CN113342938B - Corn seed selection and precise sowing decision-making method combining knowledge graph - Google Patents
Corn seed selection and precise sowing decision-making method combining knowledge graph Download PDFInfo
- Publication number
- CN113342938B CN113342938B CN202110684919.7A CN202110684919A CN113342938B CN 113342938 B CN113342938 B CN 113342938B CN 202110684919 A CN202110684919 A CN 202110684919A CN 113342938 B CN113342938 B CN 113342938B
- Authority
- CN
- China
- Prior art keywords
- corn
- sowing
- decision
- data
- level
- 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.)
- Active
Links
- 240000008042 Zea mays Species 0.000 title claims abstract description 114
- 235000002017 Zea mays subsp mays Nutrition 0.000 title claims abstract description 114
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 title claims abstract description 104
- 235000005822 corn Nutrition 0.000 title claims abstract description 104
- 238000009331 sowing Methods 0.000 title claims abstract description 81
- 238000000034 method Methods 0.000 title claims abstract description 40
- 239000002689 soil Substances 0.000 claims abstract description 37
- 235000015097 nutrients Nutrition 0.000 claims abstract description 29
- 238000004519 manufacturing process Methods 0.000 claims abstract description 14
- 238000005516 engineering process Methods 0.000 claims abstract description 13
- 238000012216 screening Methods 0.000 claims abstract description 6
- 230000008569 process Effects 0.000 claims abstract description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 27
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 claims description 24
- 239000003337 fertilizer Substances 0.000 claims description 21
- 238000010899 nucleation Methods 0.000 claims description 21
- 229910052757 nitrogen Inorganic materials 0.000 claims description 12
- XZKIHKMTEMTJQX-UHFFFAOYSA-N 4-Nitrophenyl Phosphate Chemical compound OP(O)(=O)OC1=CC=C([N+]([O-])=O)C=C1 XZKIHKMTEMTJQX-UHFFFAOYSA-N 0.000 claims description 11
- 241000196324 Embryophyta Species 0.000 claims description 11
- 238000012937 correction Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 9
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims description 8
- 230000035558 fertility Effects 0.000 claims description 8
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 claims description 7
- ZLMJMSJWJFRBEC-UHFFFAOYSA-N Potassium Chemical compound [K] ZLMJMSJWJFRBEC-UHFFFAOYSA-N 0.000 claims description 7
- 229910052698 phosphorus Inorganic materials 0.000 claims description 7
- 239000011574 phosphorus Substances 0.000 claims description 7
- 239000011591 potassium Substances 0.000 claims description 7
- 229910052700 potassium Inorganic materials 0.000 claims description 7
- 201000010099 disease Diseases 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 230000035784 germination Effects 0.000 claims description 5
- 239000013598 vector Substances 0.000 claims description 5
- 241000607479 Yersinia pestis Species 0.000 claims description 3
- 230000009471 action Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 230000004720 fertilization Effects 0.000 claims description 3
- 238000005469 granulation Methods 0.000 claims description 3
- 230000003179 granulation Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000013136 deep learning model Methods 0.000 claims description 2
- 238000005259 measurement Methods 0.000 claims description 2
- 238000007637 random forest analysis Methods 0.000 claims description 2
- 230000011218 segmentation Effects 0.000 claims description 2
- 238000012549 training Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims 1
- 239000005416 organic matter Substances 0.000 claims 1
- 235000016383 Zea mays subsp huehuetenangensis Nutrition 0.000 abstract description 10
- 235000009973 maize Nutrition 0.000 abstract description 10
- 238000004422 calculation algorithm Methods 0.000 abstract description 5
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 abstract 1
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 abstract 1
- 230000009286 beneficial effect Effects 0.000 abstract 1
- 230000003993 interaction Effects 0.000 abstract 1
- 235000013339 cereals Nutrition 0.000 description 13
- 238000010586 diagram Methods 0.000 description 7
- 238000012935 Averaging Methods 0.000 description 3
- 208000003643 Callosities Diseases 0.000 description 3
- 208000035240 Disease Resistance Diseases 0.000 description 3
- 206010020649 Hyperkeratosis Diseases 0.000 description 3
- 208000022602 disease susceptibility Diseases 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 230000035800 maturation Effects 0.000 description 3
- 208000035475 disorder Diseases 0.000 description 2
- 210000005069 ears Anatomy 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000001502 supplementing effect Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 239000003570 air Substances 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 235000021049 nutrient content Nutrition 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01G—HORTICULTURE; CULTIVATION OF VEGETABLES, FLOWERS, RICE, FRUIT, VINES, HOPS OR SEAWEED; FORESTRY; WATERING
- A01G22/00—Cultivation of specific crops or plants not otherwise provided for
- A01G22/20—Cereals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Animal Behavior & Ethology (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Botany (AREA)
- Environmental Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a corn seed selection and precise sowing decision-making method combining a knowledge graph, which comprises the following steps: step a), corn variety data are obtained through a web crawler, and variety characteristic information is extracted; step b), establishing a maize seed selection knowledge framework and constructing a knowledge graph; step c), primarily screening varieties according to sowing areas, and determining the varieties by utilizing a Markov distance algorithm according to soil nutrients; step d), acquiring sowing decision data through a GPS (global positioning system) and field acquisition technology; step e) dividing grids by utilizing Kerling interpolation and determining interpolation points; step f) determining grid sowing quantity according to a decision formula; step g) generating a shapefile format sowing quantity prescription chart. The beneficial effects are that: the method completely comprises the seed selection and sowing processes in the corn sowing decision, the seed selection decision is carried out on corn by constructing a knowledge graph, the corn sowing quantity is generated based on the interaction of farmland environment and variety characteristics, and the influence of production factors on the corn sowing decision is comprehensively utilized.
