CN110414115A - A kind of wavelet neural network tomato yield prediction technique based on genetic algorithm - Google Patents
A kind of wavelet neural network tomato yield prediction technique based on genetic algorithm Download PDFInfo
- Publication number
- CN110414115A CN110414115A CN201910663789.1A CN201910663789A CN110414115A CN 110414115 A CN110414115 A CN 110414115A CN 201910663789 A CN201910663789 A CN 201910663789A CN 110414115 A CN110414115 A CN 110414115A
- Authority
- CN
- China
- Prior art keywords
- model
- neural network
- value
- input
- individual
- 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
- 235000007688 Lycopersicon esculentum Nutrition 0.000 title claims abstract description 35
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 29
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 28
- 230000002068 genetic effect Effects 0.000 title claims abstract description 25
- 238000000034 method Methods 0.000 title claims abstract description 16
- 240000003768 Solanum lycopersicum Species 0.000 title abstract description 33
- 238000012545 processing Methods 0.000 claims abstract description 10
- 238000012216 screening Methods 0.000 claims abstract description 9
- 238000012549 training Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 22
- 230000000694 effects Effects 0.000 claims description 11
- 238000003062 neural network model Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 9
- 230000006978 adaptation Effects 0.000 claims description 8
- 238000003973 irrigation Methods 0.000 claims description 5
- 230000002262 irrigation Effects 0.000 claims description 5
- 238000013519 translation Methods 0.000 claims description 4
- KWYUFKZDYYNOTN-UHFFFAOYSA-M Potassium hydroxide Chemical compound [OH-].[K+] KWYUFKZDYYNOTN-UHFFFAOYSA-M 0.000 claims description 3
- 230000004075 alteration Effects 0.000 claims description 3
- 239000003337 fertilizer Substances 0.000 claims description 3
- 238000010353 genetic engineering Methods 0.000 claims description 3
- 238000005286 illumination Methods 0.000 claims description 3
- 230000035772 mutation Effects 0.000 claims description 3
- 239000000618 nitrogen fertilizer Substances 0.000 claims description 3
- 239000002686 phosphate fertilizer Substances 0.000 claims description 3
- 229940072033 potash Drugs 0.000 claims description 3
- 235000015320 potassium carbonate Nutrition 0.000 claims description 3
- BWHMMNNQKKPAPP-UHFFFAOYSA-L potassium carbonate Substances [K+].[K+].[O-]C([O-])=O BWHMMNNQKKPAPP-UHFFFAOYSA-L 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 241000227653 Lycopersicon Species 0.000 claims 3
- 230000010354 integration Effects 0.000 abstract description 3
- 230000005284 excitation Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 5
- 238000012360 testing method Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 2
- 238000009423 ventilation Methods 0.000 description 2
- 230000008649 adaptation response Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000003967 crop rotation Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 235000021049 nutrient content Nutrition 0.000 description 1
- 230000001766 physiological effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 235000013311 vegetables Nutrition 0.000 description 1
- 238000009333 weeding Methods 0.000 description 1
Classifications
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Economics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Tourism & Hospitality (AREA)
- Evolutionary Biology (AREA)
- Quality & Reliability (AREA)
- Genetics & Genomics (AREA)
- Physiology (AREA)
- Primary Health Care (AREA)
- Mining & Mineral Resources (AREA)
- Marine Sciences & Fisheries (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
Abstract
The wavelet neural network tomato yield prediction technique based on genetic algorithm that the invention discloses a kind of, belongs to technical field of agricultural information.This method first chooses a certain amount of parameter as input variable, passes through the classification, processing and the screening that carry out to these parameters and optimizes training to it to using genetic algorithm, obtains suitable data population.Then the number of nodes of network input layer, hidden layer and output layer is determined, replace the excitation function of the hiding node layer of BP neural network using wavelet basis function, and it introduces shift factor and scale factor is constantly adjusted weight, by repeatedly trained and iteration, making the error of model prediction constantly reduces, it is enabled to export result constantly close to measured value, by control errors in reasonable range, the precision of prediction and function convergence performance of model are improved, provides more reliable theories integration for crop yield in prediction greenhouse.
