Digital twinborn irrigation area automatic irrigation method and system based on AI

Abstract

The invention discloses an AI-based digital twinborn irrigation area automatic irrigation method and system, and relates to the technical field of artificial intelligence, and the method comprises the following steps: collecting data, and constructing a digital twinborn model; irrigation demand prediction is carried out according to the digital twinborn model; making an irrigation control decision according to the irrigation demand prediction result; performing automatic irrigation based on the irrigation control decision content; crop growth evaluation is carried out according to the irrigation execution result; performing irrigation strategy optimization according to a growth evaluation result; maintaining and updating the irrigation system according to the irrigation strategy; and performing water resource management based on the planning result. According to the method, the digital twin model can be updated in real time and adapt to environmental changes by introducing the data stream based on the Internet of Things and a reinforcement learning mechanism. The feedback mechanism based on AI and reinforcement learning enables the digital twinborn model to more accurately simulate an irrigation environment and provide an optimal irrigation strategy under different environmental conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence, and specifically provides an AI-based automatic irrigation method and system for digital twin irrigation areas. Background Technique

[0002] With the gradual development of agricultural production towards intelligence and precision, irrigation management, as a key link in agriculture, is undergoing a technological revolution. Traditional irrigation management methods mostly rely on manual experience and simple automation equipment, with low efficiency, poor accuracy, and difficulty in coping with complex and changing agricultural environments. In recent years, the rapid development of artificial intelligence (AI), digital twin technology, and deep learning has provided new solutions for irrigation management.

[0003] In this context, an automatic irrigation system for irrigation areas based on a digital twin model has emerged. Digital twin technology provides a virtual model synchronized with the real world by collecting and virtually reconstructing the physical environment of the irrigation area in real time (such as soil moisture, crop growth status, climate conditions, etc.). By highly restoring the dynamic changes of the real environment, this model enables irrigation decisions to be adjusted and optimized based on real-time data, greatly improving the water resource utilization efficiency.

[0004] In the prior art, digital twin models mainly rely on static mapping of the physical environment and prediction based on historical data. However, in practical applications, due to the complexity and dynamic changes of the environment, traditional digital twin models often cannot accurately reflect changing irrigation demands in real time. For example, traditional soil moisture models may not fully consider various factors such as irrigation time, soil type, and crop growth status, resulting in inaccurate control of irrigation decisions. In addition, existing digital twin models mostly rely on a single physical law or statistical method and lack sufficient adaptability to cope with changing climate and environmental conditions. Summary of the Invention

[0005] Aiming at the deficiencies of the prior art, the present invention provides an AI-based automatic irrigation method and system for digital twin irrigation areas to solve the problems raised in the above background technique.

[0006] To achieve the above object, the present invention provides the following technical solutions: In the first aspect, an embodiment of the present invention provides an AI-based automatic irrigation method for digital twin irrigation areas, including the following steps: S1. Data collection and construction of a digital twin model; Collect data through multi-source sensors and construct a digital twin model according to the collected data; S2. Predict irrigation demand according to the digital twin model; Based on the digital twin model, use deep learning technology to predict the irrigation demand of the irrigation area and obtain the irrigation demand prediction result; S3. Make an irrigation control decision according to the irrigation demand prediction result; Input the irrigation demand prediction result into the decision-making system based on reinforcement learning to make an irrigation control decision and obtain the irrigation control decision content; S4. Perform automatic irrigation execution based on the irrigation control decision content; Based on the irrigation control decision content, use the automatic irrigation system to perform irrigation operations and obtain the irrigation execution result; S5. Evaluate the crop growth according to the irrigation execution result; Based on the irrigation execution result, use a drone to take pictures of the crops, automatically identify the growth situation of the crops, and evaluate the health status of the leaves, roots and the overall crops to obtain the growth evaluation result; S6. Optimize the irrigation strategy according to the growth evaluation result; Optimize the irrigation strategy according to the growth evaluation result and by analyzing the relationship between the crop growth state and the irrigation amount to obtain the optimized irrigation strategy; S7. Maintain and update the irrigation system according to the irrigation strategy; According to the optimized irrigation strategy, use the evolutionary algorithm to carry out long-term irrigation planning and obtain the planning result.

[0007] S8. Conduct water resource management based on the planning result; Reasonably allocate the water resources in the irrigation area according to the planning result.