Description
Technical Field
The invention relates to the field of intelligent agriculture precise agriculture, in particular to a corn seed selection and precise sowing decision-making method combining a knowledge graph.
Background
The accurate seeding is a basic link of accurate agriculture, the adaptive planting variety is selected according to the agricultural production environment, and meanwhile, the optimal utilization of farmland related resources such as soil fertility, planting space, air illumination and the like is realized according to the established production planting model, so that the current optimum seeding amount of the farmland is achieved. However, the corn planting is not used for determining the planting variety according to the actual production conditions, the planting model is rough, the factors considered by the planting model are incomplete, and the decision result is not ideal.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a corn seed selection and precise sowing decision method combining a knowledge graph, which is realized by the following technical scheme:
the system comprises:
the corn seed selection decision system based on the knowledge graph outputs corn varieties suitable for planting by users;
the soil rapid measurement system outputs farmland water and fertilizer information;
the GPS positioning system outputs the boundary position coordinates of the land block;
the precise corn sowing decision making system is used for collecting the water and fertilizer information and the position coordinates through a 4G module, analyzing and processing the water and fertilizer information and the position coordinates and outputting a vector corn sowing quantity decision prescription chart;
the decision method for corn seed selection and accurate sowing by combining the knowledge graph specifically comprises the following steps:
obtaining corn variety structured and unstructured data through a web crawler, and establishing a corn seed information knowledge graph;
based on production environment factors and combining with a knowledge graph, outputting a corn variety suitable for planting by a user;
acquiring sowing decision data through a GPS (global positioning system) and field acquisition technology;
grid division of the land parcels is achieved by utilizing Kriging interpolation, and a certain number of points are generated in the grid;
determining the seeding quantity of each point according to a decision formula, and obtaining the seeding quantity of the grid by averaging the seeding quantities of all points in the grid;
and generating a sowing quantity prescription chart according to the shapefile format.
The process comprises:
step 1) obtaining corn variety structured data and unstructured data through a crawler technology, wherein the corn variety structured data and unstructured data comprise variety names, variety characteristics, planting areas, required soil nutrients, required weather environmental conditions, maturation period, general acre yield, general planting density, disease resistance, susceptibility and other information. Selecting neo4j database as non-relational database and establishing maize variety knowledge map by semi-automatization method;
step 2) based on the corn seed information knowledge graph, primarily screening suitable seeds according to the easily-developed conditions of the sowing area, and then selecting corn seeds which are most suitable for sowing according to soil nutrients by utilizing a mahalanobis distance algorithm to finish corn seed selection and decision making;
step 3) grid division is carried out on the water and fertilizer information and the position coordinates by utilizing a Kriging interpolation method, and grid division data are output;
step 4) constructing a corn precise sowing decision model, carrying out sowing decision on the water and fertilizer information, the position coordinates and the grid data, and outputting corn sowing decision data of the grid;
and 5) processing the grid sowing decision data according to the shapefile format, and outputting a corn sowing quantity decision prescription chart.
The corn seed selection decision method combining the knowledge patterns comprises the steps of obtaining and processing corn seed information in the step 1) and constructing the knowledge patterns, and the corn seed selection decision based on the knowledge patterns in the step 2).
The step 1) specifically comprises the following steps:
step 1-1) obtaining corn seed structured data and unstructured data through a web crawler technology. And converting the structured data into multi-element group data by using tools such as D2RD, virtuoso, MOrph, storing the multi-element group data into a csv file, and storing the multi-element group data into a neo4j graph database by a load csv method.
