CN115952421A - High-precision time-space simulation method for coupling ecological process model and machine learning algorithm - Google Patents
High-precision time-space simulation method for coupling ecological process model and machine learning algorithm Download PDFInfo
- Publication number
- CN115952421A CN115952421A CN202310049924.XA CN202310049924A CN115952421A CN 115952421 A CN115952421 A CN 115952421A CN 202310049924 A CN202310049924 A CN 202310049924A CN 115952421 A CN115952421 A CN 115952421A
- Authority
- CN
- China
- Prior art keywords
- model
- machine learning
- process model
- data
- simulation
- 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.)
- Pending
Links
Images
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
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a high-precision space-time simulation method for coupling an ecological process model and a machine learning algorithm, and belongs to the field of researches related to ecology, agriculture and global change. The method comprises the following steps: calibrating relevant parameters of the ecosystem process model by using long-term positioning test data in the research area; randomly selecting a plurality of grid points, randomly generating a management measure and climate change scene combination for each grid point, and simulating the state variable change of the ecological system at each grid point and under the scene by using the calibrated process model; training and evaluating the performance of machine learning, and obtaining an optimal integrated model according to a weighting evaluation method of different model weights; and taking the optimal machine learning integrated model as a proxy model of the process model, and performing proxy simulation on changes of crop yield, soil organic carbon and the like under a high-resolution spatiotemporal scale. The invention can predict the space-time change rule of the state variable of a certain ecological system under high space-time resolution and the response of the space-time change rule to different management measures and climate change more quickly and efficiently.
Description
Technical Field
The invention belongs to the field of ecological, agricultural and global change related researches, and particularly relates to a high-precision space-time simulation method for coupling an ecological process model and a machine learning algorithm.
Background
The process-based ecosystem model is a major tool for addressing global sustainable development and ecological environmental challenges (e.g., climate change, food safety, biodiversity protection, and natural resource management) by comprehensively considering the interaction of ecosystem processes with the environment (e.g., climate, soil, etc.) and human activities (e.g., land utilization, agricultural management). However, the application of the process model on a large scale has the problems of high calculation cost, high model uncertainty and poor data availability, and the rapid simulation under high-precision space-time resolution is difficult to complete within an acceptable time, so that the universality and operability of the model on the large scale are limited.
The process model simplifies the ecosystem process by using a mathematical equation, closed calculation can be carried out only by inputting a large number of parameters, the parameters are estimated based on test data of a field observation station or a field test station, but on the large scale that the environmental conditions are more complex and diversified, some parameters of the ecological process cannot be well estimated, and the model uncertainty is large. Machine learning can achieve greater accuracy and simulation efficiency on a large scale than process models, but it can produce results that are contrary to the ecosystem process, requiring the provision of a sufficient amount of high quality training data. In this context, the "coupled process model and machine learning integrated simulation" method arose. The process model can be calibrated at site scale, and the calibrated model can produce a large amount of simulation data under enough environment-management situations, so that a large amount of high-quality data is improved for the machine learning model. Based on the simulation data of the process models, the machine learning model can automatically learn the intrinsic relationship between the process and the environment-management of the complex ecosystem, and can realize rapid generalized simulation. The priori knowledge implied by the process model and the learning capability of machine learning are combined, and the rapid simulation of the high-resolution spatiotemporal scale can be flexibly realized by using a small amount of station observation data. However, due to the difference of algorithms, different machine learning methods have different abilities to understand ecological processes and model generalization. In order to reduce the uncertainty of machine learning integration as much as possible, the invention uses adaptive screening to eliminate models with poor learning capability, and constructs a machine learning model with weight and strong weighted average learning capability by calculating the root mean square error of different models, thereby improving the simulation precision and reducing the uncertainty of an integrated agent model.