[0008] Further optimize this technical solution. The data collected in step S1 includes: Collect soil moisture content, air temperature, precipitation, wind speed, and air humidity through multi-source sensors.

[0009] Further optimize this technical solution. The twin model in step S1 includes constructing three-dimensional coordinates and obtaining the collected data, constructing a water flow equation, and constructing a soil moisture change equation.

[0010] Further optimize this technical solution. The construction of three-dimensional coordinates and obtaining the collected data includes: Take the ground point at the bottom left corner of the irrigation area as the origin, take the east-west direction as the x-axis, with west being negative and east being positive; take the north-south direction as the y-axis, with south being negative and north being positive; take the direction perpendicular to the ground as the z-axis, with up being positive and down being negative, and establish the z-axis; Obtain the soil moisture content of each collection point of.

[0011] Further optimize this technical solution. The construction of the water flow equation includes: ; Among them, is the soil moisture content at position and time t, with the unit of ; is the diffusion coefficient of water, reflecting the ability of water to spread in the soil, obtained through experimental measurement and artificially set; is the inflow rate of water flow, with the unit of ; is the Laplace operator, representing the expansion of water in space; The inflow rate of water flow is calculated through this equation .

[0012] To further optimize this technical solution, the construction of the soil humidity change equation includes: ; is the permeability of the soil, representing the ability of the soil to absorb and flow water, obtained through experimental measurement and artificially set, with the unit of ; is the consumption rate of soil moisture, representing the absorption rate of water by crops .

[0013] The absorption rate of water by crops is calculated through this equation .

[0014] To further optimize this technical solution, the deep learning technology in step S2 includes: Based on the long short-term memory network LSTM, combined with the crop type, real-time environmental data including temperature, precipitation, wind speed, and according to the calculated water absorption rate and water flow inflow rate, predict the optimal irrigation demand of the crop under this environmental data to obtain the predicted irrigation demand And the growth state G of the crop, the predicted irrigation demand includes the water requirement of the crop, soil humidity, temperature, air humidity, and the growth state of the crop includes the water absorption rate of the crop.

[0015] To further optimize this technical solution, the reinforcement learning in step S3 includes: Construct a state vector , describing the current environmental state including soil humidity, crop growth state, temperature and air humidity, predicted irrigation demand; Construct an action vector , which is the decision made by the irrigation system at each moment, including the irrigation amount and irrigation time; Construct a reward function to measure the pros and cons of the current irrigation strategy; Construct a policy function to represent the probability distribution of selecting an action when a given state is presented; Construct a value function, which represents the expected total reward obtained in the future under a state and evaluates the importance of each state, expressed as: ; represents the expectation; is the discount factor, used to balance long-term and short-term rewards; is the reward function; T is the maximum duration.

[0016] To further optimize this technical solution, the evolutionary algorithm in step S7 includes: Initialize the population: Generate initial solutions, and each initial solution represents an irrigation strategy; Fitness evaluation: Calculate the fitness of each initial solution, which is closely related to crop growth and irrigation efficiency; Selection operation: Select excellent crop individuals for reproduction according to the fitness; Crossover operation: Generate new solutions through the crossover operation, simulating the gene recombination process in nature; Mutation operation: Mutate the new solutions, simulating the gene mutation process in nature; Replacement operation: Compare the new solutions with the old solutions, select the surviving crop individuals according to the fitness, and form a new generation of population; Stop condition: When the preset stop condition is reached, terminate the algorithm and output the optimal solution.

[0017] To further optimize this technical solution, the functional modules of the automatic irrigation system include a data acquisition and sensing module, a digital twin modeling and simulation module, an irrigation demand prediction module, a reinforcement learning decision system module, an irrigation execution and automation control module, a crop growth evaluation and analysis module, an irrigation strategy optimization and adjustment module, an irrigation system maintenance and update module, and a water resource management and optimal scheduling module.

[0018] In a second aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program instructions are executed by the processor, the steps of an AI-based digital twin irrigation area automatic irrigation method and system as described in the first aspect of the present invention are implemented.

[0019] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, where: when the computer program instructions are executed by a processor, the steps of an AI-based digital twin irrigation area automatic irrigation method and system as described in the first aspect of the present invention are implemented.