Step 1-2) training the unstructured data by using a skip gram model in a deep learning model to generate word vectors as an agricultural corn word stock;
step 1-3), a knowledge framework of corn seed selection is established, the data are segmented by utilizing a Jieba word segmentation tool and an agricultural corn word stock, and entity category classification is carried out on the segmented words by utilizing a random forest model member named entity recognition classifier. And extracting the relation of the data through a defined relation template, converting the data into multi-element group data by combining the entity, storing the data into a csv file, and storing the data into a neo4j graph database through a load csv method, namely supplementing the entity and the relation into the knowledge frame to form a complete maize variety knowledge graph.
The step 2) specifically comprises the following steps:
step 2-1), preliminarily screening proper seeds according to the seeding area and the area easy-to-develop diseases by utilizing a Cypher statement to obtain a seed list;
step 2-2) carrying out similarity calculation on soil nutrients through a mahalanobis distance on the corn variety list recommended in the step 2-1), wherein the mahalanobis distance formula is as follows:
wherein Σ is the covariance matrix of the multidimensional random variable, x is the soil nutrient, y i Is soil nutrient of different varieties in the knowledge graph. And finally obtaining the minimum value of the Male distance as the decision recommended corn variety.
The decision method for precisely sowing the corn is characterized in that the decision model for precisely sowing the corn in the step 4) outputs decision sowing quantity after the water and fertilizer information, the position coordinates and the grid data are subjected to model operation.
The step 4) specifically comprises the following steps:
step 4-1) utilizing the nearly three years of yield of the sowing land and combining the ecological environment of the land and the production technology level of farmers to establish a proper target yield AMY (kg/ha) for the farmers, namely:
AMY＝AVY×(1+YIC) (1)
W＝(R-R mean )/R mean (4)
T＝(T max +T min )/2 (6)
NSL＝0.67×SFL+0.33×FML (7)
wherein AVY is the average yield (kg/ha) of the sowing plots for nearly three years; YIC is yield coefficient, maxY is historical highest yield of the sowing land, f (w) is moisture correction function, f (T) is temperature correction function, NSL is nutrient supply level, PTL is production technology level; w is the relative amount of water supply, R is the rainfall in the current year, R mean Average rainfall over many years; t is the average air temperature of 15 days before and after sowing day, T max Is the highest temperature of the day, T min The day minimum air temperature; SFL is soil fertility level and FML is fertilization management level.
PTL＝(CML+PCL)/2 (14)
The model quantifies nitrogen, phosphorus and potassium and organic matters which have great influence on the growth of corns in soil, namely f (AN) is the content level of quick-acting nitrogen (in soil fertility indexes, the quick-acting nitrogen content is better than 150 mg/kg), f (AP) is the content level of quick-acting phosphorus (the content is better than 40 mg/kg), f (AK) is the content level of quick-acting potassium (the content is better than 200 mg/kg), f (TN) is the content level of total nitrogen (the content is better than 2 g/kg), f (SOM) is the content level of organic matters (the content is better than 40 g/kg), and the action proportion of each nutrient parameter on the growth of corns is equal; CML is the cultivation technical level, PCL is the disease and pest control level, FML, CML, PCL is divided into low, medium and high according to the actual management technical level of farmers and local areas, and the values are 0.6, 0.8 and 1.0 in sequence.
Step 4-2) the model determines the planting density SPD (ten thousand plants/ha) of the corn according to the basic principle of 'to produce fixed ears and to fix plants', and the planting density SPD is expressed as follows:
SPD＝SPNP/PNPP (1)
PNPP＝1+DER (3)
wherein SPNP is the suitable spike number (wan spike/ha) per unit area, PNPP is the effective spike number (spike/strain) of a single plant, and 10 5 Is a unit conversion coefficient; VSGN is the number of grains per spike (grains/spike); the VHGW is hundred-grain weight (g/hundred grains); DER is the double spike rate (%); TLI is the effective granulation (%), WSL is the moisture supply level equivalent to the moisture correction function f (w), NSL is the nutrient supply level.
On the basis of the proper planting density, the sowing quantity SQP (kg/ha) in unit area is determined according to the pit number N (grains/pit), the hundred grain weight VHGW, the planting area S (ha) and the germination rate SGR (%), and is expressed as follows:
SQP＝SPD×N×VHGW×S×10 -5 /SGR (5)
wherein, the number N of the holes is within 1-2, and the holes are selected according to actual conditions of farmers and local places; considering that the shape of a sowing land block is irregular in actual sowing, the situation that the grid sizes are inconsistent can occur when grid division is carried out, and 1ha is taken from the planting area S for being convenient for a sowing machine to read.