Description
Technical field
The invention belongs to technical field of agricultural information, in particular to a kind of wavelet neural network tomato based on genetic algorithm
Production prediction method.
Background technique
Tomato is one of the main vegetables of Northern Part of China facility cultivation, the indoor tomato planting of northern area sunlight
The cultivation mode of efficient crop rotation is mostly used, the physiological property of the environmental factor in greenhouse, the nutrient content of soil and tomato is directly determined
The yield of tomato is determined.The indoor tomato yield of prediction temperature can be to determine suitable planting scheme, reasonable arrangement dose,
Irrigation volume, takes the measures such as deinsectization, weeding in time, periodically takes ventilation, heats, the measures such as light filling, is the healthy growth of tomato
It creates a favorable environment, and lays the foundation for the good quality and high output of greenhouse tomato.The research that forefathers predict tomato yield is very wide
It is general, but be mostly focused on and the indoor tomato yield of temperature is predicted using crop mechanism model or crop growth model,
The parameter being related to is relatively fewer, and calculation method has certain limitation.Using genetic algorithm in conjunction with wavelet neural network
Mode carries out the prediction of greenhouse tomato yield, can be according to historical data over the years, and passes through each in continuous correction model
Kind parametric variable improves and Optimized model, lift scheme precision of prediction make it be more applicable for the actual rings of Greenhouse in North
Border has more wide application prospect.
Summary of the invention
The present invention provides a kind of wavelet neural network tomato yield prediction technique based on genetic algorithm, can predict the north
Tomato yield in energy saving sunlight greenhouse can be determining suitable planting scheme.
A kind of wavelet neural network tomato yield prediction model based on genetic algorithm, which is characterized in that this model it is defeated
Entering variable includes following characteristics parameter: environment temperature；Ambient humidity；Irrigation volume；Nitrogenous fertilizer input amount；Phosphate fertilizer input amount；Potash fertilizer is thrown
Enter amount；CO2Concentration；Intensity of illumination；
The model is established by following steps:
An initial population X is generated at randomm×n:
N=s1×s2+s2×s3+s2+s3 (1)
Wherein: m is initial population quantity, and n is individual lengths, and individual lengths had both represented the genic value quantity of each individual,
Also the initial weight quantity an of neural network is represented；s1For input layer number；s2For node in hidden layer；s3For output layer
Number of nodes；
The output error value E that genetic algorithm passes through each individual in calculating initial populationi, fitness value fi, and according to individual
The value of fitness value is assessed, and individual of the fitness value in preset range A in initial population is selected to continue into sub- population
Optimize training:
In sub- population, i-th of individual intersect or the probability of mutation operation is pi, according to crossing-over rate pcAnd aberration rate
pmAuto-adaptive function come judge the individual whether carry out intersect or genetic manipulation:
In formula: kc、kmIt is the real number less than 1, fcFor the ideal adaptation angle value to be intersected, fmIt is suitable for the individual to be made a variation
Answer angle value, fmax、Maximum adaptation angle value and average fitness value respectively in population,For the degree of convergence of population；
The data that input variable large deviations are more than its annual mean ± 10% are rejected, using genetic algorithm to data-optimized,
Data dynamic range≤± 5% is considered as that treatment effect is preferable, will be by screening with the data after optimization processing as wavelet neural
The input data of network simultaneously carries out subsequent arithmetic；
Treated characteristic parameter will be passed through as the input of wavelet-neural network model, greenhouse tomato yield conduct
Output, wherein include the following steps:
The mother wavelet function of the model selection are as follows:
G (x)=cos (1.75x) exp (- x2/2) (7)
Mother wavelet function is subjected to scale and translation transition structure wavelet basis function:
Wherein, aj、bjThe scale factor and shift factor of respectively j-th hidden layer node；
ykFor the output of model:
Wherein, xi(i=1,2 ..., I) is the input of i-th of node of input layer, yk(j=1,2 ..., k) be output layer
The output of j-th of node；
Define error function E are as follows:
Wherein, ykFor the reality output of k-th of node of output layer, tkIt is exported for the target of k-th of node of output layer；
Adjustment to weight, scale factor, shift factor includes:
Wherein, learning rate: η (η > 0), factor of momentum: μ (0 < μ < 1).
