A modified Box-Cox transformed Rayleigh (MBR) distribution for the wave crests is presented in this paper. The basic idea contains two steps: firstly, non-Gaussian elevations are transformed into Gaussian elevations by a modified Box-Cox transformation. Afterwards the crests of the transformed elevations are described by the Rayleigh distribution. Comparison between the MBR and Tayfun distributions shows that these two models are the same if terms smaller than O(ε²) are ignored, where ε is the elevations’ skewness. Finally, verifications and comparisons are presented for two sets of field measurement. Based on the results, the MBR distribution gives better results than the Forristall model for very large wave crests. Moreover, the MBR model matches the observations better than the Tayfun distribution since it obtains larger probabilities for larger crest in addition to smaller probabilities for small crests.

本文提出了一种改进的Box-Cox变换波峰瑞利（MBR）分布。基本思想包括两个步骤：首先，通过修改后的Box-Cox变换将非高斯标高转换为高斯标高。然后，通过瑞利分布描述转换后的高程的波峰。MBR和的Tayfun分布节目之间的比较，这两种模式是相同的，如果术语小于Ô（ε ²）将被忽略，其中ε是高程的偏度。最后，对两组现场测量结果进行了验证和比较。根据结果​​，对于非常大的波峰，MBR分布比Forristall模型具有更好的结果。而且，MBR模型比Tayfun分布更好地匹配了观测值，因为它除了获得较小波峰的较小概率外，还获得了较大波峰的较大概率。