Fertilizing method for realizing multi-economic-target yield increase of gleditsia sinensis

Through a multi-objective fertilization method based on mixed effect model and optimization algorithm, the accuracy of soil fertilizer application of acacia tree is solved, the optimization of various economic indicators and the optimal state of soil nutrients is achieved, and the multi-economic fertilization of acacia tree is improved. Target efficiency.

CN118901366AActive Publication Date: 2024-11-08SHANXI ACAD OF FORESTRY & GRASSLAND SCI

Patent Information

Application Number
CN202411083452.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-11-08
Estimated Expiration
2044-08-08

AI Technical Summary

Technical Problem

The prior art is difficult to accurately calculate the optimal fertilizer application amount based on soil conditions and fruit tree characteristics, which makes it difficult to achieve the optimization of economic goals such as soapberry fruit, seeds, fruit shells, polysaccharide gels, cyst acids, and saponins.

Method used

Using a multi-objective fertilization method based on mixed effect model and optimization algorithm, the soil total nitrogen, total phosphorus, total potassium content and fruit tree economic indicators are determined, and the optimal solution for each period is calculated using nonlinear optimization iterative machine learning. , determine the fertilization standards, fertilizer quantity and ratio.

Benefits of technology

The optimal solution to the comprehensive traits of the acacia tree in each period was achieved, the yield and quality of economic indicators were improved, the soil nutrients were ensured within the optimal range, and the efficiency of multiple economic goals was improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118901366A_ABST
    Figure CN118901366A_ABST
Patent Text Reader

Abstract

The invention discloses a fertilizing method for realizing multi-economic target yield increase of Chinese honey locust, and relates to the technical field of Chinese honey locust planting, the contents of total nitrogen, total phosphorus and total potassium of soil in the crown range of each tree are respectively measured in the germination stage, the flowering stage, the fruit setting stage and the mature stage of Chinese honey locust trees, and multiple economic indexes of each tree are measured after picking; recording the ground diameter of the fruit tree; using a mixed effect polynomial regression method to construct a function relationship between the total nitrogen content, the total phosphorus content and the total potassium content of the soil in each period and each economic index, and a function relationship between the total nitrogen content, the total phosphorus content and the total potassium content of the soil; performing nonlinear optimization iterative machine learning on each yield index function of each period and the relation function of total nitrogen, total phosphorus and total potassium of soil to obtain an optimal solution of each yield index; and calculating the optimal solution of the comprehensive characters in each period according to the optimal solution, and determining the amount and proportion of the fertilizer in each period according to the contents of total nitrogen, total phosphorus and total potassium in the soil of the place to be fertilized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of honey locust planting, and more particularly to a fertilization method for achieving multiple economic target yield increase of honey locust. Background Art

[0002] Gleditsia sinensis is an important economic tree species. Its fruit has unique medicinal value and is also widely used in industry, food processing, environmental protection and other industries. In current agricultural production, how to accurately calculate the best fertilizer application amount based on soil conditions and fruit tree characteristics to achieve the optimization of multiple economic goals of pods, seeds, shells, polysaccharides, echinocystic acid and saponins is a long-standing problem that needs to be solved urgently.

[0003] Therefore, how to perform multi-objective fertilization on Gleditsia sinensis based on a mixed effect model and an optimization algorithm is a technical problem that those skilled in the art need to solve urgently. Summary of the invention

[0004] In view of this, the present invention is proposed.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A multi-objective fertilization method for Gleditsia sinensis based on a mixed effect model and an optimization algorithm, the process comprising:

[0007] (1) The total nitrogen, total phosphorus, and total potassium contents in the soil within the canopy of each tree were measured during the budding, flowering, fruiting, and ripening stages of the honey locust tree. Multiple economic indicators of each tree were measured after picking, and the ground diameter of the fruit tree was recorded. The mixed-effect polynomial regression method was used to construct the functional relationship between the total nitrogen, total phosphorus, and total potassium contents in the soil at each period and various economic indicators, as well as the functional relationship between the total nitrogen, total phosphorus, and total potassium contents in the soil;

[0008] (2) Using the optimization algorithm, nonlinear optimization iterative machine learning is performed on each yield index function in each period, as well as the relationship functions of soil total nitrogen, total phosphorus, and total potassium, to obtain the optimal solution for each yield index;

[0009] (3) Based on the optimal solution of soil total nitrogen, total phosphorus and total potassium for each economic indicator in each period, the optimal solution of the comprehensive characteristics of each period is calculated as the soil testing and fertilization standard for each period. Then, in each period, the amount and ratio of fertilizer are determined based on the total nitrogen, total phosphorus and total potassium content of the soil at the proposed fertilization site.