[0020] Compared with the prior art, the present invention provides an AI-based digital twin irrigation area automatic irrigation method and system, which has the following beneficial effects: This AI-based digital twin irrigation area automatic irrigation method and system enable the digital twin model to be updated in real time and adapt to environmental changes by introducing Internet of Things-based data streams and reinforcement learning mechanisms. Through the input of real-time sensor data (such as soil humidity, crop health status, etc.), the system can adjust the model parameters through dynamic simulation after each irrigation decision, enabling it to generate dynamic feedback on external factors such as climate change and soil condition changes. This feedback mechanism based on AI and reinforcement learning enables the digital twin model to more accurately simulate the irrigation environment and provide optimal irrigation strategies under different environmental conditions. Through this dynamic adaptability, irrigation decisions are more precise, thus significantly improving water resource utilization efficiency, reducing resource waste, and ensuring the healthy growth of crops. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0022] Figure 1 It is a schematic flowchart of an AI-based digital twin irrigation area automatic irrigation method proposed by the present invention; Figure 2 It is a schematic diagram of a reinforcement learning algorithm of an AI-based digital twin irrigation area automatic irrigation method proposed by the present invention; Figure 3 It is a schematic diagram of the modules of an AI-based digital twin irrigation area automatic irrigation system proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings of the specification.

[0024] In the following description, numerous specific details are set forth to provide a thorough understanding of the present invention. However, the present invention may be implemented in other ways different from this description, and those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0025] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is separate from or mutually exclusive of other embodiments.

[0026] Embodiment 1: Referring to Figures 1 to 2 , this is the first embodiment of the present invention. This embodiment provides an AI-based automatic irrigation method for a digital twin irrigation area, including the following steps: S1. Data collection and construction of a digital twin model; Collect data through multi-source sensors and construct a digital twin model based on the collected data.

[0027] The collected data includes: Collect soil moisture content, air temperature, precipitation, wind speed, and air humidity through multi-source sensors.

[0028] The twin model includes constructing three-dimensional coordinates and obtaining the collected data, constructing a water flow equation, and constructing an equation for soil moisture change.

[0029] Constructing the water flow equation includes: ; Wherein, is the soil moisture content at position and time t, with the unit of ; is the diffusion coefficient of water, which reflects the ability of water to spread in the soil and is obtained through experimental measurement and artificially set; is the introduction rate of water flow, with the unit of ; is the Laplace operator, indicating the spread of moisture in space; Calculate the introduction rate of water flow through this equation .

[0030] Constructing the equation for soil moisture change includes: ; is the permeability of the soil, indicating the soil's ability to absorb and conduct water. It is obtained through experimental measurement and is artificially set, with the unit of ; is the consumption rate of soil moisture, representing the absorption rate of water by crops .

[0031] The absorption rate of water by crops is calculated through this equation .

[0032] In this embodiment, these two equations are coupled with each other to form a comprehensive model that can simulate the dynamic changes of water flow and soil moisture during irrigation. Through this model, the effects of different irrigation strategies on soil moisture and water distribution can be simulated in a digital twin environment, and then the irrigation plan can be adjusted.

[0033] The water flow equation simulates the water flow diffusion process in the irrigation system.

[0034] The soil moisture change equation combines the water demand of crops to reflect the water change in the soil and helps the system optimize the irrigation amount.

[0035] In actual use, these two equations are discretized through numerical solution methods (such as the finite difference method or the finite element method), combined with real-time data (such as soil moisture, temperature, precipitation, etc.) fed back by sensors, continuously update the model parameters, and make dynamic irrigation decisions based on this.

[0036] For example, if the model predicts that the soil moisture in a certain area is insufficient (i.e., less than a certain threshold), the system will automatically adjust the irrigation strategy and increase the water flow rate to supplement water and ensure that crops can fully absorb water.

[0037] S2. Predict the irrigation demand according to the digital twin model; Based on the digital twin model, use deep learning technology to predict the irrigation demand of the irrigation area and obtain the irrigation demand prediction result.

[0038] Deep learning technology includes: Based on the long short-term memory network LSTM, combined with crop types, real-time environmental data including temperature, precipitation, wind speed, and according to the calculated water absorption rate and water flow introduction rate, predict the optimal irrigation demand of the crop under this environmental data to obtain the predicted irrigation demand And the growth state G of the crop. The predicted irrigation demand includes the water demand of the crop, soil moisture, temperature, air humidity, and the growth state of the crop includes the water absorption rate of the crop.

