基于均衡流量的城市公交网络系统优化模型及算法

发布时间：2018-01-05 13:21

本文关键词：基于均衡流量的城市公交网络系统优化模型及算法　出处：《北京交通大学》2014年硕士论文　论文类型：学位论文

【摘要】：伴随着我国城市化进程的不断加快,城市化水平的不断提升以及城市人口的不断增加,城市出行需求大幅度地增加。与此同时,小汽车的快速发展一方面带来了出行的便捷性,另一方面其数量的急剧上升也大大增加了现有路网的压力。现有交通路网的规划与建设远远满足不了经济发展的步伐,城市交通问题层出不穷,城市交通拥堵问题成为了城市发展的绊脚石,亦成为现在不得不解决的关键问题。城市公交系统具备自身运力大、方便性等特点,大力发展城市公交成为解决城市交通问题的首选。基于此,本文围绕城市公交系统的优化设计进行了以下几个方面的研究。 1、城市公交客流需求是城市公交系统建设的重要依据,合理的预测、模拟城市公交客流量在公交网络中的分布,有利于奠定公交系统优化设计的出行者数据基础。城市公交客流具有多样性、多变性、复杂性等特点,合理的构建客流分配模型直接影响着对网络中的客流分布的模拟效果。基于此本文建立了考虑换乘因素的城市公交系统随机均衡配流模型,考虑换乘费用和换乘次数对于乘客出行路径选择的影响,采用随机均衡配流模型更好地模拟乘客的出行路径选择行为,较好地模拟了乘客在公交网络中的出行分布。 2、发车频率优化是城市公交系统优化设计最重要的工作之一,其设置的合理性不仅影响着城市公交对广大出行者的服务效率,也直接关系着公交运营企业自身的效益。本文从乘客和公交企业双方的利益出发,建立基于均衡流量的城市公交系统发车频率优化模型,以乘客出行总费用最小、公交运营企业收益最大为上层优化目标,以考虑换乘费用的随机均衡配流模型为下层优化目标,采用改进的遗传算法进行模型求解,改进后的公交发车频率设置更加符合乘客和公交企业的利益。 3、城市公交线网规划受到多方面条件、因素的制约,本论文介绍了城市公交线网优化设计原则、目标及影响因素,在此基础上,考虑公交系统中线路长度的限制、公交车运行的最小客流限制、非直线限制、断面流量限制、站点距离限制、发车频率限制等约束条件,以乘客直达率最大、公交运营企业收益最大为上层优化目标,建立以城市公交客流随机均衡分配模型为下层模型的基于均衡流量的城市公交线网优化设计模型,尽最大可能地顾及了乘客和运营者的双方面利益。 4、遗传算法在优化问题的求解中具有明显的优势,本论文在采用遗传算法的基础上对遗传算法做出了相应的改进,且在对遗传算法自身改进的基础上,针对公交线网优化这一具体问题,提出了遗传算法与其他优化算法(模拟退火算法)结合的新算法,有效地避免了遗传算法在求解过程中的缺点,改进的混合遗传算法具有更好的收敛性和求解效率
[Abstract]:Along with the accelerating urbanization process of our country , the rising urbanization level and the increasing of the urban population , the urban travel demand is greatly increased . At the same time , the rapid development of the small car brings convenience to the travel , on the other hand , the rapid rise of the quantity of the small car greatly increases the pressure of the existing road network . Based on this , the paper studies the optimization design of urban public transport system in the following aspects . 1 . The urban public transport passenger flow demand is an important basis for the construction of the urban public transport system . The reasonable forecast and simulation of the distribution of the urban public transport passenger flow in the public transport network will help to lay a foundation for the simulation of the passenger flow distribution in the public transport system . The reasonable construction of the passenger flow distribution model affects the passenger flow distribution in the network directly . Based on this paper , a stochastic equilibrium distribution model is established to simulate the passenger ' s travel route selection behavior better , and the distribution of passengers in the public transport network is simulated well . 2 . The optimization of vehicle frequency is one of the most important tasks of urban public transport system optimization design . The rationality not only affects the service efficiency of urban public transport to the large number of travelers , but also directly affects the efficiency of the public transport operators . In this paper , based on the interests of both passengers and public transport enterprises , this paper sets up an optimized model of urban public transport system based on equilibrium flow . The model is solved by using the improved genetic algorithm . The improved bus departure frequency setting is more consistent with the interests of passengers and public transport enterprises . 3 . The urban public transport network planning is restricted by many conditions and factors . This paper introduces the principle , goal and influencing factors of urban bus network optimization design . On the basis of this , considering the limitation of the length of the line in the bus system , the minimum passenger flow restriction , the non - linear limitation , the cross - section flow restriction , the station distance limitation , the departure frequency limit and so on , the urban public transport network optimization design model based on the equilibrium flow of the urban public transport passenger flow stochastic equilibrium distribution model is established . 4 . Genetic algorithm has obvious advantages in solving the optimization problem . This paper improves the genetic algorithm on the basis of genetic algorithm , and proposes a new algorithm combining genetic algorithm and other optimization algorithms ( simulated annealing algorithm ) on the basis of the improvement of genetic algorithm . The disadvantages of genetic algorithm and other optimization algorithms ( simulated annealing algorithm ) are proposed . The improved hybrid genetic algorithm has better convergence and solving efficiency .

【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.17

【参考文献】