热带地理

热带地理 ›› 2017, Vol. 37 ›› Issue (3): 383-391.doi: 10.13284/j.cnki.rddl.002911

• 论文 • 上一篇    下一篇

基于自适应双态粒子群的应急物资配送空间优化

陈通利1，马世发2，徐舒苑3，黎海波1   

  1. （1．东莞市地理信息与规划编制研究中心，广东 东莞 523129；2．广东省国土资源技术中心，广州 510075； 3．中山大学 地理科学与规划学院，广州 510275）
  • 出版日期:2017-05-05 发布日期:2017-05-05
  • 通讯作者: 马世发（1985―），男，湖北宜昌人，博士，主要从事地理模拟和空间规划决策支持研究，（E-mail）410803097@qq.com。
  • 作者简介:陈通利（1990―），男，浙江温州人，硕士，主要从事城市规划和地理建模的研究，（E-mail）tony_tiger@foxmail.com
  • 基金资助:
    国家自然科学基金项目（41301418）

Spatial Optimization for Emergency Materials Allocation Based on a Self-adaptive Binary-state PSO Algorithm

CHEN Tongli1，MA Shifa2，XU Shuyuan3，LI Haibo1   

  1. （1．Dongguan Geographic Information and Urban Planning Researching Center，Dongguan 523129，China；2．Land Resources Technology Center of Guangdong Provence，Guangzhou 510075，China；3．School of Geography and Planning，Sun Yat-sen University，Guangzhou 510275，China）
  • Online:2017-05-05 Published:2017-05-05

PDF (PC)

赞

被引次数

3

摘要： 应急物资的调度是一个典型的社会服务空间优化问题，将粒子群优化算法与GIS相结合建立应急物资空间分配智能优化模型。首先，针对传统标准粒子群算法随维度增高而极易陷入局部最优的特点，模型将粒子群进化曲线的曲率定义为种群多样性，采用“扑食”和“探索”2个状态建立自适应双态进化机制；其次，改进的算法只针对全局最优粒子进行变异，很好地控制了群体“集群飞行”与“外空探索”之间的协调；第三，利用动态递归和生物智能的随机特性建立起了约束处理机制。通过低维和高维理论模型测试，验证了模型在低维优化空间获取了绝对最优解；而在高维空间也达到了非常高的优化精度。最后，选择某市物资供应为案例，利用该模型分析了试验区物资分配格局。研究表明，耦合生物智能的GIS空间优化模型在智慧城市建设中具有重要的应用意义。

关键词: 粒子群, 空间优化, 自适应, 应急物资, 智慧城市

Abstract: The allocation of emergency materials is a typical spatial optimization decision problem for social service. A hybrid model was proposed in this paper to allocate the emergency materials based on GIS and particle swarm optimization algorithm. In view of the drawbacks that it is easy to fall into local best solutions with the increasing dimensions for classical PSO, the evolution curvature was defined as population diversity, and a binary-state evolution mechanism, which includes the “predation” and “exploration” states, was adopted. In addition, the variation operator was designed just only for the elite particles, which can well balance the group learning and exploration behavior in outer space. Furthermore, the constraint for this spatial optimization was set by the combination of dynamic recurrent and the random characteristic of swarm intelligent algorithms. This model has been tested by using a theoretical datasets with low dimensions and high dimensions respectively. It has shown that the spatial optimization model designed in this paper can get global optimum solution for fewer and also perform well for larger dataset. Finally, a case study was implemented in a city. Results demonstrate that the spatial optimization model coupled with GIS and swarm intelligence algorithm would be an important application for the smart city’s construction.

Key words: particle swarm optimization, spatial optimization, self-adaptive, emergency materials, smart city