CN109934400A - Based on the collection rain readjust-loss water demand of crop prediction technique for improving neural network - Google Patents
Based on the collection rain readjust-loss water demand of crop prediction technique for improving neural network Download PDFInfo
- Publication number
- CN109934400A CN109934400A CN201910173728.7A CN201910173728A CN109934400A CN 109934400 A CN109934400 A CN 109934400A CN 201910173728 A CN201910173728 A CN 201910173728A CN 109934400 A CN109934400 A CN 109934400A
- Authority
- CN
- China
- Prior art keywords
- crop
- neural network
- factor
- water
- water demand
- 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
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
- Y02A20/152—Water filtration
Abstract
The collection rain readjust-loss water demand of crop prediction technique based on Modified Elman Neural Network that the invention discloses a kind of, this method chooses the input factor for influencing more apparent factor as prediction model to the water demand of crop respectively in crop different bearing stage, then it is predicted using the Elman neural network with dynamic modeling ability, simultaneously, the best connection weight and threshold value that Elman neural network is chosen using genetic algorithm, obtain the Elman neural network of state optimization.Prediction model of the invention can effectively predict the water demand of crop, and readjust-loss trickle irrigation is carried out while making full use of rainfall, realize multiple water-saving, save water resource to a greater extent.
Description
Technical field
The collection rain readjust-loss water demand of crop prediction technique based on Modified Elman Neural Network that the present invention relates to a kind of, belongs to
Water resources management and technical field of communication network.
Background technique
China is the large agricultural country of a water resource critical shortage.It is sustainable that the scarcity of water resource has become restriction China
The bottleneck factor of development.In each water-using sector, agricultural is the maximum industry of water resources consumption, is occupied an leading position with water,
Arid area, irrigation water capacity account for 90% of Water Consumption in Agriculture or so, China's irrigation water capacity account for about the 80% of total water consumption with
On.According to scholarly forecast, to the year two thousand twenty front and back, China's agricultural water shortage will be up to 1000 × 108m3, water shortage situation will be severeer.Cause
This, Spreading Water-saving Irrigation Technology acts out one's plan with water, carries out intelligent irrigation, and the situation for accelerating agricultural modernization step seems ten
Divide urgent.Research Study on Crop Water Requirement Rules seeks suitable water conservation mode, improves the efficiency of water application of crops, realizes precision irrigation,
It is particularly significant to the shortage of alleviation agricultural water and water resources crisis, guarantee China's grain security, ecological safety and society can be held
Supervention exhibition is of great significance.
Precision irrigation is the highest goal of water-saving irrigation, and meaning is accurately to estimate to pour water necessary to crop growth
Amount, and accurately equably poured into this water in crop root layer soil using efficient water-saving irrigation technique.Therefore, it is desirable to
Rational utilization of water resources realizes Precision Irrigation, it is necessary to scientific and reasonable prediction is carried out to crop water requirement.
Currently, researcher largely biases toward the base of conventional tillage mode in the research of water demand of crop prediction technique
In the optimization innovation of the prediction algorithm of meteorologic factor, edphic factor and plant physiology information are not introduced, according to crop growth rank
Section chooses the prediction because usually carrying out the water demand of crop of the Different Effects water demand of crop.However the water demand of crop and plant growth
The factor relations such as stage, plant physiology are close.Choose different influence factors according to the crop growth stage, combined high precision it is pre-
Method of determining and calculating could more accurately predict the water demand of crop.
For prediction water demand of crop model, existing technical solution is: Zhang Bing et al. is for BP neural network convergence speed
Slow disadvantage is spent, L-M optimization algorithm BP neural network is applied in Model for predicting crop water requirements, and to tennessee,USA
The green pepper water demand of crop in university, laboratory, plateau has carried out prediction instance analysis.Although the program improves the convergence speed of network
Degree, can obtain smaller training error, but its input sample is only meteorological data, and be not introduced into plant physiology information, also not
Different influence factors is chosen according to the crop growth stage to carry out water demand of crop prediction.Shang Zhigen et al. proposes random gloomy
The prediction water demand of crop model of woods and MLP Artificial neural network ensemble, the model can get better precision of prediction, but mode input
Sample is still only meteorological data, and also the factor for influencing the water demand of crop is not analyzed and chosen according to the crop growth stage.
