Agricultural irrigation control method, system and equipment driven by AI large model and medium

Abstract

The invention provides an agricultural irrigation control method, system and device driven by an AI large model, and a medium, and belongs to the technical field of agricultural irrigation control, and the method comprises the steps: building an irrigation knowledge base and an irrigation decision set for a target crop based on the AI large model; constructing a target crop parameter set and a water demand prediction model, and training the water demand prediction model by using the target crop parameter set in combination with the irrigation knowledge base; target crop parameters collected in real time are input into the trained water demand prediction model, and target crop water demand prediction is carried out; and screening out an irrigation strategy from the irrigation knowledge base and the irrigation decision set according to the predicted value of the water demand of the target crop, and controlling the irrigation equipment based on the irrigation strategy. Through the combination of the AI large model and the deep learning model, the water demand of the crops is accurately predicted, the irrigation strategy is generated, the irrigation decision accuracy is improved, the waste of water resources is reduced, and the irrigation efficiency is improved.

Description

Technical Field

[0001] This application belongs to the field of agricultural irrigation control technology, and specifically relates to an agricultural irrigation control method, system, equipment and medium driven by an AI large model. Background Art

[0002] With the development of agriculture, agricultural irrigation plays a vital role in improving crop yield and quality. Traditional irrigation methods, such as flooding, not only waste significant water resources but can also lead to problems such as soil compaction and oxygen deprivation in crop roots. With the advancement of modern agriculture, intelligent irrigation control systems based on sensor data collection and pre-set irrigation rules are gaining popularity. Existing intelligent irrigation control systems use soil moisture sensors to monitor soil moisture in real time, utilize weather stations to obtain meteorological data, and control the opening and closing of irrigation equipment according to pre-set irrigation rules. However, this approach has numerous drawbacks.

[0003] First, irrigation demand forecasting relies primarily on pre-set rules, lacking the ability to dynamically respond to environmental factors and crop growth status. Environmental factors such as temperature, sunlight, wind speed, and precipitation are constantly changing, and climatic conditions vary significantly across regions. For example, during high temperatures in summer, crop transpiration intensifies, significantly increasing water demand. However, pre-set rules make it difficult to accurately and timely adjust irrigation demand based on these temperature fluctuations.

[0004] Secondly, crops have vastly different water requirements at different growth stages. From the seedling stage to maturity, their root development, plant size, and physiological metabolic activities are constantly changing. Fixed preset rules cannot adapt to these dynamic changes, resulting in large deviations between irrigation demand forecasts and actual demand. Furthermore, different crop types have unique water requirements. Even different varieties of the same crop may have different water requirements and irrigation method preferences at the same growth stage. Furthermore, soil conditions, including soil texture, fertility, and water holding capacity, vary across different plots and even across different areas of the same plot. Existing, relatively fixed irrigation strategies struggle to fully account for these diverse factors and cannot be flexibly adjusted according to actual conditions, resulting in poor irrigation results and an inability to fully meet the needs of crop growth, which in turn affects crop yield and quality. Summary of the Invention

[0005] In a first aspect, an embodiment of the present application provides an agricultural irrigation control method driven by an AI large model, comprising the following steps: S1. Build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; S2. Construct a target crop parameter set and a water requirement prediction model, and train the water requirement prediction model using the target crop parameter set combined with an irrigation knowledge base; S3. Input the real-time collected target crop parameters into the trained water demand prediction model to predict the water demand of the target crop; S4. According to the predicted value of the water requirement of the target crop, an irrigation strategy is selected from the irrigation knowledge base and the irrigation decision set, and the irrigation equipment is controlled based on the irrigation strategy.

[0006] Furthermore, the specific steps of step S1 are as follows: S11. Determine the target crop, target historical period, target region, target crop parameters, soil moisture, weather data, and irrigation operation data to construct a query instruction; S12. Input the query command into the AI ​​big model to initiate the collection of data related to target crop parameters for the target crop in the target area and within the target historical period, as well as the collection of knowledge documents related to target crop parameters, soil moisture, meteorological data, and irrigation operation data; S13. The AI ​​big model analyzes the relationship between the target crop water requirement and the target crop parameters, soil moisture, meteorological data, and irrigation operation data based on the collected data and knowledge documents, generates an irrigation knowledge base and irrigation decision set, and returns it.