[0010] Preferably, in step (1), the multiple economic indicators include pod yield, seed yield, husk yield, polysaccharide yield, echinocystic acid yield, and saponin yield per plant.

[0011] Preferably, in step (1), the functional relationship between the total nitrogen, total phosphorus and total potassium content of the soil in each period and various economic indicators is constructed as follows: the total nitrogen, total phosphorus and total potassium content of each period is used as a fixed effect, the ground diameter is used as a covariate, and the individual plant is used as a random effect. The model is set as follows:

[0012] Y=β0+β1X+β2X 2 +D+b+ε

[0013] Among them, Y is the yield index; X is the total nitrogen, total phosphorus and total potassium content of the soil; β0, β1, β2 are fixed effect parameters; D is the ground diameter, b is the random effect; ε is the error term. Use maximum likelihood estimation to estimate the fixed effect parameters β0, β1, β2, and the variance of the random effect. Test the fitting effect of the model and perform a significance test of the model by calculating the goodness of fit. For models that are not successfully fitted, supplement the measurement data until all models are successfully fitted.

[0014] Preferably, in step (1), the method for constructing the functional relationship between the total nitrogen, total phosphorus and total potassium contents in the soil is as follows: a linear regression model is set, total nitrogen is fitted with the total phosphorus and total potassium of the soil in each period, and total phosphorus is fitted with the total potassium, and the functional relationship between the total nitrogen, total phosphorus and total potassium in the soil is established through three equations.

[0015] The three equations are:

[0016] Total nitrogen = β0 + β1 × total phosphorus + ε;

[0017] Total nitrogen = β0 + β1 × total potassium + ε;

[0018] Total phosphorus = β0 + β1 × total potassium + ε.

[0019] The functional relationship between total nitrogen, total phosphorus and total potassium in soil is: Y = β0 + β1X + ε, where Y represents total nitrogen or total phosphorus, and X represents total phosphorus or total potassium. β0 and β1 are parameters, and ε is the error term.

[0020] Preferably, in step (2), the specific calculation steps of nonlinear optimization iterative machine learning are as follows:

[0021] 1) Initialization: Select an initial point x0, set the number of iterations n, and the storage length m of historical information;

[0022] 2) Iteration process:

[0023] a. Calculate the gradient g of the current point k ;

[0024] b. Calculate the search direction p k , use the information of the past m steps to approximate the inverse of the Hessian matrix, thereby obtaining the search direction;

[0025] c. Determine the step size α k , so that it satisfies the following conditions:

[0026] Armijo Conditions: and

[0027] Curvature conditions:

[0028] d. Update the current point x k+1 =x k +α k p k ;

[0029] 3) Check the termination condition and the gradient reaches the preset maximum number of iterations. If the termination condition is not met, return to step 2), otherwise, output the current point as the optimal solution.

[0030] Preferably, the method for calculating the optimal solution of the comprehensive traits in each period in step (3) includes:

[0031] 1) The direct mean method was used to solve the centroid of the optimal solution of soil total nitrogen, total phosphorus, and total potassium for each economic indicator, and the optimal solution of the comprehensive trait was calculated; specifically, the optimal soil total nitrogen content of all economic indicators was composed into a data set, and the mean was calculated. The mean of total phosphorus and total potassium was calculated in the same way.

[0032] 2) Combine the three calculated means of total nitrogen, total phosphorus and total potassium into a vector, which is the center of mass.

