基于遗传动态模糊聚类的道路交通状态判定方法

针对当前城市道路交通状态判定时采用的模糊c均值聚类(FCM)和c均值聚类等方法的有效性较差的状况,提出了一种基于遗传算法的动态模糊聚类方法,通过计算交通流参数之间的模糊相似性,不失真地反映它们之间的内在关联,同时将交通流参数之间的模糊相似性映射到样本之间的欧氏距离,即将高维样本映射到二维平面,利用遗传算法不断优化两者之间的映射,使交通流参数之间的欧氏距离逐步趋近于其模糊相似性,实现动态模糊聚类,仿真实验结果证明了该方法的可行性和有效性.

Identification Method of the Road Traffic Conditions Based on Genetic-Dynamic Fuzzy Clustering Algorithm

The existing C-means and Fuzzy C-means have their disadvantages in identifying city road traffic conditions. In this status, a dynamic fuzzy clustering method was presented based on the genetic algorithm. By calculating the fuzzy similarity of traffic parameters the essential associations among the samples are modeled exactly. The fuzzy similarity between two samples was mapped into their Euclidean distance, that is, the high dimensional sample was mapped into a two dimensional plane. The mapping is globally optimized by the genetic algorithm, in which the coordinates of each sample and the Euclidean distance were adjusted approximate to the fuzzy similarity between samples gradually. Simulated experiments show the feasibility and availability of the proposed method.