基于网络Voronoi图启发式和群智能的最大覆盖空间优化  

Maximal Covering Spatial Optimization Based on Network Voronoi Diagrams Heuristic and Swarm Intelligence

作　　者：谢顺平[1] 冯学智[1] 都金康[1] 

机构地区：[1]南京大学地理与海洋科学学院,江苏南京210093

出　　处：《测绘学报》2011年第6期778-784,共7页Acta Geodaetica et Cartographica Sinica

基　　金：国家863计划(2008AA12Z106)

摘　　要：提出一种基于网络Voronoi面域图的最大覆盖选址模型及相应的粒子群优化方法,为城市化区域响应敏感型公共服务设施的空间优化提供技术方法。考虑设施功能沿交通网络传导以及需求非均匀连续分布情形,对设施在网络连续空间上进行布局优化,选址模型采用网络Voronoi面域图划分布局设施的功能辐射域,以启发空间优化最小化重叠覆盖。模型最大化设施利用效率,设施功能对覆盖半径以内的需求完全覆盖,对覆盖半径以外的需求部分覆盖。提出一种集成遗传机制和广义Voronoi图的改进粒子群算法,以提高连续网络空间内的空间优化性能。对南京市消防站最大覆盖选址优化的试验表明,该研究取得较为理想的结果。A maximal covering location model based on network Voronoi area diagrams and particle swam optimization is proposed to provide the spatial optimization means for response sensitive public service facilities in urbanized area. It is taken into account that facilities function conducts along traffic network and variable demands distribute continuously, the facilities optimized can be located in continuous network space. The network Voronoi area diagrams are used to simulate the service areas of facilities in the maximal covering location model, which has heuristic to minimize overlap coverage in spatial optimization. The proposed model maximizes utilization of facilities, the demands within coverage radius are covered completely and the demands beyond coverage radius are covered partially by facilities function. An improved particle swam optimization algorithm integrated with genetic mechanism and generalized Voronoi diagram is proposed to enhance the optimization performance in continuous network space. The computational experiment for location optimization of fire stations in Nanjingshows that the proposed model and optimization algorithm have achieved desired results.

关 键 词：网络Vo ronoi面域图 空间优化 最大覆盖选址模型 Voronoi图启发式 粒子群算法 

分 类 号：P208[天文地球—地图制图学与地理信息工程]