Although Dahua Tang et al. using genetic algorithm optimization the neural network prediction water demand of crop when, input data consider
Two plant physiology indexs of leaf area index and plant height in addition to meteorologic factor, but do not consider plant physiology information
In crop canopy temperature.
For plant, blade has multi-functional, and the temperature of blade is the important indicator of plant physiology sensing, can be with
Reflect the water regime of crop, a large number of studies show that, crop canopy temperature is a good index for reflecting crop water status,
And the water regime of crop and the water demand of crop are closely bound up.When the acquisition of crop canopy temperature also overcomes other parameters measurement
The larger and time-consuming disadvantage of existing sampling error.Therefore, it is necessary to consider to make for the input factor of prediction water demand of crop model
Object canopy surface temperature.In addition, the planting patterns in above-mentioned all documents is not the collection rain readjust-loss trickle irrigation for crop yet
Mode.And it is all be related in the prediction model of the water demand of crop not according to the crop growth stage respectively to influence crop
Water requirement factor is analyzed and is chosen.
Summary of the invention
It is analyzed, the method that the discovery prior art has the following disadvantages: 1) most of prediction water demand of crop mainly considers
Meteorologic factor lacks the input factor for considering crop canopy temperature as plant physiology information.2) lack forecast set rain readjust-loss
The water demand of crop model of trickle irrigation mode.3) lack to analyze respectively according to the crop growth stage and choose different input factor prediction works
The model of object water requirement.
Therefore, the present invention proposes that a kind of collection rain readjust-loss cropping pattern descends the crop water based on Modified Elman Neural Network
Prediction technique is measured, the input factor of this method is by meteorologic factor and crop factors composition.Wherein, meteorologic factor mainly considers day
The data such as temperature on average, sunshine time, per day relative humidity, vapour pressure；Crop factor mainly considers the leaf area of crop
Index, plant height, canopy surface temperature.This method is first, in accordance with the crop growth stage, according to experimental data over the years to influence crop
The factor of water requirement is analyzed, then chosen respectively in crop different bearing stage on the water demand of crop influence it is more apparent because
The input factor of the element as prediction model.Then data are normalized, then using has dynamic modeling ability
Elman neural network predicted, meanwhile, Elman neural network is chosen using genetic algorithm (Genetic Algorithm)
Best connection weight and threshold value obtain the Elman neural network of state optimization.
The present invention uses following technical solutions:
A kind of water demand of crop prediction technique based on Modified Elman Neural Network, this method, which uses, has dynamic modeling
The Elman neural network model of ability is predicted, it is characterised in that: is chosen respectively in crop different bearing stage different
The input factor of the influence factor as prediction model；Using genetic algorithm choose Elman neural network best connection weight and
Threshold value.
The factor is inputted by meteorologic factor and crop factors composition, meteorologic factor includes daily mean temperature, sunshine time, puts down day
Equal relative humidity, daily maximum temperature；Crop factor includes leaf area index, plant height, canopy surface temperature.
When crop is green pepper, growing stage includes seedling stage, the phase of bearing fruit of blooming, result peak period and result later period.According to history
Data, influence factor and the water demand of crop to each growing stage water demand of crop of green pepper carry out correlation analysis, according to correlation
Property sequence choose different bearing stage the input factor.
Preferably, 5 input factors are respectively chosen after seedling stage, the phase of bearing fruit of blooming, result peak period and result, respectively seedling stage is selected
Take daily maximum temperature, daily mean temperature, sunshine time, canopy surface temperature, leaf area index；Phase of bearing fruit of blooming chooses per day gas
Temperature, daily maximum temperature, leaf area index, plant height, sunshine time；As a result when peak period chooses daily mean temperature, canopy surface temperature, sunshine
Number, leaf area index, plant height；As a result the later period chooses daily mean temperature, canopy surface temperature, daily maximum temperature, sunshine time, per day
Relative humidity.
When being predicted, all sample datas are normalized, are transformed into 0-1 section using following formula:
x'p=(xp-xmin)/(xmax-xmin)
In formula, xp(p=1,2 ..., P) is sample data, xmax=max { xp},xmin=min { xp}；
After Elman neural network model, reduction calculating is carried out to network output data, restores actual value:
xyp=x'yp(xy max-xy min)+xy min
In formula, xypIt is after being calculated for reduction as a result, i.e. actual value；x'ypFor the prediction numerical value of network output, xy maxFor net
Network exports the maximum value of prediction result, xy minThe minimum value of prediction result is exported for network.