The invention has the following advantages:
(1) A corn variety selection method combining with a knowledge graph is provided. Selecting neo4j database as non-relational database and establishing maize variety knowledge map by semi-automatization method; the corn variety structured data and unstructured data are obtained through a crawler technology, wherein the corn variety structured data and unstructured data comprise variety names, variety characteristics, planting areas, required soil nutrients, required weather environmental conditions, maturation period, general acre yield, general planting density, disease resistance, susceptibility and other information. Establishing a maize variety knowledge graph, primarily screening suitable seeds according to the easily-developed conditions of the sowing area, and then selecting the maize seeds which are most suitable for sowing according to soil nutrients by utilizing a mahalanobis distance algorithm so as to achieve suitable land.
(2) The corn sowing planting model of the predecessor is optimized and is reconstructed into a corn precise sowing decision model. The contents of nitrogen, phosphorus, potassium, organic matters and the like in farmland soil water and fertilizer are quantized and subdivided according to the first-level proportion in the grading index by utilizing the soil nutrient content index, so that a fuzzy value taking method for the soil water and fertilizer content condition in the original corn sowing planting model is replaced, and the accurate corn sowing decision is more accurate; the Kriging interpolation algorithm in the spatial interpolation algorithm is selected to realize the interpolation and the acquisition of important data required by decisions such as farmland plot soil water and fertilizer content, historical yield and the like, provides relevant decision data of each grid for the precise seeding decision model, and ensures the realization of variable seeding in precise seeding decisions of corns.
Drawings
Fig. 1 is a schematic diagram of a maize seed selection knowledge framework.
Fig. 2 is a schematic diagram of a corn seed selection knowledge graph.
FIG. 3 is a decision tree diagram of corn sowing quantity
FIG. 4 is a grid partition and prescription chart of a precise seeding decision example
Detailed Description
The present application is further described below with reference to the accompanying drawings.
The invention discloses a corn seed selection and accurate sowing decision method combining a knowledge graph, which comprises corn variety selection and accurate sowing decision combining the knowledge graph. The knowledge map is established by integrating the knowledge of varieties, soil nutrients, planting areas and the like, and the optimal planting varieties in the areas are provided for corn growers on the basis of the knowledge map; the seeding decision data are obtained through GPS and field acquisition technology, then grid division of land parcels is achieved through Kriging interpolation, a certain number of points are generated in the grids, the seeding quantity of each point is determined according to a decision formula, the seeding quantity of the grids is obtained by averaging the seeding quantity of all the points in the grids, and finally a seeding quantity prescription diagram is generated according to shapefile format.
Specifically, the method for selecting corn varieties based on the knowledge-graph in one embodiment of the planting of the Huchang corn specifically comprises the following steps:
step 1) obtaining corn variety structured data and unstructured data in hundred degrees encyclopedia, 114 seed nets and the like through a crawler technology, wherein the corn variety structured data and unstructured data comprise variety names, variety characteristics, planting areas, required soil nutrients, required weather environmental conditions, a maturation period, general acre yield, general planting density, disease resistance, susceptibility and other information.
The neo4j database is selected as a non-relational database, and a maize variety knowledge graph is established through a semi-automatization method.
In addition, in one embodiment, it should be further noted that the selecting the neo4j database as the non-relational database and establishing the maize variety knowledge map by the semi-automatic method includes: establishing a knowledge framework of corn seed selection, and forming a corn variety knowledge body as shown in fig. 1; automatically extracting entities and relations in the knowledge frame through natural language processing and a machine learning method; and generating a CSV document from the acquired entity and relationship data, supplementing the entity and relationship into the knowledge frame through a load CSV method, and forming a complete maize variety knowledge graph as shown in fig. 2.
Step 2) performing corn variety selection based on the corn variety knowledge graph. Specifically, in one embodiment, performing corn variety selection based on a corn variety knowledge map includes:
first pass Cypher statement in Neo4j
MATCH (n: region { name: ' planting region ' }) < - [: ' suitable planting region ] - (m)
RETURN m
Inputting a planting area, and returning to a proper variety list of the area;
then pass through
MATCH (m) - [: resistance ] - > (k: "disorder" { name: 'disorder' })
RETURN k
Inputting the recent easy-to-develop diseases in the planting area, screening the variety list, and returning to a recommended corn variety list;
and calculating the similarity of the recommended corn variety list to soil nutrients through the mahalanobis distance, wherein the mahalanobis distance formula is as follows:
wherein Σ is the covariance matrix of the multidimensional random variable, x is the soil nutrient, y i Is soil nutrient of different varieties in the knowledge graph.
