Improving Mobility and Safety in Traditional and Intelligent Transportation Systems Using Computational and Mathematical Modeling

In traditional transportation systems, park-and-ride (P&R) facilities have been introduced to mitigate the congestion problems and improve mobility. This study in the second chapter, develops a framework that integrates a demand model and an optimization model to study the optimal placement of P&R facilities. The results suggest that the optimal placement of P&R facilities has the potential to improve network performance, and reduce emission and vehicle kilometer traveled. In intelligent transportation systems, autonomous vehicles are expected to bring smart mobility to transportation systems, reduce traffic congestion, and improve safety of drivers and passengers by eliminating human errors. The safe operation of these vehicles highly depends on the data they receive from their external and on-board sensors. Autonomous vehicles like other cyber-physical systems are subject to cyberattacks and may be affected by faulty sensors. The consequent anomalous data can risk the safe operation of autonomous vehicles and may even lead to fatal crashes. Hence, in the third chapter, we develop an unsupervised/semi-supervised machine learning approach to address this gap. Specifically, this approach incorporates an additional autoencoder module into a generative adversarial network, which enables effective learning of the distribution of non-anomalous data. We term our approach GAN-enabled autoencoder for anomaly detection (GAAD). We evaluate the proposed approach using the Lyft Level 5 dataset and demonstrate its superior performance compared to state-of-the-art benchmarks. The prediction of a safe collision-free trajectory is probably the most important factor preventing the full adoption of autonomous vehicles in a public road. Despite recent advancements in motion prediction utilizing machine learning approaches for autonomous driving, the field is still in its early stages and necessitates further development of more effective methods to accurately estimate the future states of surrounding agents. Hence, in the fourth chapter, we introduce a novel deep learning approach for detecting the future trajectory of surrounding vehicles using a high-resolution semantic map and aerial imagery. Our proposed approach leverages integrated spatial and temporal learning to predict future motion. We assess the efficacy of our proposed approach on the Lyft Level 5 prediction dataset and achieve a comparable performance on various motion prediction metrics.