Study of Deck Motion Prediction Based on Adaptive Forgetting Least Squares Algorithm

This paper proposed an Adaptive Forgetting Least Squares algorithm based on the Auto-regressive model for the very short-term deck motion prediction, which aims to increase the accuracy and convergence rate of system parameter identification. The main idea of our research is adaptively adjust the forgetting factor according to the input signal. By means of adaptive forgetting, the algorithm can improve the identification and prediction accuracy by 4.3%, and improve the convergence rate by 5.5%. Finally, the results show that the algorithm is available for complex sea states, even the system signal excitation ability is weak.