Multi-target water and fertilizer collaborative optimization method and system for scutellaria baicalensis

Abstract

The invention relates to the technical field of traditional Chinese medicinal material planting, and discloses a multi-target water and fertilizer collaborative optimization method and system for scutellaria baicalensis, and the method comprises the steps: generating a plurality of candidate decision variables based on a target optimization problem through an RVEA-II algorithm, calling a preset multi-target optimization model to determine prediction data corresponding to each candidate decision variable, and carrying out the multi-target water and fertilizer collaborative optimization on the basis of the prediction data. Determining a plurality of Pareto optimal solutions for indicating a water and fertilizer strategy of the scutellaria baicalensis from the plurality of candidate decision variables; then, in response to a selection instruction triggered by a user for a target recommendation control in the multiple recommendation controls, based on a recommendation strategy corresponding to the target recommendation control, selecting a target Pareto optimal solution from the multiple Pareto optimal solutions; wherein the target optimization problem is to search a Pareto optimal solution, so that the yield, the quality and the resource utilization rate of the scutellaria baicalensis are maximized, the environmental risk is minimized, and all constraint conditions are met; therefore, the application effect of the obtained optimization strategy in practical application can be ensured.

Description

Technical Field

[0001] This invention relates to the field of Chinese medicinal herb cultivation technology, and relates to, but is not limited to, a multi-objective water and fertilizer synergistic optimization method and system for Scutellaria baicalensis. Background Technology

[0002] As an important traditional Chinese medicine, the content and yield of its main active ingredients, such as baicalin and baicalein, directly affect its medicinal value and economic benefits.

[0003] Currently, the yield or quality of Scutellaria baicalensis can be predicted using regression models or empirical formulas. However, the optimization objectives of these models are singular, resulting in limited effectiveness of optimization strategies in practical applications. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a multi-objective water and fertilizer synergistic optimization method and system for Scutellaria baicalensis, which can take into account both the yield and quality of Scutellaria baicalensis, while improving resource utilization and reducing environmental risks, so as to ensure the application effect of the obtained optimization strategy in practical applications.

[0005] The specific technical solutions of this invention are as follows: This invention provides a multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis, the method comprising: The objective optimization problem is obtained and input into the RVEA-II algorithm. The objective optimization problem is determined based on multiple preset objective functions and constraints. The objective optimization problem is to find the Pareto optimal solution that maximizes the yield, quality and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. The RVEA-II algorithm is used to generate multiple candidate decision variables based on the objective optimization problem. A pre-defined multi-objective optimization model is then called to determine the prediction data corresponding to each candidate decision variable. Based on the prediction data, multiple Pareto optimal solutions are determined from the multiple candidate decision variables. The Pareto optimal solutions are used to indicate the water and fertilizer strategies for Scutellaria baicalensis. The prediction data includes the output values ​​of multiple objective functions. The multi-objective optimization model is used to simulate the dynamic growth process of Scutellaria baicalensis under different water and fertilizer management strategies. In response to a user's selection command triggered by a target recommended control among multiple recommended controls, the target Pareto optimal solution is selected from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control.

[0006] In some embodiments, the plurality of objective functions include a first objective function, a second objective function, a third objective function, and a fourth objective function; wherein, the first objective function is used to maximize the yield of Scutellaria baicalensis; the second objective function is used to maximize the quality of Scutellaria baicalensis; the third objective function is used to maximize the resource utilization rate of Scutellaria baicalensis; and the fourth objective function is used to minimize the environmental risk of Scutellaria baicalensis.

[0007] In some embodiments, the constraints include at least: The root zone water content of Scutellaria baicalensis is greater than or equal to the water content at the wilting point of Scutellaria baicalensis and less than or equal to the field capacity. And / or, the soil electrical conductivity in the root zone is less than or equal to the highest salt concentration that Scutellaria baicalensis can tolerate; And / or, the number of times the first fertilizer is applied is less than or equal to the first preset number of times, the number of times the second fertilizer is applied is less than or equal to the second preset number of times, and the total number of times the fertilizer is applied is less than or equal to the third preset number of times; And / or, the total irrigation water volume is less than or equal to the total irrigation volume threshold and the total nitrogen fertilizer amount is less than or equal to the nitrogen fertilizer amount threshold; The total irrigation threshold is determined based on the water requirement of Scutellaria baicalensis, and the nitrogen fertilizer threshold is determined based on the target yield and the basic nitrogen supply from the soil.

[0008] In some embodiments, the recommended control includes a first control; when the target recommended control is the first control, the recommendation strategy is a comprehensive scoring method. Based on the recommendation strategy corresponding to the target recommendation control, the target Pareto optimal solution is selected from multiple Pareto optimal solutions, including: The weights corresponding to the output values ​​of multiple objective functions included in the prediction data are determined by the entropy weight method. Based on the weights corresponding to the output values ​​of multiple objective functions, the comprehensive score of each Pareto optimal solution is determined and ranked. The Pareto optimal solution with the highest overall score is determined as the target Pareto optimal solution.

[0009] In some embodiments, the recommended control includes a second control; when the target recommended control is the second control, the recommendation strategy is the knee-first method. Based on the recommendation strategy corresponding to the target recommendation control, the target Pareto optimal solution is selected from multiple Pareto optimal solutions, including: Based on the prediction data corresponding to multiple Pareto optimal solutions, a Pareto front is constructed in a multidimensional objective space; Determine the vector angle formed between three consecutive Pareto optimal solutions on the Pareto front, and determine the position with the largest vector angle on the Pareto front as the knee point; The Pareto optimal solution corresponding to the knee point is determined as the target Pareto optimal solution.

