This paper focuses on a new method of converting multidimensional time series into complex networks. Compared with the previous studies, this method puts different variables into the same high-dimensional system, not only can study the properties of individual variables but also can investigate the correlation between different variables and its dynamic evolution process. Taking the monthly crude oil price data of 23 regions in the United States as a sample, a complex network diagram of crude oil market prices was constructed in this paper. Through the study of network node degree, clustering coefficient and betweenness, this paper analyzes the linkage between crude oil markets. It is shown that the Illinois region is very important in the interaction of oil prices in the 23 regions of the USA. North Slope region in the network diagram is the middle junction of different node groups. This research can provide theoretical support for policy-making in the energy market.

本文着重介绍了一种将多维时间序列转换为复杂网络的新方法。与以往的研究相比，该方法将不同的变量放入同一个高维系统中，不仅可以研究单个变量的性质，而且可以研究不同变量与其动态演化过程之间的相关性。以美国23个地区的月度原油价格数据为样本，构建了原油市场价格的复杂网络图。通过研究网络节点度，聚类系数和相互之间的联系，分析了原油市场之间的联系。结果表明，伊利诺伊州地区在美国23个地区的油价互动中非常重要。网络图中的北坡区域是不同节点组的中间交汇点。该研究可以为能源市场的决策提供理论支持。