设施农业能源互联网智能预警理论：评述与展望

农业信息化与电气化是现代农业发展的必由之路，建设能源驱动的现代设施农业成为防范农业气象灾害的一种有效途径。然而，农业环境监控系统与供能系统的运行和管理相互割裂，无法应对温室停电、农电负荷过载等连锁风险。因此，该文将农业科学、信息科学与电力科学相关理论融合，考虑设施农业环境、农作物和能源系统状态，从系统科学的角度出发研究人工智能在农业工程和电力工程交叉学科的应用，对基于多源数据融合的设施农业能源互联网智能预警理论进行评述与展望。该研究可为保障设施农业的安全生产，推动智慧农业技术与农业智能装备的升级和发展，实现农业园区的信息化和电气化提供参考。

Intelligent early warning theory of the facility agricultural energy internet: Review and prospect

Abstract: Electrification and informatization of agricultural machinery will be the only way for the development of modern agriculture. The energy-driven modern agriculture has also become an effective way to prevent meteorological disasters. However, the agricultural environment monitoring and the energy supply system are operated separately from each other at present. It cannot cope with the chain risks, such as greenhouse power outages and agricultural power overload. An intelligent early warning of agricultural energy internet is basically unexplored so far. Therefore, it is very necessary to integrate the relevant theories of agriculture, informatics, and electrical science, particularly for the facility agriculture, crops, and energy system. This review was focused on the application of artificial intelligence technology in the interdisciplinary of agricultural and power engineering, in terms of system science. An intelligent early warning of agricultural energy internet was also explored using multi-source data fusion. The current research progress of the intelligent early warning was analyzed from three aspects: the interconnection between energy and agriculture, the correlation between energy and agricultural security, and the intelligent early warning of agricultural production. Among them, the interconnection was caused the existence of collaborative security, whereas, the intelligent early warning was one of the measures to solve the collaborative security. After that, the key technologies were also introduced from the fundamental and applied system. Specifically, the concept of the agricultural energy internet was introduced to clarify the difference and connection with general concepts, such as the agricultural Internet of Things, as well as the interaction between facility agriculture and the energy system. The applied system included three aspects, such as the current research, key issues, and difficulties, together with the future directions. It was found that the current intelligent early warning was focused on the specific issues in the various branches of agricultural informatization and electrical engineering. Some specific research was relatively mature, including the plant phenotype monitoring, the facility agricultural environment monitoring, and the static safety analysis of the energy system. Therefore, the key issue of the intelligent early warning can be widely expected to integrate the multi-source sensor data in the agricultural energy internet. As such, the early warning of environmental factors can be realized in the protected agriculture and energy system. The following brief ideas can be proposed. 1) To break through the early warning of the facility agricultural energy system. Decision-making can also be provided for the operation and control of the energy system in intensive agricultural production. 2) To propose the risk propagation in the agricultural energy system, particularly for the unified modeling of the facility agricultural environment and the integrated energy supply system. It can provide a model basis for the early warning of energy internet. 3) To break through the early warning using data fusion of energy and agriculture, particularly for the early warning of crop physiological state. It can provide the decision-making information for the collaborative control of protected facility agriculture environment and integrated energy supply system. This finding can provide a strong reference to the safe production of facility agriculture, especially on the upgrading and development of smart agricultural technology and intelligent equipment for agricultural informatization and electrification.