[0033] In step (3), in each period, the process of determining the amount and ratio of fertilizers according to the total nitrogen, total phosphorus and total potassium content of the soil at the intended fertilization location is as follows: first calculate and obtain the optimal nitrogen, phosphorus and potassium fertilizer amounts for each period of the soap locust; then, in the period when fertilization is required, detect the total nitrogen, total phosphorus and total potassium content of the soil at the rhizosphere of the target tree, and compare them with the optimal nitrogen, phosphorus and potassium content of the corresponding period calculated previously. When a certain element is insufficient, apply the corresponding single fertilizer according to the difference. When a certain element is in excess, do not apply fertilizer, and adopt methods such as soil replacement and water flushing to make the content of this element in the soil fall within the calculated optimal value range. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 The attached figure shows the functional relationship between soil nitrogen, phosphorus and potassium content and pod yield per plant.

[0035] Figure 2 The attached figure shows the functional relationship between soil nitrogen, phosphorus and potassium content and single-plant fruit shell yield.

[0036] Figure 3 The attached figure shows the functional relationship between soil nitrogen, phosphorus and potassium content and seed yield per plant.

[0037] Figure 4 The attached figure shows the functional relationship between soil nitrogen, phosphorus and potassium content and polysaccharide content.

[0038] Figure 5 The attached figure shows the functional relationship between soil nitrogen, phosphorus and potassium content and echinocystic acid content.

[0039] Figure 6 The attached figure shows the functional relationship between soil nitrogen, phosphorus and potassium content and saponin content.

[0040] Figure 7 The attached figure shows the optimal value of the yield prediction of honey locust based on nonlinear optimization iterative machine learning; each point represents a calculation, the green point represents the optimal value, and the farther away from the optimal value, the darker the color, the same below. Figure a shows the calculation result of the pod yield of a single plant; Figure b shows the calculation result of the shell yield of a single plant; Figure c shows the calculation result of the seed yield of a single plant.

[0041] Figure 8 The attached figure shows the predicted values ​​of polysaccharide gum, echinocystic acid, and saponin based on nonlinear optimization iterative machine learning. a is the calculation result of polysaccharide gum; b is the calculation result of echinocystic acid; c is the calculation result of saponin.

[0042] Fig. 9 The attached figure shows the prediction of yield and quality of a single Gleditsia sinensis plant based on the optimal solution algorithm for comprehensive traits. Note: The red dots represent the optimal values, and the green dots represent the input data values. DETAILED DESCRIPTION

[0043] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] Test site overview and test materials

[0045] The test site is located in Liangmo Village, Fencheng Town, Xiangfen County, Linfen City, Shanxi Province, at 35°48′52″N, 111°11′18″E, and 530m above sea level. The test site has a temperate monsoon climate, with an average annual temperature of 12.73℃, an average temperature of 4-5℃ in January, an average temperature of 26℃ in July, an average annual rainfall of 513mm, a frost-free period of 185 days, a pH value of 8.49, and soil nitrogen, phosphorus, and potassium contents are shown in Table 1. The test material is a 4-year-old grafted seedling of a good variety of fruit honey locust "Shuaiding" (the rootstock is 8 years old, and the scion of the same variety is 4 years old). The seedlings are in good growth condition and have basically the same specifications.

[0046] Table 1 Nutrient content in the soil layer of the test site before the test (g / kg)

[0047] Nutritional elements content N 0.52±0.11 P 0.14±0.04 K 9.9±0.25

[0048] Index determination method

[0049] Example 1 Fertilizer Application Amount Prediction

[0050] (1) Determination of soil nitrogen, phosphorus and potassium content

[0051] During the budding, flowering, fruiting and ripening stages of the honey locust tree, a stainless steel soil drill was used to dig out the soil from the 0-20 cm soil layer under each tree as a sample. After digging, the soil was put into a ziplock bag and taken back to the laboratory to remove plant residues and stones and other debris in the soil. After drying, the soil was ground and sieved using a 100-mesh standard test sieve. 50 g of the sieved soil sample was taken as a sample and mixed for later use. After being digested with concentrated sulfuric acid, each nutrient element was determined. The total nitrogen was determined by the Kjeldahl method, the total phosphorus content was determined by the alkali fusion molybdenum antimony spectrophotometry, and the total potassium was determined by the flame photometer method. The results are shown in Table 2. The functional relationship between total nitrogen, total phosphorus and total potassium in the soil is shown in formulas (1)-(3).