The invention combines an ecosystem model based on a process with various machines/deep learning, establishes an optimal integrated agent model through a self-adaptive screening and weighted average mode, realizes the rapid and accurate simulation of high-precision space-time resolution, and provides a technical method for the simulation and management decision of the ecosystem on a fine scale.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a high-precision space-time simulation method for coupling an ecological process model and a machine learning algorithm. The method couples the ecological system model based on the process with various machine/deep learning models, establishes an optimal integrated agent model through a self-adaptive screening and weighted average mode, and accurately and quickly realizes the simulation of the ecological system state variable with high-precision space-time resolution.
The invention adopts the following specific technical scheme:
the invention provides a high-precision space-time simulation method for coupling an ecological process model and a machine learning algorithm, which comprises the following specific steps:
s1: calibrating and verifying relevant parameters of the ecological system process model by using long-term positioning test data in a target research area;
s2: randomly selecting a plurality of grid points capable of covering the climate type and soil characteristic spatial variation of the research area in the research area, and randomly generating a management measure and climate variation scene combination for each grid point; simulating the state variable change of the ecological system at each grid point and under the scene by using the ecological system process model verified in the step S1 based on the soil data and the historical climate data at the scene corresponding to each grid point;
s3: integrating input data and output data of the ecosystem process model scene simulation in the step S2, and training different types of machine learning models to enable the machine learning models to fully learn and simulate state variable changes of the ecosystem process model simulation; evaluating the performance of machine learning, adaptively eliminating a machine learning model with poor performance, and obtaining an optimal integration model according to a weighting evaluation method of different model weights;
s4: and taking the optimal integrated model as a proxy model of the ecosystem process model, and performing proxy simulation on the basis of spatial data to generate a high-precision digital mapping product required by a research purpose under a high-resolution spatiotemporal scale.
Preferably, the step S1 is as follows:
aiming at a specific research purpose, determining a research area to be simulated and a suitable ecosystem process model; acquiring observation data of the ecological system in-situ site in the research area according to the requirements of the ecological system process model, and correcting and verifying key parameters of the ecological system process model by using a differential evolution algorithm.
Furthermore, the observation data of the ecosystem in-situ site comprise one or more of meteorological data, soil attribute data, farmland ecosystem data and ecosystem state variable data.
Further, the meteorological data comprises daily temperature, precipitation and solar radiation, the soil attribute data comprises soil organic carbon, pH and soil volume weight, the farmland ecosystem data comprises management data and yield data of seeding period, fertilization and irrigation stages, and the ecosystem state variable data comprises soil organic carbon and greenhouse gas emission observation data.
Further, the key parameters include crop variety parameters, soil organic carbon and N in the farmland ecosystem 2 O an emission process parameter.
Preferably, the step S2 is specifically as follows:
in the research area, generating spatial grids according to a set spatial resolution, randomly selecting a plurality of grid points capable of covering the climate type and soil characteristic spatial variation of the research area from all the grids, and randomly generating a management measure and climate change scene combination at each grid point; and (3) collecting and preparing soil data and historical climate data corresponding to the selected grid points, and driving the ecosystem process model verified in the step 1 to simulate interested state variables under historical and future situations.
Furthermore, the management measure scenes refer to various possible human activity management, and in a farmland ecosystem, the management measure scenes comprise a crop system scene, a crop sowing time scene, a fertilization scene, an irrigation scene and a straw returning scene; the climate change situation refers to a future climate change situation and is selected and determined from a global climate model and a shared economic path emission situation.
Preferably, in step S3, the machine learning model includes a LASSO model, a support vector machine, a random forest, an XGBoost, a convolutional neural network, and a long-short term memory network.
Preferably, in step S3, the root mean square error RMSE and the determination coefficient R are calculated separately 2 To evaluate the performance of each machine learning model to predict the process model output by setting a screening criterion, R 2 Discarding the models with the ranking lower than the average level of each model, and screening out a plurality of models of which the models are represented on the average level; and calculating the weight of each screened model according to the model expression, wherein the weight is 1/RMSE, and constructing a weighted average integration model to obtain an optimal integration model.
Preferably, the step S4 is as follows:
and collecting and preparing high-precision spatial data, driving the constructed agent model, and realizing rapid simulation through parallel operation to generate a high-precision digital mapping product required by the research purpose.
