Intelligent irrigation decision-making method, device and equipment and computer storage medium

Abstract

The embodiment of the invention provides an intelligent irrigation decision-making method, device and equipment and a computer storage medium. The intelligent irrigation decision-making method comprises the following steps: constructing a causal diagram irrigation model; the causal graph irrigation model is a causal graph model used for predicting crop water demand information and soil humidity information at a first time point according to a causal relationship between nodes; the first time point is a time point of a first time period after the current time point; based on the acquired soil humidity information, rainfall information and temperature information, utilizing a causal diagram irrigation model to predict the crop water demand information and the soil humidity information at the first time point to obtain a predicted value of the crop water demand information and a predicted value of the soil humidity information at the first time point; and determining the irrigation control information based on the predicted value of the crop water demand information and the predicted value of the soil humidity information at the first time point, so that the consumption of computing resources can be reduced, a dynamically changing agricultural environment can be quickly responded, and the interpretability of an irrigation decision is improved.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the technical field of intelligent irrigation, and particularly relates to an intelligent irrigation decision method, device, equipment and computer storage medium. BACKGROUND

[0002] With the continuous development of smart agriculture, the application of intelligent irrigation technology is becoming more and more widespread. At present, the research and application of intelligent irrigation technology are continuously deepening, aiming to improve the efficiency of water resource utilization and crop yield. Through multi-source data collection, model prediction and algorithm optimization, intelligent precision irrigation can be realized. SUMMARY

[0003] The embodiments of the present disclosure provide an intelligent irrigation decision method, device, equipment and computer storage medium, which can reduce the consumption of computing resources, quickly respond to dynamic changes in the agricultural environment, and improve the explainability of irrigation decision.

[0004] In a first aspect, the embodiments of the present disclosure provide an intelligent irrigation decision method, which comprises: constructing a causal diagram irrigation model; the causal diagram irrigation model is a causal diagram model for predicting crop water requirement information and soil moisture information at a first time point according to the causal relationship between the input nodes, intermediate nodes and output nodes in the causal diagram irrigation model; the first time point is a time point after a first time period from the current time point; based on the obtained soil moisture information, rainfall information and temperature information, predicting the crop water requirement information and the soil moisture information at the first time point by using the causal diagram irrigation model to obtain a predicted value of the crop water requirement information and a predicted value of the soil moisture information at the first time point; determining irrigation control information based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point.

[0005] In an implementable embodiment, constructing the causal diagram irrigation model comprises: constructing a soil moisture change equation based on rainfall variables, crop water requirement variables and temperature variables; constructing a crop water requirement calculation equation based on potential evapotranspiration variables and crop growth stage coefficients; constructing the causal diagram irrigation model based on the soil moisture change equation and the crop water requirement equation; wherein the causal diagram irrigation model is a causal diagram model with rainfall nodes and temperature nodes as input nodes, a soil moisture node as a state node, and a crop water requirement node as an output node.

[0006] In an implementable embodiment, the rainfall variable is rainfall information at a second time point; the temperature variable is temperature information at a third time point; the crop water requirement variable is crop water requirement information at a fourth time point; the soil moisture variable in the crop water requirement calculation equation is crop water requirement information at a fifth time point; and the second time point, the third time point, the fourth time point and the fifth time point are all time points before the current time point.

[0007] In an implementable embodiment, the first weight coefficient is a weight coefficient of the rainfall variable; the second weight coefficient is a weight coefficient of the crop water requirement variable; and the third weight coefficient is a weight coefficient of the temperature variable; and the method further comprises dynamically updating the first weight coefficient, the second weight coefficient and the third weight coefficient.

[0008] In an implementable embodiment, the method further comprises: performing dynamic pruning processing on the causal diagram irrigation model to obtain a pruned causal diagram irrigation model; performing weight distillation processing on the pruned causal diagram irrigation model to obtain a compressed causal diagram irrigation model; and updating the causal diagram irrigation model according to the compressed causal diagram irrigation model.

[0009] In an implementable embodiment, the dynamic pruning processing on the causal diagram irrigation model to obtain a pruned causal diagram irrigation model comprises: calculating a causal contribution degree of each node in the causal diagram irrigation model; determining a redundancy degree of each sub-tree in the causal diagram irrigation model according to the causal contribution degree of each node in the sub-tree; and performing pruning processing on the sub-tree to obtain a pruned causal diagram irrigation model when the redundancy degree is greater than a first pruning threshold.

[0010] In an implementable embodiment, the method further comprises: dynamically updating the first pruning threshold according to a crop growth stage coefficient; and an initial preset value of the first pruning threshold is determined according to a crop type.

[0011] In an implementable embodiment, the calculation of the causal contribution degree of the node comprises: multiplying the weights of all edges on a first path to obtain a weight of the first path; the first path is a path from an input node to an output node and passing through the node; and summing the weights of all first paths in the causal diagram irrigation model to obtain the causal contribution degree of the node.

[0012] In an implementable embodiment, the method further comprises: scoring an irrigation result of irrigation according to the irrigation control information to obtain an irrigation effect score; and visually displaying the irrigation result and the irrigation effect score.

[0013] In an implementable embodiment, the irrigation result of irrigation according to the irrigation control information is scored to obtain an irrigation effect score, including: determining soil humidity change information according to the soil humidity information after irrigation and the target soil humidity information; and calculating the irrigation effect score according to the normalized difference vegetation index and the soil humidity change information.

[0014] In a second aspect, the embodiments of the present disclosure provide an intelligent irrigation decision device, which comprises: a construction module configured to construct a causal diagram irrigation model; the causal diagram irrigation model is a causal diagram model configured to predict crop water requirement information and soil humidity information at a first time point according to causal relationships between input nodes, intermediate nodes and output nodes in the causal diagram irrigation model; the first time point is a time point after a first time period from a current time point; a prediction module configured to predict the crop water requirement information and the soil humidity information at the first time point based on the obtained soil humidity information, rainfall information and temperature information by using the causal diagram irrigation model to obtain a predicted value of the crop water requirement information and a predicted value of the soil humidity information at the first time point; a decision module configured to determine irrigation control information based on the predicted value of the crop water requirement information and the predicted value of the soil humidity information at the first time point.

