The validation of fast-time air traffic simulations in practice
- PDF / 721,379 Bytes
- 8 Pages / 595 x 842 pts (A4) Page_size
- 0 Downloads / 206 Views
#2000 Operational Research Society Ltd. All rights reserved. 0160-5682/00 $15.00 www.stockton-press.co.uk/jors
The validation of fast-time air traf®c simulations in practice MR Phillips1 and DT Marsh*2 1
Logica PLC and 2National Air Traf®c Services Limited, Christchurch
Computer simulation is an important tool in the development of air traf®c control procedures and systems, ranging from fast-time simulations (requiring no manual intervention) through a variety of real-time, controller-in-the-loop simulations of increasing size and ®delity. This paper presents a case study of an airspace re-design project and uses this to illustrate a number of issues about the validation of fast-time simulations. The two principal observations are: (1) that there are speci®c circumstances when `seeing is believing', that is, when it is appropriate to use visualisation of simulations as part of validation; and (2) that the distinction between validation activities and generating project results is not as sharp as descriptions in the literature typically show. Keywords: simulation; validation; air transport
Introduction Computer simulations are an important tool in the development of air traf®c control (ATC) procedures and systems. They range from fast-time simulations (those not requiring manual intervention during a simulation run) through a variety of real-time, controller-in-the-loop simulations of increasing size and ®delity, some involving dozens of controllers and other staff. The simulations are used to assess new airspace designs, new control concepts and new tools for controllers. This paper describes a fast-time simulation study of a new structure for part of the UK's airspace and uses this to illustrate a number of issues about the validation of fasttime simulations. Validation of a model is `substantiating that the model, within its domain of applicability, behaves with satisfactory accuracy consistent with the study objectives'.1 Therefore validation may be thought of as asking whether the model is the correct one for the task in hand. This is in contrast to `veri®cation', which is testing whether the model has been built according to the speci®cation. Some of the validation tools described here can be used in veri®cation, but that will not discussed here. It is not news that validation is essential to a successful simulation project, in ATC2 or anywhere else.3 Balci1 and Robinson and Pidd4 present more up-to-date accounts of the literature on validation techniques and reasons for successful simulations respectively. There are also a
*Correspondence: Dr D. Marsh, Air Traf®c Management Development Centre, National Air Traf®c Services Limited, Bournemouth Airport, Christchurch, Dorset BH23 6DF UK. E-mail: [email protected]
number of tutorials and other papers on validation from recent Winter Simulation Conferences, such as Balci,5 Sargent6 and Robinson.7 Indeed, the approach adopted for this study is succinctly summarised by Robinson7: `It is also useful to demonstrate the model, formally and informally, to thos
Data Loading...
