Optimum shuttle scheduling — a GPSS simulation

Abstract

This paper studies the operating policies of a two- station shuttle system consisting of a fleet of finite size and vehicles of finite capacity which transport passengers between two terminals (each terminal having dispatching control). Passengers arrive at either station according to independent negative exponential processes, and dispatching deci sions are made on the basis of fixed-time intervals or queue lengths (greater than zero). The objective is minimization of a total opportunity cost function defined by passenger waiting time, nonutilization of vehicles, and unused passenger seats. The travel time is probabilistic with known distribution and mean. This paper shows by GPSS simulation techniques that optimal values for any or all of the three variable parameters (fleet size, vehicle capacity, dispatching interval) can be empirically determined through sensitivity analysis. The simulation model constructed is flexible and can be adapted to a wide variety of other problems of the shuttle-system genre.