CN107665244A - A kind of method and device for obtaining relation between grain yield and fertilizer application amount - Google Patents
A kind of method and device for obtaining relation between grain yield and fertilizer application amount Download PDFInfo
- Publication number
- CN107665244A CN107665244A CN201710816507.8A CN201710816507A CN107665244A CN 107665244 A CN107665244 A CN 107665244A CN 201710816507 A CN201710816507 A CN 201710816507A CN 107665244 A CN107665244 A CN 107665244A
- Authority
- CN
- China
- Prior art keywords
- sample point
- rbf neural
- grain yield
- fertilizer application
- application amount
- 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.)
- Granted
Links
- 239000003337 fertilizer Substances 0.000 title claims abstract description 104
- 238000000034 method Methods 0.000 title claims abstract description 44
- 230000001537 neural effect Effects 0.000 claims abstract description 106
- 239000002689 soil Substances 0.000 claims abstract description 45
- 235000021049 nutrient content Nutrition 0.000 claims abstract description 44
- 238000005457 optimization Methods 0.000 claims abstract description 18
- 241001269238 Data Species 0.000 claims abstract description 12
- 238000003860 storage Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 5
- 230000001052 transient effect Effects 0.000 claims description 5
- 239000003016 pheromone Substances 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 230000005055 memory storage Effects 0.000 claims description 2
- 210000004218 nerve net Anatomy 0.000 claims description 2
- 230000008043 neural expression Effects 0.000 abstract description 3
- 239000000523 sample Substances 0.000 description 123
- 239000000126 substance Substances 0.000 description 28
- 238000003062 neural network model Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 8
- 210000002569 neuron Anatomy 0.000 description 7
- 230000007935 neutral effect Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 210000005036 nerve Anatomy 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- OAICVXFJPJFONN-UHFFFAOYSA-N Phosphorus Chemical compound [P] OAICVXFJPJFONN-UHFFFAOYSA-N 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000019637 foraging behavior Effects 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000011664 nicotinic acid Substances 0.000 description 2
- 235000015097 nutrients Nutrition 0.000 description 2
- 238000013488 ordinary least square regression Methods 0.000 description 2
- 239000005416 organic matter Substances 0.000 description 2
- 229910052698 phosphorus Inorganic materials 0.000 description 2
- 239000011574 phosphorus Substances 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- ZLMJMSJWJFRBEC-UHFFFAOYSA-N Potassium Chemical compound [K] ZLMJMSJWJFRBEC-UHFFFAOYSA-N 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 239000011591 potassium Substances 0.000 description 1
- 229910052700 potassium Inorganic materials 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training 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
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Marine Sciences & Fisheries (AREA)
- Animal Husbandry (AREA)
- Mining & Mineral Resources (AREA)
- Agronomy & Crop Science (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Fertilizers (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of method and device for obtaining relation between grain yield and fertilizer application amount, and this method includes：S1, a variety of input datas are obtained according to the unit area fertilizer application amount of each sample point, soil pH and each nutrient content, output data is obtained according to the unit area grain yield in city where each sample point and each sample point, the RBF neural of multiple ant group algorithms optimizations is built according to each input data and output data；S2, the mean square error of each RBF neural is obtained, the relation between grain yield and fertilizer application amount is represented using the minimum RBF neural of mean square error.By the present invention in that the RBF neural of multiple ant group algorithms optimizations is built with a variety of input datas, the minimum RBF neural of mean square error is therefrom selected to represent the relation between grain yield and fertilizer application amount, so as to realize that, using the relation between more accurate RBF neural expression grain yield and fertilizer application amount, the relation between the grain yield and fertilizer application amount of acquisition is also more accurate.
Description
Technical field
The present invention relates to deep learning field, is obtained more particularly, to one kind between grain yield and fertilizer application amount
The method and device of relation.
Background technology
Nowadays, in order to which the sustainability of ecosystem development, researcher are probing into applying quantity of chemical fertilizer and grain yield
Between relation in terms of done many work, such as using homing method build regression model.Wang Qi in 2013 et al. is used
OLS (Ordinary Least Square, common least square method) method is returned, and analyzes China's grain yield with changing
Quantitative relationship between fertile usage amount；Zhao Zhi heavily fortified points in 2012 et al. obtain the chemical fertilizer input and grain in Hunan Province using metering model
Relation between output；Old seedling in 2013 grade people's utilization space panel econometric model inquire into Grain in China yield influence
Factor, each dominant type study of grain yield fluctuation reason in the whole nation and its action rule are disclosed, aid in determining whether that each region grain increases
The main restricting factor and Promoting Approach of production；Gu Lemin in 2013 is carried out by least absolute deviation to exponential type production function parameters
Fitting, find some rules for meeting Grain in China change；Zhao sea English in 2013 et al. is with Eviews6.0 softwares to Chinese grain
Correlation between food yield and applying quantity of chemical fertilizer has carried out linear regression analysis, establishes Linear Regression Model in One Unknown；2016
Year Lang Gui flies et al. based on VAR (Vector Autoregression, vector auto regression) model to China's grain yield and chemical fertilizer
Amount of application has carried out proof analysis.
