CN117408829A - Method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics - Google Patents
Method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics Download PDFInfo
- Publication number
- CN117408829A CN117408829A CN202311408679.3A CN202311408679A CN117408829A CN 117408829 A CN117408829 A CN 117408829A CN 202311408679 A CN202311408679 A CN 202311408679A CN 117408829 A CN117408829 A CN 117408829A
- Authority
- CN
- China
- Prior art keywords
- data
- sample
- index
- value
- partition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005192 partition Methods 0.000 title claims abstract description 103
- 238000000034 method Methods 0.000 title claims abstract description 64
- 230000004888 barrier function Effects 0.000 title abstract description 10
- 230000001939 inductive effect Effects 0.000 title abstract description 4
- 239000002689 soil Substances 0.000 claims abstract description 67
- 238000005070 sampling Methods 0.000 claims abstract description 55
- 238000013441 quality evaluation Methods 0.000 claims abstract description 49
- 238000000638 solvent extraction Methods 0.000 claims abstract description 17
- 238000011160 research Methods 0.000 claims abstract description 15
- 238000003745 diagnosis Methods 0.000 claims description 17
- 230000008569 process Effects 0.000 claims description 15
- 238000003971 tillage Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000009826 distribution Methods 0.000 claims description 9
- 238000009440 infrastructure construction Methods 0.000 claims description 9
- 238000004519 manufacturing process Methods 0.000 claims description 9
- 239000003337 fertilizer Substances 0.000 claims description 8
- 230000006698 induction Effects 0.000 claims description 8
- 238000003973 irrigation Methods 0.000 claims description 8
- 230000002262 irrigation Effects 0.000 claims description 8
- 239000000575 pesticide Substances 0.000 claims description 8
- 239000000126 substance Substances 0.000 claims description 7
- 238000012544 monitoring process Methods 0.000 claims description 6
- 230000006855 networking Effects 0.000 claims description 6
- 238000009933 burial Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 4
- 230000003749 cleanliness Effects 0.000 claims description 4
- 238000007621 cluster analysis Methods 0.000 claims description 4
- 238000011835 investigation Methods 0.000 claims description 4
- 239000004016 soil organic matter Substances 0.000 claims description 4
- 238000012876 topography Methods 0.000 claims description 4
- 239000011573 trace mineral Substances 0.000 claims description 4
- 235000013619 trace mineral Nutrition 0.000 claims description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 4
- 230000001351 cycling effect Effects 0.000 claims description 3
- 239000006185 dispersion Substances 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000011282 treatment Methods 0.000 claims description 3
- 230000006872 improvement Effects 0.000 description 6
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 5
- ZLMJMSJWJFRBEC-UHFFFAOYSA-N Potassium Chemical compound [K] ZLMJMSJWJFRBEC-UHFFFAOYSA-N 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 5
- 239000005416 organic matter Substances 0.000 description 5
- 239000011574 phosphorus Substances 0.000 description 5
- 229910052698 phosphorus Inorganic materials 0.000 description 5
- 239000011591 potassium Substances 0.000 description 5
- 229910052700 potassium Inorganic materials 0.000 description 5
- 239000000463 material Substances 0.000 description 4
- 230000033228 biological regulation Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000001149 cognitive effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000009472 formulation Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000013316 zoning Methods 0.000 description 2
- 241000282414 Homo sapiens Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06395—Quality analysis or management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Bioinformatics & Computational Biology (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- General Health & Medical Sciences (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Health & Medical Sciences (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
A method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics belongs to the technical field of agricultural engineering. In order to solve the problems that the existing farmland quality partitioning scheme is divided according to administrative division and local space autocorrelation methods, the partitioning cannot be carried out according to objective conditions of cultivated lands, so that the land in the same partition may have large similarity difference, the characteristics and barrier factors in the partition are not extracted based on cultivated lands with similar characteristics, and the extracted barrier factors have large differences from actual conditions. According to a pre-established arable soil quality evaluation index system, arable soil quality evaluation index data of a typical sample area in a research area are firstly obtained, sampling units in the research area are clustered to obtain a plurality of sampling unit sets, and a plurality of subareas of the arable soil are obtained; and then, aiming at each area, obtaining the cultivated land characteristics of the subareas, and simultaneously adopting the index obstacle degree to identify the obstacle factor of each subarea.
Description
Technical Field
The invention belongs to the technical field of agricultural engineering, and relates to an automatic induction and obstacle factor diagnosis method for regional farmland protection partition and characteristics.
Background
Cultivated lands are basic stones for sustainable development of agriculture and even human beings, so that the cultivation quality represented by black lands in China is significant and is imperative to be protected and promoted. Objective and accurate farmland protection partition and farmland obstacle factor diagnosis are the basis for carrying out farmland quality regulation and control work, and are important for carrying out farmland quality early warning and establishing personalized farmland quality improvement measures in a targeted manner. However, the related work currently carried out in China has the following main problems:
(1) The conventional farmland protection partition is divided by adopting administrative division (such as administrative villages, villages and towns, counties and the like) and a local space autocorrelation method, and the soil area blocks which are determined by adopting the administrative division or the local space autocorrelation method are subjectively considered to have similarity based on the sampling result of the sampling unit.
(2) Most of the existing partition obstacle factor identification is extracted in the traditional partition in the step (1), and the judgment of the obstacle factor is not accurate enough due to different or even larger difference of the characteristics of cultivated lands in the partition, so that deviation is generated in the formulation of a regulation strategy. In addition, the traditional obstacle factor diagnosis extremely depends on manual identification and statistics, and has the problem of heavy cognitive load and large workload.
Disclosure of Invention
The invention aims to solve the problems that the existing cultivated quality partitioning scheme is divided according to administrative division and local space autocorrelation methods, the partitioning cannot be carried out according to objective conditions of cultivated lands, and the land in the same partition is likely to have large similarity difference, so that the characteristics and barrier factors in the partition are not extracted based on cultivated lands with similar characteristics, and the extracted barrier factors have large differences from actual conditions.
