摘要
为了解决包含禁行路线路网的最优路径快速求解问题,研究了不含禁行路线路网和包含禁行路线路网的特点,建立了相应的路网数学模型。通过路网转化法把包含禁行路线的路网转化为不含禁行路线的路网,降低了最优路径求解的难度。研究了霍普费尔特神经网络(Hopfield Neural Network,HNN)的特点,设计了适合求解路网最优路径的HNN算法,在算法中采用动态邻接矩阵,节省了计算机内存,减少了运算时间。将所研究的路网转化方法和设计的HNN算法应用于所研发的车辆诱导系统中,并进行了实际路网测试,结果表明应用该方法能够在包含禁行路线路网中求解最优路径,且比经典算法的运算效率高。
In order to solve the problem of fast computing optimal path in road network with restricted routes, characteristics of road network without and with restricted routes are analyzed, and the corresponding mathematic models of road network are built. Road network with restricted routes is translated into road network without restricted routes according to the studied method, which reduced the complexity of the optimal path solution.The characteristics of HNN （Hopfield neural network） are analyzed, and a HNN algorithm to solve optimal path in road network according to the characteristics of Hopfield neural network is designed. The method for transforming road network and the HNN algorithm are applied to the studied vehicle guidance system to solve optimal path and are tested in the actual road network.Test results show that the methods can find correct optimal path in road network with restricted routes, and the HNN algorithm is more efficient in calculation than classical algorithms.
出处
《公路交通科技》
CAS
CSCD
北大核心
2007年第6期97-101,共5页
Journal of Highway and Transportation Research and Development
基金
国家"十五"科技攻关资助项目(2002BA404A01)
山东省教育厅中青年学术骨干基金资助项目(A2002-107)
关键词
智能运输系统
路网
HNN算法
最优路径
车辆诱导系统
禁行路线
Intelligent Transport Systems
road network
Hopfield neural network algorithm
optimal path
vehicle guidance system
restricted route
