Predicting Roll Angular Motion

船舶的横摇角运动（RAM）表示船舶绕其纵轴的振荡，主要由波浪运动引起。 预测RAM的能力具有重要的实际效用。 例如，在与国防有关的工作中，它在确定武器系统的准确性方面发挥作用。 我们提出了一种预测RAM的技术。 我们的方法借用参数和非参数统计，在正弦模型拟合数据，但只有在很短的时间间隔。 我们将展示如何评估和纠正错误。 特别是，我们提出了附加概率的预测精度的方法。

The roll angular motion, or RAM, of a ship denotes its oscillation about its longitudinal axis, primarily caused by wave motion. The ability to predict RAM is of significant practical utility. For example, in defence-related work it plays a role in determining accuracy of weapons systems. We suggest a technique for predicting RAM. Our method borrows from both parametric and nonparametric statistics, in that a sinusoidal model is fitted to data but only over a short time interval. We show how to both assess and correct error. In particular, we propose methods for attaching probabilities to the accuracy of predictions.