[0039] The core of the LSTM network model is the memory cell, which is responsible for storing and transmitting information. The memory cell consists of a linear unit and a non-linear unit. The linear unit is a simple adder used to add the memory cell at the previous moment and the input at the current moment. The non-linear unit is a sigmoid function ( ), which is used to control the flow of information.

[0040] S3. Make an irrigation control decision based on the irrigation demand prediction result; Input the irrigation demand prediction result into the decision-making system based on reinforcement learning to make an irrigation control decision and obtain the content of the irrigation control decision.

[0041] Reinforcement learning includes: Construct a state vector , which describes the current environmental state: ; Among them, is the growth state of the crop at position at time t, is the current temperature and air humidity,

[0042] is the predicted irrigation demand. Construct an action vector , which represents the decision made by the irrigation system at each moment and is expressed as: Among them, is the irrigation amount, with the unit of ; is the irrigation time, with the unit of hour; Construct a reward function , which measures the quality of the current irrigation strategy and is expressed as: ; Among them, is the current growth state of the crop; is the water resource consumption of the irrigation system, which represents the difference between the actual irrigation amount and the required irrigation amount and reflects the efficiency of water resource utilization; is the maintenance cost of the irrigation system, including the energy consumption of the system and the funds consumed by the wear degree of the equipment; is an important coefficient for balancing the importance of each objective, which is obtained through expert evaluation and set artificially; Construct a policy function that, given a state , selects an action from a probability distribution, denoted as: ; where is the decision function at state , and its expression is ; where is 's i-th element; Construct a value function, which represents the expected total future reward obtained in state and evaluates the importance of each state, denoted as: ; represents the expectation; is the discount factor, which is used to balance long-term and short-term rewards; is the reward function; T is the maximum duration.

[0043] In this embodiment, the use of the model includes: Training process: During the training process, the reinforcement learning model optimizes the policy through interaction with the environment (implemented through the digital twin model and simulation technology). Each action taken (irrigation amount and irrigation duration) affects the soil moisture, crop growth status, etc., thereby generating reward feedback. The system learns through feedback and gradually improves the decision-making strategy.

[0044] Decision execution: In practical applications, when the system receives real-time environmental state information (such as soil moisture, crop growth, meteorological data, etc.), based on the current state , the reinforcement learning model will calculate the optimal irrigation amount and duration according to the policy , and execute the corresponding irrigation operation.

[0045] Feedback and optimization: After each irrigation, the system updates the reward and its value function , and adjusts the future policy according to the reward feedback. In this way, the reinforcement learning system continuously adapts to environmental changes and optimizes the irrigation policy.

[0046] In a decision-making system based on reinforcement learning, the digital twin model plays a crucial role. Digital Twin refers to the virtualization and real-time mapping of the physical world. In an irrigation decision-making system, the digital twin model connects the irrigation environment (including soil moisture, crop growth status, climate conditions, etc.) with the virtual model through real-time monitoring, data collection, and simulation technologies, providing a dynamic and real-time updated system view. This real-time and interactive virtual model enables the reinforcement learning system to obtain data on environmental changes in real time and make feedback, thus continuously adjusting irrigation decisions. The relationship between digital twin and reinforcement learning is mainly reflected in the following aspects: The relationship between digital twin and reinforcement learning: The reinforcement learning system relies on interactions with the environment to optimize decisions, and the digital twin model provides a high-fidelity and real-time updated environmental model for this process. The states, actions, and rewards in reinforcement learning are all based on the simulated environment in the digital twin model. Through simulation, the reinforcement learning model can conduct "virtual training" in the digital twin environment, simulating the impacts of different irrigation decisions on factors such as soil moisture and crop growth, thereby providing data support for actual decisions.

[0047] Specifically: State: The irrigation scenarios (such as soil moisture, crop growth, meteorological conditions, etc.) established through the digital twin model in step S1 constitute the input states in reinforcement learning .

[0048] Action: Each action (such as irrigation volume, duration, etc.) of the reinforcement learning decision-making system will predict the effects under different strategies based on the simulation results of the digital twin model.

[0049] Reward: The reward function is designed based on the outputs of the digital twin model (such as changes in soil moisture, crop health, irrigation efficiency, etc.) to guide the optimization of the reinforcement learning system.

