Search

Browse

• Communities
• Authors
• Year
• Titles
• Subjects

My Space

• Login

Real-time traffic control using queue estimation from connected vehicle data

Alexaki Sofia-Anna

Full record

---

Summary

Traffic congestion is a critical societal problem that causes an increase in road traffic accidents and environmental pollution as well as waste of time. The aim of this work is to investigate whether the adoption of the new tendency of Internet of things could improve real-time traffic control with the involvement of connected vehicles in the role of the unique detector of the traffic conditions, instead of spot sensors ( e.g loop detectors, radars, video sensors etc.). The first key point of this thesis was the implementation of an algorithm which uses a mathematical approach that achieves realtime queue estimation (in vehicles) in every link of an urban road network, based on connected vehicles measurements. Then, the algorithm was integrated into two real-time traffic control strategies, the Max-Pressure algorithm (decentralized control) and the TUC strategy (centralized control) to investigate whether the information from connected vehicles can be considered reliable; and to what extent the percentage of connected vehicles influences this reliability. In order to draw conclusions, the AIMSUN microscopic simulator was used for the urban network of Chania. Statistical analysis of simulation investigations results showed that the estimation approach leads to reliable queue estimation for all penetration rates tested. Furthermore, as far as it concerns control strategies performance after the integration of connected vehicle approach; it turned out that TUC strategy could work efficiently by taking information from connected vehicles while Max-Pressure algorithm which uses second by second measurements to take second by second decisions cannot work properly for low penetration rates of connected vehicles because of the limited real time information.

Available Files

Services