STOCHASTIC MODELLING OF PASSENGER AND BAGGAGE FLOWS THROUGH AN AIRPORT TERMINAL

The Q Solve set of stochastic queueing models has been developed to describe air passenger behaviour for design purposes in both the enplaning and deplaning processes at medium-hub airports. The sequential set of models has been tested at Manchester International Airport and has been found adequate for macroscopic planning purposes at the normal level of required accuracy. The model sequence is adequate for overall design and for setting space requirements for processing, holding, waiting and concessionary areas. It is felt that sequential queueing models can offer some advantages over simulation models in the overall design process, in that computer time can be considerably reduced in gaining estimates of such variables as average flows, processing times and the variations about these average values necessary for design solutions responsive to demand variations. For design of individual facilities, in some cases the sequential queueing models provide difficulties in obtaining adequate fit to observed data and non-sequential models using Erlangian distributions are found to give more accurate representations.