Training algorithm using the gradient decreasing function of momentum and adaptive learning rate as Elman neural network, is adopted
It is the transmission function of middle layer neuron with S function, it is 1 dimensional vector that output vector, which is the water demand of crop, and output layer is only one
Neuron uses linear transmission function for the transmission function of output neuron.It is defeated using genetic algorithm optimization Elman neural network
Enter threshold of the layer to hidden layer, hidden layer to output layer and hidden layer to the connection weight and hidden layer and output layer for accepting layer
Value.
Preferably, to the crop of collection rain readjust-loss drip irrigation in result later period moderate water deficit treatment, other breeding times are normal
The case where sufficiently pouring water carries out Water Demand Prediction.
(3) beneficial effect
Using Hebei University Of Engineering, experimental field data emulate the validity of the proposed GA-Elman method of the present invention
Verifying, the results showed that, prediction model of the invention can carry out the water demand of crop under collection rain readjust-loss trickle irrigation mode effective
Prediction, performance be substantially better than the crop time of infertility only input meteorologic factor or the crop time of infertility only choose fixed input because
Element, and predicted also superior to single use Elman neural network.In addition, being carried out while making full use of rainfall
Readjust-loss trickle irrigation, realizes multiple water-saving, has saved water resource to a greater extent.
Detailed description of the invention
Fig. 1 is collection rain readjust-loss cropping pattern schematic diagram；
Fig. 2 is collection rain readjust-loss trickle irrigation plantation schematic top plan view；
Fig. 3 is the green pepper water demand of crop prediction technique flow chart based on Modified Elman Neural Network；And
Fig. 4 is Elman neural network model figure.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with specific embodiment, and reference
Attached drawing, the present invention is described in more detail.
For green pepper crop of the present invention with collection rain readjust-loss trickle irrigation, collection rain mode uses " curvature of the spinal column formula " furrow collection rain, on ridge
It covers unregistered land film (film thickness about 0.012mm), ditch ridge ratio is set as 60:40.And it is arranged that tradition is flat to make (B is used to indicate) as comparing.
" Guan Yihang ", dripping end flow use 2.2L/h, water dropper spacing 30cm.Plant line-spacing 50cm, spacing in the rows 30cm.Collect rain readjust-loss plantation
Pattern diagram is as shown in Fig. 1, Fig. 2.The entire breeding time of green pepper is divided into 4 stages, is respectively as follows: seedling stage, blooms and bear fruit
Phase, result peak period, result later period.Each breeding time is respectively set two kinds of water deficit treatments, respectively moderate depletion (indicated with M:
60%~70% field capacity) and slight depletion (being indicated with L: 70%~80% field capacity), control group is complete
Breeding time normally sufficiently pours water and (is indicated with N).Plant that (R indicates that collection rain, B indicate to put down with flat in addition, again planting drip irrigation with catchment of rainfall
Make) it compares, specifically it is shown in Table 1.
The water deficit of 1 different growing Different Ways of Planting of table is handled
By obtaining to data collected by site test: under considerable moisture readjust-loss processing, drip irrigation with catchment of rainfall plants green pepper
Growing way and yield are substantially better than flat work.And readjust-loss cropping pattern is obviously than sufficiently irrigating water conservation, and significantly affects crop and need
Water.By variance analysis, green pepper yield, quality, irrigation water utilization efficiency etc. are comprehensively considered, choose collection rain cropping pattern
Under green pepper result later period moderate water deficit treatment, the planting patterns that other breeding times normally sufficiently pour water is best, fills than flat
Divide to irrigate and save water about 30% or more, and green pepper yield can be significantly improved, while improving green pepper Vc content.
As shown in figure 3, the green pepper water demand of crop prediction technique based on Modified Elman Neural Network is mainly inputted by data
And pretreatment and prediction algorithm and result export two parts composition.Meteorologic factor, crop factor and edphic factor all can be to crops
Water requirement affects, selection when predicted for the green pepper water demand of crop of collection rain readjust-loss trickle irrigation this present invention
The data such as daily mean temperature, sunshine time, per day relative humidity, daily maximum temperature in meteorologic factor, in crop factor
Input data as prediction model of leaf area index, plant height, canopy surface temperature.It is utilized within green pepper crop each breeding time
Origin software is analyzed respectively water demand of crop factor is influenced.
