CN117392556A - Carbon sink monitoring method based on improved LSTM algorithm - Google Patents

Carbon sink monitoring method based on improved LSTM algorithm Download PDF

Info

Publication number
CN117392556A
CN117392556A CN202311444389.4A CN202311444389A CN117392556A CN 117392556 A CN117392556 A CN 117392556A CN 202311444389 A CN202311444389 A CN 202311444389A CN 117392556 A CN117392556 A CN 117392556A
Authority
CN
China
Prior art keywords
carbon
data
remote sensing
monitoring
sensing data
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
Application number
CN202311444389.4A
Other languages
Chinese (zh)
Inventor
王茂发
周翔
杨凤山
安嘉诺
林英茂
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guilin University of Electronic Technology
Original Assignee
Guilin University of Electronic Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guilin University of Electronic Technology filed Critical Guilin University of Electronic Technology
Priority to CN202311444389.4A priority Critical patent/CN117392556A/en
Publication of CN117392556A publication Critical patent/CN117392556A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • G06V20/13—Satellite images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16C—COMPUTATIONAL CHEMISTRY; CHEMOINFORMATICS; COMPUTATIONAL MATERIALS SCIENCE
    • G16C20/00—Chemoinformatics, i.e. ICT specially adapted for the handling of physicochemical or structural data of chemical particles, elements, compounds or mixtures
    • G16C20/10—Analysis or design of chemical reactions, syntheses or processes
    • 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
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/80—Management or planning
    • Y02P90/84—Greenhouse gas [GHG] management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Chemical & Material Sciences (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Remote Sensing (AREA)
  • Astronomy & Astrophysics (AREA)
  • Medical Informatics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Databases & Information Systems (AREA)
  • Analytical Chemistry (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a carbon sink monitoring method based on an improved LSTM model, which comprises a remote sensing data acquisition module, a data processing and feature extraction module and a carbon storage monitoring and quantifying module which are sequentially connected. The main function of the remote sensing data acquisition module is to acquire multispectral remote sensing data, and the data processing and feature extraction module is responsible for selecting and identifying the multispectral remote sensing data so as to extract carbon storage monitoring data. Finally, the carbon storage monitoring and quantifying module predicts the carbon storage monitoring data by applying a machine learning model, thereby obtaining a carbon storage amount, and then calculates a carbon sink amount according to the carbon storage amount. The method fully utilizes the information of multiple aspects of remote sensing data, and realizes the intelligent and accurate calculation and analysis of the carbon sequestration.

Description

Carbon sink monitoring method based on improved LSTM algorithm
Technical Field
The invention relates to the technical field of satellite remote sensing, in particular to a carbon sink monitoring method based on an improved LSTM algorithm.
Background
In recent years, global concerns about climate change and environmental protection have increased, and many innovative solutions have been induced, some of which involve technologies to reduce greenhouse gas emissions and increase carbon sequestration. Carbon sink refers to a process or mechanism for reducing the concentration of greenhouse gases in the atmosphere by planting trees, performing forest management, recovering vegetation, and the like, absorbing carbon dioxide in the atmosphere by photosynthesis of plants, and fixing the carbon dioxide in the vegetation and soil.
Currently, the prior art for computing and monitoring carbon sinks mostly requires on-site data collection in the monitored area, and then fitting and computing based on these data. This method requires a lot of manpower and material resources, and has a low level of intelligence. Furthermore, these methods have limited accuracy because they typically rely on theoretical formulas or previous experience for analysis. Another method is to construct a fitting model by machine learning, but this still requires collecting field data of the monitored area, which cannot meet the requirement of intellectualization.
And a model is built by using a deep learning technology, such as a neural network, so that carbon sink monitoring is performed in a more automatic and accurate manner. Through analysis and model training of large-scale remote sensing data, meteorological data and land utilization data, the deep learning can improve the monitoring accuracy, and reduce the requirement of relying on-site data acquisition, so that the carbon sink monitoring is more intelligent and efficient.
Disclosure of Invention
In order to overcome the defects of the current carbon sink monitoring method, a carbon sink monitoring method based on an improved LSTM model is provided. The method extracts multidimensional information by processing remote sensing data, and effectively and accurately calculates and monitors carbon sink in an intelligent mode. This means that we use the remote sensing data to obtain information from multiple angles and then by comprehensively analyzing this information to conduct intelligent monitoring and calculation of carbon sequestration.
The technical scheme for realizing the aim of the invention is as follows:
a method for monitoring carbon sink based on an improved LSTM model, comprising:
predicting carbon deposit monitoring data using the modified LSTM model to obtain carbon deposits and calculating carbon sink based on the carbon deposits includes the steps of:
step S1, collecting remote sensing data;
step S2, screening, selecting and identifying remote sensing data, and extracting required information including carbon storage monitoring data;
s3, analyzing and predicting the carbon storage monitoring data by utilizing an improved LSTM model to estimate the carbon storage;
and S4, calculating the carbon sink based on the estimation result of the carbon reserves.
Further, the step S1 includes the steps of first obtaining initial telemetry data, wherein the initial telemetry data is a multispectral telemetry image. The multispectral remote sensing image is then preprocessed to obtain remote sensing data, including image enhancement and correction.
Further, the step S2 is performed by selecting the remote sensing data according to different wavebands to obtain the remote sensing data under different wavebands, such as
Remote sensing data of red wave band, green wave band and infrared wave band. And analyzing and calculating the remote sensing data under different wave bands to obtain carbon storage monitoring data, wherein the carbon storage monitoring data comprise vegetation types, vegetation conditions, surface characteristics and other information.
Further, the step S3 includes the step of predicting the carbon storage monitoring data using the modified LSTM model. The difference in carbon reserves, in particular Net Ecosystem Productivity (NEP), is then calculated based on a specific scaling factor, which is the difference in the respiration of the vegetation primary productivity (NPP) soil microorganisms, i.e. the net absorption or net storage of the ecosystem carbon cycle, is widely used to estimate the vegetation carbon sink/source. The NEP calculation formula is as follows:
NEP＝NPP-R H (1)
wherein R is H For microbial respiration, a model calculation established by Pei can be adopted. R is R H The calculation formula of (2) is as follows:
R H ＝0.22×(e (0.0912T) +ln(0.3415R+1))×30×46.5％ (2)
where T is the annual average temperature (. Degree.C.) and R is the annual average precipitation (mm). When NEP is greater than 0, the fixed carbon of the vegetation is greater than the carbon of the soil emission, indicating that the carbon cycle of the vegetation ecosystem is carbon sink and vice versa.
Further, the step S4 involves the steps of acquiring historical carbon storage monitoring data, and filling and integrating the data to obtain filled and integrated data. Based on these filled integrated data, a machine learning model is trained and validated by mean square error, resulting in a trained improved LSTM model. Finally, the trained modified LSTM model is used to predict carbon storage monitoring data. This series of steps helps to improve the accuracy and reliability of the carbon reserves and carbon sink.
The beneficial effects of the invention are as follows:
according to the invention, satellite remote sensing data are utilized to analyze different data, and carbon storage monitoring data capable of reflecting various information is extracted. These monitoring data are then fitted by improving the LSTM model, and the carbon sink is finally calculated accurately and efficiently. The invention does not need to collect data on site, but only needs to acquire satellite remote sensing data, thereby improving the level of intelligence. Unlike traditional fixed calculation method, the method does not need to use specific theoretical formula, but learns the relation between the carbon storage monitoring data and the carbon storage through improving LSTM model, thereby accurately predicting the carbon storage and realizing accurate monitoring of carbon sink.
Drawings
FIG. 1 is a diagram of an LSTM model in accordance with an embodiment of the present invention;
FIG. 2 is a schematic flow chart of the method of the present invention;
FIG. 3 is a schematic diagram of the framework of the present invention.
Detailed Description
The carbon sink monitoring method based on the improved LSTM model is based on analysis of large-scale carbon sink data and environmental characteristics, and a complex relation between carbon sink and environmental parameters is established through a machine learning model. The method utilizes the correlation between environmental factors (such as vegetation coverage, soil type, precipitation, etc.) and carbon sequestration data, and the trend of carbon reserves under specific environmental conditions. By training the existing carbon sink data, the influence of different environmental factors on the carbon sink can be identified and understood by improving the LSTM model, so that the change trend of the carbon sink in the future can be predicted. The method does not need to collect data in the field, only needs to acquire satellite remote sensing data or other environment monitoring data, and therefore is simple to operate and has high accuracy. The method can be widely applied to the carbon sink monitoring field, and provides powerful support for environmental protection and carbon emission control.
Specifically, referring to fig. 2, the method specifically includes:
step S1, collecting remote sensing data;
more specifically, the initial telemetry data collected is the multispectral telemetry image. The multispectral remote sensing image is preprocessed to obtain remote sensing data, including image enhancement and correction.
Step S2, screening, selecting and identifying remote sensing data, and extracting required information including carbon storage monitoring data;
more specifically, the remote sensing data is selected according to different wave bands, so as to obtain the remote sensing data in different wave bands, such as the remote sensing data in red wave band, green wave band and infrared wave band. And analyzing and calculating the remote sensing data under different wave bands to obtain carbon storage monitoring data, wherein the carbon storage monitoring data comprise vegetation types, vegetation conditions, surface characteristics and other information.
S3, analyzing and predicting the carbon storage monitoring data by utilizing an improved LSTM model to estimate the carbon storage;
more specifically, the improved LSTM model is used to predict carbon storage monitoring data. Then, calculating a difference in carbon reserves based on the specific conversion coefficient; the improved LSTM model phase involves the following parts:
step S301: LSTM is a variant of a Recurrent Neural Network (RNN) for processing and modeling sequence data. LSTM aims to solve the long-term dependency problem in conventional RNNs, which have a specific structure that can better capture and memorize long-term dependencies in input sequences. The structure of the LSTM model is shown in fig. 1, and its forward calculation can be expressed as:
i t ＝σ(W i ·[h t-1 ，x t ]+b i ) (3)
f t ＝σ(W f ·[h t-1 ，x t ]+b f ) (4)
o t ＝σ(W o ·[h t-1 ，x t ]+b o ) (7)
h t ＝o t *tanh(C t ) (8)
in the equations, i, f, C and o represent input gates, forget gates, cell states and output gates, respectively. W and b are the corresponding weight matrix and bias term, respectively. Sigma and tanh are the sigmoid function and hyperbolic tangent activation function, respectively. The state of a cell can be considered a way of transmitting information, allowing the information to be transferred in sequence.
Step S302: based on the characteristics of the LSTM model itself, the LSTM model is improved by the following aspects. Including upgrading from single-layer LSTM to three-layer LSTM models, adding Dropout layers, selecting activation functions, and loss function selection.
The upgrade from the single-layer LSTM to the three-layer LSTM is to upgrade the model into a three-layer LSTM model in the improvement process. This means that now three LSTM layers are stacked together in order to process the input sequence;
the Dropout layers are added to mitigate the problem of overfitting, with three Dropout layers added between each LSTM layer. Dropout is a regularization technique that randomly closes a portion of neurons during training, thereby reducing the excessive dependence of the model on training data and improving generalization ability. In this case, the drop rate (Dropout rate) of each Dropout layer is 0.3, which means that about 30% of neurons will be randomly turned off in each training step;
the selection activation function is that in the improvement process, the ultra tangent (Tanh) is selected as the activation function of the neurons in the LSTM model. Tanh is an S-type activation function with an output ranging from-1 to 1, which helps control the flow of information. It is commonly used in LSTM to activate neurons;
the loss function selection is to select a mean square error as the loss function of the model. MSE is commonly used for regression problems to measure the square error between the model's predictions and the actual targets, helping the model optimize parameters to reduce the error.
And S4, calculating the carbon sink based on the estimation result of the carbon reserves.
More specifically, calculating the carbon sink comprises the following steps:
step S401: net Ecosystem Productivity (NEP) is the difference in the respiration of primary plant growth productivity (NPP) soil microorganisms, i.e., net absorption or net storage of ecosystem carbon circulation, is widely used to estimate vegetation carbon sink/source. The NEP calculation formula is as follows:
NEP＝NPP-R H (9)
step S402: wherein R is H For microbial respiration, a model calculation established by Pei can be adopted. R is R H The calculation formula of (2) is as follows:
R H ＝0.22×(e (0.0912T) +ln(0.3415R+1))×30×46.5％ (10)
where T is the annual average temperature (. Degree.C.) and R is the annual average precipitation (mm). When NEP is greater than 0, the fixed carbon of the vegetation is greater than the carbon of the soil emission, indicating that the carbon cycle of the vegetation ecosystem is carbon sink and vice versa.
Specifically, historical carbon monitoring data is acquired and populated and integrated to obtain populated integrated data. Based on these filled integrated data, a machine learning model is trained and validated by mean square error, resulting in a trained improved LSTM model. Finally, the trained modified LSTM model is used to predict carbon storage monitoring data. This series of steps helps to improve the accuracy and reliability of the carbon reserves and carbon sink.
In addition, to enhance the performance of the improved LSTM model, it is contemplated that historical carbon monitoring data may be acquired and data populated and integrated to construct a more comprehensive data set. Then, the improved LSTM model is trained and verified by methods such as error placement and the like, so that high accuracy and robustness in predicting the carbon storage monitoring data are ensured.