[0052] Table 2 Survey table of soil nitrogen, phosphorus and potassium content

[0053] index average value Standard Deviation Maximum Minimum Coefficient of variation (%) N(g / kg) 3.45 1.62 6.70 0.61 0.47 P(g / kg) 0.20 0.09 0.39 0.05 0.46 K(g / kg) 5.96 2.39 10.83 1.50 0.40

[0054] N=0.57×K (1)

[0055] N = 16.39 × P (2)

[0056] P=0.03×K (3)

[0057] Pod, husk and seed yield determination

[0058] After the pods matured, the pods were collected from individual plants, brought back to the laboratory, rinsed with distilled water, and then placed in a cool place at 25±5℃ to air dry. After one month, the following indicators were measured using an electronic balance:

[0059] Single fruit weight: weight of a single fruit, g;

[0060] Pod yield per plant: sum of weights of all single fruits per plant, kg;

[0061] Shell yield per plant: Peel the pods manually, peel the soap pods, extract the shells and weigh them, kg;

[0062] Seed yield per plant: Peel the pods manually, peel the soap pods, extract the seeds and weigh them, kg.

[0063] The results are shown in Table 3.

[0064] Table 3 Single plant yield survey table

[0065] index average value Standard Deviation Maximum Minimum Coefficient of variation (%) Pod yield per plant (kg) 0.73 0.77 2.66 0.01 1.05 Shell yield per plant (kg) 0.62 0.66 2.19 0.01 1.07 Seed yield per plant (kg) 0.11 0.13 0.47 * 1.11

[0066] Determination of polysaccharide content

[0067] Remove damaged and worm-eaten seeds from the honey locust seeds, select honey locust seeds with full particles and basically the same size, crush them with a pulverizer, pass them through a 100-mesh sieve for later use, weigh 10g of honey locust seed powder for each plant using an electronic balance and put it into a 100ml beaker, then add deionized water and stir according to a solid-liquid ratio (g / mL) of 1:10, then put it into a water bath and heat it at 60°C, stir after heating, centrifuge the stirred sample on a centrifuge for 10min (5000r / min) to extract the supernatant, precipitate honey locust polysaccharide gum with 95% ethanol, the volume of ethanol is equal to the volume of the supernatant, slowly pour it into the supernatant while stirring, and white floccules are gradually separated. Put the large pieces of white flocs into the culture dish, and centrifuge the remaining white flocs that are difficult to remove with the liquid in a centrifuge with a centrifugal speed of 5000r / min for 10min, take out the sediment, spread it flat and put it into the previous culture dish, put the flattened white flocs into an electric constant temperature blast drying oven at 50℃ for drying, weigh them after drying, and calculate the glue rate. The calculation formula is as follows:

[0068] A=B / C×100

[0069] In the formula: A is the content of Gleditsia sinensis polysaccharide gum, %; B is the weight of Gleditsia sinensis polysaccharide gum after drying, g; C is the weight of weighed Gleditsia sinensis seed powder, g.

[0070] Determination of echinocystic acid content

[0071] Prepare standard curve: prepare methanol solution at 10 mg / ml, absorb 10 μL, 20 μL, 30 μL, 40 μL, 50 μL respectively in 10 ml test tubes, heat to evaporate the solvent and cool the test tubes in ice water, then add 0.5 ml of 8% vanillin ethanol solution and 5 ml of 77% sulfuric acid, shake well, heat in a 60°C constant temperature water bath for 30 minutes and then cool down, after bringing to room temperature, use a UV absorption spectrophotometer to perform full wavelength scan within the range of 454 nm to determine the highest absorption peak of each sample. Measure each tube 3 times and calculate the average value. Accurately weigh 10 mg of echinocystic acid and prepare a 2 mg / ml methanol solution. Take 300 mg of sample, add methanol, dilute to 30 ml, absorb 20 μL and place in a 10 ml test tube, and the remaining operations are the same as above. Measure the absorption value according to the standard curve determination steps.

[0072] Saponin content determination

[0073] The saponins in the soapberry were extracted by the traditional water extraction method. The pods were ground with liquid nitrogen. After grinding, 10 g of soapberry powder was weighed for each plant. After being soaked in distilled water at 25°C for 10 hours, the powder was extracted in a water bath, filtered, and the residue was filtered and the above process was repeated twice. The extracts were combined and concentrated under reduced pressure to obtain an extract. 10 times the volume of isopropanol was added. The dissolved solution was centrifuged at 4000r / min for 10 minutes, the precipitate was discarded, and the supernatant was distilled under reduced pressure to recover isopropanol. The purification was repeated 3 times to obtain a light yellow saponin viscous liquid. Vacuum freeze drying gave a light yellow solid. Weigh and calculate the content. See Table 4.

