Full-life-cycle water and fertilizer management method and system for blueberry planting

Abstract

The invention provides a full-life-cycle water and fertilizer management method and system for blueberry planting, and relates to the field of crop fertilization, and the method comprises the steps: obtaining water and fertilizer management experiment data of a plurality of blueberry growth stages; based on the water and fertilizer management experimental data of the plurality of blueberry growth stages, establishing a growth-related weighted undirected graph of the plurality of blueberry growth stages; determining a key growth factor and a key water and fertilizer management factor of each blueberry growth stage based on the water and fertilizer management experimental data of the plurality of blueberry growth stages and the growth-related weighted undirected graph of the plurality of blueberry growth stages; based on the water and fertilizer management experiment data, the key growth factors and the key water and fertilizer management factors of the blueberry growth stage, determining candidate water and fertilizer management schemes of the blueberry growth stage; based on the candidate water and fertilizer management scheme of each blueberry growth stage, a water and fertilizer management scheme for the whole life cycle of blueberry planting is generated, and the method has the advantages that accurate supply and intelligent regulation of water and fertilizer are achieved, and the water and fertilizer utilization rate is increased.

Description

Technical Field

[0001] The present invention relates to the field of crop fertilization, and in particular to a water and fertilizer management method and system for the entire life cycle of blueberry planting. Background Art

[0002] Blueberries, scientifically known as Vaccinium vitis-idaea, belong to the genus Vaccinium in the Ericaceae family and are perennial deciduous or evergreen shrubs. Rich in antioxidants such as anthocyanins, they are widely used in health foods, pharmaceuticals, cosmetics, and other fields, with market demand continuing to rise. As consumers' interest in healthy foods continues to grow, market demand for blueberries is rapidly increasing, driving the blueberry cultivation industry toward large-scale, intensive, and efficient production.

[0003] Blueberry is an oligotrophic plant. A good fertilization method can not only increase yield and promote plant growth, but also improve the quality of blueberries.

[0004] In the existing technology, there is a lack of scientific irrigation time planning, and irrigation is often carried out based on the experience of growers. The irrigation time and amount cannot be adjusted in time according to the water demand patterns of blueberries in different growth stages and the soil moisture conditions. For example, during the flowering and fruit expansion period of blueberries, they are more sensitive to water demand. If irrigation is not timely or the irrigation amount is insufficient, it will affect flower bud differentiation and fruit development, resulting in a decrease in yield; and during the dormant period of blueberries, if irrigation is too much, it may cause diseases and affect the wintering of the plants. In addition, there is a lack of scientific methods for calculating the amount of fertilizer to be applied. The amount of fertilizer to be applied is often determined based on experience, and there is a problem of too much or too little fertilizer. Excessive fertilization will not only cause fertilizer waste and increase production costs, but may also cause soil salinization and cause poisoning to the blueberry root system; too little fertilization cannot meet the nutrient needs of blueberry growth and development, affecting the yield and quality of blueberries. At the same time, traditional water and fertilizer management methods often separate irrigation and fertilization, and water and fertilizer cannot be supplied synchronously, resulting in low fertilizer utilization. During the irrigation process, the fertilizer cannot be dissolved in time and transported to the blueberry root system with water. Some fertilizer will remain on the soil surface or be washed away by rain, causing fertilizer waste and environmental pollution.

[0005] Therefore, it is necessary to provide a water and fertilizer management method and system for the entire life cycle of blueberry cultivation, so as to achieve precise supply and intelligent regulation of water and fertilizer and improve water and fertilizer utilization efficiency. Summary of the Invention

[0006] The application provides a water and fertilizer management method for the whole life cycle of blueberry planting, comprising: obtaining water and fertilizer management experimental data of multiple blueberry growth stages, wherein the water and fertilizer management experimental data comprises water and fertilizer management schemes of multiple experimental planting areas, soilless cultivation substrate detection data and blueberry growth data, and the water and fertilizer management schemes of any two experimental planting areas are different; based on the water and fertilizer management experimental data of the multiple blueberry growth stages, a growth-related weighted undirected graph of the multiple blueberry growth stages is established; based on the water and fertilizer management experimental data of the multiple blueberry growth stages and the growth-related weighted undirected graph of the multiple blueberry growth stages, key growth factors and key water and fertilizer management factors of each blueberry growth stage are determined; for each blueberry growth stage, based on the water and fertilizer management experimental data of the blueberry growth stage, the key growth factors and the key water and fertilizer management factors, a candidate water and fertilizer management scheme of the blueberry growth stage is determined; and based on the candidate water and fertilizer management scheme of each blueberry growth stage, a water and fertilizer management scheme for the whole life cycle of blueberry planting is generated.

[0007] Further, based on the water and fertilizer management experimental data of the multiple blueberry growth stages, the growth-related weighted undirected graph of the multiple blueberry growth stages is established, comprising: determining growth monitoring factors of each blueberry growth stage; for any two blueberry growth stages, based on the water and fertilizer management experimental data of the two blueberry growth stages, factor values of each growth monitoring factor of the two blueberry growth stages in each experimental planting area are determined, and based on the factor values of each growth monitoring factor of the two blueberry growth stages in each experimental planting area, a growth correlation coefficient of any one growth monitoring factor of one blueberry growth stage and any one growth monitoring factor of another blueberry growth stage is calculated; and based on the growth correlation coefficient of any one growth monitoring factor of one blueberry growth stage and any one growth monitoring factor of another blueberry growth stage corresponding to any two blueberry growth stages, the growth-related weighted undirected graph of the multiple blueberry growth stages is established.