More preferably, error amount is expressed as a percentage, then fitness value is in 0~1 range, A=fi> 0.667.
The present invention provides a kind of wavelet neural network tomato yield prediction technique based on genetic algorithm, by data into
Capable classification, processing and screening simultaneously optimize training to it using genetic algorithm, obtain suitable data population.Then it determines
The number of nodes of network input layer, hidden layer and output layer replaces BP neural network to hide swashing for node layer using wavelet basis function
Function is encouraged, and introduces shift factor and scale factor is constantly adjusted weight, by repeatedly trained and iteration, makes model
The error of prediction constantly reduces, and enables its output result constantly close to measured value, can be by control errors in reasonable range.
Detailed description of the invention
Fig. 1 is a kind of small echo of the wavelet neural network tomato yield prediction technique based on genetic algorithm provided by the invention
Neural network structure figure；
Fig. 2 is GA-WNN forecast result of model figure；
Fig. 3 is GA-WNN model error percentage curve；
Fig. 4 is WNN forecast result of model figure；
Fig. 5 is WNN model error percentage curve；
Fig. 6 is BP neural network forecast result of model；
Fig. 7 is BP neural network model error percentage curve；
Fig. 8 is a kind of process of the wavelet neural network tomato yield prediction technique based on genetic algorithm provided by the invention
Figure.
Specific embodiment
With reference to the accompanying drawing, the specific embodiment of the present invention is described in detail, it is to be understood that of the invention
Protection scope be not limited by the specific implementation.
The establishment process of tomato yield prediction model: first to data carry out classification, processing and screening and to application lose
Propagation algorithm optimizes training to it, obtains suitable data population.Then network input layer, hidden layer and output layer are determined
Number of nodes replaces BP neural network to hide the excitation function of node layer, and introduces shift factor and ruler using wavelet basis function
The degree factor is constantly adjusted weight, by repeatedly trained and iteration, reduces the error of model prediction constantly, enables its output
As a result constantly close to measured value, by control errors in reasonable range, detailed process is as shown in Figure 8.
GA-WNN algorithm specifically executes step:
This model chooses following parameter as input variable: 1. environment temperature；2. ambient humidity；3. irrigation volume；4. nitrogenous fertilizer
Input amount；5. phosphate fertilizer input amount；6. potash fertilizer input amount；7. CO2 concentration；8. intensity of illumination, using tomato yield data as the mould
The output variable of type.Firstly, optimizing processing using genetic algorithm as the preposition means of WNN algorithm to initial data, obtaining
The data of dynamic range smaller (dynamic range of data≤± 5%), will be by screening with the data after optimization as subsequent
The initial value of WNN algorithm.Then, traditional BP neural network is improved using wavelet function, by more points of wavelet function
Distinguish that the characteristic of analysis is combined with the adaptive ability of neural network, wavelet structure neural network model, and it is constantly corrected
With it is perfect, just greatly improve the precision of prediction and function convergence performance of model in this way, for prediction greenhouse in crop yield mention
For more reliable theories integration.
Specific operation process is as follows:
(1) screening and optimization processing are carried out to 8 kinds of selected parameters using genetic algorithm.Firstly, generating one at random
Initial population Xm×nProcessing is optimized to initial data, wherein m be initial population quantity (because have chosen 8 kinds of variables,
Take m=8), individual lengths n had both represented the genic value quantity of each individual, also represented the initial weight quantity an of neural network
(hidden layer number of nodes can constantly change with network training, and weight quantity also changes correspondingly, initially set 8 input layers,
5 hiding node layers, 1 output node layer, initial weight quantity are 46) and genic value in individual and neural network
Initial weight corresponds.This research encodes genic value using real coding mode, can avoid decoding process, improves
Training effectiveness.
N=s1×s2+s2×s3+s2+s3 (1)
In formula: n is individual lengths；s1For input layer number；s2For node in hidden layer；s3For output layer number of nodes.