[0010] In some embodiments, the recommended control includes a third control; when the target recommended control is a third control, the recommendation strategy is a robustness-first approach. Based on the recommendation strategy corresponding to the target recommendation control, the target Pareto optimal solution is selected from multiple Pareto optimal solutions, including: Using a multi-objective optimization model, the predicted data for each Pareto optimal solution under multiple preset meteorological scenarios were determined; Based on the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios, the coefficient of variation is determined; The Pareto optimal solution with the smallest coefficient of variation is determined as the target Pareto optimal solution.

[0011] In some embodiments, the recommended control includes a fourth control; when the target recommended control is the fourth control, the recommendation strategy is context matching. Based on the recommendation strategy corresponding to the target recommendation control, the target Pareto optimal solution is selected from multiple Pareto optimal solutions, including: Based on the preset weight vectors corresponding to the output values ​​of multiple objective functions included in the predicted data, the comprehensive score of each Pareto optimal solution is determined. The Pareto optimal solution with the highest overall score is determined as the target Pareto optimal solution.

[0012] In some embodiments, the method further includes: Obtain actual data on planting Scutellaria baicalensis based on the target Pareto optimal solution in the current cycle; Determine the deviation between the actual data and the predicted data corresponding to the target Pareto optimal solution; If the deviation exceeds the preset deviation threshold, the parameters in the multi-objective optimization model are iteratively updated based on the meteorological data, soil data and field data corresponding to the current cycle to obtain a new multi-objective optimization model, and a new target Pareto optimal solution is determined based on the new multi-objective optimization model; the new target Pareto optimal solution is used for planting Scutellaria baicalensis in the next cycle.

[0013] This invention also provides a multi-objective water and fertilizer synergistic optimization system for Scutellaria baicalensis, the system comprising: The acquisition module is used to acquire the target optimization problem and input it into the RVEA-II algorithm. The target optimization problem is determined based on multiple preset objective functions and constraints. The target optimization problem is to find the Pareto optimal solution that maximizes the yield, quality and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. The determination module is used to generate multiple candidate decision variables based on the objective optimization problem using the RVEA-II algorithm, and to call a preset multi-objective optimization model to determine the prediction data corresponding to each candidate decision variable. Based on the prediction data, multiple Pareto optimal solutions are determined from the multiple candidate decision variables. The Pareto optimal solutions are used to indicate the water and fertilizer strategies for Scutellaria baicalensis. The prediction data includes the output values ​​of multiple objective functions. The multi-objective optimization model is used to simulate the dynamic growth process of Scutellaria baicalensis under different water and fertilizer management strategies. The selection module is used to respond to the user's selection command triggered by the target recommended control among multiple recommended controls, and select the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control.

[0014] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: In this embodiment of the invention, a target optimization problem is obtained and input into the RVEA-II algorithm. The target optimization problem is to find a Pareto optimal solution that maximizes the yield, quality, and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. Then, the RVEA-II algorithm generates multiple candidate decision variables based on the target optimization problem and calls a preset multi-objective optimization model to determine the predicted data corresponding to each candidate decision variable. Based on the predicted data, multiple Pareto optimal solutions are determined from the multiple candidate decision variables. The Pareto optimal solution is used to indicate the water and fertilizer strategy for Scutellaria baicalensis. Finally, in response to the user's selection instruction triggered by the target recommended control among multiple recommended controls, the target Pareto optimal solution is selected from the multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control. In this way, the yield and quality of Scutellaria baicalensis can be balanced, while improving resource utilization and reducing environmental risk, ensuring the effectiveness of the obtained optimization strategy in practical applications. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein: Figure 1 This is a flowchart illustrating the multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a multi-objective water and fertilizer synergistic optimization system for Scutellaria baicalensis provided in an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0018] It should be noted that the terms "first, second, and third" used in the embodiments of the present invention are only used to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first, second, and third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present invention described herein can be implemented in an order other than that illustrated or described herein.

[0019] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which these embodiments of the invention pertain. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art and should not be interpreted in an idealized or overly formal sense unless specifically defined as herein.

[0020] Figure 1 A flowchart illustrating a multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis is shown. This method can be executed via a control device, which may include at least one of a personal computer, laptop computer, smartphone, tablet computer, and portable wearable device; however, this embodiment does not limit the specific device used.

[0021] like Figure 1 As shown, the multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis provided in this embodiment of the invention may include S101-S103.

[0022] S101. Obtain the objective optimization problem and input it into the RVEA-II algorithm.

[0023] In some embodiments, before obtaining the target optimization problem, the user can first construct a multi-objective optimization model and pre-store the multi-objective optimization model in the memory of the control device for easy access by the optimization algorithm (i.e., the RVEA-II algorithm). The multi-objective optimization model is used to simulate the dynamic growth process of Scutellaria baicalensis under different water and fertilizer management strategies.

