FORECASTING TIME SERIES DATA IN THE PROBLEM OF MODELING INTERACTION OF ICE AND A VESSEL MOVING IN ICE CONDITIONS

During designing mathematical models of interaction system of a vessel and equipment towed by it, carried out on the basis of vibration measurements of the model of a vessel-tow wire-equipment connected system (scale 1: 100), towed in the ice basin, an insufficiency in duration of the model's endurance time in the ice basin, in order to correctly assess the results, was revealed. To eliminate this contradiction, a virtual extension of the model's path length was applied by predicting the time series of travel of the equipment and the vessel in the statsmodels module of the Python programming language. Designed models and the forecasts issued by them quite well convey the nature of the oscillations of the system. The results obtained with their help are applicable for vibration analysis of a vessel with towed equipment in ice. The introduction of a method for predicting the time series of ship models and towed equipment travel significantly expands the capabilities of ice modeling technologies for testing ship models, which, in turn, simplifies the task of designing methods for conducting under-ice seismic-acoustic exploration using an ice-class vessels and developing special technical means for it implementation.