An automatic decision support system has been tested during the Recognised Environmental Picture 2010 experiment to support glider operations. METOC data (winds, currents and waves) and vessel traffic information were fused by using a hybrid Fuzzy/Bayesian approach in order to assess the impact of the environment on operations and decide a course of action to safely operate. The impact of the input data uncertainty was assessed by using METOC probabilistic forecasts (ensemble and super-ensemble models) and a methodology based on the Unscented Transform (UT) to propagate the data uncertainty through the decision making system. The Fuzzy/Bayesian system was integrated with a model based decision support system making use of a glider kinematic model and the Unscented Kalman filter to predict the glider path considering the sea current field in which it operates. Three dimensional probabilistic forecasts of the sea current speed were used to predict glider position mean and covariance matrix and calculate risk indexes helping the responsible of the mission in minimising operation risks. Results of the investigation will be reported and discussed.
Decision Support during REP'10
COCOCCIONI, MARCO;
2010-01-01
Abstract
An automatic decision support system has been tested during the Recognised Environmental Picture 2010 experiment to support glider operations. METOC data (winds, currents and waves) and vessel traffic information were fused by using a hybrid Fuzzy/Bayesian approach in order to assess the impact of the environment on operations and decide a course of action to safely operate. The impact of the input data uncertainty was assessed by using METOC probabilistic forecasts (ensemble and super-ensemble models) and a methodology based on the Unscented Transform (UT) to propagate the data uncertainty through the decision making system. The Fuzzy/Bayesian system was integrated with a model based decision support system making use of a glider kinematic model and the Unscented Kalman filter to predict the glider path considering the sea current field in which it operates. Three dimensional probabilistic forecasts of the sea current speed were used to predict glider position mean and covariance matrix and calculate risk indexes helping the responsible of the mission in minimising operation risks. Results of the investigation will be reported and discussed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