[0074] Table 4 Questionnaire on Gleditsia sinensis polysaccharide gum, echinocystic acid and saponin

[0075] index average value Standard Deviation Maximum Minimum Coefficient of variation (%) Polysaccharide gum content (%) 25.49 8.25 39.16 8.17 0.32 Echinocystic acid content (%) 0.40 0.13 0.81 0.16 0.33 Saponin content (%) 8.28 2.70 16.67 3.38 0.33

[0076] Measure the ground diameter at the same time:

[0077] The average ground diameter of the plants was measured to be 6.86±1.63cm.

[0078] According to the above calculation steps, the optimal soil total nitrogen, total phosphorus and total potassium content of the honey locust tree in the budding period, flowering period, fruiting period and maturity period were calculated. The total nitrogen, total phosphorus and total potassium content of each period were used as fixed effects, the ground diameter was used as a covariate, and the individual plant was used as a random effect. The model was set as: Y = β0 + β1X + β2X 2 +D+b+ε;

[0079] Among them, Y is the yield index; X is the total nitrogen, total phosphorus and total potassium content of the soil; β0, β1, β2 are fixed effect parameters; D is the ground diameter, b is the random effect; ε is the error term. Use maximum likelihood estimation to estimate the fixed effect parameters β0, β1, β2, and the variance of the random effect. Test the fitting effect of the model, calculate the goodness of fit, and perform a significance test of the model. For models that are not successfully fitted, supplement the measurement data until all models are successfully fitted to the soil.

[0080] Then a linear regression model was set up to fit total nitrogen with total phosphorus and total potassium in the soil in each period, and then fit total phosphorus with total potassium. The functional relationship between total nitrogen, total phosphorus and total potassium in the soil was established through three equations:

[0081] Total nitrogen = β0 + β1 × total phosphorus + ε;

[0082] Total nitrogen = β0 + β1 × total potassium + ε;

[0083] Total phosphorus = β0 + β1 × total potassium + ε.

[0084] The results are as follows: The functional relationship between nitrogen, phosphorus and potassium content and pod yield per plant is shown in Figure 1 The functional relationship between soil nitrogen, phosphorus and potassium content and single-plant husk yield is shown in Figure 2The functional relationship between soil nitrogen, phosphorus and potassium content and seed yield per plant is shown in Figure 3 The functional relationship between soil nitrogen, phosphorus and potassium content and echinocystic acid content is shown in Figure 4 The functional relationship between soil nitrogen, phosphorus and potassium content and echinocystic acid content is shown in Figure 5 The functional relationship between soil nitrogen, phosphorus and potassium content and saponin content is shown in Figure 6 .

[0085] (2) Calculation of the optimal nitrogen, phosphorus and potassium content of multiple economic traits of Gleditsia sinensis

[0086] The optimal value of honey locust yield prediction based on nonlinear optimization iterative machine learning is as follows:

[0087] 1) Initialization: Select an initial point x0, set the number of iterations n, and the storage length m of historical information;

[0088] 2) Iteration process:

[0089] a. Calculate the gradient g of the current point k ;

[0090] b. Calculate the search direction p k , use the information of the past m steps to approximate the inverse of the Hessian matrix, thereby obtaining the search direction;

[0091] c. Determine the step size α k , so that it satisfies the following conditions:

[0092] Armijo Conditions: and

[0093] Curvature conditions:

[0094] d. Update the current point x k+1 =x k +α k p k ;

[0095] 3) Check the termination condition and the gradient reaches the preset maximum number of iterations. If the termination condition is not met, return to step 2), otherwise, output the current point as the optimal solution.

[0096] Results Figure 7 and Figure 8 .

