改进模糊聚类方法的物流园交通小区划分

摘要

To obtain the best classification results of logistics park, the qualitative division of logistics park traffic zones is carried out, then the weighted fuzzy similar matrix is constructed, and the adjacency matrix is used to amend the weighted fuzzy similar matrix. The improved fuzzy clustering method is employed to the clustering of merger of logistics park traffic zones. In addition, the F index is constructed to determine the optimal number of class based on the distance in the class and the distance between the class. The results show that the improved fuzzy clustering method could greatly reduce the amount of logistics park traffic zones, and it can quickly get the better logistics park traffic zones classification and avoid dividing logistics park traffic zones which are in the non-adjacent geographical position into the same class.%为获取物流园交通小区的最佳分类结果,首先对物流园进行定性划分,然后构造加权模糊相似矩阵,再利用邻接矩阵对加权模糊相似矩阵进行修正,采用改进的模糊聚类方法对小区进行聚类合并,并且基于类内距离和类间距离构造F指标以确定最佳分类数.实验结果表明,基于改进的模糊聚类方法能够较大程度地减少物流园交通小区数量,并且能够快速得到较好的小区分类,还能避免将地理位置不相邻的小区划分为同一类.