Mathematical Model and Cloud Computing of Road Network Operations under Non-Recurrent Events

Abstract

Optimal traffic control under incident-driven congestion is crucial for road safety and maintaining network performance. Over the last decade, prediction and simulation of road traffic play important roles in network operation. This dissertation focuses on development of a machine learning-based prediction model, a stochastic cell transmission model (CTM), and an optimisation model. Numerical studies were performed to evaluate the proposed models. The results indicate that proposed models are helpful for road management during road incidents.