[0015] In a third aspect, the embodiments of the present disclosure provide an electronic device, which comprises a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement any of the intelligent irrigation decision methods in the above embodiments.

[0016] In a fourth aspect, the embodiments of the present disclosure provide a computer storage medium, which stores computer program instructions; the computer program instructions are executed by a processor to implement any of the intelligent irrigation decision methods in the above embodiments.

[0017] In a fifth aspect, the embodiments of the present disclosure provide a computer program product, which comprises a computer program; the computer program is executed by a processor to implement any of the intelligent irrigation decision methods in the above embodiments.

[0018] The intelligent irrigation decision-making method, apparatus, device, and computer storage medium of this disclosure construct a causal graph irrigation model to predict crop water requirement information and soil moisture information at a first time point based on the causal relationships between nodes. Based on acquired soil moisture, rainfall, and temperature information, the causal graph irrigation model is used to predict crop water requirement information and soil moisture information at the first time point, obtaining predicted values ​​for both. Based on these predicted values, irrigation control information is determined, reducing computational resource consumption, enabling rapid response to dynamically changing agricultural environments, and improving the interpretability of irrigation decisions. Constructing a causal graph irrigation model based on soil moisture change equations and crop water requirement equations makes the model more consistent with crop response mechanisms, improving its generalization ability and dynamic adaptability, and enhancing its rapid response capability to dynamic and ever-changing agricultural environments. Dynamic pruning is applied to the causal graph irrigation model to obtain a pruned causal graph irrigation model. Different pruning thresholds can be determined according to crop type, and the pruning thresholds are dynamically updated based on the crop growth stage coefficient. This improves the generalization ability and dynamic adaptability of the causal graph irrigation model. Weight distillation is then performed on the pruned causal graph irrigation model to obtain a compressed causal graph irrigation model. The causal graph irrigation model is updated based on the compressed model. This ensures that the parameter information of nodes with high causal contribution is retained after pruning while compressing the model size, thus maintaining the performance of the causal graph irrigation model, further reducing computational resource consumption, and improving response speed. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the embodiments of this disclosure will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating an intelligent irrigation decision-making method provided in an embodiment of this disclosure; Figure 2 This is a flowchart illustrating an intelligent irrigation decision-making method provided in an embodiment of this disclosure; Figure 3 This is a flowchart illustrating an intelligent irrigation decision-making method provided in an embodiment of this disclosure; Figure 4 This is a flowchart illustrating an intelligent irrigation decision-making method provided in an embodiment of this disclosure; Figure 5 This is a schematic diagram of the structure of an intelligent irrigation decision-making system provided in an embodiment of this disclosure; Figure 6is a structural schematic diagram of an intelligent irrigation decision device provided by an embodiment of the present disclosure. Figure 7 is a structural schematic diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0021] The features and exemplary embodiments of various aspects of the present application will be described below in detail, in order to make the purposes, technical solutions and advantages of the present application more clear and apparent, the present application will be further described in detail below in combination with the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, but not to limit the present application. For those skilled in the art, the present application can be implemented without some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by showing examples of the present application.

[0022] It should be noted that, in this paper, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the elements defined by the statement "include" do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0023] In order to solve the problems in the prior art, the present disclosure provides an intelligent irrigation decision method, device, equipment and computer storage medium.

[0024] First, the intelligent irrigation decision method provided by the present disclosure will be introduced.

[0025] Figure 1 is a flowchart of an intelligent irrigation decision method provided by an embodiment of the present disclosure. As shown in Figure 1 , the method can include the following steps: S101, constructing a causal diagram irrigation model; wherein the causal diagram irrigation model is a causal diagram model for predicting crop water requirement information and soil moisture information at a first time point according to causal relationships between input nodes, intermediate nodes and output nodes in the causal diagram irrigation model; the first time point is a time point after a first time period from the current time point; S102, based on the obtained soil moisture information, rainfall information and temperature information, the crop water requirement information and the soil moisture information at the first time point are predicted by using the causal diagram irrigation model, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; S103, based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point, the irrigation control information is determined.

[0026] It should be noted that the causal diagram model is a model for representing the causal relationship between variables by graphical method. In some embodiments, the causal diagram irrigation model is a directed acyclic graph. Illustratively, the causal diagram irrigation model is a directed acyclic graph with nodes as variables, edges as causal relationships and weights as formula coefficients.

[0027] In some embodiments, the causal diagram irrigation model can calculate the prediction value required for irrigation decision based on real-time input data and state variables to generate irrigation control information to guide crop irrigation.

[0028] In some embodiments, the input nodes in the causal diagram irrigation model include rainfall nodes and temperature nodes; the intermediate nodes include soil moisture nodes, and the output nodes include crop water requirement nodes. Illustratively, the causal diagram irrigation model is a causal diagram model for predicting the crop water requirement information and the soil moisture information at the first time point according to the causal relationship between the rainfall nodes, the temperature nodes, the soil moisture nodes and the crop water requirement nodes in the causal diagram irrigation model.

[0029] In some embodiments, meteorological data is collected by meteorological equipment; soil moisture data is collected by soil moisture equipment; and the meteorological data, the soil moisture data and future precipitation data are preprocessed to obtain the soil moisture information, the rainfall information and the temperature information.

[0030] In some embodiments, the irrigation control information includes irrigation water quantity.

[0031] In some embodiments, a causal diagram irrigation model capable of calculating prediction values according to causal relationships is constructed; based on the obtained soil moisture information, rainfall information and temperature information, the crop water requirement information and the soil moisture information at the first time point are predicted by using the causal diagram irrigation model, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; and based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point, the irrigation control information is determined, which can reduce the consumption of computing resources and quickly respond to the dynamically changing agricultural production environment according to the causal relationship between environmental factors and crop water requirement, thereby improving the explainability of irrigation decision.

