The joint distribution of successive wave periods plays a very important role in the study of resonant effects on coastal and marine structures. Until recently, studies on the statistics of consecutive periods have almost exclusively focused on single-wave systems. This paper proposes a parametric model established from a combination of a mixture lognormal distribution and Gaussian copula to describe individual successive wave periods in combined sea states. This new model, together with two additional distributions based on the conditional modelling method and the copula function, respectively, are compared, with reference to secondary wave data collected in laboratory experiments and simulated data obtained by a six-parameter Ochi-Hubble model. Nine types of combined sea states are considered and discussed. The conditional probability of the wave period, given by the previous wave period, is used to assess the application of the adopted models to the resonance study. Conventional models are unsuitable when the patterns exhibit multimodal characteristics and involve two-wave systems whose spectral peaks are widely separated. The mixture model provides a more accurate description of the bivariate distribution and improved performance in the analysis of the resonant effects on marine structures than the other two approaches.

连续波浪周期的联合分布在研究海岸和海洋结构的共振效应中起着非常重要的作用。直到最近，有关连续周期统计的研究几乎完全集中在单波系统上。本文提出了一种由混合对数正态分布和高斯copula组合建立的参数模型，用于描述组合海况下的单个连续波浪周期。参照实验室实验中收集的二次波数据和通过六参数Ochi-Hubble模型获得的模拟数据，将该新模型与分别基于条件建模方法和copula函数的两个其他分布进行了比较。考虑并讨论了九种类型的联合海洋状态。前一个波浪周期给出的波浪周期的条件概率用于评估所采用模型在共振研究中的应用。当模式表现出多峰特性并且涉及光谱峰被广泛分离的两波系统时，常规模型是不合适的。与其他两种方法相比，混合模型提供了对双变量分布的更准确描述，并在分析对海洋结构的共振影响时提供了改进的性能。