Real-Time Prediction of Incoming Wave Profile Surrounding Floating Offshore Wind Turbine Using Kalman Filter

Abstract The present research discusses the possibility of employing a Kalman filter algorithm to predict a real-time data of the incoming wave elevations over a floating offshore wind turbine (FOWT). This is suggested to address some problems in the existing safety management system of FOWT’s operation and maintenance, from the perspective of personnel safety. In this case, knowing immediate wave elevation from a couple cycle ahead is very beneficial to elevate the safety factor during the crew transfer process from the boat to the FOWT. This can be done when the incoming waves are able to be predicted. This time, the prediction is carried out centering around the elementary wave coefficients estimation. Hence, these coefficients are prescribed to be the state variables in the Kalman filter. Meanwhile, two types of observation data are introduced. The first one retrieved its values from numerically generated data, while the other one is data from experiment. Transfer functions that are particular over some specific measurements are used as the bridge between the state model and observation model. Filter capability to predict wave profiles from unidirectional and bidirectional spectrum is explored. Overall, prediction results obtained through numerically generated data show a good agreement with its references. Experiment-based prediction results however, exhibit less favorable results.