engleski

Wind forecast correction in road traffic management decision support system

We present an ongoing work of development of a new algorithm for suggesting road closures due to strong winds (especially Bora on Adriatic coast). The algorithm will improve the current decision support system that is used by the road traffic managers who keep the road closures optimally balanced between traffic safety and economic losses. We use neural networks and as features we use: weighted multiple successive forecasts for the next 12-hours period, the overall shapes during respective strong wind periods, and their congruence with the latest measurements. The planned algorithm's application potential is wider than traffic control management so it can serve as as an example of using previous successive forecasts to more accurately estimate the wind power production or offering decision support in different areas that depend on other meteorological elements (e.g. temperature, air pressure, precipitation etc.)

Wind forecast correction, Neural networks, Traffic management

nije evidentirano