Compared with the prior art, the invention has the following beneficial effects:
1) The ecological system process model is calibrated by acquiring ecological system observation data of site scale, and then a large amount of management-environment scene simulation is carried out through the process model. And (3) simulating the scene simulation result of the process model by using various machine learning models, fully learning the internal relation in the process model, obtaining an optimal integrated model by self-adaptive screening and weighted average, and finally quickly simulating the state variable of the ecological system on a higher-resolution space-time scale.
2) The process model and the machine learning are innovatively adopted for mixed modeling, the mechanicalness of the model is guaranteed, the model precision and the simulation efficiency of the model under the fine scale are improved, the weight method is adopted for integrating the model, the system error is further reduced, the data accuracy and the reliability are improved, and the influence of future climate change simulation and different management measures on an ecological system is favorably understood and predicted.
3) The invention is suitable for simulating the state attribute of any ecosystem in any region, at any time, such as ecosystem productivity, crop yield, leaf area index, soil organic carbon, N on the global scale, national scale or regional scale, historical or future situation 2 And indexes such as O emission are simulated, so that the state attribute simulation of the ecological system under high-precision space-time resolution and the response of the ecological system to different human activity management and climate change can be predicted more efficiently and accurately. The wide application of the technology can provide scientific and technical support for the simulation of the ecosystem on a high-precision space-time scale.
Drawings
FIG. 1 is a schematic flow diagram of the process of the present invention.
Detailed Description
The invention is further illustrated and described below with reference to the drawings and the detailed description. The technical features of the embodiments of the present invention can be combined correspondingly without mutual conflict.
The invention provides a high-precision space-time simulation method for coupling an ecological process model and a machine learning algorithm, which comprises the following steps of: 1) Calibrating relevant parameters of the ecosystem process model by using long-term positioning test data in the research area; 2) Randomly selecting a plurality of grid points in a research area, covering the climate type and soil characteristic space variation of the research area, randomly generating a management measure and climate change scene combination for each grid point, and simulating the state variable change of the ecological system (such as crop yield, soil organic carbon and the like in a farmland ecological system) under each grid point and scene by using a calibrated process model; 3) Integrating input and output data of the process model scene simulation, and training different types of machine learning models to ensure that the machine learning models fully learn and simulate the state variable change of the process model simulation; evaluating the performance of machine learning, adaptively eliminating the poor-performance machine learning model, and obtaining an optimal integrated model according to a weighting evaluation method of different model weights; 4) And taking the optimal machine learning integrated model as a proxy model of the process model, and performing proxy simulation on changes of crop yield, soil organic carbon and the like under a high-resolution spatiotemporal scale.
The respective steps will be specifically described below.
The method comprises the following steps: correction and verification of ecosystem process model
Firstly, aiming at a specific research purpose, determining a research area to be simulated and a proper ecosystem process simulation, and acquiring ecosystem in-situ site observation data in the research area according to the needs of a model, wherein the data generally comprises meteorological data such as daily temperature, rainfall, solar radiation and the like, soil attribute data such as soil organic carbon, pH, soil volume weight and the like, management data and yield data for farmland ecosystems possibly also comprise sowing date, fertilization, irrigation and the like, and other interested ecosystem state variable data (such as soil organic carbon, greenhouse gas emission observation data and the like). Using differential evolution algorithm to model key parameters (such as crop variety parameters, soil organic carbon and N in farmland ecosystem) 2 O-emission process parameters, etc.) for calibration and verification.
Step two: scenario generation and ecosystem process model-based simulation
In order for a machine learning model to adequately learn and simulate the inherent relationships of a process model, it is necessary to simulate enough data with a site-level calibrated process model. The method comprises three substeps, namely scenario generation, input data preparation, process model simulation and output.
