space contains the actions that the agent can perform to control the irrigation equipment; the reward function is defined to represent each agent Consider the irrigation effect of the current area and combine it with the collaborative benefits with other intelligent agents; Agent Integrate other agents during local optimization Strategy , optimize the sub-region irrigation plan; intelligent agent When selecting actions in the action space, collaborative decision-making is performed based on the time-based cooperative game method; Agent When making irrigation decisions, a deep Q-network is used to update the Q-value and adjust it using the strategy information of the neighboring areas; each agent adjusts its own irrigation strategy based on its updated Q-value , and finally choose the strategy that maximizes the Q value; When performing local updates, the agent Update irrigation decisions for the entire farmland by coordinating with other agents .

[0011] As a preferred solution of the digital village-based intelligent irrigation method of the present invention, the global collaborative optimization includes: According to current environmental data and predicted irrigation needs , select an initial irrigation strategy Perform game strategy calculations; each agent calculates a sub-region strategy based on the game model to ensure maximum local benefits; Add uncertainty modeling to the farmland ecological model and predict the intelligent agent by simulating different climate change scenarios. Under different environmental conditions The future benefits under the condition of Bayesian optimization are used to make probability predictions and estimate the agent Rewards for different strategy choices; The agent shares strategy data with other agents and adjusts the strategy of the sub-region to optimize the global benefit. Based on the time cooperation game method, the benefit function of each agent is calculated. , according to other agents Strategy Calculate the equilibrium of the game and obtain the benefit function by weighted summing the agent’s reward and predicted future benefits ; Input the strategy set of all agents , and obtain the benefits of each agent ; Dynamically adjust the game equilibrium, each agent By solving the game equilibrium, an optimal strategy is found to maximize one's own benefits; the game model determines the best strategy for each agent by solving the Nash equilibrium.

[0012] As a preferred solution of the digital village-based intelligent irrigation method of the present invention, wherein: the irrigation of the execution sub-area includes: each intelligent body Independently execute irrigation plans within the sub-areas it controls; optimize global irrigation decisions based on reinforcement learning and game model optimization for each agent , select the irrigation strategy for the sub-region , and convert it into device control commands; The equipment control commands include irrigation plan parameters, irrigation amount, irrigation time, control equipment parameters, water pump control signals, valve opening control signals, sprinkler irrigation frequency and duration.

[0013] As a preferred solution of the digital village-based intelligent irrigation system described in the present invention, it includes: a data acquisition module, an intelligent agent decision module, a collaboration and communication module, and a control and execution module; The data acquisition module is used to monitor the real-time environmental data of each area of ​​the farmland and transmit the data to the corresponding intelligent agent; The agent decision module is used to analyze irrigation needs based on collected data and optimize irrigation strategies using reinforcement learning algorithms. It also shares information with neighboring agents through local decision-making for collaborative optimization. The collaboration and communication module is used to realize communication and collaboration between intelligent agents; The control and execution module is used to control the irrigation equipment according to the optimized irrigation plan of each intelligent agent under global coordination.

[0014] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a step of an intelligent irrigation method based on a digital village.

[0015] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of an intelligent irrigation method based on a digital village.

[0016] Beneficial effects of the present invention: The intelligent irrigation method based on digital villages provided by the present invention divides the farmland into multiple sub-areas. Each area is independently decided by an intelligent agent and works in collaboration with other intelligent agents. This can achieve the optimal allocation of water resources for the entire farmland, avoiding the problem of mismatch between local optimization and global optimization in the existing technology. The use of ecological data modeling (including soil type, meteorological data, etc.) makes irrigation plans more accurate, and can dynamically adjust irrigation strategies based on real-time environmental data, thereby reducing water resource waste. Based on multi-agent reinforcement learning and game optimization models, the present invention can adapt to different environments and climate changes, predict future irrigation needs, and ensure that efficient decisions can be made under uncertain conditions. Information sharing and collaborative optimization among various intelligent agents enable the system to have efficient resource scheduling capabilities, thereby improving the reliability and stability of the overall irrigation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is an overall flow chart of an intelligent irrigation method based on a digital village provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0019] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that 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 persons in this field without creative work should fall within the scope of protection of the present invention.

[0020] Example 1, reference Figure 1 , as an embodiment of the present invention, provides an intelligent irrigation method based on a digital village, comprising: S1: Collect ecological data of the farmland area and model it, divide the farmland into multiple sub-areas based on ecological constraints, and assign intelligent agents.

