Date of Award

12-2018

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Industrial Engineering

Committee Member

Dr. Sandra D. Eksioglu, Committee Chair

Committee Member

Dr. Pamela Marie Murray-Tuite, Co-Chair

Committee Member

Dr. J. Cole Smith

Abstract

Emergency response vehicles (ERVs) need to reach their destinations as fast as possible. Road congestion and unpredictable movement of non-emergency vehicles (non-ERVs) makes it challenging for the ERV to move quickly. By using the autonomous/connected vehicle environment, instructions can be disseminated to the non-ERVs in the vicinity of the ERV to facilitate its passage within a link. In this thesis, an extension to a previously developed mathematical program is proposed to enable the ERV to use a contraflow lane when considerable speed gains can be potentially achieved. An experimental analysis is conducted to evaluate the sensitivity of the model’s output to traffic congestion, downstream non-ERV positions, ERV starting position, road composition, road segment length, and the length of the feasible stopping range for every non-ERV. Results showed that usage of contraflow was provided the least travel times for the ERV when it started in the left-most lane of the normal direction. Also, when the normal direction of the road was heavily congested as compared to the contraflow segment, the usage of contraflow by the ERV provided it the least travel times. In addition, a comparative study is performed to compare the proposed formulation with previously developed non-contraflow strategies as well as the currently adopted strategy requiring vehicles to move to the nearest edge. Results showed that the use of contraflow by the ERV provides improved travel times and average ERV speeds in many situations when the contraflow segment volume was sparse whereas the normal direction was congested. However, the computation times for the newly developed contraflow strategy were greater than the previously developed non-contraflow strategies. So, a heuristic was developed to reduce computational effort by cutting off the solver at a specified point, which was decided by how far the current feasible solution found was from the possible optimal solution (optimality gap). This heuristic not only provided improved computation times, but also results which did not statistically differ from the optimal results. The paths provided by the heuristic were also similar with the only difference being the points at which the lane changes happened. Hence, the utilization of this approach can potentially save lives due to reduced emergency response times.

Share

COinS