[0008] Furthermore, based on the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage corresponding to any two blueberry growth stages, a growth-related weighted undirected graph of multiple blueberry growth stages is established, including: for any two blueberry growth stages, based on the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage, determining a growth monitoring factor pair corresponding to the two blueberry growth stages, wherein the growth monitoring factor pair includes a growth monitoring factor for one blueberry growth stage and a growth monitoring factor for another blueberry growth stage; based on the growth monitoring factor pair corresponding to any two blueberry growth stages, establishing a growth-related weighted undirected graph of any two blueberry growth stages, wherein one vertex in the growth-related weighted undirected graph represents one growth monitoring factor, and the weight of the edge connecting two vertices of the growth monitoring factor pair is the absolute value of the growth correlation coefficient of the growth monitoring factor pair.

[0009] Furthermore, based on the water and fertilizer management experimental data of multiple blueberry growth stages and the growth-related weighted undirected graphs of multiple blueberry growth stages, the key growth factors and key water and fertilizer management factors of each blueberry growth stage are determined, including: determining the weight of the growth monitoring factor of each blueberry growth stage based on the growth-related weighted undirected graphs of multiple blueberry growth stages; determining the key growth factor of each blueberry growth stage based on the weight of the growth monitoring factor of each blueberry growth stage; for each blueberry growth stage, determining the key water and fertilizer management factors of the blueberry growth stage based on the water and fertilizer management experimental data of the blueberry growth stage and the key growth factors of the blueberry growth stage.

[0010] Furthermore, based on the growth-related weighted undirected graph of multiple blueberry growth stages, the weight of the growth monitoring factor of each blueberry growth stage is determined, including: S11, based on the growth-related weighted undirected graph of multiple blueberry growth stages, a weight optimization path is determined; S12, for each blueberry growth stage, based on the number of growth monitoring factors, the weight of each growth monitoring factor is initialized by the mean; S13, according to the weight optimization path, the current growth monitoring factor is determined; S14, according to the weight optimization path, the weight of the edge connecting the vertex corresponding to the current growth monitoring factor in the growth-related weighted undirected graph of multiple blueberry growth stages is determined. The weight of the growth monitoring factor connected by the edge is determined, and the weight correction value of the current growth monitoring factor is determined. According to the weight correction value of the current growth monitoring factor, the weight of the current growth monitoring factor is corrected; S15, determine whether the traversal of the weight optimization path is completed, if so, execute S16, if not, execute S13; S16, calculate the total amount of weight correction; S17, according to the total amount of weight correction, determine whether the correction end condition is met, if so, output the weight of the growth monitoring factor of each blueberry growth stage, if not, execute S18; S18, reset the weight optimization path and execute S13.

[0011] Furthermore, based on the experimental data of water and fertilizer management in the blueberry growth stage and the key growth factors of the blueberry growth stage, the key water and fertilizer management factors in the blueberry growth stage are determined, including: determining the key soilless cultivation substrate state factors based on the experimental data of water and fertilizer management in the blueberry growth stage and the key growth factors of the blueberry growth stage; determining the key water and fertilizer management factors in the blueberry growth stage based on the experimental data of water and fertilizer management in the blueberry growth stage and the key soilless cultivation substrate state factors.

[0012] Furthermore, based on the water and fertilizer management experimental data, key growth factors and key water and fertilizer management factors of the blueberry growth stage, candidate water and fertilizer management schemes for the blueberry growth stage are determined, including: obtaining blueberry yield and quality test data of multiple experimental planting areas; determining the optimal value range of the key growth factors for each blueberry growth stage based on the blueberry yield and quality test data of multiple experimental planting areas; determining the optimal value range of the key water and fertilizer management factors based on the optimal value range of the key growth factors in the blueberry growth stage, and generating candidate water and fertilizer management schemes for the blueberry growth stage.

[0013] Furthermore, based on the candidate water-fertilizer management schemes for each blueberry growth stage, a water-fertilizer management scheme for the entire life cycle of blueberry planting is generated, including: constructing a multi-objective optimization function; based on the candidate water-fertilizer management schemes for each blueberry growth stage, generating multiple candidate water-fertilizer management schemes for the entire life cycle of blueberry planting; calculating the multi-objective optimization function value of each candidate water-fertilizer management scheme for the entire life cycle of blueberry planting, and generating a water-fertilizer management scheme for the entire life cycle of blueberry planting.

[0014] Furthermore, the multi-objective optimization function is at least related to blueberry yield, blueberry quality and water and fertilizer costs.

[0015] The present invention provides a full-life cycle water and fertilizer management system for blueberry planting, which applies the above-mentioned full-life cycle water and fertilizer management method for blueberry planting, including: a data acquisition module, used to acquire water and fertilizer management experimental data for multiple blueberry growth stages, wherein the water and fertilizer management experimental data include water and fertilizer management plans for multiple experimental planting areas, soilless culture substrate detection data and blueberry growth data, and the water and fertilizer management plans for any two experimental planting areas are different; a data analysis module, used to establish a growth-related weighted undirected graph for any two blueberry growth stages based on the water and fertilizer management experimental data for the multiple blueberry growth stages, and determine key growth factors and key water and fertilizer management factors for each blueberry growth stage based on the water and fertilizer management experimental data for the multiple blueberry growth stages and the growth-related weighted undirected graph for the multiple blueberry growth stages; and a solution optimization module, used to determine, for each blueberry growth stage, a candidate water and fertilizer management plan for the blueberry growth stage based on the water and fertilizer management experimental data, key growth factors and key water and fertilizer management factors for the blueberry growth stage, and generate a water and fertilizer management plan for the full-life cycle of blueberry planting based on the candidate water and fertilizer management plans for each blueberry growth stage.

