WBGT Index Forecast Using Time Series Models in Smart Cities

WBGT (Wet Bulb Globe Temperature) was originally championed by the United States military and has been widely implemented in daily training to prevent casualties or injuries among soldiers due to unfavorable high temperature and humidity conditions during the summer, and it has been widely used in various fields, such as marathons, military training, and travel. Starting from the daily periodicity of WBGT, this paper uses the Holt-Winters 24 h, and discusses the feasibility of its autocorrelation prediction. The prediction results were evaluated using the time series cross-validation method and RMSE. Two experiments were conducted with the dataset acquired by NCSCO and the self-collected Dongguan University of Technology dataset(DGUT-D). First, a preliminary experiment was conducted using NCSCO data to explore the feasibility of WBGT autocorrelation pre-diction, and then the DGUT-D was used to predict the 24-h WBGT of the Songshan Lake Campus of the Dongguan University of Technology (DGUT).