[0007] Furthermore, the specific steps of step S2 are as follows: S21. Determine the target irrigation range and target time period, and deploy sensors within the target irrigation range to collect target crop parameters, soil moisture, and meteorological data within the target time period to construct a target crop parameter set; S22. Build a water demand prediction model driven by the irrigation knowledge base and train it using the target crop parameter set. Adjust the parameters of the water demand prediction model based on the error between the water demand prediction value output by the water demand prediction model and the water demand analysis value in the irrigation knowledge base until the error meets the requirements.

[0008] Furthermore, the specific steps of step S22 are as follows: S221. Constructing a water demand prediction model based on convolutional neural networks, bidirectional long short-term memory networks, and attention mechanism; S222. Use the target crop parameter set to train the water demand prediction model, calculate the error between the water demand prediction value output by the water demand prediction model and the water demand analysis value in the corresponding irrigation knowledge base, and define the error as a loss function. Then adjust the parameters of the water demand prediction model according to the loss function until the training is completed.

[0009] Furthermore, the specific steps of step S3 are as follows: S31. Use sensors to collect real-time target crop parameters, soil moisture, and weather data within the target irrigation range; S32. Input the real-time target crop parameters, soil moisture, and meteorological data into the water demand prediction model to obtain the predicted water demand value of the target crop in the set time period.

[0010] Furthermore, the specific steps of step S4 are as follows: S41. Filtering a corresponding irrigation mode from an irrigation knowledge base based on the predicted water demand value; S42. Based on the water demand forecast and real-time soil moisture and meteorological data, query the water demand correction coefficient from the irrigation decision set, and then select the optimal irrigation strategy from the irrigation decision set based on the target crop parameters; S43. Analyze the optimal irrigation strategy to obtain irrigation operation data, generate control instructions based on the irrigation operation data, and issue the control instructions to the irrigation equipment corresponding to the irrigation mode to execute irrigation.

[0011] Furthermore, the specific steps of step S43 are as follows: The specific steps of step S43 are as follows: S431. Analyze the optimal irrigation strategy to obtain irrigation operation data and target soil moisture; the irrigation operation data includes irrigation time, irrigation amount and irrigation method; S432. Based on the target soil moisture and real-time soil moisture data, use the PID controller to dynamically adjust the operating parameters of the irrigation equipment and calculate the control output value of the PID controller; S433. The calculated PID control output value is converted into a control instruction for the irrigation device, so that the irrigation device performs irrigation according to the set irrigation amount and irrigation time; S434. Send control instructions to the irrigation equipment corresponding to the irrigation mode to perform irrigation operations and monitor the soil moisture and operating status of the irrigation equipment during the irrigation process in real time; S435. Feedback the execution results of the irrigation operation to the AI ​​big model, evaluate the execution effect of the irrigation strategy, and update and optimize the irrigation knowledge base and irrigation strategy set based on the execution effect of the irrigation strategy.

[0012] In a second aspect, the embodiments of the present application further provide an agricultural irrigation control system driven by an AI large model, comprising: Irrigation knowledge and strategy collection module, used to build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; The water demand prediction model construction module is used to construct a target crop parameter set and a water demand prediction model, and use the target crop parameter set combined with the irrigation knowledge base to train the water demand prediction model; The water demand prediction module is used to input the real-time collected target crop parameters into the trained water demand prediction model to predict the water demand of the target crop; The irrigation control module is used to select irrigation strategies from the irrigation knowledge base and irrigation decision set according to the predicted value of target crop water demand, and control the irrigation equipment based on the irrigation strategy.

[0013] In a third aspect, an embodiment of the present application further provides an electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the high-reuse method of the agricultural irrigation control method driven by the AI ​​large model as described in the first aspect are implemented.

[0014] In a fourth aspect, an embodiment of the present application further provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the agricultural irrigation control method driven by the AI ​​large model as described in the first aspect are implemented.

[0015] It can be seen from the above technical solutions that this application has the following advantages: The AI ​​big model-driven agricultural irrigation control method, system, equipment and medium provided in this application use the AI ​​big model to integrate a large amount of data to construct an irrigation knowledge base and decision set. Combined with the water demand prediction model, it can accurately predict the water demand of crops, meet the irrigation strategy required by crop growth, avoid excessive or insufficient irrigation, and improve the efficiency of water resource utilization; it can adjust the irrigation amount according to real-time weather and soil moisture to meet the water demand of crops at different growth stages and ensure crop growth; through feedback on irrigation execution results, it can continuously optimize the irrigation knowledge base and decision set, improve the accuracy and adaptability of irrigation control, and reduce the cost of manual intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the present application, the following is a brief introduction to the drawings required for the description. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0017] Figure 1 This is a flow chart of the agricultural irrigation control method driven by the AI ​​large model of the present invention.

[0018] Figure 2 This is a flow chart of the agricultural irrigation control system driven by the AI ​​large model of the present invention. DETAILED DESCRIPTION

[0019] The specific steps of the AI ​​large model-driven agricultural irrigation control method will be described in detail below, and various embodiments of the present disclosure will be described more comprehensively. The present disclosure can have various embodiments, and adjustments and changes can be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but rather that the present disclosure should be understood to cover all adjustments, equivalents, and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0020] For example, with the development of agriculture, agricultural irrigation plays a vital role in ensuring improved crop yields and quality. Traditional irrigation methods, such as flooding, not only result in significant water waste but can also cause a range of problems, such as soil hardening and impeded crop root respiration. Modern agriculture has seen the emergence of intelligent irrigation control systems, which rely on sensor data collection and pre-set irrigation logic. However, despite these advances, current intelligent irrigation control systems still face multiple challenges.

[0021] The primary problem lies in the mechanisms used to predict irrigation demand. These systems rely primarily on pre-set rules and lack the ability to respond sensitively to environmental variables and crop growth dynamics. Environmental factors such as temperature, light intensity, wind speed, and precipitation are constantly changing, and climatic conditions vary significantly across regions. For example, during high temperatures in summer, crop transpiration rates accelerate, leading to a surge in water demand. However, pre-set rules often fail to accurately and timely adjust irrigation strategies based on temperature fluctuations. Secondly, crops exhibit distinct water requirements at different stages of their life cycle. From seedling to maturity, root development, plant size, and physiological metabolic activity undergo significant changes. Fixed pre-set rules struggle to adapt to these dynamics, resulting in significant discrepancies between predicted and actual irrigation demand. Furthermore, different crop species and varieties exhibit unique water requirements. Even different varieties of the same crop may exhibit different water requirements and irrigation method preferences at the same growth stage. Furthermore, soil conditions are highly diverse, including soil type, fertility, and water-holding capacity. These conditions can vary across different areas of the same farmland and even between adjacent fields. The current relatively rigid irrigation strategies are unable to fully consider these diverse factors and cannot be flexibly adjusted according to actual conditions, resulting in low irrigation efficiency and failure to meet the actual needs of crop growth, which in turn has an adverse impact on crop yield and quality.

[0022] In summary, the current intelligent irrigation control system still has shortcomings in the accurate prediction and flexible adjustment of irrigation demand, and further technological innovation and optimization are needed.

[0023] To address the above issues, this embodiment provides an agricultural irrigation control method driven by an AI large model. By combining the AI ​​large model with a deep learning model, it accurately predicts crop water requirements and generates irrigation strategies, thereby improving the accuracy of irrigation decisions, reducing water resource waste, and improving irrigation efficiency. It also achieves automated management of the entire irrigation process, reduces the cost of manual intervention, and improves the intelligence level of the irrigation system.

[0024] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0025] See also Figure 1 FIG. 1 is a flow chart of an agricultural irrigation control method driven by an AI large model in a specific embodiment, the method comprising the following steps: S1. Build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; It should be noted that the AI ​​model integrates multi-source information such as target crop parameters, soil moisture, meteorological data, and irrigation operation data, identifies relationships between data, and provides a knowledge base for irrigation decision-making. It also generates knowledge bases and decision sets for different target crops, regions, and historical periods, enabling irrigation decisions to adapt to diverse agricultural production scenarios. S2. Construct a target crop parameter set and a water requirement prediction model, and train the water requirement prediction model using the target crop parameter set combined with an irrigation knowledge base; It should be noted that by collecting data within the target irrigation range and time period to construct a parameter set, adaptive training data is provided for the water demand prediction model, improving the model's accuracy. Combining the irrigation knowledge base with the training model enables the model to learn from historical experience and knowledge, optimize the prediction effect, and adapt to the growth needs of different crops. S3. Input the real-time collected target crop parameters into the trained water demand prediction model to predict the water demand of the target crop; It should be noted that water demand forecasting based on real-time data can promptly reflect the actual needs of crops and provide a basis for irrigation decisions. By adapting to the dynamic changes in crop growth, water demand forecasts can be adjusted in a timely manner to ensure that crops receive appropriate water supply at different growth stages. S4. Based on the predicted water requirement of the target crop, select an irrigation strategy from the irrigation knowledge base and the irrigation decision set, and control the irrigation equipment based on the irrigation strategy; It should be noted that according to the predicted water demand value, combined with the factors in the knowledge base and decision set, the optimal irrigation strategy is screened out to improve the rationality of irrigation; according to the irrigation strategy, control instructions are generated to control the irrigation equipment to ensure the accurate implementation of irrigation operations, which can achieve water-saving irrigation and ensure crop growth.

[0026] This embodiment can more accurately predict the water requirements of crops and dynamically adjust irrigation strategies based on real-time environmental conditions and crop growth status, thereby improving the efficiency of water resource utilization, ensuring the healthy growth of crops, and realizing intelligent control of the agricultural irrigation process.

[0027] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another AI large model-driven agricultural irrigation control method is provided, which includes the following steps: S1. Build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; The specific steps of step S1 are as follows: S11. Determine the target crop, target historical period, target region, target crop parameters, soil moisture, weather data, and irrigation operation data to construct a query instruction; Irrigation operation data include irrigation time, irrigation amount and irrigation method; S12. Input the query command into the AI ​​big model to initiate the collection of data related to target crop parameters for the target crop in the target area and within the target historical period, as well as the collection of knowledge documents related to target crop parameters, soil moisture, meteorological data, and irrigation operation data; S13. The AI ​​model analyzes the relationship between target crop water requirements and target crop parameters, soil moisture, meteorological data, and irrigation operation data based on collected data and knowledge documents, and generates and returns an irrigation knowledge base and irrigation decision set. The irrigation knowledge base contains the correlation between target crop parameters, soil moisture, meteorological data and target crop water requirements and irrigation strategies; The irrigation strategy set includes the calculation constraints of target crop water requirements and the selection method of irrigation strategy; S2. Construct a target crop parameter set and a water requirement prediction model, and train the water requirement prediction model using the target crop parameter set combined with an irrigation knowledge base; The specific steps of step S2 are as follows: S21. Determine the target irrigation range and target time period, and deploy sensors within the target irrigation range to collect target crop parameters, soil moisture, and meteorological data within the target time period to construct a target crop parameter set; S22. Build a water demand prediction model driven by the irrigation knowledge base and train it using the target crop parameter set. Adjust the parameters of the water demand prediction model based on the error between the water demand prediction value output by the water demand prediction model and the water demand analysis value in the irrigation knowledge base until the error meets the requirements. S3. Input the real-time collected target crop parameters, soil moisture, and meteorological data into the trained water demand prediction model to predict the water demand of the target crop; The specific steps of step S3 are as follows: S31. Use sensors to collect real-time target crop parameters, soil moisture, and weather data within the target irrigation range; S32. Input the real-time target crop parameters, soil moisture and meteorological data into the water demand prediction model to obtain the predicted water demand of the target crop in the set time period; S4. Based on the predicted water requirement of the target crop, select an irrigation strategy from the irrigation knowledge base and the irrigation decision set, and control the irrigation equipment based on the irrigation strategy; The specific steps of step S4 are as follows: S41. Filtering a corresponding irrigation mode from an irrigation knowledge base based on the predicted water demand value; S42. Based on the water demand forecast and real-time soil moisture and meteorological data, query the water demand correction coefficient from the irrigation decision set, and then select the optimal irrigation strategy from the irrigation decision set based on the target crop parameters; S43. Analyze the optimal irrigation strategy to obtain irrigation operation data, generate control instructions based on the irrigation operation data, and issue the control instructions to the irrigation equipment corresponding to the irrigation mode to execute irrigation.

[0028] In an embodiment of the present invention, based on step S22, a possible embodiment will be given below to illustrate its specific implementation scheme in a non-limiting manner.

[0029] The specific steps of step S22 are as follows: S221. Constructing a water demand prediction model based on convolutional neural networks, bidirectional long short-term memory networks, and attention mechanism; Specifically, the water demand prediction model is constructed by sequentially connecting a convolutional neural network (CNN) layer, a bidirectional long short-term memory (BiLSTM) layer, an attention mechanism (Attention) layer, and an output layer. The convolutional neural network (CNN) layer extracts local features from the input target crop parameter set and constructs a feature map of the target crop parameters. Taking the three-dimensional target crop parameter set X as an example, , where R represents the target crop parameter set, N is the number of samples, T is the time step, and F is the feature dimension. The feature map C is obtained by the following convolution operation, and ;

