SHORT-TERM PREDICTION OF WEATHER PARAMETERS USING ONLINE WEATHER FORECASTS

摘要

While people need to know tomorrow’s weather todecide suitable activities and precautions, so do the“intelligent” building management systems. Theaccuracy of the short-term prediction of the ambientconditions is particularly import for the developmentof predictive control strategies. Although the short-termprediction methods for outside air temperaturehave been extensively studied, reliable predictionmethods for solar radiation are yet to be established.This paper discusses the possibility of usingmeteorological forecast data with local observationdata for short-term prediction of the ambientconditions. Based on the analysis of the collecteddata, a new method has been developed fortemperature and solar radiation prediction. It isfound that a linear combination of the observationand the forecast data can effectively improve theaccuracy of short-term (24 hours) prediction. Themethod is also suitable for the prediction ofhumidity and wind speed/direction.