CN-121980820-A - Fitting analysis method for matching nutrient release of potassium magnesium sulfate fertilizer with crop demands

CN121980820ACN 121980820 ACN121980820 ACN 121980820ACN-121980820-A

Abstract

The invention discloses a fitting analysis method for matching nutrient release and crop demands of a potassium magnesium sulfate fertilizer, which relates to the technical field of intelligent agriculture and comprises the following steps of S1 collecting multi-condition nutrient release of the fertilizer, nutrient demand data of the whole growth period of the crop and standardized pretreatment, S2 constructing a fertilizer nutrient release dynamic curve by a nonlinear fitting method, S3 constructing a crop nutrient demand dynamic curve by piecewise fitting, S4 constructing a dynamic fitting analysis model of an integrated multi-module, S5 calculating curve matching degree and identifying time offset and intensity difference, S6 establishing an evaluation system rating and generating a fertilization optimization suggestion, and S7 combining with crop actual data verification and iterative optimization model. The invention realizes full-period accurate matching analysis of fertilizer nutrient release and crop demands, has comprehensive data acquisition, accurate curve fitting and refined analysis results, can quantitatively evaluate suitability and generate actual fertilization suggestions, can continuously optimize a model, and provides scientific data support for accurate fertilization.

Inventors

HAN CHUNXIAO
LI JIAHANG
HUANG LILI
JIANG YAN
LI SHOUJIANG
HOU JIANHUA
DENG XIAOYONG
ZHANG YIPENG
HE YONGFENG
Yuan Kaize
LIU JINGSEN
WANG JIALONG

Assignees

国投(四川)农业科技有限责任公司
国投新疆罗布泊钾盐有限责任公司

Dates

Publication Date: 20260505
Application Date: 20260403

Claims (10)