In prediction algorithm module, the connection weight of Elman neural network and each layer threshold value are carried out using genetic algorithm
Then optimum choice is trained improved Elman neural network with the data that data input and preprocessing module obtain
And test, it was demonstrated that the validity of algorithm.
The input of 1 data and pretreatment
The input of 1.1 data
Usually prediction most of water demand of crop model only considers meteorologic factor, rare while considering meteorologic factor and work
Object is because of the green pepper water demand of crop under usually forecast set rain readjust-loss condition of drip irrigation.Meanwhile not distinguishing according to the crop growth stage
Analysis, selection influence the input factor of the water demand of crop.Therefore, the prediction green pepper water demand of crop is descended to collection rain readjust-loss condition of drip irrigation
Input data for, consider meteorologic factor while consider crop factor, according to growing stage analysis, selection input factor pair
Realize that more accurate water demand of crop prediction is very necessary.Green pepper growing stage division in Hebei Handan District is shown in Table 2 institutes
Show.
The division (Handan in Hebei province) of 2 green pepper growing stage of table
The present invention utilizes test data of the Origin software according to Handan District in recent years, to each growing stage crop of green pepper
The influence factor and the water demand of crop of water requirement carry out correlation analysis.It was found that this area influences the factor of the green pepper water demand of crop
According to the difference in crop growth stage, influence factor also can be different from the correlation size of the water demand of crop.In collection rain readjust-loss drop
Fill under cropping pattern, in green pepper different bearing stage, influence the correlation of water demand of crop factor and the green pepper water demand of crop from
Small sequence is arrived greatly to be shown in Table 3.
The correlation analysis of table 3 green pepper crop influence factor and the water demand of crop
1.2 data prediction
To guarantee that the nonlinear interaction of neuron will be normalized for the learning sample of numeric type, not lose one
As property, be normalized with formula (1), all sample datas be transformed into 0-1 section.In this way, can take lesser
The problem of connection weight W of the number as network, calculation overflow will not occur for network query function.
x′p=(xp-xmin)/(xmax-xmin) (1)
In formula, xp(p=1,2 ..., P) is sample data, xmax=max { xp},xmin=min { xp}。
2 prediction algorithms (Elman neural network improves module) and result output
2.1Elman neural network
Elman neural network is made of input layer, hidden layer, undertaking layer and four layers of output layer, to be held because which increase one
Layer is connect, as step delay operator, to achieve the purpose that memory, thus make network that there is the ability for adapting to time-varying characteristics, it can be straight
The reversed network characteristic for reflecting dynamic process, achievees the purpose that dynamic modeling, and data biggish for fluctuation have preferable
Prediction effect.Elman neural network model of the present invention is as shown in Figure 4.
Although Elman neural network promotes traditional neural network performance, in the design process, still
It is faced with the problem of all neural networks all exist: the selection of training algorithm, transfer function, network structure and connection weight.This
Training algorithm of the invention using the gradient decreasing function of momentum and adaptive learning rate as Elman neural network, using normal
S function is the transmission function of middle layer neuron, and output vector is the water demand of crop, is 1 dimensional vector, and output layer is only one
A neuron uses linear transmission function for the transmission function of output neuron.The structure of Elman neural network is (including implicit
Number, each node layer number etc. of layer) and connection weight selection be to the performance of whole network it is highly important, the present invention adopts
The connection weight of Elman neural network and each layer threshold value are in optimized selection with genetic algorithm.
2.2 have Nonlinear Mapping, self study etc. excellent based on genetic algorithm modified Delphi approach neural network
Performance, but this must be set up on using suitable connection weight and threshold value, the purpose of this part is selected using genetic algorithm
Select the network connection weight and threshold value for making Elman neural network best performance.Genetic algorithm is a kind of by simulation natural evolution
The method of process searches optimal solution has efficient ability of searching optimum, shows very in solving complicated optimum problem
Excellent performance and very high efficiency.
