基于客流预测的公交调度优化研究

发布时间：2018-02-20 05:09

本文关键词： 客流预测灰色模型发车间隔调度优化　出处：《郑州大学》2017年硕士论文　论文类型：学位论文

【摘要】：近几年,由于城市机动车保有量的增长速度与道路建设速度的严重不匹配,交通拥堵、环境污染已成为制约城市健康有序发展的重要问题。加快公共交通系统建设和发展速度,是提高道路资源利用效率、缓解城市交通拥堵压力的有效办法之一。公交调度是提高公交企业运营效率和管理水平的关键所在,科学的公交调度可以提高公共交通资源利用效率,提升公交服务品质,提高公共交通对出行市民的吸引力。公交客流是进行公交调度优化的数据基础和前提条件。本文结合客流在时间和空间上的分布特性,在深入分析大量客流数据的基础上,研究客流的周期性变化规律,并以公交客流在时间上的分布为有序样本,采用Fisher有序聚类算法,进行客流时段划分,然后在时段划分的基础上,结合灰色预测理论,建立分时段的公交客流预测模型,对客流在一个运营日的分布情况进行预测,为公交调度优化提供数据支持。公交车辆发车时间间隔是否合适是决定调度工作是否科学合理的关键。本文在客流预测的基础上建立了以公交企业运营成本与乘客等待成本加权和最小为目标函数,以发车间隔、车辆满载率、企业经济效益为约束条件,以发车时间间隔为决策变量的公交调度优化模型。该模型在客流数据的基础上,计算一个公交运营日各时段的发车时间间隔。公交调度优化模型是一个有约束条件的、多变量的、非线性模型,本文引入粒子群算法求解模型,并给出了求解过程中引入罚函数处理约束条件的具体办法。本文将公交客流预测模型应用于郑州公交45路,并在预测数据的基础上进行优化调度,得出了侧重不同利益主体的调度方案。结果表明,对不同利益主体的侧重会对发车间隔、运营日总社会成本、乘客等待时间带来不同的影响,该优化调度模型可为公交实际运营中调度人员制定发车时刻表、调整行车计划提供参考方案。
[Abstract]:In recent years, due to the serious mismatch between the growth rate of urban motor vehicle ownership and the speed of road construction, traffic congestion and environmental pollution have become an important problem restricting the healthy and orderly development of the city. It is one of the effective methods to improve the utilization efficiency of road resources and relieve the pressure of urban traffic jams. Public transport dispatch is the key to improve the operation efficiency and management level of public transport enterprises. Scientific public transport scheduling can improve the efficiency of public transport resources, improve the quality of public transport services, Public transport is the data base and prerequisite for bus scheduling optimization. This paper combines the distribution characteristics of passenger flow in time and space, based on in-depth analysis of a large number of passenger flow data. This paper studies the regularity of periodic change of passenger flow, and takes the distribution of bus passenger flow in time as an ordered sample, adopts Fisher orderly clustering algorithm to divide the passenger flow period, and then combines the grey forecasting theory with the time division of passenger flow. The forecast model of bus passenger flow is established, and the distribution of passenger flow on a operation day is forecasted. It is the key to decide whether the dispatching work is scientific and reasonable or not. Based on the forecast of passenger flow, this paper establishes the operation cost and ride of the public transport enterprise based on the forecast of passenger flow. The guest waits for the cost weighted sum to be minimized as the objective function, Based on the passenger flow data, the bus dispatch optimization model is based on the train departure interval, the vehicle full load rate, the economic benefit of the enterprise, and the departure time interval as the decision variable. The bus scheduling optimization model is a constrained, multivariable, nonlinear model. The particle swarm optimization algorithm is introduced to solve the model. The method of introducing penalty function to deal with the constraint condition is given. In this paper, the forecast model of bus passenger flow is applied to Zhengzhou bus Route 45, and the optimal dispatching is carried out on the basis of the forecast data. The results show that the emphasis on different stakeholders will have different effects on the departure interval, the total social cost of operation day, and the waiting time of passengers. The optimal dispatching model can provide a reference scheme for the dispatcher to make the departure schedule and adjust the train plan in the actual operation of public transportation.
【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.17;U492.22

【相似文献】