CN117333321A - Agricultural irrigation water consumption estimation method, system and medium based on machine learning - Google Patents
Agricultural irrigation water consumption estimation method, system and medium based on machine learning Download PDFInfo
- Publication number
- CN117333321A CN117333321A CN202311260303.2A CN202311260303A CN117333321A CN 117333321 A CN117333321 A CN 117333321A CN 202311260303 A CN202311260303 A CN 202311260303A CN 117333321 A CN117333321 A CN 117333321A
- Authority
- CN
- China
- Prior art keywords
- irrigation water
- water consumption
- agricultural irrigation
- machine learning
- model
- 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
- 239000003621 irrigation water Substances 0.000 title claims abstract description 153
- 238000010801 machine learning Methods 0.000 title claims abstract description 93
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 28
- 238000011160 research Methods 0.000 claims abstract description 27
- 238000003973 irrigation Methods 0.000 claims abstract description 25
- 230000002262 irrigation Effects 0.000 claims abstract description 25
- 238000010206 sensitivity analysis Methods 0.000 claims abstract description 21
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 36
- 239000002689 soil Substances 0.000 claims description 20
- 238000010276 construction Methods 0.000 claims description 15
- 238000007637 random forest analysis Methods 0.000 claims description 10
- 238000011156 evaluation Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 8
- 238000012544 monitoring process Methods 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000013473 artificial intelligence Methods 0.000 claims description 5
- 230000006870 function Effects 0.000 claims description 5
- 238000001556 precipitation Methods 0.000 claims description 5
- 230000005855 radiation Effects 0.000 claims description 5
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 238000013145 classification model Methods 0.000 claims description 4
- 238000012706 support-vector machine Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 claims description 3
- 238000011835 investigation Methods 0.000 claims description 3
- 230000007787 long-term memory Effects 0.000 claims description 2
- 238000003491 array Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006735 deficit Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- GVVPGTZRZFNKDS-JXMROGBWSA-N geranyl diphosphate Chemical compound CC(C)=CCC\C(C)=C\CO[P@](O)(=O)OP(O)(O)=O GVVPGTZRZFNKDS-JXMROGBWSA-N 0.000 description 1
- 230000015654 memory Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000002352 surface water Substances 0.000 description 1
- 238000013014 water-saving technology Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- 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
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Software Systems (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Marketing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Primary Health Care (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Water Supply & Treatment (AREA)
- Quality & Reliability (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Evolutionary Biology (AREA)
- Public Health (AREA)
- Operations Research (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Agronomy & Crop Science (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a machine learning-based agricultural irrigation water consumption estimation method, a machine learning-based agricultural irrigation water consumption estimation system and a machine learning-based agricultural irrigation water consumption estimation medium. The method comprises the steps of collecting relevant data of agricultural irrigation water consumption and field scale data of a research area; constructing a preliminary agricultural irrigation water consumption prediction model, performing sensitivity analysis, adjusting variables participating in training, and determining contribution of the variables to the agricultural irrigation water consumption estimation precision; constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning; constructing a regression prediction model of agricultural irrigation water consumption based on machine learning; and comparing the regression prediction models of the agricultural irrigation water consumption based on different machine learning algorithms, and determining the most robust regression prediction model with highest efficiency. The method aims to estimate the agricultural irrigation water consumption data with high spatial resolution (field scale) and high time resolution (daily scale), and improves the space-time resolution of the prior art.
Description
Technical Field
The invention belongs to the technical field of computer vision, and particularly relates to a machine learning-based agricultural irrigation water consumption estimation method, a machine learning-based agricultural irrigation water consumption estimation system and a machine learning-based agricultural irrigation water consumption estimation medium.
Background
Early agricultural irrigation water consumption estimation research is mainly based on typical investigation and quota calculation, has high experience dependence on statistics staff, is greatly interfered by artificial factors, and has low accuracy. The irrigation water calculating method based on water balance has the problem that the accuracy is still to be improved because the accurate regional water storage variable, water consumption and other key elements are difficult to obtain on the ground. The water consumption of the regional farm irrigation is directly calculated according to the single crop irrigation quota of the sample point irrigation area from the regional crop planting structure analysis, but the method is greatly limited by regions, subjective willingness of farmers to irrigate is not fully considered, and the uncertainty is large.
At present, more than 95% of surface water irrigated areas are not provided with field irrigation water monitoring equipment, most motor-pumped wells are not provided with metering devices, effective monitoring of regional agricultural water is not provided, an agricultural water monitoring network is not provided, and particularly monitoring of underground water serious super-mining areas is not provided, so that effective control of the total amount of regional agricultural water and the water consumption intensity of unit area is affected. The irrigation quantity of agricultural water is highly related to subjective intention of farmers, the water quantity of farmland irrigation is accurately monitored, the establishment of a water-saving excitation and compensation mechanism is facilitated, and the enthusiasm of farmers for water saving and the initiative of using a water-saving technology are improved.