1. A fitting analysis method for matching nutrient release of a potassium magnesium sulfate fertilizer with crop demands is characterized by comprising the following steps: s1, multi-dimensional data acquisition and pretreatment, namely collecting nutrient release data of a potassium magnesium sulfate fertilizer under different application conditions, synchronously acquiring nutrient demand data of a target crop in the whole growth period, and carrying out outlier rejection, missing value complementation and standardization treatment on the acquired data; S2, constructing a nutrient release curve, constructing a potassium magnesium sulfate fertilizer nutrient release dynamic curve by adopting a nonlinear fitting algorithm based on the preprocessed nutrient release data, optimizing curve parameters by a least square method, and outputting a release curve equation and characteristic parameters of each element; S3, constructing a crop demand curve, constructing a full-growth-period nutrient demand dynamic curve according to nutrient demand data of target crops and combining with the growth period duration of the crops, adopting a piecewise fitting mode, respectively fitting curve segments aiming at demand characteristics of different growth stages, and outputting a demand curve equation and characteristic parameters of each element; S4, constructing a dynamic fitting analysis model, namely constructing the dynamic fitting analysis model by taking a nutrient release curve and a crop demand curve as core inputs, and integrating three functional modules of time alignment, strength matching and deviation analysis; s5, matching degree calculation and difference identification, namely calculating the matching degree of a nutrient release curve and a crop demand curve through a dynamic fitting analysis model; S6, suitability evaluation and optimization suggestion generation, namely establishing a suitability evaluation index system according to a matching degree result, a time offset and an intensity difference value, and classifying suitability into four grades of excellent, good, general and poor according to a grading result; And S7, model verification and iterative optimization, namely collecting crop growth index data and yield data in an actual application scene, taking the crop growth index data and the yield data as verification indexes, and based on a verification result, adopting a gradient descent algorithm to iteratively optimize model parameters, and updating a curve fitting algorithm and matching degree calculation logic.
2. The fitting analysis method for matching the nutrient release of the potash magnesium sulphate fertilizer with the crop demands according to claim 1, further comprising an overall matching degree accurate calculation step, wherein the step is executed in the step S5, and the overall matching degree of the release and the demands is quantified through a formula, wherein the specific formula is as follows: for the degree of overall matching, As a node of the time it is, As the number of total time nodes to be counted, Is that The nutrient release amount at the moment, Is that The demand of crops at any time, Is that Time weighting coefficients.
3. The fitting analysis method for matching the nutrient release of the potash magnesium sulphate fertilizer with the crop demands according to claim 1, further comprising an application condition sensitivity analysis step, wherein the application condition sensitivity analysis step is executed after S1 and before S2, the influence degree of each application condition on a nutrient release curve is analyzed through a controlled variable method, gradient change values of soil type, soil humidity, soil temperature, fertilization amount and fertilization depth are set, the change rate of characteristic parameters of the nutrient release curve when single conditions are changed is calculated respectively, and key influence factors are determined according to the change rate.
4. The fitting analysis method for matching the nutrient release of the potassium magnesium sulfate fertilizer with the crop demands is characterized in that in S1 multidimensional data acquisition and pretreatment, nutrient release data are monitored in real time through a soil nutrient sensor, monitoring periods cover the whole period from the time when the potassium magnesium sulfate fertilizer is applied to the time when the crops are mature, crop demand data are obtained through a mode of combining field tests with literature investigation, outlier rejection adopts a Graibus criterion, missing value complementation adopts a linear interpolation method, and standardized treatment adopts a Z-score standardized method.
5. The fitting analysis method for matching the nutrient release of the potassium magnesium sulfate with the crop demands is characterized in that in the construction of an S2 nutrient release curve, a Logistic growth curve algorithm is selected as a nonlinear fitting algorithm, a curve equation is in an S-shaped curve form, and in the process of curve parameter optimization, characteristic parameter extraction is obtained by deriving and calculating curve inflection points and extreme points.
6. The fitting analysis method for matching the nutrient release of the potassium magnesium sulfate with the crop demands is characterized in that in S3 crop demand curve construction, an adaptive fitting algorithm is selected for different growth stages in a segmented fitting mode, a linear fitting algorithm is selected for a seedling stage and a mature stage, a quadratic polynomial fitting algorithm is selected for a jointing stage and a flowering stage, an exponential fitting algorithm is selected for a grouting stage, and characteristic parameters are obtained through curve integration and extremum solving through smooth transition treatment.
7. The fitting analysis method for matching the nutrient release of the potassium magnesium sulfate fertilizer with the crop demands is characterized in that in S4 dynamic fitting analysis model construction, a time alignment module adopts a dynamic time warping algorithm to finish time sequence synchronization of two curves through a stretching or compressing time axis, an intensity matching module adopts a mode of combining correlation analysis and amplitude proportion analysis to calculate a pearson correlation coefficient and an amplitude proportion coefficient between the curves, and a deviation analysis module calculates a time offset and an intensity difference value time by time interval through a sliding window method.
8. The fitting analysis method for matching the nutrient release of the potash magnesium sulphate fertilizer with the crop demand according to claim 1, wherein in the S5 matching degree calculation and the difference identification, the local matching degree calculation is respectively carried out for each growth stage, the matching degree score of each element in each stage is output, a positive value indicates that the nutrient release peak is later than the demand peak, a negative value indicates that the nutrient release peak is earlier than the demand peak, the intensity difference value calculation comprises an absolute difference value and a relative difference value, the absolute difference value is the difference between the release amount and the demand amount, and the relative difference value is the ratio of the absolute difference value and the demand amount.
9. The fitting analysis method for matching the nutrient release of the potash magnesium sulphate fertilizer with the crop demands is characterized in that in the generation of S6 suitability evaluation and optimization suggestions, a suitability evaluation index system adopts a analytic hierarchy process to determine each index weight, when the optimization suggestions are generated, the fertilization time is adjusted in advance or quick release type potash magnesium sulphate fertilizer is selected when the time offset is positive for the situation that the absolute value of the time offset is more than or equal to 3 days, the fertilization time is adjusted in delay or slow release type potash magnesium sulphate fertilizer is selected when the time offset is negative, the fertilization amount is adjusted for the situation that the absolute value of the relative difference value of the intensity is more than or equal to 20%, and the fertilization depth or soil improvement measure is optimized for the situation that the soil condition is sensitive.
10. The fitting analysis method for matching the nutrient release of the potash magnesium sulphate fertilizer with the crop demands is characterized in that in S7 model verification and iterative optimization, a comparison test method is adopted for verification index evaluation, an optimization group and a control group are set, the optimization group adopts an application strategy generated by a model, the control group adopts a conventional application strategy, the model effect is verified by comparing differences between two groups of crop growth indexes and yield data, and when the model is subjected to iterative optimization, parameter updating is carried out once every 10 groups of new test data are collected.

Description

Fitting analysis method for matching nutrient release of potassium magnesium sulfate fertilizer with crop demands Technical Field The invention relates to the technical field of intelligent agriculture, in particular to a fitting analysis method for matching nutrient release of a potassium magnesium sulfate fertilizer with crop demands. Background The potassium magnesium sulfate fertilizer is used as a high-efficiency multi-element composite potassium fertilizer, has the essential nutrient elements of three crops of potassium, magnesium and sulfur, has mild fertilizer efficiency and balanced nutrition, can supplement nutrients required by the growth of crops, promote photosynthetic efficiency and yield quality, improve the physicochemical property of soil and relieve the problems of lack of nutrients and hardening of soil, has wider application in large-scale planting of grains, fruits and vegetables and economic crops, and is one of core fertilizer seeds of a modern agriculture precise fertilization system. The method realizes the accurate matching of the nutrient release rhythm of the potassium magnesium sulfate fertilizer and the nutrient demand rhythm of the crops in the whole growth period, is a key for maximally exerting the fertilizer efficiency, improving the fertilizer utilization rate and reducing the agricultural production cost, is an important research direction in the fields of digital agriculture and intelligent planting, and is an important measure for promoting the green development of agriculture and reducing the non-point source pollution. In the current agricultural production, the application of the potassium magnesium sulfate fertilizer still depends on the experience judgment of a planter, and the common problems of the fertilizer application amount, the fertilizer application time, the fertilizer application mode and the dislocation of the actual nutrient requirements of crops exist. In some scenes, the fertilizer nutrients are released too quickly, which is easy to cause seedling burning and nutrient leaching loss of crops in the seedling stage, while in other scenes, the nutrients are released too slowly, which can cause insufficient nutrient supply in the key growth stage of crops such as flowering, grouting and the like, and directly affect the growth and yield formation of the crops. The empirical fertilization mode not only greatly reduces the utilization efficiency of the potassium magnesium sulfate fertilizer and increases the planting cost, but also is easy to cause ecological problems of soil physical and chemical property deterioration, water eutrophication and the like due to nutrient loss, and is contrary to the development requirements of modern agriculture precision and greenization. The existing analysis method aiming at matching of fertilizer nutrients and crop demands has been introduced into data fitting and simple algorithm analysis initially, but a plurality of technical short plates still exist. Most methods focus on static matching analysis of single nutrient or single growth stage of crops, do not fully capture dynamic rules of potassium magnesium fertilizer potassium, magnesium and sulfur multielement synergistic release, do not combine staged and differential characteristics of nutrient requirements of crops in whole growth period, and are difficult to realize double accurate matching of time dimension and intensity dimension. Meanwhile, the influence of the conventional method on key application conditions such as soil type, soil temperature and humidity, fertilization depth and the like is not considered enough, the scene suitability and generalization capability of the fitting analysis model are weak, a perfect suitability evaluation index system is lacking, the obtained analysis result is mostly a qualitative conclusion, and the method is difficult to be directly converted into a fertilization optimization strategy which can be operated in the field. Along with the acceleration of agricultural digital transformation, the precise fertilization technology based on big data and algorithm models becomes a development trend, a dynamic fitting analysis model is constructed by utilizing a digital means, and full-period and multi-dimensional matching analysis of the nutrient release of the potassium magnesium sulfate fertilizer and the crop demands is realized, so that the method has become an urgent demand for solving the defects of the traditional fertilization. However, the related technology still has the problems of single data acquisition dimension, insufficient curve fitting precision, lack of comprehensiveness in matching degree calculation, lack of model verification and iteration mechanisms and the like, and restricts the practical floor application of the system in agricultural production. Therefore, the research and development of the analysis method capable of realizing multidimensional data fusion, dynamic curve fitting, accurate matching degree ca