And finally, obtaining the minimum value of the Male distance as a decision recommended corn variety, and finally selecting a relaxation order 9953 as the recommended variety.
And 3) acquiring relevant data required by decision making, wherein the soil water and fertilizer data comprise nitrogen, phosphorus and potassium in soil, organic matters, pH value and water content, and the soil water and fertilizer data are acquired by using a field sampling mode. And acquiring water and fertilizer data of a limited point, and obtaining soil water and fertilizer data distributed in the whole land by using a spatial interpolation method.
The farmland nutrient data is interpolated by using a kriging interpolation method in the example. The method comprises the following specific steps:
step 3-1), coordinate data of boundary points of farmland plots are obtained through GPS and other equipment, and an approximate shape of the farmland plots is formed in a system map;
step 3-2) acquiring nutrient data of limited points in farmland plots and corresponding coordinate data thereof by utilizing a certain adoption rule and a traditional field sampling technology;
step 3-3) carrying out grid division on farmlands according to the operation requirements of users and the actual operation conditions to obtain coordinate data of all vertexes of each grid;
step 3-4) performing Kriging interpolation according to the coordinate data of each grid vertex and farmland nutrient data to obtain interpolation data of a certain number of points in each grid, and averaging the interpolation data in each grid to obtain a nutrient reference value of each land.
Step 4) according to the constructed accurate corn sowing decision model, sowing decisions are carried out on the water and fertilizer information, the position coordinates and the grid data, and corn sowing quantity decision data of the grid are output;
as shown in fig. 3, the main steps of the corn sowing decision model are as follows:
step 4-1) determining the target yield, wherein the seeding model is the target yield of the land block to be established first, and the determination of the target yield is required to be established jointly through the annual yield, the planting environment and the production level of the land block. The annual yield of the land parcels, particularly the yield of the land parcels of nearly three years, can indirectly display the production potential of the current land parcels, and has important reference significance for establishing a seeding model and establishing target yield. The planting environment of the land parcels mainly comprises factors such as soil water and fertilizer, air temperature and the like; the production level mainly refers to the production planting level of farmers. The formula of the target yield AMY (kg/ha) is:
AMY＝AVY×(1+YIC) (1)
W＝(R-R mean )/R mean (4)
T＝(T max +T min )/2 (6)
NSL＝0.67×SFL+0.33×FML (7)
PTL＝(CML+PCL)/2 (14)
wherein AVY is the average yield (kg/ha) of the sowing plots for nearly three years; YIC is a yield increase coefficient, maxY is the historical highest yield of a sowing land, AVY is the average yield (kg/ha) of nearly three years, f (W) is a moisture correction function, f (T) is a temperature correction function, NSL is a nutrient supply level, and PTL is a production technology level; w is the relative amount of water supply, R is the rainfall in the current year, R mean Average rainfall over many years; t is the average air temperature of 15 days before and after sowing day, T max Is the highest temperature of the day, T min The day minimum air temperature; SFL is soil fertility level, FML is fertilization management level; sowing model of the former [13-14 ]]In the process, SFL is taken more generally, and is mainly taken according to personal subjectivity, and certain uncertainty exists, wherein nitrogen, phosphorus, potassium and organic matters with great influence on corn growth in soil are quantified, namely f (AN) is quick-acting nitrogen (soil fertility index [ 16)]Wherein the content of quick-acting nitrogen is 150mg/kg, the content of f (AP) is the content of quick-acting phosphorus (the content is 40mg/kg, the content of f (AK) is the content of quick-acting potassium (the content is 200mg/kg, the content of f (TN) is the content of total nitrogen (the content is 2g/kg, the content of f (SOM) is the content of organic matters (the content is 40g/kg, the preferred content) and the action proportion of each nutrient parameter to the corn growth is 1, so as to take the value of SFL; CML is the cultivation technical level, PCL is the disease and pest control level, FML, CML, PCL is divided into low, medium and high according to the actual management technical level of farmers and local areas, and the values are 0.6, 0.8 and 1.0 in sequence.
Step 4-2), determining the seeding quantity, namely determining the planting density firstly, wherein the proper planting density can well coordinate productivity contradiction between corn single plants and populations, and the overall production potential of corn is exerted to the greatest extent, so that the optimal yield is obtained. Too large or too small a planting density will have a certain effect on the yield: the excessive planting density can lead to the reduction of the final yield of insufficient corn productivity in unit area, and simultaneously wastes a great amount of manpower and material resources, and the input cost is inversely proportional to the income; too small planting density can cause insufficient productivity of corn single plants in unit area, so that optimal utilization of land parcels can not be realized, and the final yield is indirectly reduced. The planting density SPD (ten thousand plants/ha) of the corn is determined according to the basic principle of 'to produce fixed ears and to spike and fix plants', and is expressed as follows:
SPD＝SPNP/PNPP (15)
PNPP＝1+DER (17)
wherein SPNP is the suitable spike number (wan spike/ha) per unit area, PNPP is the effective spike number (spike/strain) of a single plant, and 10 5 Is a unit conversion coefficient; VSGN is the number of grains per spike (grains/spike); the VHGW is hundred-grain weight (g/hundred grains); DER is the double spike rate (%); TLI is the effective granulation (%), WSL is the moisture supply level equivalent to the moisture correction function f (W), NSL is the nutrient supply level.