At present, neutral net is widely used in fields such as signal transacting, feature extraction, pattern-recognition, nonlinear predictions.No
Same neutral net is applied to different fields.Application at this stage is the most frequently BP (Back Propagation, backward biography
Broadcast) neutral net, but the convergence of BP neural network learning process and initial value are closely related, and learning process is it is also possible to go out
Existing local convergence, this is knotty problem in actual applications.But RBF (Radial Basis Function, radial direction base letter
Number) neutral net is a kind of feed-forward type neutral net of good performance, it is that the external world is reacted based on the neuronal cell of human brain
Locality and propose, be a kind of novel effective feed-forward type neutral net, and there is higher arithmetic speed.RBF nerves
Network generally has the function of three layers of BP neural network, but 10^3~10^4 times faster than BP method of pace of learning.Particularly it has
There is stronger non-linear mapping capability, a nonlinear function can be approached with the arbitrary accuracy overall situation.Bavin outstanding person in 2002 et al. probes into
Advantage of the RBF neural in terms of Function approximation capabilities.By means of spy of the RBF neural in terms of Function approximation capabilities
Property, RBF neural is widely used in fitting prediction.Wang Xinjun in 2008 et al. is using RBF neural to insurance wealth
Production loss distribution carries out the measuring and calculating of estimation and future property loss.
In the prior art, on the one hand, it is few by means of neural network model between grain yield and applying quantity of chemical fertilizer
Relation studied, and then contribute to the sustainable development of grain yield；On the other hand, in structure RBF neural network model
When, RBF models are built merely with single input or multi input, and the RBF nerves that single input and multi input are built in varied situations
The precision of network is different, i.e., single input is higher than the precision for the RBF neural that multi input is built under certain situation, and other
In the case of, multi input is higher than the precision for the RBF neural that single input is built.Therefore, built merely with single input or multi input
RBF neural network model, the precision of RBF neural network model can be influenceed.
The content of the invention
It is above-mentioned using only the problem of single input or multi input structure RBF model influence precision or at least in part to overcome
Solve the above problems, the invention provides a kind of method and device for obtaining relation between grain yield and fertilizer application amount.
According to the first aspect of the invention, there is provided a kind of method for obtaining relation between grain yield and fertilizer application amount,
Including：
S1, a variety of input numbers are obtained according to the unit area fertilizer application amount of each sample point, soil pH and each nutrient content
According to according to the unit area grain yield acquisition output data in city where each sample point and each sample point, according to each defeated
Enter data and output data builds the RBF neural of multiple ant group algorithm optimizations；
S2, the mean square error of each RBF neural is obtained, represented using the minimum RBF neural of mean square error
Relation between grain yield and fertilizer application amount.
Specifically, the step S1 is specifically included：
Using the unit area fertilizer application amount of each sample point as input data, the unit area of each sample point
Grain yield builds the first RBF neural as output data；
Using the soil pH, each nutrient content and unit area fertilizer application amount of each sample point as input data, institute
The unit area grain yield of each sample point is stated as output data, builds the second RBF neural；
Each sample point is clustered according to the soil pH of each sample point and nutrient content, in all kinds of
The sample point, the 3rd RBF neural and the 4th RBF neural are built respectively.
Specifically, according to the sample point in all kinds of, the 3rd RBF neural and the 4th RBF god are built respectively
Specifically included through the step of network：
The pH average values of all sample point soil, each nutrient content average value and the unit in place city in will be all kinds of
The average value of area fertilizer application amount is as input data, the unit area grain in city where all sample points in all kinds of
The average value of yield builds the 3rd RBF neural as output data；
The unit area fertilizer application amount of the soil pH of each sample point, each nutrient content and place city described in will be all kinds of
As input data, it is all kinds of described in the unit area grain yield in city where each sample point be used as output data, structure institute
State the 4th RBF neural.
Specifically, the step S1 is specifically included：
S11, the input data of each sample point is clustered using the ant group algorithm, obtain cluster centre, will
Center of the cluster centre as the RBF neural；
S12, the weights in the RBF neural from hidden layer to output layer are obtained using back-propagation algorithm；
S13, according to the output of the implicit unit of the hidden layer, the implicit unit is cut.
Specifically, the step S11 is specifically included：
S111, according to the routing information amount between any two sample point, obtain a sample in described two sample points
Point cluster is to the probability of another sample point, if judging, the probability is more than the first predetermined threshold value, by described two sample points
It is divided into one kind；
S112, the overall error per a kind of cluster centre and all classes is obtained, the overall error is less than or equal to if judging
Second predetermined threshold value, the then center using the cluster centre as the RBF neural；Or
If judging, the overall error is more than second predetermined threshold value, according to the sample point to the cluster centre
Distance and amended Pheromone Dauer property coefficient, obtain new routing information amount, use the new routing information amount iteration
Perform cluster and determine the operation of cluster centre, until the global error is less than or equal to second predetermined threshold value.
Specifically, the step S13 is specifically included：
The output valve of each implicit unit of the hidden layer is obtained, and the output valve is standardized；
If judging, the output valve after standardization is less than the 3rd predetermined threshold value, removes and list is implied corresponding to the output valve
Member.
Specifically, the step of being clustered according to the soil pH of each sample point and nutrient content to each sample point
Specifically include：
According to the soil pH and nutrient content of each sample point, using fuzzy K mean cluster algorithm to each sample
Point is clustered.
According to the second aspect of the invention, there is provided a kind of device for obtaining relation between grain yield and fertilizer application amount,
Including：
Construction unit, obtained for the unit area fertilizer application amount according to each sample point, soil pH and each nutrient content
A variety of input datas, output number is obtained according to the unit area grain yield in city where each sample point and each sample point
According to the RBF neural optimized according to each input data and the multiple ant group algorithms of output data structure；
Computing unit, for obtaining the mean square error of each RBF neural, use the RBF god that mean square error is minimum
Through the relation between network representation grain yield and fertilizer application amount.
