Pilot Testing of SHRP 2 Reliability Data and Analytical Products: Southern California Pilot Site

In this project, CLR tested the FreeVal tool from the SHRP2 L38 toolset using the I-5 study site, which we have a well-calibrated microsimulation model developed for another project and thus also we were very familiar with the traffic condition. 
 
FreeVal-RL is an excel based tool to evaluate the reliability based on macroscopic traffic flow model. It is the most complex tool among all reliability tool sets from SHRP2. Although it is designed to be a simple tool for end users because of the use of Excel as the model platform, we found that it was not easy to use, especially for a complicated network with multiple lanes and both mixed-flow and carpool lane facilities. With the help from the developer, we managed to devlop the model inputs as shown below. 
 
 
We found that we couldn't use the default parameters from the tool to generate reasonable results. We had to calibrate its parameter, Capacity Adjustment Factor (CAF), for each segment, which led to a lot of work, by analyzing the traffic flow fundamental diagram based on data. It meant that default capacity in the tool did not work. Even though we did it, we still couldn't get a very good match between observation and simulation. This is because FreeVal-RL only allowed one capacity value for a road segment. However, the actual capacity for a segment in the real world is not fixed due to the existence of capacity drop. 
 
 
At the end of the project, the research findings include the following:
  • FreeVal-RL has more data input requirement
  • FreeVal-RL model may need to be calibrated
  • FreeVal-RL is relatively harder to use. It is easy to end up with various error messages.
  • FreeVal-RL scenario testing results don’t match with PeMS results, even we think we have better seed file results. 
 
 
Sponsor: National Academy of Sciences (NAS) 
 
Prime Contractor: System Metrics Group (SMG)
Study Period: March 2013 - Dec 2013