Claims (5)

1. The carbon sink monitoring method based on the improved LSTM model is characterized by comprising the following steps of:
step S1, collecting remote sensing data;
step S2, screening, selecting and identifying remote sensing data, and extracting required information including carbon storage monitoring data;
s3, analyzing and predicting the carbon storage monitoring data by utilizing an improved LSTM model to estimate the carbon storage;
and S4, calculating the carbon sink based on the estimation result of the carbon reserves.
2. The method for monitoring carbon sink based on the modified LSTM model of claim 1, wherein the step S1 specifically includes:
s101, collecting multispectral remote sensing data, wherein the multispectral remote sensing data comprises a multi-band remote sensing image obtained by using satellites or other remote sensing technologies so as to capture different characteristic information of the earth surface;
step S102, preprocessing the multispectral remote sensing image to obtain available remote sensing data. This preprocessing includes image enhancement to improve the quality and contrast of the image, and image correction to eliminate image distortions that may be caused by sensors, the atmosphere, or other factors, ensuring the accuracy and consistency of the remote sensing data. These preprocessing steps help ensure the accuracy of subsequent data analysis and carbon storage monitoring.
3. The method for monitoring carbon sink based on the modified LSTM model as claimed in claim 1, wherein the step S2 involves the steps of:
step S201, selecting corresponding information from the remote sensing data, wherein the information comprises the remote sensing data of red wave band, green wave band and infrared wave band. In the step, remote sensing data of different wave bands can be selected according to different vegetation characteristics and surface characteristics so as to ensure that rich and various information is obtained;
and step S202, independent analysis and calculation are carried out on remote sensing data in different wave bands. This includes statistical, mathematical modeling and feature extraction of the data for each band to obtain quantitative and qualitative information about vegetation and the earth's surface;
and S203, extracting carbon storage monitoring data from remote sensing data of different wave bands, wherein the carbon storage monitoring data comprise vegetation types, vegetation conditions, surface features and other information. This step involves data mining and pattern recognition techniques to identify different vegetation categories, evaluate their health, and detect surface features, providing useful data for subsequent monitoring of the carbon reservoirs.
4. The method for monitoring carbon sink based on the modified LSTM model as claimed in claim 1, wherein the step S3 is performed as follows:
and step S301, predicting the carbon storage monitoring data by using the improved LSTM model. In the step, the model deduces future carbon storage monitoring data by learning the trend and the mode of the historical data, so that the time sequence prediction of the carbon storage monitoring is realized;
step S302, calculating the difference of the carbon reserves according to the specific conversion coefficient. Specifically, net Ecosystem Productivity (NEP) is the difference in the respiration of vegetation primary productivity (NPP) soil microorganisms, i.e., net absorption or net storage of ecosystem carbon circulation, is widely used to estimate vegetation carbon sink/source. The NEP calculation formula is as follows:
NEP＝NPP-R H (1)
wherein R is H For microbial respiration, a model calculation established by Pei can be adopted. R is R H The calculation formula of (2) is as follows:
R H ＝0.22×(e (0.0912T) +ln(0.3415R+1))×30×46.5％ (2)
where T is the annual average temperature (. Degree.C.) and R is the annual average precipitation (mm). When NEP is greater than 0, the fixed carbon of the vegetation is greater than the carbon discharged from the soil, which indicates that the carbon cycle of the vegetation ecosystem is a carbon sink and vice versa;
step S303, finally, obtaining carbon sink, which is the result of calculating the difference of the carbon reserves. This step provides information about the carbon sink level in a particular region or time period to enable monitoring and assessment of carbon reserves.
5. The method of claim 1, wherein the predicting of the carbon storage monitoring data by the modified LSTM model comprises: historical carbon monitoring data are collected and data filling and integration are performed on the historical data to obtain a filled and integrated data set. Based on the filled and integrated data set, the improved LSTM model is trained and validated using a mean square error approach. Training results in a validated modified LSTM model. The carbon storage monitoring data is predicted using a trained modified LSTM model.
CN202311444389.4A 2023-11-01 2023-11-01 Carbon sink monitoring method based on improved LSTM algorithm Pending CN117392556A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311444389.4A CN117392556A (en) 2023-11-01 2023-11-01 Carbon sink monitoring method based on improved LSTM algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311444389.4A CN117392556A (en) 2023-11-01 2023-11-01 Carbon sink monitoring method based on improved LSTM algorithm