Genetic algorithm optimization Elman neural network input layer to hidden layer, hidden layer to output layer and hidden layer to accept
The connection weight of layer and the threshold value of hidden layer and output layer.The parameter setting of genetic algorithm is as shown in table 4.Elman nerve net
The initial parameter setting of network is as shown in table 5.
The setting of 4 genetic algorithm key parameter of table
The setting of 5 Elman neural network initial parameter of table
The output of 2.3 results
Reduction calculating should be carried out to the output data of network, restore actual value.Reductive agent is that the inverse of formula (1) calculated
Journey uses formula (2):
xyp=x'yp(xy max-xy min)+xy min (2)
In formula, xypIt is after being calculated for reduction as a result, i.e. actual value.x'ypFor the prediction numerical value of network output, xy maxFor net
Network exports the maximum value of prediction result, xy minThe minimum value of prediction result is exported for network.The present invention uses Handan in Hebei province
Experimental field the meteorological data of automatic weather station acquisition and the experimental data acquired over the years are verified for Hebei University Of Engineering,
In include green pepper Crop Information.For the present invention, by long term test, to the green pepper yield under different year different disposal, water
Point utilization efficiency and Vc content carry out variance analysis respectively and obtain, different year experimental data otherness is at test
Reason, rather than year border difference.It moreover has been found that yield, quality and the water use efficiency of furrow collection rain green pepper significantly improve.By green pepper
Water use efficiency, yield, quality (Vc content) as evaluation high yield, high-quality, Water Saving Irrigation System assessment indicator system,
Overall merit is carried out to the production status of the lower green pepper of different in moisture processing, using Principal Component Analysis to the principal component of each factor
It extracts respectively, carries out overall merit, the best test process technical solution of selection is: after ditch ridge is than the result for 60:40
The technical solution of interim degree depletion.Readjust-loss is carried out in the case where making full use of rainwater, the scheme and tradition of selection are flat to be made sufficiently
It pours water and compares, 30% or more can be saved water.
Green pepper canopy surface temperature is had chosen in input crop factor as the input factor, by correlation analysis, in green pepper
Crops seedling stage and result peak period, the result later period greatly, therefore it is wanted with the correlation specific leaf area index and plant height of the water demand of crop
It is necessary to introduce its input factors as the water demand of crop.
Water demand of crop emulation is carried out using technical solution of the genetic algorithm modified Delphi approach method to selection
Analysis.Using 2015-2017 data as the training sample of Elman neural network, data in 2018 are as test specimens
This.According to the crop growth stage, respectively to water demand of crop factor progress correlation analysis is influenced, discovery is different in green pepper crop
In breeding time, influences water demand of crop factor and changed with water demand of crop correlation size, therefore, in different bearing stage
It is interior to choose the different prediction model input factors, the precision of Model for predicting crop water requirements can be improved, is conducive to more accurately pre-
Survey the water demand of crop.The present invention utilizes root-mean-square error (Root Mean Square Error, RMSE), mean absolute error
(mean absolute error, MAE) and the coefficient of determination (coefficient of determination, R2) to each model
Prediction result is analyzed, PiFor predicted value, OiFor observation,To predict mean value,To observe mean value, N is observation number
(sample size in test set).Interpretation of result is as shown in table 6 (since the crop time of infertility only considers the model R of meteorologic factor2
It is lower, about 0.8 or so, therefore, table 6 omits the analysis to it).
As can be seen from Table 6, it (is chosen from big to small according to correlation defeated when inputting the input factor of equal amount
Enter the factor, choose 5 input factors), the time of infertility considers the Elman neural network precision and degree of fitting of identical input factor
Minimum, the followed by time of infertility GA-Elman neural network that only considers the identical input factor, precision and degree of fitting is higher is
Consider the Elman neural network of the input factor respectively according to the crop growth stage.Precision and degree of fitting are highest for according to crop
Growing stage considers to input the GA-Elman neural network of the factor respectively.As can be seen that needle is for the purpose of the present invention, in different bearing
The influence of the different input factor pair model accuracies of selection is greater than only Modified Elman Neural Network and generates to model accuracy in phase
Influence.Method therefor of the present invention has had reached a satisfactory result.