At present, three main problems exist in the estimation research based on the water consumption of agricultural irrigation: the irrigation water consumption estimation time resolution is coarse, and estimation research can not be carried out by taking day or week as a unit; the estimated spatial resolution of irrigation water consumption is low, the accuracy in certain specific areas is low, the applicability of the summarized rules to other areas is low, and the model prediction is not fine; the uncertainty of irrigation water consumption estimation is large, and the error is large.
Disclosure of Invention
The invention aims to overcome the defects and shortcomings of the prior art and provides a machine learning-based agricultural irrigation water consumption estimation method, a machine learning-based agricultural irrigation water consumption estimation system and a machine learning-based agricultural irrigation water consumption estimation medium.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
in one aspect of the invention, there is provided a machine learning-based agricultural irrigation water consumption estimation method, comprising the steps of:
collecting agricultural irrigation water consumption data of the field scale of a research area;
acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area;
constructing a preliminary agricultural irrigation water consumption prediction model based on an artificial intelligence method;
performing sensitivity analysis by using the constructed agricultural irrigation water consumption prediction model, adjusting the variables participating in training, and determining the contribution of the variables to the estimation accuracy of the agricultural irrigation water consumption;
constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning;
constructing a regression prediction model of agricultural irrigation water consumption based on machine learning;
and comparing the regression prediction models of the agricultural irrigation water consumption based on different machine learning algorithms, and determining the most robust regression prediction model with highest efficiency.
As an optimal technical scheme, the agricultural irrigation water consumption data of the field scale of the research area is collected, and specifically comprises the following steps:
climate conditions including precipitation, temperature, relative humidity, wind speed, radiation, and saturated water vapor pressure difference of the area under study, crop planting type, and daily scale agricultural irrigation water consumption data are obtained.
As a preferable technical scheme, the method for acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of the field scale of the research area comprises the following steps:
obtaining leaf area index, total primary productivity, leaf temperature and soil water content data of the field scale of a research area through actual monitoring or simulation by adopting an ecosystem model;
the ecological system model simulation is specifically as follows:
and inputting the collected agricultural irrigation water consumption data of the field scale of the research area into an ecological system model, calibrating relevant model parameters through the model, and obtaining leaf area index, total primary productivity, leaf temperature and soil water content data through simulation.
As an optimal technical scheme, the construction of a preliminary agricultural irrigation water consumption prediction model based on machine learning is specifically as follows:
based on an artificial intelligence method, a preliminary agricultural irrigation water consumption prediction model is built, and a data set preprocessed by leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area is input to train the model, specifically:
randomly dividing the data set into a training data set and a test data set;
training the model by using a training data set, continuously adjusting model parameters, and finding out a proper parameter combination to reduce the loss function to the greatest extent; the training purpose is to find a set of parameters so that the model can fit the training data well and exhibit good generalization performance on the test data set; the confusion matrix is then used, and the receiver operates the characteristic curve to initially evaluate the model performance.
As an optimal technical scheme, the sensitivity analysis is carried out by using the constructed machine learning model, the variables participating in training are adjusted, and the contribution of the variables to the estimation precision of the agricultural irrigation water consumption is determined, specifically:
performing sensitivity analysis on the obtained agricultural irrigation water consumption related variables, wherein the agricultural irrigation water consumption related variables comprise climate conditions, crop planting types, agricultural irrigation water consumption, leaf area indexes of field scale, total primary productivity, leaf temperature and soil water content; the climatic conditions include precipitation, temperature, relative humidity, wind speed, radiation, and saturated water vapor pressure difference;
the sensitive variable corresponding to the predicted value of the agricultural irrigation water consumption is screened out, reserved and the insensitive variable is removed, specifically:
and (3) removing and/or adding a certain variable, verifying whether the variable can influence the predicted value of the agricultural irrigation water consumption, comparing the new predicted value with the original predicted value, observing the change of the predicted value, and if the ratio of the predicted value change to the predicted value exceeds a set threshold value, considering the variable as a sensitive variable, and finally obtaining a sensitive variable group corresponding to the predicted value of the agricultural irrigation water consumption.
