求解Logit随机用户均衡问题的改进人工鱼群算法.docVIP

  • 3
  • 0
  • 约6.31千字
  • 约 9页
  • 2017-03-02 发布于北京
  • 举报

求解Logit随机用户均衡问题的改进人工鱼群算法.doc

求解Logit随机用户均衡问题的改进人工鱼群算法.doc

求解Logit随机用户均衡问题的改进人工鱼群算法   摘 要: 针对人工鱼群算法求解大型优化问题时存在探索能力差以及搜索盲目性大的缺点,设计一种定向搜索变异的改进人工鱼群算法,该算法在迭代过程中不仅保证鱼群在当前状态下能够自适应变异,并且还可以使其向当前最优位置移动。随后将这种改进人工鱼群算法应用于求解Logit随机用户均衡问题,构建了随机用户均衡交通分配问题新的模型和求解方法。仿真结果表明,该方法具有较好的稳定性和收敛速度,具有在大型城市交通分配问题中应用的潜力。   关键词: 人工鱼群算法; 自适应变异; 交通分配; Logit随机用户均衡   中图分类号: TN911?34; TP18 文献标识码: A 文章编号: 1004?373X(2016)03?0127?04   Improved artificial fish swarm algorithm for solving Logit stochastic user equilibrium problem   LIU Bingquan1, DU Wei2   (1. College of Mathematics and Information Science, Weinan Normal University, Weinan 714099, China;   2. College of Transportation, Nantong University, Nantong 226019, China)   Abstract: Since the artificial fish swarm algorithm has the disadvantages of poor exploration ability and big searching blindness for solving the large optimization problem, an improved artificial fish swarm algorithm based on directed search variation was designed. The algorithm in the iterative process can ensure the adaptive variation of the fish swarm in current situation, and make the fish swarm move towards the optimal position. This algorithm is applied to solving the Logit stochastic user equilibrium problem, and the new model and solving method for stochastic user equilibrium traffic assignment were constructed. The simulation results show that the method has good stability and fast convergence rate, and has the application potential for traffic assignment problem in big cities.   Keywords: artificial fish swarm algorithm; adaptive variation; traffic assignment; Logit stochastic user equilibrium   0 引 言   人工鱼群算法(AFSA)是一种模拟鱼群行为进行随机搜索的群体智能优化算法,该算法对初始值和参数的选择不敏感,并且使用灵活,鲁棒性强,由于在算法中仅考察适应度函数,无须了解所求问题的内部信息,因此该算法具有良好的克服局部极值,取得全局极值的能力[1?2]。AFSA进行迭代的基本机理是通过群体之间的信息共享和个体自身调整求取全局最优解。其主要利用鱼的觅食、聚群和追尾行为,从构造单个鱼的底层行为开始,通过鱼群中各个体的局部寻优达到全局最优,该算法已经在交通规划、电网系统中取得了良好的应用效果[3?4]。但是人工鱼群算法在解决大型优化问题时,其保持探索与开发平衡的能力较差,并且算法运行后期搜索的盲目性较大,当一部分人工鱼处于漫无目的的随机移动或人工鱼在非全局极值点出现较严重聚集情况时,其收敛速度将大大减慢,甚至有时会陷入局部极值,搜索性能劣化,使得该算法得到的是满意解,其精度不是很高,从而影响了算法的质量和效率[5?6]。因此

文档评论(0)

1亿VIP精品文档

相关文档