The output error value E that genetic algorithm passes through each individual in calculating initial populationi, fitness value fi, and according to individual
The value of fitness value is assessed, and selects in initial population that larger (error amount is expressed as a percentage fitness value, then fitness value
In 0~1 range, by fiThe value of > 0.667 is considered as the preferable numerical value of fitness) individual enter sub- population continue it is excellent
Change training.
In sub- population, the selected intersection or the probability of mutation operation of carrying out of i-th of individual is pi, and according to crossing-over rate
pcWith aberration rate pmAuto-adaptive function come judge the individual whether carry out intersect or genetic manipulation, pcWith pmValue can root
According to individual fitness fiValue and change adaptive response, keep population that there is diversity always with this.
In formula: kc、kmIt is the real number less than 1, fcFor the ideal adaptation angle value to be intersected, fmIt is suitable for the individual to be made a variation
Answer angle value, fmax、Maximum adaptation angle value and average fitness value respectively in population,For the degree of convergence of population.According to
Upper genetic algorithm accordingly carries out screening and optimization processing to initial data, and rejecting deviation is larger, and (certain variable is more than its annual
Value ± 10% is accordingly to be regarded as the biggish value of deviation, and constrains its value with this bound, is defined as boundary value and carries out subsequent meter
Calculate), using genetic algorithm to data-optimized, the dynamic range of data can be made to reduce, data dynamic range≤± 5% is considered as place
Reason effect is preferable, as the input data of wavelet neural network and carries out subsequent fortune with the data after optimization processing for by screening
It calculates.
(2) it handles by upper step with 8 kinds of characteristic parameters after optimization as the defeated of wavelet neural network (WNN) model
Enter, for greenhouse tomato yield as output, the structure of network is most important for the precision of prediction of model, stability.Construct model
When first have to consider the structure of model, can meet test requirements document, guarantee model stability and reliability, then application misses
Difference function analysis there are the problem of, continue Optimized model, promoted precision of prediction.
The mother wavelet function of the model selection are as follows:
G (x)=cos (1.75x) exp (- x2/2) (7)
Mother wavelet function is subjected to scale and translation transition structure wavelet basis function:
Wherein, aj、bjThe scale factor and shift factor of respectively j-th hidden layer node.ykFor the output of model:
xiThe input of (i=1,2 ..., I) i-th of node of-input layer, yk(j=1,2 ..., k)-j-th of output layer section
The output of point.
By being constantly adjusted to the weight in this model, scale factor, shift factor, the stability of lift scheme and
Precision of prediction enables this model be more applicable for actual site environment, and constantly improve and optimize the model.Define error letter
Number E are as follows:
Wherein, ykFor the reality output of k-th of node of output layer, tkIt is exported for the target of k-th of node of output layer.To power
The adjustment of value, scale factor, shift factor includes:
Wherein, learning rate: η (η > 0), factor of momentum: μ (0 < μ < 1).By constantly being adjusted to the parameter in formula
It is whole, thus the precision of prediction of lift scheme, and emulated using MATLAB software, verify the prediction effect of model.
Then by the parameters such as GA parameter, BP parameter, weight, scale factor, shift factor in this model constantly into
Row adjustment, the stability and precision of prediction of lift scheme enable this model be more applicable for actual site environment, and constantly complete
It is apt to and optimizes the model.
The data that this test is selected 2010~2015 years are trained model, examine using 2016~2018 years data
Test the forecast result of GA-WNN model.This network is trained using 6 years data, the continuous weight for adjusting model,
The model parameters such as scale factor, shift factor, learning rate, momentum factor and the number of iterations, simultaneously by multiple comparison test result
Error is calculated, constantly makes improvements and optimizes, obtained the preferable model of prediction effect, and its error is in reasonable model
In enclosing (percentage error is within 5%), predicted value differs very little with measured value, realizes precisely to greenhouse tomato yield substantially
Prediction, the prediction effect of the GA-WNN model such as following figure.The result shows that this model predication value is preferable to the tracking effect of measured value,
And greenhouse tomato yield can be effectively predicted in the high stability of model.Experimental data and comparison of test results table are as follows:
In order to verify the prediction effect of model, GA-WNN model and WNN model, BP neural network model are compared,
Prediction effect comparative situation see the table below.Through analysis it is found that predicted value and actual value using two kinds of models have certain error,
GA-WNN model average relative error is that 0.66%, WNN model average relative error is 1.02%, and BP neural network model is put down
Equal relative error is that 2.42%, GA-WNN forecast result of model is ideal.GA-WNN model convergence rate is better than WNN and BP mind
Through network model, had reached by 208 step prediction effects optimal.In conclusion by by genetic algorithm, wavelet analysis and BP
The convergence rate that neural network is combined the GA-WNN model constructed is very fast, precision of prediction is higher, realizes tomato yield
Accurate prediction, can for the indoor crop species of reasonable arrangement temperature, formulate irrigate and fertilising etc. governing plans foundation be provided.
