Search

This paper presents an adaptive optimization framework for dynamic rail transit network operations with use of a rollout surrogate-approximate dynamic programming method. The optimization algorithm derives coordinated decisions of service schedules and train unit deployment with respect to prevailing passenger demand. Considering the computational effectiveness needed for real-time applications, a state-dependent surrogate function is incorporated to approximate the costs associated with operational decisions over future stages. The surrogate approximation is updated iteratively via a temporal difference learning process with feeding of observations made from the transit network. The proposed framework is implemented and tested on a real-world scenario in Hong Kong Light Rail Transit (LRT) network. The results reveal that the proposed framework is able to reduce significantly the total passengers' waiting times over existing plans with reasonable computational time via use of the surrogate approximation. This suggests the potential of the proposed optimizer for real time applications in large-scale rail transit networks. • A dynamic transit network optimization model for flexible train unit and schedule assignment. • A novel surrogate approximate dynamic programming with a cross-entropy solution framework. • A novel real-time solution framework with use of cross-entropy method incorporating state-dependent constraints. • A case study demonstrating the efficiency of the proposed framework for passengers in real-time. [ABSTRACT FROM AUTHOR]

This paper presents a Pareto optimization framework for routing and scheduling dynamic rail transit network services. We first develop an event-based dynamic transit network model that can capture the evolution of passenger demand and service operations with incorporation of passengers' transfers and recirculation of limited number of train vehicles over different service lines. A multi-objective optimizer is then built upon the transit model which seeks jointly the settings of service lines and schedules that could minimize passengers' journey times, transfer rates, and operator's cost. The problem is solved by a cross-entropy method (CEM) which samples potential solutions from statistically tractable distribution models with iterative updates via maximum likelihood. The operational constraints are explicitly incorporated in the solution process which enhances the feasibility of the sampled solutions and hence effectiveness of the computational process compared with other metaheuristics used in the literature. A CEM-based ranking algorithm is further developed for deriving the Pareto-frontiers for the multi-objective transit network routing and scheduling. The proposed framework is applied and tested on the Hong Kong Light Rail Transit (LRT) network using real world scenario data. The results reveal new insights on how the existing operational settings could be improved. The present study contributes to urban transit network service planning with advanced computational techniques. • A dynamic transit network model with passengers' transfer and vehicles' circulation. • A novel cross-entropy solution framework incorporating operational constraints. • A novel solution framework that can derive generic Pareto frontiers with multiple conflicting objectives. • A case study demonstrating the efficiency and practicality from both passengers' and operators' perspectives. [ABSTRACT FROM AUTHOR]