Abstract

Microscopic traffic simulation serves as a powerful tool for comprehensive modeling and analysis of individual vehicle behaviors and traffic flow dynamics. However, ensuring the accuracy and reliability of simulation outcomes necessitates meticulous model calibration. This calibration process involves refining the model parameters to closely align with real-world observations, thereby enhancing the simulation's ability to replicate actual traffic scenarios. Although various studies have explored methods to improve the calibration performance of car following (CF) and lane changing (LC) models, the majority have focused on developing algorithms applicable to specific traffic conditions, leaving a significant gap in a systematic framework for simultaneous CF and LC model calibration. Consequently, this thesis proposes a novel approach to calibrate CF and LC models within a microscopic traffic simulation framework, leveraging drone-based vehicle trajectories. The approach encompasses several key steps: firstly, a clustering method is introduced to classify observed trajectories into distinct CF groups, enabling optimization of CF model parameters by calibrating individual vehicle trajectories. Secondly, the LC model is calibrated using heuristic algorithms to minimize discrepancies between observed and simulated LC timing, with the calibrated CF model acting as the controller for longitudinal movement simulation. Finally, the correlation between CF and LC model parameters is examined, leading to the derivation of a joint parameter distribution. When simulating stochastic traffic flow, each generated vehicle is assigned a parameter set from this distribution. Evaluation of the calibration method on the CitySim FreewayB dataset demonstrates improved performance compared to conventional calibration methods, as evidenced by enhanced vehicle trajectory simulations. The proposed framework thus offers valuable insights into achieving accurate CF and LC model calibration.

Notes

If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu

Graduation Date

2023

Semester

Summer

Advisor

Abdel-Aty, Mohamed

Degree

Master of Science (M.S.)

College

College of Engineering and Computer Science

Department

Civil, Environmental, and Construction Engineering

Degree Program

Civil Engineering; Smart Cities Track

Identifier

CFE0009812; DP0027920

URL

https://purls.library.ucf.edu/go/DP0027920

Language

English

Release Date

August 2024

Length of Campus-only Access

1 year

Access Status

Masters Thesis (Campus-only Access)

Download

Restricted to the UCF community until August 2024; it will then be open access.

Share

COinS