MPC
Research Projects (2007-08)

Identifying Number

MPC-282

Project Title

Express Lane Genetic Algorithm Microsimulation Evaluation, Part 2

University

University of Utah

Project Investigator

Peter T. Martin

Description of Project Abstract

The purpose of this project is to build, calibrate, and validate VISSIM model of I-15 from SR 201 (or 600 N) to University Parkway. UDOT has started developing a VISSIM microsimulation model for evaluation of the HOT lanes on I-15 from SR 201 to University Parkway. Microsimulation models are required tools for evaluation of HOV and HOT facilities. However, microsimulation models require much more details when building and calibrating the models. The calibration of microsimulation parameters (e.g. car-following parameters, speed and acceleration distributions) is very essential to validate simulations results with the observed performance measures. When traffic engineers build sophisticated traffic micro-simulations of their networks, the process is detailed and time consuming. Calibration of the networks is the process whereby algorithmic controls are manipulated so that the model reflects reality realistically. In the past, this process was quick and usually heuristic. Today, as the tools are increasingly complex, the process is much more demanding. This project develops a state-of-the-art approach to calibration through the application of a Genetic Algorithm which will calibrate a network rigorously.

Project Objectives

1. Identify the proper calibration methodologies considering various possible scenarios
2. Design data collection plan and conduct the data collection in the pre-selected test beds
3. Compare and evaluate simulated and measured travel variables and make recommendations

Project Approach/Methods

1. Develop detailed project scope
2. Prepare brief literature review
3. Propose research methodology (data collection, calibration, validation)
4. Integrate material and data already developed and gathered by UDOT
5. Collect data (Flows & speeds (UTL - real time connection to the TMS data), Travel time runs - GPS data , etc.)
6. Calibrate VISSIM model by using Genetic Algorithm or other optimization searching tools
7. Validate VISSIM model for an independent data set (not used in calibration)
8. Report findings to UDOT
9. Deploy Genetic Algorithm calibration tool in UDOT Planning Division.

Technology Transfer Activities

Data sharing; database development

Time Duration

July 1, 2007 – June 30, 2008

Yearly and Total Project Cost

$78,600 Total: $157,250

TRB Keywords

Traffic operations, traffic control, safety, emergency planning