广州市重要通道高峰期交通模式聚类分析

基于广州市重要通道2016年11-12月的交通指数，通过划分重要通道的拥堵与非拥堵状态，确定拥堵状态通道的中心趋势值，建立交通模式的聚类分析模型，采取变异系数对模型进行评价，得出交通模式聚类优选模型。结果显示：广州市重要通道早高峰呈现向心型交通拥堵模式，处于交通拥堵状态的通道可划分为6种交通模式，共3种几何形态类型：滞后型、对称型和递增型。晚高峰的交通拥堵通道分布广泛，区际主干道和天河商业区内的主干道交通拥堵更加严重；处于交通拥堵状态的通道可将其划分为9种交通模式，共4种几何形态类型：前锋型、滞后型、平缓型和对称型。

Clustering Analysis of Traffic Patterns at the Peak Period of Important Thoroughfares in Guangzhou

With the increasing traffic congestion in metropolitan areas, the decision-makers of some cities have paid much attention to the congestion characteristics of different passageways in cities. Based on the important passageways' traffic index of Guangzhou City in November-December, 2016, this paper set up the traffic pattern clustering evaluation model through the following 4 aspects: classified the important passageways congestion and non-congestion state, calculated the central trend value of the congestion state passageways, set up the traffic cluster analysis model, and used the coefficient of variation to evaluate the model, then got the traffic pattern clustering optimization model .Traffic pattern clustering optimization model analysis showed that the early peak of Guangzhou’s important passageways showed a heart shaped traffic congestion model. In the mode of heart passageways traffic congestion state could be divided into six kinds of traffic patterns and a total of three kinds of geometric patterns, which were backward type, symmetric type and incremental type. The traffic congestion passageways were widely distributed in the late peak hours, there was no obvious centrifugal or centripetal traffic congestion trend. The main road traffic congestion in the inter-regional main roads and Tianhe commercial district was more serious. The passageways in the traffic congestion state could be divided into nine traffic modes, which made total four geometric types, including forward type, lagging type, gentle type and symmetrical type. From the spatial distribution of traffic patterns at early peak hours and late peak hours, we knew that the traffic patterns of different driving directions of the same passageways were different, and the traffic patterns of the same driving directions in different passageways were similar. Through the interpretation of the spatial distribution of traffic patterns, we can grasp the spatial and temporal distribution characteristics of urban traffic congestion from a global perspective and provide a reference for the decision-making of traffic congestion control in urban areas.