Estimation of the Capacity in Freeway Weaving Areas for Traffic Management and Operations

This research developed an online procedure to estimate the weaving capacity through time for a simple ramp-weave section, the most common type of weaving areas in the Twin Cities? freeway network. The field observations and the analysis of the traffic data collected from a sample weaving section indicate that the freeway-to-ramp and ramp-to-freeway vehicles first merge and travel together at the beginning portion of the auxiliary lane before they split to the mainline or exit ramp. The length of the shared portion of the auxiliary lane, called an "effective weaving zone," varied depending on the length of an auxiliary lane and the amount of weaving volume. The above merge-split behavior and the resulting mixed flow on the auxiliary lane for a short time period explains the fact that the maximum possible weaving volume in a simple ramp-weave section equals the maximum through volume that the auxiliary lane can handle. Researchers used a Kalman Filter to obtain estimated weaving volume data from three weaving, which supported this observation. Based on the above findings, an online procedure was developed to estimate the maximum possible weaving volume for a given ramp-weave area through time using the volume and occupancy measurements from the loop detectors. The proposed procedure assumes that the maximum possible weaving volume for a given time interval is a function of downstream traffic conditions that can be quantified by estimating the time-variant merging and diverging capacities of a given weaving section. Test results with the five-minute data from a ramp-weave site indicate that the maximum possible weaving volume can be estimated with reasonable accuracy during congested peak periods.