Design, Construction, and Evaluation of an Intelligent Frost Forecasting and Warning System

Abstract

Every year, frost causes the loss of many agricultural products. There is numerous equipment to protect plants against frost. Late turning on these equipment causes inefficiency in raising the air temperature, and early turning them on will increases energy consumption and costs. Therefore, accurately forecasting frost is crucial for turning on the equipment on time. In this research, an intelligent radiation frost forecasting and warning system (IFFS) based on the Internet of Things (IoT) technology was designed and constructed. This system comprises a wireless sensor, computing, and intelligent forecasting based on deep learning methods and warning announcements to the farmer by a message. Intelligent forecasting based on forecasting dew point temperature for the next three hours according to the in-situ measurement of temperature and relative humidity of the air. The meteorological data of the studied regain from 2011-2021 were used to train the network. The IFFS Performance was evaluated. Based on the obtained results, the system accuracy in measuring temperature and relative humidity of the air was 99% and 98%, respectively. The F-score of the IFFS obtained 96%, and the system accuracy in the warning announcement obtained 100%. Finally, applying the IFFS for better protection of plants is recommended.

Keywords

• deep learning
• frost
• internet of things
• intelligent agriculture
• wireless sensor network