As a preferable technical scheme, the construction of the 0-1 classification prediction model for judging whether irrigation occurs based on machine learning is specifically as follows:
adopting an artificial neural network ANN, a random forest RF, a support vector machine SVM and a long-term memory LSTM method;
firstly, classifying zero-variable based on daily-scale agricultural irrigation water consumption data, wherein 'zero' and 'one' respectively represent 'none' and 'have', namely, no irrigation event exists in the same day and irrigation events exist in the same day, the characteristics of binary classification problems are represented, and the values of the zero-variable are learned and predicted to realize classification tasks;
evaluating the performance of each machine learning algorithm classification model by applying the confusion matrix and the receiver operating characteristic curve;
the confusion matrix specifically comprises:
wherein, accuracy represents an Accuracy index, precision represents an Accuracy index, recall represents a Recall index, and TP represents a positive sample predicted to be positive by the model; TN represents a negative sample predicted negative by the model; FP represents the negative samples predicted to be positive by the model; FN represents positive samples that are model predicted negative.
As an optimal technical scheme, the construction of a regression prediction model of agricultural irrigation water consumption based on machine learning is specifically as follows:
screening a data set with the agricultural irrigation water consumption not being zero based on the result of the 0-1 classification prediction model for judging whether irrigation occurs or not, and training and constructing an agricultural irrigation water consumption regression prediction model based on ANN, RF, SVM, LSTM four machine learning algorithms; determining optimal parameters of the four models by a grid search method, selecting a classifier by an estimator, and transmitting other parameters except the optimal parameters to be determined; determining the parameter value needing to be optimized by a grid search method; and determining an evaluation standard of the model by setting scoring parameters, and finally screening to obtain optimal parameters of four models respectively, thereby completing construction of the agricultural irrigation water consumption regression prediction model based on ANN, RF, SVM, LSTM four machine learning algorithms.
As an optimal technical scheme, the comparison is based on agricultural irrigation water consumption regression prediction models of different machine learning algorithms, and the most robust regression prediction model with highest efficiency is determined specifically as follows:
the method comprises the steps of respectively predicting and obtaining estimated values of agricultural irrigation water consumption based on regression prediction models of agricultural irrigation water consumption of different machine learning algorithms; by determining the coefficient R 2 Comparing the performances of regression prediction models of the agricultural irrigation water consumption constructed by different machine learning algorithms with mean square error MSE and average absolute error MAE indexes:
determining coefficient R 2 The mean square error MSE, mean absolute error MAE are as follows:
wherein SSR is regression square sum, SST is total square sum; y is Y i Representing the true value of the irrigation water record,a predicted value representing the amount of agricultural irrigation water;
determining the coefficient R 2 The closer to 1, the closer to 0 is the mean square error MSE, the smaller is the mean absolute error MAE, and the better is the fitting effect of the model on the observed data, and the higher is the prediction precision; after normalizing the three indexes, the three indexes use the same weight to determine the most robust and most efficient regression prediction model of the irrigation water consumption in different machine learning algorithms.
The invention also provides an agricultural irrigation water consumption estimation system based on machine learning, which is applied to the agricultural irrigation water consumption estimation method based on machine learning, and comprises a data acquisition module, a model construction module, a sensitivity analysis module and a model evaluation module;
the data acquisition module is used for collecting agricultural irrigation water consumption data of the field block scale of the research area; acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area;
the model construction module is used for constructing a preliminary agricultural irrigation water consumption prediction model based on machine learning; constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning; constructing a regression prediction model of agricultural irrigation water consumption based on machine learning;
the sensitivity analysis module is used for carrying out sensitivity analysis on the constructed agricultural irrigation water consumption prediction model, adjusting the variables participating in training and determining the contribution of the variables to the agricultural irrigation water consumption estimation precision;
the model evaluation module is used for comparing agricultural irrigation water consumption regression prediction models based on different machine learning algorithms to determine the regression prediction model with the highest robustness and efficiency.
In another aspect of the present invention, there is also provided a storage medium storing a program which, when executed by a processor, implements the above-described machine learning-based agricultural irrigation water consumption estimation method.
Compared with the prior art, the invention has the following advantages and beneficial effects:
(1) According to the invention, the related data of the agricultural irrigation water consumption is collected, the ecological system model is adopted to obtain simulation data, a preliminary machine model is constructed, the sensitivity analysis is carried out on the variables, the prediction of zero-variable is carried out, a machine learning classification model is constructed, the classified data is analyzed, a machine learning regression model is constructed, the optimal parameters of each model are determined, and finally the estimation of the agricultural irrigation water consumption is realized.
(2) The invention aims to estimate agricultural irrigation water consumption data with high spatial resolution (field scale) and high time resolution (daily scale), and improves the space-time resolution of the prior art.
