Investigating the effect of distribution of mixed vehicle fleet

• Main Author: Tan, Jack
• Other Authors: Zhu Feng
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Civil engineering::Transportation
• Online Access: https://hdl.handle.net/10356/159017
• Institution: Nanyang Technological University

In recent years, the development of Automated Vehicles (AVs) has gained much attention from mainstream media due to significant technology breakthroughs in autopilot features in vehicles. In the near future, AVs will begin sharing current freeways with Human-driven Vehicles (HVs). With that, the introduction of AVs will certainly affect the efficiency of the traffic flow and road capacity. Hence, extensive research is needed to understand the characteristics of mixed traffic flow in order to plan and prepare the coexistence of AVs and HVs on our public roads. However, there is one aspect that has not been heavily studied which is, the effect of distribution of mixed vehicle fleet. In this project, we will attempt to investigate the effect of distribution of mixed vehicle fleet. In order to achieve this feat, a simulation framework is proposed using suitable car-following models: Cooperative Adaptive Cruise Control (CACC) and Adaptive Cruise Control (ACC) models representing Connected Automated Vehicles (CAVs) and Intelligent Driving Model (IDM) representing HVs. Python programming language is implemented in the simulation to execute the objectives of the project. A 10-vehicle traffic stream is then created to serve as a base scenario for observation and analysis. The simulation code is tweaked accordingly through trial and error to provide accurate results for the three key metrics: Average mean speed, standard deviation and amplitude. After that, these results were analyzed in different segments to make conclusions on the effect of the distribution of mixed vehicle fleet. Later, 8-vehicle traffic stream scenario is simulated to compare and reinforce the findings by the 10-vehicle traffic stream scenario. It is found that when the leading vehicle is a HV, the closer the CAV vehicle is to the leading HV vehicle, the worse off the traffic flow performance. Another key finding is that CACCs contribute to a much better traffic flow performance followed by IDMs and lastly, ACCs. This is due to the observations of alternating vehicles that have much worse traffic flow performance due to the formations of ACCs.