Publications (1)

Publication Number Publication Date
CN117392556A true CN117392556A (en) 2024-01-12

Family

ID=89435653

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311444389.4A Pending CN117392556A (en) 2023-11-01 2023-11-01 Carbon sink monitoring method based on improved LSTM algorithm

Country Status (1)

Country Link
CN (1) CN117392556A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117953389A (en) * 2024-01-29 2024-04-30 西北师范大学 Grassland carbon sink potential estimation method and system

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117953389A (en) * 2024-01-29 2024-04-30 西北师范大学 Grassland carbon sink potential estimation method and system

Similar Documents

Publication Publication Date Title
US20220067614A1 (en) Apparatus and method for crop yield prediction
Sharma et al. Wheat crop yield prediction using deep LSTM model
CN110991625B (en) Surface anomaly remote sensing monitoring method and device based on recurrent neural network
CN114648705B (en) Carbon sink monitoring system and method based on satellite remote sensing
CN117392556A (en) Carbon sink monitoring method based on improved LSTM algorithm
CN114821315B (en) Remote sensing image cultivated land block extraction method combining edge detection and multitask learning
CN117114913A (en) Intelligent agricultural data acquisition system based on big data
Balti et al. Big data based architecture for drought forecasting using LSTM, ARIMA, and Prophet: Case study of the Jiangsu Province, China
CN111460733A (en) Mountain fire early warning method based on deep learning
Zhang et al. Using an object-based machine learning ensemble approach to upscale evapotranspiration measured from eddy covariance towers in a subtropical wetland
Tang et al. The use of serially complete station data to improve the temporal continuity of gridded precipitation and temperature estimates
CN113537645A (en) Soybean yield prediction method based on machine learning fusion satellite and weather data
CN112464172B (en) Active and passive remote sensing inversion method and device for growth parameters
CN113343783A (en) Intelligent crop identification and growth prediction method and system
Zheng et al. Application of complete ensemble empirical mode decomposition based multi-stream informer (CEEMD-MsI) in PM2. 5 concentration long-term prediction
Kumawat et al. Time-Variant Satellite Vegetation Classification Enabled by Hybrid Metaheuristic-Based Adaptive Time-Weighted Dynamic Time Warping
CN117423003A (en) Winter wheat seedling condition grading remote sensing monitoring method in overwintering period
Mu et al. PIRT: A physics-informed red tide deep learning forecast model considering causal-inferred predictors selection
CN117132884A (en) Crop remote sensing intelligent extraction method based on land parcel scale
Miralles et al. Bayesian modeling of insurance claims for hail damage
WO2023245399A1 (en) Rice production potential simulation method based on land system and climate change coupling
CN111325384B (en) NDVI prediction method combining statistical characteristics and convolutional neural network model
Cao et al. Probabilistic runoff forecasting considering stepwise decomposition framework and external factor integration structure
Balamurugan et al. Artificial Intelligence Based Smart Farming and Data Collection Using Deep Learning
Bafti et al. Automated actual evapotranspiration estimation: Hybrid model of a novel attention based U-Net and metaheuristic optimization algorithms

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