[0016] Compared with the existing technology, the water and fertilizer management method and system for the whole life cycle of blueberry planting provided by the present invention have at least the following beneficial effects: 1. Traditional methods often independently design water and fertilizer strategies for a single growth stage (such as the bulking stage and the flower bud differentiation stage), which can easily lead to parameter conflicts between stages (such as the contradiction between high potassium in the bulking stage and nitrogen control in the flower bud differentiation stage). The present invention integrates data from the entire cycle and establishes a cross-stage parameter linkage model to ensure the synergy of water and fertilizer plans in each stage in the time dimension. By collecting and analyzing multi-regional and differentiated water and fertilizer management experimental data, empirical bias is eliminated and the scientific nature and universality of the plan are improved. Based on a growth-related weighted undirected graph, the correlation between growth factors in each stage is quantified, key growth factors are accurately located, and stage-by-stage precise management is achieved. Combining experimental data with the graph model, key growth factors and water and fertilizer management elements for each growth stage are dynamically screened to avoid resource waste and improve management efficiency. Candidate plans for each stage are integrated to form a water and fertilizer management strategy covering the entire life cycle of blueberries, ensuring growth continuity and improving overall planting efficiency.

[0017] 2. By constructing a growth-related weighted undirected graph, the correlations between multi-stage growth monitoring factors are quantified in the form of edge weights, eliminating the problem of neglecting implicit relationships between factors in traditional methods and achieving dynamic visualization of factor influence. An iterative correction algorithm is used to calculate factor weights. Through a closed-loop process of "initialization - path traversal - correction value calculation - total amount verification", subjective assignment bias is avoided, and weight distribution is more closely aligned with actual growth response patterns. Key growth factors are screened based on weight ranking, further traced back to key substrate status factors, and ultimately key water and fertilizer management factors are identified, forming a causal traceability chain of "growth performance - substrate status - water and fertilizer operation", enhancing management objectives by over 50%.

[0018] 3. Based on multi-regional yield and quality testing data, statistical analysis is used to determine the optimal range of key growth factors, avoiding data bias in a single region and ensuring that management targets are more aligned with actual production needs. By linking key growth factors with key water and fertilizer management factors, optimal ranges for water and fertilizer parameters are reversely derived, forming a precise mapping of "growth target-water and fertilizer operation" and reducing resource waste caused by blind regulation. Based on candidate solutions for each stage, a full-cycle solution is generated using a multi-objective optimization function to maximize benefits throughout the entire growth period. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein: Figure 1 is a flow chart of a method for water and fertilizer management throughout the life cycle of blueberry cultivation according to some embodiments of this specification; Figure 2 is a schematic diagram of a growth-related weighted undirected graph according to some embodiments of this specification; Figure 3 is a schematic diagram of a process for determining the weight of growth monitoring factors for each blueberry growth stage according to some embodiments of this specification; Figure 4 is a schematic diagram of a process for determining a weight optimization path according to some embodiments of this specification; Figure 5 This is a module diagram of a water and fertilizer management system for the entire life cycle of blueberry cultivation according to some embodiments of this specification. DETAILED DESCRIPTION

[0020] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0021] Figure 1 is a flow chart of a method for water and fertilizer management throughout the life cycle of blueberry cultivation according to some embodiments of this specification, such as Figure 1 As shown, the water and fertilizer management method for the whole life cycle of blueberry cultivation can include the following steps.

[0022] Step 110: Acquire water and fertilizer management experimental data for multiple blueberry growth stages.

[0023] Among them, the water and fertilizer management experimental data includes water and fertilizer management plans for multiple experimental planting areas, soilless cultivation substrate detection data and blueberry growth data. The water and fertilizer management plans for any two experimental planting areas are different.

[0024] Specifically, the multiple blueberry growth stages to which this method is applied may include at least: Vegetative growth period (1-2 years after planting): The root system and branches and leaves develop rapidly, and high nitrogen fertilizer is required to promote growth.

[0025] Flower bud differentiation period (late autumn to early winter): water and nitrogen need to be controlled, and phosphorus and potassium fertilizers should be increased to promote flower bud formation.

[0026] Flowering and fruiting period (spring): A stable water supply is required, and trace elements such as boron and calcium are supplemented to prevent fruit drop.

[0027] Fruit expansion period (30-60 days after flowering): high potassium fertilizer is needed to improve fruit quality, and moderate water control is required to increase sugar accumulation.

[0028] Post-harvest recovery period: Nitrogen and phosphorus fertilizers need to be added to restore the tree's vigor and reserve nutrients for the next year's growth.

[0029] For each blueberry growing stage, the water and fertilizer management plan for each experimental plot must differ in at least one core variable (e.g., irrigation frequency, nitrogen, phosphorus, and potassium ratios, and the ratio of organic to inorganic fertilizers). All other conditions (e.g., light intensity, temperature, and soilless culture medium type) must be kept as consistent as possible. Management plans should be randomly assigned to experimental plots to avoid location effects (e.g., differences in light intensity at the edges of the plot) that could interfere with the results.

[0030] The water and fertilizer management plan can be a comprehensive irrigation and fertilization strategy developed for different growth stages of blueberries, including irrigation strategies (for example, the frequency and single water volume of drip irrigation, micro-sprinkler irrigation or flooding (such as 2-3L / plant each time)), fertilizer ratio (for example, the ratio of nitrogen, phosphorus and potassium, etc.), concentration, fertilization frequency, etc.

[0031] For each blueberry growth stage, only one or two key parameters of the water and fertilizer management plan (such as irrigation frequency or fertilizer ratio) were varied within each experimental plot, while other conditions (such as substrate type and light intensity) remained consistent. The parameters of the water and fertilizer management plan should cover the actual application range (e.g., irrigation frequency settings of 3, 5, and 7 days). The experimental plots were randomly assigned numbers to avoid location effects (e.g., differences in ventilation between marginal areas).

[0032] As an example only, the water and fertilizer management plans for the three experimental planting areas during the fruit expansion period can be shown in Table 1.

[0033] Table 1 ; The water and fertilizer management schemes for multiple experimental planting plots at multiple blueberry growth stages can be determined in any manner. For example, the schemes can be set manually based on experience. Another example is setting the water and fertilizer management schemes for multiple experimental planting plots at multiple blueberry growth stages using orthogonal experiments.

