公交规划客流OD数据生成方法研究

发布时间：2018-06-29 10:50

本文选题：公交客流OD矩阵 + OD调查　；参考：《大连海事大学》2012年硕士论文

【摘要】：随着城市建设迅速扩张和市区人口的急剧增长,城市公交问题越发成为关系城市发展和居民舒适生活的关键。良好的公交体系能够大幅缓解城市交通压力,为城市综合发展起到疏通筋络的作用。因此无论是已成相当规模的大城市还是正在建设的中小型城市,都在积极建设、完善、优化着公交网络。公交线网优化是改善城市公交的重要课题。在公交线网优化研究中,从经验判断法到现在的数学寻优法,经历了无数发展与更新。当代,在计算机强大计算能力的支持下,应用数学知识分析研究公交线网优化问题成为主流。而这些方法都会应用到一种数据模型,就是OD矩阵。OD矩阵作为一种理想输入,在公交线网优化中占有重要地位。然而OD矩阵数据非天然呈现,它是通过调查并对调查后的数据进行科学的处理后才能获得。本文以获取完整公交客流OD矩阵数据的方法为核心内容进行课题研究,为公交线网优化奠定重要基础。 OD矩阵数据调查方法多种多样,针对城市大小、调查人员多少和对调查结果精度的要求差别会有很大的不同,对方法的选择是个多方面量化权衡的问题。本文重点研究适用于中小城市以及在较少调查资源的前提下的调查方法,应用数理统计学尤其是抽样理论方面的知识,对方法的可行性、可靠性、效率与成本进行分析,并结合玉溪市公交OD调查工作进行实例分析,创新性地采用多阶段抽样调查方法,依次对时间、车辆、乘客进行合理抽样并设计出与抽样方法相应的调查问卷,并证明了该方法在实际应用中的优越性。而后对调查所得的数据进行详细的统计分析,依然应用统计学原理,首先针对数据质量设计相应的插补方法,讨论调研误差对数据结果的影响。然后对数据进行预测分析,启发式地运用抽样调查得到的部分OD数据与流量数据结合,扩样估计全体数据,得到能够作为线网优化良好输入的完整的OD矩阵数据。最后结合玉溪市项目输出实例的OD矩阵结果。在抽样理论支持下,本文的调查方法与相应的数据处理方法切合实际并带有启发性,多步骤方法之间的连接清晰合理,输出结果良好。
[Abstract]:With the rapid expansion of urban construction and the rapid growth of urban population, urban public transport has become the key to urban development and residents' comfortable life. A good public transport system can greatly relieve the pressure of urban traffic and play a role in the comprehensive development of the city. Therefore, both large cities and small and medium-sized cities are actively building, perfecting and optimizing the public transport network. The optimization of bus network is an important task to improve urban public transportation. In the research of bus network optimization, from the empirical judgment method to the present mathematical optimization method, it has experienced numerous developments and updates. Nowadays, with the support of powerful computing power of computer, the application of mathematical knowledge to the analysis of bus network optimization has become the mainstream. These methods will be applied to a data model, namely OD matrix. OD matrix, as an ideal input, plays an important role in bus network optimization. However, OD matrix data is not natural, it is through investigation and scientific processing of the data after the survey can be obtained. In this paper, the method of obtaining complete OD matrix data of bus passenger flow is taken as the core of the research, which lays an important foundation for the optimization of bus network. There are many methods of OD matrix data investigation, aiming at the size of the city. The number of investigators is different from the requirement of the accuracy of survey results, and the choice of methods is a question of quantitative trade-off in many aspects. This paper focuses on the investigation methods suitable for small and medium-sized cities and under the premise of less investigation resources, and applies mathematical statistics, especially the knowledge of sampling theory, to analyze the feasibility, reliability, efficiency and cost of the methods. Combining with the actual case study of bus OD investigation in Yuxi City, the author innovatively adopts multi-stage sampling survey method, carries on the reasonable sampling to the time, the vehicle, the passenger in turn, and designs the questionnaire corresponding to the sampling method. The superiority of this method in practical application is proved. Then the detailed statistical analysis of the data obtained from the survey is carried out, and the principle of statistics is still applied. Firstly, the corresponding interpolation method is designed for the quality of the data, and the influence of the survey error on the data results is discussed. Then the data are predicted and analyzed, and some OD data are combined with traffic data in heuristic way, and the whole data are expanded to obtain the complete OD matrix data which can be used as a good input for network optimization. Finally, the OD matrix result of Yuxi project output example is combined. With the support of sampling theory, the investigation method and the corresponding data processing method are practical and enlightening, the connection between the multi-step methods is clear and reasonable, and the output results are good.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F224;F570

【参考文献】