Traffic density estimation with consideration of lane-changing.

The Kalman filtering methodology developed in the past for the density estimation of a roadway section is extended to include explicit consideration of lane-changing on a multilane freeway. It is found that the estimation error may be reduced in this fashion, the reduction ranging up to as much as 40 percent. As is to be expected, the reduction of error is greater when the roadway section is longer and hence the number of lane-changes greater. (A)