[0030] Among them, M is the height of the convolution kernel, N is the width of the convolution kernel, is the convolution kernel, and , b is the bias term, and , k represents the input feature dimension index, l represents the output feature dimension index; The BiLSTM layer receives the feature map C output by the convolutional neural network CNN layer, performs time series modeling to obtain time series data, and uses the forward LSTM as a hidden layer to capture the previous information of the current data in the sequence data and the backward LSTM as a hidden layer to capture the subsequent information in the sequence data. The previous information, current data, and subsequent information are then concatenated to obtain the output of the BiLSTM layer. ;

[0031] in, , is the dimension of the hidden layer; The attention mechanism Attention layer performs weighted aggregation on the output features of the convolutional neural network (CNN) layer and the bidirectional long short-term memory (BiLSTM) layer;

[0032]

[0033] in, is the attention weight, , is the unnormalized attention score, 、 、 is the weight matrix, is the bias term, C is the output of the convolutional neural network CNN layer, is the output of the BiLSTM layer of the bidirectional long short-term memory network, is the length of the time series; The output of the final attention layer is as follows:

[0034] in ; The output layer is the water demand prediction value of the attention mechanism Attention layer through a fully connected layer ;

[0035] in, is the weight of the output layer, is the bias of the output layer, , ; S222. Train the water demand prediction model using the target crop parameter set, calculate the error between the water demand prediction value output by the water demand prediction model and the water demand analysis value in the corresponding irrigation knowledge base, define the error as a loss function, and then adjust the parameters of the water demand prediction model according to the loss function until the training is completed; Specifically, determine whether the loss function value is less than the set error threshold or reaches the set number of iterations; If not, calculate the gradient of the loss function with respect to the water demand prediction model, adjust the learning rate of the training process according to the gradient, and continue iterative training using the updated learning rate; If yes, complete the training; Specifically, the loss function is defined as follows:

[0036] in, is the mean square error, MAE is the mean absolute error, is the weight coefficient of different growth stages of target crops, 、 、 is the weight parameter of the loss function.

[0037] In an embodiment of the present invention, based on step S43, a possible embodiment will be given below to illustrate its specific implementation scheme in a non-limiting manner.

[0038] The specific steps of step S43 are as follows: S431. Analyze the optimal irrigation strategy to obtain irrigation operation data and target soil moisture; the irrigation operation data includes irrigation time, irrigation amount and irrigation method; S432. Based on the target soil moisture and real-time soil moisture data, use the PID controller to dynamically adjust the operating parameters of the irrigation equipment and calculate the control output value of the PID controller; Specifically, the PID algorithm is used according to the real-time soil moisture and target soil moisture Dynamically adjust the operating status of irrigation equipment, such as valve opening, PID controller output as follows:

[0039]

[0040] in, is the proportionality coefficient, is the integration coefficient, is the differential coefficient; S433. The calculated PID control output value is converted into a control instruction for the irrigation device, so that the irrigation device performs irrigation according to the set irrigation amount and irrigation time; For example, if the irrigation equipment controls the flow rate through a valve, the output of the PID controller is Mapped to the valve opening range [0,1], the valve opening is obtained :

[0041] in, is the minimum value of the PID controller output, is the maximum value of the PID controller output; S434. Send control instructions to the irrigation equipment corresponding to the irrigation mode to perform irrigation operations and monitor the soil moisture and operating status of the irrigation equipment during the irrigation process in real time; S435. Feedback the execution results of the irrigation operation to the AI ​​big model, evaluate the execution effect of the irrigation strategy, and update and optimize the irrigation knowledge base and irrigation strategy set based on the execution effect of the irrigation strategy.

