Search

Transit priority based in exclusive right-of-way is a low-cost way of improving transit service by minimizing delays caused by interaction with other vehicles. This effect can increase the share of public transit against private cars in the mode preferences of commuters and consequently alleviate heavy congestion resulting from the dominance of single-occupancy cars. In this work we propose a modeling and optimization framework for the problem of Dedicated Bus Lanes (DBL) location selection in bi-modal urban networks with time-varying traffic congestion. Traffic flow is replicated by dynamic macroscopic traffic model with queueing characteristics instead of steady state models, to capture potential spill-back effects caused by poor DBL planning. A combinatorial optimization problem is formulated aiming at minimizing total passenger delay. Optimization is performed by Local Search (LS) and Neighborhood Search (NS) algorithms, while a network decomposition technique is proposed for improved computational cost. The results show the proposed algorithms effective in significantly improving an initial DBL plan with reasonable computational cost.

Urban mobility is facing a paradigm shift towards providing more convenient, environmentally friendly and on-demand services. Satisfying customer needs in a cost-efficient way has been the goal of many ride-sharing systems. Taxi ride sharing is considered nowadays an effective service for reducing traffic congestion and pollution; however, the operational strategies that can be used to optimize on-demand ride sharing have not been well investigated. Moreover, only a few studies in the literature provide reliable insights about capacitated ride-sharing systems. In the current work, we focus on solving the on-demand ride sharing service in a real-time frame-work, considering different optimization techniques. Furthermore, by investigating different decision variables and cost functions, we evaluate various management strategies. Furthermore,we study the sensitivity of the solutions to different ride sharing capacities. In this framework, we develop an event-based simulation engine that can be used in order to propose a real-time taxi ride sharing search algorithm. The aim of the algorithm is to quickly decide between competing taxi candidates that satisfy both the user inquiries and the problem constraints. This simulation engine can provide valuable insights regarding different cost functions and parameters variations. Moreover, by utilizing millions of real trip data from the New York City taxi database, we evaluate the feasibility of the proposed framework and evaluate the results for different strategies and optimization techniques.

Road space distribution among multiple modes of transport in urban networks has attracted the interest of several researchers and policy planners. Traditionally, these problems are solved by ignoring operational and dynamic characteristics of multi-modal congestion due to the computational burden involved. Revealing the relation between road space share and global performance of the network can lead to more efficient transport system design. The introduction of dedicated bus lanes in some parts of the network has been proposed as a measure to allow high occupancy vehicles to travel through regions with high traffic load without long delays. In this way, the bus transit system operates more efficiently and is competitive to the private car option in terms of travel time. This effect can stimulate a significant mode shift from car to bus and help alleviate congestion by reducing the number of low occupancy vehicles. In this work we address the problem of optimal allocation of exclusive bus lanes in a multi-modal urban network of fixed total road infrastructure and passenger demand, with the aim of minimizing the total passenger hours travelled (PHT) for all travel modes, within a framework which is consistent with the dynamics of congestion. A queueing theory based traffic flow model with proper treatment of spillbacks and traffic signal settings, known as the Store-and-Forward (SaF) model, is extended to simulate the dynamics of congestion inside the network by keeping track of the queues inside all links over time. The SaF simulation model can be used to evaluate the performance of different allocation policies.

Increasing urbanization and car ownership rates in developed countries result in hyper-congested cities with long delays in everyday commutes. Since scarcity of space forbids road space expansion, increasing the effectiveness and attractiveness of mass transit seems as a viable way to increase the systems supply. Dedicated Bus Lanes (DBL) provide exclusive space for public transit vehicles and therefore reduce mean travel time as also the congestion-related travel time variance; however, they decrease road capacity for regular traffic that may lead to severe congestion propagation due to the spillback effect. In this work, Variable Neighborhood Search is applied to seek for the optimal DBL plan in terms of total passenger hours travelled. A queueing theory based traffic flow model with proper treatment of spill-backs and traffic signal settings called "Store-and-Forward" is extended to simulate the evolution of traffic congestion in the network. The evaluation of the resulting optimal solution in comparison to state-of-practice solutions through microsimulation can reveal the degree of improvement in terms of decreasing total passenger delay.

Collective transport has been seen for long as a proper solution to fight congestion. While collectivity has been successful with respect to subways, it is not the norm in road traffic. Dedicated Bus Lanes (DBL) have been proposed as a measure to reduce the impact of traffic congestion in the travel time and accuracy of buses, by providing them with exclusive road space. However, DBL presence decreases road capacity, which may induce local congestion. Balancing this inherent trade-off of a DBL network, by carefully selecting the location of DBLs while considering the dynamics of congestion propagation is challenging. This work aims at finding a reliable modeling and optimization methodology to address this problem. An adjusted version of Store-and-Forward queuing model and microscopic simulation are used to simulate the traffic dynamics in presence of DBLs and assess the global network performance, while a local search algorithm is used to improve some state-of-practice solutions.