Firstly, scene generation is carried out, a spatial grid is generated according to a set spatial resolution, a plurality of grid points (covering all climate types and soil characteristic spatial variations of a research area as much as possible) are randomly selected from all the grids, and management-climate change scenes are randomly generated at the grid points. The management scenario refers to various possible human activity management, and in a farmland ecosystem, the management scenario comprises a crop system scenario, a crop seeding stage scenario, a fertilization scenario, an irrigation scenario (such as an irrigation method, an irrigation depth, an irrigation period and the like), a straw returning scenario and the like. The climate scenarios refer to future climate change scenarios and may be selected from Global Climate Models (GCMs) and shared economic path emissions scenarios (SSP). And secondly, collecting and preparing soil data and historical climate data corresponding to the selected grid points to drive the verified ecosystem process model in the step one and simulate the interested state variables such as crop yield, soil organic carbon content and the like in the farmland ecosystem under historical and future scenes.
Step three: machine/deep learning agent model
In the step, for screening and integrating the machine/deep learning model capable of acting on the process model, the input and output results of the grid-management-climate scene combination in the step two are integrated to train the machine learning model (such as machine learning of an LASSO model, a support vector machine, a random forest, XGboost and the like and deep learning models of a convolutional neural network or a long-short term memory network and the like), so that the machine learning model learns and simulates the internal relation of the process model. The data grouping is performed before model training, wherein 80% of data (namely output of the process model in the step two) train the model and 20% of data verify the model. Separately calculating Root Mean Square Error (RMSE) and coefficient of determination (R) 2 ) Evaluating the performance of the output of the prediction process model of each machine learning model by setting a screening criterion, R 2 Models ranked below the mean of the six models are discarded, and several models whose models are represented above the mean are screened. And according to the model expression, calculating the weight of each screened model, wherein the weight is 1/RMSE, and constructing a weighted average integrated model, namely the proxy model of the process model.
Step four: rapid high-precision space-time simulation based on proxy model
High-precision spatial data is collected and prepared, the constructed agent model is driven, rapid simulation is realized through parallel operation, and a high-precision digital drawing product required by the research purpose is generated.
Example 1
The research content of this embodiment is the simulation of crop yield and soil organic carbon change in a north China plain 1km spatial scale winter wheat/summer corn rotation system in different management modes, as shown in fig. 1, and specifically includes the following steps:
1) Taking winter wheat-summer corn rotation system in North China plain as research object to obtain soil, climate, management measures and yield, soil organic carbon and N of Xingji, goldenrain city, constant water, laiyang, ipomoea and Xuzhou and other station scales 2 And (4) calibrating parameters of an agricultural ecosystem model (taking an APSIM model as an example) through a differential evolution algorithm according to observation data such as O emission and the like.
2) Generating 45 ten thousand farmland cultivated land grids according to 1km spatial resolution of North China plain, randomly selecting 6000 grids from the grids, and randomly generating corresponding management scenes and climate scenes for each grid, wherein the management scenes comprise a sowing period scene (taking 1 day as an interval and from 30 days before to 30 days after a reference sowing period), and a nitrogen fertilizer scene (the nitrogen fertilizer application amount of nitrogen fertilizer is 1kg ha) -1 yr -1 In units of 0kg ha -1 yr -1 Increased to 400kg ha -1 yr -1 Since the regional fertilization method is not changed greatly, the nitrogen dressing period and the base dressing ratio are set to be in the traditional mode), and the organic fertilizer scene (the organic fertilizer is represented by farmyard manure, 10kg is taken as a unit, and 0kg ha is used as a unit) -1 Increased to 4000kg ha -1 ) A straw returning scene (straw returning amount is increased from 0% to 100% in a unit of 10%, and a total of 11 straw returning scenes), an irrigation scene [ irrigation depth 50cm, and wheat season irrigation settings are: 0 time, 1 time (joint removing), 2 times (joint removing and grouting), 3 times (green turning, joint removing and grouting), 4 times (green turning, joint removing, spike pregnancy and grouting), irrigation for corn season is set as 0 time, 1 time (jointing), 2 times (jointing and emasculation), 3 times (jointing, emasculation and grouting), 4 times (jointing, large horn mouth period,Emasculation and grouting) and automatic irrigation (start irrigation when field capacity is below 30%, 40%, 50%, 60%, 70% and 80%, respectively)]And the like. Climate scenarios were randomly selected from 30 Global Climate Models (GCMs) and shared economic path emissions scenarios (SSP) in the CMIP6, yielding a total of 6000 grid-management-climate scenario combinations.
3) And (3) simulating 6000 grid-management-climate scene combinations by using the calibrated agricultural ecosystem calibration model, and simulating the yield of wheat and corn and the soil organic carbon.
4) 6000 grid-management-climate scenario combination simulation results were integrated, 80% as training set and 20% as validation set. Training 6 different types of machine learning models in a training set, and obtaining a model with optimal training through parameter optimization and super-parameter adjustment. And simulating the data of the verification set by the trained model for evaluating the performance of the model.
5) Calculating a statistical indicator (R) for evaluating the performance of the machine learning simulation process model based on the process model on the verification set and the simulation value of the machine learning 2 And RMSE) according to R 2 The machine learning model below the average level is eliminated, the weight is calculated according to the RMSE, and the optimal integrated model is obtained through weighted average. In this example, the LASSO model and the support vector machine model simulate the R for wheat yield, corn yield, and soil organic carbon changes 2 0.76, 0.59-0.62, 0.68, respectively, below average levels were discarded. R of four indexes of simulation of random forest, XGboost, convolutional neural network and long-short term memory network 2 Are all above 0.93 and are reserved for use. The four models respectively calculate RMSE of the verification set, and 1/RMSE is used as weight to construct an optimal integrated model for the next fine-scale simulation. For example, in the example, the RMSE of the random forest, the XGboost, the convolutional neural network and the long-short term memory network for simulating the wheat yield is respectively 0.43Mg ha -1 、0.29Mg ha -1 、0.35Mg ha -1 、0.35Mg ha -1 The weights corresponding to the four models are 2.33, 3.45, 2.86 and 2.86 respectively, and the final result is obtained according to weighted average of the weights.
6) Ready history (1995-2014) and future(2030 s, 2060 s) North China plain 1km scale grid input data, the format of the input data is consistent with that of the machine learning model, the prepared data is used for driving the optimal machine learning integration model, different management levels are set at the same time, and the yield of wheat and corn, organic carbon in soil and N in the North China plain under historical and future situations are obtained through simulation 2 Spatial profile of O emissions. For this example, under historically optimal nitrogen fertilizer, irrigation and straw return management, the yield of northern China plain wheat was at the historical stage
Maize yields are >>
SOC is
The wheat yield is basically kept unchanged in the future, and the corn yield and the soil organic carbon change show a descending trend.
Therefore, the method can more quickly and efficiently predict the space-time change rule of state variables (such as crop yield, soil organic carbon and the like in a farmland ecosystem) of a certain ecosystem under high space-time resolution and the response of the space-time change rule to different management measures and climate change.
The above-described embodiments are merely preferred embodiments of the present invention, and are not intended to limit the present invention. Various changes and modifications may be made by one of ordinary skill in the pertinent art without departing from the spirit and scope of the present invention. Therefore, the technical scheme obtained by adopting the mode of equivalent replacement or equivalent transformation is within the protection scope of the invention.
Claims (10)
1. A high-precision space-time simulation method for coupling an ecological process model and a machine learning algorithm is characterized by comprising the following steps:
s1: calibrating and verifying relevant parameters of the ecosystem process model by using long-term positioning test data in a target research area;
s2: randomly selecting a plurality of grid points capable of covering the climate type and soil characteristic spatial variation of the research area in the research area, and randomly generating a management measure and climate variation scene combination for each grid point; simulating the state variable change of the ecological system at each grid point and under the scene by using the ecological system process model verified in the step S1 based on the soil data and the historical climate data at the scene corresponding to each grid point;
s3: integrating input data and output data of the ecosystem process model scene simulation in the step S2, and training different types of machine learning models to enable the machine learning models to fully learn and simulate state variable changes of the ecosystem process model simulation; evaluating the performance of machine learning, adaptively eliminating the poor-performance machine learning model, and obtaining an optimal integrated model according to a weighting evaluation method of different model weights;
s4: and taking the optimal integrated model as a proxy model of the process model of the ecosystem, and performing proxy simulation on the basis of spatial data to generate a high-precision digital drawing product required by a research purpose under a high-resolution spatiotemporal scale.