[0021] Furthermore, the ecological constraints include collecting ecological data of the farmland area through soil, meteorological, and temperature sensors, and constructing an ecological model of the farmland area based on the ecological data; the ecological data specifically includes soil type, geographical location, climate data, crop types, and planting distribution; It should be noted that farmland ecological constraints are constructed based on soil types, climate conditions and topography in the ecological model, and farmland sub-areas are divided according to the ecological constraints.

[0022] S2: Each agent independently formulates a sub-region irrigation strategy and shares strategy data.

[0023] Furthermore, the independent formulation of the sub-region irrigation strategy includes configuring an intelligent agent for each sub-region, the intelligent agent being responsible for controlling the irrigation equipment in the sub-region and making independent decisions; The agent collects environmental data within the sub-regions in real time, including soil moisture, air temperature, and crop growth status. It analyzes this data and calculates the sub-region's irrigation needs. The irrigation needs are a function of soil moisture, temperature, and crop status parameters, representing the required irrigation amount and duration for each sub-region. Each agent By analyzing the environmental data of this area and irrigation needs , calculate the local irrigation strategy of the sub-area of ​​the agroecological model ; Local strategy The goal is to meet the irrigation needs of the region.

[0024] Among them, the collaborative optimization goal is to make the global irrigation strategy of all agents Achieve optimal resource allocation for the entire farmland.

[0025] It should be noted that the shared strategy data includes the intelligent agents sharing the environmental data and irrigation needs of the sub-area through the communication protocol; the intelligent agents update the irrigation strategy through collaborative optimization based on the current state and the state of the adjacent intelligent agents, and Irrigation needs and environmental data Make adjustments.

[0026] S3: Build a multi-agent optimization model based on time-cooperative game to perform global collaborative optimization of irrigation strategies for the entire farmland area.

[0027] Furthermore, the multi-agent optimization model includes introducing a local collaborative reinforcement learning algorithm to define the state space , represents each agent The environmental information of the area and the status information of the neighboring areas, specifically including soil moisture, temperature, crop growth status and status information of the neighboring areas of the neighboring intelligent agents.

[0028] State Space: Each agent The state space Contains the environment information of its area and the state information of the neighboring areas. It can be expressed as:

[0029] in, is soil moisture, is the temperature, The crop growth state, is the set of neighboring agents, It is the status information of the neighboring area.

[0030] Reward function, each agent The reward function Not only the irrigation effect of the current area is considered, but also the collaborative benefits with other agents are taken into account. The reward function can be expressed as:

[0031] in, is the irrigation demand of the agent, is the information of neighboring agents, It is a prediction of future irrigation benefits.

[0032] Define the action space, each agent The action space contains the actions that the agent can perform to control the irrigation equipment; the reward function is defined to represent each agent Consider the irrigation effect of the current area and combine the benefits of collaboration with other agents.

[0033] Agent Integrate other agents during local optimization Strategy , optimize the sub-region irrigation plan. When selecting actions in the action space, collaborative decision-making is performed based on the time-based cooperative game method.

[0034] The agent performs the selected action , and updates the state based on the result of executing the action. Depends on the current state and action:

[0035] in, Is a function that indicates that in the current state Next, execute the action After the environment state changes. Reward calculation and Q value update Each agent receives a reward at each time step , used to evaluate the effect of the current action. The reward not only considers the irrigation effect of the local area, but also the collaborative benefits with other agents: Reward function ( ),award Agent-based In time Actions taken when As well as the current irrigation effect, the reward function also includes the influence of neighboring areas because each agent collaborates with other agents to make irrigation decisions:

[0036] in, It is an intelligent agent Irrigation needs of the current area, is the state information of the neighboring agents, is the agent’s prediction of future irrigation benefits.

[0037] Q value update, agent Using the Q learning algorithm in reinforcement learning, the Q value function formula is updated by the following formula:

[0038] in, is the learning rate, which controls the step size of Q value update, Is the discount factor, which indicates the importance of future rewards. Is the next state The Q value of the optimal action.

[0039] Agent When making irrigation decisions, a deep Q-network is used to update the Q-value and adjust it using the policy information of neighboring areas. Each agent adjusts its own irrigation strategy based on its updated Q-value. , and finally choose the strategy that maximizes the Q value.

[0040] When performing local updates, the agent Update irrigation decisions for the entire farmland by coordinating with other agents .

[0041] The global collaborative optimization includes: According to current environmental data and predicted irrigation needs , select an initial irrigation strategy Perform game strategy calculations; each agent calculates sub-region strategies based on the game model to ensure maximum local benefits.

[0042] Add uncertainty modeling to the farmland ecological model and predict the intelligent agent by simulating different climate change scenarios. Under different environmental conditions The future benefits under the condition of Bayesian optimization are used to make probability predictions and estimate the agent Rewards in different strategy choices.

[0043] Dynamic adjustment of game equilibrium: Agents calculate equilibrium strategies through cooperative game play to ensure optimal global performance. Game equilibrium is typically achieved by calculating Nash equilibrium, ensuring that each agent's decision is optimal and that it is impossible to improve performance by unilaterally changing strategies given the strategies of other agents.

[0044] Game equilibrium adjustment, initial strategy selection: Each agent initially selects an irrigation strategy based on initial environmental data and predicted benefits.

[0045] Cooperative Game Calculation: All agents use a cooperative game model to calculate the globally optimal irrigation strategy. Each agent adjusts its strategy based on the results of the strategies of other agents. During the game, agents continuously optimize their irrigation strategies to achieve the maximum overall benefit.

[0046] Game equilibrium update: The agent adjusts its strategy based on the game results and seeks the optimal solution for the system through interaction with other agents. Game equilibrium updates can be performed in the following ways: Each agent Adjustments are made based on the current strategy and the strategies of neighboring agents to ensure maximum global benefits.

[0047] Calculate the feedback during the equilibrium process and adjust the strategies of each agent so that each agent can obtain the optimal benefits within the framework of cooperation.

[0048] The agent shares strategy data with other agents and adjusts the strategy of the sub-region to optimize the global benefit. Based on the time cooperation game method, the benefit function of each agent is calculated. , according to other agents Strategy Calculate the equilibrium of the game and obtain the benefit function by weighted summing the agent’s reward and predicted future benefits ; Input the strategy set of all agents , and obtain the benefits of each agent .

[0049] Each agent selects an initial irrigation strategy based on current environmental data (e.g., soil moisture, climate change) and predicted irrigation needs. .

[0050] Enter current environment data , output the initial strategy .

[0051] Cooperative game calculation, game strategy calculation, each agent calculates its strategy based on the game model to ensure maximum benefit. Agents adjust their strategies to optimize global benefits by interacting with other agents. Calculate the benefit function of each agent , and according to the strategies of other agents Compute the equilibrium of the game.

[0052]

[0053] in, is the agent’s reward, It is a forecast of future benefits. is the time weight, is the discount factor. Represents the time period, which is the time range for benefit calculation. Indicates the initial time. Indicates the moment of decision making in the next time step.

[0054] Each agent Adjust based on the current strategy and the strategies of other agents. The agent solves the game equilibrium and finds an optimal strategy that maximizes its own benefits. Input the current strategy, the strategies of other agents, and output the updated optimal strategy.

[0055] In agricultural environments, factors such as climate change and soil moisture fluctuations can affect the execution of irrigation plans. Therefore, cooperative game models need to account for these uncertainties. Monte Carlo simulation methods can be used to simulate different climate change scenarios and estimate the benefits of each agent under different environmental conditions. Each agent then makes decisions based on these simulation results during the game. Bayesian optimization methods can be used to combine historical data (such as climate change and soil moisture) with current conditions. Agents use Bayesian optimization to make probabilistic predictions and estimate the uncertainty of future benefits. By modeling uncertainty, agents not only consider current and future expected benefits but also adjust based on these uncertainties, making their decisions more flexible and adaptable to changing circumstances.

[0056] Considering the influence of uncertain factors, the benefit function is updated and the formula is expressed as:

[0057] in, It represents an expected estimate of future benefits to deal with uncertainty.

[0058] Dynamically adjust the game equilibrium, each agent By solving the game equilibrium, an optimal strategy is found to maximize one's own benefits; the game model determines the best strategy for each agent by solving the Nash equilibrium.

[0059] Based on the Timed Cooperative Game (TCG) model, agents optimize their cooperation with other agents through game strategies. At each time t, agents adjust their strategies based on the equilibrium state of the cooperative game. By adjusting their strategies and cooperative game strategies, agents achieve game equilibrium:

[0060] in, It is an intelligent agent The benefit function at time t is based on the current reward and forecasts of future benefits .

[0061] It should be noted that after each round of the game, all agents update their strategies by comparing their actual rewards with their predicted rewards, allowing the system to gradually converge to the globally optimal irrigation plan. After multiple iterations, when all agents' strategies no longer change significantly, the model is considered converged and ultimately outputs the optimal irrigation strategy.

[0062] S4: Execute irrigation of the sub-area according to the globally coordinated optimized irrigation strategy.

[0063] Furthermore, the irrigation of the execution sub-area includes each agent Independently execute irrigation plans within the sub-areas it controls; optimize global irrigation decisions based on reinforcement learning and game model optimization for each agent , select the irrigation strategy for the sub-region , and convert it into device control commands.

[0064] The equipment control commands include irrigation plan parameters, irrigation amount, irrigation time, control equipment parameters, water pump control signals, valve opening control signals, sprinkler irrigation frequency and duration.

[0065] Each agent In the area under its control The core of the control process is to convert the irrigation decisions made by each agent, optimized through reinforcement learning and game models in the previous steps, into actual equipment control commands. This process includes the following: Irrigation plan parameters: Each agent uses the collected environmental data to and optimized irrigation needs Generate irrigation plans An irrigation plan typically includes the following parameters: Irrigation volume :Each agent calculates the required amount of water based on the irrigation demand. Irrigation time :The irrigation duration is calculated based on the required irrigation volume and water source flow. Control equipment parameters: The agent adjusts the irrigation time according to the irrigation plan. Adjust equipment within the area (e.g., pumps, valves, sprinkler systems, etc.) to achieve the desired irrigation results.

[0066] It should be noted that the parameters of the control equipment typically include: Pump control signal: determines whether the pump is on or off; Valve opening control signal: adjusts the valve opening to control the amount of water flow; Sprinkler frequency and duration: sets the sprinkler system to ensure uniform water coverage in each area.

[0067] Example 2 is an embodiment of the present invention, which provides an intelligent irrigation system based on digital villages, including a data acquisition module, an intelligent agent decision module, a collaboration and communication module, and a control and execution module.

[0068] The data acquisition module is used to monitor the real-time environmental data of each area of ​​the farmland in real time and transmit the data to the corresponding intelligent agent.

[0069] The intelligent agent decision module is used to analyze irrigation needs based on the collected data, and use the reinforcement learning algorithm to optimize the irrigation strategy. It shares information with neighboring intelligent agents through local decision-making to perform collaborative optimization.

[0070] The collaboration and communication module is used to realize communication and collaboration between intelligent agents.

[0071] The control and execution module is used to control the irrigation equipment according to the optimized irrigation plan of each intelligent agent under global coordination.

[0072] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0073] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0074] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting, or processing it in another suitable manner as necessary, and then storing it in a computer memory.

[0075] It should be understood that various components of the present invention may be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof may be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0076] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. The intelligent irrigation method based on digital village is characterized by: include: Collect ecological data of farmland areas and model them, divide the farmland into multiple sub-areas based on ecological constraints, and assign agents to them; Each agent independently formulates a sub-region irrigation strategy and shares strategy data; Construct a multi-agent optimization model based on time-cooperative game to perform global collaborative optimization of irrigation strategies for the entire farmland area; The irrigation of the sub-areas is performed according to the irrigation strategy after global collaborative optimization.

2. The intelligent irrigation method based on digital village according to claim 1, characterized in that: The ecological constraints include collecting ecological data of the farmland area through soil, weather, and temperature sensors, and constructing an ecological model of the farmland area based on the ecological data; the ecological data specifically includes soil type, geographical location, climate data, crop types, and planting distribution; Farmland ecological constraints are constructed based on soil type, climate conditions and topography in the ecological model, and farmland sub-areas are divided according to the ecological constraints.