The analysis of 6 prediction result of table
For the green pepper water demand of crop for lacking the canopy surface temperature in consideration crop factor in the case where collecting rain readjust-loss drip irrigation
Prediction model, the present invention other than considering meteorologic factor, joined in crop factor except leaf area index, strain in input data
Crop canopy temperature except height；And on this basis, according to the crop growth stage, according to correlation analysis as a result, selecting respectively
Different influence water demand of crop factors is taken, the green pepper under collection rain readjust-loss mode is carried out using modified Delphi approach and is made
Object Water Demand Prediction.
Particular embodiments described above has carried out further in detail the purpose of the present invention, technical scheme and beneficial effects
It describes in detail bright, it should be understood that the above is only a specific embodiment of the present invention, is not intended to restrict the invention, it is all
Within the spirit and principles in the present invention, any modification, equivalent substitution, improvement and etc. done, including but not limited to: predicted time
The change of scale (when, the moon, year etc.), increasing or reducing for prediction data, predicts the change in place, predicts the variation of crop species
Deng should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of water demand of crop prediction technique based on Modified Elman Neural Network, this method, which uses, has dynamic modeling energy
The Elman neural network model of power is predicted, it is characterised in that: chooses different shadows respectively in crop different bearing stage
The input factor of the factor of sound as prediction model；The best connection weight and threshold of Elman neural network are chosen using genetic algorithm
Value.
2. the method according to claim 1, the input factor is by meteorologic factor and crop factors composition.
3. method according to claim 2, the meteorologic factor include daily mean temperature, sunshine time, per day relative humidity,
Daily maximum temperature；The crop factor includes leaf area index, plant height, canopy surface temperature.
4. according to the method in claim 3, crop be green pepper, growing stage include seedling stage, the phase of bearing fruit of blooming, result peak period and
As a result the later period.
5. method according to claim 4, according to historical data, to the influence factor of each growing stage water demand of crop of green pepper with
The water demand of crop carries out correlation analysis, and the input factor of different bearing stage is chosen according to relevance ranking.
6. all sample datas are normalized in method according to claim 5, it is transformed into 0-1 section:
x'p=(xp-xmin)/(xmax-xmin)
In formula, xp(p=1,2 ..., P) is sample data, xmax=max { xp},xmin=min { xp}；
After Elman neural network model, reduction calculating is carried out to network output data, restores actual value:
xyp=x'yp(xymax-xymin)+xymin
In formula, xypIt is after being calculated for reduction as a result, i.e. actual value；x'ypFor the prediction numerical value of network output, xymaxFor network output
The maximum value of prediction result, xyminThe minimum value of prediction result is exported for network.
7. method according to claim 6 is made using the present invention using the gradient decreasing function of momentum and adaptive learning rate
For the training algorithm of Elman neural network, use S function for the transmission function of middle layer neuron, output vector needs for crop
Water is 1 dimensional vector, and output layer is only a neuron, uses linear transmission function for the transmission function of output neuron.
8. method according to claim 7, genetic algorithm optimization Elman neural network input layer to hidden layer, hidden layer are to defeated
The threshold value of layer and hidden layer to the connection weight and hidden layer and output layer for accepting layer out.
9. method according to claim 8, seedling stage, the phase of bearing fruit of blooming, result peak period and result later period respectively choose 5 inputs because
Son, respectively seedling stage choose daily maximum temperature, daily mean temperature, sunshine time, canopy surface temperature, leaf area index；It blooms and bears fruit
Phase chooses daily mean temperature, daily maximum temperature, leaf area index, plant height, sunshine time；As a result peak period selection daily mean temperature,
Canopy surface temperature, sunshine time, leaf area index, plant height；As a result the later period choose daily mean temperature, canopy surface temperature, daily maximum temperature,
Sunshine time, per day relative humidity.
10. according to any of the above-described the method for claim, to the crop of collection rain readjust-loss drip irrigation in result later period moderate depletion
The case where processing, other breeding times normally sufficiently pour water, carries out Water Demand Prediction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910173728.7A CN109934400B (en) | 2019-03-08 | 2019-03-08 | Rain collecting, regulating and deficiency crop water demand prediction method based on improved neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910173728.7A CN109934400B (en) | 2019-03-08 | 2019-03-08 | Rain collecting, regulating and deficiency crop water demand prediction method based on improved neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109934400A true CN109934400A (en) | 2019-06-25 |
| CN109934400B CN109934400B (en) | 2023-05-16 |
Family
ID=66986782
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910173728.7A Active CN109934400B (en) | 2019-03-08 | 2019-03-08 | Rain collecting, regulating and deficiency crop water demand prediction method based on improved neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109934400B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112966221A (en) * | 2021-03-04 | 2021-06-15 | 江西省水利科学院 | Method for converting total water consumption of assessment year in southern rich water region |
| CN114201923A (en) * | 2022-02-16 | 2022-03-18 | 中国地质调查局水文地质环境地质调查中心 | Processing method and device of landslide displacement prediction model |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101169627A (en) * | 2007-11-27 | 2008-04-30 | 中国水利水电科学研究院 | On-line crop water stress irrigation decision monitoring system |
| US20120109387A1 (en) * | 2009-04-06 | 2012-05-03 | Smartfield, Inc. | Remote analysis and correction of crop condition |
| US20120123817A1 (en) * | 2010-11-12 | 2012-05-17 | Smartfield, Inc. | Agricultural management using biological signals |
| CN102565150A (en) * | 2012-01-13 | 2012-07-11 | 北京盈胜泰科技术有限公司 | Detection device and monitoring system for soil relative humidity and soil fertility change |
| CN103886392A (en) * | 2014-03-26 | 2014-06-25 | 中国农业科学院农田灌溉研究所 | Winter wheat water consumption predicting method based on weather forecast information |
| CN106212217A (en) * | 2016-08-02 | 2016-12-14 | 浙江恒合景观规划设计有限公司 | Intelligent precision irrigation control system |
| US20170038749A1 (en) * | 2015-08-05 | 2017-02-09 | Iteris, Inc. | Customized land surface modeling for irrigation decision support in a crop and agronomic advisory service in precision agriculture |
| CN106707757A (en) * | 2017-01-23 | 2017-05-24 | 中国农业大学 | Dynamic regulation and control method and system for irrigation time |
| WO2017106962A1 (en) * | 2015-11-17 | 2017-06-29 | Universite Laval | System and method for real-time water management |
| CN107025505A (en) * | 2017-04-25 | 2017-08-08 | 无锡中科智能农业发展有限责任公司 | A kind of paddy water requirement prediction method based on principal component analysis and neutral net |
| CN107302875A (en) * | 2016-04-20 | 2017-10-31 | 青岛农业大学 | A kind of water-saving irrigation method based on the crucial moisture information of soil-plant system |
| CN107358294A (en) * | 2017-07-21 | 2017-11-17 | 河北工程大学 | A kind of water demand prediction method based on Elman neutral nets |
| CN107507096A (en) * | 2017-08-14 | 2017-12-22 | 陈守敏 | A kind of irrigation rainwater harvesting cellar management information system based on big data |
| CN107945042A (en) * | 2017-11-29 | 2018-04-20 | 上海华维节水灌溉股份有限公司 | A kind of plant growth irrigation decision control system |
-
2019
- 2019-03-08 CN CN201910173728.7A patent/CN109934400B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101169627A (en) * | 2007-11-27 | 2008-04-30 | 中国水利水电科学研究院 | On-line crop water stress irrigation decision monitoring system |
| US20120109387A1 (en) * | 2009-04-06 | 2012-05-03 | Smartfield, Inc. | Remote analysis and correction of crop condition |
| US20120123817A1 (en) * | 2010-11-12 | 2012-05-17 | Smartfield, Inc. | Agricultural management using biological signals |
| CN102565150A (en) * | 2012-01-13 | 2012-07-11 | 北京盈胜泰科技术有限公司 | Detection device and monitoring system for soil relative humidity and soil fertility change |
| CN103886392A (en) * | 2014-03-26 | 2014-06-25 | 中国农业科学院农田灌溉研究所 | Winter wheat water consumption predicting method based on weather forecast information |
| US20170038749A1 (en) * | 2015-08-05 | 2017-02-09 | Iteris, Inc. | Customized land surface modeling for irrigation decision support in a crop and agronomic advisory service in precision agriculture |