[0034] Soilless culture media (e.g., coconut coir, peat, perlite, rockwool, sawdust, etc., single or mixed) are solid growing media that replace soil, providing physical support and nutrient supply for blueberry roots. Soilless culture media testing data can include values ​​for soilless culture media status factors (e.g., pH, conductivity, porosity, etc.).

[0035] Blueberry growth data is used to quantitatively describe the morphology, physiological state, and yield of blueberry plants. This data can include values ​​for multiple growth factors, including morphological factors (e.g., plant height, crown width, shoot length, leaf number), physiological factors (e.g., photosynthetic rate, stomatal conductance, leaf nutrient content (e.g., nitrogen, phosphorus, and potassium concentrations), and yield factors (e.g., number of fruits per plant, fruit weight, and total yield). Different growth factors can correspond to different blueberry growth stages, as shown in Table 2.

[0036] Table 2 ; Step 120: establishing a growth-related weighted undirected graph for multiple blueberry growth stages based on the water and fertilizer management experimental data for multiple blueberry growth stages.

[0037] Specifically include: Determine growth monitoring factors for each blueberry growth stage; For any two blueberry growth stages, based on the water and fertilizer management experimental data of the two blueberry growth stages, determine the factor value of each growth monitoring factor in each experimental planting area of ​​the two blueberry growth stages; based on the factor value of each growth monitoring factor in each experimental planting area of ​​the two blueberry growth stages, calculate the growth correlation coefficient between any growth monitoring factor in one blueberry growth stage and any growth monitoring factor in another blueberry growth stage; Based on the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage corresponding to any two blueberry growth stages, a growth correlation weighted undirected graph of multiple blueberry growth stages is established.

[0038] Specifically, the factor value of each growth monitoring factor in each experimental planting area during a blueberry growth stage can be the value at a certain moment (e.g., the end moment) during the blueberry growth stage. Any growth monitoring factor in one blueberry growth stage and any growth monitoring factor in another blueberry growth stage can be used as two dependent variables. The factor values ​​of the dependent variables in each experimental planting area can be substituted into a correlation coefficient calculation formula (e.g., Pearson correlation coefficient, Spearman rank correlation coefficient, etc.) to calculate the growth correlation coefficient between any growth monitoring factor in one blueberry growth stage and any growth monitoring factor in another blueberry growth stage.

[0039] As an example only, taking the length of new shoots during the vegetative growth period and the transverse diameter of the fruit during the fruit expansion period as an example, the length of new shoots during the vegetative growth period and the transverse diameter of the fruit during the fruit expansion period of each experimental planting area are substituted into the correlation coefficient formula to calculate the growth correlation coefficient of the length of new shoots during the vegetative growth period and the transverse diameter of the fruit during the fruit expansion period.

[0040] Preferably, based on the growth correlation coefficient between any one growth monitoring factor of one blueberry growth stage and any one growth monitoring factor of another blueberry growth stage corresponding to any two blueberry growth stages, a growth correlation weighted undirected graph of multiple blueberry growth stages is established, including: For any two blueberry growth stages, based on the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage, a growth monitoring factor pair corresponding to the two blueberry growth stages is determined, wherein the growth monitoring factor pair includes a growth monitoring factor of one blueberry growth stage and a growth monitoring factor of another blueberry growth stage. For example, if the absolute value of the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage is greater than a preset threshold (for example, 0.5), then the two growth monitoring factors are used as the growth monitoring factor pair corresponding to the two blueberry growth stages. For example, if the absolute value of the growth correlation coefficient between the new shoot length in the vegetative growth period and the transverse diameter of the fruit in the fruit expansion period is greater than the preset threshold (for example, 0.5), then a growth monitoring factor pair between the vegetative growth period and the fruit expansion period is: new shoot length - fruit transverse diameter. It can be understood that any two growth monitoring factor pairs can have at most one identical growth monitoring factor. Based on the growth monitoring factor pairs corresponding to any two blueberry growth stages, a growth-related weighted undirected graph of any two blueberry growth stages is established, where Figure 2 As shown, in the growth-related weighted undirected graph, one vertex represents one growth monitoring factor, and the weight of the edge connecting two vertices of a growth monitoring factor pair is the absolute value of the growth correlation coefficient of the growth monitoring factor pair.

[0041] Step 130 : determining key growth factors and key water and fertilizer management factors for each blueberry growth stage based on the water and fertilizer management experimental data for multiple blueberry growth stages and the growth-related weighted undirected graphs for multiple blueberry growth stages.

[0042] Specifically include: Based on a growth-related weighted undirected graph of multiple blueberry growth stages, the weight of the growth monitoring factor of each blueberry growth stage is determined; Based on the weights of the growth monitoring factors at each blueberry growth stage, the key growth factors at each blueberry growth stage are determined; For each blueberry growth stage, the key water and fertilizer management factors of the blueberry growth stage are determined based on the water and fertilizer management experimental data of the blueberry growth stage and the key growth factors of the blueberry growth stage.