On the basis of the proper planting density, the sowing quantity SQP (kg/ha) in unit area is determined according to the pit number N (grains/pit), the hundred grain weight VHGW, the planting area S (ha) and the germination rate SGR (%), and is expressed as follows:
SQP＝SPD×N×VHGW×S×10 -5 /SGR (19)
wherein, the number N of the holes is within 1-2, and the holes are selected according to actual conditions of farmers and local places; considering the irregular shape of the sowing land block during actual sowing, the inconsistent grid size can occur during grid division
For the convenience of the seeder to read, the planting area S is 1ha.
Finally, the corn variety Yudan 9953 has 544.3 grains/spike, the hundred grains weight of 30.23 g/hundred grains, the double spike rate of 1.45 percent and the germination rate of 86.7 percent. Historical yield calculation formula for each grid:
AVY i ＝AVY(SFL i -SFL+1) (20)
wherein AVY is the average yield (kg/ha) of the whole sowing plot for nearly three years; AVY (Audio video Y) i (i=1, 2, 3.) is the average yield of each grid over the last three years; SFL is the average soil fertility level of the entire sown plot; SFL (Small form-factor pluggable) i Soil fertility level for each grid.
The total area of the planting land for the many and the chang is 30 mu, the whole land is divided into 12 grids of 4 rows and 3 columns according to the needs of farmers and the actual local situation, the grid division form is shown in fig. 4, the specific seeding amount of each grid is shown in table 1, and all grids are single-seed seeding.
Table 1 Xu Chang seed quantity decision example
Step 5) in the decision method, the sowing prescription data generated by the decision model are utilized to generate a sowing prescription diagram in a shape format by utilizing shape and shape libraries of python, and the prescription diagram is transmitted to the precision planter in an on-line or off-line mode. Meanwhile, the sowing prescription data can be displayed in the form of a block thermodynamic diagram in the decision model system for users to intuitively understand the sowing data of the whole land block.
The present invention is not limited to the above-mentioned embodiments, and any changes or substitutions that can be easily understood by those skilled in the art within the technical scope of the present invention are intended to be included in the scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.
Claims (6)
1. A corn seed selection and accurate sowing decision-making method combining a knowledge graph includes:
the corn seed selection decision system based on the knowledge graph outputs corn varieties suitable for planting by users;
the soil rapid measurement system outputs farmland water and fertilizer information;
the GPS positioning system outputs the boundary position coordinates of the land block;
the precise corn sowing decision making system is used for collecting the water and fertilizer information and the position coordinates through a 4G module, analyzing and processing the water and fertilizer information and the position coordinates and outputting a vector corn sowing quantity decision prescription chart;
the decision method comprises the following specific processes:
step 1) extracting and storing corn seed information on a network and a literature to construct a knowledge graph;
step 2) based on the knowledge graph, making a decision on corn seed selection;
step 3) grid division is carried out on the water and fertilizer information and the position coordinates by utilizing a Kriging interpolation method, and grid division data are output;
step 4) constructing a corn precise sowing decision model, carrying out sowing decision on the water and fertilizer information, the position coordinates and the grid data, and outputting corn sowing decision data of the grid;
and 5) processing the corn sowing decision data of the grid according to the shapefile format, and outputting a vector corn sowing quantity decision prescription chart.
2. The corn seed selection and precise sowing decision-making method combining the knowledge patterns according to claim 1 is characterized in that in the step 1), the whole corn variety knowledge patterns are constructed by using neo4j through processing structured and unstructured corn data acquired by a crawler, and in the step 2), the corn seed selection decision-making is carried out based on the knowledge patterns.
