CN-122020142-A - Method for calculating black soil structural obstacle factor and threshold value of rhizoma zingiberis

CN122020142ACN 122020142 ACN122020142 ACN 122020142ACN-122020142-A

Abstract

The invention relates to the technical field of soil quality evaluation and improvement, in particular to a method for calculating a black soil structural obstacle factor and a threshold value of a sand ginger, which comprises the steps of firstly screening 22 structural obstacle factors covering physical, chemical and biological properties of soil, constructing combined weights by adopting an entropy weight method, a principal component analysis method and an average weight method, calculating membership degrees of all factors by combining membership functions of a ring-up type, a peak type and an S type, obtaining a comprehensive index of the soil structural obstacle by a cumulative method, dividing obstacle grades, taking a gradient lifting regression tree as a core, combining T test and random forest verification, determining all factor thresholds, and finally constructing a minimum data set based on principal component analysis and clustering analysis, thereby solving the problems of strong subjectivity of obstacle factor screening, single threshold value calculation method and difficulty in popularization in the conventional method.

Inventors

GUO ZHIBIN
LI YONGXIN
WANG WENJUN
Zhan Linchuan
HUA KEKE
WANG DAOZHONG
JIN LIN
HE CHUANLONG

Assignees

安徽省农业科学院土壤肥料研究所

Dates

Publication Date: 20260512
Application Date: 20251205

Claims (10)

1. The method for calculating the black soil structural obstacle factor and the threshold value of the rhizoma zingiberis is characterized by comprising the following steps of: S1, screening out black soil structural obstacle factors of the sand ginger, wherein the obstacle factors comprise quick-acting nitrogen, quick-acting potassium, quick-acting phosphorus, organic carbon, total nitrogen, total phosphorus, total potassium, pH, soil volume weight, microbial carbon, microbial nitrogen, large agglomerates, medium agglomerates, small agglomerates, soil water content, sucrase, urease, sticky particles, powder particles, sand grains, soil penetration resistance and plough layer thickness; S2, constructing a combination weight, and fusing an entropy weight method, a principal component analysis method and an average weight method according to a ratio of 1:1:1 to obtain a combination weight Ci of each obstacle factor; S3, determining membership functions of each obstacle factor, wherein the membership functions comprise a ring-up type, a peak type and an S type, and the membership functions correspond to the obstacle factors with different influence rules respectively; s4, calculating the comprehensive index P of the black soil structural disorder of the rhizoma zingiberis by adopting an accumulation method, wherein the formula is Wherein Ci is the combination weight of an entropy weight method, a principal component analysis method and an average weight method of the ith evaluation index, and Fi is the membership of the ith index; s5, grading the disorder according to the comprehensive index P, wherein the disorder comprises mild disorder, moderate disorder, severe disorder and extremely severe disorder; S6, calculating threshold values of various obstacle factors by taking a gradient lifting regression tree as a core and combining T test and random forest verification; S7, constructing a minimum data set for evaluating the black soil structural disorder of the rhizoma zingiberis based on principal component analysis and cluster analysis.
2. The method for calculating the structural obstacle factor and threshold of the black soil of rhizoma zingiberis, as set forth in claim 1, wherein the large agglomerates in S1 are agglomerates with a particle size of > 2mm, the medium agglomerates are agglomerates with a particle size of 0.25-2mm, the small agglomerates are agglomerates with a particle size of 0.05-0.25mm, the sticky particles are particles with a particle size of < 0.002mm, the powder particles are particles with a particle size of 0.002-0.05mm, and the sand particles are particles with a particle size of 0.05-2 mm.
3. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis, as set forth in claim 1, wherein in S2 The calculation process of the entropy weight method comprises data standardization processing, factor characteristic proportion calculation, information entropy calculation, difference coefficient calculation and objective weight determination; The calculation process of the principal component analysis algorithm comprises data standardization processing, covariance matrix construction, eigenvalue and eigenvector solving, variance contribution rate calculation and objective weight determination.
4. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis recens according to claim 1, wherein in the step S3 The upper model function is applicable to quick-acting nitrogen, quick-acting potassium and quick-acting phosphorus; the peak function applies to organic carbon, total nitrogen, total phosphorus, total potassium, pH, soil bulk weight, microbial carbon, microbial nitrogen, large agglomerates, medium agglomerates, small agglomerates, soil moisture, sucrase, urease, mucilage, powder, grit, and tilled layer thickness; The sigmoid function is suitable for soil penetration resistance.
5. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis, as set forth in claim 1, wherein the value range of the comprehensive index P in the S4 is 0-1, and the greater the value of P, the more serious the structural obstacle of the black soil of the rhizoma zingiberis.
6. The method for calculating a structural obstacle factor and threshold of black soil of rhizoma Zingiberis recens as set forth in claim 1, wherein the dividing criteria of obstacle level in S5 is A slight disorder of 0.2440 < P <0.3508; Light disorder, 0.3508 is less than or equal to P <0.3768; medium disorder, 0.3768 is less than or equal to P <0.4148; The serious disorder is 0.4148 to be less than or equal to P <0.4560; extremely severe disorder is 0.4560 to be less than or equal to P.
7. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis recens according to claim 1, wherein in the step S6 The calculation process of the gradient lifting regression tree comprises the steps of data preprocessing, residual fitting model construction, iterative training of a weak learner and factor threshold interval determination; the T test is used for verifying the significance of the threshold value, and the random forest is used for verifying the stability of the threshold value.
8. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis recens according to claim 1, wherein in the step S7 Principal component analysis selects principal components with characteristic values larger than 1 and accumulated variance contribution rate larger than 78.78%; the clustering analysis adopts K-means clustering to divide factors into 4 major classes, and each class selects 1 representative factor; The minimum data set includes 8-10 factors and has a correlation with the overall factor evaluation result of greater than 0.545.
9. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis recens according to claim 1, wherein the method is characterized in that The method comprises the steps that obstacle factors are acquired through soil sample acquisition and testing, sampling points are distributed and combined with soil census sampling points and supplementary sampling points, and sampling depth covers a cultivation layer, a plough layer and a core soil layer; the test method comprises a ring knife method, a wet screening method, a potential method, a potassium dichromate oxidation-external heating method and a chloroform fumigation-leaching method.
10. The method for calculating the structural obstacle factor and the threshold value of the black soil of the rhizoma zingiberis, according to claim 8, wherein the factors of the minimum data set comprise plough layer thickness, pH, clay content, soil penetration resistance, 0.25-2mm aggregate, soil volume weight, field water holding capacity and organic carbon, and the correlation with the total factor evaluation result reaches 0.632.

Description

Method for calculating black soil structural obstacle factor and threshold value of rhizoma zingiberis Technical Field The invention relates to the technical field of soil quality evaluation and improvement, in particular to a method for calculating a black soil structural obstacle factor and a threshold value of rhizoma zingiberis. Background The problems of drought, waterlogging, stain, thin and stiff are the main low-yield reasons for restricting the black soil of the sand ginger, and the problems of drought, waterlogging and stain of the black soil of the sand ginger are effectively controlled through facilities such as water conservancy, ditches, roads, bridges, culverts and the like for a long time, the problems of thin of the black soil of the sand ginger are effectively alleviated by long-term straw returning and scientific fertilization, but the problems of stiff still exist, and the characteristics of caking during dry, muddy during wet, short suitable tillage period, large tillage resistance, poor tillage quality and small aggregate content severely restrict the local agricultural and economic development. In order to pertinently improve the black soil of the rhizoma zingiberis, obstacle evaluation and threshold calculation are required. At present, the existing method for evaluating and calculating the black soil obstacle of the rhizoma zingiberis mainly has the following limitations: 1. In the aspect of screening barrier factors, the barrier factors are focused on single physical or chemical properties of the soil, and biological indexes are not integrated, so that the integrity of the soil structure cannot be comprehensively reflected; 2. The traditional method depends on expert experience empowerment, so that objective information of data is easy to ignore, and evaluation result deviation is caused; 3. when the threshold value is calculated, a simple statistical method is mostly adopted, the nonlinear fitting capability of a machine learning model is not combined, and the characteristic of strong spatial heterogeneity of the ginger black soil is difficult to adapt; 4. The traditional calculation method does not construct a simplified minimum data set, so that a large number of indexes need to be measured, the operation cost is high, and the method is difficult to popularize in agricultural production. Therefore, a systematic, objective and high-precision method for screening and calculating the structural obstacle factor of the black soil of the rhizoma zingiberis is needed to meet the actual requirement of the accurate improvement of the black soil of the rhizoma zingiberis. Disclosure of Invention The invention aims to provide a method for calculating a black soil structural obstacle factor and a threshold value of a sand ginger, which solves the problems of strong subjectivity, single threshold value calculation method and difficulty in popularization of obstacle factor screening and weight determination in the prior art. The invention solves the technical problems by adopting the following technical scheme that the method for calculating the black soil structural obstacle factor and the threshold value of the rhizoma zingiberis comprises the following steps: S1, screening out black soil structural obstacle factors of the sand ginger, wherein the obstacle factors comprise quick-acting nitrogen, quick-acting potassium, quick-acting phosphorus, organic carbon, total nitrogen, total phosphorus, total potassium, pH, soil volume weight, microbial carbon, microbial nitrogen, large agglomerates, medium agglomerates, small agglomerates, soil water content, sucrase, urease, sticky particles, powder particles, sand grains, soil penetration resistance and plough layer thickness; s2, constructing a combination weight, and fusing an entropy weight method, a principal component analysis method and an average weight method according to the proportion of 1:1:1 to obtain a combination weight Ci of each obstacle factor; S3, determining membership functions of each obstacle factor, wherein the membership functions comprise a ring-up type, a peak type and an S type, and the membership functions correspond to the obstacle factors with different influence rules respectively; s4, calculating the comprehensive index P of the black soil structural disorder of the rhizoma zingiberis by adopting an accumulation method, wherein the formula is Wherein Ci is the combination weight of an entropy weight method, a principal component analysis method and an average weight method of the ith evaluation index, and Fi is the membership of the ith index; s5, grading the disorder according to the comprehensive index P, wherein the disorder comprises mild disorder, moderate disorder, severe disorder and extremely severe disorder; S6, calculating threshold values of various obstacle factors by taking a gradient lifting regression tree as a core and combining T test and random forest verification; S7, constructing a minimum data se