This study targets to develop a computational procedure to predict the structural response of a ship voyaging through irregular seaways taking into account the relevant uncertainties from probability perspective. To achieve the goal, ship structural response under random wave excitation was assumed to be linear one and represented by linear Volterra series, which is expanded by linear combination of Laguerre polynomials. Then the unknown Laguerre coefficients were treated as random variables, the probability of which was sought by solving Bayesian linear regression model using prepared data sets. For the validation of the proposed methodology, a single DOF linear oscillator model with artificial damping uncertainties was introduced and time series of the system response was predicted probabilistically. For more practical and realistic application, 400,000 DWT VLOC model ship experimental data was analyzed and vertical bending moment time series were probabilistically predicted using the proposed method. On top of probabilistic time series prediction of model ship, the fatigue damage was also estimated based on the stochastic time series obtained using predicted probabilistic time series data.

这项研究的目标是开发一种计算程序，以从概率的角度考虑相关的不确定性来预测通过不规则航道航行的船舶的结构响应。为了实现这一目标，假定随机波激励下的船舶结构响应为线性，并以线性Volterra级数表示，并通过拉盖尔多项式的线性组合进行扩展。然后将未知的Laguerre系数视为随机变量，通过使用准备好的数据集求解贝叶斯线性回归模型来寻找其概率。为了验证所提出的方法，引入了具有人工阻尼不确定性的单个自由度线性振荡器模型，并概率性地预测了系统响应的时间序列。为了更实际和现实的应用，分析了400,000 DWT VLOC模型船的实验数据，并使用所提出的方法对垂直弯矩时间序列进行了概率预测。除了模型船的概率时间序列预测之外，还基于使用预测的概率时间序列数据获得的随机时间序列来估计疲劳损伤。