[0032] Figure 2is a flowchart of an intelligent irrigation decision-making method provided by an embodiment of the present disclosure. As shown in Figure 2 the method can include the following steps: S201, constructing a soil moisture change equation based on a rainfall variable, a crop water requirement variable, and a temperature variable; S202, constructing a crop water requirement calculation equation based on a potential evapotranspiration variable and a crop growth stage coefficient; S203, constructing a causal diagram irrigation model based on the soil moisture change equation and the crop water requirement equation; S102, based on the obtained soil moisture information, rainfall information, and temperature information, using the causal diagram irrigation model to predict the crop water requirement information and the soil moisture information at the first time point, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; S103, determining irrigation control information based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point.

[0033] It should be noted that in some embodiments, the soil moisture change equation can be used to predict the soil moisture information at a future time.

[0034] In some embodiments, the soil moisture change equation includes a rainfall input term, a crop consumption term, and a temperature term. The rainfall input term includes a rainfall variable, which is used to represent the influence of rainfall input on soil moisture; the crop consumption term includes a crop water requirement variable, which is used to represent the influence of crop consumption on soil moisture; and the temperature term includes a temperature variable, which is used to represent the influence of evaporation on soil moisture.

[0035] For example, the rainfall input term is the product of a first weight coefficient and the rainfall variable; the crop consumption term is the product of a second weight coefficient, the crop water requirement information, and a crop water stress coefficient; the temperature term is the product of a third weight coefficient and the temperature variable; and the soil moisture change equation is constructed based on the current soil moisture information, the rainfall input term, the crop consumption term, and the temperature term.

[0036] For example, the soil moisture change equation can be represented by the following formula (1): (1) wherein, is the predicted value of the soil moisture information at the first time point; is the first time period; is the current soil moisture information; is the current rainfall information; is the first weight coefficient; is the predicted value of the crop water requirement information; is the crop water stress coefficient; is a second weight coefficient; is current temperature information; is a third weight coefficient.

[0037] The first weight coefficient may be a rainfall infiltration coefficient, which can have a value range of 0.6 to 0.8; the second weight coefficient may be a crop water absorption efficiency, which can have a value range of 0.4 to 0.6; and the third weight coefficient may be a temperature-driven evaporation coefficient, which can have a value range of 0.05 to 0.1.

[0038] It should be noted that in the traditional calculation method, the relationship between crop water requirement and soil moisture information is generally assumed to be a linear relationship.

[0039] In some embodiments, the non-linear mutation effect of water stress is expressed in the soil moisture change equation by using an S-shaped function, which can make the causal irrigation model more consistent with the crop response mechanism. Exemplarily, the crop water stress coefficient can be represented by the following formula (2): (2) wherein, is a curvature parameter; is a crop water stress threshold value. Exemplarily, the value of the curvature parameter may be 0.1; and the value of the crop water stress threshold value may be 50%.

[0040] In some embodiments, a crop water requirement calculation equation is constructed based on a potential evapotranspiration variable and a crop growth stage coefficient by using the relationship between the potential evapotranspiration variable and a temperature variable.

[0041] Exemplarily, the relationship between the potential evapotranspiration variable and the temperature variable can be represented by the following formula (3): (3) wherein, is current potential evapotranspiration information.

[0042] Exemplarily, the crop water requirement calculation equation can be represented by the following formula (4): (4) wherein, is a crop growth stage coefficient. Exemplarily, the value range of the crop growth stage coefficient may be 0.3 to 1.2. Among them, at the sowing period of the crop, the crop growth stage coefficient The value of the crop growth stage coefficient can be 0.3; at the mature stage of the crop, the crop growth stage coefficient The value of the crop growth stage coefficient can be 1.2.

[0043] In some embodiments, based on the soil moisture change equation and the crop water requirement equation, a causal diagram irrigation model is constructed. Exemplarily, the causal diagram irrigation model takes a rainfall node and a temperature node as input nodes, a soil moisture node as a state node, and a crop water requirement node as an output node. The causal diagram irrigation model includes a first edge from the rainfall node to the soil moisture node, a second edge from the temperature node to the soil moisture node, a third edge from the soil moisture node to the crop water requirement node, and a fourth edge from the temperature node to the crop water requirement node. That is, the first edge corresponds to the causal relationship that rainfall increases and soil moisture increases; the second edge corresponds to the causal relationship that temperature rises and evaporation accelerates, and soil moisture decreases; the third edge corresponds to the causal relationship that soil moisture affects crop water requirement through water stress; and the fourth edge corresponds to the causal relationship that temperature indirectly affects crop water requirement through potential evapotranspiration.

[0044] It should be noted that the influence of rainfall on soil moisture has a delay effect, and introducing a time lag term in the soil moisture change equation can better conform to the crop response mechanism.

[0045] In some embodiments, a time lag term is added to the rainfall variable, the crop water requirement variable, and the temperature variable in the soil moisture change equation, and the time lag term is used to express the causal delay in time. Exemplarily, the second time point, the third time point, and the fourth time point are all time points before the current time point; the value of the rainfall variable is the rainfall information at the second time point; the value of the temperature variable is the temperature information at the third time point; and the value of the crop water requirement variable is the crop water requirement information at the fourth time point.

[0046] Exemplarily, in some embodiments, the second time point, the third time point, and the fourth time point are all time points of a first time period before the current time point, and the soil moisture change equation after adding the lag term can be represented by the following formula (5): (5) wherein, the rainfall information at the second time point is the temperature information at the third time point is the crop water requirement information at the fourth time point is the crop water stress coefficient at the fourth time point is

[0047] It should be noted that in some embodiments, since the penetration speed of water in different types of soil is different, the value of the time lag term can be adjusted according to the soil type.