A method for protecting and partitioning a cultivated land, comprising the following steps:
s1, acquiring typical sample farmland quality evaluation index data:
setting sampling units in a region to be partitioned of cultivated land according to a pre-established cultivated land quality evaluation index system, wherein the sampling units are a land grid with a fixed area, and selecting typical sample areas in each sampling unit to obtain sample land data, wherein the sample land data comprise land utilization types, soil type data, soil physicochemical property data, topography data, infrastructure construction data, natural condition data, production condition data, soil profile property data, fertilizer and pesticide application amount and agricultural machinery total power data; dividing the continuous numerical value into a continuous numerical variable and a discrete type according to whether each index in the sample plot data is a continuous numerical value;
s2, automatic partition of cultivated land:
s2.1, constructing a study area tillage quality evaluation index data set D= { x based on the acquired sample tillage quality evaluation index data 1 ,x 2 ,...,x n (i-th pattern data x) i ＝{x i1 ,x 2 ,...,x ip ,x i(p+1) ,...,x im I is more than or equal to 1 and less than or equal to n, n is the number of sampling units, and m is the number of data types of sample area data, namely the number of indexes; the first p indexes are continuous numerical variables, and the p+1 to m indexes are discrete classification variables;
s2.2, d= { x 1 ,x 2 ,...,x n The n sampling units are subjected to cluster analysis by taking the n sampling units as input, and the specific process is as follows:
1) Is provided withDetermining the searching range [ k ] of the number of clusters min ,k max ]，k min ,k max Respectively the minimum value and the maximum value of the clustering number; let k=k min ；
2) Calculate d= { x 1 ,x 2 ,...,x n Mean center c of } 0 ＝{c 01 ,c 02 ,...,c 0p ,c 0(p+1) ,...,c 0m The average value corresponding to the attribute of the first p continuous numerical variables is determined by calculating the average value of the values of the attribute on n samples; the average value corresponding to each of the p+1 to m discrete classification variables is calculated by calculating the expected value E (T w ) Determining;
3) Judging that k is less than or equal to k max If yes, go to 4), otherwise go to 8);
4) Determining an initial cluster center c= { c 1 ,c 2 ,...,c k The procedure is as follows:
s2241, select c 0 The nearest sample is taken as the first initial cluster center c 1 ；
S2242, let d=d- { c 1 Respectively calculating the remaining n-1 samples to c 1 At a distance from c 1 The sample with the farthest distance is taken as a second initial clustering center c 2 ；
S2243, continuously searching an initial clustering center, wherein for the first initial clustering center, l is more than or equal to 3 and less than or equal to k, and D=D- { c 1 ,c 2 ,...,c l-1 Respectively calculating each sample in the remaining n- (l-1) samples to c 1 、c 2 、…、c l-1 Distance d of (2) i′j′ 1.ltoreq.i '. Ltoreq.n- (l-1), 1.ltoreq.j'. Ltoreq.l-1; sample i' to c 1 、c 2 、…、c l-1 The minimum value of the distance is denoted as d i′ ＝min(d i′1 ,d i′2 ,...,d i′(l-1) ) With argmax { d ] 1 ,d 2 ,...,d n-(l-1) The corresponding sample is used as the first initial clustering center c l ；
Repeating the process of searching the initial cluster centers until all k initial cluster centers are found;
5) Partitional clustering C (k) ＝{C 1 ,C 2 ,...,C k The procedure is as follows:
(1) let D' = { x 1 ,x 2 ,...,x q D= { x } is 1 ,x 2 ,...,x n Removing c= { c } 1 ,c 2 ,...,c k A sample set of }; let c' = { c 1 ',c 2 ',...,c k ' represents C 1 、C 2 、…、C k Is a mean center set of (1); initially, let C 1 ＝{c 1 },C 2 ＝{c 2 },...,C k ＝{c k }；
(2) For D' = { x 1 ,x 2 ,...,x q Sample x in } i Respectively calculate its to c 1 '、c 2 '、…、c k Distance d of ij ' i is equal to or less than 1 and equal to or less than q, j is equal to or less than 1 and equal to or less than k, the distance is calculated according to a formula (2), and the ith sample is taken to c 1 '、c 2 '、…、c k The minimum value of the distance is denoted as d i '＝min(d i1 ',d i2 ',...,d ik ')；
(3) Sample x i Divided into distance d i ' corresponding class C r Wherein r is more than or equal to 1 and less than or equal to k;
(4) cycling steps (2) to (3) until D' = { x 1 ,x 2 ,...,x q All samples in the sequence are divided;
(5) update C 1 、C 2 、…、C k Set and recalculate mean center set c' = { c 1 ',c 2 ',...,c k '}；
(6) Repeating steps (2) to (5) until c' = { c 1 ',c 2 ',...,c k ' no change occurs any more;
6) Calculating a cluster effectiveness index avgBWP (k);
7) Let k=k+1, turn 3);
8) Determining the number of best clustersCorresponding to itThe best clustering result is obtained;
s2.3, orderFinally obtaining k farmland subareas P= { P of the research area 1 ,P 2 ,...,P k }。
Further, the land utilization type and soil type data are obtained from a land utilization map and a soil map, the soil physicochemical property data are obtained from monitoring point sampling data, the topographic data are obtained from DEM raster data, the infrastructure construction data are obtained from remote sensing images, and the natural condition, the production condition and the soil profile property data are obtained from field investigation data; the total power data of the chemical fertilizer and the pesticide is obtained from the social and economic statistical data.
Further, the soil physicochemical property data comprise soil organic matter content, pH value, trace elements and soil volume weight; the terrain data comprises a terrain part, a gradient, a slope direction and an elevation; the infrastructure construction data comprise field block breaking degree and field forest networking degree; the natural condition data comprise temperature distribution, rainfall, underground water burial depth and biodiversity, the production condition data comprise field block regularity, field road accessibility, irrigation capacity, drainage capacity, cleanliness and soil salinization degree, and the soil profile property data comprise plough layer thickness, plough layer texture and texture configuration.
Further, the average value corresponding to each discrete classification variable is obtained by calculating the expected value E (T w ) The determination process of (2) is as follows:
wherein T is w The index corresponding to the w-th discrete classification variable; f is an index T w The number of the values of (a); t (T) wf As index T w The f-th value of (2); p (T) wf ) For T in the sample set wf Probability of occurrence.
Further, index T w By encoding the indicator as its value.
Further, the cluster effectiveness index avgBWP (k) is as follows:
baw(h,i)＝b(h,i)+w(h,i)
bsw(h,i)＝b(h,i)-w(h,i)
wherein, avgBWP (k) is a cluster effectiveness index when data is divided into k classes; BWP (h, i) is an inter-class division of the ith sample of the h-th class; bsw (h, i) is the cluster dispersion distance of the ith sample of the h class; baw (h, i) is the cluster distance of the ith sample of the h class; b (h, i) is the minimum inter-class distance of the ith sample of the h class, i.e. the minimum value of the average distance of the sample i from the samples in each of the other classes; w (h, i) is the intra-class distance of the ith sample of the h-th class, i.e. the average distance of sample i from all other samples in the h-th class; n is n r 、n h The number of samples in class r and class h, respectively;the property of the ith sample in the ith class and the property of the ith sample in the ith class are respectively; p is the number of continuous numerical attributes; m is the total number of sample attributes.
Further, step S2241 selects the separation c 0 The nearest sample is taken as the first initial cluster center c 1 Is too much to (a)In the course, two samples x i And x j Distance d between ij The calculation formula of (2) is as follows:
wherein p is the number of continuous numerical attributes; m is the total number of attributes in the data set; x is x it A nth attribute that is an ith sample; x is x jt A jth attribute for a jth sample; μ is the weighting factor of the classification attribute.