Drawings
FIG. 1 is a flow chart of a machine learning based agricultural irrigation water usage estimation method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the structure of an agricultural irrigation water consumption estimation system based on machine learning according to an embodiment of the present invention;
fig. 3 is a schematic structural view of a storage medium according to an embodiment of the present invention.
Detailed Description
In order to enable those skilled in the art to better understand the present application, the following description will make clear and complete descriptions of the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
Examples
As shown in fig. 1, the present embodiment provides a machine learning-based agricultural irrigation water consumption estimation method, which includes the following steps:
and 1, collecting data related to agricultural irrigation water consumption of the field scale of a research area.
Climate conditions such as precipitation (P), temperature (T), relative humidity (Relative Humidity, RH), wind speed (w), radiation (Rn), saturated water vapor pressure difference (Vapor Pressure Deficit, VPD), and agricultural irrigation water usage data for the type and daily scale of crop planting are obtained.
And 2, acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of the field scale of the research area.
Leaf area index, total primary productivity, leaf temperature and soil moisture data of the field scale of the research area are obtained through actual monitoring or simulation by adopting an ecosystem model (ecosys), and the leaf area index, total primary productivity, leaf temperature and soil moisture data are specifically as follows:
ecosys is an advanced agricultural ecosystem model based on biophysical and biochemical mechanisms that tracks water, energy, carbon and nutrient circulation on an hour scale. Inputting the data collected in the step 1 into an ecological system model, and calibrating relevant model parameters through the model to enable the model output result to be close to a true value. Leaf Area Index (LAI), total primary productivity (Gross Primary Productivity, GPP), leaf temperature (Canopy Temperature, T) C ) And soil moisture content (Soild Water Content, SWC) data.
And 3, constructing a preliminary agricultural irrigation water consumption prediction model based on an artificial intelligence method.
Preprocessing the data obtained in the step 2, and dividing the data into a training data set and a testing data set;
in the embodiment, a random forest algorithm is selected to build a preliminary agricultural irrigation water consumption prediction model based on machine learning;
training the model by using a training data set, continuously adjusting model parameters, and finding out a proper parameter combination to reduce the loss function to the greatest extent; the training purpose is to find a set of parameters so that the model can fit the training data well and exhibit good generalization performance on the test data set; then, using a confusion matrix, and primarily evaluating the model performance by using a Receiver Operating Characteristic (ROC) curve to prepare for the sensitivity analysis of the variables;
and 4, performing sensitivity analysis by using the machine learning model constructed in the step 3, adjusting the variables participating in training, and determining the contribution of the variables to the estimation accuracy of the agricultural irrigation water consumption.
Carrying out sensitivity analysis on the related variable of the agricultural irrigation water consumption obtained in the step 1 and the step 2, screening out the sensitive variable corresponding to the predicted value of the agricultural irrigation water consumption, reserving the sensitive variable, and removing the insensitive variable, wherein the method specifically comprises the following steps:
and removing and/or adding a certain variable, verifying whether the variable can cause larger influence on the predicted value of the agricultural irrigation water consumption, comparing the new predicted value with the original predicted value, and observing the change of the predicted value, if the ratio of the predicted value change to the predicted value exceeds a set threshold (20% in the embodiment), considering the variable as a sensitive variable, and finally obtaining a sensitive variable group corresponding to the predicted value.
And 5, constructing a 0-1 classification prediction model for judging whether irrigation occurs based on machine learning.
In this embodiment, artificial neural network (Artificial Neural Network, ANN), random Forest (RF), support vector machine (Support Vector Machine, SVM), and Long Short-Term Memory (LSTM) methods are used.
The method comprises the steps of firstly classifying zero-variable based on daily-scale agricultural irrigation water consumption data, wherein 'zero' and 'one' respectively represent 'none' and 'none', namely no irrigation event on the same day and no irrigation event on the same day, and can represent the characteristics of binary classification problems, and learning and predicting the value of the zero-variable to realize classification tasks. By applying the confusion matrix, ROC curves evaluate the performance of each machine learning algorithm classification model.
Through the confusion matrix, the indexes of Accuracy (Accuracy), precision (Precision) and Recall (Recall) can be obtained, and the formulas are as follows:
where TP represents the positive sample predicted to be positive by the model; TN represents a negative sample predicted negative by the model; FP represents the negative samples predicted to be positive by the model; FN represents positive samples that are model predicted negative.
And 6, constructing a regression prediction model of the agricultural irrigation water consumption based on machine learning.
