Optimization of Road Network Operations under Uncertain Events

Abstract

The project concerns the control and management of freeway traffic. Traffic flow is formulated as an optimal control problem. Theoretical analysis is established and a smooth optimization model with ramp metering and variable speed limits is developed. A data-driven deep Koopman model is also developed to characterize the dynamical freeway system with ramp metering. Numerical examples are also conducted to investigate the performance of developed models using actual and simulated traffic data under various scenarios.