This research and utilization GA-WNN model predicts the tomato yield in Greenhouse in North, has comprehensively considered influence
The factor of greenhouse tomato yield is handled and is screened to parameter using means such as genetic algorithms, and the feature of data is then extracted
Vector constructs greenhouse tomato Production Forecast Models as mode input, and utilizes the wavelet neural network mesoscale factor and translation
The factor carries out weighed value adjusting, the precision of prediction of lift scheme.Simulation result shows that GA-WNN model average relative error is
0.66%, WNN model average relative error are 1.02%, and the average relative error of BP neural network model is 2.42%, GA-
WNN forecast result of model is better than WNN and BP neural network model, has reached by 208 step prediction effects optimal, it was demonstrated that GA-
WNN model has preferable practical application value.Compared with traditional WNN and BP neural network model, GA-WNN model is put down
Relative error is smaller, precision is higher, stability is stronger, more meets greenhouse tomato production forecast, can be tomato species in greenhouse
The formulation for planting decision provides foundation, and provides measures such as irrigation, fertilising, light filling and the ventilations taken during tomato planting
Certain theories integration, and facilitation is played to the improvement and optimization of Greenhouse System.
Disclosed above is only several specific embodiments of the invention, and still, the embodiment of the present invention is not limited to this, is appointed
What what those skilled in the art can think variation should all fall into protection scope of the present invention.
Claims (2)
1. a kind of wavelet neural network tomato yield prediction model based on genetic algorithm, which is characterized in that the input of this model
Variable includes following characteristics parameter: environment temperature；Ambient humidity；Irrigation volume；Nitrogenous fertilizer input amount；Phosphate fertilizer input amount；Potash fertilizer investment
Amount；CO2Concentration；Intensity of illumination；
The model is established by following steps:
An initial population X is generated at randomm×n:
N=s1×s2+s2×s3+s2+s3 (1)
Wherein: m is initial population quantity, and n is individual lengths, and individual lengths had both represented the genic value quantity of each individual, also generation
The initial weight quantity of one neural network of table；s1For input layer number；s2For node in hidden layer；s3To export node layer
Number；
The output error value E that genetic algorithm passes through each individual in calculating initial populationi, fitness value fi, and according to ideal adaptation
The value of angle value is assessed, and individual of the fitness value in preset range A in initial population is selected to continue into sub- population
Optimization training:
In sub- population, i-th of individual intersect or the probability of mutation operation is pi, according to crossing-over rate pcWith aberration rate pmFrom
Fitness function come judge the individual whether carry out intersect or genetic manipulation:
In formula: kc、kmIt is the real number less than 1, fcFor the ideal adaptation angle value to be intersected, fmFor the individual adaptation degree to be made a variation
Value, fmax、Maximum adaptation angle value and average fitness value respectively in population,For the degree of convergence of population；
The data that input variable large deviations are more than its annual mean ± 10% are rejected, using genetic algorithm to data-optimized, data
Dynamic range≤± 5% is considered as that treatment effect is preferable, will be by screening with the data after optimization processing as wavelet neural network
Input data and carry out subsequent arithmetic；
Using by treated characteristic parameter, as the input of wavelet-neural network model, greenhouse tomato yield is as defeated
Out, wherein include the following steps:
The mother wavelet function of the model selection are as follows:
G (x)=cos (1.75x) exp (- x2/2) (7)
Mother wavelet function is subjected to scale and translation transition structure wavelet basis function:
Wherein, aj、bjThe scale factor and shift factor of respectively j-th hidden layer node；
ykFor the output of model:
Wherein, xi(i=1,2 ..., I) is the input of i-th of node of input layer, yk(j=1,2 ..., k) be j-th of output layer
The output of node；
Define error function E are as follows:
Wherein, ykFor the reality output of k-th of node of output layer, tkIt is exported for the target of k-th of node of output layer；
Adjustment to weight, scale factor, shift factor includes:
Wherein, learning rate: η (η > 0), factor of momentum: μ (0 < μ < 1).