[0048] In some embodiments, the second time point, the third time point and the fourth time point are determined according to the soil type. For example, the second time point, the third time point and the fourth time point are all time points one first time period before the current time point; the soil type can include sandy soil, loam and clay; in the case of the soil type being sandy soil, the first time period can be 1 hour, and the second time point, the third time point and the fourth time point are all time points 1 hour before the current time point; in the case of the soil type being loam, the first time period can be 3 hours, and the second time point, the third time point and the fourth time point are all time points 3 hours before the current time point; in the case of the soil type being clay, the first time period can be 6 hours, and the second time point, the third time point and the fourth time point are all time points 6 hours before the current time point.

[0049] It should be noted that the influence of soil moisture on crop water requirement also has a time delay effect, and introducing a time lag term in the crop water requirement calculation equation can be more in line with the crop response mechanism.

[0050] In some embodiments, a time lag term is added to the soil moisture variable in the crop water requirement calculation equation, which is used to express the causal delay in time. For example, the time lag term is a first time period; the fifth time point is a time point one first time period before the current time point; and the value of the soil moisture variable is the rainfall information at the fifth time point. The crop water requirement calculation equation with the lag term can be represented by the following formula (6): (6) Wherein, ) is the soil moisture information at the fifth time point.

[0051] In some embodiments, the irrigation control information includes the irrigation water quantity. For example, the irrigation water quantity can be determined according to the predicted value of the soil moisture information at the first time point. For example, the irrigation water quantity can be calculated by the following formula (7): (7) Wherein, is the target soil moisture information.

[0052] It should be noted that in some embodiments, a dynamic calibration mechanism is used to dynamically adjust the weight coefficients of the formulas in the causal graph irrigation model to further improve the irrigation accuracy.

[0053] In some embodiments, the first weight coefficient, the second weight coefficient and the third weight coefficient are dynamically updated.

[0054] For example, the first weight coefficient is dynamically updated based on the ratio of the historical error of the soil moisture information and the rainfall information by using a sliding window. For example, the dynamic updating of the first weight coefficient can be represented by the following formula (8): (8) wherein, is the updated first weight coefficient at the current time point; is the size of the sliding window, which can be 7 days; is the actual soil moisture information at the time point t-1; is the predicted value of the soil moisture information at the time point t-1.

[0055] For example, the second weight coefficient is processed by mean inverse correction to obtain a corrected second weight coefficient; and the second weight coefficient is updated according to the corrected second weight coefficient. For example, the mean inverse correction of the second weight coefficient can be represented by the following formula (9): (9) wherein, is the updated second weight coefficient at the current time point; is the historical mean of the actual crop water requirement information; is the historical mean of the predicted value of the crop water requirement information.

[0056] For example, the third weight coefficient is processed by mean inverse correction to obtain a corrected third weight coefficient; and the third weight coefficient is updated according to the corrected third weight coefficient. For example, the mean inverse correction of the third weight coefficient can be represented by the following formula (10): (10) wherein, is the updated third weight coefficient at the current time point; is the historical mean of the actual evaporated soil moisture information; is the historical mean of the predicted value of the soil moisture information.

[0057] ​​In some embodiments, based on the rainfall variable, the crop water requirement variable and the temperature variable, a soil moisture change equation is constructed; based on the potential evapotranspiration variable and the crop growth stage coefficient, a crop water requirement calculation equation is constructed; based on the soil moisture change equation and the crop water requirement equation, a causal diagram irrigation model is constructed, which can make the model more in line with the crop response mechanism, improve the generalization ability and dynamic adaptability of the model, and further improve the rapid response capability to the dynamic and variable agricultural environment; and based on the obtained soil moisture information, rainfall information and temperature information, the causal diagram irrigation model is used to predict the crop water requirement information and the soil moisture information at the first time point, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point, the irrigation control information is determined, which can further improve the explainability of the irrigation decision.

[0058] It should be noted that the traditional pruning method may cause misdeletion of key sub-trees due to the neglect of the causal relationship between nodes. According to the crop type, the crop growth stage coefficient and other irrigation related parameters, the redundant logic in the causal diagram irrigation model can be identified, the redundant sub-trees can be removed, and the model complexity can be reduced.

[0059] Figure 3 is a flowchart of an intelligent irrigation decision method provided by the embodiments of the present disclosure. As shown in Figure 3 the method can include the following steps: S101, constructing a causal diagram irrigation model; wherein the causal diagram irrigation model is a causal diagram model for predicting crop water requirement information and soil moisture information at a first time point according to the causal relationship between input nodes, intermediate nodes and output nodes in the causal diagram irrigation model; the first time point is a time point after a first time period from the current time point; S102, based on the obtained soil moisture information, rainfall information and temperature information, using the causal diagram irrigation model to predict the crop water requirement information and the soil moisture information at the first time point, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; S103, determining irrigation control information based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; S301, performing dynamic pruning processing on the causal diagram irrigation model to obtain a pruned causal diagram irrigation model; S302, performing weight distillation processing on the pruned causal diagram irrigation model to obtain a compressed causal diagram irrigation model; S303, updating the causal diagram irrigation model according to the compressed causal diagram irrigation model.

[0060] In some embodiments, the causal graph irrigation model takes a rainfall node and a temperature node as input nodes, a soil moisture node as a state node, and a crop water requirement node as an output node. The causal graph irrigation model includes a first edge from the rainfall node to the soil moisture node, a second edge from the temperature node to the soil moisture node, a third edge from the soil moisture node to the crop water requirement node, and a fourth edge from the temperature node to the crop water requirement node.

[0061] For example, the weight of the first edge from the rainfall node to the soil moisture node can be , i.e., a rainfall infiltration coefficient; the weight of the second edge from the temperature node to the soil moisture node can be ; the weight of the third edge from the soil moisture node to the crop water requirement node can be .

[0062] It should be noted that in some embodiments, the causal contribution of a feature to an irrigation decision is quantified by the weight of the node and the edge in the causal graph irrigation model, which can avoid the misdeletion of key sub-trees caused by the traditional pruning method due to the neglect of causal correlation.

[0063] In some embodiments, for each node in the causal graph irrigation model, the causal contribution of the node is calculated. For example, the first path is a path from the input node to the output node and passing through the node; the weights of all edges in the first path are multiplied to obtain the weight of the first path; and the weights of all first paths in the causal graph irrigation model are summed to obtain the causal contribution of the node.