[0024] For example, users can first construct a water dynamics model, a nutrient dynamics model, a biological model, and an environmental risk model. Then, basic parameters are obtained through literature review and field trials, and these parameters are substituted into the corresponding models to enable simulation based on actual conditions. Next, by synchronizing time steps and defining data flows, the water dynamics model, nutrient dynamics model, biological model, and environmental risk model are coupled together. The coupled model is then encapsulated as a "black box" function that can be called by the optimization algorithm (equivalent to the RVEA-II algorithm mentioned above), i.e., a multi-objective optimization model. The input data of this multi-objective optimization model consists of a decision vector (also known as a water and fertilizer strategy vector) and environmental data, while the output data is the objective function output value (e.g., yield, quality, resource utilization rate, environmental risk).

[0025] In some application scenarios, the data flow between the water dynamics model, nutrient dynamics model, biological model, and environmental risk model can be defined as follows: using the root zone soil water content output by the water dynamics model to generate water stress factors, and using the root zone available nutrient (such as nitrogen, phosphorus, and potassium) concentration output by the nutrient dynamics model to generate nutrient stress factors; then combining the water stress factors, nutrient stress factors, and environmental data (such as temperature and photosynthetically active radiation) to drive the growth sub-model to calculate biomass increment; after obtaining the biomass increment, using the biomass increment, water stress factors, nitrogen stress factors, and phenological parameters to drive the quality sub-model to calculate baicalin accumulation, and simultaneously using the Scutellaria baicalensis root biomass output by the growth sub-model to adjust the root absorption coefficient in the water dynamics model and the nutrient dynamics model.

[0026] In some embodiments, the water dynamics model can be constructed based on the soil water balance equation to describe the dynamic changes in soil moisture and calculate the soil water content and available water in the Scutellaria baicalensis root zone. The soil water balance equation is shown in Formula 1 below: (Formula 1) in, This represents the soil moisture content at time t; This represents the soil moisture content at time t-1; This represents the amount of irrigation at time t; This represents the effective rainfall at time t; This represents the amount of evapotranspiration at time t; This represents the amount of deep seepage at time t; Represents the surface runoff at time t; Indicates soil volume.

[0027] The nutrient dynamics model, based on the nutrient mass balance equation and kinetic absorption model, is used to simulate the dynamic changes of nitrogen (N), phosphorus (P), and potassium (K) in the soil, and to calculate the nutrient concentration and crop uptake in the rhizosphere of Scutellaria baicalensis. Specifically, the root uptake rate and effective nitrogen concentration in the rhizosphere can be determined using the Michaelis-Menten equation, which is shown in Equation 2 below: (Formula 2) in, This indicates the nitrogen uptake rate of Scutellaria baicalensis roots; This indicates the maximum nitrogen absorption rate of Scutellaria baicalensis roots; This indicates the effective nitrogen concentration in the root zone soil; This represents the Michaelis constant.

[0028] The biological model includes a growth sub-model and a quality sub-model. The growth sub-model, based on the photosynthesis-respiration balance, is used to simulate the biomass accumulation process of *Scutellaria baicalensis* and determine its daily biomass increment. The quality sub-model, based on the relationship between *Scutellaria baicalensis* quality and the ecological environment, is used to simulate the dynamic changes in *Scutellaria baicalensis* quality and determine the content of effective components (such as baicalin and baicalein) in *Scutellaria baicalensis*. The daily biomass increment can be calculated using the following formula: (Formula 3) in, Indicates daily biological increment; PAR represents photosynthetically active radiation. Indicates the amount of biomass lost through photorespiration; Indicates the temperature correction factor; Indicates water stress factor; Indicates nitrogen stress factor; Indicates phosphorus stress factor; This indicates a potassium stress factor.

[0029] The content of active ingredients in Scutellaria baicalensis can be calculated using the following formula 4: (Formula 4) in, Indicates the instantaneous accumulation rate of the active ingredient; Cumulative content of active ingredients; Represents the stress response function; Indicates water stress factor; Indicates nitrogen stress factor; Indicates light intensity; This represents the biomass at time t.

[0030] Environmental risk models are constructed based on information such as meteorological data, soil characteristic data, and crop growth data to assess environmental risks during the cultivation of Scutellaria baicalensis. For example, environmental risk models can predict the impact of potential droughts, floods, nutrient imbalances, and other problems on the growth and quality of Scutellaria baicalensis.

[0031] It should be noted that multi-objective optimization models can be constructed according to actual needs. This application embodiment does not limit the type and number of models included in the multi-objective optimization model, nor the construction method of the multi-objective optimization model.

[0032] In some embodiments, after constructing a multi-objective optimization model, the user can first define the decision variables, namely the water and fertilizer strategy.

[0033] For example, the entire optimization cycle of Scutellaria baicalensis can be divided into 5 cycles. Assuming a 150-day growing season for Scutellaria baicalensis, the optimization cycle can be divided into: seedling stage (1-30 days), branching stage (31-60 days), flowering stage (61-90 days), early rhizome enlargement stage (91-120 days), and late rhizome enlargement stage (121-150 days). During each growth stage, water and fertilizer application must remain consistent; therefore, the decision variable vector is defined as... Where t represents the period number, t=1 is the seedling stage, t=2 is the branching stage, t=3 is the flowering stage, t=4 is the early stage of rhizome enlargement, and t=5 is the late stage of rhizome enlargement; N represents the irrigation amount for the t-th cycle; t P represents the amount of nitrogen fertilizer applied in the t-th cycle. t K represents the amount of phosphate fertilizer applied in the t-th cycle. t This represents the amount of potassium fertilizer applied in the t-th cycle.

