Temporal dynamics of ice loads, which are measured with high sampling rate in experiments in ice tank, usually show strong autocorrelation. At the moment it is lack of study on autocorrelation function (ACF) of ice loads time series. In this paper the stochastic processes approach will be applied to analyze autocorrelation of ice loads.
Stochastic models for ice loads, developed earlier by the author, allow determining the distribution law and stationarity (in wide sense) for some of such time series in statistically confident manner. That allows conducting further study of those time series ACF.
For analysis and correct sampling rate choice it is important to know the time interval, which separates two statistically independent data points in time series. The algorithm for finding of such interval for time series with normal and lognormal distribution was developed in the paper. That algorithm was applied to find independence distance for global loads records, obtained in experiments with cylindrical models in ice tank of Krylov State Research Centre (St. Petersburg). The independence distance for those time series occurred to be 0,07–0,35 sec. That distance had increased with increasing of indenter diameter. Obtained results are discussed.