2. a kind of wavelet neural network tomato yield prediction technique based on genetic algorithm as described in claim 1, feature
It is, error amount is expressed as a percentage, then fitness value is in 0~1 range, A=fi> 0.667.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910663789.1A CN110414115B (en) | 2019-07-13 | 2019-07-13 | Wavelet neural network tomato yield prediction method based on genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910663789.1A CN110414115B (en) | 2019-07-13 | 2019-07-13 | Wavelet neural network tomato yield prediction method based on genetic algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110414115A true CN110414115A (en) | 2019-11-05 |
| CN110414115B CN110414115B (en) | 2023-01-20 |
Family
ID=68362450
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910663789.1A Expired - Fee Related CN110414115B (en) | 2019-07-13 | 2019-07-13 | Wavelet neural network tomato yield prediction method based on genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110414115B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113158470A (en) * | 2020-11-25 | 2021-07-23 | 中国石油大学(华东) | Oil reservoir automatic history fitting system and method based on transfer learning |
| CN113379188A (en) * | 2021-05-06 | 2021-09-10 | 贵州省烟草公司贵阳市公司 | Tobacco crop rotation planting method and system based on Internet of things |
| CN113439520A (en) * | 2021-07-21 | 2021-09-28 | 中国农业科学院农业环境与可持续发展研究所 | Intelligent decision-making method and system for crop irrigation and fertilization |
| CN115907366A (en) * | 2022-11-15 | 2023-04-04 | 淮阴工学院 | Agricultural product growth environment optimal regulation and control method and equipment based on flamingo algorithm |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150142837A1 (en) * | 2013-11-20 | 2015-05-21 | Qliktech International Ab | Methods And Systems For Wavelet Based Representation |
| CN106651012A (en) * | 2016-12-02 | 2017-05-10 | 东华大学 | Crop transpiration prediction method based on improved extreme learning machine |
| US20180202264A1 (en) * | 2016-02-08 | 2018-07-19 | Rs Energy Group Topco, Inc. | Method for estimating oil/gas production using statistical learning models |
| CN109345508A (en) * | 2018-08-31 | 2019-02-15 | 北京航空航天大学 | A kind of Assessing Standards For Skeletal method based on two stages neural network |
| CN109359741A (en) * | 2018-09-27 | 2019-02-19 | 华南师范大学 | A kind of wastewater treatment influent quality timing variations intelligent Forecasting |
-
2019
- 2019-07-13 CN CN201910663789.1A patent/CN110414115B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150142837A1 (en) * | 2013-11-20 | 2015-05-21 | Qliktech International Ab | Methods And Systems For Wavelet Based Representation |
| US20180202264A1 (en) * | 2016-02-08 | 2018-07-19 | Rs Energy Group Topco, Inc. | Method for estimating oil/gas production using statistical learning models |
| CN106651012A (en) * | 2016-12-02 | 2017-05-10 | 东华大学 | Crop transpiration prediction method based on improved extreme learning machine |
| CN109345508A (en) * | 2018-08-31 | 2019-02-15 | 北京航空航天大学 | A kind of Assessing Standards For Skeletal method based on two stages neural network |
| CN109359741A (en) * | 2018-09-27 | 2019-02-19 | 华南师范大学 | A kind of wastewater treatment influent quality timing variations intelligent Forecasting |
Non-Patent Citations (3)
| Title |
|---|
| 潘玉民等: "基于量子粒子群优化的小波神经网络预测模型", 《信息与控制》 * |
| 靳然等: "基于小波神经网络的麦蚜发生程度预测模型", 《昆虫学报》 * |
| 魏周会等: "《作物产量预测的时间序列神经网络模型》", 《节水灌溉》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113158470A (en) * | 2020-11-25 | 2021-07-23 | 中国石油大学(华东) | Oil reservoir automatic history fitting system and method based on transfer learning |
| CN113379188A (en) * | 2021-05-06 | 2021-09-10 | 贵州省烟草公司贵阳市公司 | Tobacco crop rotation planting method and system based on Internet of things |