2. The method for high-precision spatiotemporal simulation coupled with an ecological process model and a machine learning algorithm according to claim 1, wherein the step S1 is specifically as follows:
aiming at a specific research purpose, determining a research area to be simulated and a suitable ecosystem process model; acquiring observation data of the ecological system in-situ site in the research area according to the requirements of the ecological system process model, and correcting and verifying key parameters of the ecological system process model by using a differential evolution algorithm.
3. The method of high-precision spatiotemporal simulation coupling an ecological process model and a machine learning algorithm of claim 2, wherein the ecosystem in-situ site observation data comprises one or more of meteorological data, soil property data, farmland ecosystem data, ecosystem state variable data.
4. The method for high-precision spatiotemporal simulation of a coupled ecological process model and machine learning algorithm of claim 3, wherein the meteorological data includes daily temperature, precipitation and solar radiation, the soil property data includes soil organic carbon, pH and soil volume weight, the field ecosystem data includes management data and yield data for stages of seeding, fertilizing and irrigating, and the ecosystem state variable data includes soil organic carbon and greenhouse gas emission observations.
5. The method of high-precision spatiotemporal simulation of a coupled ecological process model and machine learning algorithm of claim 2, in which the key parameters include crop variety parameters, soil organic carbon and N in a field ecosystem 2 O-exhaust process parameters.
6. The method for high-precision spatio-temporal simulation of a coupled ecological process model and a machine learning algorithm according to claim 1, wherein the step S2 is specifically as follows:
in the research area, generating spatial grids according to a set spatial resolution, randomly selecting a plurality of grid points capable of covering the climate type and soil characteristic spatial variation of the research area from all the grids, and randomly generating a management measure and climate change scene combination at each grid point; and (3) collecting and preparing soil data and historical climate data corresponding to the selected grid points, and driving the ecosystem process model verified in the step 1 to simulate interested state variables under historical and future situations.
7. The method for high-precision spatiotemporal simulation of a coupled ecological process model and machine learning algorithm of claim 6, wherein the management measure scenarios refer to various possible human activity management, in a field ecosystem, the management measure scenarios include a crop system scenario, a crop seeding stage scenario, a fertilization scenario, an irrigation scenario, and a straw returning scenario; the climate change situation refers to a future climate change situation and is selected and determined from a global climate model and a shared economic path emission situation.
8. A high-precision space-time simulation method coupling an ecological process model and a machine learning algorithm according to claim 1, wherein in the step S3, the machine learning model comprises a LASSO model, a support vector machine, a random forest, an XGBoost, a convolutional neural network, and a long-short term memory network.
9. The method for high-precision spatio-temporal simulation of coupled ecological process model and machine learning algorithm as claimed in claim 1, wherein in step S3, the root mean square error RMSE and the decision coefficient R are calculated respectively 2 To evaluate the performance of each machine learning model to predict the process model output by setting a screening criterion, R 2 The models with the ranking lower than the average level of each model are abandoned, and a plurality of models with the models represented on the average level are screened out; and according to the model expression, calculating the weight of each screened model, wherein the weight is 1/RMSE, and constructing a weighted average integration model to obtain an optimal integration model.
10. The method for high-precision spatio-temporal simulation of a coupled ecological process model and a machine learning algorithm according to claim 1, wherein the step S4 is specifically as follows:
and collecting and preparing high-precision spatial data, driving the constructed agent model, and realizing rapid simulation through parallel operation to generate a high-precision digital mapping product required by the research purpose.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310049924.XA CN115952421A (en) | 2023-02-01 | 2023-02-01 | High-precision time-space simulation method for coupling ecological process model and machine learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310049924.XA CN115952421A (en) | 2023-02-01 | 2023-02-01 | High-precision time-space simulation method for coupling ecological process model and machine learning algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115952421A true CN115952421A (en) | 2023-04-11 |
Family
ID=87290752
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310049924.XA Pending CN115952421A (en) | 2023-02-01 | 2023-02-01 | High-precision time-space simulation method for coupling ecological process model and machine learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115952421A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117333321A (en) * | 2023-09-27 | 2024-01-02 | 中山大学 | Agricultural irrigation water consumption estimation method, system and medium based on machine learning |
| CN117542443A (en) * | 2023-09-27 | 2024-02-09 | 中国农业大学 | Method and device for balancing yield and relieving nitrogen pollution and electronic equipment |
-
2023
- 2023-02-01 CN CN202310049924.XA patent/CN115952421A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117333321A (en) * | 2023-09-27 | 2024-01-02 | 中山大学 | Agricultural irrigation water consumption estimation method, system and medium based on machine learning |
| CN117542443A (en) * | 2023-09-27 | 2024-02-09 | 中国农业大学 | Method and device for balancing yield and relieving nitrogen pollution and electronic equipment |
| CN117542443B (en) * | 2023-09-27 | 2024-05-24 | 中国农业大学 | Method and device for balancing yield and relieving nitrogen pollution and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109829234B (en) | A kind of across scale Dynamic High-accuracy crop condition monitoring and yield estimation method based on high-definition remote sensing data and crop modeling | |
| Lv et al. | Climate change impacts on regional winter wheat production in main wheat production regions of China | |
| Huang et al. | A multiple crop model ensemble for improving broad-scale yield prediction using Bayesian model averaging | |
| CN115952421A (en) | High-precision time-space simulation method for coupling ecological process model and machine learning algorithm | |
| CN110909933B (en) | Agricultural drought rapid diagnosis and evaluation method coupling crop model and machine learning language | |
| CN107392376A (en) | A kind of crops Meteorological Output Forecasting Methodology and system | |
| WO2023179167A1 (en) | Crop irrigation water demand prediction method based on aquacrop model and svr | |
| CN112598277B (en) | Method for evaluating yield difference of trans-regional reduced winter wheat and improving nitrogen fertilizer efficiency | |
| Yang et al. | Impacts of environment and human activity on grid-scale land cropping suitability and optimization of planting structure, measured based on the MaxEnt model | |
| Tian et al. | Improving performance of Agro-Ecological Zone (AEZ) modeling by cross-scale model coupling: An application to japonica rice production in Northeast China | |
| CN113361191A (en) | Pixel scale winter wheat yield remote sensing estimation method based on multi-scenario simulation | |
| CN106485002B (en) | In the method for complicated landform climatic province estimation sugarcane potential production | |
| CN110633841B (en) | Provincial range plot scale data assimilation yield prediction method based on set sampling | |
| US12112105B1 (en) | Soil-climate intelligent type determining method for rice target yield and nitrogen fertilizer amount | |
| CN110826797B (en) | Method for determining optimal agricultural planting system based on multi-target comprehensive evaluation system | |
| CN115640956B (en) | Future water resource supply and demand balance analysis method | |
| CN109800921A (en) | A kind of Regional Fall Wheat yield estimation method based on remote sensing phenology assimilation and particle swarm optimization algorithm | |
| CN110705182A (en) | Crop breeding adaptive time prediction method coupling crop model and machine learning | |
| CN113822479A (en) | Multi-objective optimization method for regional agricultural planting structure considering production, environment and economic benefits | |
| CN105468854B (en) | Key factor contribution to yield computational methods based on plant growth mechanism model | |
| Cheng et al. | Improving soil available nutrient estimation by integrating modified WOFOST model and time-series earth observations | |
| CN111915096B (en) | Crop yield early-stage forecasting technology based on crop model, remote sensing data and climate forecasting information | |
| Causarano et al. | Improved modeling of soil organic carbon in a semiarid region of Central East Kazakhstan using EPIC | |
| CN109359862A (en) | A kind of real-time yield estimation method of cereal crops and system | |
| CN109614763A (en) | A kind of area crops yield estimation method correcting crop modeling based on multi-source information substep |
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 |