微观交通仿真查询算法与换道模型研究

发布时间：2018-08-08 15:01

【摘要】：城市交通拥堵已成为世界性的难题。交通拥堵不仅造成了运输成本增加、居民出行延误、空气质量下降,还给整个社会造成了巨大的生态环境污染、资源浪费、道路事故率增加等无谓损失。作为交通优化设计的主要工具,交通仿真通过对路网的设计、交通流的管控、信号灯控制等对城市交通状况进行仿真模拟,评估和优化交通方案,为达到提高城市道路的利用率,缓解城市交通拥堵的目的提供了理论依据和决策支持。微观交通仿真能直观的刻画车辆、行人等交通个体行为,随着对交通仿真精度、效率、规模和准确性等需求的日益增长,因而已成为分析交通流特性的关键技术,备受研究者的关注,成为近年来研究的热点。面对微观交通仿真中的近邻车辆查询的效率问题、安全换道模型中的纵向加速度自适应变化问题和基于公交IC卡数据的大规模OD矩阵推导问题已成为实际应用的困扰,本文从分析现有微观交通仿真技术出发,针对微观交通仿真的近邻车辆查询算法的可扩展性和效率问题、纵向加速度自适应调整的情况下车辆的安全换道距离问题,以及基于公交IC卡数据的大规模OD矩阵推导问题进行了深入研究。论文的主要创新成果如下:(1)提出了一种适用于微观交通仿真的基于局部索引的近邻查询算法。针对微观交通仿真中的多一维的有向近邻查询问题,提出了一种基于局部索引的近邻车辆查询算法。该算法在B+树的基础上,通过维护仿真单元的局部位置索引,使其同时具有B+树与线性方法的优点。对算法时间复杂度和期望查询长度的分析结果显示,该算法可满足大规模、拥堵的微观交通仿真近邻车辆查询需求,提高了仿真系统的仿真效率。(2)提出了一种新型微观交通仿真的加速度自适应的安全换道模型。针对车辆换道过程中加速度自适应变化的问题,提出了一种纵向加速度自适应的安全换道模型。该模型通过定义车辆换道的临界碰撞时间节点,将换道过程分为四个阶段,并结合了加速度变化和模型参数对临界碰撞时间节点以及安全换道距离的影响,使其能更真实地表征车辆换道的情况。该模型具有较高的精度,可有效模拟车辆换道过程的加速度变化,使得安全换道距离的计算更为准确。(3)提出了一种基于公交IC卡数据补全的大规模OD矩阵推导算法。针对城市公交的大规模OD矩阵推导问题,提出了一种基于公交IC卡数据补全的大规模OD矩阵推导算法。该算法在不关联公交调度和GPS数据情况下,通过上车站点标注和行驶方向标注子算法,将刷卡记录与实际公交站点进行匹配。同时,将全局公交数据补全问题映射为图论中的节点的遍历问题,利用贪心生长算法和广度优先遍历策略实现局部最优,并在基于公交出行链的假设下挖掘城市居民出行OD矩阵。该算法可高效地完成大规模OD矩阵推导,为交通需求分析提供理论依据。(4)设计并实现了一个轻量级微观交通仿真原型系统。为了验证查询算法的效率和换道模型的精度,设计并实现了一个微观交通仿真原型系统。该系统采用基于公交IC卡数据的OD矩阵设置,通过真实数据初始化仿真环境,实现了局部索引的近邻车辆查询算法和加速度自适应的安全换道模型。仿真结果显示该近邻查询算法可有效提高仿真系统的查询效率;而该换道模型能提供更准确的安全换道距离。综上,本文对微观交通仿真技术走向实际应用时面临的几类重点挑战性问题进行了针对性研究,为微观交通仿真技术的推广应用提供了一个高效可行的解决方案。本文的研究成果能有效地提高微观交通仿真系统的仿真性能和精度,并为微观交通仿真系统的研究和应用提供了理论和技术支持,具有良好的可扩展性,能够推广到更广泛的交通仿真分析应用场景。
[Abstract]:Traffic congestion has become a worldwide problem. Traffic congestion not only causes the increase of transportation costs, travel delays, and air quality, but also causes enormous ecological environment pollution, waste of resources and increase of road accident rate to the whole society. As the main tool for optimizing design, traffic simulation passes through the road. The design of the network, the control of traffic flow, the control of the signal light and so on are used to simulate the urban traffic conditions, evaluate and optimize the traffic plan. It provides the theoretical basis and decision support for the purpose of improving the utilization of urban roads and alleviating the traffic congestion of the city. With the increasing demand for precision, efficiency, scale and accuracy of traffic simulation, it has become a key technology for analyzing traffic flow characteristics. It has attracted much attention of researchers and became a hot spot in recent years. Facing the efficiency of nearest neighbor vehicle query in micro traffic simulation, the longitudinal acceleration adaptive in the security change model is adaptive. The problem of change and the derivation of large scale OD matrix based on bus IC card data have become a problem of practical application. In this paper, based on the analysis of the existing micro traffic simulation technology, the scalability and efficiency of the nearest neighbor vehicle query algorithm for microscopic traffic simulation, and the vehicle safety change under the adaptive adjustment of the longitudinal acceleration degree The distance problem, and the large-scale OD matrix derivation problem based on the bus IC card data are deeply studied. The main achievements of this paper are as follows: (1) a near neighbor query algorithm based on local index is proposed for microscopic traffic simulation. A neighbor vehicle query algorithm based on local index. Based