[0043] Figure 3 is a flow chart of determining the weight of the growth monitoring factor for each blueberry growth stage according to some embodiments of this specification, such as Figure 3 As shown, preferably, based on a weighted undirected graph of growth-related factors of multiple blueberry growth stages, the weight of the growth monitoring factor of each blueberry growth stage is determined, including: S11. Determine a weighted optimization path based on a growth-related weighted undirected graph of multiple blueberry growth stages; S12. For each blueberry growth stage, initialize the weight of each growth monitoring factor based on the number of growth monitoring factors. For example, the initial value of the weight of each growth monitoring factor = 1 / n, where n is the total number of growth monitoring factors in the blueberry growth stage. S13. Determine the current growth monitoring factor based on the weight optimization path; S14. Determine a weight correction value of the current growth monitoring factor based on the weight of the edge connected to the vertex corresponding to the current growth monitoring factor and the weight of the growth monitoring factor connected to the edge in a growth-related weighted undirected graph of multiple blueberry growth stages. Correct the weight of the current growth monitoring factor based on the weight correction value of the current growth monitoring factor. Specifically, perform a weighted summation on the weight of the edge connected to the vertex corresponding to the current growth monitoring factor and the weight of the growth monitoring factor connected to the edge to obtain the weight correction value of the current growth monitoring factor. Summing the weight of the current growth monitoring factor and the weight correction value obtains the corrected weight of the current growth monitoring factor. S15: Determine whether the traversal of the weight optimization path is completed, that is, whether the last growth monitoring factor in the weight optimization path is accessed. If so, execute S16; if not, execute S13; S16. Calculate the total amount of weight correction, wherein the total amount of weight correction is the sum of the absolute values ​​of the difference between the weight before correction and the weight after correction of each growth monitoring factor; S17. Determine whether a correction end condition is met based on the total weight correction amount. If so, output the weight of the growth monitoring factor for each blueberry growth stage. If not, execute S18. The correction end condition may be that the number of iterations reaches a maximum value (e.g., 20 times), the total weight correction amount of two consecutive iterations is less than a weight correction total threshold (e.g., 2), etc. S18. After completing one iteration, the weight optimization path is reset, i.e., the weight optimization path is traversed from the beginning, and S13 is executed.

[0044] For example, the weight correction value of the current growth monitoring factor can be calculated based on the following formula: ; in, is the weight correction value of the current growth monitoring factor (i.e., the i-th growth monitoring factor), is the weight of the nth edge of the vertex corresponding to the current growth monitoring factor in the growth-related weighted undirected graph, The weight of the growth monitoring factor corresponding to the vertex connected by the nth edge of the vertex corresponding to the current growth monitoring factor in the growth-related weighted undirected graph, The total number of edges corresponding to the vertices of the current growth monitoring factor in the growth-related weighted undirected graph.

[0045] It is understandable that the above process uses edge weights and iterative correction mechanisms in a weighted undirected graph to output a weight value that directly quantifies the contribution of each growth monitoring factor to blueberry growth. Based on the growth-related weighted undirected graph, the direct correlation between monitoring factors is directly quantified using edge weights, avoiding subjective assignment bias. The weights are initialized by the mean (1 / n ) ensures fairness and provides a neutral starting point for subsequent optimization. A weighted neighbor factor correction mechanism is employed: the weight correction value of the current factor is calculated by the weighted sum of its neighbor factor weights and the connecting edge weights. This comprehensively considers both direct connections (edge ​​weights) and indirect control (neighbor factor weight transfer), avoiding bias caused by reliance on a single path. Multiple iterations gradually smooth data noise, allowing weights to converge to a global optimum and enhancing robustness. This process achieves objectivity, precision, and adaptability in weight allocation through quantitative associations, dynamic optimization, and staged adaptation.

[0046] Figure 4 is a flow chart of determining a weight optimization path according to some embodiments of this specification, such as Figure 4 As shown, determining the weight optimization path may include the following steps: S21. Generate multiple weighted undirected graphs, and determine the optimal weight optimization path of each weighted undirected graph; S22. For each sample weighted undirected graph, determine the values ​​of multiple graph characteristic factors (e.g., the number of connected edges, the sum of the weights of the connected edges, the average number of edges of connected vertices, the average of the sum of the weights of the edges of connected vertices, etc.) for each vertex and the sequence number of the vertex in the optimal weighted optimization path. For each graph characteristic factor, substitute the value of the graph characteristic factor of each vertex and the sequence number of the vertex in the optimal weighted optimization path into a calculation formula of a correlation coefficient (e.g., a Pearson correlation coefficient, a Spearman rank correlation coefficient, etc.) to calculate the correlation coefficient between the graph characteristic factor and the sequence number corresponding to the sample weighted undirected graph; S23. For each graph feature factor, calculate the mean of the correlation coefficients between the graph feature factor and the sequence number corresponding to each sample weighted undirected graph, and use this as the mean correlation coefficient of the graph feature factor; S24. Graph feature factors whose absolute values ​​of the correlation coefficient mean are greater than a mean threshold (e.g., 0.4) are used as key graph feature factors, and the correlation coefficient mean is used as a weight of the key graph feature factor. S25. For each vertex in the growth-related weighted undirected graph, determine a factor value of a key graph characteristic factor of the vertex, perform weighted summation on the factor values ​​of the key graph characteristic factor of the vertex according to the weight of the key graph characteristic factor, and calculate a ranking value of the vertex; S26. Sort the vertices in the growth-related weighted undirected graph in descending order according to the ranking value of each vertex in the growth-related weighted undirected graph to generate a weighted optimization path.

[0047] As can be understood, the above process generates multiple weighted undirected graphs and determines the optimal path for each. By comparing multiple paths, the limitations of a single graph structure or optimization strategy are avoided, global search capabilities are enhanced, and the robustness and adaptability of the final path are ensured. The mean correlation coefficient between each graph feature and the optimal path sequence is calculated, and key features with absolute values ​​greater than a threshold are screened. Factors strongly correlated with path ranking are retained, while noisy features are removed. Through multi-graph exploration, feature quantification, correlation screening, and dynamic sorting, a scientific, precise, and interpretable weight optimization path is achieved, providing a complete solution for assigning weights to blueberry growth monitoring factors, from graph structure analysis to path generation.

