Multistep ahead prediction of temperature and humidity in solar greenhouse based on FAM-LSTM model

Yang, Yongxia; Gao, Pan; Sun, Zhangtong; Wang, Haoyu; Lu, Miao; Liu, Yingying; Hu, Jin

doi:10.1016/j.compag.2023.108261

Cited by 9 publications

(4 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This mechanism dynamically adjusts weight distribution, prioritizing more significant information, thereby effectively enhancing the model's capacity to process time series data. In contrast to the method proposed by Yang et al [40], where a feedforward attention mechanism is directly added to LSTM, our approach, which combines an encoder-decoder with residual learning, not only augments the model's ability to mine complex data but also improves the accuracy and performance of time series prediction. This indicates that through carefully designed feature fusion and attention mechanisms, the LSTM model can be further optimized to accommodate the increasingly complex and variable demands of time series analysis.…”

Section: Discussionmentioning

confidence: 99%

“…Attention mechanisms can capture the most important data for current tasks in complex datasets [45], achieving tremendous success in computer vision tasks [46]. In large-scale data processing, compared to conventional attention mechanisms, feedforward attention mechanisms can demonstrate better performance in dealing with medium to long-term time series issues [40].…”

Section: Feedforward Attention Mechanismmentioning

confidence: 99%

“…In particular, we focused on testing the model's performance in predicting the next day's surface soil moisture (0-7 cm depth). Validation results indicated that the REDF-LSTM model showed significant performance improvements in complex soil moisture time series forecasting tasks, compared to traditional LSTM models and currently more advanced models such as EDLSTM [39], FAMLSTM [40], and GANBiLSTM [41].…”

Section: Introductionmentioning

confidence: 98%

See 2 more Smart Citations

Enhancing Soil Moisture Forecasting Accuracy with REDF-LSTM: Integrating Residual En-Decoding and Feature Attention Mechanisms

Li,

Zhang,

et al. 2024

Water

View full text Add to dashboard Cite

This study introduces an innovative deep learning model, Residual-EnDecode-Feedforward Attention Mechanism-Long Short-Term Memory (REDF-LSTM), designed to overcome the high uncertainty challenges faced by traditional soil moisture prediction methods. The REDF-LSTM model, by integrating a residual learning encoder–decoder LSTM layer, enhanced LSTM layers, and feedforward attention, not only captures the deep features of time series data but also optimizes the model’s ability to identify key influencing factors, including land surface features, atmospheric conditions, and other static environmental variables. Unlike existing methods, the innovation of this model lies in its first-time combination of the residual learning encoder–decoder and feedforward attention mechanisms in the soil moisture prediction field. It delves into the complex patterns of time series through the encoder–decoder structure and accurately locates key influencing factors through the feedforward attention mechanism, significantly improving predictive performance. The choice to combine the feedforward attention mechanism and encoder–decoder with the LSTM model is to fully leverage their advantages in processing complex data sequences and enhancing the model’s focus on important features, aiming for more accurate soil moisture prediction. After comparison with current advanced models such as EDLSTM, FAMLSTM, and GANBiLSTM, our REDF-LSTM demonstrated the best performance. Compared to traditional LSTM models, it achieved an average improvement of 13.07% in R2, 20.98% in RMSE, 24.86% in BIAS, and 11.1% in KGE key performance indicators, fully proving its superior predictive capability and potential application value in precision agriculture and ecosystem management.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Feedforward Attention Mechanismmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

Enhancing Soil Moisture Forecasting Accuracy with REDF-LSTM: Integrating Residual En-Decoding and Feature Attention Mechanisms

Li,

Zhang,

et al. 2024

Water

View full text Add to dashboard Cite

show abstract

“…With the development of sensor technology, the amount of data about the greenhouse environment is increasing. The use of data-driven time series prediction model construction is gaining attention [12].…”

Section: Introductionmentioning

confidence: 99%

Environmental Prediction Model of Solar Greenhouse Based on Improved Harris Hawks Optimization-CatBoost

Yang,

Ren,

Wang

et al. 2024

Sustainability

View full text Add to dashboard Cite

Solar greenhouses provide a favorable climate environment for the production of counter-seasonal crops in northern China. The greenhouse environment is a key factor affecting crop growth, so accurate prediction of greenhouse environment changes helps to precisely regulate the crop growth environment and helps to promote the growth of fruits and vegetables. In this study, an environmental prediction model based on the combination of a gradient boosting tree and the Harris hawk optimization algorithm (IHHO-Catboost) is constructed, and in response to the problems of the HHO algorithm, such as the fact that the adjustment of the search process is not flexible enough, it cannot be targeted to carry out a stage search, and sometimes it will fall into the local optimum to make the algorithm’s search accuracy relatively poor, an algorithm based on the improved Harris hawk optimization (IHHO) algorithm-based parameter identification method is constructed. The model considers the internal and external environmental and regulatory factors affecting crop growth, which include indoor temperature and humidity, light intensity, carbon dioxide concentration, soil temperature and humidity, outdoor temperature and humidity, light intensity, carbon dioxide concentration, wind direction, wind speed, and opening and closing of upper and lower air openings of the cotton quilt, and is input into a prediction model with a time series for training and testing. The experimental results show that the MAE (mean absolute error) values of temperature, relative humidity, carbon dioxide concentration, and light intensity of the model are reduced to 49.8%, 35.3%, 72.7%, and 32.1%, respectively, compared with LSTM (Long Short-Term Memory), which is a significant decrease in error. It shows that the proposed multi-parameter prediction model for solar greenhouse environments presents an effective method for accurate prediction of environmental data in solar greenhouses. The model not only improves prediction accuracy but also reduces dependence on large data volumes, reduces computational costs, and improves the transparency and interpretability of the model. Through this approach, an effective tool for greenhouse agriculture is provided to help farmers optimize the use of resources, reduce waste, and improve crop yield and quality, ultimately leading to a more efficient and environmentally friendly agricultural production system.

show abstract

Spatial temperature distribution and ground thermal storage in the plastic greenhouse: An experimental and modeling study

Li,

Jian,

Wang

et al. 2024

Journal of Energy Storage

View full text Add to dashboard Cite

Multistep ahead prediction of temperature and humidity in solar greenhouse based on FAM-LSTM model

Cited by 9 publications

References 50 publications

Enhancing Soil Moisture Forecasting Accuracy with REDF-LSTM: Integrating Residual En-Decoding and Feature Attention Mechanisms

Enhancing Soil Moisture Forecasting Accuracy with REDF-LSTM: Integrating Residual En-Decoding and Feature Attention Mechanisms

Environmental Prediction Model of Solar Greenhouse Based on Improved Harris Hawks Optimization-CatBoost

Spatial temperature distribution and ground thermal storage in the plastic greenhouse: An experimental and modeling study

Contact Info

Product

Resources

About