01663982

Component

Uncontrolled bus schedules tend to be highly unreliable; bus headways are typically susceptible to external disturbances (e.g. traffic congestion), creating gaps in the system that grow and eventually lead to bunching. Transit agencies have traditionally been adding slacks in their schedule to prevent this phenomenon from happening. Many strategies have been implemented to mitigate the effect of randomness on bus schedules, most of which, however, penalize the transit users with added dwell time or waiting time. In this paper, the authors analyze an alternative bus insertion strategy in which the agency positions standby buses to take over early and late buses so as to contain their deviations from schedule. This method does not penalize the transit users, is easy to implement, and only requires Global Positioning Systems (GPS) units and communication devices. The authors develop a discrete-time infinite-horizon model, as well as a non-myopic approximate dynamic programming approach to solve this model. It is shown through numerical examples that schedule deviations can be controlled by regularly inserting standby buses, that different insertion strategies can be implemented depending on the resources available to the agency, and that the proposed method outperforms the conventional slack-based schedule control.

This paper was sponsored by TRB committee ADB00 Section - Travel Analysis Methods.

18-05910

English

6p

2018

Transportation Research Board 97th Annual Meeting

Digital/other

Figures; References; Tables

Bunching; Bus transit; Bus transit operations; Delays; Dynamic programming; Global Positioning System; Headways; Optimization; Passengers; Scheduling; Stochastic programming; Traffic models

Planning and Forecasting; Public Transportation

Transportation Research Board Annual Meeting 2018 Paper #18-05910

TRIS, TRB, ATRI

Jan 8 2018 11:31AM