[0048] Specifically, the optimal weight optimization path of each weighted undirected graph can be determined according to the following process: S31. Randomly generate multiple weight optimization paths according to the weighted undirected graph. Specifically, randomly determine the sequence number of each vertex in the weight optimization path from the weighted undirected graph, thereby generating the weight optimization path; S32. Setting multiple path evaluation indicators, such as an iteration count indicator, a mean indicator of the total weight correction of a single iteration, etc.; S33. Based on the process of S13-S18, determine the correction result of each weight optimization path, wherein the correction result includes the number of iterations and the average of the total weight correction amount of a single iteration; S34. For each weighted optimization path, determine the value of each path evaluation index of the weighted optimization path, wherein a greater number of iterations results in a smaller score for the number of iterations index, a greater mean value of the total weight corrections for a single iteration, and a smaller score for the mean value index of the total weight corrections for a single iteration. Sum the values ​​of each path evaluation index for the weighted optimization path to obtain an optimized value for the weighted optimization path; S35. The weighted optimization path with the largest optimization value is taken as the optimal weighted optimization path.

[0049] Understandably, by randomly assigning vertex numbers to generate multiple paths, we can overcome the limitations of a single initial path, cover a wider solution space, avoid being trapped in local optimality, and improve the probability of global optimization. Metrics such as the number of iterations and the average total amount of single corrections are set to evaluate the quality of paths from the dual dimensions of efficiency and stability. For example, a high number of iterations may indicate slow path convergence, while a large average total amount of corrections suggests drastic weight fluctuations, both of which need to be suppressed. By using a comprehensive score to select the path with the highest optimization value, we ensure that the final path performs best in the global search and provide a data foundation for the subsequent determination of the weight optimization path.

[0050] For each blueberry growth stage, the weights of the growth monitoring factors of each blueberry growth stage can be normalized, and the growth monitoring factors whose normalized weights are greater than a weight threshold (for example, 0.2) are used as key growth monitoring factors.

[0051] Preferably, based on the experimental data of water and fertilizer management during the blueberry growth stage and the key growth factors during the blueberry growth stage, the key water and fertilizer management factors during the blueberry growth stage are determined, including: Based on the experimental data of water and fertilizer management during the growth stage of blueberry and the key growth factors of blueberry during the growth stage, the key soilless culture substrate status factors are determined; Based on the experimental data of water and fertilizer management during the blueberry growth stage and the key soilless cultivation substrate status factors, the key water and fertilizer management factors during the blueberry growth stage were determined.

[0052] Specifically, referring to the calculation method of the growth correlation coefficient of the two growth monitoring factors, the correlation analysis of the soilless culture matrix state factor and the key growth factor is performed, and the correlation coefficient of the soilless culture matrix state factor and the key growth factor is calculated.

[0053] For each soilless cultivation medium state factor, the mean of the absolute values ​​of the correlation coefficients between the soilless cultivation medium state factor and each key growth factor is calculated, and the soilless cultivation medium state factor with a mean greater than the mean threshold (e.g., 0.4) is taken as the key soilless cultivation medium state factor.

[0054] Referring to the calculation method of the growth correlation coefficients of the two growth monitoring factors, a correlation analysis was conducted on the water and fertilizer management factors and the key soilless culture substrate status factors, and the correlation coefficients between the soilless culture substrate status factors and the key growth factors were calculated.

[0055] For each water and fertilizer management factor, the mean of the absolute values ​​of the correlation coefficients between the soilless culture medium status factor and each key growth factor was calculated, and the water and fertilizer management factors with a mean greater than the mean threshold (e.g., 0.4) were regarded as key water and fertilizer management factors.

[0056] Step 140 : For each blueberry growth stage, determine a candidate water and fertilizer management plan for the blueberry growth stage based on the water and fertilizer management experimental data, key growth factors, and key water and fertilizer management factors for the blueberry growth stage.

[0057] Specifically include: Obtain blueberry yield and quality testing data from multiple experimental planting areas; Based on the blueberry yield and quality test data of multiple experimental planting areas, the optimal value range of the key growth factor at each blueberry growth stage is determined. Specifically, the blueberry yield and quality test data of multiple experimental planting areas are used to determine the preferred experimental planting area. For example, the data of the top 20% in yield / quality are used as the preferred experimental planting area. The factor values ​​of the key growth factors at each blueberry growth stage of the preferred experimental planting area are obtained, and the maximum and minimum values ​​of the key growth factors are taken to form the optimal value range of the key growth factors; Based on the preferred value range of key growth factors in the blueberry growth stage, the preferred value range of key water and fertilizer management factors is determined, and candidate water and fertilizer management plans for the blueberry growth stage are generated. Specifically, the experimental planting areas where the factor values ​​of the key growth factors in the blueberry growth stage are within the preferred value range are used as candidate experimental planting areas. According to the maximum and minimum values ​​of the key water and fertilizer management factors in the water and fertilizer management plans of the candidate experimental planting areas, the preferred value range of the key water and fertilizer management factors is determined.

[0058] For each key water and fertilizer management factor, a value is sampled from the preferred value interval of the key water and fertilizer management factor as the value of the key water and fertilizer management factor in the candidate water and fertilizer management scheme, thereby generating multiple candidate water and fertilizer management schemes for the blueberry growth stage.

[0059] Step 150 : generating a water and fertilizer management plan for the entire life cycle of blueberry planting based on the candidate water and fertilizer management plans for each blueberry growth stage.

[0060] Specifically include: Constructing a multi-objective optimization function, wherein the multi-objective optimization function is at least related to blueberry yield, blueberry quality, and water and fertilizer costs; Based on the candidate water-fertilizer management plans for each blueberry growth stage, multiple candidate water-fertilizer management plans for the entire life cycle of blueberry cultivation are generated, wherein, for each blueberry growth stage, a candidate water-fertilizer management plan is sampled from the multiple candidate water-fertilizer management plans for the blueberry growth stage, and the sampled candidate water-fertilizer management plans corresponding to each blueberry growth stage are combined to generate a candidate water-fertilizer management plan for the entire life cycle of blueberry cultivation; Calculate the multi-objective optimization function value of each candidate water and fertilizer management plan for the entire life cycle of blueberry planting, and generate a water and fertilizer management plan for the entire life cycle of blueberry planting.