[0034] In some embodiments, the objective optimization problem is determined based on a plurality of preset objective functions and constraints, and the prediction data includes the output values ​​of the plurality of objective functions.

[0035] For example, users can define objective functions and constraints according to actual needs, and determine the objective optimization problem based on the defined objective functions and constraints.

[0036] In some embodiments, the multiple objective functions include a first objective function, a second objective function, a third objective function, and a fourth objective function. The first objective function is used to maximize the yield of Scutellaria baicalensis; the second objective function is used to maximize the quality of Scutellaria baicalensis; the third objective function is used to maximize the resource utilization rate of Scutellaria baicalensis; and the fourth objective function is used to minimize the environmental risk of Scutellaria baicalensis.

[0037] For example, the first objective function can be defined as follows: Where X is the decision variable and D is the environmental data, the second objective function can be defined as: Similar to the first objective function, where X is the decision variable and D is the environmental data, the third objective function can be defined as: ,in This represents the total water consumption during the growing season (tons per acre). , , These represent the total nitrogen, phosphorus, and potassium application rates during the growing season. All are weighting coefficients. The fourth objective function can be defined as: , where X is the decision variable.

[0038] In some embodiments, the constraints include at least the following: the root zone water content of Scutellaria baicalensis is greater than or equal to the water content corresponding to the wilting point of Scutellaria baicalensis and less than or equal to the field capacity; and / or, the root zone soil electrical conductivity is less than or equal to the highest salt concentration tolerated by Scutellaria baicalensis; and / or, the number of applications of the first fertilizer is less than or equal to the first preset number of applications, the number of applications of the second fertilizer is less than or equal to the second preset number of applications, and the total number of applications is less than or equal to the third preset number of applications; and / or, the total irrigation water volume is less than or equal to the total irrigation volume threshold and the total nitrogen fertilizer volume is less than or equal to the nitrogen fertilizer volume threshold.

[0039] In some embodiments, the wilting point of Scutellaria baicalensis is the minimum water content threshold to prevent wilting, and the field capacity is the maximum water content threshold to prevent root hypoxia. The total irrigation threshold is determined based on the water requirement of Scutellaria baicalensis, and the nitrogen fertilizer threshold is determined based on the target yield and the soil's basic nitrogen supply. The first fertilizer application can be nitrogen fertilizer, and the second fertilizer application can be phosphorus fertilizer or potassium fertilizer. The first, second, and third preset number of applications are all preset values; for example, the first preset number of applications can be 3 times, the second preset number of applications can be 2 times, and the third preset number of applications can be 5 times. This application does not limit the specific values ​​of the first, second, and third preset number of applications.

[0040] S102. Using the RVEA-II algorithm, generate multiple candidate decision variables based on the objective optimization problem, and call the preset multi-objective optimization model to determine the prediction data corresponding to each candidate decision variable. Based on the prediction data, determine multiple Pareto optimal solutions from the multiple candidate decision variables.

[0041] In some embodiments, the objective optimization problem is to find a Pareto optimal solution that maximizes the yield, quality, and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. The Pareto optimal solution is used to indicate the water and fertilizer strategy for Scutellaria baicalensis; the prediction data includes the output values ​​of multiple objective functions.

[0042] For example, after receiving the objective optimization problem, the RVEA-II algorithm generates multiple candidate decision variables based on the objective optimization problem. These candidate decision variables include combinations of different water-fertilizer ratios and irrigation strategies. Then, the RVEA-II algorithm invokes a multi-objective optimization model to simulate the dynamic growth process of *Scutellaria baicalensis* based on each candidate decision variable, thereby obtaining the output values ​​of the first, second, third, and fourth objective functions, i.e., the predicted data corresponding to each candidate decision variable. Finally, based on the above predicted data, multiple combinations of water-fertilizer ratios and irrigation strategies are selected from the multiple candidate decision variables that maximize the yield, quality, and resource utilization of *Scutellaria baicalensis* while minimizing environmental risks, under the premise of satisfying various constraints, thus obtaining multiple Pareto optimal solutions.

[0043] In some embodiments, as shown in Table 1, when the RVEA-II algorithm calls a multi-objective optimization model to simulate the dynamic growth process of Scutellaria baicalensis, the first objective function can be characterized by outputting the dry weight of the rhizome to represent the yield of Scutellaria baicalensis; the second objective function can be characterized by first outputting the total amount of baicalin accumulated in the rhizome, and then converting the total amount of baicalin accumulated in the rhizome into the baicalin content in the rhizome through the formula baicalin content = total amount of baicalin accumulated ÷ dry weight of rhizome × 100%, to represent the quality of Astragalus membranaceus.

[0044] Table 1 Simulation results of the dynamic growth process of Scutellaria baicalensis It is understood that the embodiments of this application use the RVEA-II algorithm based on a preset multi-objective optimization model to comprehensively evaluate the yield, quality, resource utilization and environmental risk of Scutellaria baicalensis under different water and fertilizer strategies. Through iteration and optimization, the water and fertilizer supply strategy (i.e. Pareto optimal solution) is determined. This not only maximizes the balance between the yield and quality of Scutellaria baicalensis, but also improves resource utilization and reduces environmental risk. Furthermore, it is scientific and accurate, ensuring the application effect of the water and fertilizer strategy in practical applications.