And 5, based on the result of the 0-1 classification prediction model for judging whether irrigation occurs or not in the step 5, screening out a data set with the agricultural irrigation water consumption not being zero, and training and constructing the agricultural irrigation water consumption regression prediction model based on ANN, RF, SVM, LSTM four machine learning algorithms. Determining optimal parameters of the four models by a grid search method, selecting a classifier by an estimator, and transmitting other parameters except the optimal parameters to be determined; determining the parameter value needing to be optimized through a grid search method param_grid; and determining an evaluation standard of the model by setting scoring parameters, and finally screening to obtain optimal parameters of four models respectively, thereby completing construction of the agricultural irrigation water consumption regression prediction model based on ANN, RF, SVM, LSTM four machine learning algorithms.
And 7, comparing the agricultural irrigation water consumption regression prediction models based on different machine learning algorithms, and determining the most robust regression prediction model with the highest efficiency.
Agricultural irrigation water consumption regression prediction based on different machine learning algorithmsAnd respectively predicting and obtaining estimated values of the agricultural irrigation water consumption by the model. By determining coefficients (Coefficient of Determination, R 2 ) The mean square error (Mean Squared Error, MSE) and average absolute error (Mean Absolute Error, MAE) indexes are used for comparing the performances of the agricultural irrigation water consumption regression prediction models constructed based on different machine learning algorithms, and the most robust and most efficient agricultural irrigation water consumption regression prediction model in the different machine learning algorithms is determined so as to obtain the most accurate agricultural irrigation water consumption prediction value.
Determining coefficient R 2 The mean square error MSE, mean absolute error MAE are as follows:
wherein SSR is regression square sum, SST is total square sum; y is Y i Representing the true value of the irrigation water record,a predicted value representing the amount of agricultural irrigation water;
the judgment standard is as follows: determining the coefficient R 2 The closer to 1, the closer to 0 is the mean square error MSE, the smaller is the mean absolute error MAE, which means that the better is the fitting effect of the model to the observed data, and the higher is the prediction accuracy.
After normalizing the three indexes, the three indexes use the same weight to determine the most robust and most efficient regression prediction model of the agricultural irrigation water consumption in different machine learning algorithms.
In another embodiment of the present application, as shown in fig. 2, there is provided a machine learning-based agricultural irrigation water consumption estimation system, which includes a data acquisition module, a model construction module, a sensitivity analysis module, and a model evaluation module;
the data acquisition module is used for collecting agricultural irrigation water consumption data of the field block scale of the research area; acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area;
the model construction module is used for constructing a preliminary agricultural irrigation water consumption prediction model based on machine learning; constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning; constructing a regression prediction model of agricultural irrigation water consumption based on machine learning;
the sensitivity analysis module is used for carrying out sensitivity analysis on the constructed agricultural irrigation water consumption prediction model, adjusting the variables participating in training and determining the contribution of the variables to the agricultural irrigation water consumption estimation precision;
the model evaluation module is used for comparing agricultural irrigation water consumption regression prediction models based on different machine learning algorithms to determine the regression prediction model with the highest robustness and efficiency.
It should be noted that, the system provided in the above embodiment is only exemplified by the division of the above functional modules, and in practical application, the above functional allocation may be performed by different functional modules according to needs, that is, the internal structure is divided into different functional modules to perform all or part of the functions described above, and the system is the agricultural irrigation water consumption estimation method based on machine learning applied to the above embodiment.
As shown in fig. 3, in another embodiment of the present application, there is further provided a storage medium storing a program, which when executed by a processor, implements a machine learning-based agricultural irrigation water consumption estimation method, specifically:
collecting agricultural irrigation water consumption data of the field scale of a research area;
acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area;
constructing a preliminary agricultural irrigation water consumption prediction model based on machine learning;
performing sensitivity analysis by using the constructed agricultural irrigation water consumption prediction model, adjusting the variables participating in training, and determining the contribution of the variables to the estimation accuracy of the agricultural irrigation water consumption;
constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning;
constructing a regression prediction model of agricultural irrigation water consumption based on machine learning;
and comparing the regression prediction models of the agricultural irrigation water consumption based on different machine learning algorithms, and determining the most robust regression prediction model with highest efficiency.
It is to be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of the following techniques, as is well known in the art: discrete logic circuits having logic gates for implementing logic functions on data signals, application specific integrated circuits having suitable combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
The above examples are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above examples, and any other changes, modifications, substitutions, combinations, and simplifications that do not depart from the spirit and principle of the present invention should be made in the equivalent manner, and the embodiments are included in the protection scope of the present invention.
Claims (10)
1. The agricultural irrigation water consumption estimation method based on machine learning is characterized by comprising the following steps of:
collecting agricultural irrigation water consumption data of the field scale of a research area;
acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area;
constructing a preliminary agricultural irrigation water consumption prediction model based on an artificial intelligence method;
performing sensitivity analysis by using the constructed agricultural irrigation water consumption prediction model, adjusting the variables participating in training, and determining the contribution of the variables to the estimation accuracy of the agricultural irrigation water consumption;
constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning;
constructing a regression prediction model of agricultural irrigation water consumption based on machine learning;
and comparing the regression prediction models of the agricultural irrigation water consumption based on different machine learning algorithms, and determining the most robust regression prediction model with highest efficiency.
2. The machine learning based agricultural irrigation water yield estimation method according to claim 1, wherein the collection of agricultural irrigation water yield data of the field scale of the investigation region is specifically:
climate conditions including precipitation, temperature, relative humidity, wind speed, radiation, and saturated water vapor pressure difference of the area under study, crop planting type, and daily scale agricultural irrigation water consumption data are obtained.
3. The machine learning based agricultural irrigation water consumption estimation method according to claim 1, wherein the obtaining of leaf area index, total primary productivity, leaf temperature and soil moisture content data of the field scale of the investigation region is specifically:
obtaining leaf area index, total primary productivity, leaf temperature and soil water content data of the field scale of a research area through actual monitoring or simulation by adopting an ecosystem model;
the ecological system model simulation is specifically as follows:
and inputting the collected agricultural irrigation water consumption data of the field scale of the research area into an ecological system model, calibrating relevant model parameters through the model, and obtaining leaf area index, total primary productivity, leaf temperature and soil water content data through simulation.
4. The machine learning-based agricultural irrigation water consumption estimation method according to claim 1, wherein the construction of the preliminary machine learning-based agricultural irrigation water consumption prediction model is specifically as follows:
based on an artificial intelligence method, a preliminary agricultural irrigation water consumption prediction model is built, and a data set preprocessed by leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area is input to train the model, specifically:
randomly dividing the data set into a training data set and a test data set;
training the model by using a training data set, continuously adjusting model parameters, and finding out a proper parameter combination to reduce the loss function to the greatest extent; the training purpose is to find a set of parameters so that the model can fit the training data well and exhibit good generalization performance on the test data set; the confusion matrix is then used, and the receiver operates the characteristic curve to initially evaluate the model performance.
5. The machine learning-based high spatial-temporal resolution agricultural irrigation water consumption estimation method according to claim 1, wherein the sensitivity analysis is performed by using the constructed machine learning model, the variables participating in training are adjusted, and the contribution of the variables to the agricultural irrigation water consumption estimation precision is determined, specifically:
performing sensitivity analysis on the obtained agricultural irrigation water consumption related variables, wherein the agricultural irrigation water consumption related variables comprise climate conditions, crop planting types, agricultural irrigation water consumption, leaf area indexes of field scale, total primary productivity, leaf temperature and soil water content; the climatic conditions include precipitation, temperature, relative humidity, wind speed, radiation, and saturated water vapor pressure difference;
the sensitive variable corresponding to the predicted value of the agricultural irrigation water consumption is screened out, reserved and the insensitive variable is removed, specifically:
and (3) removing and/or adding a certain variable, verifying whether the variable can influence the predicted value of the agricultural irrigation water consumption, comparing the new predicted value with the original predicted value, observing the change of the predicted value, and if the ratio of the predicted value change to the predicted value exceeds a set threshold value, considering the variable as a sensitive variable, and finally obtaining a sensitive variable group corresponding to the predicted value of the agricultural irrigation water consumption.
6. The machine learning-based agricultural irrigation water consumption estimation method according to claim 1, wherein the construction of the machine learning-based 0-1 classification prediction model for judging whether irrigation occurs is specifically as follows:
adopting an artificial neural network ANN, a random forest RF, a support vector machine SVM and a long-term memory LSTM method;
firstly, classifying zero-variable based on daily-scale agricultural irrigation water consumption data, wherein 'zero' and 'one' respectively represent 'none' and 'have', namely, no irrigation event exists in the same day and irrigation events exist in the same day, the characteristics of binary classification problems are represented, and the values of the zero-variable are learned and predicted to realize classification tasks;
evaluating the performance of each machine learning algorithm classification model by applying the confusion matrix and the receiver operating characteristic curve;
the confusion matrix specifically comprises:
wherein, accuracy represents an Accuracy index, precision represents an Accuracy index, recall represents a Recall index, and TP represents a positive sample predicted to be positive by the model; TN represents a negative sample predicted negative by the model; FP represents the negative samples predicted to be positive by the model; FN represents positive samples that are model predicted negative.