[0061] Specifically, blueberry yield is quantified by fruit weight per unit area (kg / mu). Quality is based on indicators such as sugar content (°Brix), firmness (kg / cm²), and anthocyanin content (mg / 100g). Principal component analysis is used to weight the scores and calculate a comprehensive score. Water and fertilizer costs include direct expenses such as fertilizer procurement, irrigation energy consumption, and manual application.

[0062] The normalized weighted summation method is used to construct the multi-objective function and eliminate the dimensional difference.

[0063] The result prediction model can be used to predict blueberry yield and quality based on candidate water and fertilizer management plans for the entire blueberry cultivation lifecycle. The result prediction model can be a long-short-term memory network model. It uses a bidirectional structure to capture the forward effects of water and fertilizer inputs (e.g., early nitrogen fertilizer promotes vegetative growth) and backward feedback (e.g., the constraints of later quality on earlier water and fertilizer inputs). The hidden layer dimension is set to 64, and the output dimension is 128 (bidirectional concatenation). Yield and quality (e.g., sugar content, firmness, anthocyanin content, etc.) are predicted separately using independent fully connected layers. 200 sets of historical planting data were collected, and 800 sets of simulation data were generated. The training, validation, and test sets were divided into 8:1:1 ratios. The yield task uses mean square error, and the quality task uses weighted mean square error of multi-dimensional indicators (for example, sugar content weight 0.5, hardness 0.3, anthocyanin 0.2). The total loss is the weighted sum of the two tasks (weight 1:0.8). The Adam optimizer (learning rate 0.001) is used with learning rate decay (decaying to 0.9 times every 10 rounds). The batch size is set to 32, the training rounds are 200 rounds, and early stopping (Patience=20) is used to prevent overfitting.

[0064] Based on the output of the result prediction model, the price and amount of fertilizers and water usage used in the candidate water-fertilizer management plans, the multi-objective optimization function value of the candidate water-fertilizer management plans for the entire life cycle of blueberry cultivation is calculated, and the candidate water-fertilizer management plan for the entire life cycle of blueberry cultivation with the largest multi-objective optimization function value is used as the water-fertilizer management plan for the entire life cycle of blueberry cultivation.

[0065] Figure 5 This is a module diagram of a water and fertilizer management system for the entire life cycle of blueberry cultivation according to some embodiments of this specification, such as Figure 5 As shown, the water and fertilizer management system for the entire life cycle of blueberry planting may include a data acquisition module, a data analysis module, and a solution optimization module.

[0066] A data acquisition module is used to obtain water and fertilizer management experimental data for multiple blueberry growth stages, wherein the water and fertilizer management experimental data includes water and fertilizer management plans for multiple experimental planting areas, soilless culture substrate detection data, and blueberry growth data. The water and fertilizer management plans for any two experimental planting areas are different; A data analysis module is used to establish a growth-related weighted undirected graph between any two blueberry growth stages based on the water and fertilizer management experimental data of multiple blueberry growth stages, and to determine the key growth factors and key water and fertilizer management factors for each blueberry growth stage based on the water and fertilizer management experimental data of multiple blueberry growth stages and the growth-related weighted undirected graph of multiple blueberry growth stages; The scheme optimization module is used to determine the candidate water and fertilizer management schemes for each blueberry growth stage based on the water and fertilizer management experimental data, key growth factors and key water and fertilizer management factors of the blueberry growth stage, and generate a water and fertilizer management scheme for the entire life cycle of blueberry cultivation based on the candidate water and fertilizer management schemes for each blueberry growth stage.

[0067] The water and fertilizer management system for the whole life cycle of blueberry cultivation can apply the above-mentioned water and fertilizer management method for the whole life cycle of blueberry cultivation, which will not be repeated here.

[0068] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A water and fertilizer management method for the entire life cycle of blueberry cultivation, characterized in that: include: Obtain water and fertilizer management experimental data for multiple blueberry growth stages. The water and fertilizer management experimental data includes water and fertilizer management plans for multiple experimental planting areas, soilless culture substrate testing data, and blueberry growth data. The water and fertilizer management plans for any two experimental planting areas are different. Based on the experimental data of water and fertilizer management in multiple blueberry growth stages, a growth-related weighted undirected graph of multiple blueberry growth stages was established; Based on the experimental data of water and fertilizer management in multiple blueberry growth stages and the growth-related weighted undirected graphs of multiple blueberry growth stages, the key growth factors and key water and fertilizer management factors of each blueberry growth stage are determined; For each blueberry growth stage, determine the candidate water and fertilizer management scheme for the blueberry growth stage based on the water and fertilizer management experimental data, key growth factors and key water and fertilizer management factors of the blueberry growth stage; Based on the candidate water and fertilizer management plans for each blueberry growth stage, a water and fertilizer management plan for the entire life cycle of blueberry cultivation is generated.

2. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 1, characterized in that: Based on the experimental data of water and fertilizer management in multiple blueberry growth stages, a growth-related weighted undirected graph of multiple blueberry growth stages was established, including: Determine growth monitoring factors for each blueberry growth stage; For any two blueberry growth stages, based on the water and fertilizer management experimental data of the two blueberry growth stages, determine the factor value of each growth monitoring factor in each experimental planting area of ​​the two blueberry growth stages; based on the factor value of each growth monitoring factor in each experimental planting area of ​​the two blueberry growth stages, calculate the growth correlation coefficient between any growth monitoring factor in one blueberry growth stage and any growth monitoring factor in another blueberry growth stage; Based on the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage corresponding to any two blueberry growth stages, a growth correlation weighted undirected graph of multiple blueberry growth stages is established.

3. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 2, characterized in that: Based on the growth correlation coefficient between any growth monitoring factor of one blueberry growth stage and any growth monitoring factor of another blueberry growth stage corresponding to any two blueberry growth stages, a growth correlation weighted undirected graph of multiple blueberry growth stages is established, including: For any two blueberry growth stages, determining a growth monitoring factor pair corresponding to the two blueberry growth stages based on a growth correlation coefficient between any growth monitoring factor in one blueberry growth stage and any growth monitoring factor in another blueberry growth stage, wherein the growth monitoring factor pair includes a growth monitoring factor in one blueberry growth stage and a growth monitoring factor in another blueberry growth stage; Based on the growth monitoring factor pairs corresponding to any two blueberry growth stages, a growth-related weighted undirected graph of any two blueberry growth stages is established, wherein one vertex in the growth-related weighted undirected graph represents one growth monitoring factor, and the weight of the edge connecting the two vertices of the growth monitoring factor pair is the absolute value of the growth correlation coefficient of the growth monitoring factor pair.

4. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 3, characterized in that: Based on the experimental data of water and fertilizer management at multiple blueberry growth stages and the growth-related weighted undirected graphs of multiple blueberry growth stages, the key growth factors and key water and fertilizer management factors for each blueberry growth stage are determined, including: Based on a growth-related weighted undirected graph of multiple blueberry growth stages, the weight of the growth monitoring factor of each blueberry growth stage is determined; Based on the weights of the growth monitoring factors at each blueberry growth stage, the key growth factors at each blueberry growth stage are determined; For each blueberry growth stage, the key water and fertilizer management factors of the blueberry growth stage are determined based on the water and fertilizer management experimental data of the blueberry growth stage and the key growth factors of the blueberry growth stage.

5. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 4, characterized in that: Based on the growth-related weighted undirected graph of multiple blueberry growth stages, the weight of the growth monitoring factor for each blueberry growth stage is determined, including: S11. Determine a weighted optimization path based on a growth-related weighted undirected graph of multiple blueberry growth stages; S12. For each blueberry growth stage, initialize the weight of each growth monitoring factor based on the number of growth monitoring factors; S13. Determine the current growth monitoring factor based on the weight optimization path; S14. Determine a weight correction value of the current growth monitoring factor based on the weights of the edges connected to the vertices corresponding to the current growth monitoring factor and the weights of the growth monitoring factors connected to the edges in a growth-related weighted undirected graph of multiple blueberry growth stages, and correct the weight of the current growth monitoring factor based on the weight correction value of the current growth monitoring factor; S15: Determine whether the traversal of the weight optimization path is completed. If so, execute S16; if not, execute S13; S16. Calculate the total amount of weight correction; S17. Determine whether the correction end condition is met based on the total weight correction amount. If so, output the weight of the growth monitoring factor for each blueberry growth stage. If not, execute S18. S18: Reset the weight optimization path and execute S13.

6. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 4, characterized in that: Based on the experimental data of water and fertilizer management during the blueberry growth stage and the key growth factors of blueberry growth stage, the key water and fertilizer management factors during the blueberry growth stage are determined, including: Based on the experimental data of water and fertilizer management during the growth stage of blueberry and the key growth factors of blueberry during the growth stage, the key soilless culture substrate status factors are determined; Based on the experimental data of water and fertilizer management during the blueberry growth stage and the key soilless cultivation substrate status factors, the key water and fertilizer management factors during the blueberry growth stage were determined.

7. The water and fertilizer management method for the whole life cycle of blueberry planting according to any one of claims 1 to 6, characterized in that: Based on the experimental data of water and fertilizer management during the blueberry growth stage, key growth factors and key water and fertilizer management factors, candidate water and fertilizer management plans for the blueberry growth stage are determined, including: Obtain blueberry yield and quality testing data from multiple experimental planting areas; Based on the blueberry yield and quality test data from multiple experimental planting areas, the optimal value range of key growth factors for each blueberry growth stage was determined; Based on the optimal value ranges of key growth factors during the blueberry growth stage, the optimal value ranges of key water and fertilizer management factors are determined, and candidate water and fertilizer management plans for the blueberry growth stage are generated.

8. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 7, characterized in that: Based on the candidate water and fertilizer management plans for each blueberry growth stage, a water and fertilizer management plan for the entire life cycle of blueberry cultivation is generated, including: Construct multi-objective optimization functions; Based on the candidate water and fertilizer management plans for each blueberry growth stage, multiple candidate water and fertilizer management plans for the entire life cycle of blueberry cultivation are generated; Calculate the multi-objective optimization function value of each candidate water and fertilizer management plan for the entire life cycle of blueberry planting, and generate a water and fertilizer management plan for the entire life cycle of blueberry planting.

9. The water and fertilizer management method for the whole life cycle of blueberry planting according to claim 8, characterized in that: The multi-objective optimization function is at least related to blueberry yield, blueberry quality and water and fertilizer costs.

10. A water and fertilizer management system for the entire life cycle of blueberry cultivation, characterized in that: The water and fertilizer management method for the whole life cycle of blueberry planting according to any one of claims 1 to 9 comprises: A data acquisition module is used to obtain water and fertilizer management experimental data for multiple blueberry growth stages, wherein the water and fertilizer management experimental data includes water and fertilizer management plans for multiple experimental planting areas, soilless culture substrate detection data, and blueberry growth data. The water and fertilizer management plans for any two experimental planting areas are different; A data analysis module is used to establish a growth-related weighted undirected graph between any two blueberry growth stages based on the water and fertilizer management experimental data of multiple blueberry growth stages, and to determine the key growth factors and key water and fertilizer management factors for each blueberry growth stage based on the water and fertilizer management experimental data of multiple blueberry growth stages and the growth-related weighted undirected graph of multiple blueberry growth stages; The scheme optimization module is used to determine the candidate water and fertilizer management schemes for each blueberry growth stage based on the water and fertilizer management experimental data, key growth factors and key water and fertilizer management factors of the blueberry growth stage, and generate a water and fertilizer management scheme for the entire life cycle of blueberry cultivation based on the candidate water and fertilizer management schemes for each blueberry growth stage.

Images