Design of Agricultural Environmental Data Collection System Based on Internet of Things

An agricultural environment data collection system based on the Internet of things is designed. The system uses support vector machine regression (SVMR) algorithm to process the raw data, the predict daily air temperature, relative humidity and wind speed value, in order to help predict the spread of harmful bacteria in farmland. The experimental results show that the mean absolute error of predicted average air temperature and the actual measured values is 1 °C. For the average relative air humidity, the mean absolute error of predict values and the actual measured values is 5%. The mean absolute error of predict the average wind speed value and the actual measured values is 1 km/h. The system data are accessed through the Internet of things environment. Eventually, the system will contribute to the field crop management staff to better manage and prevent the spread of germs.