Publication Date

8-2014

Date of Final Oral Examination (Defense)

6-20-2014

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Civil Engineering

Department

Civil Engineering

Supervisory Committee Chair

Mandar Khanal, Ph.D.

Supervisory Committee Member

Jaechoul Lee, Ph.D.

Supervisory Committee Member

Yang Lu, Ph.D.

Abstract

Deterioration of freeway traffic flow condition due to bottlenecks can be ameliorated with ramp metering. A challenge in ramp metering is that it is not possible to process data in real-time and use the output in a control algorithm. This is due to the fact that by the time processing is completed and a control measure applied, the traffic state will have changed. A solution to this problem is to forecast the traffic state and implement a control measure based on the forecast.

A dual-state Kalman filter was used to forecast traffic data at two locations on a freeway (I-84). A Kalman filter is an optimal recursive data processing algorithm; predictions are based on only the previous time-step’s prediction and all previous data do not need to be stored and reprocessed with new measurements. A coordinated feedback ramp metering control logic was implemented. The closed-loop system seeks to control the traffic density on the mainline while minimizing on-ramp queues through weighting functions.

The integration of the Kalman filter with the ramp meter control logic accomplishes the ramp meter algorithmic scheme in which is proactive to changes in freeway conditions by controlling a forecasted state. In this closed-loop framework, real-time forecasts are produced with a continuously updated prediction that minimizes errors and recursively improves with each successive measurement. MATLAB was used to model the closed-loop control system as well as modify the input output constraints to evaluate and tune controller performance.

Download

Share

COinS