[0042] Take the example of intelligent irrigation control for a field of corn in a certain area: First, determine the query instruction, with corn as the target crop, the past 10 years as the target historical period, the farmland in the region as the target area, and plant physiological indicators in the same time period as the target crop parameters, such as stem flow, plant height, leaf area, stem diameter, and biomass. Soil moisture at different depths, such as 10 cm, 20 cm, and 30 cm, is collected. Meteorological data for the same time period, such as temperature, humidity, light intensity, wind speed, and precipitation, is collected. Irrigation time, irrigation amount, and irrigation method are determined as irrigation operation data. Irrigation time needs to be specific to a time point, and irrigation methods include drip irrigation and sprinkler irrigation. The query quality can be constructed as follows: "Query the stem flow, plant height, leaf area, stem diameter, biomass, soil moisture at different depths, meteorological data (temperature, humidity, light intensity, wind speed, precipitation), and related data and knowledge documents such as irrigation time, irrigation amount, and irrigation method for corn in XXX area over the past 10 years." The query is input into the AI ​​model, which then begins collecting data on target crop parameters for corn in the target area and within the target historical period from multiple sources, including weather station databases, agricultural research institute data, and farm historical records. The model also collects relevant knowledge documents, such as research reports on corn irrigation by agricultural experts. The AI ​​model analyzed the collected data and knowledge documents and found that corn's water demand increases significantly during the jointing stage when the temperature is above 30°C and the wind speed is greater than 3m / s; when soil moisture is low, drip irrigation can more effectively utilize water resources, etc. The generated irrigation knowledge base contains the water demand and corresponding irrigation strategies for corn at different growth stages, different meteorological conditions, and soil moisture. For example, in the hot and rainy season, when the soil moisture of a certain variety of corn is below a certain threshold during the jointing stage, the water demand will increase significantly, and drip irrigation is suitable. The irrigation decision set includes calculation constraints for water demand, such as correction coefficients for meteorological factors and irrigation strategy selection methods, such as selecting irrigation methods based on soil moisture and crop growth stage; determining the correction coefficient for water demand calculation based on different soil types and meteorological conditions; and formulating a priority order for irrigation strategies based on crop growth stage and meteorological disaster risks. Soil sensors, weather stations, and stalk flow sensors were deployed in this corn-growing field to collect data on stalk flow, plant height, leaf area, soil moisture at different depths, temperature, humidity, light intensity, wind speed, and precipitation over the past week to construct a target crop parameter set. After building the water demand prediction model, the target crop parameter set is used for training. The error between the water demand prediction value output by the model and the water demand analysis value in the irrigation knowledge base is calculated, and the error is defined as the loss function. For example, the error threshold is set to 0.01, and the number of iterations is capped at 100. During training, if the loss function value is greater than 0.01 and the number of iterations has not reached 100, the gradient of the loss function with respect to the model is calculated, the learning rate is adjusted based on the gradient, and iterative training continues. If the loss function value is less than 0.01 or the number of iterations has reached 100, the training is completed. Sensors are used to collect real-time data on corn stem flow, plant height, leaf area, soil moisture at different depths, temperature, humidity, light intensity, wind speed, and precipitation. This real-time data is input into a trained water demand prediction model to obtain a predicted water demand for corn in the coming week, assuming the predicted value is 500 cubic meters. According to the predicted water demand value of 500 cubic meters, the corresponding irrigation mode is selected from the irrigation knowledge base, for example, the drip irrigation mode is selected; Based on the predicted water demand of 500 cubic meters, combined with real-time soil moisture and meteorological data, the water demand correction coefficient is queried from the irrigation decision set. Assuming the correction coefficient is 1.1, the corrected water demand is 550 cubic meters. Then, based on target crop parameters such as corn plant height and leaf area, the optimal irrigation strategy is selected from the irrigation decision set. The irrigation time is determined to be from 10 pm to 2 am every day, to be completed over 5 days, with a daily irrigation volume of 110 cubic meters. The optimal irrigation strategy is analyzed to obtain the irrigation operation data as follows: Irrigation time: 10 pm to 2 am every day; Irrigation volume: 110 cubic meters per day; Irrigation method: drip irrigation; Target soil moisture: 60%; Based on the target soil moisture of 60% and real-time soil moisture data, a PID controller is used to dynamically adjust the operating parameters of the irrigation equipment, such as the flow rate and pressure of the water pump. For example, if the real-time soil moisture is 50%, the control output value of the PID controller is calculated; The calculated PID control output value is converted into control instructions for the irrigation equipment, such as the start time of the water pump and the opening degree of the valve, so that the irrigation equipment irrigates according to the set irrigation amount and irrigation time; control instructions are issued to the drip irrigation equipment to execute the irrigation operation, and the soil moisture and the operating status of the irrigation equipment are monitored in real time during the irrigation process to ensure that the soil moisture gradually approaches the target soil moisture and the irrigation equipment operates normally; The execution results of irrigation operations, such as actual irrigation volume, soil moisture changes, and crop growth conditions, are fed back to the AI ​​large model to evaluate the execution effect of the irrigation strategy. If it is found that there is a large deviation between the actual water demand and the predicted value, or the irrigation effect is poor, the irrigation knowledge base and irrigation strategy set are updated and optimized based on the execution effect, so that the next irrigation decision will be more accurate.

[0043] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0044] like Figure 2 As shown, the following is an embodiment of the agricultural irrigation control system driven by an AI large model provided by the embodiment of the present disclosure. This system and the agricultural irrigation control method driven by the AI ​​large model in the above embodiments belong to the same inventive concept. For details not fully described in the embodiment of the agricultural irrigation control system driven by the AI ​​large model, please refer to the embodiment of the agricultural irrigation control method driven by the above AI large model.

