基于灰色决策的驾驶员路径选择研究

发布时间：2018-07-11 17:21

本文选题：智能交通 + 路径选择算法　；参考：《沈阳航空航天大学》2014年硕士论文

【摘要】：随着经济的稳步增长，城市的快速发展带来了日益严重的交通拥堵和频发的交通事故，且交通污染等问题已在影响着人们的生活和社会的和谐安定。面对如此严酷的交通问题，运用先进的管理技术为出行者提供多种服务的智能交通应运而生。动态路径诱导系统作为其核心，成为当前的研究热点。本文就对存在驾驶员偏好的最优路径选择问题进行了研究。驾驶员选择路径通常是感知路段的各种属性的综合评价的结果。本文将驾驶员的个人偏好应用在路径诱导过程中，提出一个路径选择指标体系，，来体现路径诱导系统的个性化特点。而层次分析法把定性方法与定量方法有机地结合起来，将人们的思维过程数学化、系统化，因此运用层次分析法来确定驾驶员的偏好。往往路径诱导系统中收集到的信息是不完全的，且其中驾驶员有不好描述的偏好信息。路径选择是一个复杂的系统问题，为此本文运用灰色系统理论来对系统中的灰色信息进行研究。在提出的路径选择指标体系下，由层次分析法确定驾驶员的偏好后，运用基于灰色决策方法来进行路径选择。由诱导信息得到的可行路径集，本文运用基于灰色关联决策的方法和基于多目标智能加权灰靶决策的方法在其中搜索满足驾驶员个人偏好的最优路径。通过仿真实验，对所提出的算法进行验证。运用VISSIM在一个实际路网上进行路径选择仿真，Dijkstra算法得到最短路径，基于灰色关联决策和基于多目标智能加权灰靶决策的路径选择算法均得到能够反映驾驶员个人偏好的最优路径。改变仿真路网上行驶车辆使用算法的比例，研究不同算法使用比例变化对整个路网交通的影响。对于有突发事件造成交通拥堵时，运用Dijkstra算法的车辆行驶速度大幅降低，而基于多目标智能加权灰靶决策的算法能立即改变路径选择，使出行者绕过拥堵路段，行驶速度没有受到太大影响。仿真结果表明：所提出的路径选择算法具有很好的可行性和适用性。
[Abstract]:With the steady growth of economy, the rapid development of the city has brought more and more serious traffic congestion and frequent traffic accidents, and traffic pollution and other problems have affected people's lives and social harmony and stability. In the face of such severe traffic problems, intelligent transportation (its), which uses advanced management technology to provide multiple services for travelers, emerges as the times require. As the core of dynamic path guidance system, dynamic path guidance system has become a hot research topic. In this paper, the problem of optimal path selection with driver preference is studied. The driver's choice of path is usually the result of comprehensive evaluation of the various attributes of the perceived road section. In this paper, the driver's personal preference is applied to the path guidance process, and a path selection index system is proposed to reflect the personalized characteristics of the path guidance system. The Analytic hierarchy process (AHP) combines qualitative and quantitative methods organically, and makes people's thinking process mathematical and systematic. Therefore, AHP is used to determine drivers' preferences. The information collected in the path guidance system is often incomplete, and the driver has bad preference information. Path selection is a complex system problem. In this paper, grey system theory is used to study the grey information in the system. Under the proposed path selection index system, after determining the driver's preference by AHP, the path selection is based on grey decision method. Based on the set of feasible paths derived from the induced information, this paper uses the method based on grey relational decision and the method based on multi-objective intelligent weighted grey target decision to search for the optimal path satisfying the individual preference of the driver. The proposed algorithm is verified by simulation experiments. The shortest path can be obtained by using VISSIM to simulate the path selection in a real road network by using Dijkstra algorithm. The path selection algorithm based on grey relational decision and multi-objective intelligent weighted grey target decision can obtain the optimal path which can reflect the individual preference of driver. By changing the proportion of the vehicle usage algorithm on the road network, the influence of different algorithms on the whole road network traffic is studied. For traffic jams caused by unexpected events, the speed of vehicles using Dijkstra algorithm is greatly reduced, while the algorithm based on multi-objective intelligent weighted grey target decision can change the path choice immediately, and make travelers bypass the congested section. The speed of driving was not greatly affected. Simulation results show that the proposed path selection algorithm is feasible and applicable.
【学位授予单位】：沈阳航空航天大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U495

【相似文献】