Abstract
This research looked at lane-changing behavior in multilane highway modeling by a cellular automaton. The effects of speed limit and stochastic noise were analyzed. A new parameter is introduced to allow vehicles not to change lanes even when the environmental criteria are met. Without stochastic noise, the lane-changing rate vanishes in the stationary states of a homogeneous highway. With stochastic noise, vehicles change lanes frequently even when there are no slow vehicles to overtake. The lane-changing rate reflects the intrinsic fluctuations much more than the inhomogeneity of the highway. Aggressive vehicles that change lanes at every opportunity will only keep a speed slightly higher than others.