Application of Kalman filter to the uncertainty of model resistance data obtained from experiment

Standard deviation is the correct way to characterise the spread of the data and as the uncertainty associated with measurement the value of the standard deviation may be refined. The aim is to quantify the level of uncertainty in the resistance data of a model tanker obtained from towing tank tests. Kalman Filter (KF) was used to correct the standard deviation of the data, which is composed of the state-space model and least-squares method. Results of the simulations showed that KF could decrease the standard deviation of the resistance for a range of speeds (1,029-1.543 m/s). The standard deviation of filtered data is much smaller (1.3%-4.2%) than that of unfiltered data (14.7%-28.4%). The proposed filter method can therefore reduce the uncertainty of the model experiment.