基于改进FCM算法的城市交叉口交通状态判别

Identification of Traffic State at Urban Intersections Based on Improved FCM Algorithm

为准确、有效判别城市交叉口交通状态,提出了一种基于聚类算法的城市交叉口交通状态判别方法.该方法针对交叉口交通流的变化特性及拥堵强度,将交通状态分为畅通、缓行、拥堵3类,选取平均排队长度、平均延误、饱和度3个交通参数为判别指标,利用基于模糊K均值聚类(K-means)算法和熵权法加权的改进模糊C均值聚类(FCM)算法确定交通状态,并以济南市为例验证判别算法的有效性.实例仿真结果表明:与传统的FCM算法相比,改进的FCM算法的目标函数值降低了76.0%,迭代次数减少了28.6%,同时利用划分系数(PC)和划分熵系数(PE)的评价结果更优,验证了所提的方法对城市交叉口交通状态判别的准确性更高.

In order to accurately and effectively discriminate the traffic state of urban intersections, a method for identifying traffic states at urban intersections based on clustering algorithm is proposed. According to the variation characteristics of intersection traffic flow and congestion intensity, the traffic state is divided into free flow, slow and congested. Three traffic parameters such as average queue length, average delay and saturation are selected as discrimination indexes. The K-means algorithm and the improved fuzzy c-means clustering(FCM) algorithm weighted by the entropy weight method are used to determine the traffic state. Jinan City is taken as an example to verify the effectiveness of the discrimination algorithm. The simulation results of the example show that compared with the traditional FCM algorithm, the objective function value of the improved FCM algorithm is reduced by 76.0% and the number of iterations is reduced by 28.6%. At the same time, the evaluation results of partition coefficient(PC) and partition entropy coefficient(PE) are better, which verifies that the proposed method has higher accuracy in discriminating the traffic state of urban intersections.