[0064] For example, the causal contribution of the node can be calculated by the following formula (11): (11) wherein is the causal contribution of the i-th node; is the edge from the i-th node to the j-th node; is the weight of the edge; is the first path; is the set of all paths from the input node to the output node, for example, it can be the set of all paths from the rainfall node or the temperature node to the crop water requirement node.

[0065] ​​​​​​​In some embodiments, after the causal contribution of each node is calculated, for each sub-tree in the causal graph irrigation model, the redundancy of the sub-tree is determined according to the causal contribution of each node in the sub-tree.

[0066] For example, the redundancy of the sub-tree can be calculated by the following formula (12): (12) wherein, is the redundancy of the sub-tree.

[0067] In some embodiments, when the redundancy is greater than the first pruning threshold, the sub-tree is pruned to obtain a pruned causal graph irrigation model. For example, the sub-tree is replaced by a first leaf node, the value of the first leaf node being the historical average of the crop water requirement information; or a lightweight substitute function is constructed, and the node corresponding to the lightweight substitute function is used to replace the sub-tree; or when the sub-tree includes a soil moisture node, a substitute prediction function of soil moisture information is constructed, and the node corresponding to the substitute prediction function is used to replace the sub-tree.

[0068] It should be noted that in some embodiments, the first pruning threshold is dynamically updated to control the pruning intensity. For example, in the early growth stage, the pruning intensity is increased; in the late growth stage, the first pruning threshold is relaxed, and more sub-trees are retained to cope with complex water requirement changes.

[0069] In some embodiments, according to different crop types, an initial preset value of the first pruning threshold is determined; and the first pruning threshold is dynamically updated according to a crop growth stage coefficient. For example, when the crop type is wheat, the first pruning threshold is set to 0.3; when the crop type is fruit trees, the first pruning threshold is set to 0.5, which can reflect the difference in water sensitivity of different crops. When the growth stage of the crop is the seeding stage, the first pruning threshold is updated to 0.3; when the growth stage of the crop is the mature stage, the first pruning threshold is updated to 1.2. For example, the dynamic updating of the first pruning threshold according to the crop growth stage coefficient can be represented by the following formula (13): (13) wherein, is the first pruning threshold; is the initial preset value of the first pruning threshold.

[0070] ​In some embodiments, after obtaining the pruned causal graph irrigation model, a weight distillation process is performed on the pruned causal graph irrigation model to obtain a compressed causal graph irrigation model; the causal graph irrigation model is then updated based on the compressed causal graph irrigation model. For example, a weight distillation loss function is constructed based on the node parameters of each node in the pruned causal graph irrigation model; using this weight distillation loss function, the pruned causal graph irrigation model is subjected to weight distillation to obtain a compressed causal graph irrigation model, and the causal graph irrigation model is updated based on the compressed causal graph irrigation model. This ensures that the parameter information of nodes with high causal contribution is retained after pruning while compressing the model volume, thus maintaining the performance of the causal graph irrigation model. The weight distillation loss function can be expressed by the following formula (14): (14) in, The weighted distillation loss function; The causal diagram irrigation model before pruning was developed in the first... Node parameters of each node; The causal graph irrigation model after pruning is in the first... Node parameters of each node.

[0071] In some embodiments, the causal graph irrigation model is dynamically pruned to obtain a pruned causal graph irrigation model. Different pruning thresholds can be determined according to crop type, and the pruning thresholds are dynamically updated according to crop growth stage coefficients. This improves the generalization ability and dynamic adaptability of the causal graph irrigation model. The pruned causal graph irrigation model is then subjected to weight distillation to obtain a compressed causal graph irrigation model. The causal graph irrigation model is updated according to the compressed causal graph irrigation model. This can compress the model volume while ensuring that the parameter information of nodes with high causal contribution is retained after pruning, so as to maintain the performance of the causal graph irrigation model, further reduce the consumption of computing resources, and improve the response speed.

[0072] Figure 4 This is a flowchart illustrating an intelligent irrigation decision-making method provided in an embodiment of this disclosure. Figure 3 As shown, the method may include the following steps: S101, Construct a causal graph irrigation model; wherein, the causal graph irrigation model is a causal graph model used to predict crop water demand information and soil moisture information at the first time point based on the causal relationship between the input nodes, intermediate nodes and output nodes in the causal graph irrigation model. S102, based on the obtained soil moisture information, rainfall information and temperature information, the crop water requirement information and the soil moisture information at the first time point are predicted by using the causal diagram irrigation model, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point; the first time point is a time point after the current time point; S103, based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point, the irrigation control information is determined; S401, the irrigation result according to the irrigation control information is scored to obtain an irrigation effect score; S402, the irrigation result and the irrigation effect score are visually displayed.

[0073] In some embodiments, the irrigation effect score can be a multi-dimensional score for evaluating the irrigation effect after irrigation according to the irrigation decision. For example, the irrigation effect score can include irrigation efficiency score, water saving rate, crop health index and model stability; the irrigation efficiency score is used to comprehensively evaluate the irrigation effect, which is added when the soil moisture meets the standard and is reduced when the energy consumption is too high; the water saving rate is used to evaluate the water saving proportion compared with the traditional irrigation calculation method; the crop health score is used to track the long-term growth trend and evaluate the long-term impact of the irrigation decision on crop growth; the model stability is used to evaluate the negative impact of dynamic pruning on model accuracy.

[0074] In some embodiments, according to the soil moisture information after irrigation and the target soil moisture information, the soil moisture change information is determined; according to the normalized difference vegetation index and the soil moisture change information, the irrigation effect score is calculated. For example, the irrigation effect score is the irrigation efficiency score; according to the soil moisture information after irrigation and the target soil moisture information, the soil moisture change information is determined; according to the normalized difference vegetation index and the soil moisture change information, the irrigation efficiency score is calculated.