3. The intelligent irrigation method based on digital village according to claim 2, characterized in that: The independent formulation of the sub-region irrigation strategy includes configuring an intelligent agent for each sub-region, the intelligent agent being responsible for controlling the irrigation equipment in the sub-region and making independent decisions; The agent collects environmental data within the sub-regions in real time, including soil moisture, air temperature, and crop growth status. It analyzes this data and calculates the sub-region's irrigation needs. The irrigation needs are a function of soil moisture, temperature, and crop status parameters, representing the required irrigation amount and duration for each sub-region. Each agent By analyzing the environmental data of this area and irrigation needs , calculate the local irrigation strategy of the sub-area of ​​the agroecological model ; Local strategy The goal is to meet the irrigation needs of the region.

4. The intelligent irrigation method based on digital village according to claim 3, characterized in that: The shared strategy data includes the environment data and irrigation requirements of the sub-area shared by the agent through the communication protocol; the agent updates the irrigation strategy through collaborative optimization based on the current state and the state of the neighboring agents, and Irrigation needs and environmental data Make adjustments; Among them, the collaborative optimization goal is to make the global irrigation strategy of all agents Achieve optimal resource allocation for the entire farmland.

5. The intelligent irrigation method based on digital village according to claim 4, characterized in that: The multi-agent optimization model includes introducing a local collaborative reinforcement learning algorithm and defining a state space. , represents each agent Environmental information of the area and status information of the neighboring areas, including soil moisture, temperature, crop growth status and status information of the neighboring areas of the neighboring agents; Define the action space, each agent The action space contains the actions that the agent can perform to control the irrigation equipment; the reward function is defined to represent each agent Consider the irrigation effect of the current area and combine it with the collaborative benefits with other intelligent agents; Agent Integrate other agents during local optimization Strategy , optimize the sub-region irrigation plan; intelligent agent When selecting actions in the action space, collaborative decision-making is performed based on the time-based cooperative game method; Agent When making irrigation decisions, a deep Q-network is used to update the Q-value and adjust it using the strategy information of the neighboring areas; each agent adjusts its own irrigation strategy based on its updated Q-value , and finally choose the strategy that maximizes the Q value; When performing local updates, the agent Update irrigation decisions for the entire farmland by coordinating with other agents .

6. The intelligent irrigation method based on digital village according to claim 5, characterized in that: The global collaborative optimization includes: Agent According to current environmental data and predicted irrigation needs , select an initial irrigation strategy Perform game strategy calculations; each agent calculates a sub-region strategy based on the game model to ensure maximum local benefits; Add uncertainty modeling to the farmland ecological model and predict the intelligent agent by simulating different climate change scenarios. Under different environmental conditions The future benefits under the condition of Bayesian optimization are used to make probability predictions and estimate the agent Rewards for different strategy choices; The agent shares strategy data with other agents and adjusts the strategy of the sub-region to optimize the global benefit. Based on the time cooperation game method, the benefit function of each agent is calculated. , according to other agents Strategy Calculate the equilibrium of the game and obtain the benefit function by weighted summing the agent’s reward and predicted future benefits ; Input the strategy set of all agents , and obtain the benefits of each agent ; Dynamically adjust the game equilibrium, each agent By solving the game equilibrium, an optimal strategy is found to maximize one's own benefits; the game model determines the best strategy for each agent by solving the Nash equilibrium.

7. The intelligent irrigation method based on digital village according to claim 6, characterized in that: The irrigation of the execution sub-area includes: each agent Independently execute irrigation plans within the sub-areas it controls; optimize global irrigation decisions based on reinforcement learning and game model optimization for each agent , select the irrigation strategy for the sub-region , and convert it into device control commands; The equipment control commands include irrigation plan parameters, irrigation amount, irrigation time, control equipment parameters, water pump control signals, valve opening control signals, sprinkler irrigation frequency and duration.

8. A system using the digital village-based intelligent irrigation method according to any one of claims 1 to 7, characterized in that: Including data acquisition module, intelligent agent decision module, collaboration and communication module, control and execution module; The data acquisition module is used to monitor the real-time environmental data of each area of ​​the farmland and transmit the data to the corresponding intelligent agent; The agent decision module is used to analyze irrigation needs based on collected data and optimize irrigation strategies using reinforcement learning algorithms. It also shares information with neighboring agents through local decision-making for collaborative optimization. The collaboration and communication module is used to realize communication and collaboration between intelligent agents; The control and execution module is used to control the irrigation equipment according to the optimized irrigation plan of each intelligent agent under global coordination.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the smart irrigation method based on digital village according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the smart irrigation method based on digital village according to any one of claims 1 to 7 are implemented.

Images