According to the third aspect of the invention we, there is provided a kind of equipment for obtaining relation between grain yield and fertilizer application amount,
Including：
At least one processor, at least one memory and bus；Wherein,
The processor and memory complete mutual communication by the bus；
The memory storage has and by the programmed instruction of the computing device, the processor described program can be called to refer to
Order is able to carry out method as described before.
According to the fourth aspect of the invention, there is provided a kind of non-transient computer readable storage medium storing program for executing, for storing such as preceding institute
The computer program stated.
The present invention improves a kind of method and device for obtaining relation between grain yield and fertilizer application amount, and this method passes through
The RBF god of RBF neural, therefrom selection mean square error minimum that multiple ant group algorithms optimize is built using a variety of input datas
Through the relation between network representation grain yield and fertilizer application amount, represented so as to realize using more accurate RBF neural
Relation between grain yield and fertilizer application amount, the relation between the grain yield and fertilizer application amount of acquisition are also more accurate.
Brief description of the drawings
Fig. 1 is the method overall flow of relation between acquisition grain yield and fertilizer application amount provided in an embodiment of the present invention
Figure；
Fig. 2 is RBF neural schematic diagram of the prior art；
Fig. 3 is the method flow of relation between the acquisition grain yield and fertilizer application amount that further embodiment of this invention provides
Figure；
Fig. 4 is the method flow of relation between the acquisition grain yield and fertilizer application amount that another embodiment of the present invention provides
Figure；
Fig. 5 is the device overall structure of relation between acquisition grain yield and fertilizer application amount provided in an embodiment of the present invention
Schematic diagram；
Fig. 6 is the equipment overall structure of relation between acquisition grain yield and fertilizer application amount provided in an embodiment of the present invention
Schematic diagram.
Embodiment
With reference to the accompanying drawings and examples, the embodiment of the present invention is described in further detail.Implement below
Example is used to illustrate the present invention, but is not limited to the scope of the present invention.
A kind of method for obtaining relation between grain yield and fertilizer application amount is provided in one embodiment of the invention,
The method overall flow figure of Fig. 1 relations between acquisition grain yield provided in an embodiment of the present invention and fertilizer application amount, such as Fig. 1
Shown, this method includes：S1, obtained according to the unit area fertilizer application amount of each sample point, soil pH and each nutrient content more
Kind input data, output data is obtained according to the unit area grain yield in city where each sample point and each sample point,
The RBF neural of multiple ant group algorithms optimizations is built according to each input data and output data；S2, obtain each RBF god
Mean square error through network, represented using the minimum RBF neural of mean square error between grain yield and fertilizer application amount
Relation.
Specifically, in S1, the input data for the RBF neural input data, every kind of input number
The unit area fertilizer application amount or per unit fertilizer application amount of each sample point are included in.Can in a kind of data
Only to include fertilizer application amount, the pH value and nutrient content of soil can also be included.The nutrient content includes N (nitrogen), P
(phosphorus), AVP (available phosphorus), AVK (effective potassium) and OM (organic matter).An ant group algorithm corresponding to a kind of input data structure is excellent
The RBF neural of change.The number of RBF neural of the kind number of input data with building is identical.Various kinds can directly be used
The RBF neural of ant group algorithm optimization, can also be used to after the processing of each sample point data corresponding to the data structure of this point
Result structure corresponding to ant group algorithm optimization RBF neural.Optimize the various input datas as ant group algorithm
The input of RBF neural network model afterwards, made according to the unit area chemical fertilizer in city where each sample point and each sample point
Dosage obtains output data, and the output data is the data of RBF neural output.The ant group algorithm is also known as ant
Ant algorithm, it is a kind of probability type algorithm that path optimizing is found in figure.The ant group algorithm is taken the lead in by Dorigo etc. for the first time
It is proposed, be based on collective's foraging behavior of biological Ant ColonySystem and a kind of bionic optimization algorithm for growing up, there is parallel point
Cloth calculating, powerful global optimizing ability, strong adaptability and the advantages that be easy to be combined with other algorithms.The RBF nerve nets
Network is by three layers of feedforward network formed, as shown in Fig. 2 wherein first layer is input layer, node number is equal to the dimension of input；
The second layer is hidden layer, and node number is depending on the complexity of problem；Third layer is output layer, and node number is equal to output data
Dimension.x1、x2...xnFor input data, w1、w2...wnFor the weights from hidden layer to output layer.The RBF neural
Output y obtained by following formula：
Wherein, X={ x1,x2..., xnIt is input vector；ωkFor k-th of hidden layer neuron and output layer neuron
Connection weight；φkFor the output of k-th of hidden layer neuron, obtained by following formula：
μkFor the center of the RBF neural, σkFor variance.The number of hidden layer neuron is according to the complicated journey of problem
Spend to determine, although more hidden layer neuron numbers can make the accuracy of the RBF neural higher, excessive
Hidden layer neuron number can make the RBF neural training time long and produce over-fitting problem.In S2, ant colony is obtained
The mean square error of each RBF neural of algorithm optimization, i.e., use institute using individual using the input data of test sample point
The grain yield that each RBF neural calculates each sample is stated, each grain yield calculated and actual grain yield are carried out
Compare, the difference between each actual grain yield and the corresponding each grain yield calculated is obtained, to each test sample point
The squared difference after be added, then evolution, that is, obtain the mean square error of each RBF neural, use mean square error minimum
The RBF neural represents the relation between grain yield and fertilizer application amount.
The present embodiment builds the RBF neural of multiple ant group algorithm optimizations, Cong Zhongxuan by using a variety of input datas
The relation between mean square error minimum RBF neural expression grain yield and fertilizer application amount is selected, is made so as to realize
The relation between grain yield and fertilizer application amount, the grain yield and chemical fertilizer of acquisition are represented with more accurate RBF neural
Relation between usage amount is also more accurate.
On the basis of above-described embodiment, step S1 described in the present embodiment specifically includes：By the list of each sample point
Plane accumulates fertilizer application amount as input data, and the unit area grain yield of each sample point is as output data, structure
First RBF neural；Using the soil pH, each nutrient content and unit area fertilizer application amount of each sample point as input
Data, the unit area grain yield of each sample point build the second RBF neural as output data；According to described
The soil pH and nutrient content of each sample point cluster to each sample point, according to the sample point in all kinds of, respectively
Build the 3rd RBF neural and the 4th RBF neural.
Specifically, for example, sorting out 234 sample point unit area fertilizer application amounts and unit area grain yield number
According to, the input using unit area fertilizer application amount as RBF neural respectively, using unit area grain yield as output,
Build the first RBF neural network model of a single-input single-output.122 sample points are sorted out, each sample point wraps respectively
The unit area applying quantity of chemical fertilizer of soil pH containing sample point, soil nutrient content, and the sample point produces with unit area grain
Amount, form are as follows：Jiangyin nineteen eighty-two soil nutrient content and pH, unit area applying quantity of chemical fertilizer in 1980 and unit area grain
Yield；Jiangyin nineteen eighty-two soil nutrient content and pH, unit area applying quantity of chemical fertilizer in 1985 and unit area grain yield；River
Cloudy nineteen eighty-two soil nutrient content and pH, nineteen ninety unit area applying quantity of chemical fertilizer and unit area grain yield；Jiangyin 2000
Year soil nutrient content and pH, nineteen ninety unit area applying quantity of chemical fertilizer and unit area grain yield；Jiangyin soil in 2000
Nutrient content and pH, nineteen ninety-five unit area applying quantity of chemical fertilizer and unit area grain yield；Jiangyin soil nutrient in 2000 contains
Amount and pH, unit area applying quantity of chemical fertilizer in 2000 and unit area grain yield.With soil nutrient content and pH and unit
Input of the area applying quantity of chemical fertilizer as RBF neural, unit area grain yield is as output, one seven input of structure
Second RBF neural network model of one output.
On the basis of above-described embodiment, according to the sample point in all kinds of in the present embodiment, described is built respectively
The step of three RBF neurals and four RBF neurals, specifically includes：The pH of all sample point soil is put down in will be all kinds of
The average value of the unit area fertilizer application amount of average, each nutrient content average value and place city is all kinds of as input data
In the average value of the unit area grain yield in cities where all sample points be used as output data, build the described 3rd
RBF neural；The unit area chemical fertilizer of the soil pH of each sample point, each nutrient content and place city described in will be all kinds of makes
Dosage as input data, it is all kinds of described in city where each sample point unit area grain yield as output data, structure
Build the 4th RBF neural.
Specifically, for example, sorting out 170 sample points, according to the soil pH of each sample point and nutrient content to institute
State each sample point to be clustered, using cluster result, obtain the city where per a kind of sample point, then sort out sample point institute
The nineteen eighty-two in city and the soil nutrient content of 2000 and pH data and 1980,1985,1990,1995 and 2000
Unit area fertilizer application amount and unit area grain yield data.The soil nutrient content of of a sort sample point and pH are taken
Average value.3rd RBF neural network model of one output of same one seven input of structure.6194 sample points are sorted out,
Each sample point is clustered according to the soil pH of each sample point and nutrient content, will be each using cluster result
The nutrient content and pH of the sampled point included in class are organized into all as input data：Nineteen eighty-two, the soil of first sample point
PH, nutrient content, and 1980, the 1985 and 1990 unit area applying quantity of chemical fertilizer in first sample point place city,
Unit area grain yield；2000, city where soil pH, the nutrient content of the second sample point, and second sample point
1990,1995, the 2000 unit area applying quantity of chemical fertilizer in city, unit area grain yield；….Equally, in the same way
To construct the 4th RBF neural of seven inputs, one output.
Grain yield is influenced by many factors, such as, applying quantity of chemical fertilizer, weather, moisture etc..Therefore, grain yield with
There is the relation of complexity between applying quantity of chemical fertilizer, it is impossible to simply with linear model either both nonlinear model type analysis it
Between dependency relation.In addition, Application of Neural Network is extensive, especially the characteristic of RBF neural is applied to probe into unit area
Non-linear relation between grain yield and unit area applying quantity of chemical fertilizer.Can preferably it be fitted with RBF neural network model
The relation gone out between unit area applying quantity of chemical fertilizer and unit area grain yield, and according to structure in the case of different input datas
The RBF neural built is compared, inquire into increase soil nutrient input condition after to structure unit area applying quantity of chemical fertilizer with
Unit area grain yield model accuracy influences.So as to select more suitable input data structure RBF models, chemical fertilizer is represented
Relation between amount of application and grain yield, it can be seen that between applying quantity of chemical fertilizer, grain yield and the ecosystem
Relation, for protecting ecology system and agricultural production is taken into account, maintain the foundation that ecosystem balance provides science.
Fig. 3 is the method flow diagram of relation between acquisition grain yield and fertilizer application amount provided in an embodiment of the present invention,
As shown in figure 3, on the basis of above-described embodiment, step S1 described in the present embodiment specifically includes：S11, use the ant colony
Algorithm clusters to the input data of each sample point, obtains cluster centre, using the cluster centre as RBF god
Center through network；S12, the weights in the RBF neural from hidden layer to output layer are obtained using back-propagation algorithm；
S13, according to the output of the implicit unit of the hidden layer, the implicit unit is cut.
Specifically, in S11, each sample point is clustered using the ant group algorithm, obtains each sample point
Cluster centre.Center using the cluster centre as the RBF neural, the number of the cluster centre with it is described
The number at RBF neural center is identical.In S12, using cross entropy as object function, using back-propagation algorithm to described
RBF neural is trained, and obtains the weights from hidden layer to output layer in the RBF neural.In S13, according to institute
The output of each implicit unit of hidden layer is stated, the implicit unit is cut.It is described to be cut to remove the implicit list
Member.
The present embodiment according to each sample point, using the RBF neural that ant group algorithm optimizes as grain yield and
The result of relation between fertilizer application amount, fast convergence rate, network structure are simple, strong robustness and are not easy to be absorbed in local minimum
Point.
Fig. 4 is the method flow diagram of relation between acquisition grain yield and fertilizer application amount provided in an embodiment of the present invention,
As shown in figure 4, on the basis of above-described embodiment, step S11 described in the present embodiment specifically includes：S111, according to any two
Routing information amount between individual sample point, obtain a sample point cluster in described two sample points and arrive the general of another sample point
Rate, if judging, the probability is more than the first predetermined threshold value, and described two sample points are divided into one kind；S112, obtain per a kind of
Cluster centre and all classes overall error, if judging, the overall error is less than or equal to the second predetermined threshold value, will be described poly-
Center of the class center as the RBF neural；Or the overall error is more than second predetermined threshold value if judging,
According to the distance of the sample point to the cluster centre and amended Pheromone Dauer property coefficient, new routing information is obtained
Amount, perform cluster using the new routing information amount iteration and determine the operation of cluster centre, until the global error is small
In or equal to second predetermined threshold value.
Specifically, in S111, for any two sample point in the sample point, according between described two sample points
Routing information amount, obtain in described two sample points a sample point to the probability of another sample point.Sample point xiCluster
To sample point xjProbability pijObtained by following formula：
Wherein, τijFor routing information amount, n is the number of the sample point.If pijMore than the first predetermined threshold value, then by xiWith
xjIt is divided into one kind, otherwise not by xiWith xjIt is divided into one kind.
In S112, obtain and obtained per a kind of cluster centre, the cluster centre cj by following formula：
Wherein, J is and sample point xjIt is divided into the number of the sample point of one kind, xkFor the value of k-th of sample point.All classes
Global error ε is obtained by following formula：
Wherein, xkiFor with xjIt is divided into i-th of value of the k-th sample point of one kind, xjiFor xjI-th value, m is a sample
The number of the value of this point, K are the number of cluster centre.If judging, the overall error ε is less than or equal to the second predetermined threshold value, will
Center of the cluster centre as the RBF neural.If judging, the overall error is more than second predetermined threshold value,
According to the distance of the sample point to the cluster centre and amended Pheromone Dauer property coefficient, new routing information is obtained
Amount, perform cluster using the new routing information amount iteration and determine the operation of cluster centre, until the global error is small
In or equal to second predetermined threshold value.Each sample point is obtained to the distance between new cluster centre dij, modification letter
Cease the lasting property coefficient ρ of element, the new routing information amount of the formulaObtained by following formula：
Wherein,For the routing information amount in last iteration, Q is constant.
On the basis of above-described embodiment, step S13 described in the present embodiment specifically includes：Obtain the every of the hidden layer
The output valve of individual implicit unit, and the output valve is standardized；If it is pre- to judge that the output valve after standardization is less than the 3rd
If threshold value, then remove and unit is implied corresponding to the output valve.
Specifically, the output valve of each implicit unit of the hidden layer is obtained, standardization bag is carried out to the output valve
Include：The maximum in the output valve is obtained, output valve divided by the maximum with each implicit unit.If judge rule
Output valve after generalized is less than the 3rd predetermined threshold value, then removes and unit is implied corresponding to the output valve.
By being optimized to implicit unit in the present embodiment, the implicit unit that output valve is unsatisfactory for condition removes,
Ensure to simplify the structure of the RBF neural in the case of the RBF neural precision, improve arithmetic speed.
On the basis of above-described embodiment, also include before the step S111 in the present embodiment：According to any two
Euclidean distance between sample point initializes to the routing information amount between described two sample points.
Specifically, any two sample point x is calculated by following formulaiAnd xjBetween Euclidean distance dij：
dij=| | (xi-xj)||2, i, j=1,2...n,
Wherein, n is the number of the sample point.To the routing information amount τ of described two sample pointsijInitialized：
Wherein, r is the 4th predetermined threshold value.
On the basis of above-described embodiment, according to the soil pH of each sample point and nutrient content to institute in the present embodiment
The step of each sample point is clustered is stated to specifically include：According to the soil pH and nutrient content of each sample point, fuzzy K is used
Means clustering algorithm clusters to each sample point.
Specifically, clustering method is a kind of conventional data analysis technique, cluster be the set of an object is divided into it is several
Individual class, it is similar between the object in each class, but is dissimilar with the objects of other classes.Fuzzy k-means cluster
A kind of clustering algorithm based on division, it is improved based on one kind on common k mean clusters, is come pair by means of membership function
Object is divided.
A kind of dress for obtaining relation between grain yield and fertilizer application amount is provided in another embodiment of the present invention
Put, Fig. 5 is that the device overall structure of relation between acquisition grain yield and fertilizer application amount provided in an embodiment of the present invention is illustrated
Figure, as shown in figure 5, the device includes construction unit 1 and computing unit 2, wherein：
The construction unit 1 is used for more according to the acquisition of the fertilizer application amount of each sample point, the pH of soil and each nutrient content
Kind input data, output data is obtained according to the fertilizer application amount in city where each sample point and each sample point, according to institute
State the RBF neural that various input datas and output data build multiple ant group algorithm optimizations；The computing unit 2 obtains institute
The mean square error of each RBF neural is stated, grain yield and chemical fertilizer are represented using the minimum RBF neural of the mean square error
Relation between usage amount.
Specifically, the input data is the data of the input of the RBF neural, is wrapped in every kind of input data
Fertilizer application amount or average chemical fertilizer usage amount containing each sample point.The output data is the output of the RBF neural
Data.Fertilizer application amount only can be included in a kind of input data, the pH value and nutrient content of soil can also be included.One
The RBF neural of an ant group algorithm optimization corresponding to kind input data structure.The RBF of the species number of input data and structure
The number of neutral net is identical.The RBF nerves that directly can be optimized using ant group algorithm corresponding to the data structure of each sample point
Network, the RBF neural to ant group algorithm optimization corresponding to the result structure after the processing of each sample point data can also be used.
Input of the construction unit 1 using the various input datas as the RBF neural network model, by each sample point and
Output of the fertilizer application amount in city as the RBF neural where each sample point, ant group algorithm corresponding to structure optimize
RBF neural.The ant group algorithm is also known as ant algorithm, is that a kind of probability type that path optimizing is found in figure is calculated
Method.The ant group algorithm takes the lead in proposing by Dorigo etc. for the first time, is sent out based on collective's foraging behavior of biological Ant ColonySystem
A kind of bionic optimization algorithm that exhibition is got up, there is parallel distributed calculating, powerful global optimizing ability, strong adaptability and be easy to
The advantages that being combined with other algorithms.
The computing unit 2 obtains the mean square error of each RBF neural of ant group algorithm optimization, i.e., using test
The input data of sample point is each by what is calculated using the individual grain yield that each sample is calculated using each RBF neural
Grain yield obtains each actual grain yield and produced with the corresponding each grain calculated compared with the grain yield of reality
Difference between amount, to being added after the squared difference of each test sample point, then evolution, that is, obtain each RBF neural
Mean square error, the relation between grain yield and fertilizer application amount is represented using the minimum RBF neural of mean square error.
The present embodiment builds the RBF neural of multiple ant group algorithm optimizations, Cong Zhongxuan by using a variety of input datas
The relation between mean square error minimum RBF neural expression grain yield and fertilizer application amount is selected, is made so as to realize
The relation between grain yield and fertilizer application amount, the grain yield and chemical fertilizer of acquisition are represented with more accurate RBF neural
Relation between usage amount is also more accurate.
The present embodiment provides a kind of equipment for obtaining relation between grain yield and fertilizer application amount, and Fig. 6 is real for the present invention
The equipment overall structure schematic diagram of relation between acquisition grain yield and fertilizer application amount that example provides is applied, the equipment includes：Extremely
Few a processor 61, at least one memory 62 and bus 63；Wherein,
The processor 61 and memory 62 complete mutual communication by the bus 63；
The memory 62 is stored with the programmed instruction that can be performed by the processor 61, and the processor calls the journey
Sequence instruction is able to carry out the method that above-mentioned each method embodiment is provided, such as including：S1, according to the unit area of each sample point
Fertilizer application amount, soil pH and each nutrient content obtain a variety of input datas, according to where each sample point and each sample point
The unit area grain yield in city obtains output data, and it is excellent to build multiple ant group algorithms according to each input data and output data
The RBF neural of change；S2, the mean square error of each RBF neural is obtained, use the RBF nerves that mean square error is minimum
Relation between network representation grain yield and fertilizer application amount.
The present embodiment provides a kind of non-transient computer readable storage medium storing program for executing, the non-transient computer readable storage medium storing program for executing
Computer instruction is stored, the computer instruction makes the computer perform the method that above-mentioned each method embodiment is provided, example
Such as include：S1, a variety of input numbers are obtained according to the unit area fertilizer application amount of each sample point, soil pH and each nutrient content
According to according to the unit area grain yield acquisition output data in city where each sample point and each sample point, according to each defeated
Enter data and output data builds the RBF neural of multiple ant group algorithm optimizations；S2, obtain each RBF neural
Mean square error, the relation between grain yield and fertilizer application amount is represented using the minimum RBF neural of mean square error.
One of ordinary skill in the art will appreciate that：Realizing all or part of step of above method embodiment can pass through
Programmed instruction related hardware is completed, and foregoing program can be stored in a computer read/write memory medium, the program
Upon execution, the step of execution includes above method embodiment；And foregoing storage medium includes：ROM, RAM, magnetic disc or light
Disk etc. is various can be with the medium of store program codes.
The embodiment such as equipment of relation is only to illustrate between acquisition grain yield and fertilizer application amount described above
Property, wherein the unit illustrated as separating component can be or may not be it is physically separate, as unit
The part of display can be or may not be physical location, you can with positioned at a place, or can also be distributed to more
On individual NE.Some or all of module therein can be selected to realize this embodiment scheme according to the actual needs
Purpose.Those of ordinary skill in the art are not in the case where paying performing creative labour, you can to understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
Realized by the mode of software plus required general hardware platform, naturally it is also possible to pass through hardware.Based on such understanding, on
The part that technical scheme substantially in other words contributes to prior art is stated to embody in the form of software product, should
Computer software product can store in a computer-readable storage medium, such as ROM/RAM, magnetic disc, CD, including some fingers
Make to cause a computer equipment (can be personal computer, server, or network equipment etc.) to perform each implementation
Method described in some parts of example or embodiment.
Finally, the present processes are only preferable embodiment, are not intended to limit the scope of the present invention.It is all
Within the spirit and principles in the present invention, any modification, equivalent substitution and improvements made etc., the protection of the present invention should be included in
Within the scope of.
Claims (10)
- A kind of 1. method for obtaining relation between grain yield and fertilizer application amount, it is characterised in that including：S1, a variety of input datas, root are obtained according to the unit area fertilizer application amount of each sample point, soil pH and each nutrient content Output data is obtained according to the unit area grain yield in city where each sample point and each sample point, according to each input data The RBF neural of multiple ant group algorithm optimizations is built with output data；S2, the mean square error of each RBF neural is obtained, grain is represented using the minimum RBF neural of mean square error Relation between yield and fertilizer application amount.
- 2. according to the method for claim 1, it is characterised in that the step S1 is specifically included：Using the unit area fertilizer application amount of each sample point as input data, the unit area grain of each sample point Yield builds the first RBF neural as output data；It is described each using the soil pH, each nutrient content and unit area fertilizer application amount of each sample point as input data The unit area grain yield of sample point builds the second RBF neural as output data；Each sample point is clustered according to the soil pH of each sample point and nutrient content, according in all kinds of Sample point, the 3rd RBF neural and the 4th RBF neural are built respectively.
- 3. according to the method for claim 2, it is characterised in that according to the sample point in all kinds of, respectively described in structure The step of 3rd RBF neural and four RBF neurals, specifically includes：The pH average values of all sample point soil, each nutrient content average value and the unit area in place city in will be all kinds of The average value of fertilizer application amount is as input data, the unit area grain yield in city where all sample points in all kinds of Average value as output data, build the 3rd RBF neural；The unit area fertilizer application amount of the soil pH of each sample point, each nutrient content and place city described in will be all kinds of as Input data, it is all kinds of described in the unit area grain yield in city where each sample point be used as output data, build described the Four RBF neurals.
- 4. according to any described methods of claim 1-3, it is characterised in that the step S1 is specifically included：S11, the input data of each sample point is clustered using the ant group algorithm, obtain cluster centre, by described in Center of the cluster centre as the RBF neural；S12, the weights in the RBF neural from hidden layer to output layer are obtained using back-propagation algorithm；S13, according to the output of the implicit unit of the hidden layer, the implicit unit is cut.
- 5. according to the method for claim 4, it is characterised in that the step S11 is specifically included：S111, according to the routing information amount between any two sample point, obtain a sample point in described two sample points and gather For class to the probability of another sample point, if judging, the probability is more than the first predetermined threshold value, and described two sample points are divided into It is a kind of；S112, the overall error per a kind of cluster centre and all classes is obtained, the overall error is less than or equal to second if judging Predetermined threshold value, the then center using the cluster centre as the RBF neural；OrIf judging, the overall error is more than second predetermined threshold value, the distance according to the sample point to the cluster centre With amended Pheromone Dauer property coefficient, new routing information amount is obtained, is performed using the new routing information amount iteration Cluster and the operation for determining cluster centre, until the global error is less than or equal to second predetermined threshold value.
- 6. according to the method for claim 4, it is characterised in that the step S13 is specifically included：The output valve of each implicit unit of the hidden layer is obtained, and the output valve is standardized；If judging, the output valve after standardization is less than the 3rd predetermined threshold value, removes and unit is implied corresponding to the output valve.
- 7. according to the method for claim 2, it is characterised in that according to the soil pH and nutrient content pair of each sample point The step of each sample point is clustered specifically includes：According to the soil pH and nutrient content of each sample point, each sample is clicked through using fuzzy K mean cluster algorithm Row cluster.
- A kind of 8. device for obtaining relation between grain yield and fertilizer application amount, it is characterised in that including：Construction unit, obtained for the unit area fertilizer application amount according to each sample point, soil pH and each nutrient content a variety of Input data, output data, root are obtained according to the unit area grain yield in city where each sample point and each sample point The RBF neural of multiple ant group algorithm optimizations is built according to each input data and output data；Computing unit, for obtaining the mean square error of each RBF neural, use the RBF nerve nets that mean square error is minimum Network represents the relation between grain yield and fertilizer application amount.
- A kind of 9. equipment for obtaining relation between grain yield and fertilizer application amount, it is characterised in that including：At least one processor, at least one memory and bus；Wherein,The processor and memory complete mutual communication by the bus；The memory storage has can be by the programmed instruction of the computing device, and the processor calls described program instruction energy Enough perform the method as described in claim 1 to 7 is any.