on the B+ tree, the algorithm maintains the advantages of the B+ tree and the linear method by maintaining the local location index of the simulation unit. The analysis results of the time complexity and the expected query length of the algorithm show that the algorithm can satisfy the large-scale and congestion micro traffic simulation. In order to improve the simulation efficiency of the simulation system, the efficiency of the simulation system is improved. (2) a new model of the acceleration adaptive path change of the new micro traffic simulation is proposed. The critical collision time node divides the process into four stages, and combines the influence of the acceleration change and the model parameters on the critical collision time node and the safe change distance, so that it can more truly characterize the vehicle lane change. The model has a high precision and can effectively simulate the acceleration change of the vehicle changing process. The calculation of the safety change distance is more accurate. (3) a large-scale OD matrix derivation algorithm based on the bus IC card data complementation is proposed. A large-scale OD matrix derivation algorithm based on the bus IC card data complementation is proposed for the large-scale OD matrix derivation of urban public transport. This algorithm is not related to the bus scheduling and the GPS data. In the case, the card records are matched with the actual bus stations through the tagging and driving direction annotation algorithm, and the problem of the global bus data complement is mapped to the traversal problem of nodes in the graph theory. The local optimal is realized by the greedy growth algorithm and the breadth priority traversal strategy, and the bus travel chain is based on the bus travel chain. On the assumption that the urban resident travel OD matrix is excavated, the algorithm can efficiently complete the large-scale OD matrix derivation and provide the theoretical basis for the traffic demand analysis. (4) a lightweight microscopic traffic simulation prototype system is designed and implemented. In order to verify the efficiency of the query algorithm and the accuracy of the change model, a microscopic traffic imitation is designed and realized. The system uses the OD matrix set based on the bus IC card data, initializes the simulation environment through real data, realizes the local index nearest neighbor vehicle query algorithm and the acceleration adaptive security change path model. The simulation results show that the nearest neighbor query algorithm can effectively improve the query efficiency of the simulation system. The model can provide more accurate distance of safe channel change. In this paper, this paper studies several key challenges faced by the micro traffic simulation technology to practical application, and provides a efficient and feasible solution for the popularization and application of micro traffic simulation technology. The research results of this paper can effectively improve the micro level. The simulation performance and precision of the traffic simulation system provide theoretical and technical support for the research and application of the micro traffic simulation system. It has good extensibility and can be extended to a wider range of traffic simulation and analysis applications.
【学位授予单位】：电子科技大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：U495

【相似文献】