Design and evaluation of Cyber Transportation Systems
MetadataShow full item record
Cyber Transportation Systems (CTS) combine recent advance in information, communication and automation technologies in the field of road transport. CTS aim to provide advanced applications that improve road safety, traffic efficiency and sustainability of transport networks. Challenges in designing CTS exist in three aspects: the transportation system, which concerns the physical world (model traffic flow, travel demand, vehicle dynamic etc.); the cyber realm that is governed by distributed computing and communications; and human factor issues which are related to driver behavioral and decision-making. To address these challenges, we present research that has two novel and closely related thrusts. The first thrust is the design and evaluation of new applications that are made possible by the communication and data collection technologies in CTS. Communication allows vehicles and infrastructures to collaborate, and the “Big Data” makes it possible to predict the systems status. Subsequently, we could improve the control and traffic management actions. The second thrust is the design and development of an integrated-traffic-driving-networking simulator (ITDNS). Such an ITDNS is not only useful for evaluating the proposed CTS applications that require simulation models from multiple disciplines, but also is an essential tool for studying human factor issues in the design, evaluation and optimization of other advanced CTS applications. In this thesis, we study four CTS applications. First, we present a new rideshare model that allows passengers to transfer from one vehicle to another, thus improve the chance of finding a rideshare plan. Second, we apply the transfer-allowed carpooling to electric taxis, and combine rideshare scheduling with taxis’ chagrining plans. We show that such model is particularly suited for the emerging electric taxi fleets. The proposed system which could be beneficial for both passengers and taxi drivers. Third, we design and evaluate a taxi dispatch system that can efficiently provide vacant taxi with cruising guidance, so as to reduce their idle time (i.e., not serving a customer). Finally, in the fourth application, we explores the opportunity for Cooperative Vehicle and Intersection Control to contribute to a more sustainable transportation system. We propose a two-level approach that jointly decides the traffic light timing and controls vehicles’ speed, with the objective being to minimize energy consumption. In the second part of the thesis, we present our work in building a 3-in-1 integrated simulator that combines traffic, network and driving simulation models. We first describe our effort in validating the simulation framework through a field test. Then, as an example of the ITDNS’s advantage in human factor research, we study the driver’s acceptance of autonomous speed control system and obtain reference headways (which can be used for calculating route capacity in future CTS). Moreover, we link the simulator to an energy efficiency model with the eventual goal of quantifying energy related impacts of various automation technologies. In general, applications and simulation tools proposed in this thesis combine research from multiple disciplinary, and provide a holistic design for emerging applications in CTS.