| WO2017106962A1 (en) * | 2015-11-17 | 2017-06-29 | Universite Laval | System and method for real-time water management |
| CN107302875A (en) * | 2016-04-20 | 2017-10-31 | 青岛农业大学 | A kind of water-saving irrigation method based on the crucial moisture information of soil-plant system |
| CN106212217A (en) * | 2016-08-02 | 2016-12-14 | 浙江恒合景观规划设计有限公司 | Intelligent precision irrigation control system |
| CN106707757A (en) * | 2017-01-23 | 2017-05-24 | 中国农业大学 | Dynamic regulation and control method and system for irrigation time |
| CN107025505A (en) * | 2017-04-25 | 2017-08-08 | 无锡中科智能农业发展有限责任公司 | A kind of paddy water requirement prediction method based on principal component analysis and neutral net |
| CN107358294A (en) * | 2017-07-21 | 2017-11-17 | 河北工程大学 | A kind of water demand prediction method based on Elman neutral nets |
| CN107507096A (en) * | 2017-08-14 | 2017-12-22 | 陈守敏 | A kind of irrigation rainwater harvesting cellar management information system based on big data |
| CN107945042A (en) * | 2017-11-29 | 2018-04-20 | 上海华维节水灌溉股份有限公司 | A kind of plant growth irrigation decision control system |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112966221A (en) * | 2021-03-04 | 2021-06-15 | 江西省水利科学院 | Method for converting total water consumption of assessment year in southern rich water region |
| CN114201923A (en) * | 2022-02-16 | 2022-03-18 | 中国地质调查局水文地质环境地质调查中心 | Processing method and device of landslide displacement prediction model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109934400B (en) | 2023-05-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Effects of irrigation and straw mulching on microclimate characteristics and water use efficiency of winter wheat in north chinas | |
| CN102524024B (en) | Crop irrigation system based on computer vision | |
| CN101916337B (en) | Method for dynamically predicting potential productivity of paddy rice based on geographical information system | |
| CN104521699A (en) | Field intelligent irrigation on-line control management method | |
| CN108781926A (en) | Greenhouse irrigation system based on neural network prediction and method | |
| CN110069032A (en) | A kind of eggplant greenhouse intelligent checking system based on wavelet neural network | |
| CN107466816A (en) | A kind of irrigation method based on dynamic multilayer extreme learning machine | |
| CN109934400A (en) | Based on the collection rain readjust-loss water demand of crop prediction technique for improving neural network | |
| CN110490228A (en) | A kind of Hydro-Model Parameter Calibration Technology dynamic rating method based on CPP | |
| CN106530104A (en) | Method and device for evaluating production suitability of planting area | |
| Wang et al. | Cotton growth model under drip irrigation with film mulching: A case study of Xinjiang, China | |
| CN101276446B (en) | Method for measuring and calculating area crops water demand quantity | |
| CN112364710B (en) | Plant electric signal classification and identification method based on deep learning algorithm | |
| CN109657999A (en) | A kind of mixing dynamic agricultural machinery cooperation efficiency rating method based on SU-DEA Yu Malmquist index | |
| CN115081952A (en) | Regional agricultural drought identification method and system | |
| CN107330541A (en) | A kind of west and central Pacific ocean stripped tuna Predicting Center Fishery method | |
| Macabiog et al. | Soil moisture and rain prediction based irrigation controller for the strawberry farm of La Trinidad, Benguet | |
| Terjung et al. | Actual and potential yield for rainfed and irrigated wheat in China | |
| Zhu et al. | Prediction of continuous rain disaster in Henan Province based on Markov model | |
| Xing et al. | Estimating daily kiwifruit evapotranspiration under regulated deficit irrigation strategy using optimized surface resistance based model | |
| Zhang et al. | ANALYSIS OF DROUGHT-FLOOD ABRUPT ALTERNATION OF TOBACCO BASED ON PRECIPITATION AND SOIL PONDING IN SIUWEN CHINA. | |
| Terjung et al. | Potential paddy rice yields for rainfed and irrigated agriculture in China and Korea | |
| Ma et al. | The characteristics of climate change and adaptability assessment of migratory bird habitats in Wolonghu wetlands | |
| CN117391406B (en) | Irrigation area water resource optimal allocation method considering crop reseeding | |
| Hu et al. | Study on Comprehensive Efficiency Evaluation of Farmland Irrigation Water Based on Compound Fuzzy Mathematical 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 |