Using a Model of Ocean Currents to Control the Position of Vertically Profiling Marine Floats. Troesch, M.; Chien, S.; Chao, Y.; and Farrara, J. In Workshop on Planning and Robotics, International Conference on Automated Planning and Scheduling (PlanRob, ICAPS 2016), London, UK, June, 2016.
Using a Model of Ocean Currents to Control the Position of Vertically Profiling Marine Floats [pdf]Paper  abstract   bibtex   
We describe a methodology for control of vertically profiling floats that uses an imperfect predictive model of ocean currents. In this approach, the floats have control only over their depth. We combine this control authority with an imperfect model of ocean currents to force the floats to maintain position. First, we study the impact of model accuracy on this ability to station keep (e.g. maintain X-Y position) using simulated planning and nature models. In this study, we examine the impact of batch versus continuous planning. In batch planning the float depth plan is derived for an extended period of time and then executed open loop. In continuous planning the depth plan is updated with the actual position and the remainder of the plan re-planned based on the new information. In these simulation results, we show that (a) active control can significantly improve station keeping with even an imperfect predictive model and (b) continuous planning can mitigate the impact of model inaccuracy. Second, we study the effect of using heuristic path completion estimators in search. In general, using a more conservative estimator increases search quality but commensurately increases the amount of search and therefore computation time. Third, we discuss results from an April 2015 deployment int he Pacific Ocean and compare model accuracy and float control performance.
@inproceedings{troesch-chien-chao-et-al-ICAPS-2016-june,
  author = {M. Troesch and S. Chien and Y. Chao and J. Farrara},
  title = {Using a Model of Ocean Currents to Control the Position of Vertically Profiling Marine Floats},
  booktitle = {Workshop on Planning and Robotics, International Conference on Automated Planning and Scheduling (PlanRob, ICAPS 2016)},
  year = {2016},
  month = {June},
  address = {London, UK},
  abstract = { We describe a methodology for control of vertically profiling floats that uses an imperfect predictive model of ocean currents. In this approach, the floats have control only over their depth. We combine this control authority with an imperfect model of ocean currents to force the floats to maintain position. First, we study the impact of model accuracy on this ability to station keep (e.g. maintain  X-Y  position)  using  simulated  planning  and nature models. In this study, we examine the impact of batch versus continuous planning. In batch planning the float depth plan is derived for an extended period of time and then executed open loop. In continuous planning the depth plan is updated with the actual position and the remainder of the plan re-planned based on the new information. In these simulation results, we show that (a) active control can significantly improve station keeping with even an imperfect predictive model and (b) continuous planning can mitigate the impact of model inaccuracy. Second, we study the effect of using heuristic path completion estimators in search. In general, using a more conservative estimator increases search quality but commensurately increases the amount of search and therefore  computation  time.  Third,  we  discuss  results from an April 2015 deployment int he Pacific Ocean and compare model accuracy and float control performance. },
  url = {https://ai.jpl.nasa.gov/public/papers/troesch_icaps2016_model.pdf},
  clearance = {CL#16-2240},
  project = {apf},
}
Downloads: 0