[0045] S103. In response to the user's selection instruction triggered by the target recommended control among multiple recommended controls, select the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control.

[0046] In some embodiments, the target Pareto optimal solution is the Pareto optimal solution that best meets the user's needs among multiple Pareto optimal solutions.

[0047] For example, a user triggers a selection instruction for a target recommended control by clicking on one of multiple recommended controls. Upon receiving the selection instruction, the control device determines the recommendation strategy corresponding to the target recommended control and, based on that strategy, selects the target Pareto optimal solution from multiple Pareto optimal solutions.

[0048] In some embodiments, the recommended control includes a first control, a second control, a third control, and a fourth control. Wherein, when the target recommended control is the first control, the recommendation strategy is a comprehensive scoring method. When the target recommended control is the second control, the recommendation strategy is a knee-first method. When the target recommended control is the third control, the recommendation strategy is a robustness-first method. When the target recommended control is the fourth control, the recommendation strategy is a context-matching method.

[0049] For example, different recommendation strategies are suitable for different scenarios. The comprehensive rating method, the default option, is suitable for scenarios where users have no preferred goals; the knee-first method is suitable for scenarios where the goal is balanced and optimal; the robustness-first method is suitable for scenarios with large climate fluctuations; and the scenario matching method is suitable for scenarios with goal priorities (such as quality priority or production priority). Customers can choose the target recommendation control from multiple recommendation controls according to their actual needs.

[0050] In some embodiments, when the recommendation strategy is a comprehensive scoring method, selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: determining the weights corresponding to the output values ​​of multiple objective functions included in the prediction data using the entropy weight method; determining and ranking the comprehensive score of each Pareto optimal solution based on the weights corresponding to the output values ​​of multiple objective functions; and determining the Pareto optimal solution with the highest comprehensive score as the target Pareto optimal solution.

[0051] For example, after receiving a selection instruction triggered by the user on the first control, the control device, in response to the selection instruction, determines that the recommendation strategy is the comprehensive scoring method. Then, the control device uses the comprehensive scoring method to determine the target Pareto optimal solution. Specifically, the control device can use the entropy weight method to determine the weights corresponding to the output values ​​of the first, second, third, and fourth objective functions in the prediction data. Then, based on the weights corresponding to the output values ​​of the first, second, third, and fourth objective functions, the control device performs a weighted sum of these output values ​​to obtain a comprehensive score. Finally, the Pareto optimal solution with the highest comprehensive score is determined as the target Pareto optimal solution. It should be noted that the entropy weight method is existing technology and will not be elaborated upon here.

[0052] In some embodiments, when the recommendation strategy is the knee-first method, selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: constructing a Pareto front in a multidimensional target space based on the prediction data corresponding to the multiple Pareto optimal solutions; determining the vector angle formed between three consecutive Pareto optimal solutions on the Pareto front, and determining the position with the largest vector angle on the Pareto front as the knee point; and determining the Pareto optimal solution corresponding to the knee point as the target Pareto optimal solution.

[0053] For example, after receiving a selection instruction triggered by the user on the second control, the control device, in response to the selection instruction, determines that the recommended strategy is the knee-first method. Then, the control device uses the knee-first method to determine the target Pareto optimal solution. That is, the control device can determine the prediction data corresponding to each Pareto optimal solution output by the multi-objective optimization model. Then, based on the prediction data corresponding to the Pareto optimal solution, a Pareto front is constructed in the multi-dimensional objective space. The vector angle formed between three consecutive Pareto optimal solutions is determined on the Pareto front, and the position with the largest vector angle on the Pareto front is determined as the knee point. Finally, the Pareto optimal solution corresponding to the knee point is determined as the target Pareto optimal solution.

[0054] In some embodiments, when the recommendation strategy is a robustness-first approach, the target Pareto optimal solution is selected from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control. This includes: determining the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios through a multi-objective optimization model; determining the coefficient of variation based on the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios; and determining the Pareto optimal solution with the smallest coefficient of variation as the target Pareto optimal solution.

[0055] For example, after receiving a selection instruction triggered by a user on a third control, the control device, in response to the selection instruction, determines that the recommended strategy is the robustness-first method. Then, the control device uses the robustness-first method to determine the target Pareto optimal solution. That is, the control device can use a multi-objective optimization model to determine the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios (such as drought scenarios and rainy scenarios). Then, based on the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios, the control device determines the standard deviation and mean of each Pareto optimal solution under multiple preset meteorological scenarios, and uses the standard deviation divided by the mean to obtain the coefficient of variation corresponding to each Pareto optimal solution. Finally, the Pareto optimal solution with the smallest coefficient of variation is determined as the target Pareto optimal solution. The meteorological scenarios can be set according to actual needs; for example, a scenario with a 30% decrease in precipitation can be set as a drought scenario, and a scenario with a 20% increase in precipitation can be set as a rainy scenario. This application embodiment does not limit this.

[0056] In some embodiments, when the recommendation strategy is scenario matching, selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: determining the comprehensive score of each Pareto optimal solution based on the preset weight vectors corresponding to the output values ​​of the multiple objective functions included in the prediction data; and determining the Pareto optimal solution with the highest comprehensive score as the target Pareto optimal solution.

