A Reliable Routing Algorithm with Real Time and Historical Information

A traffic network with the following information structure is considered. The link travel times have probability distributions which are known from historical data. When the traveler arrives at a node, the actual travel times on the down streaming links are revealed to him. For given origin, destination, departure time and scheduled arrival time, the reliability of a routing strategy can be measured by the probability of arriving at destination before scheduled time. In this paper an algorithm is provided for generating a route by taking into account both the historical statistical information and real time information. The route generated has maximum probability of arriving at the destination before the scheduled time, under the adopted information structure