The rise of automated vehicle technologies requires traffic simulation frameworks that can accurately represent realistic naturalistic road behaviours. Traditional microscopic traffic simulation tools like PTV VISSIM uses Wiedemann car following models which are not suitable for modelling automated driving. This study introduces a framework for integrating additional vehicle behaviour models into VISSIM by replacing the standard lateral and longitudinal control models with the Intelligent Driver Model (IDM) and the Minimizing Overall Braking Induced by lane Changes (MOBIL) models. The proposed method uses a hybrid integration approach with External Driver Model implementation in C++ to override VISSIM’s default longitudinal and lateral control by adding IDM based acceleration control and MOBIL lane changing decisions.To ensure a realistic representation of vehicle interactions, parameters of each model are calibrated using particle swarm optimization (PSO) based on trajectory level data. The framework is tested using freeway case studies, which include single and multiple weaving sections for on-ramp and off-ramp interactions. Simulation experiments are carried out under various traffic conditions. Performance is evaluated using time to collision (TTC), flow density relationships, and time space diagrams to compare how traffic behaves with five different scenarios of No Automated Vehicles (No AV), Default Adaptive Cruise Control (Default ACC), Longitudinal AV (without MOBIL lane changing), Mixed conditions of 50% Human Driven Vehicles (HDV) and 50% AV (both longitudinal and lateral control), and fully automated conditions (100% AV - lateral and longitudinal). Results show that the IDM and MOBIL integration captures realistic traffic behaviour like shockwave formation, localized congestion in weaving areas, and diverse vehicle interactions. However, improvements associated with AV lane-changing behaviour are not clear especially in mixed traffic situations. This approach provides better representation of existing simulation tools aimed at capturing interactions between AVs and HDVs in different base settings. With its limitations, this research still offers a scalable and simulation based framework for modelling automated vehicle behaviour and demonstrates the characteristics of utilizing external models for assessing new traffic management strategies.

Author

• Reddygari, Sai Shivani

Language

• en

Keyword

• Intelligent Driver Model
• Longitudinal Control
• Lateral Control
• PTV VISSIM
• MOBIL
• AV

Date created

• 2026

Type of Work

• Master's Thesis

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Civil & Environmental Engineering

Degree

• M.S.

Advisor

• Hamdar, Samer H

Committee Member(s)

• Silva, Pedro
• Xu, Zhengtian

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/9c67wn85c