[0057] For example, after receiving a selection command from the user triggered by the fourth control, the control device, in response to the selection command, determines that the recommendation strategy is the scenario matching method. Then, the control device uses the scenario matching method to determine the target Pareto optimal solution. Specifically, the control device can perform a weighted sum of the output values ​​of the first, second, third, and fourth objective functions based on their respective preset weight vectors in the predicted data, to obtain a comprehensive score. Finally, the Pareto optimal solution with the highest comprehensive score is determined as the target Pareto optimal solution. The preset weight vectors corresponding to the output values ​​of the first, second, third, and fourth objective functions in the predicted data are pre-configured by the user in the control device according to the actual application scenario.

[0058] It is understood that, in the embodiments of this application, users can select the most suitable water and fertilizer supply strategy from multiple water and fertilizer supply strategies according to actual needs, thus expanding the scope of application.

[0059] In some embodiments, the multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis provided in this application may further include: obtaining actual data on planting Scutellaria baicalensis based on the target Pareto optimal solution for the current period; determining the deviation between the actual data and the predicted data corresponding to the target Pareto optimal solution; if the deviation is greater than a preset deviation threshold, iteratively updating the parameters in the multi-objective optimization model based on the meteorological data, soil data and field data corresponding to the current period to obtain a new multi-objective optimization model, and determining a new target Pareto optimal solution based on the new multi-objective optimization model.

[0060] The new objective Pareto optimal solution is used for planting Scutellaria baicalensis in the next cycle.

[0061] In some embodiments, the user cultivates Scutellaria baicalensis based on a target Pareto optimal solution and deploys sensors (such as soil moisture sensors, electrical conductivity sensors, irrigation flow sensors, etc.) and drones in the field. The control device can acquire actual data on the cultivation of Scutellaria baicalensis based on the target Pareto optimal solution through the sensors and drones. This actual data may include soil moisture, electrical conductivity, rainfall, irrigation volume, plant status, etc.

[0062] It should be noted that sensors can be deployed in the field to obtain actual data according to actual needs. This application does not limit the number or type of sensors.

[0063] In some embodiments, the entire Scutellaria baicalensis optimization cycle is divided into five cycles, each corresponding to a different planting stage. Each planting stage has a different focus, therefore the actual data obtained in different cycles will differ. For example, actual data during the seedling stage may be plant height; during the branching stage, actual data may be biomass increase; during the flowering stage, actual data may be transpiration; during the early stage of rhizome enlargement, actual data may be biomass increase; and during the later stage of rhizome enlargement, actual data may be baicalin content. Because the actual data obtained in different cycles differs, the methods for calculating the deviation between the actual data and the predicted data will also differ.

[0064] For example, when the current cycle is the seedling stage, branching stage, or flowering stage, the deviation between the actual data and the predicted data can be determined by calculating the mean absolute error (MAE); when the current cycle is the early stage or late stage of root and stem enlargement, the deviation between the actual data and the predicted data can be determined by calculating the relative error (RE).

[0065] In some embodiments, if the deviation is greater than a preset deviation threshold, it means that the parameters in the multi-objective optimization model need to be iteratively updated in order to improve the prediction accuracy of the multi-objective optimization model.

[0066] For example, based on meteorological, soil, and field data corresponding to the current period, and using the deviation between predicted and measured data as the objective function, a genetic algorithm can be used to invert and correct the parameters in the multi-objective optimization model to obtain new model parameters. A new multi-objective optimization model is then derived based on these new parameters. It should be noted that the preset deviation threshold is a preset value and can be set according to actual needs.

[0067] In some embodiments, after obtaining a new multi-objective optimization model, the new multi-objective optimization model can be invoked through the RVEA-II algorithm to determine a new Pareto optimal solution for the next cycle.

[0068] It is understood that the embodiments of this application can optimize the multi-objective optimization model according to the actual planting conditions of Scutellaria baicalensis, realizing dynamic adjustment and closed-loop control. Through the closed-loop cycle of "execution → monitoring → comparison → calibration → re-optimization → re-execution", the water and fertilizer strategy can be adapted to the dynamic changes in environmental conditions, maintaining long-term effectiveness and stability.

[0069] Based on the same inventive concept, this application also provides a multi-objective water and fertilizer synergistic optimization system for Scutellaria baicalensis, used to implement the aforementioned multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations of one or more control module embodiments provided below can be found in the limitations of the multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis described above, and will not be repeated here. Specifically, Figure 2 This is a schematic diagram of a multi-objective water and fertilizer synergistic optimization system for Scutellaria baicalensis, as described in an embodiment of this application. Figure 2 As shown, the multi-objective water and fertilizer synergistic optimization system for Scutellaria baicalensis includes: The acquisition module 201 is used to acquire the target optimization problem and input the target optimization problem into the RVEA-II algorithm; the target optimization problem is determined based on multiple preset objective functions and constraints; the target optimization problem is to find the Pareto optimal solution that maximizes the yield, quality and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. The determination module 202 is used to generate multiple candidate decision variables based on the objective optimization problem using the RVEA-II algorithm, and to call a preset multi-objective optimization model to determine the prediction data corresponding to each candidate decision variable. Based on the prediction data, multiple Pareto optimal solutions are determined from the multiple candidate decision variables. The Pareto optimal solutions are used to indicate the water and fertilizer strategies for Scutellaria baicalensis. The prediction data includes the output values ​​of multiple objective functions. The multi-objective optimization model is used to simulate the dynamic growth process of Scutellaria baicalensis under different water and fertilizer management strategies. Selection module 203 is used to respond to a user's selection instruction triggered by a target recommended control among multiple recommended controls, and select a target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control.

