This paper describes a model for (1) studying the genesis of rear-end collisions in freeway traffic and (2) estimating the potential benefits of collision warning systems in such crashes. The model is known as REAMACS (Rear-End Accident Model And Countermeasure Simulation). REAMACS uses a quasi-Monte Carlo routine to generate the initial conditions that determine whether or not a crash occurs. The model was used to estimate the benefits of a collision-avoidance algorithm which provides a warning when the distance to the vehicle ahead is less than a threshold distance which is determined by the closing rate. Applications of the model indicate that a collision warning system with a range of about 250 feet has the potential to reduce the number of all rear-end collisions by 50% and the more serious ones by over 60%. Further, the impact speeds of crashes that do occur are reduced by a 250-foot range system by as much as 38%. (A)
Samenvatting