Use this URL to cite or link to this record in EThOS:
Title: The application of vehicle classification, vehicle-to-infrastructure communication and a car-following model to single intersection traffic signal control
Author: Dodsworth, Joel Andrew
ISNI:       0000 0004 7654 9948
Awarding Body: University of Leeds
Current Institution: University of Leeds
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
On-line responsive traffic signal optimization strategies most commonly use data received from loop detectors to feed information into an underlying traffic model. The limited data available from conventional detection systems has dictated the way that current 'state-of-the-art' traffic signal control systems have been developed. Such systems tend to consider traffic as having homogenous properties to avoid the requirement for more detailed knowledge of individual vehicle properties. However, a consequence of this simplification is to limit an optimizer in achieving its objectives. The first element of this study investigates whether additional data regarding vehicle type can be reliably extracted from conventional detection to improve optimizer performance using existing infrastructure. A single detector classification algorithm is developed and it is shown that, using a modification of an existing state-of-the-art optimization method, a modest improvement in performance can be achieved. The emergence of connected vehicle technology and, in particular, Vehicle-to-Infrastructure (V2I) communications promises more comprehensive data. V2I-based optimization methods proposed in literature require a minimum penetration rate of V2I equipped vehicles before performance matches existing systems. To address this problem, the second part of the study focuses on the development of a hybrid detection model that is capable of simultaneously using information from conventional and V2I detection. It is demonstrated that the hybrid detection model can begin to realise benefits as soon as V2I data becomes available. V2I-based vehicle classification is then applied to the developed hybrid model and significant benefits are demonstrated for HGVs. The final section of the thesis introduces the use of a more sophisticated internal traffic model and a new optimization method is developed to implement it. The car-following model based optimization method addresses the lack of modelled interaction between vehicles and is shown to be capable of reducing vehicle stops over and above the developed (vertical queue based) hybrid model.
Supervisor: Shepherd, Simon P. ; Liu, Ronghui Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available