Further, in step 1), k min ＝2，
Firstly, dividing the cultivated land by using the cultivated land protection partition method, and then carrying out feature induction based on a partition result;
the process of feature induction based on the partition result comprises the following steps:
for each partition P s Extracting common characteristics of the partitioned cultivated land on the quality evaluation index as partition characteristics; aiming at the quality evaluation indexes of two different value types, the following processing is carried out:
nominal index: index T w The value set of (2) is marked as S w ＝{t w1 ,t w2 ,..,t wy P+1 is less than or equal to w is less than or equal to m, for T w Is given a value t wo O=1, 2,..y, y represents an index T w The number of possible values; statistical partition P s The number c of sampling units corresponding to the value wo ；
Numerical index: for index T w′ Finding out the partition P, wherein w' is not less than 1 and not more than P s Maximum value t of the index in each sampling unit w′_max And a minimum value t w′_min Discretizing the value of the index to obtain a plurality of discrete intervals of the value of the index to form an index T w′ Value set S of (2) w′ ＝{t w′1 ,t w′2 ,..,t w′v -a }; then, for T w′ Is given a value t w′o′ O' =1, 2, where, v, statistical partition P s The number c of sampling units corresponding to the value w'o' ；
For partition P s Each index T of the middle sample plot w And T w′ Number of sampling units c wo And c w′o′ Descending order of arrangement is carried out, and a partition P is obtained according to the arrangement result s The majority of the features presented.
The obstacle factor diagnosis method based on the cultivated land protection partition comprises the steps of firstly partitioning cultivated land by using the cultivated land protection partition method, and then performing obstacle factor diagnosis based on a partition result:
determining a weight value W of each cultivated quality evaluation index aiming at a pre-established cultivated quality evaluation index system a The method comprises the steps of carrying out a first treatment on the surface of the Meanwhile, determining the membership function F of each plough quality evaluation index by adopting a Teerfihai method a ；
For partition P s Calculating the obstacle degree OD of each quality evaluation index to the cultivated quality a ：
Wherein F is ab The membership degree calculated for the a quality evaluation index according to the b sample plot monitoring data; c (C) a A weight value of the a-th quality evaluation index; n is n s For partition P s The total number of sampling units in the system;
dividing the index obstacle degree into obstacle degree levels according to an equidistant method; then for each partition, according to the obstacle degree OD of each index a And determining the corresponding obstacle factor, namely the obstacle degree level corresponding to the obstacle factor.
The beneficial effects are that:
1. the method can objectively and scientifically partition the cultivated land from the actual expression of the cultivated land on each quality evaluation index, ensures the similarity of the land in the area, improves the accuracy of partition, and solves the problems of strong subjectivity, poor interpretability and the like and large difference of the land similarity existing in the traditional partition method.
2. The method and the device can automatically induce the remarkable characteristics of each cultivated land partition based on the partition result, obtain the main characteristics of each partition presented under a given index system, and help to explain the meaning of each partition result, thereby improving the credibility of the automatic partition result, and also provide a more targeted, more accurate and effective data basis for the development, utilization and protection strategy of the subsequent cultivated lands according to the main characteristics.
3. The invention can realize the automatic diagnosis of the partition obstacle factors, and greatly reduces the labor and material costs of the related work of the protection of the cultivated quality or the improvement. Secondly, by carrying out obstacle factor diagnosis in each partition, different obstacle factors of different partitions can be found, so that different cultivation quality protection and improvement strategies are formulated by taking the partition as a unit, and the cultivation quality protection or improvement of the working quality is improved while the cost of manpower and material resources is reduced.
Drawings
FIG. 1 shows a method for automatically inducing and diagnosing obstacle factors in a protected partition and features of a cultivated land.
Detailed Description
According to a pre-established arable soil quality evaluation index system, arable soil quality evaluation index data of a typical sample area in a research area are firstly obtained; then, clustering sampling units in the research area to obtain a plurality of sampling unit sets to form a plurality of subareas of the cultivated land of the research area, wherein a clustering effect quality evaluation index is designed aiming at the effective measurement problem of the optimal clustering number of a clustering algorithm, and the index can effectively evaluate the inter-class distance and the intra-class distance of the clusters; aiming at the problem that the randomness of the initial cluster center selection causes poor comparability of the clustering results under different clustering numbers, a sample mean point is adopted as a first initial cluster center, and an effective initial cluster center selection method is designed, so that the stability of the initial cluster center can be ensured; further, for each subarea, the cultivated land characteristics of the subarea are automatically obtained by calculating the value of each cultivated land quality evaluation index or the number of sampling units in a discrete interval; and finally, identifying the obstacle factors of each partition by adopting index obstacle degree, and automatically forming an obstacle factor diagnosis matrix to realize automatic diagnosis of the obstacle factors of the partition of the cultivated quality. The method can objectively and scientifically realize the zoning of the regional cultivated land, can automatically carry out the induction of the characteristics of the zoned cultivated land and the diagnosis of the zoning barrier factors, not only improves the scientificity of cultivated quality evaluation and the efficiency of related work, but also reduces the cognitive load, and has important significance for developing personalized cultivated quality protection and lifting work.
The first embodiment is as follows: the present embodiment will be described with reference to figure 1,
the present embodiment is a method for diagnosing a protected partition and an obstacle factor of a cultivated land, comprising the steps of:
1. obtaining typical sample land tillage quality evaluation index data:
according to the pre-established arable soil quality evaluation index system, sampling units, i.e. a fixed area (such as 5hm 2 Or 10hm 2 ) Selecting a typical sample area from each sampling unit to obtain sample area data, wherein the sample area data comprises land utilization type, soil type data, soil physicochemical property data, terrain data, infrastructure construction data, natural condition data, production condition data, soil profile property data, and data such as fertilizer and pesticide application amount, total power of agricultural machinery and the like; dividing each index in the sample plot data into a continuous numerical variable and a discrete classification variable according to whether each index is a continuous numerical value or not;
the method comprises the steps of acquiring data such as land utilization type, soil type and the like of a sampling unit from a land utilization map and a soil map in a targeted manner;
acquiring soil physical and chemical property data from sampling data of monitoring points, wherein the soil physical and chemical property data comprise soil organic matter content, pH value, trace elements, soil volume weight and the like;
obtaining terrain data from DEM raster data, wherein the terrain data comprises terrain parts, gradients, slope directions, altitudes and the like;
infrastructure construction data are obtained from the remote sensing images, wherein the infrastructure construction data comprise field block breaking degree, field forest networking degree and the like;
acquiring data such as natural conditions, production conditions, soil profile properties and the like from field investigation data; the natural condition data comprise temperature distribution, rainfall, underground water burial depth, biodiversity and the like, the production condition data comprise field block regularity, field road accessibility, irrigation capacity, drainage capacity, cleanliness, soil salinization degree and the like, and the soil profile property data comprise plough layer thickness, plough layer texture, texture configuration and the like;
and acquiring data such as the application amount of chemical fertilizer and pesticide, the total power of agricultural machinery and the like from social and economic statistical data.
The invention uses the relative data of the non-soil self attribute on the surfaces such as the fertilizer and pesticide application amount, the total power of the agricultural machinery and the like to process, in practice, the invention processes the data of the non-soil self attribute on the surfaces, but actually influences the soil self attribute, thereby being equivalent to expanding the data characterization space of the soil self attribute, not only being capable of classifying the soil more accurately, thereby improving the accuracy of the subareas, further improving the similarity of the characteristics in the subareas and the accuracy of determining the subarea tillage quality obstacle factors, being capable of providing more effective measures for the subsequent tillage protection, being closer to the implementation environment and the situation of the measures, and ensuring the effectiveness and the landing property of the measures.
The arable quality evaluation index system is constructed according to regional arable quality grade division indexes specified in GB/T33469-2016 arable quality grade, and can be expanded on the basis of GB/T33469-2016 by combining with the actual conditions of a research area.
2. Automatic partition of cultivated land:
2.1 based on the obtained sample soil texture evaluation index dataConstructing a study area tillage quality evaluation index data set D= { x 1 ,x 2 ,...,x n (i-th pattern data x) i ＝{x i1 ,x 2 ,...,x ip ,x i(p+1) ,...,x im I is more than or equal to 1 and less than or equal to n, n is the number of sampling units, and m is the number of data types of sample area data, namely the number of indexes; the first p indexes are continuous numerical variables, including effective soil layer thickness, soil organic matter content, plough layer thickness, soil pH value, soil volume weight, soil trace element content, field gradient, underground water burial depth, altitude, accumulated temperature, rainfall, chemical fertilizer and pesticide application amount, total agricultural machinery power and the like; the p+1 to m indexes are discrete classification variables such as land utilization type, soil type, terrain position, slope direction, plough layer texture, texture configuration, biodiversity, cleanliness degree, irrigation capacity, drainage capacity, farmland forest networking degree, farmland block regularity, field block crushing degree, field road accessibility degree, salinization degree and the like.
2.2, as d= { x 1 ,x 2 ,...,x n The n sampling units are subjected to cluster analysis by taking the n sampling units as input, and the specific algorithm process is as follows:
1) Setting a cluster number searching range [ k ] min ,k max ]，k min ,k max Respectively the minimum value and the maximum value of the clustering number, and k is taken in the embodiment min ＝2，Let k=k min 。
2) Calculate d= { x 1 ,x 2 ,...,x n Mean center c of } 0 ＝{c 01 ,c 02 ,...,c 0p ,c 0(p+1) ,...,c 0m The average value corresponding to the attribute of the first p continuous numerical variables is obtained by calculating the average value of the values of the attribute on n samples; the average value corresponding to each of the p+1 to m discrete classification variables is calculated by calculating the expected value E (T) w ) The calculation method comprises the following steps:
wherein T is w (p+1 is less than or equal to w is less than or equal to m) is the w-th discrete classification variable; f is a variable T w The number of the values of (a); t (T) wf As variable T w For non-numeric indicators, by encoding the indicator as its value; p (T) wf ) For T in the sample set wf Probability of occurrence, i.e. T wf The ratio of the number of occurrences to the total number of samples.
3) Judging that k is less than or equal to k max If yes, go to 4), otherwise go to 8);
4) Determining an initial cluster center c= { c 1 ,c 2 ,...,c k The procedure is as follows:
(1) selecting the distance c 0 The nearest sample is taken as the first initial cluster center c 1 ；
Two samples x i And x j Distance d between ij The calculation formula of (2) is as follows:
wherein p is the number of continuous numerical attributes; m is the total number of attributes in the data set; x is x it A nth attribute that is an ith sample; x is x jt A jth attribute for a jth sample; μ is a weighting factor for the classification attribute;
(2) let d=d- { c 1 I.e. as c is removed from sample set D 1 The set of remaining samples is still denoted by symbol D), and the remaining n-1 samples are calculated to c, respectively 1 At a distance from c 1 The sample with the farthest distance is taken as a second initial clustering center c 2 ；
(3) And continuously searching an initial clustering center, wherein for the first initial clustering center, l is more than or equal to 3 and less than or equal to k, and D=D-{c 1 ,c 2 ,...,c l-1 Respectively calculating each sample in the remaining n- (l-1) samples to c 1 、c 2 、…、c l-1 Distance d of (2) i′j′ 1.ltoreq.i '. Ltoreq.n- (l-1), 1.ltoreq.j'. Ltoreq.l-1; sample i' to c 1 、c 2 、…、c l-1 The minimum value of the distance is denoted as d i′ ＝min(d i′1 ,d i′2 ,...,d i′(l-1) ) With argmax { d ] 1 ,d 2 ,...,d n-(l-1) The corresponding sample is used as the first initial clustering center c l 。
The process of finding the initial cluster centers is repeated until all k initial cluster centers are found.
5) Partitional clustering C (k) ＝{C 1 ,C 2 ,...,C k The procedure is as follows:
(1) let D' = { x 1 ,x 2 ,...,x q D= { x } is 1 ,x 2 ,...,x n Removing c= { c } 1 ,c 2 ,...,c k A sample set of }; let c' = { c 1 ',c 2 ',...,c k ' represents C 1 、C 2 、…、C k Is a mean center set of (1); initially, let C 1 ＝{c 1 },C 2 ＝{c 2 },...,C k ＝{c k }；
(2) For D' = { x 1 ,x 2 ,...,x q Sample x in } i Respectively calculate its to c 1 '、c 2 '、…、c k Distance d of ij ' i is equal to or less than 1 and equal to or less than q, j is equal to or less than 1 and equal to or less than k, the distance is calculated according to a formula (2), and the ith sample is taken to c 1 '、c 2 '、…、c k The minimum value of the distance is denoted as d i '＝min(d i1 ',d i2 ',...,d ik ')；
(3) Sample x i Divided into distance d i ' corresponding class C r Wherein r is more than or equal to 1 and less than or equal to k;
(4) cycling steps (2) to (3) until D' = { x 1 ,x 2 ,...,x q All samples in the sequence are divided;
(5) update C 1 、C 2 、…、C k Set and recalculate mean center set c' = { c 1 ',c 2 ',...,c k '}；
(6) Repeating steps (2) to (5) until c' = { c 1 ',c 2 ',...,c k ' no change occurs.