3. The method for selecting corn seeds and accurately sowing decision-making by combining knowledge-based maps according to claim 2, wherein the step 1) specifically comprises the following steps:
step 1-1) converting the structured data into multi-element group data by using a tool, storing the multi-element group data into a csv file, and storing the multi-element group data into a neo4j graph database by a load csv method; the tool includes Virtuoso, MOrph;
step 1-2) training the unstructured data by using a skip gram model in a deep learning model to generate word vectors as an agricultural corn word stock;
step 1-3) utilizing a Jieba word segmentation tool and an agricultural corn word stock to segment unstructured data, and utilizing a random forest model member named entity recognition classifier to classify entity categories of the segmented words; and carrying out relation extraction on unstructured data through a defined relation template, converting the unstructured data into multi-element data by combining an entity, storing the data into a csv file, and storing the data into a neo4j graph database through a load csv method.
4. The method for selecting corn seeds and accurately sowing decision combined with the knowledge graph according to claim 2, wherein the step 2) specifically comprises the following steps:
step 2-1), preliminarily screening proper seeds according to the seeding area and the area easy-to-develop diseases by utilizing a Cypher statement to obtain a seed list;
step 2-2) carrying out similarity calculation on soil nutrients through a mahalanobis distance on the corn variety list recommended in the step 2-1), wherein the mahalanobis distance formula is as follows:
wherein Σ is the covariance matrix of the multidimensional random variable, x is the soil nutrient, y i Soil nutrients of different varieties in the knowledge graph; and finally obtaining the minimum value of the Male distance as the decision recommended corn variety.
5. The method for selecting corn seeds and accurately sowing decision-making combined with the knowledge graph according to claim 1, wherein the method is characterized in that the accurate corn sowing decision-making model in the step 4) outputs decision-making sowing quantity after model operation on the water and fertilizer information corresponding to the grids divided in the step 3);
the step 4) specifically comprises the following steps:
step 4-1) utilizing the nearly three years of yield of the sowing land and combining the ecological environment of the land and the production technology level of farmers to establish a proper target yield AMY for the farmers, wherein the unit is kg/ha, namely:
AMY＝AVY×(1+YIC)
W＝(R-R mean )/R mean
T＝(T max +T min )/2
NSL＝0.67×SFL+0.33×FML
wherein AVY is the average yield of the sowing plots for nearly three years, and the unit is kg/ha; YIC is yield coefficient, maxY is historical highest yield of the sowing land, f (W) is moisture correction function, f (T) is temperature correction function, NSL is nutrient supply level, PTL is production technology level; w is the relative amount of water supply, R is the rainfall in the current year, R mean Average rainfall over many years; t is the average air temperature of 15 days before and after sowing day, T max Is the highest temperature of the day, T min The day minimum air temperature; SFL is soil fertility level, FML is fertilization management level, FML is divided according to actual management technical level of farmers and local areas;
SFL is obtained by the formula:
f (AN) is the quick-acting nitrogen content level, f (AP) is the quick-acting phosphorus content level, f (AK) is the quick-acting potassium content level, f (TN) is the total nitrogen content level, f (SOM) is the organic matter content level, and the action proportion of each nutrient parameter on the corn growth is equal;
PTL is obtained by the following formula:
PTL＝(CML+PCL)/2
CML is the cultivation technical level, PCL is the disease and pest control level, and the CML and PCL are divided according to the actual management technical level of farmers and local areas;
step 4-2) determining the planting density SPD of the corn, wherein the unit of the SPD is ten thousand plants/ha, and the SPD is expressed as follows by a formula:
SPD＝SPNP/PNPP
PNPP＝1+DER
wherein SPNP is the suitable spike number per unit area and the unit is ten thousand spikes/ha; PNPP is the effective spike number of a single plant, and the unit spike/plant; 10 5 Is a unit conversion coefficient; VSGN is the number of grains per spike, grain/spike; the VHGW is hundred-grain weight, g/hundred-grain; DER is the double spike rate; TLI is effective granulation rate, which is obtained by Chinese seed information network; WSL is the water supply level equivalent water correction function f (w), NSL is the nutrient supply level;
selecting a grid with proper planting density SPD, and determining a unit area sowing quantity SQP according to the hole number N, the hundred-grain weight VHGW, the planting area S and the germination rate SGR, wherein the unit kg/ha is a unit; expressed by the formula:
SQP＝SPD×N×VHGW×S×10 -5 /SGR
in the formula, the number N of the hole grains and the planting area S are determined by actual planting conditions; hundred-grain weight VHGW and germination rate SGR are all obtained by Chinese seed information network.
