Autonomous vehicles: Evaluation of traffic management strategies in the case of an incident

Abstract

Since the invention of the first automobile, many inventions and innovations relevant to the automobiles and automobile technologies have been made in the automotive sector. As the traffic congestion becomes a worldwide problem, many studies have been conducted to determine and improve the issues. Autonomous vehicle technology is a breakthrough idea of 2000’s. Since the technology was invented, investments and developments on this subject are made continuously. Autonomous vehicles are expected to change the traffic models completely in the near future. In this thesis, it is aimed to determine the benefits and the disadvantages of autonomous vehicles where they are in a network under traffic management strategies in the case of an incident, which causes a lane to be closed for a given period. PARAMICS, the program used for the simulation model, is a microscopic traffic simulator that enables to input different calibration properties for vehicles and routes. A simulation model was specifically created for this research where Lane Control Signals (LCSs) at 2 different locations are present. It evaluates the average travel times and average speeds of the vehicles in the case of an incident for 90 different scenarios while also introducing autonomous vehicles into the system. In the simulation model, the best improvements are observed when the traffic demand is 3000 vehicle/hour for three lanes of the network. As the autonomous vehicle ratio changes from %0 to %100 on the network, without a LCS implementation, which is the best scenario, autonomous vehicles improve the average travel time by %20.26 and the average speed by %25.38. In the best scenario, where human-driven vehicles are evaluated only, LCS improves the average travel time by %16.05 and the average speed by %19.08. When LCS and autonomous vehicles are simultaneously introduced to the network, in the best scenario, which is also the best scenario among all the scenarios of the simulation model, the average travel time reduces by %21.33 and the average speed increases by %27.08.