7. The machine learning-based agricultural irrigation water consumption estimation method according to claim 1, wherein the construction of the machine learning-based regression prediction model of agricultural irrigation water consumption is specifically as follows:
screening a data set with the agricultural irrigation water consumption not being zero based on the result of the 0-1 classification prediction model for judging whether irrigation occurs or not, and training and constructing an agricultural irrigation water consumption regression prediction model based on ANN, RF, SVM, LSTM four machine learning algorithms; determining optimal parameters of the four models by a grid search method, selecting a classifier by an estimator, and transmitting other parameters except the optimal parameters to be determined; determining the parameter value needing to be optimized by a grid search method; and determining an evaluation standard of the model by setting scoring parameters, and finally screening to obtain optimal parameters of four models respectively, thereby completing construction of the agricultural irrigation water consumption regression prediction model based on ANN, RF, SVM, LSTM four machine learning algorithms.
8. The machine learning-based agricultural irrigation water consumption estimation method according to claim 1, wherein the comparing the agricultural irrigation water consumption regression prediction models based on different machine learning algorithms determines a most robust regression prediction model with highest efficiency, specifically:
the method comprises the steps of respectively predicting and obtaining estimated values of agricultural irrigation water consumption based on regression prediction models of agricultural irrigation water consumption of different machine learning algorithms; by determining the coefficient R 2 Comparing the performances of regression prediction models of the agricultural irrigation water consumption constructed by different machine learning algorithms with mean square error MSE and average absolute error MAE indexes:
determining coefficient R 2 The mean square error MSE, mean absolute error MAE are as follows:
wherein SSR is regression square sum, SST is total square sum; y is Y i Representing the true value of the irrigation water record,a predicted value representing the amount of agricultural irrigation water;
determining the coefficient R 2 The closer to 1, the closer to 0 is the mean square error MSE, the smaller is the mean absolute error MAE, and the better is the fitting effect of the model on the observed data, and the higher is the prediction precision; after normalizing the three indexes, the three indexes use the same weight to determine the most robust and most efficient regression prediction model of the irrigation water consumption in different machine learning algorithms.
9. The agricultural irrigation water consumption estimation system based on machine learning is characterized by being applied to the agricultural irrigation water consumption estimation method based on machine learning as claimed in any one of claims 1-8, and comprising a data acquisition module, a model construction module, a sensitivity analysis module and a model evaluation module;
the data acquisition module is used for collecting agricultural irrigation water consumption data of the field block scale of the research area; acquiring leaf area index, total primary productivity, leaf temperature and soil water content data of a field scale of a research area;
the model construction module is used for constructing a preliminary agricultural irrigation water consumption prediction model based on machine learning; constructing a 0-1 classification prediction model for judging whether irrigation occurs or not based on machine learning; constructing a regression prediction model of agricultural irrigation water consumption based on machine learning;
the sensitivity analysis module is used for carrying out sensitivity analysis on the constructed agricultural irrigation water consumption prediction model, adjusting the variables participating in training and determining the contribution of the variables to the agricultural irrigation water consumption estimation precision;
the model evaluation module is used for comparing agricultural irrigation water consumption regression prediction models based on different machine learning algorithms to determine the regression prediction model with the highest robustness and efficiency.
