METHODOLOGY

This study first tested the VSA algorithm in a microsimulation model to replicate the actual freeway environment and then implemented it in the field. A website for real-time traffic detector data is developed that includes flow, speed, and occupancy data from loop detectors in the field test site. Radar equipment and solar panel powered LED display equipment at seven different sites are installed along the corridor. Speed data from the radar equipment was aggregated with the detector data for input of the VSA algorithm. This data was used as the input for the VSA algorithm to calculate VSA and maximize  the throughput from the most downstream bottleneck to the upstream. Finally, the recommended speeds were displayed on the VSA signs.

To test the performance of VSA, scenarios with and without VSA were analyzed using the Performance Measurement System (PeMS) hourly data. Data collection for the “VSA-ON” scenario took place from April 9 – May 4. For the “VSA-OFF” scenario, data collection took place on March 12 – 16 and May 7 – 11.Performance measures included  Vehicle Miles Traveled, Vehicle Hours Traveled, and average speed.

FINDINGS

The average of three performance measures were reported for the AM (6-9 AM) and PM peak hours (2-7 PM) over the four weeks.

• The Vehicle Miles Traveled for the AM peak hour increased by 2.72 percent and decreased by 0.096 in the PM peak hour.
• Vehicle Hours Traveled  decreased for both AM and PM, by 6.28 and 1.47 percent respectively.
• The average speed  increased by 8.71 and 2.8 percent for both AM and PM, respectively.
• Driver compliance gradually improved as the test progressed and the increase in driver compliance was generally in line with an improvement in system performance.

The results of the performance analysis illustrated an improvement in all three performance measures for the AM peak hours. However, the PM performance measures improved only for two of the measures, Vehicle Hours Traveled and average speed.