Parameter Estimation of Wave-Induced Oscillatory Ship Motion for Wave Filtering in Dynamic Positioning

Abstract

By means of wave filtering a dynamic positioning system for ships can be developed which does not counteract high-frequency, wave induced oscillatory ship motion. For high-performance and energy-efficient operation, the wave filtering parameters must be adjusted according to the environmental conditions. This prevents unnecessary wear and tear of the propulsion system while still achieving a high level of control performance. If the characteristics of the wave disturbance vary during operation, an online update of the filter parameters is required in order to maintain the control performance. The filter parameter which is significantly influenced by the varied wave spectrum is the peak frequency of the wave-induced oscillatory ship motion. Hence, the online estimation of the peak frequency is the topic of this paper. The approach adopted is to determine the peak frequency based on the parameter identification of a low order auto regressive model. The practical problems that arise are addressed and modifications are proposed. The algorithms shown are tested and compared on the basis of real experiments on a model scaled ship with Voith-Schneider Propellers. The results show the improved reliability of the proposed detail solutions in terms of numerical robustness and capability for practical real-time application in continuous operation.