SPREADING OF TRAFFIC JAM IN A TRAFFIC FLOW MODEL

A cellular automaton (CA) model is presented to simulate the traffic jam induced by a traffic accident. The spreading of jamming cars induced by a car crash is investigated by computer simulation. An analogy is proposed between the crystal growth and the traffic-jam spreading. The scaling behavior of the traffic-jam spreading is studied. It is shown that the number N of jamming cars scales as N almost-equal-to t2.34+/-0.03 for p above the dynamical jamming transition p(c)=0.35 and N almost-equal-to t1.07 below p(c) where t is the time and p is the density of cars. The time constant t(s), which is the time required for all cars to stop, scales as t(s) almost-equal-to p-1.07+/-0.03 for p<p(c). The following scaling form is found: p-1 N almost-equal-to (p1.07t)1.07 for p<p(c).