Modelling of Dynamically Positioned vessels and managed ice-field interactions using multiple regression techniques

Empirical models based on full-scale or model scale measurements are often preferred over high-fidelity numerical models when real-time predictions for ice actions on ships and floating offshore platforms are needed. This paper presents multiple regression-based empirical models for predicting a Dynamically Position (DP) controlled vessel and managed ice-field interactions developed using a large dataset from ice basin tests of two distinct DP-controlled vessels. The authors carried out a detailed analysis of measurements from two physical modelling programs for developing a unique database. A multiple regression modelling procedure was developed to predict the time-averaged and average peaks of forces and moment of the DP-controlled vessels due to ice actions in multiple configurations of managed ice-field. The models were validated using model scale measurements for the predictions of time-averaged and average peak responses. The models were then used to evaluate the effects of each ice-field parameter on vessel forces and moment and interaction event parameters. The models presented in this paper showed reasonable accuracy in predicting the effects of several ice-field parameters on the forces and moments of two distinct DP-controlled vessels in a range of realistic operating conditions. •Presents multiple regression-based empirical models for predicting the forces on DP-controlled vessels due to ice actions.•The models are developed using a large dataset from two extensive ice basin tests of two distinct DP-controlled vessels.•The models predict the effect of several managed ice-field parameters on the global forces and moments of the DP vessels.•The models accurately predicted the parameters' effects on the DP vessels operating in realistic head-on and oblique ice.