[0045] The system includes: Irrigation knowledge and strategy collection module, used to build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; The water demand prediction model construction module is used to construct a target crop parameter set and a water demand prediction model, and use the target crop parameter set combined with the irrigation knowledge base to train the water demand prediction model; The water demand prediction module is used to input the real-time collected target crop parameters into the trained water demand prediction model to predict the water demand of the target crop; The irrigation control module is used to select irrigation strategies from the irrigation knowledge base and irrigation decision set according to the predicted value of target crop water demand, and control the irrigation equipment based on the irrigation strategy.

[0046] This embodiment realizes intelligent irrigation control by integrating an irrigation knowledge and strategy collection module, a water demand prediction model construction module, a water demand prediction module and an irrigation control module.

[0047] The AI ​​large model-driven agricultural irrigation control method provided in the embodiment of the present application can be applied to electronic devices. Those skilled in the art will understand that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently. In the embodiment of the present invention, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0048] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, a button, a camera, a display, and a SIM card interface, etc.

[0049] It is understood that the structures illustrated in the embodiments of the present application do not constitute specific limitations on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown, or combine or separate certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0050] A processor may include one or more processing units, such as a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU). Different processing units may be independent devices or integrated into one or more processors.

[0051] The processor can be the nerve center and command center of the electronic device. The controller can generate operation control signals based on the instruction opcode and timing signal to complete the control of instruction fetching and execution.

[0052] The processor may also include a memory for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. This memory can store instructions or data that the processor has just used or is reusing. If the processor needs to use the instruction or data again, it can directly call it from the memory. This avoids repeated accesses, reduces processor latency, and thus improves system efficiency.

[0053] The above-mentioned electronic device implements the agricultural irrigation control method driven by the AI ​​big model of the present application, which constructs an irrigation knowledge base and an irrigation decision set for the target crop based on the AI ​​big model; constructs a target crop parameter set and a water demand prediction model, and uses the target crop parameter set in combination with the irrigation knowledge base to train the water demand prediction model; inputs the target crop parameters collected in real time into the trained water demand prediction model to predict the water demand of the target crop; based on the predicted value of the water demand of the target crop, screens out irrigation strategies from the irrigation knowledge base and the irrigation decision set, and controls the irrigation equipment based on the irrigation strategy. This technical solution achieves the beneficial effect of being able to more accurately predict the water demand of crops, and dynamically adjust the irrigation strategy according to real-time environmental conditions and crop growth status, thereby improving the utilization efficiency of water resources.

[0054] The storage medium provided in this application stores a program product that can implement an agricultural irrigation control method driven by an AI large model.

[0055] The agricultural irrigation control method driven by the AI ​​big model includes: constructing an irrigation knowledge base and an irrigation decision set for the target crop based on the AI ​​big model; constructing a target crop parameter set and a water demand prediction model, and using the target crop parameter set in combination with the irrigation knowledge base to train the water demand prediction model; inputting the target crop parameters collected in real time into the trained water demand prediction model to predict the water demand of the target crop; according to the predicted water demand value of the target crop, screening the irrigation strategy from the irrigation knowledge base and the irrigation decision set, and controlling the irrigation equipment based on the irrigation strategy.

[0056] In some possible embodiments, the AI ​​large model-driven agricultural irrigation control method disclosed herein can be implemented in the form of a program product, which includes program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of various exemplary embodiments of the present disclosure described in the above "Exemplary Method" section of this specification.

[0057] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0058] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An AI large model driven agricultural irrigation control method, characterized in that: The steps include: S1. Build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; S2. Construct a target crop parameter set and a water requirement prediction model, and train the water requirement prediction model using the target crop parameter set combined with an irrigation knowledge base; S3. Input the real-time collected target crop parameters into the trained water demand prediction model to predict the water demand of the target crop; S4. According to the predicted value of the water requirement of the target crop, an irrigation strategy is selected from the irrigation knowledge base and the irrigation decision set, and the irrigation equipment is controlled based on the irrigation strategy.