[0070] In some embodiments, the plurality of objective functions include a first objective function, a second objective function, a third objective function, and a fourth objective function; wherein, the first objective function is used to maximize the yield of Scutellaria baicalensis; the second objective function is used to maximize the quality of Scutellaria baicalensis; the third objective function is used to maximize the resource utilization rate of Scutellaria baicalensis; and the fourth objective function is used to minimize the environmental risk of Scutellaria baicalensis.

[0071] In some embodiments, the constraints include at least: The root zone water content of Scutellaria baicalensis is greater than or equal to the water content at the wilting point of Scutellaria baicalensis and less than or equal to the field capacity. And / or, the soil electrical conductivity in the root zone is less than or equal to the highest salt concentration that Scutellaria baicalensis can tolerate; And / or, the number of times the first fertilizer is applied is less than or equal to the first preset number of times, the number of times the second fertilizer is applied is less than or equal to the second preset number of times, and the total number of times the fertilizer is applied is less than or equal to the third preset number of times; And / or, the total irrigation water volume is less than or equal to the total irrigation volume threshold and the total nitrogen fertilizer amount is less than or equal to the nitrogen fertilizer amount threshold; The total irrigation threshold is determined based on the water requirement of Scutellaria baicalensis, and the nitrogen fertilizer threshold is determined based on the target yield and the basic nitrogen supply of the soil.

[0072] In some embodiments, the recommendation control includes a first control; when the target recommendation control is the first control, the recommendation strategy is a comprehensive scoring method. The selection module 203 is also used to determine the weights corresponding to the output values ​​of the multiple objective functions included in the prediction data by using the entropy weight method; Based on the weights corresponding to the output values ​​of the multiple objective functions, the comprehensive scores of each Pareto optimal solution are determined and ranked. The Pareto optimal solution with the highest overall score is determined as the target Pareto optimal solution.

[0073] In some embodiments, the recommendation control includes a second control; when the target recommendation control is the second control, the recommendation strategy is the knee-first method. The selection module 203 is also used to construct a Pareto front in a multidimensional target space based on the prediction data corresponding to multiple Pareto optimal solutions. Determine the vector angle formed between three consecutive Pareto optimal solutions on the Pareto front, and determine the position with the largest vector angle on the Pareto front as the knee point; The Pareto optimal solution corresponding to the knee point is determined as the target Pareto optimal solution.

[0074] In some embodiments, the recommendation control includes a third control; when the target recommendation control is the third control, the recommendation strategy is a robustness-first approach. The selection module 203 is also used to determine the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios through the multi-objective optimization model; Based on the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios, the coefficient of variation is determined; The Pareto optimal solution with the smallest coefficient of variation is determined as the target Pareto optimal solution.

[0075] In some embodiments, the recommendation control includes a fourth control; when the target recommendation control is the fourth control, the recommendation strategy is a context matching method. The selection module 203 is also used to determine the comprehensive score of each Pareto optimal solution based on the preset weight vector corresponding to the output values ​​of the multiple objective functions included in the prediction data. The Pareto optimal solution with the highest comprehensive score is determined as the target Pareto optimal solution.

[0076] In some embodiments, the acquisition module 201 is further configured to acquire actual data on the planting of Scutellaria baicalensis based on the target Pareto optimal solution in the current period; The determining module 202 is further configured to determine the deviation between the actual data and the predicted data corresponding to the target Pareto optimal solution; if the deviation is greater than a preset deviation threshold, the parameters in the multi-objective optimization model are iteratively updated based on the meteorological data, soil data and field data corresponding to the current period to obtain a new multi-objective optimization model, and a new target Pareto optimal solution is determined based on the new multi-objective optimization model; wherein, the new target Pareto optimal solution is used for planting Scutellaria baicalensis in the next period.

[0077] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of the invention, the sequence numbers of the above-described processes do not imply a sequential 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 invention. The sequence numbers of the above-described embodiments of the invention are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0078] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0079] In the several embodiments provided by this invention, it should be understood that the disclosed methods can be implemented in other ways. The methods disclosed in the several method embodiments provided by this invention can be arbitrarily combined without conflict to obtain new method embodiments. The features disclosed in the several method embodiments provided by this invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0080] The above description is merely an embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A multi-objective water and fertilizer synergistic optimization method for Scutellaria baicalensis, characterized in that, The method includes: The objective optimization problem is obtained and input into the RVEA-II algorithm. The objective optimization problem is determined based on multiple preset objective functions and constraints. The objective optimization problem is to find a Pareto optimal solution that maximizes the yield, quality, and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. The RVEA-II algorithm generates multiple candidate decision variables based on the objective optimization problem, and a preset multi-objective optimization model is invoked to determine the predicted data corresponding to each candidate decision variable. Based on the predicted data, multiple Pareto optimal solutions are determined from the multiple candidate decision variables. The Pareto optimal solutions are used to indicate the water and fertilizer strategies for Scutellaria baicalensis. The predicted data includes the output values ​​of multiple objective functions. The multi-objective optimization model is used to simulate the dynamic growth process of Scutellaria baicalensis under different water and fertilizer management strategies. In response to a user's selection instruction triggered by a target recommended control among multiple recommended controls, a target Pareto optimal solution is selected from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control.

