Forecasting Adverse Weather Situations in the Road Network

Abstract

Weather is an important factor that affects traffic flow and road safety. Adverse weather situations affect the driving conditions directly; hence, drivers must be informed about the weather conditions downstream to adapt their driving. In the framework of intelligent transport systems, several systems have been developed to know the weather situations and inform drivers. However, these systems do not forecast weather in advance, and they need the support of road operators to inform drivers. This paper presents a new autonomous system to forecast weather conditions in a short time and to give users the information obtained. The system uses a set of algorithms and rules to determine the weather and to forecast dangerous situations on the road network. It has been implemented using a multiagent approach and tested with real data. Results are very promising. The system is able to forecast adverse situations with a high degree of quality. This quality makes it possible to trust in the system and to avoid the supervision of operators.