| CN113439520A (en) * | 2021-07-21 | 2021-09-28 | 中国农业科学院农业环境与可持续发展研究所 | Intelligent decision-making method and system for crop irrigation and fertilization |
| CN115907366A (en) * | 2022-11-15 | 2023-04-04 | 淮阴工学院 | Agricultural product growth environment optimal regulation and control method and equipment based on flamingo algorithm |
| CN115907366B (en) * | 2022-11-15 | 2023-09-22 | 淮阴工学院 | Agricultural product growth environment optimal regulation and control method and equipment based on flamingo algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110414115B (en) | 2023-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110414115A (en) | A kind of wavelet neural network tomato yield prediction technique based on genetic algorithm | |
| CN106373022B (en) | BP-GA-based greenhouse crop planting efficiency condition optimization method and system | |
| Priya et al. | An IoT based gradient descent approach for precision crop suggestion using MLP | |
| CN110163254A (en) | A kind of cucumber green house yield intelligent Forecasting device based on recurrent neural network | |
| CN110472840A (en) | A kind of agricultural water conservancy dispatching method and system based on nerual network technique | |
| CN107145941B (en) | Method for dynamically acquiring light demand quantity in real time based on optimal light quality and photon flux density | |
| CN104376142A (en) | Rice plant type quantitative control method integrating crop virtual growth model | |
| Hu et al. | Model for tomato photosynthetic rate based on neural network with genetic algorithm | |
| CN110119086A (en) | A kind of tomato greenhouse environmental parameter intelligent monitoring device based on ANFIS neural network | |
| CN111915062A (en) | Greenhouse crop water demand regulation and control method with water utilization rate and photosynthetic rate coordinated | |
| CN113221447A (en) | Soil humidity prediction method for optimizing BP neural network based on improved genetic algorithm | |
| Valenzuela et al. | Pre-harvest factors optimization using genetic algorithm for lettuce | |
| CN117575094B (en) | Crop yield prediction and optimization method and device based on digital twin | |
| Lyu et al. | Multi-objective winter wheat irrigation strategies optimization based on coupling AquaCrop-OSPy and NSGA-III: A case study in Yangling, China | |
| An et al. | A simulator-based planning framework for optimizing autonomous greenhouse control strategy | |
| Ding et al. | Genetic algorithm based approach to optimize phenotypical traits of virtual rice | |
| CN105913326B (en) | Constrain the Crop growing stage model Cultivar parameter optimization method of sex knowledge and the tactful Genetic Algorithm Fusion of elite individual | |
| CN113625806B (en) | Optimal-efficiency greenhouse plant regulation and control method for carbon neutralization demand | |
| CN114282702A (en) | Soil conditioning time sequence prediction method and system based on IGA-BP neural network | |
| Chen et al. | A water-saving irrigation decision-making model for greenhouse tomatoes based on genetic optimization TS fuzzy neural network | |
| Chen et al. | Multi-objective optimal regulation model and system based on whole plant photosynthesis and light use efficiency of lettuce | |
| Su et al. | Greenhouse climate setpoint optimization: An online decision strategy | |
| Gao et al. | Application of Artificial Intelligence System Design Based on Genetic Algorithm In Horticultural Cultivation | |
| CN109472320A (en) | Crop growing stage model Cultivar parameter automatically corrects frame under condition of uncertainty | |
| CN115226516B (en) | Cooperative regulation and control method for facility light and carbon dioxide environment |
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 |
Granted publication date: 20230120 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |