Modeling and optimization of environment in agricultural greenhouses for improving cleaner and sustainable crop production

Resource-use efficiency and crop yield are significant factors in the management of agricultural greenhouse. Appropriate modeling methods effectively improve the control performance and efficiency of the greenhouse system and are conducive to the design of water and energy-saving strategies. Meanwhile, the extreme environment could be forecasted in advance, which reduces pests and diseases as well as provides high-quality food. Accordingly, the interest of the scientific community in greenhouse modeling and optimizing has grown considerably. The objective of this work is to provide guidance and insight into the topic by reviewing 73 representative articles and to further support cleaner and sustainable crop production. Compared to the existing literature review, this work details the approaches to improve the greenhouse model in the aspects of parameter identification, structure and process optimization, and multi-model integration to better model complex greenhouse system. Furthermore, a statistical study has been carried out to summarize popular technology and future trends. It was found that dynamic and neural network techniques are most commonly used to establish the greenhouse model and the heuristic algorithm is popular to improve the accuracy and generalization ability of the model. Notably, deep learning, the combination of knowledge and data, and coupling between the greenhouse system elements have been considered as future valuable development. (c) 2020 Elsevier Ltd. All rights reserved.

Automated summary

Resource-use efficiency and crop yield are important factors in agricultural greenhouse management. Using appropriate modeling methods can enhance control performance, design water and energy-saving strategies, and forecast extreme environments in advance, which can reduce pests and diseases and provide high-quality food. Interest in greenhouse modeling and optimization has increased significantly, and this study aims to provide guidance by reviewing 73 representative articles. It details approaches to improve greenhouse models, such as parameter identification, structure optimization, and multi-model integration, and highlights popular technologies and future trends, such as dynamic and neural network techniques and heuristic algorithms. Future valuable developments include deep learning, knowledge-data combinations, and coupling between greenhouse system elements.