[0075] For example, the irrigation efficiency score can be calculated according to the following formula (15): (15) Wherein, is the irrigation efficiency score; is the soil moisture change information, which is the value obtained by subtracting the target soil moisture information from the soil moisture information after irrigation; is the reference value of the normalized difference vegetation index; is the normalized difference vegetation index before irrigation; is the normalized difference vegetation index after irrigation; is the weight coefficient of the normalized difference vegetation index.

[0076] In some embodiments, the weight coefficient of the normalized difference vegetation index is dynamically updated according to a crop growth stage coefficient. Illustratively, the weight coefficient of the normalized difference vegetation index can be updated according to the following formula (16): (16) wherein, is the weight coefficient of the normalized difference vegetation index at a current time point.

[0077] In some embodiments, the weight coefficient of the normalized difference vegetation index is updated to 0.8 in a case where the temperature information is greater than a first temperature threshold or the rainfall information satisfies a first condition. Illustratively, the first temperature threshold is 35 degrees Celsius; the first condition is that it has not rained for 5 consecutive days; in a case where the temperature information is greater than 35 degrees Celsius or the rainfall information satisfies the first condition, the weight coefficient of the normalized difference vegetation index is updated to 0.8 to strengthen the evaluation weight of the change in soil humidity.

[0078] In some embodiments, after obtaining the irrigation effect score, the first pruning threshold can be adjusted according to the irrigation effect score to obtain an adjusted first pruning threshold, and the first pruning threshold is updated according to the adjusted first pruning threshold.

[0079] Illustratively, in a case where the irrigation effect score is greater than or equal to a first effect threshold, the first pruning threshold is not updated; in a case where the irrigation effect score is greater than or equal to a second effect threshold and less than the first effect threshold, the first pruning threshold is adjusted to obtain an adjusted first pruning threshold, and the first pruning threshold is updated according to the adjusted first pruning threshold; in a case where the irrigation effect score is less than the second effect threshold, the causal diagram irrigation model is retrained.

[0080] In some embodiments, the first pruning threshold is updated in a case where the irrigation effect score satisfies a first relaxation condition. Illustratively, the first relaxation condition can be that the irrigation effect score is less than a second effect threshold for 5 consecutive times; in a case where the irrigation effect score is less than the second effect threshold for 5 consecutive times, the first pruning threshold is updated according to the following formula (17): (17) wherein, is the updated first pruning threshold; is the first pruning threshold before being updated; is a learning rate, which can be 0.05; is the second effect threshold.

[0081] Illustratively, in some embodiments, the water saving rate can be calculated by the following formula (18): (18) wherein, is a water saving rate.

[0082] Exemplarily, in some embodiments, the crop health index can be calculated according to the following formula (19): (19) wherein, is a crop health index; is a normalized difference vegetation index at a current time point.

[0083] Exemplarily, in some embodiments, the model stability can be calculated according to the following formula (19): (20) wherein, is a model stability. Exemplarily, the decision error can be a mean square error or a mean absolute error between a predicted value and a verified value of the crop water requirement information, for measuring a change in model prediction performance before and after pruning.

[0084] In some embodiments, irrigation results and irrigation effect scores are displayed by using a visual chart. Exemplarily, irrigation effects are converted into a visual chart for display; by using a farmland heat map, soil humidity distribution is displayed by different color depths, and by using a sliding time axis, historical changes can be traced back and historical data can be viewed; by using a contribution analysis chart, the influence weight of factors such as rainfall and temperature on irrigation decision-making can be displayed; by using a model comparison board, the performance difference of the model before and after pruning can be displayed by a column chart; by using a multi-crop comparison report, the effect comparison of crops such as wheat and fruit trees can be displayed.

[0085] In some embodiments, irrigation results according to irrigation control information are scored to obtain irrigation effect scores; by using the irrigation effect scores, the pruning threshold can be relaxed when the effect of the causal diagram irrigation model is poor, so as to realize self-correction of the model and improve the precision of irrigation decision-making; irrigation results and irrigation effect scores are visually displayed, so that irrigation effects can be intuitively displayed and support can be provided for irrigation decision optimization.

[0086] Figure 5 is a structural schematic diagram of an intelligent irrigation decision system provided by an embodiment of the present disclosure. As shown in Figure 5 , the intelligent irrigation decision system includes an application layer, a business processing layer, and a terminal device access layer.

[0087] The application layer includes a web terminal, a mobile terminal, and a cloud platform, and is used for real-time viewing of core parameter data, assisting users in understanding model decision logic, and realizing irrigation decision visualization.

[0088] The terminal device access layer is used to access the meteorological environment monitoring device, the soil moisture monitoring device, and the irrigation device.

[0089] The service layer includes a data processing module, a causal reasoning module, a dynamic pruning module, an irrigation control module, and an irrigation result analysis module.

[0090] In some embodiments, the data processing module is configured to preprocess meteorological data collected by the meteorological environment monitoring device, soil moisture data collected by the soil moisture monitoring device, and future rainfall obtained from a weather forecast.

[0091] In some embodiments, the causal reasoning module is configured to analyze causal paths of variables such as soil moisture, rainfall, crop water requirement, and temperature based on a soil moisture change equation and a crop water requirement equation, and construct a causal graph irrigation model.

[0092] In some embodiments, the dynamic pruning module is configured to accurately identify redundant logic to dynamically adjust pruning thresholds according to irrigation-related parameters such as crop type, crop growth stage coefficient, and causal influence strength, and then perform dynamic pruning and model compression processing on the causal graph irrigation model.

[0093] In some embodiments, the irrigation control module can complete irrigation of the farmland according to the calculated irrigation water quantity and the optimized irrigation strategy.

[0094] In some embodiments, the irrigation result analysis module can evaluate the irrigation effect through multi-dimensional scoring and visually display the irrigation results and irrigation effect scores.

[0095] Figure 6 is a structural schematic diagram of an intelligent irrigation decision device provided by an embodiment of the present disclosure. As shown in Figure 6 the device can include a construction module 610, a prediction module 620, and a decision module 630.

