Artificial neural systems for verglass forecast

Presents an artificial neural system used to forecast weather variables such as temperature, humidity, wind speed and direction, along a mountain highway. The final goal of the project is to forecast the ice formation in advance to allow salt scattering and to prevent the “verglass” phenomenon. The system analyses in real time data coming from weather stations placed along a highway and tries to guess the next two hours parameters. A mathematical model is able to detect the probability of ice formation when these parameters exceed a threshold