基于复杂网络聚类的最优选址模型
作者简介:戴技才(1974-),男,湖南益阳人,博士,讲师,主要从事地理信息系统模型,复杂系统研究。E-mail:daijicai@cqnu.edu.cn
收稿日期: 2012-01-09
要求修回日期: 2012-04-09
网络出版日期: 2013-02-20
基金资助
教育部人文社会科学研究青年基金(批准号: 11YJCZH023)
国家自然科学基金项目(41271411,41001102)资助
Model of Locating Optimal Sites Based on Complex Network Clustering
Received date: 2012-01-09
Request revised date: 2012-04-09
Online published: 2013-02-20
Copyright
最优选址在社会经济活动中非常重要。传统的网络聚类分析以空间两点之间的直线度量距离,而不是以空间最短网络路径作为聚类条件,无法找到复杂网络的最优选址中心。基于最短路径的复杂网络聚类模型,探索复杂道路网络中的最优选址。模型通过迭代法获取近似最优解,二分邻域分割法逼近最优解分布区,应用邻域下降法达到最优选址点。实验结果表明:本模型与穷举-Dijkstra算法相比,计算精度相当,计算速度提高了约23倍以上。模型以复杂网络聚类为基础推导,为复杂网络选址、聚类提供了一种新的理论与方法。
关键词: 选址; 网络聚类; 复杂网络; 最短路径; Dijkstra算法
戴技才 , 宗会明 . 基于复杂网络聚类的最优选址模型[J]. 地理科学, 2013 , 33(2) : 143 -149 . DOI: 10.13249/j.cnki.sgs.2013.02.143
The best location for site selection is very important for social and economical systems. Traditional clustering analysis cannot reach the optimal location allocation center based on the Euclidian, Manhattan, Chebyshev, Fu or Tanimoto distance,which does not depend on the shortest path distance as the clustering condition. Complex network clustering model based on the shortest path is proposed to explore site selection in this paper. The approximate optimal results are obtained using the iterative technology. The searching range of optimal site allocation is approached by half dividing the neighborhood. And the best location is captured applying the descend strategy. Experiments indicate that the proposed model can improve the calculating speed by at least 23 times with the same precision to approach the optimal resolution compared to the greedy Dijkstra algorithm. This model is inferred from the complex network clustering. It provides new theory and method to cluster complex network and locate sitting facilities.
Fig. 1 Flow chart of location allocation based on complex network clustering图1 复杂网络聚类选址模型流程 |
Fig. 2 Approach the approximate optimal location using iterative algorithm图2 迭代法获取近似最优解 |
Fig. 3 (a) the approximate optimal result fitting to the vertex of network, (b) space shape of the results图3 (a)近似最优解拟合至网络节点,(b)解的空间形状 |
Fig. 4 Method of half dividing the neighborhood. (a) the length and width of the blue rectangle is half of the length and width of the original rectangle of the network, (b) the second half-dividing the neighborhood, (c) the third half-dividing the neighborhood图4 二分邻域分割法。(a)蓝色矩形的长和宽分别为原网络图形长和宽的二分之一,(b)第二次二分邻域分割,(c)第三次二分邻域分割 |
Fig. 5 Principle of searching for the results based on the descend approach in the neighborhood图5 邻域下降法搜索解的原理 |
Fig. 6 (a)Principle of the descend approach in the neighborhood, (b)path of searching the results图6 (a)邻域下降法的算法原理,(b)邻域下降法搜索解的路径 |
Fig. 7 To capture the optimal location site based on complex network clustering图7 网络聚类法求解最优选址点 |
Table 1 Comparing with the results based on complex network clustering and greedy Dijkstra algorithm表1 网络聚类法与穷举-Dijkstra算法的实验对比结果 |
国家或地区 | 节点数 (个) | 弧段数 (个) | 穷举-Dijkstra法用时(h) | 网络聚类法用时 (h) | 网络聚类法是否找到了 最优解 |
---|---|---|---|---|---|
中国大陆 | 21986 | 27256 | 49.47 | 1.83 | 是 |
英国 | 3774 | 5295 | 0.23 | 0.01 | 是 |
德国 | 3799 | 5343 | 0.24 | 0.01 | 是 |
法国 | 5314 | 7056 | 0.63 | 0.01 | 是 |
意大利 | 2563 | 3469 | 0.07 | 0.003 | 是 |
印度 | 15941 | 21674 | 17.09 | 0.31 | 是 |
日本 | 6706 | 8983 | 1.27 | 0.05 | 是 |
俄罗斯 | 37705 | 43908 | 258.49 | 9.32 | 是 |
美国本土 | 28602 | 46358 | 112.26 | 3.74 | 是 |
The authors have declared that no competing interests exist.
[1] |
|
[2] |
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
|
[9] |
|
[10] |
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
[19] |
|
[20] |
|
[21] |
|
[22] |
|
[23] |
|
/
〈 |
|
〉 |