Evaluation of the Operational Effects of Autonomous and Connected Vehicles through Microsimulation

Abstract

Proper evaluation of traffic operations integrating connected and autonomous vehicles (CAVs) requires accurate representation of these emerging technologies in microscopic simulation. This paper evaluates the ability of microscopic simulator PTV-VISSIM (Version 10.0) to simulate CAVs, and presents a comprehensive CAV model extension. In addition, emissions modeling is integrated with VISSIM to calculate real-time energy and emission estimates. The evaluation of VISSIM revealed that its internal CAV modeling has several limitations, such as modeling connectivity and complex vehicle behavior. For external modeling, there are two available VISSIM interfaces. The Component Object Model (COM) Application Programming Interface (API) is the superior approach for fetching data and modeling connectivity, whereas the External Driver Model (EDM) is a better tool for lateral and longitudinal control. The simulation extension developed leveraged both interfaces. An isolated signalized intersection was simulated to demonstrate the impact of connected vehicle (CV), autonomous vehicle (AV), and CAV traffic on speed, delay, and travel time. In addition, trajectory data, combined with the Motor Vehicle Emission Simulator (MOVES) method, were utilized to obtain energy, fuel consumption, and greenhouse gas emissions. The results show that CAVs result in net improvement in travel time and speed. However, emissions did not follow the same trend. While increasing AV penetration rates resulted in emissions reductions, increasing CV and CAV penetration rates resulted in higher emissions. While the CV logic chosen for testing seeks to maximize the likelihood of vehicle arrival-on-green, the algorithm likely results in increased variations in second-by-second accelerations, leading to overall higher emissions.

Impacts of Different Types of Automated Vehicles on Traffic Flow Characteristics and Emissions: A Microscopic Traffic Simulation of Different Freeway Segments

“…Rafael et al report that introduction of AVs might increase NO x and CO 2 emissions [17]. This finding is further confirmed by Manjunatha et al [18]. Their results show that increasing AAV penetration rates results in emissions reductions, while increasing CAV penetration rates results in higher emissions.…”

“…The first stream (e.g., see [31][32][33][34][35]), following the internal modeling approach, adjusts the parameters of the VISSIM's default car-following (i.e., Wiedemann 99 algorithm) and lane-changing (i.e., a rule-based algorithm) algorithms to mimic the anticipated driving behaviors of interest. The second stream, following the external modeling approach, simulates driving behavior of AVs/CAVs through external VISSIM interfaces (e.g., Component Object Model Application Programming Interface (COM API), and External Driver Model (EDM) [36]) and user defined algorithm and code development. The COM API enables changes in vehicle movements and driving behaviors and can be developed in several programming languages (e.g., C# [37], Python [38]).…”

Insights into Simulated Smart Mobility on Roundabouts: Achievements, Lessons Learned, and Steps Ahead

“…As level 4 and 5 CAVs are not yet deployed on actual road infras-tructures, adjustment factors for CAVs are derived from microsimulation, assuming reliable operation of all communication elements [5,22]. In light of emerging challenges in the automotive market and cooperative driving technologies, this holds particular relevance for the conceptualization of road infrastructures and the evaluation of their performance efficiency [23][24][25].…”

A Microsimulation Modelling Approach to Quantify Environmental Footprint of Autonomous Buses

“…Some studies have examined the environmental footprint of AVs through microsimulation. For example, Manjunatha et al [43] integrated VISSIM with EPA's MOVES model to estimate vehicle energy consumption and exhaust emissions. They used speed and acceleration outputs from an hour long VISSIM model combined with MOVES emission factors to calculate impact of different AV penetration rates in car traffic on a small section.…”