Data-driven greenhouse climate regulation in lettuce cultivation using BiLSTM and GRU predictive control
Computers and Electronics in Agriculture · 2026-06 · 247 · 111719
DOI: 10.1016/j.compag.2026.111719 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computers and Electronics in Agriculture, 2026-06. Verfasst von Soumo Emmanuel Arnaud und 2 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: EPSRC.
Abstract (Originalsprache)
Efficient greenhouse management is essential for sustainable food production, but the high energy demand for climate regulation poses significant economic and environmental challenges. While traditional process-based greenhouse models exist, they are often too complex or imprecise for reliable control. To address this, our study introduces a novel data-driven predictive control framework using Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) neural networks within a Model Predictive Control (MPC) architecture. Training data were generated from a validated dynamic model simulating lettuce cultivation under various environmental conditions. The LSTM and GRU networks were trained to predict future greenhouse states — including temperature, humidity, CO 2 concentration, and crop dry matter — with robustness confirmed via 10-fold cross-validation. These networks were embedded into an online MPC controller to optimize heating, ventilation, and CO 2 injection, aiming to minimize energy consumption and maximize crop yield while respecting biological constraints. Results showed that both the LSTM- and GRU-based controllers significantly outperformed a conventional MPC baseline. For example, humidity violations dropped from 54.77% (MPC) to 15.45% (GRU) and 17.71% (LSTM), while day–night temperature deviations were kept below 2 ∘ C . The GRU controller further achieved up to 40% lower computation time than its LSTM counterpart, confirming its real-time feasibility. Overall, the proposed GRU-driven predictive control approach offers a robust and computationally efficient solution for intelligent greenhouse climate automation under practical operational constraints. • GRU and LSTM models were applied for greenhouse climate predictive control. • GRU reduced climate violations by 5% compared to LSTM while matching yield. • GRU required 40% less computation time than LSTM in real-time scenarios. • Both models maintained high crop yield and economic performance. • GRU offers a practical solution for energy-efficient greenhouse control.
Förderung
- EPSRC
Zitieren
APA
Arnaud, S. E., Calisti, M., & Polydoros, A. (2026). Data-driven greenhouse climate regulation in lettuce cultivation using BiLSTM and GRU predictive control. Computers and Electronics in Agriculture, 247, 111719. https://doi.org/10.1016/j.compag.2026.111719
Vancouver
Arnaud SE, Calisti M, Polydoros A. Data-driven greenhouse climate regulation in lettuce cultivation using BiLSTM and GRU predictive control. Computers and Electronics in Agriculture. 2026;247:111719. doi:10.1016/j.compag.2026.111719
BibTeX
@article{Arnaud2026,
author = {Arnaud, Soumo Emmanuel and Calisti, Marcello and Polydoros, Athanasios},
title = {Data-driven greenhouse climate regulation in lettuce cultivation using BiLSTM and GRU predictive control},
journal = {Computers and Electronics in Agriculture},
year = {2026},
volume = {247},
pages = {111719},
doi = {10.1016/j.compag.2026.111719}
}
EUVIMED