[0096] The construction module 610 is configured to construct a causal graph irrigation model. The causal graph irrigation model is a causal graph model configured to predict crop water requirement information and soil moisture information at a first time point according to causal relationships between input nodes, intermediate nodes, and output nodes in the causal graph irrigation model. The first time point is a time point after a first time period from a current time point. The prediction module 620 is configured to predict the crop water requirement information and the soil moisture information at the first time point based on the obtained soil moisture information, rainfall information, and temperature information, using the causal graph irrigation model, to obtain a predicted value of the crop water requirement information and a predicted value of the soil moisture information at the first time point. The decision module 630 is configured to determine irrigation control information based on the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point.

[0097] In some embodiments, the construction module 610 is further configured to construct a soil moisture change equation based on the rainfall variable, the crop water requirement variable, and the temperature variable; construct a crop water requirement calculation equation based on the potential evapotranspiration variable and the crop growth stage coefficient; and construct a causal diagram irrigation model based on the soil moisture change equation and the crop water requirement equation; wherein the causal diagram irrigation model is a causal diagram model taking the rainfall node and the temperature node as input nodes, the soil moisture node as a state node, and the crop water requirement node as an output node.

[0098] In some embodiments, the rainfall variable is rainfall information at a second time point; the temperature variable is temperature information at a third time point; the crop water requirement variable is crop water requirement information at a fourth time point; the soil moisture variable in the crop water requirement calculation equation is crop water requirement information at a fifth time point; and the second time point, the third time point, the fourth time point, and the fifth time point are all time points before the current time point.

[0099] In some embodiments, the device further comprises a weight coefficient updating module 604. The weight coefficient updating module 604 is configured to dynamically update the first weight coefficient, the second weight coefficient, and the third weight coefficient; the first weight coefficient is a weight coefficient of the rainfall variable; the second weight coefficient is a weight coefficient of the crop water requirement variable; and the third weight coefficient is a weight coefficient of the temperature variable.

[0100] In some embodiments, the device further comprises a pruning module 605. The pruning module 605 is configured to perform dynamic pruning processing on the causal diagram irrigation model to obtain a pruned causal diagram irrigation model; perform weight distillation processing on the pruned causal diagram irrigation model to obtain a compressed causal diagram irrigation model; and update the causal diagram irrigation model according to the compressed causal diagram irrigation model.

[0101] In some embodiments, the pruning module 605 is further configured to, for each node in the causal diagram irrigation model, calculate a causal contribution degree of the node; for each sub-tree in the causal diagram irrigation model, determine a redundancy degree of the sub-tree according to the causal contribution degree of each node in the sub-tree; and in a case where the redundancy degree is greater than a first pruning threshold, perform pruning processing on the sub-tree to obtain the pruned causal diagram irrigation model.

[0102] In some embodiments, the device further comprises a pruning threshold updating module 606. The pruning threshold updating module 606 is configured to dynamically update the first pruning threshold according to the crop growth stage coefficient; and an initial preset value of the first pruning threshold is determined according to the crop type.

[0103] In some embodiments, the pruning module 605 is further configured to multiply the weights of all edges on a first path to obtain a weight of the first path, the first path being a path from an input node to an output node and passing through the node, and sum the weights of all first paths in the causal graph irrigation model to obtain the causal contribution of the node.

[0104] In some embodiments, the apparatus further comprises a scoring module 607. The scoring module 607 is configured to score an irrigation result according to the irrigation control information to obtain an irrigation effect score, and visually display the irrigation result and the irrigation effect score.

[0105] In some embodiments, the scoring module 607 is further configured to determine soil humidity change information according to the soil humidity information after irrigation and the target soil humidity information, and calculate the irrigation effect score according to the normalized difference vegetation index and the soil humidity change information.

[0106] It should be noted that the intelligent irrigation decision apparatus provided by the embodiments of the present disclosure includes modules (or units) for performing the steps of the intelligent irrigation decision method in the above embodiments, and can achieve the same technical effects. To avoid repetition, it will not be described here.

[0107] Figure 7 FIG. 1 is a structural schematic diagram of an electronic device provided by an embodiment of the present disclosure.

[0108] As shown in FIG. 1, the electronic device can include a processor 701 and a memory 702 having computer program instructions stored therein. Figure 7

[0109] Specifically, the processor 701 can include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or can be configured as one or more integrated circuits that implement one or more embodiments of the present disclosure.

[0110] The memory 702 can include a mass storage for data or instructions. By way of example and not limitation, the memory 702 can include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In one example, the memory 702 can include a removable or non-removable (or fixed) medium, or the memory 702 is a non-volatile solid state memory. The memory 702 can be internal or external to the integrated gateway disaster recovery device.

[0111] ​In one example, the memory 702 can be a Read Only Memory (ROM). In one example, the ROM can be a mask programmed ROM, a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable PROM (EEPROM), an Electrically Alterable ROM (EAROM) or a flash memory, or a combination of two or more of these.

[0112] The memory 702 can include a Read Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk storage media device, an optical storage media device, a flash memory device, an electrical, optical, or other physical / tangible memory storage device. Thus, generally, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software comprising computer-executable instructions that, when executed (e.g., by one or more processors), are operable to perform the operations described with reference to the methods according to an aspect of the present disclosure.

[0113] The processor 701 implements any one of the intelligent irrigation decision-making methods in the embodiments of the present disclosure by reading and executing the computer program instructions stored in the memory 702.

[0114] In one example, the electronic device can further include a communication interface 703 and a bus 704. As shown, the processor 701, the memory 702, and the communication interface 703 are connected through the bus 704 and complete communication among each other. Figure 7

[0115] The communication interface 703 is mainly used to realize the communication between the modules, devices, units and / or equipment in the embodiments of the present disclosure.

[0116] ​Bus 704 includes a hardware, software, or both that couples components of the online data traffic metering device to each other. As an example without limitation, bus can include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or another suitable bus or a combination of two or more of these. Where suitable, bus 304 can include one or more buses. Although specific busses are described and illustrated in the embodiments of the present disclosure, the present disclosure contemplates any suitable bus or interconnect.

[0117] In addition, in combination with the intelligent irrigation decision method in the above embodiments, the embodiments of the present disclosure can provide a computer storage medium for implementation. The computer storage medium has computer program instructions stored thereon; the computer program instructions are executed by a processor to implement any one of the intelligent irrigation decision methods in the above embodiments.

[0118] The embodiments of the present disclosure also provide a computer program product, which includes a computer program, and the computer program is executed by a processor to implement any one of the intelligent irrigation decision methods in the above embodiments.

[0119] It needs to be clear that the present disclosure is not limited to the specific configurations and processes described above and shown in the drawings. For the sake of brevity, detailed descriptions of well-known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present disclosure are not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between steps, after understanding the spirit of the present disclosure.

[0120] The functional blocks shown in the above-described block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this disclosure are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, read-only memory (ROM), flash memory, erasable read-only memory (EROM), floppy disks, compact disc read-only memory (CD-ROM), optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0121] It should also be noted that the exemplary embodiments mentioned in this disclosure describe methods or systems based on a series of steps or apparatus. However, the present invention is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0122] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0123] The above merely describes specific implementation of the present application, and those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, module and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described herein again. It should be understood that the protection scope of the present application is not limited to this, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. A smart irrigation decision-making method, characterized in that, include: A causal graph irrigation model is constructed; the causal graph irrigation model is used to predict crop water requirement information and soil moisture information at a first time point based on the causal relationship between input nodes, intermediate nodes and output nodes in the causal graph irrigation model; the first time point is the time point of the first time period after the current time point. Based on the acquired soil moisture, rainfall and temperature information, the causal graph irrigation model is used to predict the crop water requirement information and the soil moisture information at the first time point, so as to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point. Irrigation control information is determined based on the predicted values ​​of crop water requirement information and soil moisture information at the first time point.

2. The method according to claim 1, characterized in that, The construction of the causal graph irrigation model includes: Based on rainfall, crop water requirement, and temperature variables, an equation for soil moisture variation was constructed. Based on potential evapotranspiration variables and crop growth stage coefficients, an equation for calculating crop water requirement is constructed. Based on the soil moisture change equation and the crop water requirement equation, the causal graph irrigation model is constructed. The causal graph irrigation model is a causal graph model with rainfall and temperature nodes as input nodes, soil moisture nodes as state nodes, and crop water requirement nodes as output nodes.

3. The method according to claim 2, characterized in that, The rainfall variable is the rainfall information at the second time point; the temperature variable is the temperature information at the third time point; the crop water requirement variable is the crop water requirement information at the fourth time point; the soil moisture variable in the crop water requirement calculation equation is the crop water requirement information at the fifth time point; the second time point, the third time point, the fourth time point, and the fifth time point are all time points prior to the current time point.

4. The method according to claim 2, characterized in that, The first weighting coefficient is the weighting coefficient of the rainfall variable; the second weighting coefficient is the weighting coefficient of the crop water requirement variable; The third weighting coefficient is the weighting coefficient of the temperature variable; the method further includes: The first weight coefficient, the second weight coefficient, and the third weight coefficient are dynamically updated.

5. The method according to claim 1, characterized in that, The method further includes: The causal graph irrigation model is dynamically pruned to obtain the pruned causal graph irrigation model. The pruned causal graph irrigation model is subjected to weight distillation to obtain a compressed causal graph irrigation model. The causal graph irrigation model is updated based on the compressed causal graph irrigation model.

6. The method according to claim 5, characterized in that, The dynamic pruning process of the causal graph irrigation model to obtain the pruned causal graph irrigation model includes: For each node in the causal graph irrigation model, calculate the causal contribution of that node; For each subtree in the causal graph irrigation model, the redundancy of the subtree is determined based on the causal contribution of each node in the subtree. If the redundancy is greater than the first pruning threshold, the subtree is pruned to obtain the pruned causal graph irrigation model.

7. The method according to claim 6, characterized in that, The method further includes: The first pruning threshold is dynamically updated based on the crop growth stage coefficient; the initial preset value of the first pruning threshold is determined according to the crop type.

8. The method according to claim 6, characterized in that, The calculation of the causal contribution of the node includes: The weight of the first path is obtained by multiplying the weights of all edges on the first path; the first path is the path from the input node to the output node and passing through the node. The causal contribution of each node is obtained by summing the weights of all the first paths in the causal graph irrigation model.

9. The method according to any one of claims 1-8, characterized in that, The method further includes: The irrigation results obtained by irrigating according to the irrigation control information are scored to obtain an irrigation effect score; The irrigation results and irrigation effect scores are visualized.

10. The method according to claim 9, characterized in that, The process of scoring the irrigation results according to the irrigation control information to obtain an irrigation effect score includes: Based on the soil moisture information after irrigation and the target soil moisture information, determine the soil moisture change information; The irrigation effect score is calculated based on the normalized difference vegetation index and the soil moisture change information.

11. An intelligent irrigation decision-making device, characterized in that, The device includes: The construction module is used to construct a causal graph irrigation model; the causal graph irrigation model is a causal graph model used to predict crop water requirement information and soil moisture information at the first time point based on the causal relationships between input nodes, intermediate nodes and output nodes in the causal graph irrigation model; the first time point is the time point of the first time period after the current time point; The prediction module is used to predict the crop water requirement information and the soil moisture information at the first time point based on the acquired soil moisture information, rainfall information and temperature information, using the causal graph irrigation model, to obtain the predicted value of the crop water requirement information and the predicted value of the soil moisture information at the first time point. The decision module is used to determine irrigation control information based on the predicted values ​​of the crop water requirement information and the predicted values ​​of the soil moisture information at the first time point.

12. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement the intelligent irrigation decision-making method as described in any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement the intelligent irrigation decision-making method as described in any one of claims 1-10.

14. A computer program product, characterized in that, It includes a computer program, which, when executed, implements the intelligent irrigation decision-making method as described in any one of claims 1-10.

Images