6) The cluster effectiveness index avgBWP (k) is calculated according to the following calculation formula:
baw(h,i)＝b(h,i)+w(h,i) (6)
bsw(h,i)＝b(h,i)-w(h,i) (7)
wherein, avgBWP (k) is a cluster effectiveness index when data is divided into k classes; BWP (h, i) is an inter-class division of the ith sample of the h-th class; bsw (h, i) is the cluster dispersion distance of the ith sample of the h class; baw (h, i) is the cluster distance of the ith sample of the h class; b (h, i) is the minimum inter-class distance of the ith sample of the h class, i.e. the minimum value of the average distance of the sample i from the samples in each of the other classes; w (h, i) is the intra-class distance of the ith sample of the h-th class, i.e. the average distance of sample i from all other samples in the h-th class; n is n r 、n h The number of samples in class r and class h, respectively;the t attribute of the ith sample in the nth class and the t of the ith sample in the nth class respectivelyA plurality of attributes; p is the number of continuous numerical attributes; m is the total number of sample attributes.
7) Let k=k+1, turn 3);
8) Determining the number of best clustersCorresponding to itAnd the best clustering result is obtained.
2.3 orderFinally obtaining k farmland subareas P= { P of the research area 1 ,P 2 ,...,P k }。
3. And (3) automatic induction of partition characteristics:
for each partition P s And extracting common characteristics of the partitioned cultivated land on the quality evaluation index as partition characteristics. The traditional method needs to manually summarize and sort index features of all sampling units in the partition, and is time-consuming and labor-consuming. Since the quality evaluation index data of the sample area covered in each partition is known, and the index values are generally classified into two types of a nominal type and a numerical type, the partition characteristics are automatically induced by adopting a statistical method. For each partition P s For the quality evaluation indexes of two different value types, the following treatments are respectively carried out:
(1) Nominal index: index T w The value set of (2) is marked as S w ＝{t w1 ,t w2 ,..,t wy P+1 is less than or equal to w is less than or equal to m, for T w Is given a value t wo O=1, 2,..y, y represents an index T w The number of possible values, for example, 7 types for the terrain, 7 corresponding to the value, 7 corresponding to y, and statistical partition P s The number c of sampling units corresponding to the value wo 。
(2) Numerical index: for index T w′ Finding out the partition P, wherein w' is not less than 1 and not more than P s Maximum value t of the index in each sampling unit w′_max And a minimum value t w′_min Discretizing the value of the index to obtain a plurality of discrete intervals of the value of the index to form an index T w′ Value set S of (2) w′ ＝{t w′1 ,t w′2 ,..,t w′v -a }; then, for T w′ Is given a value t w′o′ O' =1, 2, where, v, statistical partition P s The number c of sampling units corresponding to the value w'o' 。
After the above processing, the partition P can be obtained s The number distribution of sampling units at different values or intervals of different indexes is shown in table 1.
TABLE 1 partition P s Distribution of sampling unit number on different values of different indexes in a computer
For partition P s Each index T of the middle sample plot w (p+1. Ltoreq.w.ltoreq.m) and T w′ (w' is not less than 1 and not more than p), and the number of sampling units c wo (o=1, 2,) y) and c w′o′ (o' =1, 2.,. V) in descending order, partition P can be obtained s The majority of the features presented.
Through the process, after each sampling unit of the research area is automatically partitioned, the value with the largest index ratio in each partition can be further automatically induced, so that the main characteristics of each partition presented under a given index system are obtained, the interpretation of the meaning of each partition result is facilitated, the reliability of the automatic partition result is improved, and a more targeted, more accurate and effective data basis can be provided for the subsequent preparation of the tillage quality protection and lifting strategy according to the main characteristics.
4. Automatic diagnosis of partition obstacle factors:
(1) For pre-established cultivationThe weight value W of each cultivated quality evaluation index is determined by adopting a common principal component analysis method, a hierarchical analysis method, an entropy weight method and the like in the quality evaluation index system a Determining the weight value W of each plough quality evaluation index a The process of (1) can adopt a single method to acquire the index weight value, or can weight the weight value obtained by each method, and the weighted value is used as the weight value of the index;
meanwhile, determining the membership function F of each plough quality evaluation index by adopting a Teerfihai method a Can be realized by referring to the tillage quality grade.
(2) For partition P s Calculating the obstacle degree OD of each quality evaluation index to the cultivated quality a The calculation formula is as follows:
wherein F is ab The membership degree calculated for the a quality evaluation index according to the b sample plot monitoring data; c (C) a A weight value of the a-th quality evaluation index; n is n s For partition P s The total number of sampling units in the system.
(3) According to the equidistant method, the index Obstacle Degree (OD) is divided into no obvious obstacle (0-10%), slight obstacle (10-20%), moderate obstacle (20-30%), strength obstacle (30-50%) and severe obstacle>50%) 5 grades. Then for each partition, according to the obstacle degree OD of each index a Its corresponding obstacle factor, i.e. the level to which the obstacle factor corresponds, is determined.
For each partition, the combination of the tilling quality obstacle factors of each partition is divided according to the change condition of the quality evaluation index, so as to obtain a diagnosis matrix of the tilling quality obstacle factors of different partitions, as shown in table 2.
TABLE 2 matrix for diagnosis of different partition tilling quality obstacle factors
According to the invention, the sampling units in the research area are subjected to similarity calculation, a plurality of sampling areas under the constraint of a set index system are automatically divided into a plurality of subareas, the sampling areas in the same subarea have great similarity, and the sampling areas among different subareas have great differences, so that diagnosis of obstacle factors and formulation of strategies for protecting or improving the cultivated quality are not carried out by taking the sampling units as units, but by taking the subareas comprising a plurality of sampling units as units, and as each sampling unit in each subarea has great similarity, the cost of manpower and material resources for protecting or improving the cultivated quality is greatly reduced. Secondly, by carrying out obstacle factor diagnosis in each partition, different obstacle factors of different partitions can be found, so that different cultivation quality protection and improvement strategies are formulated by taking the partition as a unit, and the cultivation quality protection or improvement of the working quality is improved while the cost of manpower and material resources is reduced.
Examples
Taking a certain research area of the northeast black soil area as an example, combining the actual situation of the research area and the cultivated quality evaluation grade, the cultivated quality evaluation index system comprises: terrain part, effective soil layer thickness, organic matter content, plough layer texture, soil volume weight, texture configuration, effective phosphorus content, quick-acting potassium content, biodiversity condition, irrigation capacity, farmland forest networking degree, soil pH value and plough layer thickness, and obtaining a corresponding sample land data set D= { x 1 ,x 2 ,...,x n }. Wherein x is 1 ～x 7 The soil is a numerical index, and sequentially represents the effective soil layer thickness, the organic matter content, the soil volume weight, the effective phosphorus content, the quick-acting potassium content, the soil pH value and the plough layer thickness; x is x 8 ～x 13 Is a nominal type index, and sequentially represents the terrain part, the plough layer texture, the texture configuration, the biodiversity condition, the irrigation capacity and the farmland forest networking degree.