[0097] The prediction results show that when the soil nitrogen, phosphorus and potassium contents reach 4.11g / kg, 0.25g / kg and 7.16g / kg respectively, the yield of a single honey locust plant can reach 5.86kg; when the soil nitrogen, phosphorus and potassium contents reach 4.03g / kg, 0.24g / kg and 6.95g / kg, the shell yield of a single honey locust plant can reach 5.04kg; when the soil nitrogen, phosphorus and potassium contents reach 3.97g / kg, 0.24g / kg and 6.91g / kg, the seed yield of a single honey locust plant can reach 1. Reach 1.07kg; when the soil nitrogen, phosphorus and potassium content reaches 3.77g / kg, 0.22g / kg and 6.49g / kg respectively, the polysaccharide gum yield can reach 45%; when the soil nitrogen, phosphorus and potassium content reaches 3.72g / kg, 0.23g / kg and 6.2g / kg respectively, the saponin content can reach 33.61%; when the soil nitrogen, phosphorus and potassium content reaches 3.74g / kg, 0.23g / kg and 6.49g / kg respectively, the echinocystic acid content can reach 1.56%.

[0098] (3) According to the optimal solution of soil total nitrogen, total phosphorus and total potassium for each economic indicator in each period, the optimal solution of the comprehensive traits in each period was calculated as the soil testing and fertilization standard for each period. The centroid of the optimal soil total nitrogen, total phosphorus and total potassium contents of the above six traits was solved. It was shown that when the soil nitrogen, phosphorus and potassium contents reached 3.74 g / kg, 0.22 g / kg and 6.51 g / kg respectively, the yield and quality of soapberry could reach a better level at the same time ( Fig. 9 ). Then, in each period, determine the amount and ratio of fertilizer according to the total nitrogen, total phosphorus and total potassium content of the soil at the intended fertilization location, see Table 2.

[0099] Fertilizers were applied according to the calculated optimal soil total nitrogen, total phosphorus, and total potassium contents in the test plots during the budding, flowering, fruiting, and maturity periods. At the corresponding time point in the second year, soil nutrient contents were measured. For areas where soil total nitrogen, total phosphorus, and total potassium contents were lower than the calculated values, fertilization was increased to bring them close to the calculated soil total nitrogen, total phosphorus, and total potassium contents. For areas where soil total nitrogen, total phosphorus, and total potassium contents were higher than the calculated values, irrigation and drainage were increased to help dissolve and dilute excess nutrients in the soil to bring them close to the calculated soil total nitrogen, total phosphorus, and total potassium contents.

[0100] Table 2

[0101]

[0102]

[0103] Example 2 Effect verification of the method of Example 1

[0104] The target area was fertilized according to the method of Example 1. After one year of treatment, many economic indicators of Chinese honey locust in the study area were greatly improved, as shown in Table 2 for details.

[0105] Table 2 Increase in economic indicators of Gleditsia sinensis

[0106]

[0107] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0108] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis, characterized in that the process include: (1) The total nitrogen, total phosphorus, and total potassium contents in the soil within the canopy of each tree were measured during the budding, flowering, fruiting, and ripening stages of the honey locust tree. Multiple economic indicators of each tree were measured after picking, and the ground diameter of the fruit tree was recorded. The mixed-effect polynomial regression method was used to construct the functional relationship between the total nitrogen, total phosphorus, and total potassium contents in the soil at each period and various economic indicators, as well as the functional relationship between the total nitrogen, total phosphorus, and total potassium contents in the soil; (2) Using the optimization algorithm, nonlinear optimization iterative machine learning is performed on each yield index function in each period, as well as the relationship functions of soil total nitrogen, total phosphorus, and total potassium, to obtain the optimal solution for each yield index; (3) Based on the optimal solution of soil total nitrogen, total phosphorus and total potassium for each economic indicator in each period, the optimal solution of the comprehensive characteristics of each period is calculated as the soil testing and fertilization standard for each period. Then, in each period, the amount and ratio of fertilizer are determined based on the total nitrogen, total phosphorus and total potassium content of the soil at the proposed fertilization site.

2. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 1, characterized in that: In step (1), the multiple economic indicators include pod yield, seed yield, shell yield, polysaccharide yield, echinocystic acid yield, and saponin yield of a single plant.

3. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 2, characterized in that, In step (1), the functional relationship between the total nitrogen, total phosphorus, and total potassium content of the soil in each period and various economic indicators is constructed as follows: the total nitrogen, total phosphorus, and total potassium content of each period are used as fixed effects, the ground diameter is used as a covariate, and the individual plant is used as a random effect. The model is set as follows: Y=β0+β1X+β2X 2 +D+b+e Among them, Y is the yield index; X is the total nitrogen, total phosphorus and total potassium content of the soil; β0, β1, β2 are fixed effect parameters; D is the ground diameter, b is the random effect; ε is the error term. Use maximum likelihood estimation to estimate the fixed effect parameters β0, β1, β2, and the variance of the random effect. Test the fitting effect of the model and perform a significance test of the model by calculating the goodness of fit. For models that are not successfully fitted, supplement the measurement data until all models are successfully fitted.

4. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 3, characterized in that: In step (1), the method for constructing the functional relationship between the total nitrogen, total phosphorus and total potassium contents in the soil is as follows: a linear regression model is set up, total nitrogen is fitted using the total phosphorus and total potassium in the soil in each period, and total phosphorus is fitted using the total potassium, and the functional relationship between the total nitrogen, total phosphorus and total potassium in the soil is established through three equations.

5. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 4, characterized in that, The three equations are: Total nitrogen = β0 + β1 × total phosphorus + ε; Total nitrogen = β0 + β1 × total potassium + ε; Total phosphorus = β0 + β1 × total potassium + ε.

6. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 5, characterized in that: The functional relationship between total nitrogen, total phosphorus and total potassium in soil is: Y = β0 + β1X + ε, where Y represents total nitrogen or total phosphorus, and X represents total phosphorus or total potassium. β0 and β1 are parameters, and ε is the error term.

7. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 6, characterized in that: In step (2), the specific calculation steps of nonlinear optimization iterative machine learning are as follows: 1) Initialization: Select an initial point x0, set the number of iterations n, and the storage length m of historical information; 2) Iteration process: a. Calculate the gradient g of the current point k ; b. Calculate the search direction p k , use the information of the past m steps to approximate the inverse of the Hessian matrix, thereby obtaining the search direction; c. Determine the step size α k , so that it satisfies the following conditions: Armijo Conditions: and Curvature conditions: d. Update the current point x k+1 =x k +α k p k ; 3) Check the termination condition and the gradient reaches the preset maximum number of iterations. If the termination condition is not met, return to step 2), otherwise, output the current point as the optimal solution.

8. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 7, characterized in that: The calculation method of the optimal solution of the comprehensive traits in each period in step (3) includes: 1) The direct mean method was used to solve the centroid of the optimal solution of soil total nitrogen, total phosphorus, and total potassium for each economic indicator, and the optimal solution of the comprehensive trait was calculated; specifically, the optimal soil total nitrogen content of all economic indicators was composed into a data set, and the mean was calculated. The mean of total phosphorus and total potassium was calculated in the same way. 2) Combine the calculated three means of total nitrogen, total phosphorus and total potassium into a vector. This vector is the centroid, that is, the optimal solution for the comprehensive traits.

9. A fertilization method for achieving multiple economic target yield increase of Gleditsia sinensis according to claim 8, characterized in that: In step (3), in each period, the process of determining the amount and ratio of fertilizer according to the total nitrogen, total phosphorus and total potassium content of the soil at the intended fertilization location is as follows: calculate the optimal solution of the comprehensive shape; then, in the period when fertilization is required, detect the total nitrogen, total phosphorus and total potassium content of the soil in the rhizosphere of the target tree, and compare them with the optimal solution of the corresponding period calculated previously. When a certain element is insufficient, apply the corresponding single fertilizer according to the difference. When a certain element is excessive, do not apply fertilizer, and adopt methods such as soil replacement and water flushing to make the content of this element in the soil fall within the calculated optimal value range.

Citation Information

Patent Citations

  • Application method of nitrogen-phosphorus-potassium chemical fertilizer in full bearing period of apple trees

    CN104350855A

  • Recommended fertilizer applying method for purple paddy soil based on yield response and agronomic efficiency

    CN109661885A

  • Recommended fertilizer application method for radish

    CN110073789A

  • Field scale jujube garden variable grid fertilization diagram manufacturing method applied to agricultural machinery operation

    CN114521376A

Cited By

  • Efficient irrigation and fertilization decision support method and system for protected tomatoes

    CN120500958A

  • Method, system, and equipment for generating the optimal fertilizer blend ratio for agricultural crops.

    JP7875648B1