The purpose of this study is to classify and analyze the traffic performance characteristics of urban trunk roads. It can help travelers effectively avoid peak hours and crowded routes, and traffic managers formulate some refined management and control on roads. This paper extracts the operating characteristics of the trunk roads on macro level by analyzing the Traffic Performance Index (TPI) data. The trunk roads are divided into three types using the agglomerative hierarchical clustering algorithm. According to the research of the spatial–temporal principles and differences of various types in traffic performance, the links with frequent traffic jams are identified, and the mechanism of the different traffic conditions is illustrated. It can be concluded that the trunk road is good conditions in most cases. On weekdays, there are two obvious peak times every day. On weekends, TPI data shows that there is no obvious peak time, and the overall road traffic conditions on weekends are better than those on weekdays. In addition, there are significant differences in the change trajectory of the TPI data and spatial distribution of various types of trunk roads. The roads with frequent traffic jams mainly come from the first and third type.

本研究的目的是对城市主干道的交通性能特征进行分类和分析。它可以帮助旅行者有效地避开高峰时段和拥挤的路线，交通管理人员可以对道路进行精细的管理和控制。本文通过分析交通绩效指数（TPI）数据，从宏观层面提取主干道路的运营特征。使用聚集层次聚类算法将主干路分为三种类型。根据时空原理和各种交通性能差异的研究，确定了交通拥堵频繁的环节，并阐述了不同交通状况的机理。可以得出结论，在大多数情况下，主干道路是良好的条件。在工作日，每天有两个明显的高峰时间。在周末，TPI数据显示没有明显的高峰时间，并且周末的总体道路交通状况要好于工作日。此外，TPI数据的变化轨迹和各种主干道路的空间分布也存在显着差异。交通拥堵频繁的道路主要来自第一类和第三类。