2. For dataset d= { x 1 ,x 2 ,...,x n Cluster analysis is carried out to obtain 5 clusters of a research area, and 5 farmland protection partitions P= { P of the research area are formed 1 ,P 2 ,P 3 ,P 4 ,P 5 Each partition P s And (1) s is less than or equal to 5) comprises a plurality of sample area data.
3. For each partition P s The partition characteristics are automatically summarized. For example, for a partition P of the investigation region 1 The distribution of the number of sampling units on different index values is shown in tables 3 and 4.
TABLE 3 partition P 1 Sample cell number distribution example at different values of each index (nominal index)
TABLE 4 partition P 1 Sample cell number distribution examples at different values of each index (numerical index)
As shown in tables 3 and 4, according to partition P 1 The distribution of the number of sample sites on different indexes can be summarized as the partition characteristics: the topography is mainly the hillside land, and soil is compact type heavy loam, and the biodiversity is general, can satisfy irrigation needs and farmland forest network degree is high. The effective soil layer thickness is mainly distributed between 80 cm and 100cm, the organic matter content is high (more than or equal to 20 g/kg), the soil volume weight is moderate, the effective phosphorus content is generally (15 mg/kg to 25 mg/kg), the quick-acting potassium content is moderate, the soil pH value is moderate, and the plough layer thickness is mainly distributed between 20 cm and 25 cm.
4. Based on the generalization of the partition characteristics, for each partition P s And automatically diagnosing the obstacle factors according to the index obstacle degree respectively to form an obstacle factor diagnosis matrix. For example, for 5 partitions of the study area, the resulting obstacle factor diagnostic matrices are shown in table 5.
Table 5 example of study area obstacle factor diagnostic matrix
As shown in table 5, for partition 1, the effective phosphorus content is the partition severe obstacle factor, the plough layer texture, irrigation capacity, organic matter content and quick-acting potassium content are the intensity obstacle factors, the texture configuration, the degree of farmland forest reticulation and the plough layer thickness are the moderate obstacle factors, and the biodiversity condition and the effective soil layer thickness are the mild obstacle factors; however, for partition 2, the terrain part and the effective phosphorus content are the serious obstacle factors, the organic matter content, the quick-acting potassium content and the soil pH value are the strength obstacle factors, the effective soil layer thickness, the soil volume weight and the plough layer thickness are the moderate obstacle factors, and the biological diversity condition and the farmland forest network degree are the mild obstacle factors. By means of different barrier factors of the subareas 1 and 2, government departments can formulate different cultivation quality protection and lifting strategies according to the different barrier factors, and therefore quick, low-cost and differentiated protection of cultivation quality in the area is achieved.
The above examples of the present invention are only for describing the calculation model and calculation flow of the present invention in detail, and are not limiting of the embodiments of the present invention. Other variations and modifications of the above description will be apparent to those of ordinary skill in the art, and it is not intended to be exhaustive of all embodiments, all of which are within the scope of the invention.
Claims (10)
1. A method for protecting and partitioning a cultivated land, comprising the steps of:
s1, acquiring typical sample farmland quality evaluation index data:
setting sampling units in a region to be partitioned of cultivated land according to a pre-established cultivated land quality evaluation index system, wherein the sampling units are a land grid with a fixed area, and selecting typical sample areas in each sampling unit to obtain sample land data, wherein the sample land data comprise land utilization types, soil type data, soil physicochemical property data, topography data, infrastructure construction data, natural condition data, production condition data, soil profile property data, fertilizer and pesticide application amount and agricultural machinery total power data; dividing the continuous numerical value into a continuous numerical variable and a discrete type according to whether each index in the sample plot data is a continuous numerical value;
s2, automatic partition of cultivated land:
s2.1, constructing a study area tillage quality evaluation index data set D= { x based on the acquired sample tillage quality evaluation index data 1 ,x 2 ,...,x n (i-th pattern data x) i ＝{x i1 ,x 2 ,...,x ip ,x i(p+1) ,...,x im I is more than or equal to 1 and less than or equal to n, n is the number of sampling units, and m is the number of data types of sample area data, namely the number of indexes; the first p indexes are continuous numerical variables, and the p+1 to m indexes are discrete classification variables;
s2.2, d= { x 1 ,x 2 ,...,x n The n sampling units are subjected to cluster analysis by taking the n sampling units as input, and the specific process is as follows:
1) Setting a cluster number searching range [ k ] min ,k max ]，k min ,k max Respectively the minimum value and the maximum value of the clustering number; let k=k min ；
2) Calculate d= { x 1 ,x 2 ,...,x n Mean center c of } 0 ＝{c 01 ,c 02 ,...,c 0p ,c 0(p+1) ,...,c 0m The average value corresponding to the attribute of the first p continuous numerical variables is determined by calculating the average value of the values of the attribute on n samples; the average value corresponding to each of the p+1 to m discrete classification variables is calculated by calculating the expected value E (T w ) Determining;
3) Judging that k is less than or equal to k max If yes, go to 4), otherwise go to 8);
4) Determining an initial cluster center c= { c 1 ,c 2 ,...,c k The procedure is as follows:
s2241, select c 0 The nearest sample is taken as the first initial cluster center c 1 ；
S2242, let d=d- { c 1 Respectively calculating the remaining n-1 samples to c 1 At a distance from c 1 The sample with the farthest distance is taken as a second initial clustering center c 2 ；
S2243, continuously searching an initial clustering center, wherein for the first initial clustering center, l is more than or equal to 3 and less than or equal to k, and D=D- { c 1 ,c 2 ,...,c l-1 Respectively calculating each sample in the remaining n- (l-1) samples to c 1 、c 2 、…、c l-1 Distance d of (2) i′j′ 1.ltoreq.i '. Ltoreq.n- (l-1), 1.ltoreq.j'. Ltoreq.l-1; sample i' to c 1 、c 2 、…、c l-1 The minimum value of the distance is denoted as d i′ ＝min(d i′1 ,d i′2 ,...,d i′(l-1) ) With argmax { d ] 1 ,d 2 ,...,d n-(l-1) The corresponding sample is used as the first initial clustering center c l ；
Repeating the process of searching the initial cluster centers until all k initial cluster centers are found;
5) Partitional clustering C (k) ＝{C 1 ,C 2 ,...,C k The procedure is as follows:
(1) let D' = { x 1 ,x 2 ,...,x q D= { x } is 1 ,x 2 ,...,x n Removing c= { c } 1 ,c 2 ,...,c k A sample set of }; let c' = { c 1 ',c 2 ',...,c k ' represents C 1 、C 2 、…、C k Is a mean center set of (1); initially, let C 1 ＝{c 1 },C 2 ＝{c 2 },...,C k ＝{c k }；
(2) For D' = { x 1 ,x 2 ,...,x q Sample x in } i Respectively calculate its to c 1 '、c 2 '、…、c k ' sDistance d ij ' i is equal to or less than 1 and equal to or less than q, j is equal to or less than 1 and equal to or less than k, the distance is calculated according to a formula (2), and the ith sample is taken to c 1 '、c 2 '、…、c k The minimum value of the distance is denoted as d i '＝min(d i1 ',d i2 ',...,d ik ')；
(3) Sample x i Divided into distance d i ' corresponding class C r Wherein r is more than or equal to 1 and less than or equal to k;
(4) cycling steps (2) to (3) until D' = { x 1 ,x 2 ,...,x q All samples in the sequence are divided;
(5) update C 1 、C 2 、…、C k Set and recalculate mean center set c' = { c 1 ',c 2 ',...,c k '}；
(6) Repeating steps (2) to (5) until c' = { c 1 ',c 2 ',...,c k ' no change occurs any more;
6) Calculating a cluster effectiveness index avgBWP (k);
7) Let k=k+1, turn 3);
8) Determining the number of best clustersCorresponding to itThe best clustering result is obtained;
s2.3, orderFinally obtaining k farmland subareas P= { P of the research area 1 ,P 2 ,...,P k }。
2. The method for protecting and partitioning a cultivated land according to claim 1, wherein the land use type and the soil type data are obtained from a land use map and a soil map, the soil physicochemical property data are obtained from monitoring point sampling data, the topography data are obtained from DEM raster data, the infrastructure construction data are obtained from remote sensing images, and the natural condition, the production condition and the soil profile property data are obtained from field investigation data; the total power data of the chemical fertilizer and the pesticide is obtained from the social and economic statistical data.
3. The method for protecting and partitioning a cultivated land according to claim 2, wherein the data of physical and chemical properties of the soil comprises soil organic matter content, pH value, trace elements and soil volume weight; the terrain data comprises a terrain part, a gradient, a slope direction and an elevation; the infrastructure construction data comprise field block breaking degree and field forest networking degree; the natural condition data comprise temperature distribution, rainfall, underground water burial depth and biodiversity, the production condition data comprise field block regularity, field road accessibility, irrigation capacity, drainage capacity, cleanliness and soil salinization degree, and the soil profile property data comprise plough layer thickness, plough layer texture and texture configuration.
4. A method according to claim 1,2 or 3, wherein the average value of the discrete classification variables is calculated by calculating the expected value E (T w ) The determination process of (2) is as follows:
wherein T is w The index corresponding to the w-th discrete classification variable; f is an index T w The number of the values of (a); t (T) wf As index T w The f-th value of (2); p (T) wf ) For T in the sample set wf Probability of occurrence.
5. The method for protecting and partitioning a cultivated land according to claim 4, wherein the index T w By encoding the indicator as its value.
6. The method for protecting and partitioning a cultivated land according to claim 5, wherein the cluster effectiveness index avgBWP (k) is as follows:
baw(h,i)＝b(h,i)+w(h,i)
bsw(h,i)＝b(h,i)-w(h,i)
wherein, avgBWP (k) is a cluster effectiveness index when data is divided into k classes; BWP (h, i) is an inter-class division of the ith sample of the h-th class; bsw (h, i) is the cluster dispersion distance of the ith sample of the h class; baw (h, i) is the cluster distance of the ith sample of the h class; b (h, i) is the minimum inter-class distance of the ith sample of the h class, i.e. the minimum value of the average distance of the sample i from the samples in each of the other classes; w (h, i) is the intra-class distance of the ith sample of the h-th class, i.e. the average distance of sample i from all other samples in the h-th class; n is n r 、n h The number of samples in class r and class h, respectively;the property of the ith sample in the ith class and the property of the ith sample in the ith class are respectively; p is the number of continuous numerical attributes; m is the total number of sample attributes.
7. Root of Chinese characterThe method of claim 6, wherein step S2241 selects the distance c 0 The nearest sample is taken as the first initial cluster center c 1 In the process of (2), two samples x i And x j Distance d between ij The calculation formula of (2) is as follows:
wherein p is the number of continuous numerical attributes; m is the total number of attributes in the data set; x is x it A nth attribute that is an ith sample; x is x jt A jth attribute for a jth sample; μ is the weighting factor of the classification attribute.
8. The method for protecting and partitioning a cultivated land as claimed in claim 7, wherein in step 1), k is min ＝2，
9. An automatic feature induction method based on farmland protection and partitioning, characterized in that firstly, farmland is partitioned by using a farmland protection and partitioning method according to any one of claims 1 to 8, and then feature induction is performed based on the partitioning result;
the process of feature induction based on the partition result comprises the following steps:
for each partition P s Extracting common characteristics of the partitioned cultivated land on the quality evaluation index as partition characteristics; aiming at the quality evaluation indexes of two different value types, the following processing is carried out:
nominal index: index T w The value set of (2) is marked as S w ＝{t w1 ,t w2 ,..,t wy P+1 is less than or equal to w is less than or equal to m, for T w Is given a value t wo O=1, 2,..y, y represents an index T w The number of possible values; statistical partition P s The number c of sampling units corresponding to the value wo ；
Numerical index: for index T w′ Finding out the partition P, wherein w' is not less than 1 and not more than P s Maximum value t of the index in each sampling unit w′_max And a minimum value t w′_min Discretizing the value of the index to obtain a plurality of discrete intervals of the value of the index to form an index T w′ Value set S of (2) w′ ＝{t w′1 ,t w′2 ,..,t w′v -a }; then, for T w′ Is given a value t w′o′ O' =1, 2, where, v, statistical partition P s The number c of sampling units corresponding to the value w'o' ；
For partition P s Each index T of the middle sample plot w And T w′ Number of sampling units c wo And c w′o′ Descending order of arrangement is carried out, and a partition P is obtained according to the arrangement result s The majority of the features presented.
