Mr Kyriacos Theocharides

Unmanned Aerial Vehicles are an emerging, noninvasive, vision-based traffic sensing technology that is cost effective, rapidly deployable, and well suited for collecting traffic data over spatially dispersed urban areas. In this work, these capabilities are leveraged for regional origin-destination (OD) matrix estimation. Urban traffic dynamics are described by a regional accumulation-based model, with measurements provided by a UAV-based monitoring system. Under free-flow conditions, the static OD estimation problem is formulated as the minimization of the mismatch between measured and model predicted traffic states, while regularizing deviations from a prior OD matrix. The resulting problem is a convex quadratic program with linear equality constraints, for which an analytical solution is derived. The proposed method is evaluated on a seven-region network with known split ratios and normally distributed measurement and prior OD noise. Numerical results indicate that the proposed solution approach is effective at accurately estimating the OD matrix even in scenarios with high noise and measurement sparsity and that OD matrix estimates from UAV obtained measurements are more accurate than measurements from loop detectors of comparable sparsity.

Biography:

Kyriacos received an MEng in Mechanical Engineering from Imperial College London in 2021. He is currently a PhD Candidate at the KIOS Research and Innovation Center of Excellence, Department of Electrical and Computer Engineering, UCY. His research focuses on traffic state estimation for macroscopically modelled urban traffic networks via UAV-based sensing.

Email: