Search

We develop a methodology to optimize the schedule coordination of a full-stop service pattern and a short-turning service pattern on a bus route. To capture the influence of bus crowding and seat availability on passengers' riding experience, we develop a Markov model to describe the seat-searching process of a passenger and an approach to estimate the transition probabilities of the Markov model. An optimization model that incorporates the Markov model is proposed to design the short-turning strategy. The proposed model minimizes the total cost, which includes operational cost, passengers' waiting time cost and passengers' in-vehicle travel time cost. Algorithm is developed to produce optimal values of the decision variables. The proposed methodology is evaluated in a case study. Compared with methodologies that ignore the effect of bus crowding, the proposed methodology could better balance bus load along the route and between two service patterns, provide passengers with better riding experience and reduce the total cost. In addition, it is shown that the optimal design of the short-turning strategy is sensitive to seat capacity. Copyright © 2016 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

Bus bridging has been widely used to connect stations affected by metro disruptions such that stranded passengers could resume their journeys. Previous studies generally assumed that a bus operates exclusively on one bridging route with given frequency, which limits the service flexibility and reduce the operational efficiency. We propose a strategy to instruct buses to operate on predefined bridging routes once they are dispatched from depots. Buses are allowed to flexibly serve different bridging routes. Each bus operates based on a bridging plan that lists the stations to serve in sequence instead of route frequencies. A two-stage model is developed to optimize the bridging plans and their assignments to buses with the objectives that balance the operational priorities between minimizing bus bridging time and reducing passenger delay. A Weight Shortest Processing Time first (WSPT) rule based heuristic algorithm is developed to solve the proposed model. The developed model is further incorporated in a rolling horizon framework to handle dynamic passenger arrivals during the disruption period. The effectiveness of the proposed strategy is demonstrated in comparison with alternative strategies in real-world case studies. [ABSTRACT FROM AUTHOR]