[0050] Digital twin and feedback mechanism: The digital twin model not only provides a real-time virtual environment but also plays a very important role in the reinforcement learning process through the feedback mechanism. When the system executes an irrigation decision, the digital twin model can feedback environmental changes, such as changes in soil moisture and crop health, and these changes will be used as the reward input ( ) of the reinforcement learning to evaluate the effectiveness of the decision. Therefore, the digital twin model provides an accurate "virtual environment" for reinforcement learning, enabling the system to more precisely simulate the real world and adjust decisions.

[0051] Combined formula model: In step S1, we assume that a state equation of the irrigation environment is obtained through digital twin models (such as soil moisture dynamic models, crop growth models, etc.) (such as soil humidity, crop growth status, etc.). This state vector serves as the input to the reinforcement learning decision-making system.

[0052] Then, in step S3, the reinforcement learning decision-making system calculates the optimal irrigation amount and irrigation duration based on the current state : During the decision-making process, the model evaluates the effect of each irrigation decision through a reward function : ; S4. Automatically execute irrigation based on the irrigation control decision content; Based on the irrigation control decision content, use an automated irrigation system to perform irrigation operations and obtain irrigation execution results.

[0053] In this embodiment, an automated irrigation system (such as an Internet of Things-based intelligent irrigation system) is used to perform irrigation operations. Through intelligent valve and pump control, the irrigation system can automatically adjust parameters such as water flow rate and irrigation time to achieve precise irrigation. At the same time, the system can provide real-time feedback on the irrigation effect through soil humidity sensors and make corresponding adjustments. This technology reduces manual intervention through automated equipment, improves irrigation efficiency, and ensures the rational use of water resources.

[0054] S5. Evaluate crop growth based on the irrigation execution results; Based on the irrigation execution results, use a drone to take pictures of the crops, automatically identify the growth conditions of the crops, and evaluate the health status of the leaves, roots, and overall crops to obtain growth evaluation results.

[0055] In this embodiment, computer vision technology (such as convolutional neural network CNN) is used to evaluate the growth of crops. The computer vision technology will identify the growth conditions of the crops in the images and perform correlation analysis with factors such as irrigation amount and soil humidity to give a quantitative evaluation of the irrigation effect.

[0056] S6. Optimize the irrigation strategy based on the growth evaluation results; Optimize the irrigation strategy based on the growth evaluation results and by analyzing the relationship between crop growth status and irrigation amount to obtain an optimized irrigation strategy.

[0057] In this embodiment, the irrigation strategy is optimized through data mining techniques (such as clustering analysis and association rule analysis). The data mining techniques can reveal the optimal irrigation schemes under different crops and different climate conditions. The system will adjust the irrigation strategy through historical data, crop growth rules, and optimization algorithms to ensure the best moisture conditions for crop growth.

[0058] Clustering analysis: It is an unsupervised learning method used to automatically group samples in a dataset according to similarity. It can discover the rules, characteristics, and trends in agricultural production irrigation, thereby improving production efficiency and decision-making levels.

[0059] Association rule analysis: It is used to discover interesting associations between item sets in a dataset. Association rule analysis can help analyze the relationship between crop growth status and irrigation volume, and reveal the optimal irrigation schemes under different crops and different climate conditions. For example, by analyzing the rise and fall rules of vegetables through the association rule algorithm, the possible association relationships between different categories or different single items of vegetables can be discovered. This helps to formulate more effective irrigation strategies and achieve the rational allocation and utilization of water resources.

[0060] S7. Perform maintenance and update on the irrigation system according to the irrigation strategy; According to the optimized irrigation strategy, use the evolutionary algorithm for long-term irrigation planning to obtain the planning results.

[0061] The evolutionary algorithm includes: Initializing the population: Generate initial solutions, and each initial solution represents an irrigation strategy; Fitness evaluation: Calculate the fitness of each initial solution. The fitness is closely related to crop growth and irrigation efficiency; Selection operation: Select excellent crop individuals for reproduction according to the fitness; Crossover operation: Generate new solutions through the crossover operation, simulating the gene recombination process in nature; Mutation operation: Mutate the new solutions, simulating the gene mutation process in nature; Replacement operation: Compare the new solutions with the old solutions, and select the surviving crop individuals according to the fitness to form a new generation of population; Stop condition: When the preset stop condition is reached, terminate the algorithm and output the optimal solution.

