Nonlinear roll damping coefficients and natural frequency estimation via augmented extended Kalman filtering for floating body

Abstract

ABSTRACT The nonlinear roll damping coefficients and natural frequency of a floating body are discussed in this article using an augmented Extended Kalman Filtering (EKF) method. This paper presents a non-linear model-based observer for the combined estimation of free roll decay motion states and damping coefficients. The identification framework uses an improved extended Kalman filter (EKF) to deal with model parameter variability and noisy measurement input. The estimated error was compared to the theoretical results obtained from the covariance matrix for filtering correctness. It is found that the nonlinear damping coefficients and natural frequency obtained from the EKF approach are a more realistic representation of the roll dynamics from the general estimation procedure given by the literature. The proposed method also has the built-in capability to remove random noise from the measured roll. The influence of noise levels in the measurements on the identification accuracy was investigated and discussed.