A sensitivity quantification approach to significance analysis of thrusters in dynamic positioning operations.

Abstract

The safety of offshore operations is highly dependent on the dynamic positioning (DP) capability of a vessel. Meanwhile, DP capability comes down to the ability of the thrust generated by thrusters to counteract environmental forces. Therefore, it is significant to investigate which thrusters are important to the position-keeping ability of vessels. However, complex environmental factors make the investigation of thrusters' importance more complicated. Hence, this paper proposes a new method to identify the influence of each thruster on vessel's station-keeping capability in different sea states. The station-keeping capability is quantified by a defined synthesized positioning ability criterion comprised by vessel position, heading angle, and consumed power. Through the comparison of different machine learning approaches, support vector machine (SVM) is used for building a surrogate model between DP capability and thrusters. In order to determine the most sensitive thruster in the whole process of vessel operation, an improved sensitivity analysis (SA) called ‘PAWN’ is employed along with statistical analysis to evaluate the significance of thrusters from different perspectives. Seventeen cases are investigated with respect to different thruster failures in various sea states. The results show the proposed method is able to identify the significance of each thruster in different scenarios.