10. A method for diagnosing an obstacle factor based on a protected area of a cultivated land, characterized in that firstly the cultivated land is partitioned by a method for protecting and partitioning a cultivated land according to any one of claims 1 to 8, and then the obstacle factor diagnosis is performed based on the result of the partitioning:
determining a weight value W of each cultivated quality evaluation index aiming at a pre-established cultivated quality evaluation index system a The method comprises the steps of carrying out a first treatment on the surface of the Meanwhile, determining the membership function F of each plough quality evaluation index by adopting a Teerfihai method a ；
For partition P s Calculating the obstacle degree OD of each quality evaluation index to the cultivated quality a ：
Wherein F is ab The membership degree calculated for the a quality evaluation index according to the b sample plot monitoring data; c (C) a A weight value of the a-th quality evaluation index; n is n s For partition P s The total number of sampling units in the system;
dividing the index obstacle degree into obstacle degree levels according to an equidistant method; then for each partition, according to the obstacle degree OD of each index a And determining the corresponding obstacle factor, namely the obstacle degree level corresponding to the obstacle factor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311408679.3A CN117408829A (en) | 2023-10-27 | 2023-10-27 | Method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311408679.3A CN117408829A (en) | 2023-10-27 | 2023-10-27 | Method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117408829A true CN117408829A (en) | 2024-01-16 |
Family
ID=89499617
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311408679.3A Pending CN117408829A (en) | 2023-10-27 | 2023-10-27 | Method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117408829A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101546421A (en) * | 2009-04-01 | 2009-09-30 | 河北农业大学 | Province-wide comparison tillage quality evaluation method based on GIS |
| CN106778826A (en) * | 2016-11-25 | 2017-05-31 | 南昌航空大学 | Based on the hereditary Hybrid Clustering Algorithm with preferred Fuzzy C average of self adaptation cellular |
| CN109492796A (en) * | 2018-10-12 | 2019-03-19 | 东南大学 | A kind of Urban Spatial Morphology automatic Mesh Partition Method and system |
| CN110909792A (en) * | 2019-11-21 | 2020-03-24 | 安徽大学 | Clustering analysis method based on improved K-means algorithm and new clustering effectiveness index |
| CN111291326A (en) * | 2020-02-06 | 2020-06-16 | 武汉大学 | Clustering effectiveness index establishing method combining intra-class similarity and inter-class difference |
| CN113542017A (en) * | 2021-07-01 | 2021-10-22 | 天翊谛晟(深圳)科技有限公司 | Network fault positioning method based on network topology and multiple indexes |
| CN116778210A (en) * | 2022-03-07 | 2023-09-19 | 奥图码数码科技(上海)有限公司 | Teaching image evaluation system and teaching image evaluation method |
-
2023
- 2023-10-27 CN CN202311408679.3A patent/CN117408829A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101546421A (en) * | 2009-04-01 | 2009-09-30 | 河北农业大学 | Province-wide comparison tillage quality evaluation method based on GIS |
| CN106778826A (en) * | 2016-11-25 | 2017-05-31 | 南昌航空大学 | Based on the hereditary Hybrid Clustering Algorithm with preferred Fuzzy C average of self adaptation cellular |
| CN109492796A (en) * | 2018-10-12 | 2019-03-19 | 东南大学 | A kind of Urban Spatial Morphology automatic Mesh Partition Method and system |
| CN110909792A (en) * | 2019-11-21 | 2020-03-24 | 安徽大学 | Clustering analysis method based on improved K-means algorithm and new clustering effectiveness index |
| CN111291326A (en) * | 2020-02-06 | 2020-06-16 | 武汉大学 | Clustering effectiveness index establishing method combining intra-class similarity and inter-class difference |
| CN113542017A (en) * | 2021-07-01 | 2021-10-22 | 天翊谛晟(深圳)科技有限公司 | Network fault positioning method based on network topology and multiple indexes |
| CN116778210A (en) * | 2022-03-07 | 2023-09-19 | 奥图码数码科技(上海)有限公司 | Teaching image evaluation system and teaching image evaluation method |
Non-Patent Citations (4)
| Title |
|---|
| 孙茂凯，等: "阜平县北流河流域林地多功能评价及分区研究", 《林业与生态科学》, vol. 38, no. 1, 31 January 2023 (2023-01-31), pages 39 - 47 * |
| 段桂芹: "基于改进密度的簇内均值最小距离聚类算法", 《智能计算机与应用》, vol. 11, no. 12, 31 December 2021 (2021-12-31), pages 83 - 86 * |
| 田鹏，等: "川中丘陵区耕地质量评价指标体系的优化与应用-以四川省遂宁市为例", 《亚热带农业研究》, vol. 18, no. 2, 31 May 2022 (2022-05-31), pages 124 - 130 * |
| 陈正发，等: "基于质量评价的省域尺度坡耕地质量调控体系构建", 《农业工程学报》, vol. 37, no. 20, 31 October 2021 (2021-10-31), pages 137 - 145 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gili et al. | Comparison of three methods for delineating management zones for site-specific crop management | |
| Kelly et al. | Landscape dynamics of the spread of sudden oak death | |
| Kitchen et al. | Delineating productivity zones on claypan soil fields using apparent soil electrical conductivity | |
| Moncada et al. | Data-driven analysis of soil quality indicators using limited data | |
| Van Ranst et al. | Application of fuzzy logic to land suitability for rubber production in peninsular Thailand | |
| Dengiz et al. | Multi-criteria approach with linear combination technique and analytical hierarchy process in land evaluation studies | |
| Schmidt et al. | Soil quality under conservation practices on farm operations of the southern semiarid pampas region of Argentina | |
| Deng et al. | GIS-based assessment of land suitability for alfalfa cultivation: A case study in the dry continental steppes of northern China | |
| CN112651548B (en) | Evaluation and identification method for plateau lakeside ecological landscape restoration planning | |
| Sieben et al. | The herbaceous vegetation of subtropical freshwater wetlands in South Africa: Classification, description and explanatory environmental factors | |
| Jiang et al. | Study on delineation of irrigation management zones based on management zone analyst software | |
| CN107608938B (en) | Factor screening method for binary classification based on enhanced regression tree algorithm | |
| Wang et al. | Determining the impact of key climatic factors on geographic distribution of wild Akebia trifoliate | |
| Roel et al. | Why do some farmers get higher yields? Multivariate analysis of a group of Uruguayan rice farmers | |
| CN117408829A (en) | Method for automatically inducing and diagnosing barrier factors in farmland protection partition and characteristics | |
| Heller et al. | An inductive approach to developing ecological site concepts with existing monitoring data | |
| CN116227692B (en) | Crop heavy metal enrichment risk quantification method, system and storable medium | |
| CN115310719B (en) | Farmland soil sampling scheme design method based on three-stage k-means | |
| Huang et al. | The vulnerability assessment method for Beijing Agricultural Drought | |
| Zhang et al. | Spatial variability of soil organic matter and total nitrogen and the influencing factors in Huzhu County of Qinghai Province, China | |
| Wright et al. | On the analysis of soil variability, with an example from Spain | |
| Chen et al. | Ecological land classification: A quantitative classification and ordination of forest communities adjacent to a rapidly expanding urban area in southeast coastal China | |
| Pottier et al. | Analysing the spatial heterogeneity of emergent groups to assess ecological restoration | |
| Wang et al. | Evaluation of the spatial pattern of surface soil water content of a karst hillslope in Southwest China using a state-space approach | |
| Li et al. | Gully erosion susceptibility maps and influence factor analysis in the Lhasa River Basin on the Tibetan Plateau, based on machine learning algorithms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |