Speed stochastic processes and freeway reliability estimation: Evidence from the A22 Freeway, Italy

Abstract

In this paper, a criterion for predicting the reliability of freeway traffic flow is presented. The idea is based on an analysis of spot speed time series divided into sequences of events of random and homogeneous traffic processes. For each process, the flow rate and density were calculated; then the relationships between parameters of spot speed processes and vehicular density were obtained. Using these relationships and a simulation procedure for the spot speed process, a formulation for predicting the reliability of traffic flow moving along the offside lane on the freeway roadway was derived. Through this formulation and the measurements of flow rate and speed, the probability of instability of the roadway under examination was calculated. Furthermore, with an inverse procedure, the capacity of a lane was estimated: setting the reliability value close to 1 and considering a 15 min interval, the resulting flow rate represented the capacity. Thus, the simulation procedure was used to provide, in real time, the incoming instability conditions on the roadway and to estimate roadway capacity for the A22 Brenner Freeway, Italy.