10. A storage medium storing a program, characterized in that: the program, when executed by a processor, implements the machine learning-based agricultural irrigation water consumption estimation method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311260303.2A CN117333321B (en) | 2023-09-27 | 2023-09-27 | Agricultural irrigation water consumption estimation method, system and medium based on machine learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311260303.2A CN117333321B (en) | 2023-09-27 | 2023-09-27 | Agricultural irrigation water consumption estimation method, system and medium based on machine learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117333321A true CN117333321A (en) | 2024-01-02 |
| CN117333321B CN117333321B (en) | 2024-07-09 |
Family
ID=89274808
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311260303.2A Active CN117333321B (en) | 2023-09-27 | 2023-09-27 | Agricultural irrigation water consumption estimation method, system and medium based on machine learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117333321B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117814097A (en) * | 2024-03-06 | 2024-04-05 | 北京佳格天地科技有限公司 | Machine learning-based efficient farmland irrigation method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180348714A1 (en) * | 2017-06-01 | 2018-12-06 | Valmont Industries, Inc. | System and method for irrigation management using machine learning workflows |
| CN111369093A (en) * | 2018-12-26 | 2020-07-03 | 天云融创数据科技（北京）有限公司 | Irrigation method and device based on machine learning |
| CN113378473A (en) * | 2021-06-23 | 2021-09-10 | 中国地质科学院水文地质环境地质研究所 | Underground water arsenic risk prediction method based on machine learning model |
| CN114638146A (en) * | 2022-03-25 | 2022-06-17 | 西北农林科技大学 | Crop irrigation water demand prediction method based on AquaCrop model and SVR |
| CN115952421A (en) * | 2023-02-01 | 2023-04-11 | 浙江大学 | High-precision time-space simulation method for coupling ecological process model and machine learning algorithm |
| WO2023108213A1 (en) * | 2021-12-14 | 2023-06-22 | Rubicon Research Pty Ltd | Methods and systems for classifying and benchmarking irrigation performance |
-
2023
- 2023-09-27 CN CN202311260303.2A patent/CN117333321B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180348714A1 (en) * | 2017-06-01 | 2018-12-06 | Valmont Industries, Inc. | System and method for irrigation management using machine learning workflows |
| CN111369093A (en) * | 2018-12-26 | 2020-07-03 | 天云融创数据科技（北京）有限公司 | Irrigation method and device based on machine learning |
| CN113378473A (en) * | 2021-06-23 | 2021-09-10 | 中国地质科学院水文地质环境地质研究所 | Underground water arsenic risk prediction method based on machine learning model |
| WO2023108213A1 (en) * | 2021-12-14 | 2023-06-22 | Rubicon Research Pty Ltd | Methods and systems for classifying and benchmarking irrigation performance |
| CN114638146A (en) * | 2022-03-25 | 2022-06-17 | 西北农林科技大学 | Crop irrigation water demand prediction method based on AquaCrop model and SVR |
| CN115952421A (en) * | 2023-02-01 | 2023-04-11 | 浙江大学 | High-precision time-space simulation method for coupling ecological process model and machine learning algorithm |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117814097A (en) * | 2024-03-06 | 2024-04-05 | 北京佳格天地科技有限公司 | Machine learning-based efficient farmland irrigation method and system |
| CN117814097B (en) * | 2024-03-06 | 2024-05-03 | 北京佳格天地科技有限公司 | Machine learning-based efficient farmland irrigation method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117333321B (en) | 2024-07-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109272146B (en) | Flood prediction method based on deep learning model and BP neural network correction | |
| CN110084367B (en) | Soil moisture content prediction method based on LSTM deep learning model | |
| CN112602563B (en) | Water-saving irrigation system and accurate irrigation method | |
| Yang et al. | Classifying floods by quantifying driver contributions in the Eastern Monsoon Region of China | |
| Coopersmith et al. | Machine learning assessments of soil drying for agricultural planning | |
| CN109299812B (en) | Flood prediction method based on deep learning model and KNN real-time correction | |
| UA126650C2 (en) | Forecasting national crop yield during the growing season | |
| CN117333321B (en) | Agricultural irrigation water consumption estimation method, system and medium based on machine learning | |
| CN111665575B (en) | Medium-and-long-term rainfall grading coupling forecasting method and system based on statistical power | |
| CN111915058B (en) | Flood prediction method and device based on long-time memory network and transfer learning | |
| CN113159439B (en) | Crop yield prediction method, system, storage medium and electronic equipment | |
| CN109002604A (en) | A kind of soil moisture content prediction technique based on Bayes's maximum entropy | |
| Cappelaere et al. | Hydrologic process simulation of a semiarid, endoreic catchment in Sahelian West Niger. 2. Model calibration and uncertainty characterization | |
| CN117114374B (en) | Intelligent agricultural irrigation management system based on weather prediction | |
| CN117935081B (en) | Cultivated land change monitoring method and system based on remote sensing satellite data | |
| CN114357737B (en) | Agent optimization calibration method for time-varying parameters of large-scale hydrologic model | |
| CN115511219A (en) | Intelligent prediction method and system for growth state of wheat crop | |
| Xiujia et al. | Application of WNN-PSO model in drought prediction at crop growth stages: A case study of spring maize in semi-arid regions of northern China | |
| CN114662911A (en) | Irrigation water management method and management system based on farmland moisture monitoring | |
| CN116341994B (en) | River water ecological treatment method and system based on passenger water diversion | |
| CN117296538A (en) | Green plant maintenance method, device and system based on vegetation soil component detection | |
| WO2023245399A1 (en) | Rice production potential simulation method based on land system and climate change coupling | |
| CN110009132A (en) | A kind of short-term electric load fining prediction technique based on LSTM deep neural network | |
| Zanella et al. | Internet of things for hydrology: Potential and challenges | |
| CN117933946B (en) | Rural business management method based on big data |
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 |