6. The method for selecting corn seeds and accurately sowing decision combined with the knowledge graph according to claim 5, wherein the method is characterized by comprising the following steps:
f (AN), f (AP), f (AK), f (TN), f (SOM) are obtained by the following formulas, respectively:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110684919.7A CN113342938B (en) | 2021-06-21 | 2021-06-21 | Corn seed selection and precise sowing decision-making method combining knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110684919.7A CN113342938B (en) | 2021-06-21 | 2021-06-21 | Corn seed selection and precise sowing decision-making method combining knowledge graph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113342938A CN113342938A (en) | 2021-09-03 |
| CN113342938B true CN113342938B (en) | 2024-03-08 |
Family
ID=77478424
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110684919.7A Active CN113342938B (en) | 2021-06-21 | 2021-06-21 | Corn seed selection and precise sowing decision-making method combining knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113342938B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114467607B (en) * | 2022-02-08 | 2023-01-24 | 新疆维吾尔自治区阿克苏地区林业和草原局 | Greening planting method for oversized image-text model in arid region |
| CN115062778B (en) * | 2022-07-15 | 2024-04-23 | 中国农业大学 | Corn sowing quantity decision method based on soil organic matter content |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107832892A (en) * | 2017-11-17 | 2018-03-23 | 北京大学 | A kind of lumpily seed selection decision optimization method based on Combinatorial Optimization |
| CN111949828A (en) * | 2019-05-17 | 2020-11-17 | 南京农业大学 | Automatic construction method of rice cultivation mode knowledge graph facing semi-structured form |
| CN112115379A (en) * | 2020-09-18 | 2020-12-22 | 吉林农业大学 | Rice variety selection method and device based on knowledge graph |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20200052448A (en) * | 2018-10-30 | 2020-05-15 | 삼성전자주식회사 | System and method for integrating databases based on knowledge graph |
-
2021
- 2021-06-21 CN CN202110684919.7A patent/CN113342938B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107832892A (en) * | 2017-11-17 | 2018-03-23 | 北京大学 | A kind of lumpily seed selection decision optimization method based on Combinatorial Optimization |
| CN111949828A (en) * | 2019-05-17 | 2020-11-17 | 南京农业大学 | Automatic construction method of rice cultivation mode knowledge graph facing semi-structured form |
| CN112115379A (en) * | 2020-09-18 | 2020-12-22 | 吉林农业大学 | Rice variety selection method and device based on knowledge graph |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113342938A (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11726452B2 (en) | System and method for aggregating test plot results based on agronomic environments | |
| Vanlauwe et al. | Beyond averages: new approaches to understand heterogeneity and risk of technology success or failure in smallholder farming | |
| Liu et al. | A neural network for setting target corn yields | |
| US20200250593A1 (en) | Yield estimation in the cultivation of crop plants | |
| Timsina et al. | Improved nutrient management in cereals using Nutrient Expert and machine learning tools: Productivity, profitability and nutrient use efficiency | |
| CN113342938B (en) | Corn seed selection and precise sowing decision-making method combining knowledge graph | |
| CA2663917A1 (en) | Variable zone crop-specific inputs prescription method and systems therefor | |
| Srivastava et al. | Effect of sowing date distributions on simulation of maize yields at regional scale–A case study in Central Ghana, West Africa | |
| US20200245525A1 (en) | Yield estimation in the cultivation of crop plants | |
| CN111949828A (en) | Automatic construction method of rice cultivation mode knowledge graph facing semi-structured form | |
| CN111489092A (en) | Method and system for evaluating suitable growing area of plant cultivation and planting environment | |
| Tatsumi | Effects of automatic multi-objective optimization of crop models on corn yield reproducibility in the USA | |
| Bongiovanni | A spatial econometric approach to the economics of site-specific nitrogen management in corn production | |
| Samsuri et al. | Logistic regression model on land productivity of Pagar Alam coffee farming | |
| Meena et al. | Fertilizer Recommendations by Using Different Geospatial Technologies in Precision Farming or Nanotechnology | |
| Rao et al. | Crop yield prediction by using machine learning techniques | |
| Dua et al. | Simulation to support evaluation of the production potential of rice varieties in tropical climates | |
| Wu et al. | An integrated model to simulate sown area changes for major crops at a global scale | |
| Ed-Daoudi et al. | A Machine Learning Approach to Identify Optimal Cultivation Practices for Sustainable apple Production in Precision Agriculture in Morocco | |
| Booltink et al. | Application of simulation models and weather generators to optimize farm management strategies | |
| Akinsunmade et al. | Land suitability and crop pattern model using integrated pollination intelligence algorithm and remote sensing | |
| Satizábal et al. | Enhancing decision-making processes of small farmers in tropical crops by means of machine learning models | |
| Jiang et al. | A Comparison of Supervised Machine Learning Algorithms for Predicting Subfield Yield Variability of Maize Grain | |
| Sarron et al. | Assessing production gaps at the tree scale: definition and application to mango (Mangifera indica L.) in West Africa | |
| Muraru et al. | Development of software for optimizing the fertilization of agricultural crop |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |