A virtual-environment-based testbed offers an alternative to difficult
, costly, and possibly hazardous real-time testing and evaluation of c
ontrol algorithms for autonomous underwater vehicles (AUVs). Various m
ethods are used in the testbed to model the vehicle and its environmen
t. Pertinent control algorithms are integrated with the testbed throug
h sensor and control interfaces. A review of selected virtual environm
ent testbeds and related standards is provided. Some implementation is
sues and results are discussed. A path-planning algorithm is described
to illustrate this approach. (C) 1998 Published by Elsevier Science L
td. All rights reserved.