Recent studies on maritime traffic model the interplay between vessels and ports as a graph, which is often built using automatic identification system (AIS) data. However, only a few works explicitly study the evolution of such graphs and, when they do, generally consider coarse-grained time intervals. Our goal is to fill this gap by providing a conceptual framework for the fine-grained systematic study of maritime graphs evolution. To this end, this paper presents the month-by-month analysis of world-wide graphs built using a 3-years AIS dataset. The analysis focuses on the evolution of several topological graph features, as well as their stationarity and statistical correlation. Results have revealed some interesting seasonal and trending patterns that can provide insights in the world-wide maritime context and be used as building blocks toward the prediction of graphs topology.

最近关于海上交通的研究将船舶和港口之间的相互作用建模为图形，该图形通常使用自动识别系统 (AIS) 数据构建。然而，只有少数作品明确研究了此类图的演变，并且在研究时通常会考虑粗粒度的时间间隔。我们的目标是通过为海图演变的细粒度系统研究提供概念框架来填补这一空白。为此，本文介绍了使用 3 年 AIS 数据集构建的全球图表的逐月分析。分析侧重于几个拓扑图特征的演变，以及它们的平稳性和统计相关性。