Simplex Formations of Autonomous Vehicles and Adaptive Sampling Strategies
|
Proceedings of TAROS (Towards Autonomous Robotic Systems) '07, Aberystwyth, September 2007
Knowledge of iso-contours of the underwater terrain can be used to reconstruct it using interpolation. Identifying a set of isoclines can be more efficient and less time-intensive than sweeping a large area. In this paper, we propose a system where a small number of agile underwater vehicles cooperatively maintain a triangular formation on a plane above the terrain and use field values measured by the individual robots to implement the simplex optimization algorithm which has proven very useful for nonsmooth objective functions. In this paper, we propose general rules for implementing the reflection operation and extend it to reflexive motion over isoclines (areas of constant field value).