2. The AI ​​large model driven agricultural irrigation control method according to claim 1, characterized in that: The specific steps of step S1 are as follows: S11. Determine the target crop, target historical period, target region, target crop parameters, soil moisture, weather data, and irrigation operation data to construct a query instruction; S12. Input the query command into the AI ​​big model to initiate the collection of data related to target crop parameters for the target crop in the target area and within the target historical period, as well as the collection of knowledge documents related to target crop parameters, soil moisture, meteorological data, and irrigation operation data; S13. The AI ​​big model analyzes the relationship between the target crop water requirement and the target crop parameters, soil moisture, meteorological data, and irrigation operation data based on the collected data and knowledge documents, generates an irrigation knowledge base and irrigation decision set, and returns it.

3. The agricultural irrigation control method driven by the AI ​​large model according to claim 1 is characterized in that: The specific steps of step S2 are as follows: S21. Determine the target irrigation range and target time period, and deploy sensors within the target irrigation range to collect target crop parameters, soil moisture, and meteorological data within the target time period to construct a target crop parameter set; S22. Build a water demand prediction model driven by the irrigation knowledge base and train it using the target crop parameter set. Adjust the parameters of the water demand prediction model based on the error between the water demand prediction value output by the water demand prediction model and the water demand analysis value in the irrigation knowledge base until the error meets the requirements.

4. The AI ​​large model driven agricultural irrigation control method according to claim 3 is characterized in that: The specific steps of step S22 are as follows: S221. Constructing a water demand prediction model based on convolutional neural networks, bidirectional long short-term memory networks, and attention mechanism; S222. Use the target crop parameter set to train the water demand prediction model, calculate the error between the water demand prediction value output by the water demand prediction model and the water demand analysis value in the corresponding irrigation knowledge base, and define the error as a loss function. Then adjust the parameters of the water demand prediction model according to the loss function until the training is completed.

5. The AI ​​large model driven agricultural irrigation control method according to claim 1, characterized in that: The specific steps of step S3 are as follows: S31. Use sensors to collect real-time target crop parameters, soil moisture, and weather data within the target irrigation range; S32. Input the real-time target crop parameters, soil moisture, and meteorological data into the water demand prediction model to obtain the predicted water demand value of the target crop in the set time period.

6. The AI ​​large model driven agricultural irrigation control method according to claim 5, characterized in that: The specific steps of step S4 are as follows: S41. Filtering a corresponding irrigation mode from an irrigation knowledge base based on the predicted water demand value; S42. Based on the water demand forecast and real-time soil moisture and meteorological data, query the water demand correction coefficient from the irrigation decision set, and then select the optimal irrigation strategy from the irrigation decision set based on the target crop parameters; S43. Analyze the optimal irrigation strategy to obtain irrigation operation data, generate control instructions based on the irrigation operation data, and issue the control instructions to the irrigation equipment corresponding to the irrigation mode to execute irrigation.

7. The AI ​​large model driven agricultural irrigation control method according to claim 6, characterized in that: The specific steps of step S43 are as follows: S431. Analyze the optimal irrigation strategy to obtain irrigation operation data and target soil moisture; the irrigation operation data includes irrigation time, irrigation amount and irrigation method; S432. Based on the target soil moisture and real-time soil moisture data, use the PID controller to dynamically adjust the operating parameters of the irrigation equipment and calculate the control output value of the PID controller; S433. The calculated PID control output value is converted into a control instruction for the irrigation device, so that the irrigation device performs irrigation according to the set irrigation amount and irrigation time; S434. Send control instructions to the irrigation equipment corresponding to the irrigation mode to perform irrigation operations and monitor the soil moisture and operating status of the irrigation equipment during the irrigation process in real time; S435. Feedback the execution results of the irrigation operation to the AI ​​big model, evaluate the execution effect of the irrigation strategy, and update and optimize the irrigation knowledge base and irrigation strategy set based on the execution effect of the irrigation strategy.

8. An agricultural irrigation control system driven by an AI large model, characterized in that: include: Irrigation knowledge and strategy collection module, used to build an irrigation knowledge base and irrigation decision set for target crops based on the AI ​​big model; The water demand prediction model construction module is used to construct a target crop parameter set and a water demand prediction model, and use the target crop parameter set combined with the irrigation knowledge base to train the water demand prediction model; The water demand prediction module is used to input the real-time collected target crop parameters into the trained water demand prediction model to predict the water demand of the target crop; The irrigation control module is used to select irrigation strategies from the irrigation knowledge base and irrigation decision set according to the predicted value of target crop water demand, and control the irrigation equipment based on the irrigation strategy.

9. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the agricultural irrigation control method driven by the AI ​​large model as claimed in any one of claims 1 to 7 are implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the agricultural irrigation control method driven by the AI ​​large model as described in any one of claims 1 to 7 are implemented.

Images