Optimization and scheduling for a large-scale urban transportation system involving human factors

• Main Author: Zhang, Yi
• Other Authors: Su Rong
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/137395
• Institution: Nanyang Technological University

Many urban areas are facing heavy traffic congestion, pollution, noises and traffic fatalities. Limited financial resources and physical space cannot support continuous infrastructures expansion, which calls for other alternative solutions to address the above problems. Traffic signal control, as one of the most essential strategies to tackle this challenge, has been applied and studied over the past several decades. However, existing traffic signals generally focus on optimality of vehicle flows, but the pedestrian delay is seldom considered. Also, many adaptive traffic signal strategies focus on increasing the traffic efficiency, which may unintentionally affect the safety of the system. According to Traffic Police accident statistics, the proportion of the pedestrian fatal accidents on signalized crosswalks is 22%. Besides the traffic light control, public transportation, another effective solution to address the urban issues due to its larger ridership and sustainability on economic efficiency, environmental protection and social equity, has also been studied for several decades. Especially for Singapore, a densely-populated small city-country with a 648 km^2 main island and a total population of 5.6 million. Clearly, developing an efficient public transport system is definitely indispensable in order to solve serious traffic issues with a limited land supply and exploding transport demands. Therefore, this thesis focuses on solving the above issues by designing adaptive traffic signal controllers for both vehicles and pedestrians and developing optimal dispatching and operating strategies for public transport systems. The thesis includes three technical chapters: The first technical chapter develops a model to provide traffic light scheduling for pedestrian-vehicle mixed-flow networks, where a macroscopic pedestrian "hopping" model at an intersection is formulated, which is capable of reflecting the changing capacity during one time interval resulted from the crosswalk length and pedestrian relatively lower speed. By considering drivers' psychological response to traffic signal states, a network-based vehicle flow model is adopted. Both pedestrian performance and vehicle performance are integrated together via the weighted sum method, which could be translated into monetary values for economic considerations or delay costs for efficiency evaluation. By connecting with the commercial simulator, Vissim, experiment studies are carried out to illustrate the usefulness of our proposed solution. The second part of the thesis investigates the potential impact of the pedestrian movement to the vehicle traffic networks when pedestrian safety is considered in the system. Considering the high-incidence rate of pedestrian violations during Flashing GREEN (FG), an additional Dynamic All RED (DAR) phase is introduced at the end of each FG period, whose duration is adaptively adjusted according to the number of non-compliant pedestrians. With computational complexity being a concern for our model, an evolutionary algorithm with repairing mechanism (EARM), is proposed to solve our problem. Thirdly, we address the bus dispatching system from the tactical and operational level, with the aim to minimize passenger delay and bus vacancy by scheduling corresponding vehicles. Meanwhile, we take the stop-skipping control strategies into account and allow the bus-platoon dispatching to guarantee an efficient bus service. Also, passengers' perceived delay is firstly formulated by a nonlinear function, which describes the dissatisfaction of the passengers who have not boarded after their expected bus arrival time. Moreover, we compare our method with the traditional fixed schedule and the optimized single bus dispatching schedule, and the simulation results illustrate the efficiency of our method. Overall, the thesis proposes three major problems faced by the researchers and policy makers in the intelligent transportation field. The suggested guidelines and methods can assist in developing a comprehensive and sustainable urban transport system in order to provide citizens a smarter, safer and more interactive transportation experience.