2. The method according to claim 1, characterized in that, The plurality of objective functions include a first objective function, a second objective function, a third objective function, and a fourth objective function; wherein, the first objective function is used to maximize the yield of Scutellaria baicalensis; the second objective function is used to maximize the quality of Scutellaria baicalensis; the third objective function is used to maximize the resource utilization rate of Scutellaria baicalensis; and the fourth objective function is used to minimize the environmental risk of Scutellaria baicalensis.

3. The method according to claim 1, characterized in that, The constraints include at least the following: The root zone water content of Scutellaria baicalensis is greater than or equal to the water content at the wilting point of Scutellaria baicalensis and less than or equal to the field capacity. And / or, the soil electrical conductivity in the root zone is less than or equal to the highest salt concentration that Scutellaria baicalensis can tolerate; And / or, the number of times the first fertilizer is applied is less than or equal to the first preset number of times, the number of times the second fertilizer is applied is less than or equal to the second preset number of times, and the total number of times the fertilizer is applied is less than or equal to the third preset number of times; And / or, the total irrigation water volume is less than or equal to the total irrigation volume threshold and the total nitrogen fertilizer amount is less than or equal to the nitrogen fertilizer amount threshold; The total irrigation threshold is determined based on the water requirement of Scutellaria baicalensis, and the nitrogen fertilizer threshold is determined based on the target yield and the basic nitrogen supply of the soil.

4. The method according to claim 1, characterized in that, The recommendation control includes a first control; when the target recommendation control is the first control, the recommendation strategy is a comprehensive scoring method. The step of selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: The weights corresponding to the output values ​​of the multiple objective functions included in the predicted data are determined by the entropy weight method. Based on the weights corresponding to the output values ​​of multiple objective functions, the comprehensive score of each Pareto optimal solution is determined and ranked. The Pareto optimal solution with the highest overall score is determined as the target Pareto optimal solution.

5. The method according to claim 1, characterized in that, The recommendation control includes a second control; when the target recommendation control is the second control, the recommendation strategy is the knee-first method. The step of selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: Based on the prediction data corresponding to multiple Pareto optimal solutions, a Pareto front is constructed in a multidimensional target space; Determine the vector angle formed between three consecutive Pareto optimal solutions on the Pareto front, and determine the position with the largest vector angle on the Pareto front as the knee point; The Pareto optimal solution corresponding to the knee point is determined as the target Pareto optimal solution.

6. The method according to claim 1, characterized in that, The recommended control includes a third control; when the target recommended control is the third control, the recommendation strategy is a robustness-first approach. The step of selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: The multi-objective optimization model is used to determine the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios. Based on the prediction data of each Pareto optimal solution under multiple preset meteorological scenarios, the coefficient of variation is determined; The Pareto optimal solution with the smallest coefficient of variation is determined as the target Pareto optimal solution.

7. The method according to claim 1, characterized in that, The recommendation control includes a fourth control; when the target recommendation control is the fourth control, the recommendation strategy is a context matching method. The step of selecting the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommendation control includes: Based on the preset weight vectors corresponding to the output values ​​of the multiple objective functions included in the predicted data, the comprehensive score of each Pareto optimal solution is determined. The Pareto optimal solution with the highest overall score is determined as the target Pareto optimal solution.

8. The method according to claim 1, characterized in that, The method further includes: Obtain actual data on Scutellaria baicalensis planting in the current period based on the target Pareto optimal solution; Determine the deviation between the actual data and the predicted data corresponding to the target Pareto optimal solution; If the deviation exceeds a preset deviation threshold, the parameters in the multi-objective optimization model are iteratively updated based on the meteorological data, soil data, and field data corresponding to the current cycle to obtain a new multi-objective optimization model. A new target Pareto optimal solution is then determined based on the new multi-objective optimization model. The new target Pareto optimal solution is used for planting Scutellaria baicalensis in the next cycle.

9. A multi-objective water and fertilizer synergistic optimization system for Scutellaria baicalensis, characterized in that, The system includes: The acquisition module is used to acquire the target optimization problem and input the target optimization problem into the RVEA-II algorithm; the target optimization problem is determined based on multiple preset objective functions and constraints; the target optimization problem is to find the Pareto optimal solution that maximizes the yield, quality and resource utilization of Scutellaria baicalensis, minimizes environmental risk, and satisfies all constraints. The determination module is used to generate multiple candidate decision variables based on the objective optimization problem using the RVEA-II algorithm, and to call a preset multi-objective optimization model to determine the prediction data corresponding to each candidate decision variable. Based on the prediction data, multiple Pareto optimal solutions are determined from the multiple candidate decision variables. The Pareto optimal solutions are used to indicate the water and fertilizer strategies for Scutellaria baicalensis. The prediction data includes the output values ​​of multiple objective functions. The multi-objective optimization model is used to simulate the dynamic growth process of Scutellaria baicalensis under different water and fertilizer management strategies. The selection module is used to respond to a user's selection instruction triggered by a target recommended control among multiple recommended controls, and select the target Pareto optimal solution from multiple Pareto optimal solutions based on the recommendation strategy corresponding to the target recommended control.

Images