- 10. a kind of non-transient computer readable storage medium storing program for executing, it is characterised in that the non-transient computer readable storage medium storing program for executing is deposited Computer instruction is stored up, the computer instruction makes the computer perform the method as described in claim 1 to 7 is any.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710816507.8A CN107665244B (en) | 2017-09-12 | 2017-09-12 | Method and device for acquiring relationship between grain yield and fertilizer usage amount |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710816507.8A CN107665244B (en) | 2017-09-12 | 2017-09-12 | Method and device for acquiring relationship between grain yield and fertilizer usage amount |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107665244A true CN107665244A (en) | 2018-02-06 |
| CN107665244B CN107665244B (en) | 2020-01-14 |
Family
ID=61097394
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710816507.8A Expired - Fee Related CN107665244B (en) | 2017-09-12 | 2017-09-12 | Method and device for acquiring relationship between grain yield and fertilizer usage amount |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107665244B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111145304A (en) * | 2019-12-27 | 2020-05-12 | 新奥数能科技有限公司 | Data processing method and device, intelligent terminal and storage medium |
| CN113159265A (en) * | 2021-03-24 | 2021-07-23 | 国网河南省电力公司电力科学研究院 | Traction load parameter identification method and system based on SVM-ant colony algorithm |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6490573B1 (en) * | 2000-04-11 | 2002-12-03 | Philip Chidi Njemanze | Neural network for modeling ecological and biological systems |
| CN105138865A (en) * | 2015-10-10 | 2015-12-09 | 重庆科技学院 | Multi-parameter optimization method of giant liposome preparation process |
| CN106373022A (en) * | 2016-09-13 | 2017-02-01 | 重庆科技学院 | BP-GA-based greenhouse crop plantation efficiency condition optimization method and system |
| CN106651001A (en) * | 2016-11-08 | 2017-05-10 | 浙江理工大学 | Needle mushroom yield prediction method based on improved neural network and implementation system |
-
2017
- 2017-09-12 CN CN201710816507.8A patent/CN107665244B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6490573B1 (en) * | 2000-04-11 | 2002-12-03 | Philip Chidi Njemanze | Neural network for modeling ecological and biological systems |
| CN105138865A (en) * | 2015-10-10 | 2015-12-09 | 重庆科技学院 | Multi-parameter optimization method of giant liposome preparation process |
| CN106373022A (en) * | 2016-09-13 | 2017-02-01 | 重庆科技学院 | BP-GA-based greenhouse crop plantation efficiency condition optimization method and system |
| CN106651001A (en) * | 2016-11-08 | 2017-05-10 | 浙江理工大学 | Needle mushroom yield prediction method based on improved neural network and implementation system |
Non-Patent Citations (2)
| Title |
|---|
| 屈文婷: "基于粒子群算法及RBF神经网络技术的粮食产量预测方法", 《中国优秀硕士学位论文全文数据库 经济与管理科学辑》 * |
| 马洪伟 等: "基于蚁群聚类和裁剪方法的RBF神经网络优化算法", 《青岛大学学报（工程技术版）》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111145304A (en) * | 2019-12-27 | 2020-05-12 | 新奥数能科技有限公司 | Data processing method and device, intelligent terminal and storage medium |
| CN111145304B (en) * | 2019-12-27 | 2023-10-10 | 新奥数能科技有限公司 | Data processing method and device, intelligent terminal and storage medium |
| CN113159265A (en) * | 2021-03-24 | 2021-07-23 | 国网河南省电力公司电力科学研究院 | Traction load parameter identification method and system based on SVM-ant colony algorithm |
| CN113159265B (en) * | 2021-03-24 | 2022-09-09 | 国网河南省电力公司电力科学研究院 | Traction load parameter identification method and system based on SVM-ant colony algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107665244B (en) | 2020-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Remesan et al. | Hydrological data driven modelling | |
| Keerthana et al. | An ensemble algorithm for crop yield prediction | |
| Gong et al. | Multi-objective parameter optimization of common land model using adaptive surrogate modeling | |
| Gao et al. | Identifying preferred management options: An integrated agent-based recreational fishing simulation model with an AHP-TOPSIS evaluation method | |
| Ahmadi et al. | Learning fuzzy cognitive maps using imperialist competitive algorithm | |
| US9043326B2 (en) | Methods and systems for biclustering algorithm | |
| Pokkuluri et al. | Crop Disease Prediction with Convolution Neural Network (CNN) Augmented With Cellular Automata. | |
| CN105930531A (en) | Method for optimizing cloud dimensions of agricultural domain ontological knowledge on basis of hybrid models | |
| Czajkowski et al. | Steering the interpretability of decision trees using lasso regression-an evolutionary perspective | |
| Nadimi-Shahraki et al. | Discrete improved grey wolf optimizer for community detection | |
| Chatterjee et al. | Non-dominated sorting genetic algorithm—II supported neural network in classifying forest types | |
| CN107665244A (en) | A kind of method and device for obtaining relation between grain yield and fertilizer application amount | |
| Senthilnath et al. | FPA clust: evaluation of the flower pollination algorithm for data clustering | |
| CN108830407B (en) | Sensor distribution optimization method in structure health monitoring under multi-working condition | |
| Wijayaningrum et al. | An improved crow search algorithm for data clustering | |
| Kathiria et al. | Smart Crop Recommendation System: A Machine Learning Approach for Precision Agriculture | |
| Harifi et al. | Using Metaheuristic Algorithms to Improve k-Means Clustering: A Comparative Study. | |
| CN111612246A (en) | Method, device and equipment for predicting heavy metal content of farmland soil and storage medium | |
| Alif et al. | Crop prediction based on geographical and climatic data using machine learning and deep learning | |
| Raman et al. | Enhanced reptile search optimization with convolutional autoencoder for soil nutrient classification model | |
| Manjula et al. | Efficient prediction of recommended crop variety through soil nutrients using deep learning algorithm | |
| Owuor et al. | Ant colony optimization for mining gradual patterns | |
| Agarwal et al. | Crop Prediction Using Ensemble Learning | |
| Sajid et al. | Integrating crop simulation and machine learning models to improve crop yield prediction | |
| Jun-wei et al. | Differences in parameter estimates derived from various methods for the ORYZA (v3) Model |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200114 Termination date: 20210912 |