The management of large-scale transportation infrastructure is becoming a very
complex task for the urban areas of this century which are covering bigger geographic
spaces and facing the inclusion of connected and self-controlled vehicles. This new
system paradigm can leverage many forms of sensing and interaction, including a
high-scale mobile sensing approach. To obtain a high penetration sensing system
on urban areas more practical and scalable platforms are needed, combined with
estimation algorithms suitable to the computational capabilities of these platforms.
The purpose of this work was to develop a transportation framework that is able
to handle different kinds of sensing data (e.g., connected vehicles, loop detectors) and
optimize the traffic state on a defined traffic network. The framework estimates the
traffic on road networks modeled by a family of Lighthill-Whitham-Richards equations.
Based on an equivalent formulation of the problem using a Hamilton-Jacobi
equation and using a semi-analytic formula, I will show that the model constraints
resulting from the Hamilton-Jacobi equation are linear, albeit with unknown integer
variables. This general framework solve exactly a variety of problems arising in
transportation networks: traffic estimation, traffic control (including robust control),
cybersecurity and sensor fault detection, or privacy analysis of users in probe-based
traffic monitoring systems. This framework is very flexible, fast, and yields exact
results.
The recent advances in sensors (GPS, inertial measurement units) and microprocessors enable the development low-cost dedicated devices for traffic sensing in cities, 5 which are highly scalable, providing a feasible solution to cover large urban areas. However, one of the main problems to address is the privacy of the users of the transportation system, the framework presented here is a viable option to guarantee the
privacy of the users by design.
Date of Award | Nov 2016 |
---|
Original language | English (US) |
---|
Awarding Institution | - Computer, Electrical and Mathematical Sciences and Engineering
|
---|
Supervisor | Christian Claudel (Supervisor) |
---|
- Traffic estimation
- Mixed integer linear programming
- Highway networks
- Optimization