[0062] In this embodiment, in order to achieve irrigation optimization in the evolutionary algorithm, an innovative fitness function is designed and modeled in combination with the specific scenarios in the irrigation problem.

[0063] In the irrigation system, the optimization objectives are multi-dimensional, including: Water resource efficiency: Maximize the effective utilization of water resources and reduce water waste.

[0064] Crop growth health: Ensure that the crops receive sufficient water to promote growth, but avoid over-irrigation.

[0065] System cost control: Optimize the energy consumption and maintenance costs of irrigation.

[0066] We need to optimize the following objectives: : Water resource efficiency (unit: m³ / ha), reflecting the increased production benefit per unit of water.

[0067] : Crop growth health (unit: growth rate), reflecting the impact of irrigation on crop growth.

[0068] : System cost (unit: currency), including costs such as energy consumption and equipment wear.

[0069] ; Among them, are the actions of irrigation decision-making, namely the irrigation amount and the irrigation duration (hours).

[0070] Calculate the water resource efficiency through the relationship between the irrigation amount and the crop water requirement.

[0071] Estimate the crop health and growth rate based on data such as the growth situation of the crops and the soil moisture.

[0072] are the energy consumption and equipment costs during the irrigation process.

[0073] is the fitness function.

[0074] are the weight coefficients, used to balance the importance of each objective, and different objectives can be optimized by adjusting these coefficients.

[0075] Calculation model of water resource efficiency: ; Among them, , representing the min function, which is used to represent the minimum value of the data in the brackets. In this formula, it represents , the minimum value.

[0076] is the irrigation amount, with the unit of .

[0077] is the actual water requirement of the crop, with the unit of .

[0078] is the water resource utilization efficiency, representing the ratio of the actual irrigation amount to the demand.

[0079] Crop growth health model: Crop growth health reflects the growth rate of the crop, which is usually related to factors such as soil moisture, temperature, and irrigation amount. It is represented by the following model: ; where, is the maximum growth potential of the crop in this area, obtained through expert field experiments, and its data is set artificially.

[0080] represents the growth health status of the crop, with a value range of [0, 1], and 1 indicates healthy crop growth.

[0081] System cost model: System cost can be calculated through the energy consumption and maintenance cost of the irrigation system. The energy consumption is mainly proportional to the irrigation amount and irrigation time, while the maintenance cost is related to the equipment usage frequency. The model can be expressed as: ; where, is the energy consumption coefficient, obtained through experiments and set artificially.

[0082] is the maintenance cost coefficient, obtained through experiments and set artificially.

[0083] is the irrigation duration, with the unit of hour.

[0084] is the equipment usage frequency, with the unit of times / hour.

[0085] Implementation and use of the evolutionary algorithm, including: Initializing the population: Generating multiple initial solutions, representing different combinations of irrigation strategies, such as different configurations of irrigation amount and irrigation duration.

[0086] Fitness evaluation: Each solution is evaluated through the fitness function to calculate its water resource efficiency, crop health, and system cost.

[0087] Selection operation: Selecting individuals with higher fitness based on the fitness function for reproduction to generate the next generation of population.

[0088] Crossover and Mutation: New solutions are generated through crossover and mutation operations, simulating the gene exchange and mutation processes in nature.

[0089] Replacement Operation and Termination: Update the population according to fitness, select the optimal solution and perform the next iteration until the termination condition is met.

[0090] S8. Conduct water resource management based on the planning results; Reasonably allocate the water resources in the irrigation area according to the planning results.

[0091] In this embodiment, comprehensive water resource scheduling and optimization management are carried out through a water resource management system. This system formulates a water resource usage plan based on factors such as irrigation demand, climate change, and water source availability, and through systematic scheduling and management, ensures that the water resources in the irrigation area are reasonably allocated, maximizing the water use efficiency. This step combines various technologies to ensure the sustainability of the irrigation system during long-term operation.

[0092] Embodiment Two: Refer to Figure 3 , which is the second embodiment of the present invention. The functional modules of the AI-based digital twin irrigation area automatic irrigation system include a data acquisition and sensing module, a digital twin modeling and simulation module, an irrigation demand prediction module, a reinforcement learning decision system module, an irrigation execution and automation control module, a crop growth assessment and analysis module, an irrigation strategy optimization and adjustment module, an irrigation system maintenance and update module, and a water resource management and optimization scheduling module.

[0093] In this embodiment, the multi-source sensors in the data acquisition and sensing module collect the environmental data in the irrigation area and transmit it to the cloud platform. The data includes soil moisture content, air temperature, precipitation, wind speed, and air humidity, which are used to update the digital twin model of the irrigation area in real time to ensure that the digital twin modeling and simulation module can respond to the changing environment. The data collected is used by the digital twin modeling and simulation module to construct and dynamically update the digital twin model using 3D modeling and simulation technologies. The system module converts information such as the environmental characteristics and crop growth status of the irrigation area into a digital virtual model through virtual modeling. The irrigation demand prediction module uses deep learning technology to predict the crop irrigation demand in the irrigation area based on the environmental data and historical climate data provided by the digital twin model. The LSTM (Long Short-Term Memory Network) model is used to process time series data to ensure prediction accuracy. The reinforcement learning decision system module makes irrigation control decisions based on the reinforcement learning algorithm using the irrigation demand prediction results. The reinforcement learning system module can dynamically adjust the irrigation strategy according to the current environmental state to ensure the efficient use of water resources and the optimal growth of crops. The irrigation execution and automation control module executes the irrigation operation and controls devices such as the valves, pumps, and water flow of the irrigation system. Through the IoT intelligent irrigation system, precise water resource regulation is achieved. The crop growth assessment and analysis module automatically assesses the growth status of crops based on computer vision technology and conducts correlation analysis with factors such as irrigation volume and soil humidity to evaluate the irrigation effect. The irrigation strategy optimization and adjustment module optimizes the irrigation strategy using data mining techniques (such as clustering analysis and association rule analysis) to find the best relationship between crops and irrigation. The irrigation system maintenance and update module conducts regular inspections and maintenance on the irrigation system. By analyzing the equipment operation data and failure modes, potential failures are predicted in advance and maintained to ensure the stability and efficiency of the system. The water resource management and optimal scheduling module optimizes the scheduling and allocation of irrigation water sources to ensure the rational use and long-term sustainability of water resources.

[0094] Embodiment 3: This embodiment also provides a computer device applicable to a situation of an AI-based digital twin irrigation area automatic irrigation method and system, including a memory and a processor. The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement an AI-based digital twin irrigation area automatic irrigation method and system as proposed in the above embodiment.

[0095] This embodiment also provides a storage medium with a computer program stored thereon, and when the program is executed by a processor, it implements an AI-based digital twin irrigation area automatic irrigation method and system as proposed in the above embodiment.

[0096] The computer device may be a terminal, which includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0097] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this 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 may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0098] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or used in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0099] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which a program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing it as appropriate, and then storing it in a computer memory.

[0100] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well-known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

[0101] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An AI-based digital twin irrigation area automatic irrigation method, characterized in that: The following steps are involved: S1, data collection and construction of digital twin model; Collect data through multi-source sensors and build a digital twin model based on the collected data; S2. Predict irrigation demand based on the digital twin model; Based on the digital twin model, deep learning technology is used to predict the irrigation demand of the irrigation area and obtain the irrigation demand prediction results; S3. Make irrigation control decisions based on the irrigation demand forecast results; The irrigation demand forecast results are input into the decision-making system based on reinforcement learning to make irrigation control decisions and obtain the irrigation control decision content; S4, automatic irrigation execution based on irrigation control decision content; Based on the irrigation control decision content, the automatic irrigation system is used to perform irrigation operations and obtain irrigation execution results; S5. Evaluate crop growth based on irrigation implementation results; Based on the irrigation execution results, the crops are photographed by drones to automatically identify the growth of the crops, and the health of the leaves, roots and the overall crops is evaluated to obtain growth assessment results. S6. Optimize irrigation strategy based on growth assessment results; The irrigation strategy is optimized according to the growth assessment results and by analyzing the relationship between the crop growth status and the irrigation amount to obtain the optimal irrigation strategy; S7. Maintain and update the irrigation system according to the irrigation strategy; According to the optimized irrigation strategy, the evolutionary algorithm is used to carry out long-term irrigation planning and obtain the planning results; S8. Carry out water resource management based on planning results; Water resources in the irrigation area are rationally allocated based on the planning results.

2. According to the AI-based digital twin irrigation area automatic irrigation method of claim 1, it is characterized in that: The data collection in step S1 includes: Soil moisture content, temperature, precipitation, wind speed, and air humidity are collected through multi-source sensors.

3. According to the AI-based digital twin irrigation area automatic irrigation method of claim 1, it is characterized in that: The twin model in step S1 includes constructing three-dimensional coordinates and acquiring collected data, constructing a water flow equation, and constructing a soil moisture change equation.

4. According to claim 3, the AI-based digital twin irrigation area automatic irrigation method is characterized in that: The constructing of three-dimensional coordinates and obtaining of collected data include: Take the ground point at the bottom left corner of the irrigation area as the origin, the east-west direction as the x-axis, with the west as negative and the east as positive; the north-south direction as the y-axis, with the south as negative and the north as positive; take the direction perpendicular to the ground as the z-axis, with the upward direction as positive and the downward direction as negative, and establish the z-axis; Get each collection point of soil moisture content.

5. According to the AI-based digital twin irrigation area automatic irrigation method of claim 3, it is characterized in that: The constructing of the water flow equation comprises: ； in, is in position and soil moisture content at time t, in units of ; It is the diffusion coefficient of water, which reflects the ability of water to spread in the soil. It is obtained through experimental measurement and is set artificially. is the water introduction rate, in units of ; is the Laplace operator, representing the expansion of moisture in space; The water flow introduction rate is calculated by this equation .

6. The AI-based digital twin irrigation area automatic irrigation method according to claim 3 is characterized in that: The soil moisture variation equation is constructed as follows: ； It is the permeability of soil, which indicates the soil's ability to absorb and flow water. It is obtained through experimental measurement and is artificially set in units of ; is the soil moisture consumption rate, representing the rate at which crops absorb water ; The water absorption rate of crops can be calculated by this equation .

7. The AI-based digital twin irrigation area automatic irrigation method according to claim 1 is characterized in that: The deep learning technology in step S2 includes: Based on the long short-term learning network LSTM, combined with crop types, real-time environmental data including temperature, precipitation, wind speed, and the calculated water absorption rate and water flow introduction rate, the optimal irrigation demand of the crop under the environmental data is predicted to obtain the predicted irrigation demand. The predicted irrigation demand includes crop water requirement, soil moisture, temperature, and air humidity, and the growth status of the crop includes the crop's absorption rate of water.

8. The AI-based digital twin irrigation area automatic irrigation method according to claim 1 is characterized in that: The reinforcement learning in step S3 includes: Constructing the state vector , describing the current environmental status including soil moisture, crop growth status, temperature and air humidity, and predicted irrigation needs; Constructing Action Vectors , this vector is the decision made by the irrigation system at each moment, including irrigation amount and irrigation time; Building the Reward Function , measure the pros and cons of current irrigation strategies; Building a policy function , indicating a given state When you select Action The probability distribution of Construct a value function to represent the state Under this condition, the expected total reward in the future is obtained, and the importance of each state is evaluated, which is expressed as: ； express expectations; is a discount factor used to balance long-term and short-term rewards; is the reward function; T is the maximum duration.

9. The AI-based digital twin irrigation area automatic irrigation method according to claim 1 is characterized in that: The evolutionary algorithm in step S7 includes: Initialize the population: generate initial solutions, each of which represents an irrigation strategy; Fitness evaluation: Calculate the fitness of each initial solution, which is closely related to crop growth and irrigation efficiency; Selection operation: select excellent crop individuals for reproduction based on fitness; Crossover operation: Generate new solutions through crossover operation, simulating the gene recombination process in nature; Mutation operation: mutate the new solution to simulate the gene mutation process in nature; Alternative operation: compare the new solution with the old solution, select the surviving crop individuals according to fitness, and form a new generation of population; Stop condition: When the preset stop condition is reached, the algorithm is terminated and the optimal solution is output.

10. An AI-based digital twin irrigation area automatic irrigation system, used to implement the AI-based digital twin irrigation area automatic irrigation method according to any one of claims 1 to 9, characterized in that: The functional modules of the system include data acquisition and sensing module, digital twin modeling and simulation module, irrigation demand prediction module, reinforcement learning decision system module, irrigation execution and automation control module, crop growth assessment and analysis module, irrigation strategy optimization and adjustment module, irrigation system maintenance and update module, and water resources management and optimization scheduling module.

Images